STM32 RM

RM0090
Reference manual
STM32F405/415, STM32F407/417, STM32F427/437 and
STM32F429/439 advanced Arm®-based 32-bit MCUs
Introduction
This reference manual targets application developers. It provides complete information on
how to use the STM32F405xx/07xx, STM32F415xx/17xx, STM32F42xxx and
STM32F43xxx microcontroller memory and peripherals.
The STM32F405xx/07xx, STM32F415xx/17xx, STM32F42xxx and STM32F43xxx constitute
a family of microcontrollers with different memory sizes, packages and peripherals.
For ordering information, mechanical and electrical device characteristics please refer to the
datasheets.
For information on the Arm® Cortex®-M4 with FPU core, please refer to the Cortex®-M4 with
FPU Technical Reference Manual.
Related documents
Available from STMicroelectronics web site (http://www.st.com):
• STM32F40x and STM32F41x datasheets
• STM32F42x and STM32F43x datasheets
• For information on the Arm® Cortex®-M4 with FPU, refer to the STM32F3xx/F4xxx
Cortex®-M4 with FPU programming manual (PM0214).
February 2021 RM0090 Rev 19 1/1751

www.st.com 1
Contents RM0090
Contents
1 Documentation conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
1.1 List of abbreviations for registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
1.2 Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
1.3 Peripheral availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
2 Memory and bus architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

2.1 System architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
2.1.1 I-bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
2.1.2 D-bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
2.1.3 S-bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
2.1.4 DMA memory bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
2.1.5 DMA peripheral bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
2.1.6 Ethernet DMA bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
2.1.7 USB OTG HS DMA bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
2.1.8 LCD-TFT controller DMA bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
2.1.9 DMA2D bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
2.1.10 BusMatrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
2.1.11 AHB/APB bridges (APB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.2 Memory organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.3 Memory map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.3.1 Embedded SRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
2.3.2 Flash memory overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
2.3.3 Bit banding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
2.4 Boot configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3 Embedded Flash memory interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.2 Main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.3 Embedded Flash memory in
STM32F405xx/07xx and STM32F415xx/17xx . . . . . . . . . . . . . . . . . . . . . 74
3.4 Embedded Flash memory in STM32F42xxx and STM32F43xxx . . . . . . . 76
3.5 Read interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
3.5.1 Relation between CPU clock frequency and Flash memory read time . 80
2/1751 RM0090 Rev 19

RM0090 Contents
3.5.2 Adaptive real-time memory accelerator (ART Accelerator™) . . . . . . . . 82

3.6 Erase and program operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
3.6.1 Unlocking the Flash control register . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
3.6.2 Program/erase parallelism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.6.3 Erase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.6.4 Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
3.6.5 Read-while-write (RWW) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
3.6.6 Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
3.7 Option bytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
3.7.1 Description of user option bytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
3.7.2 Programming user option bytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.7.3 Read protection (RDP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
3.7.4 Write protections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
3.7.5 Proprietary code readout protection (PCROP) . . . . . . . . . . . . . . . . . . . 95
3.8 One-time programmable bytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
3.9 Flash interface registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
3.9.1 Flash access control register (FLASH_ACR)
for STM32F405xx/07xx and STM32F415xx/17xx . . . . . . . . . . . . . . . . . 98
for STM32F42xxx and STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . . 99
3.9.3 Flash key register (FLASH_KEYR) . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
3.9.4 Flash option key register (FLASH_OPTKEYR) . . . . . . . . . . . . . . . . . . 100
3.9.5 Flash status register (FLASH_SR) for
STM32F405xx/07xx and STM32F415xx/17xx . . . . . . . . . . . . . . . . . . . 101
STM32F42xxx and STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
3.9.7 Flash control register (FLASH_CR) for
STM32F42xxx and STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
3.9.9 Flash option control register (FLASH_OPTCR) for
3.9.10 Flash option control register (FLASH_OPTCR)
for STM32F42xxx and STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . 108
3.9.11 Flash option control register (FLASH_OPTCR1)
3.9.12 Flash interface register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
4 CRC calculation unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
RM0090 Rev 19 3/1751

39
Contents RM0090
4.1 CRC introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113

4.2 CRC main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113
4.3 CRC functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .114
4.4 CRC registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .114
4.4.1 Data register (CRC_DR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
4.4.2 Independent data register (CRC_IDR) . . . . . . . . . . . . . . . . . . . . . . . . 114
4.4.3 Control register (CRC_CR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
4.4.4 CRC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
5 Power controller (PWR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116

5.1 Power supplies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .116
5.1.1 Independent A/D converter supply and reference voltage . . . . . . . . . . 117
5.1.2 Battery backup domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
5.1.3 Voltage regulator for STM32F405xx/07xx and STM32F415xx/17xx . . 120
5.1.4 Voltage regulator for STM32F42xxx and STM32F43xxx . . . . . . . . . . . 121
5.2 Power supply supervisor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
5.2.1 Power-on reset (POR)/power-down reset (PDR) . . . . . . . . . . . . . . . . . 124
5.2.2 Brownout reset (BOR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
5.2.3 Programmable voltage detector (PVD) . . . . . . . . . . . . . . . . . . . . . . . . 125
5.3 Low-power modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
5.3.1 Slowing down system clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
5.3.2 Peripheral clock gating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
5.3.3 Sleep mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
5.3.4 Stop mode (STM32F405xx/07xx and STM32F415xx/17xx) . . . . . . . . 130
5.3.5 Stop mode (STM32F42xxx and STM32F43xxx) . . . . . . . . . . . . . . . . . 133
5.3.6 Standby mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
5.3.7 Programming the RTC alternate functions to wake up the device from
the Stop and Standby modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
5.4 Power control registers (STM32F405xx/07xx and STM32F415xx/17xx) 141
5.4.1 PWR power control register (PWR_CR)
for STM32F405xx/07xx and STM32F415xx/17xx . . . . . . . . . . . . . . . . 141
5.4.2 PWR power control/status register (PWR_CSR)
for STM32F405xx/07xx and STM32F415xx/17xx . . . . . . . . . . . . . . . . 142
5.5 Power control registers (STM32F42xxx and STM32F43xxx) . . . . . . . . . 144
4/1751 RM0090 Rev 19

RM0090 Contents
5.6 PWR register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
6 Reset and clock control for

STM32F42xxx and STM32F43xxx (RCC) . . . . . . . . . . . . . . . . . . . . . . . 150
6.1 Reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
6.1.1 System reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
6.1.2 Power reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
6.1.3 Backup domain reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
6.2 Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
6.2.1 HSE clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
6.2.2 HSI clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
6.2.3 PLL configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
6.2.4 LSE clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
6.2.5 LSI clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
6.2.6 System clock (SYSCLK) selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
6.2.7 Clock security system (CSS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
6.2.8 RTC/AWU clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
6.2.9 Watchdog clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
6.2.10 Clock-out capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
6.2.11 Internal/external clock measurement using TIM5/TIM11 . . . . . . . . . . . 158
6.3 RCC registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
6.3.1 RCC clock control register (RCC_CR) . . . . . . . . . . . . . . . . . . . . . . . . . 161
6.3.2 RCC PLL configuration register (RCC_PLLCFGR) . . . . . . . . . . . . . . . 163
6.3.3 RCC clock configuration register (RCC_CFGR) . . . . . . . . . . . . . . . . . 165
6.3.4 RCC clock interrupt register (RCC_CIR) . . . . . . . . . . . . . . . . . . . . . . . 167
6.3.5 RCC AHB1 peripheral reset register (RCC_AHB1RSTR) . . . . . . . . . . 170
6.3.8 RCC APB1 peripheral reset register (RCC_APB1RSTR) . . . . . . . . . . 174
6.3.10 RCC AHB1 peripheral clock register (RCC_AHB1ENR) . . . . . . . . . . . 180
6.3.11 RCC AHB2 peripheral clock enable register (RCC_AHB2ENR) . . . . . 182
6.3.13 RCC APB1 peripheral clock enable register (RCC_APB1ENR) . . . . . 183
6.3.15 RCC AHB1 peripheral clock enable in low power mode register
(RCC_AHB1LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
RM0090 Rev 19 5/1751

39
Contents RM0090
(RCC_AHB2LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
(RCC_AHB3LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
6.3.18 RCC APB1 peripheral clock enable in low power mode register
(RCC_APB1LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
6.3.19 RCC APB2 peripheral clock enabled in low power mode register
(RCC_APB2LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
6.3.20 RCC Backup domain control register (RCC_BDCR) . . . . . . . . . . . . . . 199
6.3.21 RCC clock control & status register (RCC_CSR) . . . . . . . . . . . . . . . . 200
6.3.22 RCC spread spectrum clock generation register (RCC_SSCGR) . . . . 202
6.3.23 RCC PLLI2S configuration register (RCC_PLLI2SCFGR) . . . . . . . . . 203
6.3.24 RCC PLL configuration register (RCC_PLLSAICFGR) . . . . . . . . . . . . 206
6.3.25 RCC Dedicated Clock Configuration Register (RCC_DCKCFGR) . . . 207
6.3.26 RCC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210

STM32F405xx/07xx and STM32F415xx/17xx(RCC) . . . . . . . . . . . . . . 213
7.1 Reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
7.1.1 System reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
7.1.2 Power reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
7.1.3 Backup domain reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
7.2 Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
7.2.1 HSE clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
7.2.2 HSI clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
7.2.3 PLL configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
7.2.4 LSE clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
7.2.5 LSI clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
7.2.6 System clock (SYSCLK) selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
7.2.7 Clock security system (CSS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
7.2.8 RTC/AWU clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
7.2.9 Watchdog clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
7.2.10 Clock-out capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
7.2.11 Internal/external clock measurement using TIM5/TIM11 . . . . . . . . . . . 222
7.3 RCC registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
7.3.1 RCC clock control register (RCC_CR) . . . . . . . . . . . . . . . . . . . . . . . . . 224
7.3.2 RCC PLL configuration register (RCC_PLLCFGR) . . . . . . . . . . . . . . . 226
7.3.3 RCC clock configuration register (RCC_CFGR) . . . . . . . . . . . . . . . . . 228
6/1751 RM0090 Rev 19

RM0090 Contents
7.3.4 RCC clock interrupt register (RCC_CIR) . . . . . . . . . . . . . . . . . . . . . . . 230

7.3.13 RCC APB1 peripheral clock enable register
(RCC_APB1ENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
(RCC_AHB1LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
(RCC_AHB2LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
(RCC_AHB3LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
(RCC_APB1LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
7.3.19 RCC APB2 peripheral clock enabled in low power mode
register (RCC_APB2LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
7.3.20 RCC Backup domain control register (RCC_BDCR) . . . . . . . . . . . . . . 259
7.3.21 RCC clock control & status register (RCC_CSR) . . . . . . . . . . . . . . . . 260
7.3.22 RCC spread spectrum clock generation register (RCC_SSCGR) . . . . 262
7.3.23 RCC PLLI2S configuration register (RCC_PLLI2SCFGR) . . . . . . . . . 263
7.3.24 RCC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
8 General-purpose I/Os (GPIO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267

8.1 GPIO introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
8.2 GPIO main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
8.3 GPIO functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
8.3.1 General-purpose I/O (GPIO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
8.3.2 I/O pin multiplexer and mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
8.3.3 I/O port control registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
8.3.4 I/O port data registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
8.3.5 I/O data bitwise handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
8.3.6 GPIO locking mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
8.3.7 I/O alternate function input/output . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
RM0090 Rev 19 7/1751

39
Contents RM0090
8.3.8 External interrupt/wakeup lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275

8.3.9 Input configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
8.3.10 Output configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
8.3.11 Alternate function configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
8.3.12 Analog configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
8.3.13 Using the OSC32_IN/OSC32_OUT pins as GPIO PC14/PC15
port pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
8.3.14 Using the OSC_IN/OSC_OUT pins as GPIO PH0/PH1 port pins . . . . 278
8.3.15 Selection of RTC_AF1 and RTC_AF2 alternate functions . . . . . . . . . . 279
8.4 GPIO registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
8.4.1 GPIO port mode register (GPIOx_MODER) (x = A..I/J/K) . . . . . . . . . . 281
8.4.2 GPIO port output type register (GPIOx_OTYPER)
(x = A..I/J/K) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
8.4.3 GPIO port output speed register (GPIOx_OSPEEDR)
(x = A..I/J/K) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
8.4.4 GPIO port pull-up/pull-down register (GPIOx_PUPDR)
(x = A..I/J/K) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
8.4.5 GPIO port input data register (GPIOx_IDR) (x = A..I/J/K) . . . . . . . . . . 283
8.4.6 GPIO port output data register (GPIOx_ODR) (x = A..I/J/K) . . . . . . . . 283
8.4.7 GPIO port bit set/reset register (GPIOx_BSRR) (x = A..I/J/K) . . . . . . . 284
8.4.8 GPIO port configuration lock register (GPIOx_LCKR)
(x = A..I/J/K) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284
8.4.9 GPIO alternate function low register (GPIOx_AFRL) (x = A..I/J/K) . . . 285
8.4.10 GPIO alternate function high register (GPIOx_AFRH)
(x = A..I/J) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
8.4.11 GPIO register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
9 System configuration controller (SYSCFG) . . . . . . . . . . . . . . . . . . . . 289

9.1 I/O compensation cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
9.2 SYSCFG registers for STM32F405xx/07xx and STM32F415xx/17xx . . 289
9.2.1 SYSCFG memory remap register (SYSCFG_MEMRMP) . . . . . . . . . . 289
9.2.2 SYSCFG peripheral mode configuration register (SYSCFG_PMC) . . 290
9.2.3 SYSCFG external interrupt configuration register 1
(SYSCFG_EXTICR1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
(SYSCFG_EXTICR2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
(SYSCFG_EXTICR3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
(SYSCFG_EXTICR4) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
8/1751 RM0090 Rev 19

RM0090 Contents
9.2.7 Compensation cell control register (SYSCFG_CMPCR) . . . . . . . . . . . 293

9.2.8 SYSCFG register maps for
9.3 SYSCFG registers for STM32F42xxx and STM32F43xxx . . . . . . . . . . . 294
9.3.1 SYSCFG memory remap register (SYSCFG_MEMRMP) . . . . . . . . . . 294
9.3.2 SYSCFG peripheral mode configuration register (SYSCFG_PMC) . . 296
(SYSCFG_EXTICR1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
(SYSCFG_EXTICR2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
(SYSCFG_EXTICR3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
(SYSCFG_EXTICR4) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
9.3.7 Compensation cell control register (SYSCFG_CMPCR) . . . . . . . . . . . 300
9.3.8 SYSCFG register maps for STM32F42xxx and STM32F43xxx . . . . . . 301
10 DMA controller (DMA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302

10.1 DMA introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
10.2 DMA main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
10.3 DMA functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
10.3.1 General description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
10.3.2 DMA transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
10.3.3 Channel selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
10.3.4 Arbiter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
10.3.5 DMA streams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
10.3.6 Source, destination and transfer modes . . . . . . . . . . . . . . . . . . . . . . . 309
10.3.7 Pointer incrementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
10.3.8 Circular mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
10.3.9 Double buffer mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
10.3.10 Programmable data width, packing/unpacking, endianess . . . . . . . . . 314
10.3.11 Single and burst transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
10.3.12 FIFO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
10.3.13 DMA transfer completion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
10.3.14 DMA transfer suspension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
10.3.15 Flow controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
10.3.16 Summary of the possible DMA configurations . . . . . . . . . . . . . . . . . . . 322
10.3.17 Stream configuration procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
RM0090 Rev 19 9/1751

39
Contents RM0090
10.3.18 Error management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323

10.4 DMA interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
10.5 DMA registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
10.5.1 DMA low interrupt status register (DMA_LISR) . . . . . . . . . . . . . . . . . . 325
10.5.2 DMA high interrupt status register (DMA_HISR) . . . . . . . . . . . . . . . . . 326
10.5.3 DMA low interrupt flag clear register (DMA_LIFCR) . . . . . . . . . . . . . . 327
10.5.4 DMA high interrupt flag clear register (DMA_HIFCR) . . . . . . . . . . . . . 327
10.5.5 DMA stream x configuration register (DMA_SxCR) (x = 0..7) . . . . . . . 328
10.5.6 DMA stream x number of data register (DMA_SxNDTR) (x = 0..7) . . . 331
10.5.7 DMA stream x peripheral address register (DMA_SxPAR) (x = 0..7) . 332
10.5.8 DMA stream x memory 0 address register (DMA_SxM0AR) (x = 0..7) 332
10.5.9 DMA stream x memory 1 address register (DMA_SxM1AR) (x = 0..7) 332
10.5.10 DMA stream x FIFO control register (DMA_SxFCR) (x = 0..7) . . . . . . 333
10.5.11 DMA register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
11 Chrom-Art Accelerator™ controller (DMA2D) . . . . . . . . . . . . . . . . . . 339

11.1 DMA2D introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
11.2 DMA2D main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
11.3 DMA2D functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
11.3.2 DMA2D control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
11.3.3 DMA2D foreground and background FIFOs . . . . . . . . . . . . . . . . . . . . 341
11.3.4 DMA2D foreground and background pixel format converter (PFC) . . . 342
11.3.5 DMA2D foreground and background CLUT interface . . . . . . . . . . . . . 344
11.3.6 DMA2D blender . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
11.3.7 DMA2D output PFC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
11.3.8 DMA2D output FIFO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
11.3.9 DMA2D AHB master port timer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
11.3.10 DMA2D transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
11.3.11 DMA2D configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
11.3.12 DMA2D transfer control (start, suspend, abort and completion) . . . . . 350
11.3.13 Watermark . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
11.3.15 AHB dead time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
11.4 DMA2D interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
11.5 DMA2D registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
10/1751 RM0090 Rev 19

RM0090 Contents
11.5.1 DMA2D control register (DMA2D_CR) . . . . . . . . . . . . . . . . . . . . . . . . 352

11.5.2 DMA2D Interrupt Status Register (DMA2D_ISR) . . . . . . . . . . . . . . . . 354
11.5.3 DMA2D interrupt flag clear register (DMA2D_IFCR) . . . . . . . . . . . . . . 355
11.5.4 DMA2D foreground memory address register (DMA2D_FGMAR) . . . 356
11.5.5 DMA2D foreground offset register (DMA2D_FGOR) . . . . . . . . . . . . . . 356
11.5.6 DMA2D background memory address register (DMA2D_BGMAR) . . 357
11.5.7 DMA2D background offset register (DMA2D_BGOR) . . . . . . . . . . . . . 357
11.5.8 DMA2D foreground PFC control register (DMA2D_FGPFCCR) . . . . . 358
11.5.9 DMA2D foreground color register (DMA2D_FGCOLR) . . . . . . . . . . . . 360
11.5.10 DMA2D background PFC control register (DMA2D_BGPFCCR) . . . . 361
11.5.11 DMA2D background color register (DMA2D_BGCOLR) . . . . . . . . . . . 363
11.5.12 DMA2D foreground CLUT memory address register
(DMA2D_FGCMAR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363
11.5.13 DMA2D background CLUT memory address register
(DMA2D_BGCMAR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364
11.5.14 DMA2D output PFC control register (DMA2D_OPFCCR) . . . . . . . . . . 364
11.5.15 DMA2D output color register (DMA2D_OCOLR) . . . . . . . . . . . . . . . . . 365
11.5.16 DMA2D output memory address register (DMA2D_OMAR) . . . . . . . . 366
11.5.17 DMA2D output offset register (DMA2D_OOR) . . . . . . . . . . . . . . . . . . 367
11.5.18 DMA2D number of line register (DMA2D_NLR) . . . . . . . . . . . . . . . . . 367
11.5.19 DMA2D line watermark register (DMA2D_LWR) . . . . . . . . . . . . . . . . . 368
11.5.20 DMA2D AHB master timer configuration register (DMA2D_AMTCR) . 368
11.5.21 DMA2D register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
12 Interrupts and events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371

12.1 Nested vectored interrupt controller (NVIC) . . . . . . . . . . . . . . . . . . . . . . 371
12.1.1 NVIC features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371
12.1.2 SysTick calibration value register . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371
12.1.3 Interrupt and exception vectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371
12.2 External interrupt/event controller (EXTI) . . . . . . . . . . . . . . . . . . . . . . . . 371
12.2.1 EXTI main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379
12.2.2 EXTI block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380
12.2.3 Wakeup event management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380
12.2.4 Functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380
12.2.5 External interrupt/event line mapping . . . . . . . . . . . . . . . . . . . . . . . . . 382
12.3 EXTI registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
12.3.1 Interrupt mask register (EXTI_IMR) . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
RM0090 Rev 19 11/1751

39
Contents RM0090
12.3.2 Event mask register (EXTI_EMR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384

12.3.3 Rising trigger selection register (EXTI_RTSR) . . . . . . . . . . . . . . . . . . 385
12.3.4 Falling trigger selection register (EXTI_FTSR) . . . . . . . . . . . . . . . . . . 385
12.3.5 Software interrupt event register (EXTI_SWIER) . . . . . . . . . . . . . . . . 386
12.3.6 Pending register (EXTI_PR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
12.3.7 EXTI register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
13 Analog-to-digital converter (ADC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388

13.1 ADC introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388
13.2 ADC main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388
13.3 ADC functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
13.3.1 ADC on-off control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390
13.3.2 ADC clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390
13.3.3 Channel selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390
13.3.4 Single conversion mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391
13.3.5 Continuous conversion mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392
13.3.6 Timing diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392
13.3.7 Analog watchdog . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392
13.3.8 Scan mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393
13.3.9 Injected channel management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394
13.3.10 Discontinuous mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395
13.4 Data alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
13.5 Channel-wise programmable sampling time . . . . . . . . . . . . . . . . . . . . . 397
13.6 Conversion on external trigger and trigger polarity . . . . . . . . . . . . . . . . 397
13.7 Fast conversion mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
13.8 Data management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
13.8.1 Using the DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
13.8.2 Managing a sequence of conversions without using the DMA . . . . . . 400
13.8.3 Conversions without DMA and without overrun detection . . . . . . . . . . 401
13.9 Multi ADC mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
13.9.1 Injected simultaneous mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404
13.9.2 Regular simultaneous mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405
13.9.3 Interleaved mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
13.9.4 Alternate trigger mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408
13.9.5 Combined regular/injected simultaneous mode . . . . . . . . . . . . . . . . . . 410
13.9.6 Combined regular simultaneous + alternate trigger mode . . . . . . . . . . 411
12/1751 RM0090 Rev 19

RM0090 Contents
13.10 Temperature sensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412

13.11 Battery charge monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
13.12 ADC interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
13.13 ADC registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
13.13.1 ADC status register (ADC_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
13.13.2 ADC control register 1 (ADC_CR1) . . . . . . . . . . . . . . . . . . . . . . . . . . . 416
13.13.3 ADC control register 2 (ADC_CR2) . . . . . . . . . . . . . . . . . . . . . . . . . . . 418
13.13.4 ADC sample time register 1 (ADC_SMPR1) . . . . . . . . . . . . . . . . . . . . 420
13.13.5 ADC sample time register 2 (ADC_SMPR2) . . . . . . . . . . . . . . . . . . . . 420
13.13.6 ADC injected channel data offset register x (ADC_JOFRx) (x=1..4) . . 421
13.13.7 ADC watchdog higher threshold register (ADC_HTR) . . . . . . . . . . . . . 421
13.13.8 ADC watchdog lower threshold register (ADC_LTR) . . . . . . . . . . . . . . 422
13.13.9 ADC regular sequence register 1 (ADC_SQR1) . . . . . . . . . . . . . . . . . 422
13.13.12 ADC injected sequence register (ADC_JSQR) . . . . . . . . . . . . . . . . . . 424
13.13.13 ADC injected data register x (ADC_JDRx) (x= 1..4) . . . . . . . . . . . . . . 425
13.13.14 ADC regular data register (ADC_DR) . . . . . . . . . . . . . . . . . . . . . . . . . 425
13.13.15 ADC Common status register (ADC_CSR) . . . . . . . . . . . . . . . . . . . . . 426
13.13.16 ADC common control register (ADC_CCR) . . . . . . . . . . . . . . . . . . . . . 427
13.13.17 ADC common regular data register for dual and triple modes
(ADC_CDR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
13.13.18 ADC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
14 Digital-to-analog converter (DAC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433

14.1 DAC introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
14.2 DAC main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
14.3 DAC functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
14.3.1 DAC channel enable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
14.3.2 DAC output buffer enable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
14.3.3 DAC data format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
14.3.4 DAC conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
14.3.5 DAC output voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
14.3.6 DAC trigger selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
14.3.7 DMA request . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438
14.3.8 Noise generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438
14.3.9 Triangle-wave generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
RM0090 Rev 19 13/1751

39
Contents RM0090
14.4 Dual DAC channel conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440

14.4.1 Independent trigger without wave generation . . . . . . . . . . . . . . . . . . . 441
14.4.2 Independent trigger with single LFSR generation . . . . . . . . . . . . . . . . 441
14.4.3 Independent trigger with different LFSR generation . . . . . . . . . . . . . . 441
14.4.4 Independent trigger with single triangle generation . . . . . . . . . . . . . . . 442
14.4.5 Independent trigger with different triangle generation . . . . . . . . . . . . . 442
14.4.6 Simultaneous software start . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442
14.4.7 Simultaneous trigger without wave generation . . . . . . . . . . . . . . . . . . 443
14.4.8 Simultaneous trigger with single LFSR generation . . . . . . . . . . . . . . . 443
14.4.9 Simultaneous trigger with different LFSR generation . . . . . . . . . . . . . 443
14.4.10 Simultaneous trigger with single triangle generation . . . . . . . . . . . . . . 444
14.4.11 Simultaneous trigger with different triangle generation . . . . . . . . . . . . 444
14.5 DAC registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445
14.5.1 DAC control register (DAC_CR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445
14.5.2 DAC software trigger register (DAC_SWTRIGR) . . . . . . . . . . . . . . . . 448
14.5.3 DAC channel1 12-bit right-aligned data holding register
(DAC_DHR12R1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448
14.5.4 DAC channel1 12-bit left aligned data holding register
(DAC_DHR12L1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449
14.5.5 DAC channel1 8-bit right aligned data holding register
(DAC_DHR8R1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449
(DAC_DHR12R2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450
(DAC_DHR12L2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450
(DAC_DHR8R2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450
14.5.9 Dual DAC 12-bit right-aligned data holding register
(DAC_DHR12RD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
14.5.10 DUAL DAC 12-bit left aligned data holding register
(DAC_DHR12LD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
14.5.11 DUAL DAC 8-bit right aligned data holding register
(DAC_DHR8RD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452
14.5.12 DAC channel1 data output register (DAC_DOR1) . . . . . . . . . . . . . . . . 452
14.5.13 DAC channel2 data output register (DAC_DOR2) . . . . . . . . . . . . . . . . 452
14.5.14 DAC status register (DAC_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
14.5.15 DAC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
15 Digital camera interface (DCMI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
14/1751 RM0090 Rev 19

RM0090 Contents
15.1 DCMI introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455

15.2 DCMI main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
15.3 DCMI pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
15.4 DCMI clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
15.5 DCMI functional overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456
15.5.1 DMA interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
15.5.2 DCMI physical interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
15.5.3 Synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459
15.5.4 Capture modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 462
15.5.5 Crop feature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463
15.5.6 JPEG format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
15.5.7 FIFO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
15.6 Data format description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
15.6.1 Data formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
15.6.2 Monochrome format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
15.6.3 RGB format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
15.6.4 YCbCr format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
15.7 DCMI interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467
15.8 DCMI register description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467
15.8.1 DCMI control register 1 (DCMI_CR) . . . . . . . . . . . . . . . . . . . . . . . . . . 467
15.8.2 DCMI status register (DCMI_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470
15.8.3 DCMI raw interrupt status register (DCMI_RIS) . . . . . . . . . . . . . . . . . 471
15.8.4 DCMI interrupt enable register (DCMI_IER) . . . . . . . . . . . . . . . . . . . . 472
15.8.5 DCMI masked interrupt status register (DCMI_MIS) . . . . . . . . . . . . . . 473
15.8.6 DCMI interrupt clear register (DCMI_ICR) . . . . . . . . . . . . . . . . . . . . . . 474
15.8.7 DCMI embedded synchronization code register (DCMI_ESCR) . . . . . 475
15.8.8 DCMI embedded synchronization unmask register (DCMI_ESUR) . . 476
15.8.9 DCMI crop window start (DCMI_CWSTRT) . . . . . . . . . . . . . . . . . . . . . 477
15.8.10 DCMI crop window size (DCMI_CWSIZE) . . . . . . . . . . . . . . . . . . . . . . 477
15.8.11 DCMI data register (DCMI_DR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478
15.8.12 DCMI register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478
16 LCD-TFT controller (LTDC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480

16.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480
16.2 LTDC main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480
16.3 LTDC functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 481
RM0090 Rev 19 15/1751

39
Contents RM0090
16.3.1 LTDC block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 481

16.3.2 LTDC reset and clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 481
16.3.3 LCD-TFT pins and signal interface . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
16.4 LTDC programmable parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
16.4.1 LTDC Global configuration parameters . . . . . . . . . . . . . . . . . . . . . . . . 483
16.4.2 Layer programmable parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . 486
16.5 LTDC interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 490
16.6 LTDC programming procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 492
16.7 LTDC registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 493
16.7.1 LTDC Synchronization Size Configuration Register (LTDC_SSCR) . . 493
16.7.2 LTDC Back Porch Configuration Register (LTDC_BPCR) . . . . . . . . . . 493
16.7.3 LTDC Active Width Configuration Register (LTDC_AWCR) . . . . . . . . . 494
16.7.4 LTDC Total Width Configuration Register (LTDC_TWCR) . . . . . . . . . . 495
16.7.5 LTDC Global Control Register (LTDC_GCR) . . . . . . . . . . . . . . . . . . . . 495
16.7.6 LTDC Shadow Reload Configuration Register (LTDC_SRCR) . . . . . . 497
16.7.7 LTDC Background Color Configuration Register (LTDC_BCCR) . . . . 497
16.7.8 LTDC Interrupt Enable Register (LTDC_IER) . . . . . . . . . . . . . . . . . . . 498
16.7.9 LTDC Interrupt Status Register (LTDC_ISR) . . . . . . . . . . . . . . . . . . . . 499
16.7.10 LTDC Interrupt Clear Register (LTDC_ICR) . . . . . . . . . . . . . . . . . . . . . 499
16.7.11 LTDC Line Interrupt Position Configuration Register (LTDC_LIPCR) . 500
16.7.12 LTDC Current Position Status Register (LTDC_CPSR) . . . . . . . . . . . . 500
16.7.13 LTDC Current Display Status Register (LTDC_CDSR) . . . . . . . . . . . . 501
16.7.14 LTDC Layerx Control Register (LTDC_LxCR) (where x=1..2) . . . . . . . 502
16.7.15 LTDC Layerx Window Horizontal Position Configuration Register
(LTDC_LxWHPCR) (where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . 503
16.7.16 LTDC Layerx Window Vertical Position Configuration Register
(LTDC_LxWVPCR) (where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . 504
16.7.17 LTDC Layerx Color Keying Configuration Register (LTDC_LxCKCR)
(where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505
16.7.18 LTDC Layerx Pixel Format Configuration Register (LTDC_LxPFCR)
(where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505
16.7.19 LTDC Layerx Constant Alpha Configuration Register (LTDC_LxCACR)
(where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 506
16.7.20 LTDC Layerx Default Color Configuration Register (LTDC_LxDCCR)
(where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 506
16.7.21 LTDC Layerx Blending Factors Configuration Register (LTDC_LxBFCR)
(where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 508
16.7.22 LTDC Layerx Color Frame Buffer Address Register (LTDC_LxCFBAR)
(where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 509
16/1751 RM0090 Rev 19

RM0090 Contents
16.7.23 LTDC Layerx Color Frame Buffer Length Register (LTDC_LxCFBLR)

(where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 509
16.7.24 LTDC Layerx ColorFrame Buffer Line Number Register
(LTDC_LxCFBLNR) (where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . 510
16.7.25 LTDC Layerx CLUT Write Register (LTDC_LxCLUTWR)
(where x=1..2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 511
16.7.26 LTDC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 512
17 Advanced-control timers (TIM1 and TIM8) . . . . . . . . . . . . . . . . . . . . . 515

17.1 TIM1 and TIM8 introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515
17.2 TIM1 and TIM8 main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 516
17.3 TIM1 and TIM8 functional description . . . . . . . . . . . . . . . . . . . . . . . . . . 518
17.3.1 Time-base unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 518
17.3.2 Counter modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 520
17.3.3 Repetition counter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 529
17.3.4 Clock selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 532
17.3.5 Capture/compare channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535
17.3.6 Input capture mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538
17.3.7 PWM input mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 539
17.3.8 Forced output mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 539
17.3.9 Output compare mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 540
17.3.10 PWM mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 541
17.3.11 Complementary outputs and dead-time insertion . . . . . . . . . . . . . . . . 544
17.3.12 Using the break function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 546
17.3.13 Clearing the OCxREF signal on an external event . . . . . . . . . . . . . . . 549
17.3.14 6-step PWM generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 550
17.3.15 One-pulse mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551
17.3.16 Encoder interface mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 552
17.3.17 Timer input XOR function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 555
17.3.18 Interfacing with Hall sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 555
17.3.19 TIMx and external trigger synchronization . . . . . . . . . . . . . . . . . . . . . . 557
17.3.20 Timer synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 560
17.3.21 Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 560
17.4 TIM1 and TIM8 registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561
17.4.1 TIM1 and TIM8 control register 1 (TIMx_CR1) . . . . . . . . . . . . . . . . . . 561
17.4.3 TIM1 and TIM8 slave mode control register (TIMx_SMCR) . . . . . . . . 565
17.4.4 TIM1 and TIM8 DMA/interrupt enable register (TIMx_DIER) . . . . . . . . 567
RM0090 Rev 19 17/1751

39
Contents RM0090
17.4.5 TIM1 and TIM8 status register (TIMx_SR) . . . . . . . . . . . . . . . . . . . . . . 569

17.4.6 TIM1 and TIM8 event generation register (TIMx_EGR) . . . . . . . . . . . . 570
17.4.7 TIM1 and TIM8 capture/compare mode register 1 (TIMx_CCMR1) . . 572
17.4.8 TIM1 and TIM8 capture/compare mode register 2 (TIMx_CCMR2) . . 575
17.4.9 TIM1 and TIM8 capture/compare enable register (TIMx_CCER) . . . . 576
17.4.10 TIM1 and TIM8 counter (TIMx_CNT) . . . . . . . . . . . . . . . . . . . . . . . . . . 580
17.4.11 TIM1 and TIM8 prescaler (TIMx_PSC) . . . . . . . . . . . . . . . . . . . . . . . . 580
17.4.12 TIM1 and TIM8 auto-reload register (TIMx_ARR) . . . . . . . . . . . . . . . . 580
17.4.13 TIM1 and TIM8 repetition counter register (TIMx_RCR) . . . . . . . . . . . 581
17.4.14 TIM1 and TIM8 capture/compare register 1 (TIMx_CCR1) . . . . . . . . . 581
17.4.18 TIM1 and TIM8 break and dead-time register (TIMx_BDTR) . . . . . . . 583
17.4.19 TIM1 and TIM8 DMA control register (TIMx_DCR) . . . . . . . . . . . . . . . 585
17.4.20 TIM1 and TIM8 DMA address for full transfer (TIMx_DMAR) . . . . . . . 586
17.4.21 TIM1 and TIM8 register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
18 General-purpose timers (TIM2 to TIM5) . . . . . . . . . . . . . . . . . . . . . . . . 589

18.1 TIM2 to TIM5 introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
18.2 TIM2 to TIM5 main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
18.3 TIM2 to TIM5 functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
18.3.1 Time-base unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
18.3.2 Counter modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592
18.3.3 Clock selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602
18.3.6 PWM input mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608
18.3.9 PWM mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 610
18.3.10 One-pulse mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 613
18.3.11 Clearing the OCxREF signal on an external event . . . . . . . . . . . . . . . 614
18.3.12 Encoder interface mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615
18.3.13 Timer input XOR function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 618
18.3.14 Timers and external trigger synchronization . . . . . . . . . . . . . . . . . . . . 618
18.3.15 Timer synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 621
18/1751 RM0090 Rev 19

RM0090 Contents
18.3.16 Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626

18.4 TIM2 to TIM5 registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 627
18.4.1 TIMx control register 1 (TIMx_CR1) . . . . . . . . . . . . . . . . . . . . . . . . . . 627
18.4.2 TIMx control register 2 (TIMx_CR2) . . . . . . . . . . . . . . . . . . . . . . . . . . 629
18.4.3 TIMx slave mode control register (TIMx_SMCR) . . . . . . . . . . . . . . . . . 630
18.4.4 TIMx DMA/Interrupt enable register (TIMx_DIER) . . . . . . . . . . . . . . . . 632
18.4.5 TIMx status register (TIMx_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 633
18.4.6 TIMx event generation register (TIMx_EGR) . . . . . . . . . . . . . . . . . . . . 635
18.4.7 TIMx capture/compare mode register 1 (TIMx_CCMR1) . . . . . . . . . . . 636
18.4.8 TIMx capture/compare mode register 2 (TIMx_CCMR2) . . . . . . . . . . . 639
18.4.9 TIMx capture/compare enable register (TIMx_CCER) . . . . . . . . . . . . 640
18.4.10 TIMx counter (TIMx_CNT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 642
18.4.11 TIMx prescaler (TIMx_PSC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 642
18.4.12 TIMx auto-reload register (TIMx_ARR) . . . . . . . . . . . . . . . . . . . . . . . . 642
18.4.13 TIMx capture/compare register 1 (TIMx_CCR1) . . . . . . . . . . . . . . . . . 643
18.4.17 TIMx DMA control register (TIMx_DCR) . . . . . . . . . . . . . . . . . . . . . . . 645
18.4.18 TIMx DMA address for full transfer (TIMx_DMAR) . . . . . . . . . . . . . . . 646
18.4.19 TIM2 option register (TIM2_OR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 647
18.4.20 TIM5 option register (TIM5_OR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 647
18.4.21 TIMx register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 648
19 General-purpose timers (TIM9 to TIM14) . . . . . . . . . . . . . . . . . . . . . . . 650

19.1 TIM9 to TIM14 introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650
19.2 TIM9 to TIM14 main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650
19.2.1 TIM9/TIM12 main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650
19.2.2 TIM10/TIM11 and TIM13/TIM14 main features . . . . . . . . . . . . . . . . . . 651
19.3 TIM9 to TIM14 functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . 653
19.3.1 Time-base unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 653
19.3.2 Counter modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655
19.3.3 Clock selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 658
19.3.6 PWM input mode (only for TIM9/12) . . . . . . . . . . . . . . . . . . . . . . . . . . 663
RM0090 Rev 19 19/1751

39
Contents RM0090

19.3.9 PWM mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665
19.3.10 One-pulse mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 666
19.3.11 TIM9/12 external trigger synchronization . . . . . . . . . . . . . . . . . . . . . . . 668
19.3.12 Timer synchronization (TIM9/12) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671
19.3.13 Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671
19.4 TIM9 and TIM12 registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 672
19.4.1 TIM9/12 control register 1 (TIMx_CR1) . . . . . . . . . . . . . . . . . . . . . . . . 672
19.4.2 TIM9/12 slave mode control register (TIMx_SMCR) . . . . . . . . . . . . . . 673
19.4.3 TIM9/12 Interrupt enable register (TIMx_DIER) . . . . . . . . . . . . . . . . . 674
19.4.4 TIM9/12 status register (TIMx_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . 676
19.4.5 TIM9/12 event generation register (TIMx_EGR) . . . . . . . . . . . . . . . . . 677
19.4.6 TIM9/12 capture/compare mode register 1 (TIMx_CCMR1) . . . . . . . . 678
19.4.7 TIM9/12 capture/compare enable register (TIMx_CCER) . . . . . . . . . . 681
19.4.8 TIM9/12 counter (TIMx_CNT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 682
19.4.9 TIM9/12 prescaler (TIMx_PSC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 682
19.4.10 TIM9/12 auto-reload register (TIMx_ARR) . . . . . . . . . . . . . . . . . . . . . 682
19.4.11 TIM9/12 capture/compare register 1 (TIMx_CCR1) . . . . . . . . . . . . . . 683
19.4.12 TIM9/12 capture/compare register 2 (TIMx_CCR2) . . . . . . . . . . . . . . 683
19.4.13 TIM9/12 register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684
19.5 TIM10/11/13/14 registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 686
19.5.1 TIM10/11/13/14 control register 1 (TIMx_CR1) . . . . . . . . . . . . . . . . . . 686
19.5.2 TIM10/11/13/14 Interrupt enable register (TIMx_DIER) . . . . . . . . . . . . 687
19.5.3 TIM10/11/13/14 status register (TIMx_SR) . . . . . . . . . . . . . . . . . . . . . 687
19.5.4 TIM10/11/13/14 event generation register (TIMx_EGR) . . . . . . . . . . . 688
19.5.5 TIM10/11/13/14 capture/compare mode register 1 (TIMx_CCMR1) . . 688
19.5.6 TIM10/11/13/14 capture/compare enable register (TIMx_CCER) . . . . 691
19.5.7 TIM10/11/13/14 counter (TIMx_CNT) . . . . . . . . . . . . . . . . . . . . . . . . . 692
19.5.8 TIM10/11/13/14 prescaler (TIMx_PSC) . . . . . . . . . . . . . . . . . . . . . . . . 692
19.5.9 TIM10/11/13/14 auto-reload register (TIMx_ARR) . . . . . . . . . . . . . . . . 692
19.5.10 TIM10/11/13/14 capture/compare register 1 (TIMx_CCR1) . . . . . . . . . 693
19.5.11 TIM11 option register 1 (TIM11_OR) . . . . . . . . . . . . . . . . . . . . . . . . . . 693
19.5.12 TIM10/11/13/14 register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 694
20 Basic timers (TIM6 and TIM7) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 696

20.1 TIM6 and TIM7 introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 696
20.2 TIM6 and TIM7 main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 696
20/1751 RM0090 Rev 19

RM0090 Contents
20.3 TIM6 and TIM7 functional description . . . . . . . . . . . . . . . . . . . . . . . . . . 697

20.3.1 Time-base unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 697
20.3.2 Counting mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 699
20.3.3 Clock source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 701
20.3.4 Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 702
20.4 TIM6 and TIM7 registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 702
20.4.3 TIM6 and TIM7 DMA/Interrupt enable register (TIMx_DIER) . . . . . . . 704
20.4.4 TIM6 and TIM7 status register (TIMx_SR) . . . . . . . . . . . . . . . . . . . . . . 705
20.4.5 TIM6 and TIM7 event generation register (TIMx_EGR) . . . . . . . . . . . . 705
20.4.6 TIM6 and TIM7 counter (TIMx_CNT) . . . . . . . . . . . . . . . . . . . . . . . . . . 705
20.4.7 TIM6 and TIM7 prescaler (TIMx_PSC) . . . . . . . . . . . . . . . . . . . . . . . . 706
20.4.8 TIM6 and TIM7 auto-reload register (TIMx_ARR) . . . . . . . . . . . . . . . . 706
20.4.9 TIM6 and TIM7 register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707
21 Independent watchdog (IWDG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708

21.1 IWDG introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708
21.2 IWDG main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708
21.3 IWDG functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708
21.3.1 Hardware watchdog . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708
21.3.2 Register access protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708
21.3.3 Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 709
21.4 IWDG registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 710
21.4.1 Key register (IWDG_KR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 710
21.4.2 Prescaler register (IWDG_PR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 710
21.4.3 Reload register (IWDG_RLR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 711
21.4.4 Status register (IWDG_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 711
21.4.5 IWDG register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 712
22 Window watchdog (WWDG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 713

22.1 WWDG introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 713
22.2 WWDG main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 713
22.3 WWDG functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 713
22.4 How to program the watchdog timeout . . . . . . . . . . . . . . . . . . . . . . . . . . 715
22.5 Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 716
RM0090 Rev 19 21/1751

39
Contents RM0090
22.6 WWDG registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 717

22.6.1 Control register (WWDG_CR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 717
22.6.2 Configuration register (WWDG_CFR) . . . . . . . . . . . . . . . . . . . . . . . . . 718
22.6.3 Status register (WWDG_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 718
22.6.4 WWDG register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 719
23 Cryptographic processor (CRYP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 720

23.1 CRYP introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 720
23.2 CRYP main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 720
23.3 CRYP functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 722
23.3.1 DES/TDES cryptographic core . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 723
23.3.2 AES cryptographic core . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 728
23.3.3 Data type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 739
23.3.4 Initialization vectors - CRYP_IV0...1(L/R) . . . . . . . . . . . . . . . . . . . . . . 742
23.3.5 CRYP busy state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 743
23.3.6 Procedure to perform an encryption or a decryption . . . . . . . . . . . . . . 744
23.3.7 Context swapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745
23.4 CRYP interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 747
23.5 CRYP DMA interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 748
23.6 CRYP registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 748
23.6.1 CRYP control register (CRYP_CR) for STM32F415/417xx . . . . . . . . . 748
23.6.2 CRYP control register (CRYP_CR) for STM32F415/417xx . . . . . . . . . 750
23.6.3 CRYP status register (CRYP_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 753
23.6.4 CRYP data input register (CRYP_DIN) . . . . . . . . . . . . . . . . . . . . . . . . 754
23.6.5 CRYP data output register (CRYP_DOUT) . . . . . . . . . . . . . . . . . . . . . 755
23.6.6 CRYP DMA control register (CRYP_DMACR) . . . . . . . . . . . . . . . . . . . 756
23.6.7 CRYP interrupt mask set/clear register (CRYP_IMSCR) . . . . . . . . . . . 756
23.6.8 CRYP raw interrupt status register (CRYP_RISR) . . . . . . . . . . . . . . . 757
23.6.9 CRYP masked interrupt status register (CRYP_MISR) . . . . . . . . . . . . 757
23.6.10 CRYP key registers (CRYP_K0...3(L/R)R) . . . . . . . . . . . . . . . . . . . . . 758
23.6.11 CRYP initialization vector registers (CRYP_IV0...1(L/R)R) . . . . . . . . . 760
23.6.12 CRYP context swap registers (CRYP_CSGCMCCM0..7R and
CRYP_CSGCM0..7R) for STM32F42xxx and STM32F43xxx . . . . . . . 762
23.6.13 CRYP register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 763
24 Random number generator (RNG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767

24.1 RNG introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767
22/1751 RM0090 Rev 19

RM0090 Contents
24.2 RNG main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767

24.3 RNG functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767
24.3.1 Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 768
24.4 RNG registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 768
24.4.1 RNG control register (RNG_CR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 769
24.4.2 RNG status register (RNG_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 769
24.4.3 RNG data register (RNG_DR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 770
24.4.4 RNG register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771
25 Hash processor (HASH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 772

25.1 HASH introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 772
25.2 HASH main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 772
25.3 HASH functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 773
25.3.1 Duration of the processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 775
25.3.2 Data type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 775
25.3.3 Message digest computing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 777
25.3.4 Message padding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 778
25.3.5 Hash operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 779
25.3.6 HMAC operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 779
25.3.7 Context swapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 780
25.3.8 HASH interrupt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 782
25.4 HASH registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 782
25.4.1 HASH control register (HASH_CR) for STM32F415/417xx . . . . . . . . . 782
25.4.2 HASH control register (HASH_CR) for STM32F43xxx . . . . . . . . . . . . 785
25.4.3 HASH data input register (HASH_DIN) . . . . . . . . . . . . . . . . . . . . . . . . 788
25.4.4 HASH start register (HASH_STR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789
25.4.5 HASH digest registers (HASH_HR0..4/5/6/7) . . . . . . . . . . . . . . . . . . . 790
25.4.6 HASH interrupt enable register (HASH_IMR) . . . . . . . . . . . . . . . . . . . 792
25.4.7 HASH status register (HASH_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 793
25.4.8 HASH context swap registers (HASH_CSRx) . . . . . . . . . . . . . . . . . . . 794
25.4.9 HASH register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 795
26 Real-time clock (RTC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 798

26.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 798
26.2 RTC main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 799
RM0090 Rev 19 23/1751

39
Contents RM0090
26.3 RTC functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 801

26.3.1 Clock and prescalers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 801
26.3.2 Real-time clock and calendar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 802
26.3.3 Programmable alarms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 802
26.3.4 Periodic auto-wakeup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 802
26.3.5 RTC initialization and configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . 803
26.3.6 Reading the calendar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 805
26.3.7 Resetting the RTC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 806
26.3.8 RTC synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 806
26.3.9 RTC reference clock detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 807
26.3.10 RTC coarse digital calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 808
26.3.11 RTC smooth digital calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 809
26.3.12 Timestamp function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 811
26.3.13 Tamper detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 811
26.3.14 Calibration clock output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 813
26.3.15 Alarm output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 813
26.4 RTC and low-power modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 814
26.5 RTC interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 814
26.6 RTC registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 816
26.6.1 RTC time register (RTC_TR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 816
26.6.2 RTC date register (RTC_DR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 817
26.6.3 RTC control register (RTC_CR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 818
26.6.4 RTC initialization and status register (RTC_ISR) . . . . . . . . . . . . . . . . . 820
26.6.5 RTC prescaler register (RTC_PRER) . . . . . . . . . . . . . . . . . . . . . . . . . 823
26.6.6 RTC wakeup timer register (RTC_WUTR) . . . . . . . . . . . . . . . . . . . . . . 823
26.6.7 RTC calibration register (RTC_CALIBR) . . . . . . . . . . . . . . . . . . . . . . . 824
26.6.8 RTC alarm A register (RTC_ALRMAR) . . . . . . . . . . . . . . . . . . . . . . . . 825
26.6.9 RTC alarm B register (RTC_ALRMBR) . . . . . . . . . . . . . . . . . . . . . . . . 826
26.6.10 RTC write protection register (RTC_WPR) . . . . . . . . . . . . . . . . . . . . . 827
26.6.11 RTC sub second register (RTC_SSR) . . . . . . . . . . . . . . . . . . . . . . . . . 827
26.6.12 RTC shift control register (RTC_SHIFTR) . . . . . . . . . . . . . . . . . . . . . . 828
26.6.13 RTC time stamp time register (RTC_TSTR) . . . . . . . . . . . . . . . . . . . . 828
26.6.14 RTC time stamp date register (RTC_TSDR) . . . . . . . . . . . . . . . . . . . . 829
26.6.15 RTC timestamp sub second register (RTC_TSSSR) . . . . . . . . . . . . . . 830
26.6.16 RTC calibration register (RTC_CALR) . . . . . . . . . . . . . . . . . . . . . . . . . 830
26.6.17 RTC tamper and alternate function configuration register
(RTC_TAFCR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 832
24/1751 RM0090 Rev 19

RM0090 Contents
26.6.18 RTC alarm A sub second register (RTC_ALRMASSR) . . . . . . . . . . . . 834

26.6.19 RTC alarm B sub second register (RTC_ALRMBSSR) . . . . . . . . . . . . 835
26.6.20 RTC backup registers (RTC_BKPxR) . . . . . . . . . . . . . . . . . . . . . . . . . 836
26.6.21 RTC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 836
27 Inter-integrated circuit (I2C) interface . . . . . . . . . . . . . . . . . . . . . . . . . 839

27.1 I2C introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 839
27.2 I2C main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 839
27.3 I2C functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 840
27.3.1 Mode selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 840
27.3.2 I2C slave mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 843
27.3.3 I2C master mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 846
27.3.4 Error conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 851
27.3.5 Programmable noise filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 852
27.3.6 SDA/SCL line control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 853
27.3.7 SMBus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 853
27.3.8 DMA requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 856
27.3.9 Packet error checking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 857
27.4 I2C interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 858
27.5 I2C debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 860
27.6 I2C registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 860
27.6.1 I2C Control register 1 (I2C_CR1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 860
27.6.2 I2C Control register 2 (I2C_CR2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 862
27.6.3 I2C Own address register 1 (I2C_OAR1) . . . . . . . . . . . . . . . . . . . . . . . 864
27.6.4 I2C Own address register 2 (I2C_OAR2) . . . . . . . . . . . . . . . . . . . . . . . 864
27.6.5 I2C Data register (I2C_DR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865
27.6.6 I2C Status register 1 (I2C_SR1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865
27.6.7 I2C Status register 2 (I2C_SR2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 868
27.6.8 I2C Clock control register (I2C_CCR) . . . . . . . . . . . . . . . . . . . . . . . . . 870
27.6.9 I2C TRISE register (I2C_TRISE) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 871
27.6.10 I2C FLTR register (I2C_FLTR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 871
27.6.11 I2C register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 872
28 Serial peripheral interface (SPI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 873

28.1 SPI introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 873
28.2 SPI and I2S main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 874
RM0090 Rev 19 25/1751

39
Contents RM0090
28.2.1 SPI features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 874

28.2.2 I2S features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 875
28.3 SPI functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 876
28.3.2 Configuring the SPI in slave mode . . . . . . . . . . . . . . . . . . . . . . . . . . . 880
28.3.3 Configuring the SPI in master mode . . . . . . . . . . . . . . . . . . . . . . . . . . 882
28.3.4 Configuring the SPI for half-duplex communication . . . . . . . . . . . . . . . 884
28.3.5 Data transmission and reception procedures . . . . . . . . . . . . . . . . . . . 885
28.3.6 CRC calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 891
28.3.7 Status flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 893
28.3.8 Disabling the SPI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 894
28.3.9 SPI communication using DMA (direct memory addressing) . . . . . . . . 895
28.3.10 Error flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 897
28.3.11 SPI interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 898
28.4 I2S functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 899
28.4.1 I2S general description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 899
28.4.2 I2S full duplex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 900
28.4.3 Supported audio protocols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 901
28.4.4 Clock generator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 907
28.4.5 I2S master mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 909
28.4.6 I2S slave mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 911
28.4.7 Status flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 913
28.4.8 Error flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 914
28.4.9 I2S interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 915
28.4.10 DMA features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 915
28.5 SPI and I2S registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 916
28.5.1 SPI control register 1 (SPI_CR1) (not used in I2S mode) . . . . . . . . . . 916
28.5.2 SPI control register 2 (SPI_CR2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 918
28.5.3 SPI status register (SPI_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 919
28.5.4 SPI data register (SPI_DR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 920
28.5.5 SPI CRC polynomial register (SPI_CRCPR) (not used in I2S
mode) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 921
28.5.6 SPI RX CRC register (SPI_RXCRCR) (not used in I2S mode) . . . . . . 921
28.5.7 SPI TX CRC register (SPI_TXCRCR) (not used in I2S mode) . . . . . . . 922
28.5.8 SPI_I2S configuration register (SPI_I2SCFGR) . . . . . . . . . . . . . . . . . . 922
28.5.9 SPI_I2S prescaler register (SPI_I2SPR) . . . . . . . . . . . . . . . . . . . . . . . 924
28.5.10 SPI register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 925
26/1751 RM0090 Rev 19

RM0090 Contents
29 Serial audio interface (SAI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 926

29.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 926
29.2 Main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 927
29.3 Functional block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 928
29.4 Main SAI modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 929
29.5 SAI synchronization mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 930
29.6 Audio data size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 930
29.7 Frame synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 930
29.7.1 Frame length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 931
29.7.2 Frame synchronization polarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 931
29.7.3 Frame synchronization active level length . . . . . . . . . . . . . . . . . . . . . . 932
29.7.4 Frame synchronization offset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 932
29.7.5 FS signal role . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 932
29.8 Slot configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 933
29.9 SAI clock generator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 935
29.10 Internal FIFOs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 936
29.11 AC’97 link controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 939
29.12 Specific features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 940
29.12.1 Mute mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 940
29.12.2 MONO/STEREO function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 940
29.12.3 Companding mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 941
29.12.4 Output data line management on an inactive slot . . . . . . . . . . . . . . . . 942
29.13 Error flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 944
29.13.1 FIFO overrun/underrun (OVRUDR) . . . . . . . . . . . . . . . . . . . . . . . . . . . 944
29.13.2 Anticipated frame synchronisation detection (AFSDET) . . . . . . . . . . . 946
29.13.3 Late frame synchronization detection . . . . . . . . . . . . . . . . . . . . . . . . . 946
29.13.4 Codec not ready (CNRDY AC’97) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 947
29.13.5 Wrong clock configuration in master mode (with NODIV = 0) . . . . . . . 947
29.14 Interrupt sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 948
29.15 Disabling the SAI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 948
29.16 SAI DMA interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 949
29.17 SAI registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 950
29.17.1 SAI xConfiguration register 1 (SAI_xCR1) where x is A or B . . . . . . . . 950
29.17.2 SAI xConfiguration register 2 (SAI_xCR2) where x is A or B . . . . . . . . 953
29.17.3 SAI xFrame configuration register (SAI_XFRCR) where x is A or B . . 955
RM0090 Rev 19 27/1751

39
Contents RM0090
29.17.4 SAI xSlot register (SAI_xSLOTR) where x is A or B . . . . . . . . . . . . . . 957

29.17.5 SAI xInterrupt mask register (SAI_xIM) where x is A or B . . . . . . . . . . 958
29.17.6 SAI xStatus register (SAI_xSR) where x is A or B . . . . . . . . . . . . . . . . 960
29.17.7 SAI xClear flag register (SAI_xCLRFR) where X is A or B . . . . . . . . . 962
29.17.8 SAI xData register (SAI_xDR) where x is A or B . . . . . . . . . . . . . . . . . 963
29.17.9 SAI register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 963
30 Universal synchronous asynchronous receiver

transmitter (USART) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 965
30.1 USART introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 965
30.2 USART main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 965
30.3 USART functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 966
30.3.1 USART character description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 969
30.3.2 Transmitter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 970
30.3.3 Receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 973
30.3.4 Fractional baud rate generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 978
30.3.5 USART receiver tolerance to clock deviation . . . . . . . . . . . . . . . . . . . . 988
30.3.6 Multiprocessor communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 989
30.3.7 Parity control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 991
30.3.8 LIN (local interconnection network) mode . . . . . . . . . . . . . . . . . . . . . . 992
30.3.9 USART synchronous mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 994
30.3.10 Single-wire half-duplex communication . . . . . . . . . . . . . . . . . . . . . . . . 996
30.3.11 Smartcard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 997
30.3.12 IrDA SIR ENDEC block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 999
30.3.13 Continuous communication using DMA . . . . . . . . . . . . . . . . . . . . . . . 1001
30.3.14 Hardware flow control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1003
30.4 USART interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1006
30.5 USART mode configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1007
30.6 USART registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1007
30.6.1 Status register (USART_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1007
30.6.2 Data register (USART_DR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1010
30.6.3 Baud rate register (USART_BRR) . . . . . . . . . . . . . . . . . . . . . . . . . . . 1010
30.6.4 Control register 1 (USART_CR1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1010
30.6.7 Guard time and prescaler register (USART_GTPR) . . . . . . . . . . . . . 1017
30.6.8 USART register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1018
28/1751 RM0090 Rev 19

RM0090 Contents
31 Secure digital input/output interface (SDIO) . . . . . . . . . . . . . . . . . . . 1019

31.1 SDIO main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1019
31.2 SDIO bus topology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1020
31.3 SDIO functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1022
31.3.1 SDIO adapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1023
31.3.2 SDIO APB2 interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1033
31.4 Card functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1034
31.4.1 Card identification mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1034
31.4.2 Card reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1034
31.4.3 Operating voltage range validation . . . . . . . . . . . . . . . . . . . . . . . . . . 1034
31.4.4 Card identification process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1035
31.4.5 Block write . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1036
31.4.6 Block read . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1037
31.4.7 Stream access, stream write and stream read
(MultiMediaCard only) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1037
31.4.8 Erase: group erase and sector erase . . . . . . . . . . . . . . . . . . . . . . . . 1039
31.4.9 Wide bus selection or deselection . . . . . . . . . . . . . . . . . . . . . . . . . . . 1039
31.4.10 Protection management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1039
31.4.11 Card status register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1042
31.4.12 SD status register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1045
31.4.13 SD I/O mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1049
31.4.14 Commands and responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1050
31.5 Response formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1054
31.5.1 R1 (normal response command) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1054
31.5.2 R1b . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1054
31.5.3 R2 (CID, CSD register) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1054
31.5.4 R3 (OCR register) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1055
31.5.5 R4 (Fast I/O) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1055
31.5.6 R4b . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1056
31.5.7 R5 (interrupt request) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1056
31.5.8 R6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1057
31.6 SDIO I/O card-specific operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1057
31.6.1 SDIO I/O read wait operation by SDIO_D2 signaling . . . . . . . . . . . . 1058
31.6.2 SDIO read wait operation by stopping SDIO_CK . . . . . . . . . . . . . . . 1058
31.6.3 SDIO suspend/resume operation . . . . . . . . . . . . . . . . . . . . . . . . . . . 1058
31.6.4 SDIO interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1058
RM0090 Rev 19 29/1751

39
Contents RM0090
31.7 CE-ATA specific operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1059

31.7.1 Command completion signal disable . . . . . . . . . . . . . . . . . . . . . . . . . 1059
31.7.2 Command completion signal enable . . . . . . . . . . . . . . . . . . . . . . . . . 1059
31.7.3 CE-ATA interrupt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1059
31.7.4 Aborting CMD61 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1059
31.8 HW flow control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1060
31.9 SDIO registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1060
31.9.1 SDIO power control register (SDIO_POWER) . . . . . . . . . . . . . . . . . . 1060
31.9.2 SDI clock control register (SDIO_CLKCR) . . . . . . . . . . . . . . . . . . . . 1061
31.9.3 SDIO argument register (SDIO_ARG) . . . . . . . . . . . . . . . . . . . . . . . . 1062
31.9.4 SDIO command register (SDIO_CMD) . . . . . . . . . . . . . . . . . . . . . . . 1062
31.9.5 SDIO command response register (SDIO_RESPCMD) . . . . . . . . . . 1063
31.9.6 SDIO response 1..4 register (SDIO_RESPx) . . . . . . . . . . . . . . . . . . 1064
31.9.7 SDIO data timer register (SDIO_DTIMER) . . . . . . . . . . . . . . . . . . . . 1064
31.9.8 SDIO data length register (SDIO_DLEN) . . . . . . . . . . . . . . . . . . . . . 1065
31.9.9 SDIO data control register (SDIO_DCTRL) . . . . . . . . . . . . . . . . . . . . 1066
31.9.10 SDIO data counter register (SDIO_DCOUNT) . . . . . . . . . . . . . . . . . 1067
31.9.11 SDIO status register (SDIO_STA) . . . . . . . . . . . . . . . . . . . . . . . . . . . 1068
31.9.12 SDIO interrupt clear register (SDIO_ICR) . . . . . . . . . . . . . . . . . . . . . 1069
31.9.13 SDIO mask register (SDIO_MASK) . . . . . . . . . . . . . . . . . . . . . . . . . . 1071
31.9.14 SDIO FIFO counter register (SDIO_FIFOCNT) . . . . . . . . . . . . . . . . . 1073
31.9.15 SDIO data FIFO register (SDIO_FIFO) . . . . . . . . . . . . . . . . . . . . . . . 1074
31.9.16 SDIO register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1074
32 Controller area network (bxCAN) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1076

32.1 bxCAN introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1076
32.2 bxCAN main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1076
32.3 bxCAN general description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1077
32.3.1 CAN 2.0B active core . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1077
32.3.2 Control, status and configuration registers . . . . . . . . . . . . . . . . . . . . 1078
32.3.3 Tx mailboxes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1078
32.3.4 Acceptance filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1078
32.4 bxCAN operating modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1079
32.4.1 Initialization mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1080
32.4.2 Normal mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1080
32.4.3 Sleep mode (low-power) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1080
30/1751 RM0090 Rev 19

RM0090 Contents
32.5 Test mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1081

32.5.1 Silent mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1081
32.5.2 Loop back mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1082
32.5.3 Loop back combined with silent mode . . . . . . . . . . . . . . . . . . . . . . . . 1082
32.6 Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1083
32.7 bxCAN functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1083
32.7.1 Transmission handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1083
32.7.2 Time triggered communication mode . . . . . . . . . . . . . . . . . . . . . . . . . 1085
32.7.3 Reception handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1085
32.7.4 Identifier filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1087
32.7.5 Message storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1091
32.7.6 Error management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1093
32.7.7 Bit timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1093
32.8 bxCAN interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1095
32.9 CAN registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1097
32.9.1 Register access protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1097
32.9.2 CAN control and status registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1097
32.9.3 CAN mailbox registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1107
32.9.4 CAN filter registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1114
32.9.5 bxCAN register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1118
33 Ethernet (ETH): media access control (MAC) with

DMA controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1122
33.1 Ethernet introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1122
33.2 Ethernet main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1122
33.2.1 MAC core features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1123
33.2.2 DMA features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1124
33.2.3 PTP features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1124
33.3 Ethernet pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1125
33.4 Ethernet functional description: SMI, MII and RMII . . . . . . . . . . . . . . . .1126
33.4.1 Station management interface: SMI . . . . . . . . . . . . . . . . . . . . . . . . . 1126
33.4.2 Media-independent interface: MII . . . . . . . . . . . . . . . . . . . . . . . . . . . 1129
33.4.3 Reduced media-independent interface: RMII . . . . . . . . . . . . . . . . . . 1131
33.4.4 MII/RMII selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1132
33.5 Ethernet functional description: MAC 802.3 . . . . . . . . . . . . . . . . . . . . . .1133
33.5.1 MAC 802.3 frame format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1134
RM0090 Rev 19 31/1751

39
Contents RM0090
33.5.2 MAC frame transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1137

33.5.3 MAC frame reception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1144
33.5.4 MAC interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1149
33.5.5 MAC filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1150
33.5.6 MAC loopback mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1153
33.5.7 MAC management counters: MMC . . . . . . . . . . . . . . . . . . . . . . . . . . 1153
33.5.8 Power management: PMT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1154
33.5.9 Precision time protocol (IEEE1588 PTP) . . . . . . . . . . . . . . . . . . . . . . 1157
33.6 Ethernet functional description: DMA controller operation . . . . . . . . . . .1163
33.6.1 Initialization of a transfer using DMA . . . . . . . . . . . . . . . . . . . . . . . . . 1164
33.6.2 Host bus burst access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1164
33.6.3 Host data buffer alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1165
33.6.4 Buffer size calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1165
33.6.5 DMA arbiter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1166
33.6.6 Error response to DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1166
33.6.7 Tx DMA configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1166
33.6.8 Rx DMA configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1178
33.6.9 DMA interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1189
33.7 Ethernet interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1190
33.8 Ethernet register descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1191
33.8.1 MAC register description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1191
33.8.2 MMC register description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1210
33.8.3 IEEE 1588 time stamp registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1215
33.8.4 DMA register description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1223
33.8.5 Ethernet register maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1236
34 USB on-the-go full-speed (OTG_FS) . . . . . . . . . . . . . . . . . . . . . . . . . 1240

34.1 OTG_FS introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1240
34.2 OTG_FS main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1241
34.2.1 General features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1241
34.2.2 Host-mode features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1242
34.2.3 Peripheral-mode features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1242
34.3 OTG_FS functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1243
34.3.1 OTG pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1243
34.3.2 OTG full-speed core . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1243
34.3.3 Full-speed OTG PHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1244
32/1751 RM0090 Rev 19

RM0090 Contents
34.4 OTG dual role device (DRD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1245

34.4.1 ID line detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1245
34.4.2 HNP dual role device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1245
34.4.3 SRP dual role device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1246
34.5 USB peripheral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1246
34.5.1 SRP-capable peripheral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1247
34.5.2 Peripheral states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1247
34.5.3 Peripheral endpoints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1248
34.6 USB host . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1250
34.6.1 SRP-capable host . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1251
34.6.2 USB host states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1251
34.6.3 Host channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1253
34.6.4 Host scheduler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1254
34.7 SOF trigger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1255
34.7.1 Host SOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1255
34.7.2 Peripheral SOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1256
34.8 OTG low-power modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1256
34.9 Dynamic update of the OTG_FS_HFIR register . . . . . . . . . . . . . . . . . . 1257
34.10 USB data FIFOs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1258
34.11 Peripheral FIFO architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1259
34.11.1 Peripheral Rx FIFO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1259
34.11.2 Peripheral Tx FIFOs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1260
34.12 Host FIFO architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1260
34.12.1 Host Rx FIFO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1260
34.12.2 Host Tx FIFOs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1261
34.13 FIFO RAM allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1261
34.13.1 Device mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1261
34.13.2 Host mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1262
34.14 USB system performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1262
34.15 OTG_FS interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1263
34.16 OTG_FS control and status registers . . . . . . . . . . . . . . . . . . . . . . . . . . 1265
34.16.1 CSR memory map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1266
34.16.2 OTG_FS global registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1271
34.16.3 Host-mode registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1292
34.16.4 Device-mode registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1302
RM0090 Rev 19 33/1751

39
Contents RM0090
34.16.5 OTG_FS power and clock gating control register

(OTG_FS_PCGCCTL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1325
34.16.6 OTG_FS register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1326
34.17 OTG_FS programming model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1335
34.17.1 Core initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1335
34.17.2 Host initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1336
34.17.3 Device initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1336
34.17.4 Host programming model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1337
34.17.5 Device programming model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1353
34.17.6 Operational model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1355
34.17.7 Worst case response time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1373
34.17.8 OTG programming model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1374
35 USB on-the-go high-speed (OTG_HS) . . . . . . . . . . . . . . . . . . . . . . . . 1381

35.1 OTG_HS introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1381
35.2 OTG_HS main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1381
35.2.1 General features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1382
35.2.2 Host-mode features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1383
35.2.3 Peripheral-mode features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1383
35.3 OTG_HS functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1383
35.3.1 OTG pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1384
35.3.2 High-speed OTG PHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1384
35.3.3 Embedded Full-speed OTG PHY . . . . . . . . . . . . . . . . . . . . . . . . . . . 1385
35.4 OTG dual-role device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1385
35.4.1 ID line detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1385
35.4.2 HNP dual role device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1385
35.4.3 SRP dual-role device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1386
35.5 USB functional description in peripheral mode . . . . . . . . . . . . . . . . . . 1386
35.5.1 SRP-capable peripheral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1386
35.5.2 Peripheral states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1387
35.5.3 Peripheral endpoints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1388
35.6 USB functional description on host mode . . . . . . . . . . . . . . . . . . . . . . 1391
35.6.1 SRP-capable host . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1391
35.6.2 USB host states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1392
35.6.3 Host channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1393
35.6.4 Host scheduler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1395
34/1751 RM0090 Rev 19

RM0090 Contents
35.7 SOF trigger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1396

35.7.1 Host SOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1396
35.7.2 Peripheral SOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1396
35.8 OTG_HS low-power modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1397
35.9 Dynamic update of the OTG_HS_HFIR register . . . . . . . . . . . . . . . . . 1398
35.10 FIFO RAM allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1399
35.10.1 Peripheral mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1399
35.10.2 Host mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1399
35.11 OTG_HS interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1400
35.12 OTG_HS control and status registers . . . . . . . . . . . . . . . . . . . . . . . . . 1402
35.12.1 CSR memory map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1402
35.12.2 OTG_HS global registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1407
35.12.3 Host-mode registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1430
35.12.4 Device-mode registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1443
35.12.5 OTG_HS power and clock gating control register
(OTG_HS_PCGCCTL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1472
35.12.6 OTG_HS register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1472
35.13 OTG_HS programming model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1487
35.13.1 Core initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1487
35.13.2 Host initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1488
35.13.3 Device initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1489
35.13.4 DMA mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1489
35.13.5 Host programming model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1489
35.13.6 Device programming model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1515
35.13.7 Operational model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1517
35.13.8 Worst case response time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1536
35.13.9 OTG programming model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1538
36 Flexible static memory controller (FSMC) . . . . . . . . . . . . . . . . . . . . 1544

36.1 FSMC main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1544
36.2 Block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1545
36.3 AHB interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1545
36.3.1 Supported memories and transactions . . . . . . . . . . . . . . . . . . . . . . . 1546
36.4 External device address mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1547
36.4.1 NOR/PSRAM address mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1547
36.4.2 NAND/PC Card address mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . 1548
RM0090 Rev 19 35/1751

39
Contents RM0090
36.5 NOR Flash/PSRAM controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1549

36.5.1 External memory interface signals . . . . . . . . . . . . . . . . . . . . . . . . . . . 1550
36.5.3 General timing rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1553
36.5.4 NOR Flash/PSRAM controller asynchronous transactions . . . . . . . . 1554
36.5.5 Synchronous transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1571
36.5.6 NOR/PSRAM control registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1577
36.6 NAND Flash/PC Card controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1584
36.6.2 NAND Flash / PC Card supported memories and transactions . . . . . 1587
36.6.3 Timing diagrams for NAND and PC Card . . . . . . . . . . . . . . . . . . . . . 1587
36.6.4 NAND Flash operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1588
36.6.5 NAND Flash prewait functionality . . . . . . . . . . . . . . . . . . . . . . . . . . . 1589
36.6.6 Computation of the error correction code (ECC)
in NAND Flash memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1590
36.6.7 PC Card/CompactFlash operations . . . . . . . . . . . . . . . . . . . . . . . . . . 1591
36.6.8 NAND Flash/PC Card control registers . . . . . . . . . . . . . . . . . . . . . . . 1593
36.6.9 FSMC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1600
37 Flexible memory controller (FMC) . . . . . . . . . . . . . . . . . . . . . . . . . . . 1602

37.1 FMC main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1602
37.2 Block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1603
37.3 AHB interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1604
37.4 External device address mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1606
37.4.1 NOR/PSRAM address mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1607
37.4.2 NAND Flash memory/PC Card address mapping . . . . . . . . . . . . . . . 1608
37.4.3 SDRAM address mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1609
37.5 NOR Flash/PSRAM controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1612
37.5.3 General timing rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1617
37.5.4 NOR Flash/PSRAM controller asynchronous transactions . . . . . . . . 1617
37.5.5 Synchronous transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1634
37.5.6 NOR/PSRAM controller registers . . . . . . . . . . . . . . . . . . . . . . . . . . . 1640
37.6 NAND Flash/PC Card controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1647
36/1751 RM0090 Rev 19

RM0090 Contents
37.6.2 NAND Flash / PC Card supported memories and transactions . . . . . 1650

37.6.3 Timing diagrams for NAND Flash memory and PC Card . . . . . . . . . 1650
37.6.4 NAND Flash operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1651
37.6.5 NAND Flash prewait functionality . . . . . . . . . . . . . . . . . . . . . . . . . . . 1652
37.6.6 Computation of the error correction code (ECC)
in NAND Flash memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1653
37.6.7 PC Card/CompactFlash operations . . . . . . . . . . . . . . . . . . . . . . . . . . 1654
37.6.8 NAND Flash/PC Card controller registers . . . . . . . . . . . . . . . . . . . . . 1656
37.7 SDRAM controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1663
37.7.1 SDRAM controller main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1663
37.7.2 SDRAM External memory interface signals . . . . . . . . . . . . . . . . . . . . 1663
37.7.3 SDRAM controller functional description . . . . . . . . . . . . . . . . . . . . . . 1664
37.7.4 Low power modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1671
37.7.5 SDRAM controller registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1674
37.8 FMC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1680
38 Debug support (DBG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1683

38.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1683
38.2 Reference Arm® documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1684
38.3 SWJ debug port (serial wire and JTAG) . . . . . . . . . . . . . . . . . . . . . . . . 1684
38.3.1 Mechanism to select the JTAG-DP or the SW-DP . . . . . . . . . . . . . . . 1685
38.4 Pinout and debug port pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1685
38.4.1 SWJ debug port pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1686
38.4.2 Flexible SWJ-DP pin assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1686
38.4.3 Internal pull-up and pull-down on JTAG pins . . . . . . . . . . . . . . . . . . . 1687
38.4.4 Using serial wire and releasing the unused debug pins as GPIOs . . 1688
38.5 STM32F4xx JTAG TAP connection . . . . . . . . . . . . . . . . . . . . . . . . . . . 1688
38.6 ID codes and locking mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1690
38.6.1 MCU device ID code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1690
38.6.2 Boundary scan TAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1691
38.6.3 Cortex®-M4 with FPU TAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1691
38.6.4 Cortex®-M4 with FPU JEDEC-106 ID code . . . . . . . . . . . . . . . . . . . . 1691
38.7 JTAG debug port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1691
38.8 SW debug port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1693
38.8.1 SW protocol introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1693
38.8.2 SW protocol sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1693
RM0090 Rev 19 37/1751

39
Contents RM0090
38.8.3 SW-DP state machine (reset, idle states, ID code) . . . . . . . . . . . . . . 1694

38.8.4 DP and AP read/write accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1694
38.8.5 SW-DP registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1695
38.8.6 SW-AP registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1696
38.9 AHB-AP (AHB access port) - valid for both JTAG-DP
and SW-DP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1696
38.10 Core debug . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1697
38.11 Capability of the debugger host to connect under system reset . . . . . 1698
38.12 FPB (Flash patch breakpoint) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1698
38.13 DWT (data watchpoint trigger) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1699
38.14 ITM (instrumentation trace macrocell) . . . . . . . . . . . . . . . . . . . . . . . . . 1699
38.14.1 General description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1699
38.14.2 Time stamp packets, synchronization and overflow packets . . . . . . . 1699
38.15 ETM (Embedded Trace Macrocell™) . . . . . . . . . . . . . . . . . . . . . . . . . . 1701
38.15.1 ETM general description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1701
38.15.2 ETM signal protocol and packet types . . . . . . . . . . . . . . . . . . . . . . . . 1701
38.15.3 Main ETM registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1702
38.15.4 ETM configuration example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1702
38.16 MCU debug component (DBGMCU) . . . . . . . . . . . . . . . . . . . . . . . . . . 1702
38.16.1 Debug support for low-power modes . . . . . . . . . . . . . . . . . . . . . . . . . 1702
38.16.2 Debug support for timers, watchdog, bxCAN and I2C . . . . . . . . . . . . 1703
38.16.3 Debug MCU configuration register . . . . . . . . . . . . . . . . . . . . . . . . . . 1703
38.16.4 Debug MCU APB1 freeze register (DBGMCU_APB1_FZ) . . . . . . . . 1705
38.16.5 Debug MCU APB2 Freeze register (DBGMCU_APB2_FZ) . . . . . . . . 1706
38.17 TPIU (trace port interface unit) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1707
38.17.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1707
38.17.2 TRACE pin assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1708
38.17.3 TPUI formatter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1709
38.17.4 TPUI frame synchronization packets . . . . . . . . . . . . . . . . . . . . . . . . . 1710
38.17.5 Transmission of the synchronization frame packet . . . . . . . . . . . . . . 1710
38.17.6 Synchronous mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1710
38.17.7 Asynchronous mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1711
38.17.8 TRACECLKIN connection inside the STM32F4xx . . . . . . . . . . . . . . . 1711
38.17.9 TPIU registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1711
38.17.10 Example of configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1712
38.18 DBG register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1713
38/1751 RM0090 Rev 19

RM0090 Contents
39 Device electronic signature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1714

39.1 Unique device ID register (96 bits) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1714
39.2 Flash size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1715
40 Revision history . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1716
RM0090 Rev 19 39/1751

39
List of tables RM0090
List of tables
Table 1. STM32F4xx register boundary addresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

Table 2. Boot modes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Table 3. Memory mapping vs. Boot mode/physical remap
in STM32F405xx/07xx and STM32F415xx/17xx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
in STM32F42xxx and STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Table 5. Flash module organization (STM32F40x and STM32F41x) . . . . . . . . . . . . . . . . . . . . . . . . 75
Table 6. Flash module - 2 Mbyte dual bank organization (STM32F42xxx and STM32F43xxx) . . . . 77
Table 7. 1 Mbyte Flash memory single bank vs dual bank organization
(STM32F42xxx and STM32F43xxx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Table 8. 1 Mbyte single bank Flash memory organization
(STM32F42xxx and STM32F43xxx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Table 9. 1 Mbyte dual bank Flash memory organization (STM32F42xxx and STM32F43xxx) . . . . 79
Table 10. Number of wait states according to CPU clock (HCLK) frequency
(STM32F405xx/07xx and STM32F415xx/17xx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
(STM32F42xxx and STM32F43xxx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Table 12. Program/erase parallelism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Table 13. Flash interrupt request . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Table 14. Option byte organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Table 15. Description of the option bytes (STM32F405xx/07xx and STM32F415xx/17xx) . . . . . . . . 89
Table 16. Description of the option bytes
Table 17. Access versus read protection level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Table 18. OTP area organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Table 19. Flash register map and reset values
(STM32F405xx/07xx and STM32F415xx/17xx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Table 20. Flash register map and reset values (STM32F42xxx and STM32F43xxx) . . . . . . . . . . . . 111
Table 21. CRC calculation unit register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Table 22. Voltage regulator configuration mode versus device operating mode . . . . . . . . . . . . . . . 122
Table 23. Low-power mode summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Table 24. Sleep-now entry and exit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Table 25. Sleep-on-exit entry and exit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
Table 26. Stop operating modes
Table 27. Stop mode entry and exit (for STM32F405xx/07xx and STM32F415xx/17xx) . . . . . . . . . 132
Table 28. Stop operating modes (STM32F42xxx and STM32F43xxx) . . . . . . . . . . . . . . . . . . . . . . . 134
Table 29. Stop mode entry and exit (STM32F42xxx and STM32F43xxx) . . . . . . . . . . . . . . . . . . . . 136
Table 30. Standby mode entry and exit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
Table 31. PWR - register map and reset values for STM32F405xx/07xx and STM32F415xx/17xx. 149
Table 32. PWR - register map and reset values for STM32F42xxx and STM32F43xxx . . . . . . . . . 149
Table 33. RCC register map and reset values for STM32F42xxx and STM32F43xxx . . . . . . . . . . . 210
Table 34. RCC register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
Table 35. Port bit configuration table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
Table 36. Flexible SWJ-DP pin assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
Table 37. RTC_AF1 pin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
Table 38. RTC_AF2 pin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280
Table 39. GPIO register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
40/1751 RM0090 Rev 19

RM0090 List of tables
Table 40. SYSCFG register map and reset values (STM32F405xx/07xx and STM32F415xx/17xx) 294
Table 41. SYSCFG register map and reset values (STM32F42xxx and STM32F43xxx) . . . . . . . . . 301
Table 42. DMA1 request mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
Table 43. DMA2 request mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
Table 44. Source and destination address . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Table 45. Source and destination address registers in Double buffer mode (DBM=1). . . . . . . . . . . 314
Table 46. Packing/unpacking & endian behavior (bit PINC = MINC = 1) . . . . . . . . . . . . . . . . . . . . . 315
Table 47. Restriction on NDT versus PSIZE and MSIZE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
Table 48. FIFO threshold configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 318
Table 49. Possible DMA configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
Table 50. DMA interrupt requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
Table 51. DMA register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
Table 52. Supported color mode in input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342
Table 53. Data order in memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
Table 54. Alpha mode configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344
Table 55. Supported CLUT color mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
Table 56. CLUT data order in memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
Table 57. Supported color mode in output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
Table 58. Data order in memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
Table 59. DMA2D interrupt requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
Table 60. DMA2D register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
Table 61. Vector table for STM32F405xx/07xx and STM32F415xx/17xx. . . . . . . . . . . . . . . . . . . . . 372
Table 62. Vector table for STM32F42xxx and STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
Table 63. External interrupt/event controller register map and reset values. . . . . . . . . . . . . . . . . . . 387
Table 64. External interrupt/event controller register map and reset values. . . . . . . . . . . . . . . . . . . 387
Table 65. ADC pins. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390
Table 66. Analog watchdog channel selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393
Table 67. Configuring the trigger polarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
Table 68. External trigger for regular channels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
Table 69. External trigger for injected channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
Table 70. ADC interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
Table 71. ADC global register map. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
Table 72. ADC register map and reset values for each ADC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
Table 73. ADC register map and reset values (common ADC registers) . . . . . . . . . . . . . . . . . . . . . 432
Table 74. DAC pins. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434
Table 75. External triggers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
Table 76. DAC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
Table 77. DCMI pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
Table 78. DCMI signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
Table 79. Positioning of captured data bytes in 32-bit words (8-bit width) . . . . . . . . . . . . . . . . . . . . 458
Table 80. Positioning of captured data bytes in 32-bit words (10-bit width) . . . . . . . . . . . . . . . . . . . 458
Table 83. Data storage in monochrome progressive video format . . . . . . . . . . . . . . . . . . . . . . . . . . 466
Table 84. Data storage in RGB progressive video format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
Table 85. Data storage in YCbCr progressive video format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467
Table 86. DCMI interrupts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467
Table 87. DCMI register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478
Table 88. LTDC registers versus clock domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 482
Table 89. LCD-TFT pins and signal interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
Table 90. Pixel Data mapping versus Color Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487
Table 91. LTDC interrupt requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491
RM0090 Rev 19 41/1751

46
Table 92. LTDC register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 512

Table 93. Counting direction versus encoder signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 553
Table 94. TIMx Internal trigger connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567
Table 95. Output control bits for complementary OCx and OCxN channels with
break feature. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579
Table 96. TIM1 and TIM8 register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
Table 97. Counting direction versus encoder signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 616
Table 98. TIMx internal trigger connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 632
Table 99. Output control bit for standard OCx channels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 641
Table 100. TIM2 to TIM5 register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 648
Table 101. TIMx internal trigger connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 674
Table 103. TIM9/12 register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684
Table 105. TIM10/11/13/14 register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 694
Table 106. TIM6 and TIM7 register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707
Table 107. Min/max IWDG timeout period (in ms) at 32 kHz (LSI). . . . . . . . . . . . . . . . . . . . . . . . . . . 709
Table 108. IWDG register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 712
Table 109. Minimum and maximum timeout values at 30 MHz (fPCLK1). . . . . . . . . . . . . . . . . . . . . . . 716
Table 110. WWDG register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 719
Table 111. Number of cycles required to process each 128-bit block
(STM32F415/417xx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 720
(STM32F43xxx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 720
Table 113. Data types. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 740
Table 114. CRYP register map and reset values for STM32F415/417xx . . . . . . . . . . . . . . . . . . . . . . 763
Table 115. CRYP register map and reset values for STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . 764
Table 116. RNG register map and reset map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771
Table 117. HASH register map and reset values on STM32F415/417xx . . . . . . . . . . . . . . . . . . . . . . 795
Table 118. HASH register map and reset values on STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . 796
Table 119. Effect of low-power modes on RTC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 814
Table 120. Interrupt control bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 815
Table 121. RTC register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 836
Table 122. Maximum DNF[3:0] value to be compliant with Thd:dat(max) . . . . . . . . . . . . . . . . . . . . . 852
Table 123. SMBus vs. I2C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 854
Table 124. I2C Interrupt requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 858
Table 125. I2C register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 872
Table 126. SPI interrupt requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 898
Table 127. Audio frequency precision (for PLLM VCO = 1 MHz or 2 MHz) . . . . . . . . . . . . . . . . . . . . 909
Table 128. I2S interrupt requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 915
Table 129. SPI register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 925
Table 130. Example of possible audio frequency sampling range . . . . . . . . . . . . . . . . . . . . . . . . . . . 936
Table 131. Interrupt sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 948
Table 132. SAI register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 963
Table 133. Noise detection from sampled data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 977
Table 134. Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK = 12 MHz,
oversampling by 16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 980
Table 135. Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK =12 MHz,
oversampling by 8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 981
oversampling by 16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 981
42/1751 RM0090 Rev 19

oversampling by 8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 982
oversampling by 16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 983
oversampling by 8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 983
oversampling by 16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 984
oversampling by 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 985
Table 142. Error calculation for programmed baud rates at fPCLK = 42 MHz or fPCLK = 84 Hz,
oversampling by 16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 986
oversampling by 8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 987
Table 144. USART receiver’s tolerance when DIV fraction is 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 988
Table 145. USART receiver tolerance when DIV_Fraction is different from 0 . . . . . . . . . . . . . . . . . . 989
Table 146. Frame formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 991
Table 147. USART interrupt requests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1006
Table 148. USART mode configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1007
Table 149. USART register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1018
Table 150. SDIO I/O definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1023
Table 151. Command format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1027
Table 152. Short response format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1028
Table 153. Long response format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1028
Table 154. Command path status flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1028
Table 155. Data token format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1031
Table 156. Transmit FIFO status flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1032
Table 157. Receive FIFO status flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1033
Table 158. Card status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1043
Table 159. SD status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1046
Table 160. Speed class code field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1047
Table 161. Performance move field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1047
Table 162. AU_SIZE field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1048
Table 163. Maximum AU size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1048
Table 164. Erase size field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1048
Table 165. Erase timeout field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1049
Table 166. Erase offset field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1049
Table 167. Block-oriented write commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1051
Table 168. Block-oriented write protection commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1052
Table 169. Erase commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1052
Table 170. I/O mode commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1053
Table 171. Lock card . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1053
Table 172. Application-specific commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1053
Table 173. R1 response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1054
Table 174. R2 response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1055
Table 175. R3 response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1055
Table 176. R4 response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1055
Table 177. R4b response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1056
Table 178. R5 response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1056
Table 179. R6 response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1057
Table 180. Response type and SDIO_RESPx registers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1064
Table 181. SDIO register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1074
Table 182. Transmit mailbox mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1091
RM0090 Rev 19 43/1751

46
Table 183. Receive mailbox mapping. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1091

Table 184. bxCAN register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1118
Table 185. Alternate function mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1125
Table 186. Management frame format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1127
Table 187. Clock range. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1129
Table 188. TX interface signal encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1130
Table 189. RX interface signal encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1131
Table 190. Frame statuses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1146
Table 191. Destination address filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1152
Table 192. Source address filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1153
Table 193. Receive descriptor 0 - encoding for bits 7, 5 and 0 (normal descriptor
format only, EDFE=0) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1184
Table 194. Time stamp snapshot dependency on registers bits . . . . . . . . . . . . . . . . . . . . . . . . . . . 1217
Table 195. Ethernet register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1236
Table 196. OTG_FS input/output pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1243
Table 197. Compatibility of STM32 low power modes with the OTG . . . . . . . . . . . . . . . . . . . . . . . . 1256
Table 198. Core global control and status registers (CSRs). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1266
Table 199. Host-mode control and status registers (CSRs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1267
Table 200. Device-mode control and status registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1268
Table 201. Data FIFO (DFIFO) access register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1269
Table 202. Power and clock gating control and status registers . . . . . . . . . . . . . . . . . . . . . . . . . . . 1270
Table 203. TRDT values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1276
Table 204. Minimum duration for soft disconnect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1304
Table 205. OTG_FS register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1326
Table 206. OTG_HS input/output pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1384
Table 207. Compatibility of STM32 low power modes with the OTG . . . . . . . . . . . . . . . . . . . . . . . . 1397
Table 208. Core global control and status registers (CSRs). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1403
Table 209. Host-mode control and status registers (CSRs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1404
Table 210. Device-mode control and status registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1405
Table 211. Data FIFO (DFIFO) access register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1407
Table 212. Power and clock gating control and status registers . . . . . . . . . . . . . . . . . . . . . . . . . . . 1407
Table 213. TRDT values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1414
Table 214. Minimum duration for soft disconnect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1446
Table 215. OTG_HS register map and reset values. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1472
Table 216. NOR/PSRAM bank selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1548
Table 217. External memory address . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1548
Table 218. Memory mapping and timing registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1548
Table 219. NAND bank selections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1549
Table 220. Programmable NOR/PSRAM access parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1550
Table 221. Nonmultiplexed I/O NOR Flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1550
Table 222. Multiplexed I/O NOR Flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1551
Table 223. Nonmultiplexed I/Os PSRAM/SRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1551
Table 224. Multiplexed I/O PSRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1552
Table 225. NOR Flash/PSRAM controller: example of supported memories and transactions . . . . 1552
Table 226. FSMC_BCRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1555
Table 227. FSMC_BTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1556
Table 230. FSMC_BWTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1558
44/1751 RM0090 Rev 19


Table 246. Programmable NAND/PC Card access parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1585
Table 247. 8-bit NAND Flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1585
Table 248. 16-bit NAND Flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1586
Table 249. 16-bit PC Card . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1586
Table 250. Supported memories and transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1587
Table 251. 16-bit PC-Card signals and access type. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1592
Table 252. ECC result relevant bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1599
Table 253. FSMC register map. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1600
Table 254. NOR/PSRAM bank selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1607
Table 255. NOR/PSRAM External memory address . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1608
Table 256. NAND/PC Card memory mapping and timing registers . . . . . . . . . . . . . . . . . . . . . . . . . 1608
Table 257. NAND bank selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1609
Table 258. SDRAM bank selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1609
Table 259. SDRAM address mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1609
Table 260. SDRAM address mapping with 8-bit data bus width. . . . . . . . . . . . . . . . . . . . . . . . . . . . 1610
Table 261. SDRAM address mapping with 16-bit data bus width. . . . . . . . . . . . . . . . . . . . . . . . . . . 1611
Table 262. SDRAM address mapping with 32-bit data bus width. . . . . . . . . . . . . . . . . . . . . . . . . . . 1611
Table 263. Programmable NOR/PSRAM access parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1613
Table 264. Non-multiplexed I/O NOR Flash memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1614
Table 265. 16-bit multiplexed I/O NOR Flash memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1614
Table 266. Non-multiplexed I/Os PSRAM/SRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1614
Table 267. 16-Bit multiplexed I/O PSRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1615
Table 268. NOR Flash/PSRAM: Example of supported memories and transactions . . . . . . . . . . . . 1616
Table 269. FMC_BCRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1619
Table 270. FMC_BTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1619
Table 273. FMC_BWTRx bit fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1622
RM0090 Rev 19 45/1751

46

Table 289. Programmable NAND Flash/PC Card access parameters . . . . . . . . . . . . . . . . . . . . . . . 1648
Table 290. 8-bit NAND Flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1648
Table 291. 16-bit NAND Flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1649
Table 292. 16-bit PC Card . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1649
Table 293. Supported memories and transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1650
Table 294. 16-bit PC-Card signals and access type. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1655
Table 295. ECC result relevant bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1662
Table 296. SDRAM signals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1663
Table 297. FMC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1680
Table 298. SWJ debug port pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1686
Table 299. Flexible SWJ-DP pin assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1686
Table 300. JTAG debug port data registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1691
Table 301. 32-bit debug port registers addressed through the shifted value A[3:2] . . . . . . . . . . . . 1692
Table 302. Packet request (8-bits) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1693
Table 303. ACK response (3 bits). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1694
Table 304. DATA transfer (33 bits) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1694
Table 305. SW-DP registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1695
Table 306. Cortex®-M4 with FPU AHB-AP registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1696
Table 307. Core debug registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1697
Table 308. Main ITM registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1700
Table 309. Main ETM registers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1702
Table 310. Asynchronous TRACE pin assignment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1708
Table 311. Synchronous TRACE pin assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1708
Table 312. Flexible TRACE pin assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1709
Table 313. Important TPIU registers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1711
Table 314. DBG register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1713
Table 315. Document revision history . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1716
46/1751 RM0090 Rev 19

RM0090 List of figures
List of figures
Figure 1. System architecture for STM32F405xx/07xx and STM32F415xx/17xx devices. . . . . . . . . 60

Figure 2. System architecture for STM32F42xxx and STM32F43xxx devices . . . . . . . . . . . . . . . . . 62
Figure 3. Flash memory interface connection inside system architecture
Figure 5. Sequential 32-bit instruction execution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Figure 6. RDP levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Figure 7. PCROP levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Figure 8. CRC calculation unit block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Figure 9. Power supply overview for STM32F405xx/07xx and STM32F415xx/17xx . . . . . . . . . . . . 116
Figure 10. Power supply overview for STM32F42xxx and STM32F43xxx. . . . . . . . . . . . . . . . . . . . . 117
Figure 11. Backup domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Figure 12. Power-on reset/power-down reset waveform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
Figure 13. BOR thresholds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
Figure 14. PVD thresholds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
Figure 15. Simplified diagram of the reset circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Figure 16. Clock tree . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
Figure 17. HSE/ LSE clock sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
Figure 18. Frequency measurement with TIM5 in Input capture mode . . . . . . . . . . . . . . . . . . . . . . . 159
Figure 19. Frequency measurement with TIM11 in Input capture mode . . . . . . . . . . . . . . . . . . . . . . 160
Figure 20. Simplified diagram of the reset circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
Figure 21. Clock tree . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
Figure 22. HSE/ LSE clock sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
Figure 23. Frequency measurement with TIM5 in Input capture mode . . . . . . . . . . . . . . . . . . . . . . . 223
Figure 24. Frequency measurement with TIM11 in Input capture mode . . . . . . . . . . . . . . . . . . . . . . 223
Figure 25. Basic structure of a five-volt tolerant I/O port bit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
Figure 26. Selecting an alternate function on STM32F405xx/07xx and STM32F415xx/17xx . . . . 272
Figure 27. Selecting an alternate function on STM32F42xxx and STM32F43xxx . . . . . . . . . . . . . . 273
Figure 28. Input floating/pull up/pull down configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
Figure 29. Output configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Figure 30. Alternate function configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Figure 31. High impedance-analog configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
Figure 32. DMA block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
Figure 33. System implementation of the two DMA controllers
Figure 34. System implementation of the two DMA controllers (STM32F42xxx and STM32F43xxx) 306
Figure 35. Channel selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
Figure 36. Peripheral-to-memory mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
Figure 37. Memory-to-peripheral mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
Figure 38. Memory-to-memory mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
Figure 39. FIFO structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
Figure 40. DMA2D block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
Figure 41. External interrupt/event controller block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380
Figure 42. External interrupt/event GPIO mapping (STM32F405xx/07xx and STM32F415xx/17xx). 382
Figure 43. External interrupt/event GPIO mapping (STM32F42xxx and STM32F43xxx) . . . . . . . . . 383
Figure 44. Single ADC block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
Figure 45. Timing diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392
RM0090 Rev 19 47/1751

56
List of figures RM0090
Figure 46. Analog watchdog’s guarded area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393

Figure 47. Injected conversion latency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394
Figure 48. Right alignment of 12-bit data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
Figure 49. Left alignment of 12-bit data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
Figure 50. Left alignment of 6-bit data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
Figure 51. Multi ADC block diagram(1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402
Figure 52. Injected simultaneous mode on 4 channels: dual ADC mode . . . . . . . . . . . . . . . . . . . . . 405
Figure 53. Injected simultaneous mode on 4 channels: triple ADC mode . . . . . . . . . . . . . . . . . . . . . 405
Figure 54. Regular simultaneous mode on 16 channels: dual ADC mode . . . . . . . . . . . . . . . . . . . . 406
Figure 55. Regular simultaneous mode on 16 channels: triple ADC mode . . . . . . . . . . . . . . . . . . . . 406
Figure 56. Interleaved mode on 1 channel in continuous conversion mode: dual ADC mode. . . . . . 407
Figure 57. Interleaved mode on 1 channel in continuous conversion mode: triple ADC mode . . . . . 408
Figure 58. Alternate trigger: injected group of each ADC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
Figure 59. Alternate trigger: 4 injected channels (each ADC) in discontinuous mode . . . . . . . . . . . . 410
Figure 60. Alternate trigger: injected group of each ADC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 410
Figure 61. Alternate + regular simultaneous . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411
Figure 62. Case of trigger occurring during injected conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
Figure 63. Temperature sensor and VREFINT channel block diagram . . . . . . . . . . . . . . . . . . . . . . 413
Figure 64. DAC channel block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434
Figure 65. Data registers in single DAC channel mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
Figure 66. Data registers in dual DAC channel mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
Figure 67. Timing diagram for conversion with trigger disabled TEN = 0 . . . . . . . . . . . . . . . . . . . . . 437
Figure 68. DAC LFSR register calculation algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
Figure 69. DAC conversion (SW trigger enabled) with LFSR wave generation. . . . . . . . . . . . . . . . . 439
Figure 70. DAC triangle wave generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440
Figure 71. DAC conversion (SW trigger enabled) with triangle wave generation . . . . . . . . . . . . . . . 440
Figure 72. DCMI block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456
Figure 73. Top-level block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
Figure 74. DCMI signal waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458
Figure 75. Timing diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 460
Figure 76. Frame capture waveforms in Snapshot mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 462
Figure 77. Frame capture waveforms in continuous grab mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463
Figure 78. Coordinates and size of the window after cropping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 464
Figure 79. Data capture waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 464
Figure 80. Pixel raster scan order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
Figure 81. LTDC block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 481
Figure 82. LCD-TFT Synchronous timings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 484
Figure 83. Layer window programmable parameters: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487
Figure 84. Blending two layers with background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 490
Figure 85. Interrupt events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491
Figure 86. Advanced-control timer block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 517
Figure 87. Counter timing diagram with prescaler division change from 1 to 2 . . . . . . . . . . . . . . . . . 519
Figure 89. Counter timing diagram, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 520
Figure 92. Counter timing diagram, internal clock divided by N. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521
Figure 93. Counter timing diagram, update event when ARPE=0 (TIMx_ARR not preloaded) . . . . . 522
Figure 94. Counter timing diagram, update event when ARPE=1 (TIMx_ARR preloaded) . . . . . . . . 522
48/1751 RM0090 Rev 19


Figure 99. Counter timing diagram, update event when repetition counter is not used . . . . . . . . . . . 526
Figure 100. Counter timing diagram, internal clock divided by 1, TIMx_ARR = 0x6 . . . . . . . . . . . . . . 527
Figure 102. Counter timing diagram, internal clock divided by 4, TIMx_ARR=0x36 . . . . . . . . . . . . . . 528
Figure 104. Counter timing diagram, update event with ARPE=1 (counter underflow) . . . . . . . . . . . . 529
Figure 105. Counter timing diagram, Update event with ARPE=1 (counter overflow) . . . . . . . . . . . . . 529
Figure 106. Update rate examples depending on mode and TIMx_RCR register settings . . . . . . . . . 531
Figure 107. Control circuit in normal mode, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . 532
Figure 108. TI2 external clock connection example. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533
Figure 109. Control circuit in external clock mode 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 534
Figure 110. External trigger input block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 534
Figure 112. Capture/compare channel (example: channel 1 input stage) . . . . . . . . . . . . . . . . . . . . . . 536
Figure 113. Capture/compare channel 1 main circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 536
Figure 114. Output stage of capture/compare channel (channel 1 to 3) . . . . . . . . . . . . . . . . . . . . . . . 537
Figure 115. Output stage of capture/compare channel (channel 4). . . . . . . . . . . . . . . . . . . . . . . . . . . 537
Figure 116. PWM input mode timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 539
Figure 117. Output compare mode, toggle on OC1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 541
Figure 118. Edge-aligned PWM waveforms (ARR=8) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542
Figure 119. Center-aligned PWM waveforms (ARR=8) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543
Figure 120. Complementary output with dead-time insertion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545
Figure 121. Dead-time waveforms with delay greater than the negative pulse. . . . . . . . . . . . . . . . . . 545
Figure 122. Dead-time waveforms with delay greater than the positive pulse. . . . . . . . . . . . . . . . . . . 545
Figure 123. Output behavior in response to a break.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 548
Figure 124. Clearing TIMx OCxREF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 549
Figure 125. 6-step generation, COM example (OSSR=1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 550
Figure 126. Example of one pulse mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551
Figure 127. Example of counter operation in encoder interface mode. . . . . . . . . . . . . . . . . . . . . . . . . 554
Figure 128. Example of encoder interface mode with TI1FP1 polarity inverted. . . . . . . . . . . . . . . . . . 554
Figure 129. Example of Hall sensor interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556
Figure 130. Control circuit in reset mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 557
Figure 131. Control circuit in gated mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 558
Figure 132. Control circuit in trigger mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 559
Figure 133. Control circuit in external clock mode 2 + trigger mode . . . . . . . . . . . . . . . . . . . . . . . . . . 560
Figure 134. General-purpose timer block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 590
Figure 141. Counter timing diagram, Update event when ARPE=0 (TIMx_ARR not preloaded). . . . . 595
Figure 142. Counter timing diagram, Update event when ARPE=1 (TIMx_ARR preloaded). . . . . . . . 595
Figure 147. Counter timing diagram, Update event . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 598
Figure 148. Counter timing diagram, internal clock divided by 1, TIMx_ARR=0x6 . . . . . . . . . . . . . . . 599
RM0090 Rev 19 49/1751

56
Figure 150. Counter timing diagram, internal clock divided by 4, TIMx_ARR=0x36 . . . . . . . . . . . . . . 600
Figure 152. Counter timing diagram, Update event with ARPE=1 (counter underflow). . . . . . . . . . . . 601
Figure 153. Counter timing diagram, Update event with ARPE=1 (counter overflow) . . . . . . . . . . . . . 601
Figure 157. External trigger input block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604
Figure 163. Output compare mode, toggle on OC1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 610
Figure 165. Center-aligned PWM waveforms (ARR=8) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 612
Figure 166. Example of one-pulse mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 613
Figure 167. Clearing TIMx OCxREF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615
Figure 168. Example of counter operation in encoder interface mode . . . . . . . . . . . . . . . . . . . . . . . . 617
Figure 169. Example of encoder interface mode with TI1FP1 polarity inverted . . . . . . . . . . . . . . . . . 617
Figure 173. Control circuit in external clock mode 2 + trigger mode . . . . . . . . . . . . . . . . . . . . . . . . . . 621
Figure 174. Master/Slave timer example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 621
Figure 175. Gating timer 2 with OC1REF of timer 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 622
Figure 176. Gating timer 2 with Enable of timer 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 623
Figure 177. Triggering timer 2 with update of timer 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 624
Figure 178. Triggering timer 2 with Enable of timer 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 625
Figure 179. Triggering timer 1 and 2 with timer 1 TI1 input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626
Figure 180. General-purpose timer block diagram (TIM9 and TIM12) . . . . . . . . . . . . . . . . . . . . . . . . 651
Figure 181. General-purpose timer block diagram (TIM10/11/13/14) . . . . . . . . . . . . . . . . . . . . . . . . 652
Figure 188. Counter timing diagram, update event when ARPE=0 (TIMx_ARR not preloaded) . . . . . 657
Figure 189. Counter timing diagram, update event when ARPE=1 (TIMx_ARR preloaded) . . . . . . . . 657
Figure 197. Output compare mode, toggle on OC1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665
Figure 199. Example of one pulse mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 667
50/1751 RM0090 Rev 19


Figure 203. Basic timer block diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 696
Figure 210. Counter timing diagram, update event when ARPE=0 (TIMx_ARR not
preloaded). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 701
Figure 211. Counter timing diagram, update event when ARPE=1 (TIMx_ARR
preloaded). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 701
Figure 213. Independent watchdog block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 709
Figure 214. Watchdog block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 714
Figure 215. Window watchdog timing diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715
Figure 216. Block diagram (STM32F415/417xx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 722
Figure 217. Block diagram (STM32F43xxx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 723
Figure 218. DES/TDES-ECB mode encryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725
Figure 219. DES/TDES-ECB mode decryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725
Figure 220. DES/TDES-CBC mode encryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 727
Figure 221. DES/TDES-CBC mode decryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 728
Figure 222. AES-ECB mode encryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 729
Figure 223. AES-ECB mode decryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 730
Figure 224. AES-CBC mode encryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 731
Figure 225. AES-CBC mode decryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 732
Figure 226. AES-CTR mode encryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 733
Figure 227. AES-CTR mode decryption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 734
Figure 228. Initial counter block structure for the Counter mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 734
Figure 229. 64-bit block construction according to DATATYPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 741
Figure 230. Initialization vectors use in the TDES-CBC encryption . . . . . . . . . . . . . . . . . . . . . . . . . . . 743
Figure 231. CRYP interrupt mapping diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 748
Figure 232. Block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767
Figure 233. Block diagram for STM32F415/417xx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 773
Figure 234. Block diagram for STM32F43xxx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 774
Figure 235. Bit, byte and half-word swapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 776
Figure 236. HASH interrupt mapping diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 782
Figure 237. RTC block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 800
Figure 238. I2C bus protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 841
Figure 239. I2C block diagram for STM32F40x/41x . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 842
Figure 240. I2C block diagram for STM32F42x/43x . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 843
Figure 241. Transfer sequence diagram for slave transmitter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 845
Figure 242. Transfer sequence diagram for slave receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 846
Figure 243. Transfer sequence diagram for master transmitter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 849
Figure 244. Transfer sequence diagram for master receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 850
Figure 245. I2C interrupt mapping diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 859
Figure 246. SPI block diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 876
Figure 247. Single master/ single slave application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 877
Figure 248. Data clock timing diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 879
Figure 249. TI mode - Slave mode, single transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 881
Figure 250. TI mode - Slave mode, continuous transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 882
Figure 251. TI mode - master mode, single transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 883
RM0090 Rev 19 51/1751

56
Figure 252. TI mode - master mode, continuous transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 884

Figure 253. TXE/RXNE/BSY behavior in Master / full-duplex mode (BIDIMODE=0 and
RXONLY=0) in case of continuous transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 887
Figure 254. TXE/RXNE/BSY behavior in Slave / full-duplex mode (BIDIMODE=0,
RXONLY=0) in case of continuous transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 888
Figure 255. TXE/BSY behavior in Master transmit-only mode (BIDIMODE=0 and RXONLY=0)
in case of continuous transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 889
Figure 256. TXE/BSY in Slave transmit-only mode (BIDIMODE=0 and RXONLY=0) in case of
continuous transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 889
Figure 257. RXNE behavior in receive-only mode (BIDIRMODE=0 and RXONLY=1)
in case of continuous transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 890
Figure 258. TXE/BSY behavior when transmitting (BIDIRMODE=0 and RXONLY=0)
in case of discontinuous transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 891
Figure 259. Transmission using DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 896
Figure 260. Reception using DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 896
Figure 261. TI mode frame format error detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 898
Figure 262. I2S block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 899
Figure 263. I2S full duplex block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 900
Figure 264. I2S Philips protocol waveforms (16/32-bit full accuracy, CPOL = 0). . . . . . . . . . . . . . . . . 902
Figure 265. I2S Philips standard waveforms (24-bit frame with CPOL = 0) . . . . . . . . . . . . . . . . . . . . . 902
Figure 266. Transmitting 0x8EAA33 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902
Figure 267. Receiving 0x8EAA33 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 903
Figure 268. I2S Philips standard (16-bit extended to 32-bit packet frame with CPOL = 0) . . . . . . . . . 903
Figure 269. Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 903
Figure 270. MSB justified 16-bit or 32-bit full-accuracy length with CPOL = 0 . . . . . . . . . . . . . . . . . . 904
Figure 271. MSB justified 24-bit frame length with CPOL = 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 904
Figure 272. MSB justified 16-bit extended to 32-bit packet frame with CPOL = 0 . . . . . . . . . . . . . . . . 904
Figure 273. LSB justified 16-bit or 32-bit full-accuracy with CPOL = 0 . . . . . . . . . . . . . . . . . . . . . . . . 905
Figure 274. LSB justified 24-bit frame length with CPOL = 0. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 905
Figure 275. Operations required to transmit 0x3478AE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 905
Figure 276. Operations required to receive 0x3478AE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 906
Figure 277. LSB justified 16-bit extended to 32-bit packet frame with CPOL = 0 . . . . . . . . . . . . . . . . 906
Figure 278. Example of LSB justified 16-bit extended to 32-bit packet frame . . . . . . . . . . . . . . . . . . . 906
Figure 279. PCM standard waveforms (16-bit) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 907
Figure 280. PCM standard waveforms (16-bit extended to 32-bit packet frame). . . . . . . . . . . . . . . . . 907
Figure 281. Audio sampling frequency definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 908
Figure 282. I2S clock generator architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 908
Figure 283. Functional block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 928
Figure 284. Audio frame . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 930
Figure 285. FS role is start of frame + channel side identification (FSDEF = TRIS = 1) . . . . . . . . . . . 933
Figure 286. FS role is start of frame (FSDEF = 0) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 933
Figure 287. Slot size configuration with FBOFF = 0 in SAI_xSLOTR . . . . . . . . . . . . . . . . . . . . . . . . . 934
Figure 288. First bit offset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 934
Figure 289. Audio block clock generator overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 935
Figure 290. AC’97 audio frame . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 939
Figure 291. Data companding hardware in an audio block in the SAI . . . . . . . . . . . . . . . . . . . . . . . . . 941
Figure 292. Tristate strategy on SD output line on an inactive slot . . . . . . . . . . . . . . . . . . . . . . . . . . . 943
Figure 293. Tristate on output data line in a protocol like I2S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 944
Figure 294. Overrun detection error. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 945
Figure 295. FIFO underrun event . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 946
Figure 296. USART block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 968
Figure 297. Word length programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 969
52/1751 RM0090 Rev 19

Figure 298. Configurable stop bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 971

Figure 299. TC/TXE behavior when transmitting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 972
Figure 300. Start bit detection when oversampling by 16 or 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 973
Figure 301. Data sampling when oversampling by 16 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 976
Figure 302. Data sampling when oversampling by 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 977
Figure 303. Mute mode using Idle line detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 990
Figure 304. Mute mode using address mark detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 990
Figure 305. Break detection in LIN mode (11-bit break length - LBDL bit is set) . . . . . . . . . . . . . . . . . 993
Figure 306. Break detection in LIN mode vs. Framing error detection. . . . . . . . . . . . . . . . . . . . . . . . . 994
Figure 307. USART example of synchronous transmission. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 995
Figure 308. USART data clock timing diagram (M=0) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 995
Figure 309. USART data clock timing diagram (M=1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 996
Figure 310. RX data setup/hold time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 996
Figure 311. ISO 7816-3 asynchronous protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 997
Figure 312. Parity error detection using the 1.5 stop bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 998
Figure 313. IrDA SIR ENDEC- block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1000
Figure 314. IrDA data modulation (3/16) -Normal mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1000
Figure 315. Transmission using DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1002
Figure 316. Reception using DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1003
Figure 317. Hardware flow control between 2 USARTs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1003
Figure 318. RTS flow control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1004
Figure 319. CTS flow control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1005
Figure 320. USART interrupt mapping diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1006
Figure 321. SDIO “no response” and “no data” operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1020
Figure 322. SDIO (multiple) block read operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1020
Figure 323. SDIO (multiple) block write operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1021
Figure 324. SDIO sequential read operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1021
Figure 325. SDIO sequential write operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1021
Figure 326. SDIO block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1022
Figure 327. SDIO adapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1023
Figure 328. Control unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1024
Figure 329. SDIO adapter command path . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1025
Figure 330. Command path state machine (CPSM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1026
Figure 331. SDIO command transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1027
Figure 332. Data path . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1029
Figure 333. Data path state machine (DPSM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1030
Figure 334. CAN network topology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1077
Figure 335. Dual CAN block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1079
Figure 336. bxCAN operating modes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1081
Figure 337. bxCAN in silent mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1082
Figure 338. bxCAN in loop back mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1082
Figure 339. bxCAN in combined mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1083
Figure 340. Transmit mailbox states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1085
Figure 341. Receive FIFO states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1086
Figure 342. Filter bank scale configuration - register organization . . . . . . . . . . . . . . . . . . . . . . . . . . 1088
Figure 343. Example of filter numbering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1089
Figure 344. Filtering mechanism - Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1090
Figure 345. CAN error state diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1092
Figure 346. Bit timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1094
Figure 347. CAN frames . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1095
Figure 348. Event flags and interrupt generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1096
Figure 349. RX and TX mailboxes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1107
RM0090 Rev 19 53/1751

56
Figure 350. ETH block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1126

Figure 351. SMI interface signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1127
Figure 352. MDIO timing and frame structure - Write cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1128
Figure 353. MDIO timing and frame structure - Read cycle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1129
Figure 354. Media independent interface signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1129
Figure 355. MII clock sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1131
Figure 356. Reduced media-independent interface signals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1132
Figure 357. RMII clock sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1132
Figure 358. Clock scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1133
Figure 359. Address field format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1135
Figure 360. MAC frame format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1136
Figure 361. Tagged MAC frame format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1137
Figure 362. Transmission bit order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1143
Figure 363. Transmission with no collision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1143
Figure 364. Transmission with collision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1144
Figure 365. Frame transmission in MMI and RMII modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1144
Figure 366. Receive bit order. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1148
Figure 367. Reception with no error. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1149
Figure 368. Reception with errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1149
Figure 369. Reception with false carrier indication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1149
Figure 370. MAC core interrupt masking scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1150
Figure 371. Wakeup frame filter register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1154
Figure 372. Networked time synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1158
Figure 373. System time update using the Fine correction method. . . . . . . . . . . . . . . . . . . . . . . . . . 1160
Figure 374. PTP trigger output to TIM2 ITR1 connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1162
Figure 375. PPS output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1163
Figure 376. Descriptor ring and chain structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1164
Figure 377. TxDMA operation in Default mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1168
Figure 378. TxDMA operation in OSF mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1170
Figure 379. Normal transmit descriptor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1171
Figure 380. Enhanced transmit descriptor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1177
Figure 381. Receive DMA operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1179
Figure 382. Normal Rx DMA descriptor structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1181
Figure 383. Enhanced receive descriptor field format with IEEE1588 time stamp enabled. . . . . . . . 1187
Figure 384. Interrupt scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1190
Figure 385. Ethernet MAC remote wakeup frame filter register (ETH_MACRWUFFR). . . . . . . . . . . 1200
Figure 386. OTG full-speed block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1243
Figure 387. OTG A-B device connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1245
Figure 388. USB peripheral-only connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1247
Figure 389. USB host-only connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1251
Figure 390. SOF connectivity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1255
Figure 391. Updating OTG_FS_HFIR dynamically . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1258
Figure 392. Device-mode FIFO address mapping and AHB FIFO access mapping . . . . . . . . . . . . . 1259
Figure 393. Host-mode FIFO address mapping and AHB FIFO access mapping . . . . . . . . . . . . . . . 1260
Figure 394. Interrupt hierarchy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1264
Figure 395. CSR memory map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1266
Figure 396. Transmit FIFO write task . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1338
Figure 397. Receive FIFO read task . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1339
Figure 398. Normal bulk/control OUT/SETUP and bulk/control IN transactions . . . . . . . . . . . . . . . . 1340
Figure 399. Bulk/control IN transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1343
Figure 400. Normal interrupt OUT/IN transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1345
Figure 401. Normal isochronous OUT/IN transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1350
54/1751 RM0090 Rev 19

Figure 402. Receive FIFO packet read . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1356

Figure 403. Processing a SETUP packet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1358
Figure 404. Bulk OUT transaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1365
Figure 405. TRDT max timing case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1374
Figure 406. A-device SRP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1375
Figure 407. B-device SRP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1376
Figure 408. A-device HNP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1377
Figure 409. B-device HNP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1379
Figure 410. USB OTG interface block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1384
Figure 411. USB host-only connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1391
Figure 412. SOF trigger output to TIM2 ITR1 connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1396
Figure 413. Updating OTG_HS_HFIR dynamically . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1398
Figure 414. Interrupt hierarchy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1401
Figure 415. CSR memory map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1403
Figure 416. Transmit FIFO write task . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1492
Figure 417. Receive FIFO read task . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1493
Figure 418. Normal bulk/control OUT/SETUP and bulk/control IN transactions - DMA
mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1494
Figure 419. Normal bulk/control OUT/SETUP and bulk/control IN transactions - Slave
mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1495
Figure 420. Bulk/control IN transactions - DMA mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1498
Figure 421. Bulk/control IN transactions - Slave mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1499
Figure 422. Normal interrupt OUT/IN transactions - DMA mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1501
Figure 423. Normal interrupt OUT/IN transactions - Slave mode . . . . . . . . . . . . . . . . . . . . . . . . . . . 1502
Figure 424. Normal isochronous OUT/IN transactions - DMA mode . . . . . . . . . . . . . . . . . . . . . . . . . 1507
Figure 425. Normal isochronous OUT/IN transactions - Slave mode . . . . . . . . . . . . . . . . . . . . . . . . 1508
Figure 426. Receive FIFO packet read in slave mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1519
Figure 427. Processing a SETUP packet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1521
Figure 428. Slave mode bulk OUT transaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1528
Figure 429. TRDT max timing case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1537
Figure 430. A-device SRP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1538
Figure 431. B-device SRP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1539
Figure 432. A-device HNP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1540
Figure 433. B-device HNP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1542
Figure 434. FSMC block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1545
Figure 435. FSMC memory banks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1547
Figure 436. Mode1 read accesses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1554
Figure 437. Mode1 write accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1555
Figure 438. ModeA read accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1556
Figure 439. ModeA write accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1557
Figure 440. Mode2 and mode B read accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1559
Figure 441. Mode2 write accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1559
Figure 442. Mode B write accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1560
Figure 443. Mode C read accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1562
Figure 444. Mode C write accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1562
Figure 445. Mode D read accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1564
Figure 446. Mode D write accesses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1565
Figure 447. Multiplexed read accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1567
Figure 448. Multiplexed write accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1567
Figure 449. Asynchronous wait during a read access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1570
Figure 450. Asynchronous wait during a write access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1570
Figure 451. Wait configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1572
RM0090 Rev 19 55/1751

56
Figure 452. Synchronous multiplexed read mode - NOR, PSRAM (CRAM) . . . . . . . . . . . . . . . . . . . 1573
Figure 453. Synchronous multiplexed write mode - PSRAM (CRAM) . . . . . . . . . . . . . . . . . . . . . . . . 1575
Figure 454. NAND/PC Card controller timing for common memory access . . . . . . . . . . . . . . . . . . . 1588
Figure 455. Access to non ‘CE don’t care’ NAND-Flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1589
Figure 456. FMC block diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1604
Figure 457. FMC memory banks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1607
Figure 458. Mode1 read access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1618
Figure 459. Mode1 write access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1618
Figure 460. ModeA read access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1620
Figure 461. ModeA write access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1621
Figure 462. Mode2 and mode B read access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1623
Figure 463. Mode2 write access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1623
Figure 464. ModeB write access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1624
Figure 465. ModeC read access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1626
Figure 466. ModeC write access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1626
Figure 467. ModeD read access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1628
Figure 468. ModeD write access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1629
Figure 469. Muxed read access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1631
Figure 470. Muxed write access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1631
Figure 471. Asynchronous wait during a read access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . 1633
Figure 472. Asynchronous wait during a write access waveforms. . . . . . . . . . . . . . . . . . . . . . . . . . . 1634
Figure 473. Wait configuration waveforms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1636
Figure 474. Synchronous multiplexed read mode waveforms - NOR, PSRAM (CRAM) . . . . . . . . . . 1636
Figure 475. Synchronous multiplexed write mode waveforms - PSRAM (CRAM). . . . . . . . . . . . . . . 1638
Figure 476. NAND Flash/PC Card controller waveforms for common memory access. . . . . . . . . . . 1651
Figure 477. Access to non ‘CE don’t care’ NAND-Flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1652
Figure 478. Burst write SDRAM access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1665
Figure 479. Burst read SDRAM access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1666
Figure 480. Logic diagram of Read access with RBURST bit set (CAS=2, RPIPE=0) . . . . . . . . . . . 1667
Figure 481. Read access crossing row boundary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1669
Figure 482. Write access crossing row boundary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1669
Figure 483. Self-refresh mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1672
Figure 484. Power-down mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1673
Figure 485. Block diagram of STM32 MCU and Cortex®-M4 with FPU-level debug support . . . . . . 1683
Figure 486. SWJ debug port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1685
Figure 487. JTAG TAP connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1689
Figure 488. TPIU block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1707
56/1751 RM0090 Rev 19

RM0090 Documentation conventions
1 Documentation conventions
The STM32F401xx devices have an Arm®(a) Cortex®-M4 with FPU.
1.1 List of abbreviations for registers

The following abbreviations are used in register descriptions:
read/write (rw) Software can read and write to these bits.

read-only (r) Software can only read these bits.
write-only (w) Software can only write to this bit. Reading the bit returns the reset
value.
read/clear (rc_w1) Software can read as well as clear this bit by writing 1. Writing ‘0’ has
no effect on the bit value.
read/clear (rc_w0) Software can read as well as clear this bit by writing 0. Writing ‘1’ has
no effect on the bit value.
read/clear by read Software can read this bit. Reading this bit automatically clears it to ‘0’.
(rc_r) Writing ‘0’ has no effect on the bit value.
read/set (rs) Software can read as well as set this bit. Writing ‘0’ has no effect on the
bit value.
read-only write Software can read this bit. Writing ‘0’ or ‘1’ triggers an event but has no
trigger (rt_w) effect on the bit value.
toggle (t) Software can only toggle this bit by writing ‘1’. Writing ‘0’ has no effect.
Reserved (Res.) Reserved bit, must be kept at reset value.
a. Arm is a registered trademark of Arm Limited (or its subsidiaries) in the US and/or elsewhere.
RM0090 Rev 19 57/1751

112
Documentation conventions RM0090
1.2 Glossary
This section gives a brief definition of acronyms and abbreviations used in this document:
• The CPU core integrates two debug ports:
– JTAG debug port (JTAG-DP) provides a 5-pin standard interface based on the
Joint Test Action Group (JTAG) protocol.
– SWD debug port (SWD-DP) provides a 2-pin (clock and data) interface based on
the Serial Wire Debug (SWD) protocol.
For both the JTAG and SWD protocols, please refer to the Cortex®-M4 with FPU
Technical Reference Manual
• Word: data/instruction of 32-bit length.
• Half word: data/instruction of 16-bit length.
• Byte: data of 8-bit length.
• Double word: data of 64-bit length.
• IAP (in-application programming): IAP is the ability to reprogram the Flash memory of a
microcontroller while the user program is running.
• ICP (in-circuit programming): ICP is the ability to program the Flash memory of a
microcontroller using the JTAG protocol, the SWD protocol or the bootloader while the
device is mounted on the user application board.
• I-Code: this bus connects the Instruction bus of the CPU core to the Flash instruction
interface. Prefetch is performed on this bus.
• D-Code: this bus connects the D-Code bus (literal load and debug access) of the CPU
to the Flash data interface.
• Option bytes: product configuration bits stored in the Flash memory.
• OBL: option byte loader.
• AHB: advanced high-performance bus.
• CPU: refers to the Cortex®-M4 with FPU core.
1.3 Peripheral availability

For peripheral availability and number across all STM32F405xx/07xx and
STM32F415xx/17xx sales types, please refer to the STM32F405xx/07xx and
STM32F415xx/17xx datasheets.
For peripheral availability and number across all STM32F42xxx and STM32F43xxx sales
types, please refer to the STM32F42xxx and STM32F43xxx datasheets.
58/1751 RM0090 Rev 19

RM0090 Memory and bus architecture
2 Memory and bus architecture
2.1 System architecture

In STM32F405xx/07xx and STM32F415xx/17xx, the main system consists of 32-bit
multilayer AHB bus matrix that interconnects:
The main system consists of 32-bit multilayer AHB bus matrix that interconnects:
• Eight masters:
– Cortex®-M4 with FPU core I-bus, D-bus and S-bus
– DMA1 memory bus
– DMA2 memory bus
– DMA2 peripheral bus
– Ethernet DMA bus
– USB OTG HS DMA bus
• Seven slaves:
– Internal Flash memory ICode bus
– Internal Flash memory DCode bus
– Main internal SRAM1 (112 KB)
– Auxiliary internal SRAM2 (16 KB)
– AHB1 peripherals including AHB to APB bridges and APB peripherals
– AHB2 peripherals
– FSMC
The bus matrix provides access from a master to a slave, enabling concurrent access and
efficient operation even when several high-speed peripherals work simultaneously. The 64-
Kbyte CCM (core coupled memory) data RAM is not part of the bus matrix and can be
accessed only through the CPU. This architecture is shown in Figure 1.
RM0090 Rev 19 59/1751

112
Memory and bus architecture RM0090
Figure 1. System architecture for STM32F405xx/07xx and STM32F415xx/17xx devices
64-Kbyte ARM GP GP MAC USB OTG

CCM data RAM Cortex-M4 DMA1 DMA2 Ethernet HS
DMA_P1
I-bus
USB_HS_M
DMA_MEM1
DMA_MEM2
D-bus
DMA_P2
ETHERNET_M
S-bus
ICODE
Flash
ACCEL
DCODE memory
SRAM1
112 Kbyte
SRAM2
16 Kbyte
AHB1
peripherals APB1
AHB2
peripherals
APB2
FSMC
Static MemCtl
Bus matrix-S
ai18490e
60/1751 RM0090 Rev 19

In the STM32F42xx and STM32F43xx devices, the main system consists of 32-bit multilayer
AHB bus matrix that interconnects:
• Ten masters:
– Cortex®-M4 with FPU core I-bus, D-bus and S-bus
– DMA1 memory bus
– DMA2 memory bus
– DMA2 peripheral bus
– Ethernet DMA bus
– USB OTG HS DMA bus
– LCD Controller DMA-bus
– DMA2D (Chrom-Art Accelerator™) memory bus
• Eight slaves:
– Internal Flash memory ICode bus
– Internal Flash memory DCode bus
– Main internal SRAM1 (112 KB)
– AHB1peripherals including AHB to APB bridges and APB peripherals
– AHB2 peripherals
– FMC
The bus matrix provides access from a master to a slave, enabling concurrent access and
efficient operation even when several high-speed peripherals work simultaneously. The 64-
Kbyte CCM (core coupled memory) data RAM is not part of the bus matrix and can be
accessed only through the CPU. This architecture is shown in Figure 2.
RM0090 Rev 19 61/1751

112
Figure 2. System architecture for STM32F42xxx and STM32F43xxx devices
64-Kbyte ARM GP GP MAC USB OTG Chrom ART Accelerator(DMA2D)

DMA1 LCD-TFT
CCM data RAM Cortex-M4 DMA2 Ethernet HS
DMA_P1
I-bus
D-bus
S-bus
DMA2D
DMA_MEM1
DMA_MEM2
DMA_P2
USB_HS_M
LCD-TFT_M
ETHERNET_M
ICODE
Flash
ACCEL
DCODE memory
SRAM1
112 Kbyte
SRAM2
16 Kbyte
SRAM3
64 Kbyte
AHB2 APB1
peripherals
AHB1
peripherals APB2
FMC external
MemCtl
Bus matrix-S
MS30421V7
2.1.1 I-bus
This bus connects the Instruction bus of the Cortex®-M4 with FPU core to the BusMatrix.
This bus is used by the core to fetch instructions. The target of this bus is a memory
containing code (internal Flash memory/SRAM or external memories through the
FSMC/FMC).
2.1.2 D-bus
This bus connects the databus of the Cortex®-M4 with FPU to the 64-Kbyte CCM data RAM
to the BusMatrix. This bus is used by the core for literal load and debug access. The target
of this bus is a memory containing code or data (internal Flash memory or external
memories through the FSMC/FMC).
2.1.3 S-bus
This bus connects the system bus of the Cortex®-M4 with FPU core to a BusMatrix. This
bus is used to access data located in a peripheral or in SRAM. Instructions may also be
fetched on this bus (less efficient than ICode). The targets of this bus are the internal
SRAM1, SRAM2 and SRAM3, the AHB1 peripherals including the APB peripherals, the
AHB2 peripherals and the external memories through the FSMC/FMC.
62/1751 RM0090 Rev 19

2.1.4 DMA memory bus

This bus connects the DMA memory bus master interface to the BusMatrix. It is used by the
DMA to perform transfer to/from memories. The targets of this bus are data memories:
internal SRAMs (SRAM1, SRAM2 and SRAM3) and external memories through the
FSMC/FMC.
2.1.5 DMA peripheral bus

This bus connects the DMA peripheral master bus interface to the BusMatrix. This bus is
used by the DMA to access AHB peripherals or to perform memory-to-memory transfers.
The targets of this bus are the AHB and APB peripherals plus data memories: internal
SRAMs (SRAM1, SRAM2 and SRAM3) and external memories through the FSMC/FMC.
2.1.6 Ethernet DMA bus

This bus connects the Ethernet DMA master interface to the BusMatrix. This bus is used by
the Ethernet DMA to load/store data to a memory. The targets of this bus are data
memories: internal SRAMs (SRAM1, SRAM2, SRAM3), internal Flash memory, and
external memories through the FSMC/FMC.
2.1.7 USB OTG HS DMA bus

This bus connects the USB OTG HS DMA master interface to the BusMatrix. This bus is
used by the USB OTG DMA to load/store data to a memory. The targets of this bus are data
memories: internal SRAMs (SRAM1, SRAM2, SRAM3), internal Flash memory, and
external memories through the FSMC/FMC.
2.1.8 LCD-TFT controller DMA bus

This bus connects the LCD controller DMA master interface to the BusMatrix. It is used by
the LCD-TFT DMA to load/store data to a memory. The targets of this bus are data
memories: internal SRAMs (SRAM1, SRAM2, SRAM3), external memories through FMC,
and internal Flash memory.
2.1.9 DMA2D bus

This bus connect the DMA2D master interface to the BusMatrix. This bus is used by the
DMA2D graphic Accelerator to load/store data to a memory. The targets of this bus are data
memories: internal SRAMs (SRAM1, SRAM2, SRAM3), external memories through FMC,
and internal Flash memory.
2.1.10 BusMatrix
The BusMatrix manages the access arbitration between masters. The arbitration uses a
round-robin algorithm.
RM0090 Rev 19 63/1751

112
2.1.11 AHB/APB bridges (APB)

The two AHB/APB bridges, APB1 and APB2, provide full synchronous connections between
the AHB and the two APB buses, allowing flexible selection of the peripheral frequency.
Refer to the device datasheets for more details on APB1 and APB2 maximum frequencies,
and to Table 1 for the address mapping of AHB and APB peripherals.
After each device reset, all peripheral clocks are disabled (except for the SRAM and Flash
memory interface). Before using a peripheral you have to enable its clock in the
RCC_AHBxENR or RCC_APBxENR register.
Note: When a 16- or an 8-bit access is performed on an APB register, the access is transformed
into a 32-bit access: the bridge duplicates the 16- or 8-bit data to feed the 32-bit vector.
2.2 Memory organization

Program memory, data memory, registers and I/O ports are organized within the same linear
4 Gbyte address space.
The bytes are coded in memory in little endian format. The lowest numbered byte in a word
is considered the word’s least significant byte and the highest numbered byte, the word’s
most significant.
For the detailed mapping of peripheral registers, please refer to the related chapters.
The addressable memory space is divided into 8 main blocks, each of 512 MB.
All the memory areas that are not allocated to on-chip memories and peripherals are
considered “Reserved”). Refer to the memory map figure in the product datasheet.
2.3 Memory map

See the datasheet corresponding to your device for a comprehensive diagram of the
memory map. Table 1 gives the boundary addresses of the peripherals available in all
STM32F4xx devices.
Table 1. STM32F4xx register boundary addresses

Boundary address Peripheral Bus Register map
FSMC control register

(STM32F405xx/07xx
and Section 36.6.9: FSMC register map on page 1600
0xA000 0000 - 0xA000 0FFF STM32F415xx/17xx)/ AHB3
Section 37.8: FMC register map on page 1680
FMC control register
(STM32F42xxx and
STM32F43xxx)
64/1751 RM0090 Rev 19

Table 1. STM32F4xx register boundary addresses (continued)

0x5006 0800 - 0x5006 0BFF RNG Section 24.4.4: RNG register map on page 771
0x5006 0400 - 0x5006 07FF HASH Section 25.4.9: HASH register map on page 795
0x5006 0000 - 0x5006 03FF CRYP Section 23.6.13: CRYP register map on page 763
AHB2
0x5005 0000 - 0x5005 03FF DCMI Section 15.8.12: DCMI register map on page 478
Section 34.16.6: OTG_FS register map on
0x5000 0000 - 0x5003 FFFF USB OTG FS
page 1326
Section 35.12.6: OTG_HS register map on
0x4004 0000 - 0x4007 FFFF USB OTG HS
page 1472
0x4002 B000 - 0x4002 BBFF DMA2D Section 11.5: DMA2D registers on page 352
Section 33.8.5: Ethernet register maps on
0x4002 8000 - 0x4002 93FF ETHERNET MAC
page 1236
0x4002 6400 - 0x4002 67FF DMA2
Section 10.5.11: DMA register map on page 335
0x4002 6000 - 0x4002 63FF DMA1
0x4002 4000 - 0x4002 4FFF BKPSRAM
Flash interface
0x4002 3C00 - 0x4002 3FFF Section 3.9: Flash interface registers
register
0x4002 3800 - 0x4002 3BFF RCC Section 7.3.24: RCC register map on page 265
0x4002 3000 - 0x4002 33FF CRC Section 4.4.4: CRC register map on page 115
AHB1
0x4002 2800 - 0x4002 2BFF GPIOK
Section 8.4.11: GPIO register map on page 287
0x4002 2400 - 0x4002 27FF GPIOJ
0x4002 2000 - 0x4002 23FF GPIOI
0x4002 1C00 - 0x4002 1FFF GPIOH
0x4002 1800 - 0x4002 1BFF GPIOG
0x4002 1400 - 0x4002 17FF GPIOF
0x4002 1000 - 0x4002 13FF GPIOE Section 8.4.11: GPIO register map on page 287
0x4002 0C00 - 0x4002 0FFF GPIOD
0x4002 0800 - 0x4002 0BFF GPIOC
0x4002 0400 - 0x4002 07FF GPIOB
0x4002 0000 - 0x4002 03FF GPIOA
0x4001 6800 - 0x4001 6BFF LCD-TFT Section 16.7.26: LTDC register map on page 512
APB2
0x4001 5800 - 0x4001 5BFF SAI1 Section 29.17.9: SAI register map on page 963
0x4001 5400 - 0x4001 57FF SPI6
APB2 Section 28.5.10: SPI register map on page 925
0x4001 5000 - 0x4001 53FF SPI5
RM0090 Rev 19 65/1751

112

0x4001 4800 - 0x4001 4BFF TIM11 Section 19.5.12: TIM10/11/13/14 register map on
0x4001 4400 - 0x4001 47FF TIM10 page 694
Section 19.4.13: TIM9/12 register map on

0x4001 4000 - 0x4001 43FF TIM9
page 684
0x4001 3C00 - 0x4001 3FFF EXTI APB2 Section 12.3.7: EXTI register map on page 387
Section 9.2.8: SYSCFG register maps for
STM32F405xx/07xx and STM32F415xx/17xx on
0x4001 3800 - 0x4001 3BFF SYSCFG page 294 and Section 9.3.8: SYSCFG register
maps for STM32F42xxx and STM32F43xxx on
page 301
0x4001 3400 - 0x4001 37FF SPI4 APB2 Section 28.5.10: SPI register map on page 925
0x4001 3000 - 0x4001 33FF SPI1 Section 28.5.10: SPI register map on page 925
0x4001 2C00 - 0x4001 2FFF SDIO Section 31.9.16: SDIO register map on page 1074
0x4001 2000 - 0x4001 23FF ADC1 - ADC2 - ADC3 Section 13.13.18: ADC register map on page 430
0x4001 1400 - 0x4001 17FF USART6 APB2
Section 30.6.8: USART register map on page 1018
0x4001 1000 - 0x4001 13FF USART1
0x4001 0400 - 0x4001 07FF TIM8 Section 17.4.21: TIM1 and TIM8 register map on
0x4001 0000 - 0x4001 03FF TIM1 page 587
0x4000 7C00 - 0x4000 7FFF UART8

APB1 Section 30.6.8: USART register map on page 1018
0x4000 7800 - 0x4000 7BFF UART7
66/1751 RM0090 Rev 19


0x4000 7400 - 0x4000 77FF DAC Section 14.5.15: DAC register map on page 453
0x4000 7000 - 0x4000 73FF PWR Section 5.6: PWR register map on page 149
0x4000 6800 - 0x4000 6BFF CAN2
Section 32.9.5: bxCAN register map on page 1118
0x4000 6400 - 0x4000 67FF CAN1
0x4000 5C00 - 0x4000 5FFF I2C3
0x4000 5800 - 0x4000 5BFF I2C2 Section 27.6.11: I2C register map on page 872
0x4000 5400 - 0x4000 57FF I2C1
0x4000 5000 - 0x4000 53FF UART5
0x4000 4C00 - 0x4000 4FFF UART4
Section 30.6.8: USART register map on page 1018
0x4000 4800 - 0x4000 4BFF USART3
0x4000 4400 - 0x4000 47FF USART2
0x4000 4000 - 0x4000 43FF I2S3ext
0x4000 3C00 - 0x4000 3FFF SPI3 / I2S3
Section 28.5.10: SPI register map on page 925
0x4000 3800 - 0x4000 3BFF SPI2 / I2S2
APB1
0x4000 3400 - 0x4000 37FF I2S2ext
0x4000 3000 - 0x4000 33FF IWDG Section 21.4.5: IWDG register map on page 712
0x4000 2C00 - 0x4000 2FFF WWDG Section 22.6.4: WWDG register map on page 719
0x4000 2800 - 0x4000 2BFF RTC & BKP Registers Section 26.6.21: RTC register map on page 836
0x4000 2000 - 0x4000 23FF TIM14 Section 19.5.12: TIM10/11/13/14 register map on
0x4000 1C00 - 0x4000 1FFF TIM13 page 694
Section 19.4.13: TIM9/12 register map on

0x4000 1800 - 0x4000 1BFF TIM12
page 684
0x4000 1400 - 0x4000 17FF TIM7 Section 20.4.9: TIM6 and TIM7 register map on
0x4000 1000 - 0x4000 13FF TIM6 page 707
0x4000 0C00 - 0x4000 0FFF TIM5

0x4000 0800 - 0x4000 0BFF TIM4
Section 18.4.21: TIMx register map on page 648
0x4000 0400 - 0x4000 07FF TIM3
0x4000 0000 - 0x4000 03FF TIM2
RM0090 Rev 19 67/1751

112
2.3.1 Embedded SRAM

The STM32F405xx/07xx and STM32F415xx/17xx feature 4 Kbytes of backup SRAM (see
Section 5.1.2: Battery backup domain) plus 192 Kbytes of system SRAM.
The STM32F42xxx and STM32F43xxx feature 4 Kbytes of backup SRAM (see
Section 5.1.2: Battery backup domain) plus 256 Kbytes of system SRAM.
The embedded SRAM can be accessed as bytes, half-words (16 bits) or full words (32 bits).
Read and write operations are performed at CPU speed with 0 wait state. The embedded
SRAM is divided into up to three blocks:
• SRAM1 and SRAM2 mapped at address 0x2000 0000 and accessible by all AHB
masters.
• SRAM3 (available on STM32F42xxx and STM32F43xxx) mapped at address
0x2002 0000 and accessible by all AHB masters.
• CCM (core coupled memory) mapped at address 0x1000 0000 and accessible only by
the CPU through the D-bus.
The AHB masters support concurrent SRAM accesses (from the Ethernet or the USB OTG
HS): for instance, the Ethernet MAC can read/write from/to SRAM2 while the CPU is
reading/writing from/to SRAM1 or SRAM3.
The CPU can access the SRAM1, SRAM2, and SRAM3 through the System Bus or through
the I-Code/D-Code buses when boot from SRAM is selected or when physical remap is
selected (Section 9.2.1: SYSCFG memory remap register (SYSCFG_MEMRMP) in the
SYSCFG controller). To get the max performance on SRAM execution, physical remap
should be selected (boot or software selection).
2.3.2 Flash memory overview

The Flash memory interface manages CPU AHB I-Code and D-Code accesses to the Flash
memory. It implements the erase and program Flash memory operations and the read and
write protection mechanisms. It accelerates code execution with a system of instruction
prefetch and cache lines.
The Flash memory is organized as follows:
• A main memory block divided into sectors.
• System memory from which the device boots in System memory boot mode
• 512 OTP (one-time programmable) bytes for user data.
• Option bytes to configure read and write protection, BOR level, watchdog
software/hardware and reset when the device is in Standby or Stop mode.
Refer to Section 3: Embedded Flash memory interface for more details.
2.3.3 Bit banding

The Cortex®-M4 with FPU memory map includes two bit-band regions. These regions map
each word in an alias region of memory to a bit in a bit-band region of memory. Writing to a
word in the alias region has the same effect as a read-modify-write operation on the
targeted bit in the bit-band region.
In the STM32F4xx devices both the peripheral registers and the SRAM are mapped to a bit-
band region, so that single bit-band write and read operations are allowed. The operations
68/1751 RM0090 Rev 19

are only available for Cortex®-M4 with FPU accesses, and not from other bus masters (e.g.
DMA).
A mapping formula shows how to reference each word in the alias region to a corresponding
bit in the bit-band region. The mapping formula is:
bit_word_addr = bit_band_base + (byte_offset x 32) + (bit_number × 4)
where:
– bit_word_addr is the address of the word in the alias memory region that maps to
the targeted bit
– bit_band_base is the starting address of the alias region
– byte_offset is the number of the byte in the bit-band region that contains the
targeted bit
– bit_number is the bit position (0-7) of the targeted bit
Example
The following example shows how to map bit 2 of the byte located at SRAM address
0x20000300 to the alias region:
0x22006008 = 0x22000000 + (0x300*32) + (2*4)
Writing to address 0x22006008 has the same effect as a read-modify-write operation on bit
2 of the byte at SRAM address 0x20000300.
Reading address 0x22006008 returns the value (0x01 or 0x00) of bit 2 of the byte at SRAM
address 0x20000300 (0x01: bit set; 0x00: bit reset).
For more information on bit-banding, please refer to the Cortex®-M4 with FPU programming
manual (see Related documents on page 1).
2.4 Boot configuration

Due to its fixed memory map, the code area starts from address 0x0000 0000 (accessed
through the ICode/DCode buses) while the data area (SRAM) starts from address
0x2000 0000 (accessed through the system bus). The Cortex®-M4 with FPU CPU always
fetches the reset vector on the ICode bus, which implies to have the boot space available
only in the code area (typically, Flash memory). STM32F4xx microcontrollers implement a
special mechanism to be able to boot from other memories (like the internal SRAM).
In the STM32F4xx, three different boot modes can be selected through the BOOT[1:0] pins
as shown in Table 2.
Table 2. Boot modes

Boot mode selection pins
Boot mode Aliasing
BOOT1 BOOT0
x 0 Main Flash memory Main Flash memory is selected as the boot space
0 1 System memory System memory is selected as the boot space
1 1 Embedded SRAM Embedded SRAM is selected as the boot space
RM0090 Rev 19 69/1751

112
The values on the BOOT pins are latched on the 4th rising edge of SYSCLK after a reset. It
is up to the user to set the BOOT1 and BOOT0 pins after reset to select the required boot
mode.
BOOT0 is a dedicated pin while BOOT1 is shared with a GPIO pin. Once BOOT1 has been
sampled, the corresponding GPIO pin is free and can be used for other purposes.
The BOOT pins are also resampled when the device exits the Standby mode. Consequently,
they must be kept in the required Boot mode configuration when the device is in the Standby
mode. After this startup delay is over, the CPU fetches the top-of-stack value from address
0x0000 0000, then starts code execution from the boot memory starting from 0x0000 0004.
Note: When the device boots from SRAM, in the application initialization code, you have to
relocate the vector table in SRAM using the NVIC exception table and the offset register.
In STM32F42xxx and STM32F43xxx devices, when booting from the main Flash memory,
the application software can either boot from bank 1 or from bank 2. By default, boot from
bank 1 is selected.
To select boot from Flash memory bank 2, set the BFB2 bit in the user option bytes. When
this bit is set and the boot pins are in the boot from main Flash memory configuration, the
device boots from system memory, and the boot loader jumps to execute the user
application programmed in Flash memory bank 2. For further details, please refer to
AN2606.
Embedded bootloader
The embedded bootloader mode is used to reprogram the Flash memory using one of the
following serial interfaces:
• USART1 (PA9/PA10)
• USART3 (PB10/11 and PC10/11)
• CAN2 (PB5/13)
• USB OTG FS (PA11/12) in Device mode (DFU: device firmware upgrade).
The USART peripherals operate at the internal 16 MHz oscillator (HSI) frequency, while the
CAN and USB OTG FS require an external clock (HSE) multiple of 1 MHz (ranging from 4 to
26 MHz).
The embedded bootloader code is located in system memory. It is programmed by ST
during production. For additional information, refer to application note AN2606.
Physical remap in STM32F405xx/07xx and STM32F415xx/17xx

Once the boot pins are selected, the application software can modify the memory
accessible in the code area (in this way the code can be executed through the ICode bus in
place of the System bus). This modification is performed by programming the Section 9.2.1:
SYSCFG memory remap register (SYSCFG_MEMRMP) in the SYSCFG controller.
The following memories can thus be remapped:
• Main Flash memory
• System memory
• Embedded SRAM1 (112 KB)
• FSMC bank 1 (NOR/PSRAM 1 and 2)
70/1751 RM0090 Rev 19


in STM32F405xx/07xx and STM32F415xx/17xx
Boot/Remap in Boot/Remap in Boot/Remap in
Addresses Remap in FSMC
main Flash memory embedded SRAM System memory
0x2001 C000 - 0x2001 FFFF SRAM2 (16 KB) SRAM2 (16 KB) SRAM2 (16 KB) SRAM2 (16 KB)
0x2000 0000 - 0x2001 BFFF SRAM1 (112 KB) SRAM1 (112 KB) SRAM1 (112 KB) SRAM1 (112 KB)
0x1FFF 0000 - 0x1FFF 77FF System memory System memory System memory System memory
0x0810 0000 - 0x0FFF FFFF Reserved Reserved Reserved Reserved
0x0800 0000 - 0x080F FFFF Flash memory Flash memory Flash memory Flash memory
FSMC bank 1
0x0400 0000 - 0x07FF FFFF Reserved Reserved Reserved NOR/PSRAM 2
(128 MB Aliased)
FSMC bank 1
0x0000 0000 - SRAM1 (112 KB) System memory
Flash (1 MB) Aliased NOR/PSRAM 1
0x000F FFFF(1)(2) Aliased (30 KB) Aliased
(128 MB Aliased)
1. When the FSMC is remapped at address 0x0000 0000, only the first two regions of bank 1 memory controller (bank 1
NOR/PSRAM 1 and NOR/PSRAM 2) can be remapped. In remap mode, the CPU can access the external memory via
ICode bus instead of System bus which boosts up the performance.
2. Even when aliased in the boot memory space, the related memory is still accessible at its original memory space.
Physical remap in STM32F42xxx and STM32F43xxx

Once the boot pins are selected, the application software can modify the memory
accessible in the code area (in this way the code can be executed through the ICode bus in
place of the System bus). This modification is performed by programming the Section 9.2.1:
SYSCFG memory remap register (SYSCFG_MEMRMP) in the SYSCFG controller.
The following memories can thus be remapped:
• Main Flash memory
• System memory
• Embedded SRAM1 (112 KB)
• FMC bank 1 (NOR/PSRAM 1 and 2)
• FMC SDRAM bank 1

in STM32F42xxx and STM32F43xxx
Addresses Remap in FMC
0x2002 0000 - 0x2002 FFFF SRAM3 (64 KB) SRAM3 (64 KB) SRAM3 (64 KB) SRAM3 (64 KB)
0x2001 C000 - 0x2001 FFFF SRAM2 (16 KB) SRAM2 (16 KB) SRAM2 (16 KB) SRAM2 (16 KB)
0x2000 0000 - 0x2001 BFFF SRAM1 (112 KB) SRAM1 (112 KB) SRAM1 (112 KB) SRAM1 (112 KB)
0x1FFF 0000 - 0x1FFF 77FF System memory System memory System memory System memory
0x0810 0000 - 0x0FFF FFFF Reserved Reserved Reserved Reserved
0x0800 0000 - 0x081F FFFF Flash memory Flash memory Flash memory Flash memory
RM0090 Rev 19 71/1751

112

in STM32F42xxx and STM32F43xxx (continued)
Addresses Remap in FMC
FMC bank 1
0x0400 0000 - 0x07FF FFFF Reserved Reserved Reserved NOR/PSRAM 2
(128 MB Aliased)
FMC bank 1
NOR/PSRAM 1
0x0000 0000 - SRAM1 (112 KB) System memory (128 MB Aliased)
Flash (2 MB) Aliased
0x001F FFFF(1)(2) Aliased (30 KB) Aliased or FMC SDRAM
bank 1 (128 MB
Aliased)
1. When the FMC is remapped at address 0x0000 0000, only the first two regions of bank 1 memory controller (bank 1
NOR/PSRAM 1 and NOR/PSRAM 2) or SDRAM bank 1 can be remapped. In remap mode, the CPU can access the
external memory via ICode bus instead of System bus which boosts up the performance.
2. Even when aliased in the boot memory space, the related memory is still accessible at its original memory space.
72/1751 RM0090 Rev 19

RM0090 Embedded Flash memory interface
3 Embedded Flash memory interface
3.1 Introduction
The Flash memory interface manages CPU AHB I-Code and D-Code accesses to the Flash
memory. It implements the erase and program Flash memory operations and the read and
write protection mechanisms.
The Flash memory interface accelerates code execution with a system of instruction
prefetch and cache lines.
3.2 Main features

• Flash memory read operations
• Flash memory program/erase operations
• Read / write protections
• Prefetch on I-Code
• 64 cache lines of 128 bits on I-Code
• 8 cache lines of 128 bits on D-Code
Figure 3 shows the Flash memory interface connection inside the system architecture.

(STM32F405xx/07xx and STM32F415xx/17xx)
AHB Flash
Cortex-M4 with FPU 32-bit memory
instruction Flash interface bus
I-Code bus I-Code bus 128 bits
Flash
Cortex D-Code
memory
core
S bus AHB
D-code bus 32-bit FLITF registers
data bus
CCM data AHB

AHB periph1
RAM
32-bit
DMA1 system bus SRAM and
External
DMA2 memories
AHB
USB HS periph2
Ethernet
Access to instruction in Flash memory

Access to data and literal pool in Flash memory
FLITF register access
MS30468V3
RM0090 Rev 19 73/1751

112
Embedded Flash memory interface RM0090

(STM32F42xxx and STM32F43xxx)
AHB Flash
Cortex-M4 with FPU 32-bit memory
instruction Flash interface bus
I-Code bus I-Code bus 128 bits
Flash
Cortex D-Code
memory
core
S bus AHB
D-code bus 32-bit FLITF registers
data bus
AHB
CCM data periph1
AHB
RAM
32-bit
DMA1 system bus SRAM and
external
DMA2
memories
DMA2D AHB
periph2
LCD-TFT
USB HS
Ethernet
Access to instruction in Flash memory

Access to data and literal pool in Flash memory MS30466V3
FLITF register access
3.3 Embedded Flash memory in

STM32F405xx/07xx and STM32F415xx/17xx
The Flash memory has the following main features:
• Capacity up to 1 Mbyte
• 128 bits wide data read
• Byte, half-word, word and double word write
• Sector and mass erase
• Memory organization
– A main memory block divided into 4 sectors of 16 Kbytes, 1 sector of 64 Kbytes,
and 7 sectors of 128 Kbytes
– System memory from which the device boots in System memory boot mode
– 512 OTP (one-time programmable) bytes for user data
The OTP area contains 16 additional bytes used to lock the corresponding OTP
data block.
– Option bytes to configure read and write protection, BOR level, watchdog
software/hardware and reset when the device is in Standby or Stop mode.
• Low-power modes (for details refer to the Power control (PWR) section of the reference
manual)
74/1751 RM0090 Rev 19

Table 5. Flash module organization (STM32F40x and STM32F41x)

Block Name Block base addresses Size
Sector 0 0x0800 0000 - 0x0800 3FFF 16 Kbytes

Sector 2 0x0800 8000 - 0x0800 BFFF 16 Kbytes
Sector 3 0x0800 C000 - 0x0800 FFFF 16 Kbytes
Sector 4 0x0801 0000 - 0x0801 FFFF 64 Kbytes
Main memory
. . .
. . .
. . .
Sector 11 0x080E 0000 - 0x080F FFFF 128 Kbytes
System memory 0x1FFF 0000 - 0x1FFF 77FF 30 Kbytes
OTP area 0x1FFF 7800 - 0x1FFF 7A0F 528 bytes

Option bytes 0x1FFF C000 - 0x1FFF C00F 16 bytes
RM0090 Rev 19 75/1751

112
3.4 Embedded Flash memory in STM32F42xxx and

STM32F43xxx
The Flash memory has the following main features:
• Capacity up to 2 Mbyte with dual bank architecture supporting read-while-write
capability (RWW)
• 128 bits wide data read
• Byte, half-word, word and double word write
• Sector, bank, and mass erase (both banks)
• Dual bank memory organization
– For each bank, a main memory block (1 Mbyte) divided into 4 sectors of 16
Kbytes, 1 sector of 64 Kbytes, and 7 sectors of 128 Kbytes
– System memory from which the device boots in System memory boot mode
– 512 OTP (one-time programmable) bytes for user data
The OTP area contains 16 additional bytes used to lock the corresponding OTP
data block.
– Option bytes to configure read and write protection, BOR level, watchdog, dual
bank boot mode, dual bank feature, software/hardware and reset when the device
is in Standby or Stop mode.
• Dual bank organization on 1 Mbyte devices
The dual bank feature on 1 Mbyte devices is enabled by setting the DB1M option bit.
To obtain a dual bank Flash memory, the last 512 Kbytes of the single bank (sectors
[8:11]) are re-structured in the same way as the first 512 Kbytes.
The sector numbering of dual bank memory organization is different from the single
bank: the single bank memory contains 12 sectors whereas the dual bank memory
contains 16 sectors (see Table 7: 1 Mbyte Flash memory single bank vs dual bank
organization (STM32F42xxx and STM32F43xxx)).
For erase operation, the right sector numbering must be considered according the
DB1M option bit.
– When the DB1M bit is reset, the erase operation must be performed on the default
sector number.
– When the DB1M bit is set, to perform an erase operation on bank 2, the sector
number must be programmed (sector number from 12 to 19). Refer to FLASH_CR
register for SNB (Sector number) configuration.
Refer to Table 8: 1 Mbyte single bank Flash memory organization (STM32F42xxx and
STM32F43xxx) and Table 9: 1 Mbyte dual bank Flash memory organization (STM32F42xxx
and STM32F43xxx) for details on 1 Mbyte single bank and 1 Mbyte dual bank
organizations.
76/1751 RM0090 Rev 19

Table 6. Flash module - 2 Mbyte dual bank organization (STM32F42xxx and STM32F43xxx)
Block Bank Name Block base addresses Size

Sector 3 0x0800 C000 - 0x0800 FFFF 16 Kbyte
Bank 1 Sector 5 0x0802 0000 - 0x0803 FFFF 128 Kbytes
- - -
- - -
- - -
Main memory
Bank 2 Sector 17 0x0812 0000 - 0x0813 FFFF 128 Kbytes
- -
- -
- -
OTP 0x1FFF 7800 - 0x1FFF 7A0F 528 bytes
Bank 1 0x1FFF C000 - 0x1FFF C00F 16 bytes
Option bytes
Bank 2 0x1FFE C000 - 0x1FFE C00F 16 bytes
RM0090 Rev 19 77/1751

112
Table 7. 1 Mbyte Flash memory single bank vs dual bank organization

1 Mbyte single bank Flash memory (default) 1 Mbyte dual bank Flash memory
DB1M=0 DB1M=1
Main memory Sector number Sector size Main memory Sector number Sector size
Sector 0 16 Kbytes Sector 0 16 Kbytes
Sector 3 16 Kbytes Bank 1 Sector 3 16 Kbytes
Sector 4 64 Kbytes 512KB Sector 4 64 Kbytes
1MB
Sector 11 128 Kbytes Bank 2 Sector 15 16 Kbytes
- - 512KB Sector 16 64 Kbytes
- - Sector 17 128 Kbytes


Main memory Single bank
Sector 9 0x080A 0000 - 0x080B FFFF 128 Kbytes
Sector 10 0x080C 0000 - 0x080D FFFF 128 Kbytes
78/1751 RM0090 Rev 19


(STM32F42xxx and STM32F43xxx) (continued)
System memory 0x1FFF 0000 - 0x1FFFF 77FF 30 Kbytes

0x1FFF C000 - 0x1FFF C00F 16 bytes
Option bytes
0x1FFE C000 - 0x1FFE C00F 16 bytes
Table 9. 1 Mbyte dual bank Flash memory organization (STM32F42xxx and STM32F43xxx)
Block Name Block base addresses Size

Sector 3 0x0800 C000 - 0x0800 FFFF 16 Kbyte
Bank 1
Main memory
Bank 2
Sector 17 0x080A 0000 - 0x080B FFFF 128 Kbytes
Sector 18 0x080C 0000 - 0x080D FFFF 128 Kbytes
Bank 1 0x1FFF C000 - 0x1FFF C00F 16 bytes
Option bytes
Bank 2 0x1FFE C000 - 0x1FFE C00F 16 bytes
RM0090 Rev 19 79/1751

112
3.5 Read interface
3.5.1 Relation between CPU clock frequency and Flash memory read time
To correctly read data from Flash memory, the number of wait states (LATENCY) must be
correctly programmed in the Flash access control register (FLASH_ACR) according to the
frequency of the CPU clock (HCLK) and the supply voltage of the device.
The prefetch buffer must be disabled when the supply voltage is below 2.1 V. The
correspondence between wait states and CPU clock frequency is given in Table 10 and
Table 11.
Note: On STM32F405xx/07xx and STM32F415xx/17xx devices:
- when VOS = '0', the maximum value of fHCLK = 144 MHz.
- when VOS = '1', the maximum value of fHCLK = 168 MHz.
On STM32F42xxx and STM32F43xxx devices:
- when VOS[1:0] = '0x01', the maximum value of fHCLK is 120 MHz.
- when VOS[1:0] = '0x10', the maximum value of fHCLK is 144 MHz. It can be extended to
168 MHz by activating the over-drive mode.
- when VOS[1:0] = '0x11, the maximum value of fHCLK is 168 MHz. It can be extended to
180 MHz by activating the over-drive mode.
- The over-drive mode is not available when VDD ranges from 1.8 to 2.1 V.
Refer to Section 5.1.4: Voltage regulator for STM32F42xxx and STM32F43xxx for details on
how to activate the over-drive mode.
HCLK (MHz)
Wait states (WS) Voltage range
(LATENCY) Voltage range Voltage range Voltage range
1.8 V - 2.1 V
2.7 V - 3.6 V 2.4 V - 2.7 V 2.1 V - 2.4 V
Prefetch OFF
0 WS (1 CPU cycle) 0 < HCLK ≤ 30 0 < HCLK ≤ 24 0 < HCLK ≤ 22 0 < HCLK ≤ 20
1 WS (2 CPU cycles) 30 < HCLK ≤ 60 24 < HCLK ≤ 48 22 < HCLK ≤ 44 20 <HCLK ≤ 40
2 WS (3 CPU cycles) 60 < HCLK ≤ 90 48 < HCLK ≤ 72 44 < HCLK ≤ 66 40 < HCLK ≤ 60
6 WS (7 CPU cycles) 144 < HCLK ≤ 168 132 < HCLK ≤ 154 120 < HCLK ≤ 140
7 WS (8 CPU cycles) 154 < HCLK ≤ 168 140 < HCLK ≤ 160
80/1751 RM0090 Rev 19

HCLK (MHz)
Wait states (WS) Voltage range
(LATENCY) Voltage range Voltage range Voltage range
1.8 V - 2.1 V
2.7 V - 3.6 V 2.4 V - 2.7 V 2.1 V - 2.4 V
Prefetch OFF
0 WS (1 CPU cycle) 0 <HCLK ≤ 30 0 <HCLK ≤ 24 0 <HCLK ≤ 22 0 < HCLK ≤ 20

1 WS (2 CPU cycles) 30 <HCLK ≤ 60 24 < HCLK ≤ 48 22 <HCLK ≤ 44 20 <HCLK ≤ 40
2 WS (3 CPU cycles) 60 <HCLK ≤ 90 48 < HCLK ≤ 72 44 < HCLK ≤ 66 40 < HCLK ≤ 60
3 WS (4 CPU cycles) 90 <HCLK ≤ 120 72 < HCLK ≤ 96 66 <HCLK ≤ 88 60 < HCLK ≤ 80
4 WS (5 CPU cycles) 120 <HCLK ≤ 150 96 < HCLK ≤ 120 88 < HCLK ≤ 110 80 < HCLK ≤ 100
5 WS (6 CPU cycles) 150 <HCLK ≤ 180 120 <HCLK ≤ 144 110 < HCLK ≤ 132 100 < HCLK ≤ 120
6 WS (7 CPU cycles) 144 <HCLK ≤ 168 132 < HCLK ≤ 154 120 < HCLK ≤ 140
7 WS (8 CPU cycles) 168 <HCLK ≤ 180 154 <HCLK ≤ 176 140 < HCLK ≤ 160
8 WS (9 CPU cycles) 176 <HCLK ≤ 180 160 < HCLK ≤ 168
After reset, the CPU clock frequency is 16 MHz and 0 wait state (WS) is configured in the
FLASH_ACR register.
It is highly recommended to use the following software sequences to tune the number of
wait states needed to access the Flash memory with the CPU frequency.
Increasing the CPU frequency

1. Program the new number of wait states to the LATENCY bits in the FLASH_ACR
register
2. Check that the new number of wait states is taken into account to access the Flash
memory by reading the FLASH_ACR register
3. Modify the CPU clock source by writing the SW bits in the RCC_CFGR register
4. If needed, modify the CPU clock prescaler by writing the HPRE bits in RCC_CFGR
5. Check that the new CPU clock source or/and the new CPU clock prescaler value is/are
taken into account by reading the clock source status (SWS bits) or/and the AHB
prescaler value (HPRE bits), respectively, in the RCC_CFGR register.
Decreasing the CPU frequency

1. Modify the CPU clock source by writing the SW bits in the RCC_CFGR register
2. If needed, modify the CPU clock prescaler by writing the HPRE bits in RCC_CFGR
3. Check that the new CPU clock source or/and the new CPU clock prescaler value is/are
taken into account by reading the clock source status (SWS bits) or/and the AHB
prescaler value (HPRE bits), respectively, in the RCC_CFGR register
4. Program the new number of wait states to the LATENCY bits in FLASH_ACR
5. Check that the new number of wait states is used to access the Flash memory by
reading the FLASH_ACR register
RM0090 Rev 19 81/1751

112
Note: A change in CPU clock configuration or wait state (WS) configuration may not be effective
straight away. To make sure that the current CPU clock frequency is the one you have
configured, you can check the AHB prescaler factor and clock source status values. To
make sure that the number of WS you have programmed is effective, you can read the
FLASH_ACR register.
3.5.2 Adaptive real-time memory accelerator (ART Accelerator™)

The proprietary Adaptive real-time (ART) memory accelerator is optimized for STM32
industry-standard Arm® Cortex®-M4 with FPU processors. It balances the inherent
performance advantage of the Arm® Cortex®-M4 with FPU over Flash memory
technologies, which normally requires the processor to wait for the Flash memory at higher
operating frequencies.
To release the processor full performance, the accelerator implements an instruction
prefetch queue and branch cache which increases program execution speed from the 128-
bit Flash memory. Based on CoreMark benchmark, the performance achieved thanks to the
ART accelerator is equivalent to 0 wait state program execution from Flash memory at a
CPU frequency up to 180 MHz.
Instruction prefetch
Each Flash memory read operation provides 128 bits from either four instructions of 32 bits
or 8 instructions of 16 bits according to the program launched. So, in case of sequential
code, at least four CPU cycles are needed to execute the previous read instruction line.
Prefetch on the I-Code bus can be used to read the next sequential instruction line from the
Flash memory while the current instruction line is being requested by the CPU. Prefetch is
enabled by setting the PRFTEN bit in the FLASH_ACR register. This feature is useful if at
least one wait state is needed to access the Flash memory.
Figure 5 shows the execution of sequential 32-bit instructions with and without prefetch
when 3 WSs are needed to access the Flash memory.
82/1751 RM0090 Rev 19

Figure 5. Sequential 32-bit instruction execution
@ F D E
WAIT
1 1 1 1
Without prefetch
@ F D E
2 2 2 2
@ F D E
3 3 3 3
@ F D E
4 4 4 4
@ F D E
WAIT
5 5 5 5
@ F D E
6 6 6 6
@ F D
7 7 7
@ F
8 8
ins 1 ins 2 ins 3 ins 4 ins 5 ins 6 ins 7 ins 8

fetch fetch fetch fetch fetch fetch fetch fetch
Read ins 1, 2, 3, 4 Gives ins 1, 2, 3, 4 Read ins 5, 6, 7, 8 Gives ins 5, 6, 7, 8
@ Wait data F D E
1 1 1 1 With prefetch
@ F D E
2 2 2 2
@ F D E
3 3 3 3
@ F D E
4 4 4 4
@ F D E
5 5 5 5
@ F D E
6 6 6
@ F D Cortex-M4 pipeline
7 7 7
@ F
@ F D E
8 8
AHB protocol
ins 1 ins 2 ins 3 ins 4 ins 5 ins 6 ins 7 ins 8
fetch fetch fetch fetch fetch fetch fetch fetch @ : address requested
F: Fetch stage
D: Decode stage
Read ins 1, 2, 3, 4 Gives ins 1, 2, 3, 4 Gives ins 5, 6, 7, 8 E: Execute stage
Read ins 5, 6, 7, 8 Read ins 9, 10, ...
MS31831V1
When the code is not sequential (branch), the instruction may not be present in the currently
used instruction line or in the prefetched instruction line. In this case (miss), the penalty in
terms of number of cycles is at least equal to the number of wait states.
RM0090 Rev 19 83/1751

112
Instruction cache memory

To limit the time lost due to jumps, it is possible to retain 64 lines of 128 bits in an instruction
cache memory. This feature can be enabled by setting the instruction cache enable (ICEN)
bit in the FLASH_ACR register. Each time a miss occurs (requested data not present in the
currently used instruction line, in the prefetched instruction line or in the instruction cache
memory), the line read is copied into the instruction cache memory. If some data contained
in the instruction cache memory are requested by the CPU, they are provided without
inserting any delay. Once all the instruction cache memory lines have been filled, the LRU
(least recently used) policy is used to determine the line to replace in the instruction memory
cache. This feature is particularly useful in case of code containing loops.
Data management
Literal pools are fetched from Flash memory through the D-Code bus during the execution
stage of the CPU pipeline. The CPU pipeline is consequently stalled until the requested
literal pool is provided. To limit the time lost due to literal pools, accesses through the AHB
databus D-Code have priority over accesses through the AHB instruction bus I-Code.
If some literal pools are frequently used, the data cache memory can be enabled by setting
the data cache enable (DCEN) bit in the FLASH_ACR register. This feature works like the
instruction cache memory, but the retained data size is limited to 8 rows of 128 bits.
Note: Data in user configuration sector are not cacheable.
3.6 Erase and program operations

For any Flash memory program operation (erase or program), the CPU clock frequency
(HCLK) must be at least 1 MHz. The contents of the Flash memory are not guaranteed if a
device reset occurs during a Flash memory operation.
Any attempt to read the Flash memory on STM32F4xx while it is being written or erased,
causes the bus to stall. Read operations are processed correctly once the program
operation has completed. This means that code or data fetches cannot be performed while
a write/erase operation is ongoing.
On STM32F42xxx and STM32F43xxx devices, two banks are available allowing read
operation from one bank while a write/erase operation is performed to the other bank.
3.6.1 Unlocking the Flash control register

After reset, write is not allowed in the Flash control register (FLASH_CR) to protect the
Flash memory against possible unwanted operations due, for example, to electric
disturbances. The following sequence is used to unlock this register:
1. Write KEY1 = 0x45670123 in the Flash key register (FLASH_KEYR)
2. Write KEY2 = 0xCDEF89AB in the Flash key register (FLASH_KEYR)
Any wrong sequence will return a bus error and lock up the FLASH_CR register until the
next reset.
The FLASH_CR register can be locked again by software by setting the LOCK bit in the
FLASH_CR register.
84/1751 RM0090 Rev 19

Note: The FLASH_CR register is not accessible in write mode when the BSY bit in the FLASH_SR
register is set. Any attempt to write to it with the BSY bit set will cause the AHB bus to stall
until the BSY bit is cleared.
3.6.2 Program/erase parallelism

The Parallelism size is configured through the PSIZE field in the FLASH_CR register. It
represents the number of bytes to be programmed each time a write operation occurs to the
Flash memory. PSIZE is limited by the supply voltage and by whether the external VPP
supply is used or not. It must therefore be correctly configured in the FLASH_CR register
before any programming/erasing operation.
A Flash memory erase operation can only be performed by sector, bank or for the whole
Flash memory (mass erase). The erase time depends on PSIZE programmed value. For
more details on the erase time, refer to the electrical characteristics section of the device
datasheet.
Table 12 provides the correct PSIZE values.
Table 12. Program/erase parallelism

Voltage range 2.7 - 3.6 V Voltage range Voltage range Voltage range Voltage range
with External VPP 2.7 - 3.6 V 2.4 - 2.7 V 2.1 - 2.4 V 1.8 V - 2.1 V
Parallelism size x64 x32 x16 x8

PSIZE(1:0) 11 10 01 00
Note: Any program or erase operation started with inconsistent program parallelism/voltage range
settings may lead to unpredicted results. Even if a subsequent read operation indicates that
the logical value was effectively written to the memory, this value may not be retained.
To use VPP, an external high-voltage supply (between 8 and 9 V) must be applied to the VPP
pad. The external supply must be able to sustain this voltage range even if the DC
consumption exceeds 10 mA. It is advised to limit the use of VPP to initial programming on
the factory line. The VPP supply must not be applied for more than an hour, otherwise the
Flash memory might be damaged.
3.6.3 Erase
The Flash memory erase operation can be performed at sector level or on the whole Flash
memory (Mass Erase). Mass Erase does not affect the OTP sector or the configuration
sector.
Sector Erase
To erase a sector, follow the procedure below:
1. Check that no Flash memory operation is ongoing by checking the BSY bit in the
FLASH_SR register
2. Set the SER bit and select the sector out of the 12 sectors (for STM32F405xx/07xx and
STM32F415xx/17xx) and out of 24 (for STM32F42xxx and STM32F43xxx) in the main
memory block you wish to erase (SNB) in the FLASH_CR register
3. Set the STRT bit in the FLASH_CR register
4. Wait for the BSY bit to be cleared
RM0090 Rev 19 85/1751

112
Bank erase in STM32F42xxx and STM32F43xxx devices

To erase bank 1 or bank 2, follow the procedure below:
FLASH_SR register
2. Set MER or MER1 bit accordingly in the FLASH_CR register
4. Wait for the BSY bit to be reset.
Mass Erase
To perform Mass Erase, the following sequence is recommended:
FLASH_SR register
2. Set the MER bit in the FLASH_CR register (on STM32F405xx/07xx and
STM32F415xx/17xx devices)
3. Set both the MER and MER1 bits in the FLASH_CR register (on STM32F42xxx and
STM32F43xxx devices).
Note: If MERx and SER bits are both set in the FLASH_CR register, mass erase is performed.
If both MERx and SER bits are reset and the STRT bit is set, an unpredictable behavior may
occur without generating any error flag. This condition should be forbidden.
3.6.4 Programming
Standard programming
The Flash memory programming sequence is as follows:
1. Check that no main Flash memory operation is ongoing by checking the BSY bit in the
FLASH_SR register.
2. Set the PG bit in the FLASH_CR register
3. Perform the data write operation(s) to the desired memory address (inside main
memory block or OTP area):
– Byte access in case of x8 parallelism
– Half-word access in case of x16 parallelism
– Word access in case of x32 parallelism
– Double word access in case of x64 parallelism
4. Wait for the BSY bit to be cleared.
Note: Successive write operations are possible without the need of an erase operation when
changing bits from ‘1’ to ‘0’. Writing ‘1’ requires a Flash memory erase operation.
If an erase and a program operation are requested simultaneously, the erase operation is
performed first.
86/1751 RM0090 Rev 19

Programming errors
It is not allowed to program data to the Flash memory that would cross the 128-bit row
boundary. In such a case, the write operation is not performed and a program alignment
error flag (PGAERR) is set in the FLASH_SR register.
The write access type (byte, half-word, word or double word) must correspond to the type of
parallelism chosen (x8, x16, x32 or x64). If not, the write operation is not performed and a
program parallelism error flag (PGPERR) is set in the FLASH_SR register.
If the standard programming sequence is not respected (for example, if there is an attempt
to write to a Flash memory address when the PG bit is not set), the operation is aborted and
a program sequence error flag (PGSERR) is set in the FLASH_SR register.
Programming and caches

If a Flash memory write access concerns some data in the data cache, the Flash write
access modifies the data in the Flash memory and the data in the cache.
If an erase operation in Flash memory also concerns data in the data or instruction cache,
you have to make sure that these data are rewritten before they are accessed during code
execution. If this cannot be done safely, it is recommended to flush the caches by setting the
DCRST and ICRST bits in the FLASH_CR register.
Note: The I/D cache should be flushed only when it is disabled (I/DCEN = 0).
3.6.5 Read-while-write (RWW)

In STM32F42xxx and STM32F43xxx devices, the Flash memory is divided into two banks
allowing read-while-write operations. This feature allows to perform a read operation from
one bank while an erase or program operation is performed to the other bank.
Note: Write-while-write operations are not allowed. As an exampled, It is not possible to perform
an erase operation on one bank while programming the other one.
Read from bank 1 while erasing bank 2

While executing a program code from bank 1, it is possible to perform an erase operation on
bank 2 (and vice versa). Follow the procedure below:
FLASH_SR register (BSY is active when erase/program operation is on going to
bank 1 or bank 2)
2. Set MER or MER1 bit in the FLASH_CR register
4. Wait for the BSY bit to be reset (or use the EOP interrupt).
Read from bank 1 while programming bank 2

While executing a program code (over the I-Code bus) from bank 1,it is possible to perform
an program operation to the bank 2 (and vice versa). Follow the procedure below:
RM0090 Rev 19 87/1751

112
FLASH_SR register (BSY is active when erase/program operation is on going on bank
1 or bank 2)
2. Set the PG bit in the FLASH_CR register
3. Perform the data write operation(s) to the desired memory address inside main
memory block or OTP area
4. Wait for the BSY bit to be reset.
3.6.6 Interrupts
Setting the end of operation interrupt enable bit (EOPIE) in the FLASH_CR register enables
interrupt generation when an erase or program operation ends, that is when the busy bit
(BSY) in the FLASH_SR register is cleared (operation completed, correctly or not). In this
case, the end of operation (EOP) bit in the FLASH_SR register is set.
If an error occurs during a program, an erase, or a read operation request, one of the
following error flags is set in the FLASH_SR register:
• PGAERR, PGPERR, PGSERR (Program error flags)
• WRPERR (Protection error flag)
• RDERR (Read protection error flag) for STM32F42xxx and STM32F43xxx devices
only.
In this case, if the error interrupt enable bit (ERRIE) is set in the FLASH_CR register, an
interrupt is generated and the operation error bit (OPERR) is set in the FLASH_SR register.
Note: If several successive errors are detected (for example, in case of DMA transfer to the Flash
memory), the error flags cannot be cleared until the end of the successive write requests.
Table 13. Flash interrupt request

Interrupt event Event flag Enable control bit
End of operation EOP EOPIE

Write protection error WRPERR ERRIE
Programming error PGAERR, PGPERR, PGSERR ERRIE
Read protection error RDERR ERRIE
3.7 Option bytes
3.7.1 Description of user option bytes

The option bytes are configured by the end user depending on the application requirements.
Table 14 shows the organization of these bytes inside the user configuration sector.
Table 14. Option byte organization

Address [63:16] [15:0]
0x1FFF C000 Reserved ROP & user option bytes (RDP & USER)
SPRMOD and Write protection nWRP bits for
0x1FFF C008 Reserved
sectors 0 to 11
88/1751 RM0090 Rev 19

Table 14. Option byte organization (continued)

Address [63:16] [15:0]
0x1FFE C000 Reserved Reserved

SPRMOD and Write protection nWRP bits for
0x1FFE C008 Reserved
sectors 12 to 23
Table 15. Description of the option bytes (STM32F405xx/07xx and

STM32F415xx/17xx)
Option bytes (word, address 0x1FFF C000)
RDP: Read protection option byte.
The read protection is used to protect the software code stored in Flash memory.
0xAA: Level 0, no protection
Bits 15:8 0xCC: Level 2, chip protection (debug and boot from RAM features disabled)
Others: Level 1, read protection of memories (debug features limited)
USER: User option byte
This byte is used to configure the following features:
– Select the watchdog event: Hardware or software
– Reset event when entering the Stop mode
– Reset event when entering the Standby mode
nRST_STDBY
Bit 7 0: Reset generated when entering the Standby mode
1: No reset generated
nRST_STOP
Bit 6 0: Reset generated when entering the Stop mode
WDG_SW
Bit 5 0: Hardware independent watchdog
1: Software independent watchdog
Bit 4 0x1: Not used
BOR_LEV: BOR reset Level
These bits contain the supply level threshold that activates/releases the reset.
They can be written to program a new BOR level value into Flash memory.
00: BOR Level 3 (VBOR3), brownout threshold level 3
Bits 3:2 01: BOR Level 2 (VBOR2), brownout threshold level 2
11: BOR off, POR/PDR reset threshold level is applied
Note: For full details on BOR characteristics, refer to the “Electrical characteristics”
section of the product datasheet.
Bits 1:0 0x3: Not used
RM0090 Rev 19 89/1751

112
Table 15. Description of the option bytes (STM32F405xx/07xx and

STM32F415xx/17xx) (continued)
Bits 15:12 0xF: Not used
nWRP: Flash memory write protection option bytes
Sectors 0 to 11 can be write protected.
nWRPi
Bits 11:0 0: Write protection active on selected sector
1: Write protection not active on selected sector

RDP: Read protection option byte.

The read protection is used to protect the software code stored in Flash memory.
0xAA: Level 0, no protection
Bit 15:8 0xCC: Level 2, chip protection (debug and boot from RAM features
disabled)
Others: Level 1, read protection of memories (debug features limited)
USER: User option byte
This byte is used to configure the following features:
Select the watchdog event: Hardware or software
Reset event when entering the Stop mode
Reset event when entering the Standby mode
nRST_STDBY
Bit 7 0: Reset generated when entering the Standby mode
nRST_STOP
Bit 6 0: Reset generated when entering the Stop mode
WDG_SW
Bit 5 0: Hardware independent watchdog
1: Software independent watchdog
BFB2: Dual bank boot
0: Boot from Flash memory bank 1 or system memory depending on boot pin
Bit 4
state (Default).
1: Boot always from system memory (Dual bank boot mode).
90/1751 RM0090 Rev 19


BOR_LEV: BOR reset Level
These bits contain the supply level threshold that activates/releases the reset.
They can be written to program a new BOR level value into Flash memory.
Bits 3:2 01: BOR Level 2 (VBOR2), brownout threshold level 2
Note: For full details on BOR characteristics, refer to the “Electrical
characteristics” section of the product datasheet.
Bits 1:0 0x3: Not used
SPRMOD: Selection of protection mode of nWPRi bits
Bit 15 0: nWPRi bits used for sector i write protection (Default)
1: nWPRi bits used for sector i PCROP protection (Sector)
DB1M: Dual bank 1 Mbyte Flash memory devices
0: 1 Mbyte single Flash memory (contiguous addresses in bank 1)
1: 1 Mbyte dual bank Flash memory. The Flash memory is organized as two banks
Bit 14 of 512 Kbytes each (see Table 7: 1 Mbyte Flash memory single bank vs dual bank
organization (STM32F42xxx and STM32F43xxx) and Table 9: 1 Mbyte dual bank
Flash memory organization (STM32F42xxx and STM32F43xxx)). To perform an
erase operation, the right sector must be programmed (see Table 7 for information
on the sector numbering scheme).
Bits 13:12 0x3: not used
nWRP: Flash memory write protection option bytes for bank 1. Sectors 0 to 11 can be write
protected.
nWRPi:
If SPRMOD is reset (default value):
0: Write protection active on sector i.
Bits 11:0 1: Write protection not active on sector i.
If SPRMOD is set (active):
0: PCROP protection not active on sector i.
1: PCROP protection active on sector i.
Option bytes (word, address 0x1FFE C000)
Bit 15:0 0xFFFF: not used
RM0090 Rev 19 91/1751

112

Option bytes (word, address 0x1FFE C008)
Bit 15:12 0xF: not used
nWRP: Flash memory write protection option bytes for bank 2. Sectors 12 to 23 can be write
protected.
nWRPi:
Bits 11: 0 1: Write protection not active on sector i.
If SPRMOD is set (active):
3.7.2 Programming user option bytes

To run any operation on this sector, the option lock bit (OPTLOCK) in the Flash option
control register (FLASH_OPTCR) must be cleared. To be allowed to clear this bit, you have
to perform the following sequence:
1. Write OPTKEY1 = 0x0819 2A3B in the Flash option key register (FLASH_OPTKEYR)
2. Write OPTKEY2 = 0x4C5D 6E7F in the Flash option key register (FLASH_OPTKEYR)
The user option bytes can be protected against unwanted erase/program operations by
setting the OPTLOCK bit by software.
Modifying user option bytes on STM32F405xx/07xx and STM32F415xx/17xx

To modify the user option value, follow the sequence below:
FLASH_SR register
2. Write the desired option value in the FLASH_OPTCR register.
3. Set the option start bit (OPTSTRT) in the FLASH_OPTCR register
4. Wait for the BSY bit to be cleared.
Note: The value of an option is automatically modified by first erasing the user configuration sector
and then programming all the option bytes with the values contained in the FLASH_OPTCR
register.
Modifying user option bytes on STM32F42xxx and STM32F43xxx

The user option bytes for bank 1 and bank 2 cannot be modified independently. They must
be updated concurrently.
To modify the user option byte value, follow the sequence below:
FLASH_SR register
2. Write the bank 2 option byte value in the FLASH_OPTCR1 register
3. Write the bank 1 option byte value in the FLASH_OPTCR register.
4. Set the option start bit (OPTSTRT) in the FLASH_OPTCR register
92/1751 RM0090 Rev 19

Note: The value of an option byte is automatically modified by first erasing the user configuration
sector (bank 1 and 2) and then programming all the option bytes with the values contained
in the FLASH_OPTCR and FLASH_OPTCR1 registers.
3.7.3 Read protection (RDP)

The user area in the Flash memory can be protected against read operations by an
entrusted code. Three read protection levels are defined:
• Level 0: no read protection
When the read protection level is set to Level 0 by writing 0xAA into the read protection
option byte (RDP), all read/write operations (if no write protection is set) from/to the
Flash memory or the backup SRAM are possible in all boot configurations (Flash user
boot, debug or boot from RAM).
• Level 1: read protection enabled
It is the default read protection level after option byte erase. The read protection Level
1 is activated by writing any value (except for 0xAA and 0xCC used to set Level 0 and
Level 2, respectively) into the RDP option byte. When the read protection Level 1 is set:
– No access (read, erase, program) to Flash memory or backup SRAM can be
performed while the debug feature is connected or while booting from RAM or
system memory bootloader. A bus error is generated in case of read request.
– When booting from Flash memory, accesses (read, erase, program) to Flash
memory and backup SRAM from user code are allowed.
When Level 1 is active, programming the protection option byte (RDP) to Level 0
causes the Flash memory and the backup SRAM to be mass-erased. As a result the
user code area is cleared before the read protection is removed. The mass erase only
erases the user code area. The other option bytes including write protections remain
unchanged from before the mass-erase operation. The OTP area is not affected by
mass erase and remains unchanged. Mass erase is performed only when Level 1 is
active and Level 0 requested. When the protection level is increased (0->1, 1->2, 0->2)
there is no mass erase.
• Level 2: debug/chip read protection disabled
The read protection Level 2 is activated by writing 0xCC to the RDP option byte. When
the read protection Level 2 is set:
– All protections provided by Level 1 are active.
– Booting from RAM or system memory bootloader is no more allowed.
– JTAG, SWV (single-wire viewer), ETM, and boundary scan are disabled.
– User option bytes can no longer be changed.
– When booting from Flash memory, accesses (read, erase and program) to Flash
memory and backup SRAM from user code are allowed.
Memory read protection Level 2 is an irreversible operation. When Level 2 is activated,
the level of protection cannot be decreased to Level 0 or Level 1.
Note: The JTAG port is permanently disabled when Level 2 is active (acting as a JTAG fuse). As a
consequence, boundary scan cannot be performed. STMicroelectronics is not able to
perform analysis on defective parts on which the Level 2 protection has been set.
RM0090 Rev 19 93/1751

112
--
Table 17. Access versus read protection level

Debug features, Boot from RAM or
Booting from Flash memory
Protection from System memory bootloader
Memory area
Level
Read Write Erase Read Write Erase
Main Flash Memory Level 1 NO NO(1) YES

and Backup SRAM
Level 2 NO YES
Level 1 YES YES
Option Bytes
Level 2 NO NO
Level 1 NO NA YES NA
OTP
Level 2 NO NA YES NA
1. The main Flash memory and backup SRAM are only erased when the RDP changes from level 1 to 0. The OTP area
remains unchanged.
Figure 6 shows how to go from one RDP level to another.
Figure 6. RDP levels
RDP /= AAh & /= CCh

Others options modified
Level 1
RDP /= AAh
RDP /= CCh
Write options default Write options
including including
RDP = CCh Write options including RDP = AAh
RDP /= CCh & /= AAh
Level 2 Level 0
RDP = CCh Write options RDP = AAh
including
RDP = CCh
RDP = AAh
Options write (RDP level increase) includes Others option(s) modified
- Options erase
- New options program
Options write (RDP level decrease) includes Options write (RDP level identical) includes
- Mass erase - Options erase
- Options erase - New options program
- New options program
MS47541V1
3.7.4 Write protections

Up to 24 user sectors in Flash memory can be protected against unwanted write operations
due to loss of program counter contexts. When the non-write protection nWRPi bit
(0 ≤ i ≤ 11) in the FLASH_OPTCR or FLASH_OPTCR1 registers is low, the corresponding
94/1751 RM0090 Rev 19

sector cannot be erased or programmed. Consequently, a mass erase cannot be performed

if one of the sectors is write-protected.
If an erase/program operation to a write-protected part of the Flash memory is attempted
(sector protected by write protection bit, OTP part locked or part of the Flash memory that
can never be written like the ICP), the write protection error flag (WRPERR) is set in the
FLASH_SR register.
On STM32F42xxx and STM32F43xxx devices, when the PCROP mode is set, the active
level of nWRPi is high, and the corresponding sector i is write protected when nWRPi is
high. A PCROP sector is automatically write protected.
Note: When the memory read protection level is selected (RDP level = 1), it is not possible to
program or erase Flash memory sector i if the CPU debug features are connected (JTAG or
single wire) or boot code is being executed from RAM, even if nWRPi = 1.
Write protection error flag

If an erase/program operation to a write protected area of the Flash memory is performed,
the Write Protection Error flag (WRPERR) is set in the FLASH_SR register.
If an erase operation is requested, the WRPERR bit is set when:
• Mass, bank, sector erase are configured (MER or MER/MER1 and SER = 1)
• A sector erase is requested and the Sector Number SNB field is not valid
• A mass erase is requested while at least one of the user sector is write protected by
option bit (MER or MER/MER1 = 1 and nWRPi = 0 with 0 ≤ i ≤ 11 bits in the
FLASH_OPTCRx register
• A sector erase is requested on a write protected sector. (SER = 1, SNB = i and
nWRPi = 0 with 0 ≤ i ≤ 11 bits in the FLASH_OPTCRx register)
• The Flash memory is readout protected and an intrusion is detected.
If a program operation is requested, the WRPERR bit is set when:
• A write operation is performed on system memory or on the reserved part of the user
specific sector.
• A write operation is performed to the user configuration sector
• A write operation is performed on a sector write protected by option bit.
• A write operation is requested on an OTP area which is already locked
• The Flash memory is read protected and an intrusion is detected.
3.7.5 Proprietary code readout protection (PCROP)

The proprietary readout protection (PCROP) is available only on STM32F42xxx and
STM32F43xxx devices.
Flash memory user sectors (0 to 23) can be protected against D-bus read accesses by
using the proprietary readout protection (PCROP).
The PCROP protection is selected as follows, through the SPRMOD option bit in the
FLASH_OPTCR register:
• SPRMOD = 0: nWRPi control the write protection of respective user sectors
• SPRMOD = 1: nWRPi control the read and write protection (PCROP) of respective
user sectors.
RM0090 Rev 19 95/1751

112
When a sector is readout protected (PCROP mode activated), it can only be accessed for
code fetch through ICODE Bus on Flash interface:
• Any read access performed through the D-bus triggers a RDERR flag error.
• Any program/erase operation on a PCROPed sector triggers a WRPERR flag error.
Figure 7. PCROP levels
Write options
SPMOD = active
and valid nWRPi*
Level 1
RDP /= 0xAA
RDP /= 0xCC
Write options default Write options No restriction on
SPMOD = active SPMOD = active Write options
and valid nWRPi* and valid nWRPi*
Level 2 Level 0
RDP = 0xCC RDP = 0xAA
Write options
SPMOD = active
Write options
and valid nWRPi*
SPMOD = active
User option sector erase and valid nWRPi*
Program new options
Global mass erase

User option section erase
Program new options
* Valid nWRPi means that none of the nWRP bits set can be reset (transition from 1 to 0)
MS30467V1
The deactivation of the SPRMOD and/or the unprotection of PCROPed user sectors can
only occur when, at the same time, the RDP level changes from 1 to 0. If this condition is not
respected, the user option byte modification is cancelled and the write error WRPERR flag
is set. The modification of the users option bytes (BOR_LEV, RST_STDBY, ..) is allowed
since none of the active nWRPi bits is reset and SPRMOD is kept active.
Note: The active value of nWRPi bits is inverted when PCROP mode is active (SPRMOD =1).
If SPRMOD = 1 and nWRPi =1, then user sector i of bank 1, respectively bank 2 is
read/write protected (PCROP).
96/1751 RM0090 Rev 19

3.8 One-time programmable bytes

Table 18 shows the organization of the one-time programmable (OTP) part of the OTP area.
Table 18. OTP area organization

Block [128:96] [95:64] [63:32] [31:0] Address byte 0
OTP0 OTP0 OTP0 OTP0 0x1FFF 7800

0
1
. . .
. . .
. . .
OTP15 OTP15 OTP15 OTP15 0x1FFF 79E0
15
OTP15 OTP15 OTP15 OTP15 0x1FFF 79F0
LOCKB15 ... LOCKB11 ... LOCKB7 ... LOCKB3 ...
Lock block 0x1FFF 7A00
LOCKB12 LOCKB8 LOCKB4 LOCKB0
The OTP area is divided into 16 OTP data blocks of 32 bytes and one lock OTP block of 16
bytes. The OTP data and lock blocks cannot be erased. The lock block contains 16 bytes
LOCKBi (0 ≤ i ≤ 15) to lock the corresponding OTP data block (blocks 0 to 15). Each OTP
data block can be programmed until the value 0x00 is programmed in the corresponding
OTP lock byte. The lock bytes must only contain 0x00 and 0xFF values, otherwise the OTP
bytes might not be taken into account correctly.
RM0090 Rev 19 97/1751

112
3.9 Flash interface registers

for STM32F405xx/07xx and STM32F415xx/17xx
The Flash access control register is used to enable/disable the acceleration features and
control the Flash memory access time according to CPU frequency.
Address offset: 0x00
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DCRST ICRST DCEN ICEN PRFTEN LATENCY[2:0]
Reserved Reserved
rw w rw rw rw rw rw rw
Bits 31:13 Reserved, must be kept cleared.

Bit 12 DCRST: Data cache reset
0: Data cache is not reset
1: Data cache is reset
This bit can be written only when the D cache is disabled.
Bit 11 ICRST: Instruction cache reset
0: Instruction cache is not reset
1: Instruction cache is reset
This bit can be written only when the I cache is disabled.
Bit 10 DCEN: Data cache enable
0: Data cache is disabled
1: Data cache is enabled
Bit 9 ICEN: Instruction cache enable
0: Instruction cache is disabled
1: Instruction cache is enabled
Bit 8 PRFTEN: Prefetch enable
0: Prefetch is disabled
1: Prefetch is enabled
Bits 2:0 LATENCY[2:0]: Latency
These bits represent the ratio of the CPU clock period to the Flash memory access time.
000: Zero wait state
001: One wait state
010: Two wait states
011: Three wait states
100: Four wait states
101: Five wait states
110: Six wait states
111: Seven wait states
98/1751 RM0090 Rev 19


for STM32F42xxx and STM32F43xxx
The Flash access control register is used to enable/disable the acceleration features and
control the Flash memory access time according to CPU frequency.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DCRST ICRST DCEN ICEN PRFTEN LATENCY[3:0]
Reserved Reserved
rw w rw rw rw rw rw rw rw

Bit 12 DCRST: Data cache reset
0: Data cache is not reset
1: Data cache is reset
This bit can be written only when the D cache is disabled.
Bit 11 ICRST: Instruction cache reset
0: Instruction cache is not reset
1: Instruction cache is reset
This bit can be written only when the I cache is disabled.
Bit 10 DCEN: Data cache enable
0: Data cache is disabled
1: Data cache is enabled
Bit 9 ICEN: Instruction cache enable
0: Instruction cache is disabled
1: Instruction cache is enabled
Bit 8 PRFTEN: Prefetch enable
0: Prefetch is disabled
1: Prefetch is enabled
Bits 3:0 LATENCY[3:0]: Latency
These bits represent the ratio of the CPU clock period to the Flash memory access time.
0000: Zero wait state
0001: One wait state
0010: Two wait states
...
1110: Fourteen wait states
1111: Fifteen wait states
RM0090 Rev 19 99/1751

112
3.9.3 Flash key register (FLASH_KEYR)

The Flash key register is used to allow access to the Flash control register and so, to allow
program and erase operations.
Access: no wait state, word access
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
KEY[31:16]
w w w w w w w w w w w w w w w w
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
KEY[15:0]
Bits 31:0 FKEYR[31:0]: FPEC key

The following values must be programmed consecutively to unlock the FLASH_CR register
and allow programming/erasing it:
a) KEY1 = 0x45670123
b) KEY2 = 0xCDEF89AB
3.9.4 Flash option key register (FLASH_OPTKEYR)

The Flash option key register is used to allow program and erase operations in the user
configuration sector.
Access: no wait state, word access
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
OPTKEYR[31:16
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OPTKEYR[15:0]
Bits 31:0 OPTKEYR[31:0]: Option byte key

The following values must be programmed consecutively to unlock the FLASH_OPTCR
register and allow programming it:
a) OPTKEY1 = 0x08192A3B
b) OPTKEY2 = 0x4C5D6E7F
100/1751 RM0090 Rev 19


The Flash status register gives information on ongoing program and erase operations.
Address offset: 0x0C
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
BSY
Reserved
r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PGSERR PGPERR PGAERR WRPERR OPERR EOP
Reserved Reserved
rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1

Bit 16 BSY: Busy
This bit indicates that a Flash memory operation is in progress. It is set at the beginning of a
Flash memory operation and cleared when the operation finishes or an error occurs.
0: no Flash memory operation ongoing
1: Flash memory operation ongoing
Bit 7 PGSERR: Programming sequence error
Set by hardware when a write access to the Flash memory is performed by the code while the
control register has not been correctly configured.
Cleared by writing 1.
Bit 6 PGPERR: Programming parallelism error
Set by hardware when the size of the access (byte, half-word, word, double word) during the
program sequence does not correspond to the parallelism configuration PSIZE (x8, x16, x32,
x64).
Bit 5 PGAERR: Programming alignment error
Set by hardware when the data to program cannot be contained in the same 128-bit Flash
memory row.
Bit 4 WRPERR: Write protection error
Set by hardware when an address to be erased/programmed belongs to a write-protected part
of the Flash memory.
RM0090 Rev 19 101/1751

112

Bit 1 OPERR: Operation error
Set by hardware when a flash operation (programming / erase /read) request is detected and
can not be run because of parallelism, alignment, or write protection error. This bit is set only if
error interrupts are enabled (ERRIE = 1).
Bit 0 EOP: End of operation
Set by hardware when one or more Flash memory operations (program/erase) has/have
completed successfully. It is set only if the end of operation interrupts are enabled (EOPIE = 1).
Cleared by writing a 1.

STM32F42xxx and STM32F43xxx
The Flash status register gives information on ongoing program and erase operations.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
BSY
Reserved
r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RDERR PGSERR PGPERR PGAERR WRPERR OPERR EOP
Reserved Reserved
rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1

Bit 16 BSY: Busy
This bit indicates that a Flash memory operation is in progress to/from one bank. It is set at the
beginning of a Flash memory operation and cleared when the operation finishes or an error
occurs.
0: no Flash memory operation ongoing
1: Flash memory operation ongoing
Bit 8 RDERR: Proprietary readout protection (PCROP) error
Set by hardware when a read access through the D-bus is performed to an address belonging
to a proprietary readout protected Flash sector.
Bit 7 PGSERR: Programming sequence error
Set by hardware when a write access to the Flash memory is performed by the code while the
control register has not been correctly configured.
102/1751 RM0090 Rev 19

Bit 6 PGPERR: Programming parallelism error

Set by hardware when the size of the access (byte, half-word, word, double word) during the
program sequence does not correspond to the parallelism configuration PSIZE (x8, x16, x32,
x64).
Bit 5 PGAERR: Programming alignment error
Set by hardware when the data to program cannot be contained in the same 128-bit Flash
memory row.
Bit 4 WRPERR: Write protection error
Set by hardware when an address to be erased/programmed belongs to a write-protected part
of the Flash memory.
Bit 1 OPERR: Operation error
Set by hardware when a flash operation (programming/erase/read) request is detected and
can not be run because of parallelism, alignment, write or read (PCROP) protection error. This
bit is set only if error interrupts are enabled (ERRIE = 1).
Bit 0 EOP: End of operation
Set by hardware when one or more Flash memory operations (program/erase) has/have
completed successfully. It is set only if the end of operation interrupts are enabled (EOPIE = 1).
Cleared by writing a 1.

The Flash control register is used to configure and start Flash memory operations.
Access: no wait state when no Flash memory operation is ongoing, word, half-word and
byte access.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
LOCK ERRIE EOPIE STRT
Reserved Reserved
rs rw rw rs
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PSIZE[1:0] SNB[3:0] MER SER PG
Reserved Res.
rw rw rw rw rw rw rw rw rw
RM0090 Rev 19 103/1751

112
Bit 31 LOCK: Lock

Write to 1 only. When it is set, this bit indicates that the FLASH_CR register is locked. It is
cleared by hardware after detecting the unlock sequence.
In the event of an unsuccessful unlock operation, this bit remains set until the next reset.
Bit 25 ERRIE: Error interrupt enable
This bit enables the interrupt generation when the OPERR bit in the FLASH_SR register is set
to 1.
0: Error interrupt generation disabled
1: Error interrupt generation enabled
Bit 24 EOPIE: End of operation interrupt enable
This bit enables the interrupt generation when the EOP bit in the FLASH_SR register goes to
1.
0: Interrupt generation disabled
1: Interrupt generation enabled
Bit 16 STRT: Start
This bit triggers an erase operation when set. It is set only by software and cleared when the
BSY bit is cleared.
Bits 9:8 PSIZE[1:0]: Program size
These bits select the program parallelism.
00 program x8
01 program x16
10 program x32
11 program x64
Bit 7 Reserved, must be kept cleared.
Bits 6:3 SNB[3:0]: Sector number
These bits select the sector to erase.
0000 sector 0
0001 sector 1
...
1011 sector 11
Others not allowed
Bit 2 MER: Mass Erase
Erase activated for all user sectors.
Bit 1 SER: Sector Erase
Sector Erase activated.
Bit 0 PG: Programming
Flash programming activated.
104/1751 RM0090 Rev 19


STM32F42xxx and STM32F43xxx
The Flash control register is used to configure and start Flash memory operations.
byte access.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
LOCK ERRIE EOPIE STRT
Reserved Reserved
rs rw rw rs
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MER1 PSIZE[1:0] SNB[4:0] MER SER PG
Reserved
rw rw rw rw rw rw rw rw rw rw rw
Bit 31 LOCK: Lock

Write to 1 only. When it is set, this bit indicates that the FLASH_CR register is locked. It is
cleared by hardware after detecting the unlock sequence.
This bit enables the interrupt generation when the OPERR bit in the FLASH_SR register is set
to 1.
0: Error interrupt generation disabled
1: Error interrupt generation enabled
Bit 24 EOPIE: End of operation interrupt enable
This bit enables the interrupt generation when the EOP bit in the FLASH_SR register goes to
1.
0: Interrupt generation disabled
1: Interrupt generation enabled
Bit 16 STRT: Start
This bit triggers an erase operation when set. It is set only by software and cleared when the
BSY bit is cleared.
Bit 15 MER1: Mass Erase of bank 2 sectors
Erase activated for bank 2 user sectors 12 to 23.
Bits 9:8 PSIZE[1:0]: Program size
These bits select the program parallelism.
00 program x8
01 program x16
10 program x32
11 program x64
RM0090 Rev 19 105/1751

112
Bits 7:3 SNB[3:0]: Sector number

These bits select the sector to erase.
0000: sector 0
0001: sector 1
...
01011: sector 11
01100: not allowed
01101: not allowed
01110: not allowed
01111: not allowed
10000: section 12
10001: section 13
...
11011 sector 23
11100: not allowed
11101: not allowed
11110: not allowed
11111: not allowed
Bit 2 MER: Mass Erase of bank 1 sectors
Erase activated of bank 1 sectors.
Bit 1 SER: Sector Erase
Sector Erase activated.
Bit 0 PG: Programming
Flash programming activated.
3.9.9 Flash option control register (FLASH_OPTCR) for

The FLASH_OPTCR register is used to modify the user option bytes.
Reset value: 0x0FFF AAED. The option bits are loaded with values from Flash memory at
reset release.
byte access.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
nWRP[11:0]
Reserved
rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
nRST_ nRST_ WDG_ OPTST OPTLO
RDP[7:0] Reserve BOR_LEV
STDBY STOP SW RT CK
d
rw rw rw rw rw rw rw rw rw rw rw rw rw rs rs
106/1751 RM0090 Rev 19


Bits 27:16 nWRP[11:0]: Not write protect
These bits contain the value of the write-protection option bytes after reset. They can be
written to program a new write protect value into Flash memory.
0: Write protection active on selected sector
1: Write protection inactive on selected sector
Bits 15:8 RDP[7:0]: Read protect
These bits contain the value of the read-protection option level after reset. They can be written
to program a new read protection value into Flash memory.
0xAA: Level 0, read protection not active
0xCC: Level 2, chip read protection active
Others: Level 1, read protection of memories active
Bits 7:5 USER[2:0]: User option bytes
These bits contain the value of the user option byte after reset. They can be written to program
a new user option byte value into Flash memory.
Bit 7: nRST_STDBY
Bit 6: nRST_STOP
Bit 5: WDG_SW
Note: When changing the WDG mode from hardware to software or from software to
hardware, a system reset is required to make the change effective.
Bit 4 Reserved, must be kept cleared. Always read as “0”.
Bits 3:2 BOR_LEV[1:0]: BOR reset Level
These bits contain the supply level threshold that activates/releases the reset. They can be
written to program a new BOR level. By default, BOR is off. When the supply voltage (VDD)
drops below the selected BOR level, a device reset is generated.
Note: For full details about BOR characteristics, refer to the “Electrical characteristics” section
in the device datasheet.
Bit 1 OPTSTRT: Option start
This bit triggers a user option operation when set. It is set only by software and cleared when
the BSY bit is cleared.
Bit 0 OPTLOCK: Option lock
Write to 1 only. When this bit is set, it indicates that the FLASH_OPTCR register is locked. This
bit is cleared by hardware after detecting the unlock sequence.
RM0090 Rev 19 107/1751

112
3.9.10 Flash option control register (FLASH_OPTCR)

The FLASH_OPTCR register is used to modify the user option bytes.
Reset value: 0x0FFF AAED. The option bits are loaded with values from Flash memory at
reset release.
byte access.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
SPR
DB1M nWRP[11:0]
MOD Reserved
rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
nRST_ nRST_ WDG_ OPTST OPTLO
RDP[7:0] BFB2 BOR_LEV
STDBY STOP SW RT CK
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rs rs
Bit 31 SPRMOD: Selection of protection mode for nWPRi bits

0: PCROP disabled. nWPRi bits used for Write protection on sector i.
1: PCROP enabled. nWPRi bits used for PCROP protection on sector i
Bit 30 DB1M: Dual-bank on 1 Mbyte Flash memory devices
0: 1 Mbyte single bank Flash memory (contiguous addresses in bank1)
1: 1 Mbyte dual bank Flash memory. The Flash memory is organized as two banks of
512 Kbytes each (see Table 7: 1 Mbyte Flash memory single bank vs dual bank organization
(STM32F42xxx and STM32F43xxx) and Table 9: 1 Mbyte dual bank Flash memory
organization (STM32F42xxx and STM32F43xxx)). To perform an erase operation, the right
sector must be programmed (see Table 7 for information on the sector numbering scheme).
Note: If DB1M is set and an erase operation is performed on Bank 2 while the default sector
number is selected (as an example, sector 8 is configured instead of sector 12), the
erase operation on Bank 2 sector is not performed.
These bits contain the value of the write-protection and read-protection (PCROP) option bytes
for sectors 0 to 11 after reset. They can be written to program a new write-protect or PCROP
value into Flash memory.
If SPRMOD is reset:
0: Write protection active on sector i
1: Write protection not active on sector i
If SPRMOD is set:
0: PCROP protection not active on sector i
1: PCROP protection active on sector i
Bits 15:8 RDP[7:0]: Read protect
These bits contain the value of the read-protection option level after reset. They can be written
to program a new read protection value into Flash memory.
0xAA: Level 0, read protection not active
0xCC: Level 2, chip read protection active
Others: Level 1, read protection of memories active
108/1751 RM0090 Rev 19

Bits 7:5 USER: User option bytes

These bits contain the value of the user option byte after reset. They can be written to program
a new user option byte value into Flash memory.
Bit 7: nRST_STDBY
Bit 6: nRST_STOP
Bit 5: WDG_SW
Note: When changing the WDG mode from hardware to software or from software to
hardware, a system reset is required to make the change effective.
Bit 4 BFB2: Dual-bank Boot option byte
0: Dual-bank boot disabled. Boot can be performed either from Flash memory bank 1 or from
system memory depending on boot pin state (default)
1: Dual-bank boot enabled. Boot is always performed from system memory.
Note: For STM32F42xx and STM32F43xx 1MB part numbers, this option bit must be kept
cleared when DB1M=0.
Bits 3:2 BOR_LEV: BOR reset Level
These bits contain the supply level threshold that activates/releases the reset. They can be
written to program a new BOR level. By default, BOR is off. When the supply voltage (VDD)
drops below the selected BOR level, a device reset is generated.
Note: For full details on BOR characteristics, refer to the “Electrical characteristics” section of
the product datasheet.
Bit 1 OPTSTRT: Option start
This bit triggers a user option operation when set. It is set only by software and cleared when
the BSY bit is cleared.
Bit 0 OPTLOCK: Option lock
Write to 1 only. When this bit is set, it indicates that the FLASH_OPTCR register is locked. This
bit is cleared by hardware after detecting the unlock sequence.
RM0090 Rev 19 109/1751

112
3.9.11 Flash option control register (FLASH_OPTCR1)

This register is available only on STM32F42xxx and STM32F43xxx.
The FLASH_OPTCR1 register is used to modify the user option bytes for bank 2.
Reset value: 0x0FFF 0000. The option bits are loaded with values from Flash memory at
reset release.
byte access.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
nWRP[11:0]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserved

These bits contain the value of the write-protection and read-protection (PCROP) option bytes
for sectors 0 to 11 after reset. They can be written to program a new write-protect or PCROP
value into Flash memory.
1: Write protection not active on sector i.
If SPRMOD is set:
110/1751 RM0090 Rev 19

3.9.12 Flash interface register map
Table 19. Flash register map and reset values

Offset Register 31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
PRFTEN
DCRST
ICRST
DCEN
ICEN
LATENCY
FLASH_ACR
[2:0]
0x00 Reserved Reserved
Reset value 0 0 0 0 0 0 0 0
FLASH_
KEY[31:16] KEY[15:0]
KEYR
0x04
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
FLASH_OPT
OPTKEYR[31:16] OPTKEYR[15:0]
KEYR
0x08
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
WRPERR
PGSERR
PGPERR
PGAERR
OPERR
Reserved
EOP
BSY
FLASH_SR
0x0C Reserved Reserved
Reset value 0 0 0 0 0 0 0
PSIZE[1:0]
EOPIE
LOCK
STRT
MER
Reserved
SER
PG
FLASH_CR SNB[3:0]
0x10 Reserved Reserved Reserved
Reset value 1 0 0 0 0 0 0 0 0 0 0
BOR_LEV[1:0]
nRST_STDBY
nRST_STOP
OPTLOCK
OPTSTRT
WDG_SW
FLASH_
Reserved
nWRP[11:0] RDP[7:0]
0x14 OPTCR Reserved
Reset value 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 0 1 0 1 0 1 1 1 1 1 0 1
Table 20. Flash register map and reset values (STM32F42xxx and STM32F43xxx)
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
PRFTEN
DCRST
ICRST
DCEN
ICEN
FLASH_ACR LATENCY[3:0]
Reset value 0 0 0 0 0 0 0 0 0
FLASH_KEYR KEY[31:16] KEY[15:0]

0x04
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
FLASH_
OPTKEYR[31:16] OPTKEYR[15:0]
OPTKEYR
0x08
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
WRPERR
PGSERR
PGPERR
PGAERR
OPERR
RDERR
Reserved
EOP
BSY
FLASH_SR
Reset value 0 0 0 0 0 0 0 0
RM0090 Rev 19 111/1751

112
Table 20. Flash register map and reset values (STM32F42xxx and STM32F43xxx) (continued)
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
PSIZE[1:0]
EOPIE
MER1
LOCK
STRT
MER
SER
PG
FLASH_CR SNB[4:0]
Reset value 1 0 0 0 0 0 0 0 0 0 0 0 0 0
BOR_LEV[1:0]
nRST_STDBY
nRST_STOP
OPTLOCK
OPTSTRT
WDG_SW
SPRMOD
DB1M
BFB2
Reserved
FLASH_OPTCR nWRP[11:0] RDP[7:0]

0x14
Reset value 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 0 1 0 1 0 1 1 1 0 1 1 0 1
FLASH_
Reserved
nWRP[11:0]
OPTCR1
0x18 Reserved
Reset value 1 1 1 1 1 1 1 1 1 1 1 1
112/1751 RM0090 Rev 19

RM0090 CRC calculation unit
4 CRC calculation unit
This section applies to the whole STM32F4xx family, unless otherwise specified.
4.1 CRC introduction

The CRC (cyclic redundancy check) calculation unit is used to get a CRC code from a 32-bit
data word and a fixed generator polynomial.
Among other applications, CRC-based techniques are used to verify data transmission or
storage integrity. In the scope of the EN/IEC 60335-1 standard, they offer a means of
verifying the Flash memory integrity. The CRC calculation unit helps compute a signature of
the software during runtime, to be compared with a reference signature generated at link-
time and stored at a given memory location.
4.2 CRC main features

• Uses CRC-32 (Ethernet) polynomial: 0x4C11DB7
– X32 + X26 + X23 + X22 + X16 + X12 + X11 + X10 +X8 + X7 + X5 + X4 + X2+ X + 1
• Single input/output 32-bit data register
• CRC computation done in 4 AHB clock cycles (HCLK)
• General-purpose 8-bit register (can be used for temporary storage)
The block diagram is shown in Figure 8.
Figure 8. CRC calculation unit block diagram
AHB bus
32-bit (read access)
Data register (output)
CRC computation (polynomial: 0x4C11DB7)
32-bit (write access)
Data register (input)
ai14968
RM0090 Rev 19 113/1751

115
CRC calculation unit RM0090
4.3 CRC functional description

The CRC calculation unit mainly consists of a single 32-bit data register, which:
• is used as an input register to enter new data in the CRC calculator (when writing into
the register)
• holds the result of the previous CRC calculation (when reading the register)
Each write operation into the data register creates a combination of the previous CRC value
and the new one (CRC computation is done on the whole 32-bit data word, and not byte per
byte).
The write operation is stalled until the end of the CRC computation, thus allowing back-to-
back write accesses or consecutive write and read accesses.
The CRC calculator can be reset to 0xFFFF FFFF with the RESET control bit in the
CRC_CR register. This operation does not affect the contents of the CRC_IDR register.
4.4 CRC registers

The CRC calculation unit contains two data registers and a control register.The peripheral
The CRC registers have to be accessed by words (32 bits).
4.4.1 Data register (CRC_DR)

Reset value: 0xFFFF FFFF
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DR [31:16]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DR [15:0]
Bits 31:0 Data register bits

Used as an input register when writing new data into the CRC calculator.
Holds the previous CRC calculation result when it is read.
4.4.2 Independent data register (CRC_IDR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
IDR[7:0]
Reserved
rw rw rw rw rw rw rw rw
114/1751 RM0090 Rev 19

RM0090 CRC calculation unit
Bits 31:8 Reserved, must be kept at reset value.

Bits 7:0 General-purpose 8-bit data register bits
Can be used as a temporary storage location for one byte.
This register is not affected by CRC resets generated by the RESET bit in the CRC_CR
register.
4.4.3 Control register (CRC_CR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RESET
Reserved
w

Bit 0 RESET bit
Resets the CRC calculation unit and sets the data register to 0xFFFF FFFF.
This bit can only be set, it is automatically cleared by hardware.
4.4.4 CRC register map

The following table provides the CRC register map and reset values.
Table 21. CRC calculation unit register map and reset values
Offset Register 31-24 23-16 15-8 7 6 5 4 3 2 1 0
CRC_DR Data register

0x00 Reset
0xFFFF FFFF
value
CRC_IDR Independent data register
0x04 Reset Reserved
0x00
value
CRC_CR RESET
0x08 Reset Reserved
0
value
RM0090 Rev 19 115/1751

115
Power controller (PWR) RM0090
5 Power controller (PWR)
5.1 Power supplies

The device requires a 1.8 to 3.6 V operating voltage supply (VDD). An embedded linear
voltage regulator is used to supply the internal 1.2 V digital power.
The real-time clock (RTC), the RTC backup registers, and the backup SRAM (BKP SRAM)
can be powered from the VBAT voltage when the main VDD supply is powered off.
Note: Depending on the operating power supply range, some peripheral may be used with limited
functionality and performance. For more details refer to section “General operating
conditions” in STM32F4xx datasheets.
Figure 9. Power supply overview for STM32F405xx/07xx and STM32F415xx/17xx
VBAT
Backup circuitry
VBAT = Power (OSC32K,RTC,
1.65 to 3.6V switch Wakeup logic
Backup registers,
backup RAM)
Level shifter
OUT
IO
GPIOs
Logic
IN
VCAP_1 Kernel logic
2 × 2.2 μF VCAP_2 (CPU, digital
& RAM)
VDD VDD
1/2/...14/15 Voltage
15 × 100 nF VSS regulator
+ 1 × 4.7 μF 1/2/...14/15
BYPASS_REG Flash memory
Reset
PDR_ON controller
VDD
VDDA
VREF
VREF+
Analog:
100 nF 100 nF VREF- ADC RCs,
+ 1 μF + 1 μF PLL,..
VSSA
MS19911V2
1. VDDA and VSSA must be connected to VDD and VSS, respectively.
116/1751 RM0090 Rev 19

RM0090 Power controller (PWR)
Figure 10. Power supply overview for STM32F42xxx and STM32F43xxx

VBAT
Backup circuitry
VBAT = Po wer swi tch (OSC32K,RTC,
1.65 to 3.6V Wakeup logic
Backup registers,
backup RAM)
Level shifter
OUT
IO
GP I/Os Logic
IN Kernel logic
(CPU,
VCAP_1
VCAP_2 digital
2 × 2.2 μF & RAM)
VDD
VDD Voltage
1/2/...14/20
regulator
19 × 100 nF VSS
+ 1 × 4.7 μF 1/2/...14/20
BYPASS_REG Flash memory

Reset
PDR_ON controller
VDD
VDDA
VREF
VREF+
100 nF Analog:
100 nF VREF- ADC RCs, PLL,
+ 1 μF + 1 μF
...
VSSA
MS30482V2
1. VDDA and VSSA must be connected to VDD and VSS, respectively.
5.1.1 Independent A/D converter supply and reference voltage

To improve conversion accuracy, the ADC has an independent power supply which can be
separately filtered and shielded from noise on the PCB.
• The ADC voltage supply input is available on a separate VDDA pin.
• An isolated supply ground connection is provided on pin VSSA.
To ensure a better accuracy of low voltage inputs, the user can connect a separate external
reference voltage ADC input on VREF. The voltage on VREF ranges from 1.8 V to VDDA.
RM0090 Rev 19 117/1751

149
5.1.2 Battery backup domain

Backup domain description
To retain the content of the RTC backup registers, backup SRAM, and supply the RTC when
VDD is turned off, VBAT pin can be connected to an optional standby voltage supplied by a
battery or by another source.
To allow the RTC to operate even when the main digital supply (VDD) is turned off, the VBAT
pin powers the following blocks:
• The RTC
• The LSE oscillator
• The backup SRAM when the low-power backup regulator is enabled
• PC13 to PC15 I/Os, plus PI8 I/O (when available)
The switch to the VBAT supply is controlled by the power-down reset embedded in the Reset
block.
Warning: During tRSTTEMPO (temporization at VDD startup) or after a PDR

is detected, the power switch between VBAT and VDD remains
connected to VBAT.
During the startup phase, if VDD is established in less than
tRSTTEMPO (Refer to the datasheet for the value of tRSTTEMPO)
and VDD > VBAT + 0.6 V, a current may be injected into VBAT
through an internal diode connected between VDD and the
power switch (VBAT).
If the power supply/battery connected to the VBAT pin cannot
support this current injection, it is strongly recommended to
connect an external low-drop diode between this power
supply and the VBAT pin.
If no external battery is used in the application, it is recommended to connect the VBAT pin to
VDD with a 100 nF external decoupling ceramic capacitor in parallel.
When the backup domain is supplied by VDD (analog switch connected to VDD), the
following functions are available:
• PC14 and PC15 can be used as either GPIO or LSE pins
• PC13 can be used as a GPIOas the RTC_AF1 pin (refer to Table 37: RTC_AF1 pin for
more details about this pin configuration)
Note: Due to the fact that the switch only sinks a limited amount of current (3 mA), the use of PI8
and PC13 to PC15 GPIOs in output mode is restricted: the speed has to be limited to 2 MHz
with a maximum load of 30 pF and these I/Os must not be used as a current source (e.g. to
drive an LED).
118/1751 RM0090 Rev 19

When the backup domain is supplied by VBAT (analog switch connected to VBAT because
VDD is not present), the following functions are available:
• PC14 and PC15 can be used as LSE pins only
• PC13 can be used as the RTC_AF1 pin (refer to Table 37: RTC_AF1 pin for more
details about this pin configuration)
• PI8 can be used as RTC_AF2
Backup domain access

After reset, the backup domain (RTC registers, RTC backup register and backup SRAM) is
protected against possible unwanted write accesses. To enable access to the backup
domain, proceed as follows:
• Access to the RTC and RTC backup registers
1. Enable the power interface clock by setting the PWREN bits in the RCC_APB1ENR
register (see Section 7.3.13 and Section 6.3.13)
2. Set the DBP bit in the Section 5.4.1 and PWR power control register (PWR_CR) for
STM32F42xxx and STM32F43xxx to enable access to the backup domain
3. Select the RTC clock source: see Section 7.2.8: RTC/AWU clock
4. Enable the RTC clock by programming the RTCEN [15] bit in the Section 7.3.20: RCC
Backup domain control register (RCC_BDCR)
• Access to the backup SRAM
1. Enable the power interface clock by setting the PWREN bits in the RCC_APB1ENR
register (see Section 7.3.13 and Section 6.3.13 for STM32F405xx/07xx and
STM32F415xx/17xx and STM32F42xxx and STM32F43xxx, respectively)
2. Set the DBP bit in the PWR power control register (PWR_CR) for STM32F405xx/07xx
and STM32F415xx/17xx and PWR power control register (PWR_CR) for
STM32F42xxx and STM32F43xxx to enable access to the backup domain
3. Enable the backup SRAM clock by setting BKPSRAMEN bit in the RCC AHB1
peripheral clock enable register (RCC_AHB1ENR).
RTC and RTC backup registers

The real-time clock (RTC) is an independent BCD timer/counter. The RTC provides a time-
of-day clock/calendar, two programmable alarm interrupts, and a periodic programmable
wakeup flag with interrupt capability. The RTC contains 20 backup data registers (80 bytes)
which are reset when a tamper detection event occurs. For more details refer to Section 26:
Real-time clock (RTC).
Backup SRAM
The backup domain includes 4 Kbytes of backup SRAM addressed in 32-bit, 16-bit or 8-bit
mode. Its content is retained even in Standby or VBAT mode when the low-power backup
regulator is enabled. It can be considered as an internal EEPROM when VBAT is always
present.
When the backup domain is supplied by VDD (analog switch connected to VDD), the backup
SRAM is powered from VDD which replaces the VBAT power supply to save battery life.
When the backup domain is supplied by VBAT (analog switch connected to VBAT because
VDD is not present), the backup SRAM is powered by a dedicated low-power regulator. This
regulator can be ON or OFF depending whether the application needs the backup SRAM
function in Standby and VBAT modes or not. The power-down of this regulator is controlled
RM0090 Rev 19 119/1751

149
by a dedicated bit, the BRE control bit of the PWR_CSR register (see Section 5.4.2: PWR
power control/status register (PWR_CSR) for STM32F405xx/07xx and
STM32F415xx/17xx).
The backup SRAM is not mass erased by an tamper event. When the Flash memory is read
protected, the backup SRAM is also read protected to prevent confidential data, such as
cryptographic private key, from being accessed. When the protection level change from
level 1 to level 0 is requested, the backup SRAM content is erased.
Figure 11. Backup domain
Power switch LP voltage regulator

Voltage regulator 3.3 -> 1.2 V
3.3 -> 1.2 V
1.2 V domain
Backup SRAM Backup SRAM

interface 1.2 V
RTC LSE 32.768 Hz
Backup domain
MS30430V1
5.1.3 Voltage regulator for STM32F405xx/07xx and STM32F415xx/17xx

An embedded linear voltage regulator supplies all the digital circuitries except for the backup
domain and the Standby circuitry. The regulator output voltage is around 1.2 V.
This voltage regulator requires one or two external capacitors to be connected to one or two
dedicated pins, VCAP_1 and VCAP_2 available in all packages. Specific pins must be
connected either to VSS or VDD to activate or deactivate the voltage regulator. These pins
depend on the package.
When activated by software, the voltage regulator is always enabled after Reset. It works in
three different modes depending on the application modes.
• In Run mode, the regulator supplies full power to the 1.2 V domain (core, memories
and digital peripherals). In this mode, the regulator output voltage (around 1.2 V) can
be scaled by software to different voltage values:
Scale 1 or scale 2 can be configured on the fly through VOS (bit 15 of the
PWR_CR register).
The voltage scaling allows optimizing the power consumption when the device is
clocked below the maximum system frequency.
• In Stop mode, the main regulator or the low-power regulator supplies to the 1.2 V
domain, thus preserving the content of registers and internal SRAM. The voltage
120/1751 RM0090 Rev 19

regulator can be put either in main regulator mode (MR) or in low-power mode (LPR).
The programmed voltage scale remains the same during Stop mode:
The programmed voltage scale remains the same during Stop mode (see
Section 5.4.1: PWR power control register (PWR_CR) for STM32F405xx/07xx
and STM32F415xx/17xx).
• In Standby mode, the regulator is powered down. The content of the registers and
SRAM are lost except for the Standby circuitry and the backup domain.
Note: For more details, refer to the voltage regulator section in the STM32F405xx/07xx and
STM32F415xx/17xx datasheets.
5.1.4 Voltage regulator for STM32F42xxx and STM32F43xxx

An embedded linear voltage regulator supplies all the digital circuitries except for the backup
domain and the Standby circuitry. The regulator output voltage is around 1.2 V.
This voltage regulator requires two external capacitors to be connected to two dedicated
pins, VCAP_1 and VCAP_2 available in all packages. Specific pins must be connected either to
VSS or VDD to activate or deactivate the voltage regulator. These pins depend on the
package.
When activated by software, the voltage regulator is always enabled after Reset. It works in
three different modes depending on the application modes (Run, Stop, or Standby mode).
• In Run mode, the main regulator supplies full power to the 1.2 V domain (core,
memories and digital peripherals). In this mode, the regulator output voltage (around
1.2 V) can be scaled by software to different voltage values (scale 1, scale 2, and scale
3 can be configured through VOS[1:0] bits of the PWR_CR register). The scale can be
modified only when the PLL is OFF and the HSI or HSE clock source is selected as
system clock source. The new value programmed is active only when the PLL is ON.
When the PLL is OFF, the voltage scale 3 is automatically selected.
The voltage scaling allows optimizing the power consumption when the device is
clocked below the maximum system frequency. After exit from Stop mode, the voltage
RM0090 Rev 19 121/1751

149
scale 3 is automatically selected.(see Section 5.4.1: PWR power control register

(PWR_CR) for STM32F405xx/07xx and STM32F415xx/17xx.
2 operating modes are available:
– Normal mode: The CPU and core logic operate at maximum frequency at a given
voltage scaling (scale 1, scale 2 or scale 3)
– Over-drive mode: This mode allows the CPU and the core logic to operate at a
higher frequency than the normal mode for the voltage scaling scale 1 and scale
2.
• In Stop mode: the main regulator or low-power regulator supplies a low-power voltage
to the 1.2V domain, thus preserving the content of registers and internal SRAM.
The voltage regulator can be put either in main regulator mode (MR) or in low-power
mode (LPR). Both modes can be configured by software as follows:
– Normal mode: the 1.2 V domain is preserved in nominal leakage mode. It is the
default mode when the main regulator (MR) or the low-power regulator (LPR) is
enabled.
– Under-drive mode: the 1.2 V domain is preserved in reduced leakage mode. This
mode is only available with the main regulator or in low-power regulator mode
(see Table 22).
• In Standby mode: the regulator is powered down. The content of the registers and
SRAM are lost except for the Standby circuitry and the backup domain.
Note: Over-drive and under-drive mode are not available when the regulator is bypassed.
For more details, refer to the voltage regulator section in the STM32F42xxx and
STM32F43xxx datasheets.
Table 22. Voltage regulator configuration mode versus device operating mode(1)
Voltage regulator
Run mode Sleep mode Stop mode Standby mode
configuration
Normal mode MR MR MR or LPR -

Over-drive
MR MR - -
mode(2)
Under-drive mode - - MR or LPR -
Power-down
- - - Yes
mode
1. ‘-’ means that the corresponding configuration is not available.
2. The over-drive mode is not available when VDD = 1.8 to 2.1 V.
122/1751 RM0090 Rev 19

Entering Over-drive mode

It is recommended to enter Over-drive mode when the application is not running critical
tasks and when the system clock source is either HSI or HSE. To optimize the configuration
time, enable the Over-drive mode during the PLL lock phase.
To enter Over-drive mode, follow the sequence below:
1. Select HSI or HSE as system clock.
2. Configure RCC_PLLCFGR register and set PLLON bit of RCC_CR register.
3. Set ODEN bit of PWR_CR register to enable the Over-drive mode and wait for the
ODRDY flag to be set in the PWR_CSR register.
4. Set the ODSW bit in the PWR_CR register to switch the voltage regulator from Normal
mode to Over-drive mode. The System will be stalled during the switch but the PLL
clock system will be still running during locking phase.
5. Wait for the ODSWRDY flag in the PWR_CSR to be set.
6. Select the required Flash latency as well as AHB and APB prescalers.
7. Wait for PLL lock.
8. Switch the system clock to the PLL.
9. Enable the peripherals that are not generated by the System PLL (I2S clock, LCD-TFT
clock, SAI1 clock, USB_48MHz clock....).
Note: The PLLI2S and PLLSAI can be configured at the same time as the system PLL.
During the Over-drive switch activation, no peripheral clocks should be enabled. The
peripheral clocks must be enabled once the Over-drive mode is activated.
Entering Stop mode disables the Over-drive mode, as well as the PLL. The application
software has to configure again the Over-drive mode and the PLL after exiting from Stop
mode.
Exiting from Over-drive mode

It is recommended to exit from Over-drive mode when the application is not running critical
tasks and when the system clock source is either HSI or HSE.There are two sequences that
allow exiting from over-drive mode:
• By resetting simultaneously the ODEN and ODSW bits bit in the PWR_CR register
(sequence 1)
• By resetting first the ODSW bit to switch the voltage regulator to Normal mode and then
resetting the ODEN bit to disable the Over-drive mode (sequence 2).
Example of sequence 1:
1. Select HSI or HSE as system clock source.
2. Disable the peripheral clocks that are not generated by the System PLL (I2S clock,
LCD-TFT clock, SAI1 clock, USB_48MHz clock,....)
3. Reset simultaneously the ODEN and the ODSW bits in the PWR_CR register to switch
back the voltage regulator to Normal mode and disable the Over-drive mode.
4. Wait for the ODWRDY flag of PWR_CSR to be reset.
RM0090 Rev 19 123/1751

149
Example of sequence 2:
1. Select HSI or HSE as system clock source.
2. Disable the peripheral clocks that are not generated by the System PLL (I2S clock,
LCD-TFT clock, SAI1 clock, USB_48MHz clock,....).
3. Reset the ODSW bit in the PWR_CR register to switch back the voltage regulator to
Normal mode. The system clock is stalled during voltage switching.
4. Wait for the ODWRDY flag of PWR_CSR to be reset.
5. Reset the ODEN bit in the PWR_CR register to disable the Over-drive mode.
Note: During step 3, the ODEN bit remains set and the Over-drive mode is still enabled but not
active (ODSW bit is reset). If the ODEN bit is reset instead, the Over-drive mode is disabled
and the voltage regulator is switched back to the initial voltage.
5.2 Power supply supervisor
5.2.1 Power-on reset (POR)/power-down reset (PDR)

The device has an integrated POR/PDR circuitry that allows proper operation starting
from 1.8 V.
The device remains in Reset mode when VDD/VDDA is below a specified threshold,
VPOR/PDR, without the need for an external reset circuit. For more details concerning the
power on/power-down reset threshold, refer to the electrical characteristics of the
datasheet.
Figure 12. Power-on reset/power-down reset waveform

VDD/VDDA
PDR
40 mV
hysteresis PDR
Temporization
tRSTTEMPO
Reset
MS30431V1
124/1751 RM0090 Rev 19

5.2.2 Brownout reset (BOR)

During power on, the Brownout reset (BOR) keeps the device under reset until the supply
voltage reaches the specified VBOR threshold.
VBOR is configured through device option bytes. By default, BOR is off. 3 programmable
VBOR threshold levels can be selected:
• BOR Level 3 (VBOR3). Brownout threshold level 3.
Note: For full details about BOR characteristics, refer to the "Electrical characteristics" section in
the device datasheet.
When the supply voltage (VDD) drops below the selected VBOR threshold, a device reset is
generated.
The BOR can be disabled by programming the device option bytes. In this case, the
power-on and power-down is then monitored by the POR/ PDR (see Section 5.2.1: Power-
on reset (POR)/power-down reset (PDR)).
The BOR threshold hysteresis is ~100 mV (between the rising and the falling edge of the
supply voltage).
Figure 13. BOR thresholds

VDD/VDDA
BOR threshold 100 mV

hysteresis
Reset
MS30433V1
5.2.3 Programmable voltage detector (PVD)

You can use the PVD to monitor the VDD power supply by comparing it to a threshold
selected by the PLS[2:0] bits in the PWR power control register (PWR_CR) for
RM0090 Rev 19 125/1751

149
STM32F405xx/07xx and STM32F415xx/17xx and PWR power control register (PWR_CR)

for STM32F42xxx and STM32F43xxx.
The PVD is enabled by setting the PVDE bit.
A PVDO flag is available, in the PWR power control/status register (PWR_CSR) for
STM32F405xx/07xx and STM32F415xx/17xx, to indicate if VDD is higher or lower than the
PVD threshold. This event is internally connected to the EXTI line16 and can generate an
interrupt if enabled through the EXTI registers. The PVD output interrupt can be generated
when VDD drops below the PVD threshold and/or when VDD rises above the PVD threshold
depending on EXTI line16 rising/falling edge configuration. As an example the service
routine could perform emergency shutdown tasks.
Figure 14. PVD thresholds

VDD
PVD threshold 100 mV

hysteresis
PVD output
MS30432V2
5.3 Low-power modes

By default, the microcontroller is in Run mode after a system or a power-on reset. In Run
mode the CPU is clocked by HCLK and the program code is executed. Several low-power
modes are available to save power when the CPU does not need to be kept running, for
example when waiting for an external event. It is up to the user to select the mode that gives
the best compromise between low-power consumption, short startup time and available
wakeup sources.
The devices feature three low-power modes:
• Sleep mode (Cortex®-M4 with FPU core stopped, peripherals kept running)
• Stop mode (all clocks are stopped)
• Standby mode (1.2 V domain powered off)
126/1751 RM0090 Rev 19

In addition, the power consumption in Run mode can be reduce by one of the following
means:
• Slowing down the system clocks
• Gating the clocks to the APBx and AHBx peripherals when they are unused.
Entering low-power mode

Low-power modes are entered by the MCU by executing the WFI (Wait For Interrupt), or
WFE (Wait for Event) instructions, or when the SLEEPONEXIT bit in the Cortex®-M4 with
FPU System Control register is set on Return from ISR.
Entering Low-power mode through WFI or WFE will be executed only if no interrupt is
pending or no event is pending.
Exiting low-power mode

The MCU exits from Sleep and Stop modes low-power mode depending on the way the low-
power mode was entered:
• If the WFI instruction or Return from ISR was used to enter the low-power mode, any
peripheral interrupt acknowledged by the NVIC can wake up the device.
• If the WFE instruction is used to enter the low-power mode, the MCU exits the low-
power mode as soon as an event occurs. The wakeup event can be generated either
by:
– NVIC IRQ interrupt:
When SEVONPEND = 0 in the Cortex®-M4 with FPU System Control register: by
enabling an interrupt in the peripheral control register and in the NVIC. When the
MCU resumes from WFE, the peripheral interrupt pending bit and the NVIC
peripheral IRQ channel pending bit (in the NVIC interrupt clear pending register)
have to be cleared. Only NVIC interrupts with sufficient priority will wakeup and
interrupt the MCU.
When SEVONPEND = 1 in the Cortex®-M4 with FPU System Control register: by
enabling an interrupt in the peripheral control register and optionally in the NVIC.
When the MCU resumes from WFE, the peripheral interrupt pending bit and when
enabled the NVIC peripheral IRQ channel pending bit (in the NVIC interrupt clear
pending register) have to be cleared. All NVIC interrupts will wakeup the MCU,
even the disabled ones.Only enabled NVIC interrupts with sufficient priority will
wakeup and interrupt the MCU.
– Event
This is done by configuring a EXTI line in event mode. When the CPU resumes
from WFE, it is not necessary to clear the EXTI peripheral interrupt pending bit or
the NVIC IRQ channel pending bit as the pending bits corresponding to the event
line is not set. It may be necessary to clear the interrupt flag in the peripheral.
The MCU exits from Standby low-power mode through an external reset (NRST pin), an
IWDG reset, a rising edge on one of the enabled WKUPx pins or a RTC event occurs (see
Figure 237: RTC block diagram).
After waking up from Standby mode, program execution restarts in the same way as after a
Reset (boot pin sampling, option bytes loading, reset vector is fetched, etc.).
RM0090 Rev 19 127/1751

149
Only enabled NVIC interrupts with sufficient priority will wakeup and interrupt the MCU.
Table 23. Low-power mode summary

Effect on
Effect on 1.2 V VDD
Mode name Entry Wakeup Voltage regulator
domain clocks domain
clocks
Sleep WFI or Return CPU CLK OFF

Any interrupt
(Sleep now or from ISR no effect on other
None ON
Sleep-on- clocks or analog
exit) WFE Wakeup event clock sources
ON or in low- power
mode (depends on
PWR power control
register (PWR_CR)
for
STM32F405xx/07x
x and
PDDS and LPDS STM32F415xx/17x
bits + Any EXTI line (configured x and PWR power
Stop SLEEPDEEP bit in the EXTI registers, control register
+ WFI, Return internal and external lines) (PWR_CR) for
from ISR or WFE HSI and STM32F405xx/07x
All 1.2 V domain HSE x and
clocks OFF oscillator STM32F415xx/17x
s OFF xPWR power
control register
(PWR_CR) for
STM32F42xxx and
STM32F43xxx
WKUP pin rising edge,
RTC alarm (Alarm A or
PDDS bit +
Alarm B), RTC Wakeup
SLEEPDEEP bit
Standby event, RTC tamper OFF
+ WFI, Return
events, RTC time stamp
from ISR or WFE
event, external reset in
NRST pin, IWDG reset
5.3.1 Slowing down system clocks

In Run mode the speed of the system clocks (SYSCLK, HCLK, PCLK1, PCLK2) can be
reduced by programming the prescaler registers. These prescalers can also be used to slow
down peripherals before entering Sleep mode.
For more details refer to Section 7.3.3: RCC clock configuration register (RCC_CFGR).
5.3.2 Peripheral clock gating

In Run mode, the HCLKx and PCLKx for individual peripherals and memories can be
stopped at any time to reduce power consumption.
To further reduce power consumption in Sleep mode the peripheral clocks can be disabled
prior to executing the WFI or WFE instructions.
128/1751 RM0090 Rev 19

Peripheral clock gating is controlled by the AHB1 peripheral clock enable register
(RCC_AHB1ENR), AHB2 peripheral clock enable register (RCC_AHB2ENR), AHB3
peripheral clock enable register (RCC_AHB3ENR) (see Section 7.3.10: RCC AHB1
peripheral clock enable register (RCC_AHB1ENR), Section 7.3.11: RCC AHB2 peripheral
clock enable register (RCC_AHB2ENR), Section 7.3.12: RCC AHB3 peripheral clock
enable register (RCC_AHB3ENR) for STM32F405xx/07xx and STM32F415xx/17xx, and
Section 6.3.10: RCC AHB1 peripheral clock register (RCC_AHB1ENR), Section 6.3.11:
RCC AHB2 peripheral clock enable register (RCC_AHB2ENR), and Section 6.3.12: RCC
AHB3 peripheral clock enable register (RCC_AHB3ENR) for STM32F42xxx and
STM32F43xxx).
Disabling the peripherals clocks in Sleep mode can be performed automatically by resetting
the corresponding bit in RCC_AHBxLPENR and RCC_APBxLPENR registers.
5.3.3 Sleep mode

Entering Sleep mode
The Sleep mode is entered according to Section : Entering low-power mode, when the
SLEEPDEEP bit in the Cortex®-M4 with FPU System Control register is cleared.
Refer to Table 24 and Table 25 for details on how to enter Sleep mode.
Note: All interrupt pending bits must be cleared before the sleep mode entry.
Exiting Sleep mode

The Sleep mode is exited according to Section : Exiting low-power mode.
Refer to Table 24 and Table 25 for more details on how to exit Sleep mode.
Table 24. Sleep-now entry and exit

Sleep-now mode Description
WFI (Wait for Interrupt) or WFE (Wait for Event) while:

– SLEEPDEEP = 0, and
– No interrupt (for WFI) or event (for WFE) is pending.
Refer to the Cortex®-M4 with FPU System Control register.
Mode entry On Return from ISR while:
– SLEEPDEEP = 0 and
– SLEEPONEXIT = 1,
– No interrupt is pending.
RM0090 Rev 19 129/1751

149
Table 24. Sleep-now entry and exit (continued)

Sleep-now mode Description
If WFI or Return from ISR was used for entry:

Interrupt: Refer to Table 61: Vector table for STM32F405xx/07xx and
STM32F415xx/17xx and Table 62: Vector table for STM32F42xxx and
STM32F43xxx
If WFE was used for entry and SEVONPEND = 0
Mode exit Wakeup event: Refer to Section 12.2.3: Wakeup event management
f WFE was used for entry and SEVONPEND = 1
Interrupt even when disabled in NVIC: refer to Table 61: Vector table for
STM32F405xx/07xx and STM32F415xx/17xx and Table 62: Vector table
for STM32F42xxx and STM32F43xxx or Wakeup event (see
Section 12.2.3: Wakeup event management).
Wakeup latency None
Table 25. Sleep-on-exit entry and exit

Sleep-on-exit Description

– No interrupt (for WFI) or event (for WFE) is pending.
Mode entry On Return from ISR while:
– SLEEPONEXIT = 1, and
– No interrupt is pending.
Interrupt: refer to Table 61: Vector table for STM32F405xx/07xx and
Mode exit STM32F415xx/17xx and Table 62: Vector table for STM32F42xxx and
STM32F43xxx
Wakeup latency None
5.3.4 Stop mode (STM32F405xx/07xx and STM32F415xx/17xx)

The Stop mode is based on the Cortex®-M4 with FPU deepsleep mode combined with
peripheral clock gating. The voltage regulator can be configured either in normal or low-
power mode. In Stop mode, all clocks in the 1.2 V domain are stopped, the PLLs, the HSI
and the HSE RC oscillators are disabled. Internal SRAM and register contents are
preserved.
By setting the FPDS bit in the PWR_CR register, the Flash memory also enters power-down
mode when the device enters Stop mode. When the Flash memory is in power-down mode,
an additional startup delay is incurred when waking up from Stop mode (see Table 26: Stop
operating modes (STM32F405xx/07xx and STM32F415xx/17xx) and Section 5.4.1: PWR
power control register (PWR_CR) for STM32F405xx/07xx and STM32F415xx/17xx).
130/1751 RM0090 Rev 19

Table 26. Stop operating modes

Stop mode LPDS bit FPDS bit Wake-up latency
STOP MR
0 0 HSI RC startup time
(Main regulator)
HSI RC startup time +

STOP MR-FPD 0 1 Flash wakeup time from Power Down
mode

STOP LP 1 0
regulator wakeup time from LP mode

Flash wakeup time from Power Down
STOP LP-FPD 1 1
mode +
regulator wakeup time from LP mode
Entering Stop mode (for STM32F405xx/07xx and STM32F415xx/17xx)

The Stop mode is entered according to Section : Entering low-power mode, when the
SLEEPDEEP bit in the Cortex®-M4 with FPU System Control register is set.
Refer to Table 27 for details on how to enter the Stop mode.
To further reduce power consumption in Stop mode, the internal voltage regulator can be put
in low-power mode. This is configured by the LPDS bit of the PWR power control register
(PWR_CR) for STM32F405xx/07xx and STM32F415xx/17xx and PWR power control
register (PWR_CR) for STM32F42xxx and STM32F43xxx.
If Flash memory programming is ongoing, the Stop mode entry is delayed until the memory
access is finished.
If an access to the APB domain is ongoing, The Stop mode entry is delayed until the APB
access is finished.
In Stop mode, the following features can be selected by programming individual control bits:
• Independent watchdog (IWDG): the IWDG is started by writing to its Key register or by
hardware option. Once started it cannot be stopped except by a Reset. See
Section 21.3 in Section 21: Independent watchdog (IWDG).
• Real-time clock (RTC): this is configured by the RTCEN bit in the Section 7.3.20: RCC
• Internal RC oscillator (LSI RC): this is configured by the LSION bit in the
Section 7.3.21: RCC clock control & status register (RCC_CSR).
• External 32.768 kHz oscillator (LSE OSC): this is configured by the LSEON bit in the
Section 7.3.20: RCC Backup domain control register (RCC_BDCR).
The ADC or DAC can also consume power during the Stop mode, unless they are disabled
before entering it. To disable them, the ADON bit in the ADC_CR2 register and the ENx bit
in the DAC_CR register must both be written to 0.
RM0090 Rev 19 131/1751

149
Note: If the application needs to disable the external clock before entering Stop mode, the HSEON
bit must first be disabled and the system clock switched to HSI.
Otherwise, if the HSEON bit is kept enabled while the external clock (external oscillator) can
be removed before entering stop mode, the clock security system (CSS) feature must be
enabled to detect any external oscillator failure and avoid a malfunction behavior when
entering stop mode.
Exiting Stop mode (for STM32F405xx/07xx and STM32F415xx/17xx)

The Stop mode is exited according to Section : Exiting low-power mode.
Refer to Table 27 for more details on how to exit Stop mode.
When exiting Stop mode by issuing an interrupt or a wakeup event, the HSI RC oscillator is
selected as system clock.
When the voltage regulator operates in low-power mode, an additional startup delay is
incurred when waking up from Stop mode. By keeping the internal regulator ON during Stop
mode, the consumption is higher although the startup time is reduced.
Table 27. Stop mode entry and exit (for STM32F405xx/07xx and STM32F415xx/17xx)
Stop mode Description

– No interrupt (for WFI) or event (for WFE) is pending,
– SLEEPDEEP bit is set in Cortex®-M4 with FPU System Control register,
– PDDS bit is cleared in Power Control register (PWR_CR),
– Select the voltage regulator mode by configuring LPDS bit in PWR_CR.
On Return from ISR:
– No interrupt is pending,
Mode entry – SLEEPDEEP bit is set in Cortex®-M4 with FPU System Control register,
– SLEEPONEXIT = 1,
– PDDS bit is cleared in Power Control register (PWR_CR).
Note: To enter Stop mode, all EXTI Line pending bits (in Pending register
(EXTI_PR)), all peripheral interrupts pending bits, the RTC Alarm
(Alarm A and Alarm B), RTC wakeup, RTC tamper, and RTC time
stamp flags, must be reset. Otherwise, the Stop mode entry
procedure is ignored and program execution continues.
132/1751 RM0090 Rev 19

Table 27. Stop mode entry and exit (for STM32F405xx/07xx and STM32F415xx/17xx)

Any EXTI lines configured in Interrupt mode (the corresponding EXTI
Interrupt vector must be enabled in the NVIC). The interrupt source can
be external interrupts or peripherals with wakeup capability. Refer to
Table 61: Vector table for STM32F405xx/07xx and STM32F415xx/17xx
on page 372 and Table 62: Vector table for STM32F42xxx and
STM32F43xxx.
If WFE was used for entry and SEVONPEND = 0
Any EXTI lines configured in event mode. Refer to Section 12.2.3:
Mode exit Wakeup event management on page 380.
If WFE was used for entry and SEVONPEND = 1:
– Any EXTI lines configured in Interrupt mode (even if the corresponding
EXTI Interrupt vector is disabled in the NVIC). The interrupt source can
be an external interrupt or a peripheral with wakeup capability. Refer to
STM32F43xxx.
– Wakeup event: refer to Section 12.2.3: Wakeup event management on
page 380.
Table 26: Stop operating modes (STM32F405xx/07xx and
Wakeup latency
STM32F415xx/17xx)
5.3.5 Stop mode (STM32F42xxx and STM32F43xxx)

The Stop mode is based on the Cortex®-M4 with FPU deepsleep mode combined with
peripheral clock gating. The voltage regulator can be configured either in normal or low-
power mode. In Stop mode, all clocks in the 1.2 V domain are stopped, the PLLs, the HSI
and the HSE RC oscillators are disabled. Internal SRAM and register contents are
preserved.
In Stop mode, the power consumption can be further reduced by using additional settings in
the PWR_CR register. However this will induce an additional startup delay when waking up
from Stop mode (see Table 28).
RM0090 Rev 19 133/1751

149
Table 28. Stop operating modes (STM32F42xxx and STM32F43xxx)

UDEN[1:0] MRUDS LPUDS LPDS FPDS
Voltage Regulator Mode Wakeup latency
bits bit bit bit bit
STOP MR
- 0 - 0 0 HSI RC startup time
(Main Regulator)
STOP MR- FPD - 0 - 0 1 Flash wakeup time from power-
down mode
Normal HSI RC startup time +

mode STOP LP - 0 0 1 0 regulator wakeup time from LP
mode
Flash wakeup time from power-
STOP LP-FPD - - 0 1 1 down mode +
regulator wakeup time from LP
mode
STOP UMR- down mode +
3 1 - 0 -
FPD Main regulator wakeup time from
under-drive mode + Core logic to
Under- nominal mode
drive
Mode HSI RC startup time +
down mode +
STOP ULP-FPD 3 - 1 1 -
regulator wakeup time from LP
under-drive mode + Core logic to
nominal mode
Entering Stop mode (STM32F42xxx and STM32F43xxx)

The Stop mode is entered according to Section : Entering low-power mode, when the
Refer to Table 29 for details on how to enter the Stop mode.
When the microcontroller enters in Stop mode, the voltage scale 3 is automatically selected.
To further reduce power consumption in Stop mode, the internal voltage regulator can be put
in low voltage mode. This is configured by the LPDS, MRUDS, LPUDS and UDEN bits of
the PWR power control register (PWR_CR) for STM32F405xx/07xx and
STM32F415xx/17xx.
If Flash memory programming is ongoing, the Stop mode entry is delayed until the memory
access is finished.
If an access to the APB domain is ongoing, The Stop mode entry is delayed until the APB
access is finished.
If the Over-drive mode was enabled before entering Stop mode, it is automatically disabled
during when the Stop mode is activated.
134/1751 RM0090 Rev 19

In Stop mode, the following features can be selected by programming individual control bits:
hardware option. Once started it cannot be stopped except by a Reset. See
• Real-time clock (RTC): this is configured by the RTCEN bit in the Section 7.3.20: RCC
• Internal RC oscillator (LSI RC): this is configured by the LSION bit in the
Section 7.3.21: RCC clock control & status register (RCC_CSR).
RCC Backup domain control register (RCC_BDCR).
The ADC or DAC can also consume power during the Stop mode, unless they are disabled
before entering it. To disable them, the ADON bit in the ADC_CR2 register and the ENx bit
in the DAC_CR register must both be written to 0.
Note: Before entering Stop mode, it is recommended to enable the clock security system (CSS)
feature to prevent external oscillator (HSE) failure from impacting the internal MCU
behavior.
Exiting Stop mode (STM32F42xxx and STM32F43xxx)

The Stop mode is exited according to Section : Exiting low-power mode.
Refer to Table 29 for more details on how to exit Stop mode.
When exiting Stop mode by issuing an interrupt or a wakeup event, the HSI RC oscillator is
selected as system clock.
If the Under-drive mode was enabled, it is automatically disabled after exiting Stop mode.
When the voltage regulator operates in low voltage mode, an additional startup delay is
incurred when waking up from Stop mode. By keeping the internal regulator ON during Stop
mode, the consumption is higher although the startup time is reduced.
When the voltage regulator operates in Under-drive mode, an additional startup delay is
induced when waking up from Stop mode.
RM0090 Rev 19 135/1751

149
Table 29. Stop mode entry and exit (STM32F42xxx and STM32F43xxx)

– No interrupt or event is pending,
– PDDS bit is cleared in Power Control register (PWR_CR),
– Select the voltage regulator mode by configuring LPDS, MRUDS, LPUDS
and UDEN bits in PWR_CR (see Table 28: Stop operating modes
(STM32F42xxx and STM32F43xxx)).
On Return from ISR while:

Mode entry – No interrupt is pending,
and
– PDDS is cleared in PWR_CR1.
Note: To enter Stop mode, all EXTI Line pending bits (in Pending register
(EXTI_PR)), all peripheral interrupts pending bits, the RTC Alarm
(Alarm A and Alarm B), RTC wakeup, RTC tamper, and RTC time
stamp flags, must be reset. Otherwise, the Stop mode entry
procedure is ignored and program execution continues.
All EXTI lines configured in Interrupt mode (the corresponding EXTI
Interrupt vector must be enabled in the NVIC). The interrupt source can
on page 372.
All EXTI Lines configured in event mode. Refer to Section 12.2.3:
Wakeup event management on page 380
Mode exit
– Any EXTI lines configured in Interrupt mode (even if the corresponding
EXTI Interrupt vector is disabled in the NVIC). The interrupt source can
STM32F43xxx.
– Wakeup event: refer to Section 12.2.3: Wakeup event management on
page 380.
Refer to Table 28: Stop operating modes (STM32F42xxx and
Wakeup latency
STM32F43xxx)
5.3.6 Standby mode

The Standby mode allows to achieve the lowest power consumption. It is based on the
Cortex®-M4 with FPU deepsleep mode, with the voltage regulator disabled. The 1.2 V
domain is consequently powered off. The PLLs, the HSI oscillator and the HSE oscillator are
also switched off. SRAM and register contents are lost except for registers in the backup
domain (RTC registers, RTC backup register and backup SRAM), and Standby circuitry (see
Figure 9).
136/1751 RM0090 Rev 19

Entering Standby mode

The Standby mode is entered according to Section : Entering low-power mode, when the
Refer to Table 30 for more details on how to enter Standby mode.
In Standby mode, the following features can be selected by programming individual control
bits:
hardware option. Once started it cannot be stopped except by a reset. See
• Real-time clock (RTC): this is configured by the RTCEN bit in the backup domain
control register (RCC_BDCR)
• Internal RC oscillator (LSI RC): this is configured by the LSION bit in the Control/status
register (RCC_CSR).
backup domain control register (RCC_BDCR)
Exiting Standby mode

The Standby mode is exited according to Section : Exiting low-power mode. The SBF status
flag in PWR_CR (see Section 5.4.2: PWR power control/status register (PWR_CSR) for
STM32F405xx/07xx and STM32F415xx/17xx) indicates that the MCU was in Standby
mode. All registers are reset after wakeup from Standby except for PWR_CR.
Refer to Table 30 for more details on how to exit Standby mode.
Table 30. Standby mode entry and exit

Standby mode Description

– SLEEPDEEP is set in Cortex®-M4 with FPU System Control register,
– PDDS bit is set in Power Control register (PWR_CR),
– No interrupt (for WFI) or event (for WFE) is pending,
– WUF bit is cleared in Power Control register (PWR_CR),
– the RTC flag corresponding to the chosen wakeup source (RTC Alarm A,
RTC Alarm B, RTC wakeup, Tamper or Timestamp flags) is cleared
Mode entry On return from ISR while:

and
– PDDS bit is set in Power Control register (PWR_CR), and
– No interrupt is pending,
– WUF bit is cleared in Power Control/Status register (PWR_SR),
– The RTC flag corresponding to the chosen wakeup source (RTC Alarm
A, RTC Alarm B, RTC wakeup, Tamper or Timestamp flags) is cleared.
WKUP pin rising edge, RTC alarm (Alarm A and Alarm B), RTC wakeup,
Mode exit
tamper event, time stamp event, external reset in NRST pin, IWDG reset.
Wakeup latency Reset phase.
RM0090 Rev 19 137/1751

149
I/O states in Standby mode

In Standby mode, all I/O pins are high impedance except for:
• Reset pad (still available)
• RTC_AF1 pin (PC13) if configured for tamper, time stamp, RTC Alarm out, or RTC
clock calibration out
• WKUP pin (PA0), if enabled
Debug mode
By default, the debug connection is lost if the application puts the MCU in Stop or Standby
mode while the debug features are used. This is due to the fact that the Cortex®-M4 with
FPU core is no longer clocked.
However, by setting some configuration bits in the DBGMCU_CR register, the software can
be debugged even when using the low-power modes extensively. For more details, refer to
Section 38.16.1: Debug support for low-power modes.
5.3.7 Programming the RTC alternate functions to wake up the device from
the Stop and Standby modes
The MCU can be woken up from a low-power mode by an RTC alternate function.
The RTC alternate functions are the RTC alarms (Alarm A and Alarm B), RTC wakeup, RTC
tamper event detection and RTC time stamp event detection.
These RTC alternate functions can wake up the system from the Stop and Standby low-
power modes.
The system can also wake up from low-power modes without depending on an external
interrupt (Auto-wakeup mode), by using the RTC alarm or the RTC wakeup events.
The RTC provides a programmable time base for waking up from the Stop or Standby mode
at regular intervals.
For this purpose, two of the three alternate RTC clock sources can be selected by
programming the RTCSEL[1:0] bits in the Section 7.3.20: RCC Backup domain control
register (RCC_BDCR):
• Low-power 32.768 kHz external crystal oscillator (LSE OSC)
This clock source provides a precise time base with a very low-power consumption
(additional consumption of less than 1 µA under typical conditions)
• Low-power internal RC oscillator (LSI RC)
This clock source has the advantage of saving the cost of the 32.768 kHz crystal. This
internal RC oscillator is designed to use minimum power.
138/1751 RM0090 Rev 19

RTC alternate functions to wake up the device from the Stop mode
• To wake up the device from the Stop mode with an RTC alarm event, it is necessary to:
a) Configure the EXTI Line 17 to be sensitive to rising edges (Interrupt or Event
modes)
b) Enable the RTC Alarm Interrupt in the RTC_CR register
c) Configure the RTC to generate the RTC alarm
• To wake up the device from the Stop mode with an RTC tamper or time stamp event, it
is necessary to:
modes)
b) Enable the RTC time stamp Interrupt in the RTC_CR register or the RTC tamper
interrupt in the RTC_TAFCR register
c) Configure the RTC to detect the tamper or time stamp event
• To wake up the device from the Stop mode with an RTC wakeup event, it is necessary
to:
modes)
b) Enable the RTC wakeup interrupt in the RTC_CR register
c) Configure the RTC to generate the RTC Wakeup event
RTC alternate functions to wake up the device from the Standby mode
• To wake up the device from the Standby mode with an RTC alarm event, it is necessary
to:
a) Enable the RTC alarm interrupt in the RTC_CR register
b) Configure the RTC to generate the RTC alarm
• To wake up the device from the Standby mode with an RTC tamper or time stamp
event, it is necessary to:
a) Enable the RTC time stamp interrupt in the RTC_CR register or the RTC tamper
interrupt in the RTC_TAFCR register
b) Configure the RTC to detect the tamper or time stamp event
• To wake up the device from the Standby mode with an RTC wakeup event, it is
necessary to:
a) Enable the RTC wakeup interrupt in the RTC_CR register
b) Configure the RTC to generate the RTC wakeup event
Safe RTC alternate function wakeup flag clearing sequence

If the selected RTC alternate function is set before the PWR wakeup flag (WUTF) is cleared,
it will not be detected on the next event as detection is made once on the rising edge.
To avoid bouncing on the pins onto which the RTC alternate functions are mapped, and exit
correctly from the Stop and Standby modes, it is recommended to follow the sequence
below before entering the Standby mode:
• When using RTC alarm to wake up the device from the low-power modes:
a) Disable the RTC alarm interrupt (ALRAIE or ALRBIE bits in the RTC_CR register)
b) Clear the RTC alarm (ALRAF/ALRBF) flag
RM0090 Rev 19 139/1751

149
c) Clear the PWR Wakeup (WUF) flag

d) Enable the RTC alarm interrupt
e) Re-enter the low-power mode
• When using RTC wakeup to wake up the device from the low-power modes:
a) Disable the RTC Wakeup interrupt (WUTIE bit in the RTC_CR register)
b) Clear the RTC Wakeup (WUTF) flag
d) Enable the RTC Wakeup interrupt
• When using RTC tamper to wake up the device from the low-power modes:
a) Disable the RTC tamper interrupt (TAMPIE bit in the RTC_TAFCR register)
b) Clear the Tamper (TAMP1F/TSF) flag
d) Enable the RTC tamper interrupt
• When using RTC time stamp to wake up the device from the low-power modes:
a) Disable the RTC time stamp interrupt (TSIE bit in RTC_CR)
b) Clear the RTC time stamp (TSF) flag
d) Enable the RTC TimeStamp interrupt
140/1751 RM0090 Rev 19

5.4 Power control registers (STM32F405xx/07xx and

STM32F415xx/17xx)

Reset value: 0x0000 4000 (reset by wakeup from Standby mode)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
VOS FPDS DBP PLS[2:0] PVDE CSBF CWUF PDDS LPDS
Res. Reserved
rw rw rw rw rw rw rw w w rw rw

Bit 14 VOS: Regulator voltage scaling output selection
This bit controls the main internal voltage regulator output voltage to achieve a trade-off
between performance and power consumption when the device does not operate at the
maximum frequency.
0: Scale 2 mode
1: Scale 1 mode (default value at reset)
Bit 9 FPDS: Flash power-down in Stop mode
When set, the Flash memory enters power-down mode when the device enters Stop mode.
This allows to achieve a lower consumption in stop mode but a longer restart time.
0: Flash memory not in power-down when the device is in Stop mode
1: Flash memory in power-down when the device is in Stop mode
Bit 8 DBP: Disable backup domain write protection
In reset state, the RCC_BDCR register, the RTC registers (including the backup registers),
and the BRE bit of the PWR_CSR register, are protected against parasitic write access. This
bit must be set to enable write access to these registers.
0: Access to RTC and RTC Backup registers and backup SRAM disabled
1: Access to RTC and RTC Backup registers and backup SRAM enabled
Bits 7:5 PLS[2:0]: PVD level selection
These bits are written by software to select the voltage threshold detected by the Power
Voltage Detector
000: 2.0 V
001: 2.1 V
010: 2.3 V
011: 2.5 V
100: 2.6 V
101: 2.7 V
110: 2.8 V
111: 2.9 V
Note: Refer to the electrical characteristics of the datasheet for more details.
RM0090 Rev 19 141/1751

149
Bit 4 PVDE: Power voltage detector enable

This bit is set and cleared by software.
0: PVD disabled
1: PVD enabled
Bit 3 CSBF: Clear standby flag
This bit is always read as 0.
0: No effect
1: Clear the SBF Standby Flag (write).
Bit 2 CWUF: Clear wakeup flag
0: No effect
1: Clear the WUF Wakeup Flag after 2 System clock cycles
Bit 1 PDDS: Power-down deepsleep
This bit is set and cleared by software. It works together with the LPDS bit.
0: Enter Stop mode when the CPU enters deepsleep. The regulator status depends on the
LPDS bit.
1: Enter Standby mode when the CPU enters deepsleep.
Bit 0 LPDS: Low-power deepsleep
This bit is set and cleared by software. It works together with the PDDS bit.
0: Voltage regulator on during Stop mode
1: Voltage regulator in low-power mode during Stop mode

Reset value: 0x0000 0000 (not reset by wakeup from Standby mode)
Additional APB cycles are needed to read this register versus a standard APB read.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
VOS
BRE EWUP BRR PVDO SBF WUF
Res RDY Reserved Reserved
r rw rw r r r r

Bit 14 VOSRDY: Regulator voltage scaling output selection ready bit
0: Not ready
1: Ready
142/1751 RM0090 Rev 19


Bit 9 BRE: Backup regulator enable
When set, the Backup regulator (used to maintain backup SRAM content in Standby and
VBAT modes) is enabled. If BRE is reset, the backup regulator is switched off. The backup
SRAM can still be used but its content will be lost in the Standby and VBAT modes. Once set,
the application must wait that the Backup Regulator Ready flag (BRR) is set to indicate that
the data written into the RAM will be maintained in the Standby and VBAT modes.
0: Backup regulator disabled
1: Backup regulator enabled
Note: This bit is not reset when the device wakes up from Standby mode, by a system reset,
or by a power reset.
Bit 8 EWUP: Enable WKUP pin
0: WKUP pin is used for general purpose I/O. An event on the WKUP pin does not wakeup
the device from Standby mode.
1: WKUP pin is used for wakeup from Standby mode and forced in input pull down
configuration (rising edge on WKUP pin wakes-up the system from Standby mode).
Note: This bit is reset by a system reset.
Bit 3 BRR: Backup regulator ready
Set by hardware to indicate that the Backup Regulator is ready.
0: Backup Regulator not ready
1: Backup Regulator ready
Note: This bit is not reset when the device wakes up from Standby mode or by a system reset
or power reset.
Bit 2 PVDO: PVD output
This bit is set and cleared by hardware. It is valid only if PVD is enabled by the PVDE bit.
0: VDD is higher than the PVD threshold selected with the PLS[2:0] bits.
1: VDD is lower than the PVD threshold selected with the PLS[2:0] bits.
Note: The PVD is stopped by Standby mode. For this reason, this bit is equal to 0 after
Standby or reset until the PVDE bit is set.
Bit 1 SBF: Standby flag
This bit is set by hardware and cleared only by a POR/PDR (power-on reset/power-down
reset) or by setting the CSBF bit in the PWR_CR register.
0: Device has not been in Standby mode
1: Device has been in Standby mode
Bit 0 WUF: Wakeup flag
This bit is set by hardware and cleared either by a system reset or by setting the CWUF bit in
the PWR_CR register.
0: No wakeup event occurred
1: A wakeup event was received from the WKUP pin or from the RTC alarm (Alarm A or
Alarm B), RTC Tamper event, RTC TimeStamp event or RTC Wakeup).
Note: An additional wakeup event is detected if the WKUP pin is enabled (by setting the
EWUP bit) when the WKUP pin level is already high.
RM0090 Rev 19 143/1751

149
5.5 Power control registers (STM32F42xxx and STM32F43xxx)

Reset value: 0x0000 C000 (reset by wakeup from Standby mode)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
ODSWE
UDEN[1:0] ODEN
Reserved N
rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
VOS[1:0] ADCDC1 MRUDS LPUDS FPDS DBP PLS[2:0] PVDE CSBF CWUF PDDS LPDS
Res.
rw rw rw rw rw rw rw rw rw rw rw rc_w1 rc_w1 rw rw

Bits 19:18 UDEN[1:0]: Under-drive enable in stop mode
These bits are set by software. They allow to achieve a lower power consumption in Stop
mode but with a longer wakeup time.
When set, the digital area has less leakage consumption when the device enters Stop mode.
00: Under-drive disable
01: Reserved
10: Reserved
11:Under-drive enable
Bit 17 ODSWEN: Over-drive switching enabled.
This bit is set by software. It is cleared automatically by hardware after exiting from Stop
mode or when the ODEN bit is reset. When set, It is used to switch to Over-drive mode.
To set or reset the ODSWEN bit, the HSI or HSE must be selected as system clock.
The ODSWEN bit must only be set when the ODRDY flag is set to switch to Over-drive
mode.
0: Over-drive switching disabled
1: Over-drive switching enabled
Note: On any over-drive switch (enabled or disabled), the system clock will be stalled during
the internal voltage set up.
Bit 16 ODEN: Over-drive enable
This bit is set by software. It is cleared automatically by hardware after exiting from Stop
mode. It is used to enabled the Over-drive mode in order to reach a higher frequency.
To set or reset the ODEN bit, the HSI or HSE must be selected as system clock. When the
ODEN bit is set, the application must first wait for the Over-drive ready flag (ODRDY) to be
set before setting the ODSWEN bit.
0: Over-drive disabled
1: Over-drive enabled
144/1751 RM0090 Rev 19

Bits 15:14 VOS[1:0]: Regulator voltage scaling output selection

These bits control the main internal voltage regulator output voltage to achieve a trade-off
between performance and power consumption when the device does not operate at the
maximum frequency (refer to the STM32F42xx and STM32F43xx datasheets for more
details).
These bits can be modified only when the PLL is OFF. The new value programmed is active
only when the PLL is ON. When the PLL is OFF, the voltage scale 3 is automatically
selected.
00: Reserved (Scale 3 mode selected)
01: Scale 3 mode
10: Scale 2 mode
11: Scale 1 mode (reset value)
Bit 13 ADCDC1:
0: No effect.
1: Refer to AN4073 for details on how to use this bit.
Note: This bit can only be set when operating at supply voltage range 2.7 to 3.6V and when
the Prefetch is OFF.
Bit 12 Reserved, must be kept at reset value.
Bit 11 MRUDS: Main regulator in deepsleep under-drive mode
0: Main regulator ON when the device is in Stop mode
1: Main Regulator in under-drive mode and Flash memory in power-down when the device is
in Stop under-drive mode.
Bit 10 LPUDS: Low-power regulator in deepsleep under-drive mode
0: Low-power regulator ON if LPDS bit is set when the device is in Stop mode
1: Low-power regulator in under-drive mode if LPDS bit is set and Flash memory in power-
down when the device is in Stop under-drive mode.
Bit 9 FPDS: Flash power-down in Stop mode
When set, the Flash memory enters power-down mode when the device enters Stop mode.
This allows to achieve a lower consumption in stop mode but a longer restart time.
0: Flash memory not in power-down when the device is in Stop mode
1: Flash memory in power-down when the device is in Stop mode
Bit 8 DBP: Disable backup domain write protection
In reset state, the RCC_BDCR register, the RTC registers (including the backup registers),
and the BRE bit of the PWR_CSR register, are protected against parasitic write access. This
bit must be set to enable write access to these registers.
0: Access to RTC and RTC Backup registers and backup SRAM disabled
1: Access to RTC and RTC Backup registers and backup SRAM enabled
RM0090 Rev 19 145/1751

149
Bits 7:5 PLS[2:0]: PVD level selection

These bits are written by software to select the voltage threshold detected by the Power
Voltage Detector
000: 2.0 V
001: 2.1 V
010: 2.3 V
011: 2.5 V
100: 2.6 V
101: 2.7 V
110: 2.8 V
111: 2.9 V
Note: Refer to the electrical characteristics of the datasheet for more details.
Bit 4 PVDE: Power voltage detector enable
0: PVD disabled
1: PVD enabled
Bit 3 CSBF: Clear standby flag
0: No effect
1: Clear the SBF Standby Flag (write).
Bit 2 CWUF: Clear wakeup flag
0: No effect
1: Clear the WUF Wakeup Flag after 2 System clock cycles
Bit 1 PDDS: Power-down deepsleep
This bit is set and cleared by software. It works together with the LPDS bit.
0: Enter Stop mode when the CPU enters deepsleep. The regulator status depends on the
LPDS bit.
1: Enter Standby mode when the CPU enters deepsleep.
Bit 0 LPDS: Low-power deepsleep
This bit is set and cleared by software. It works together with the PDDS bit.
0:Main voltage regulator ON during Stop mode
1: Low-power voltage regulator ON during Stop mode
146/1751 RM0090 Rev 19


Reset value: 0x0000 0000 (not reset by wakeup from Standby mode)
Additional APB cycles are needed to read this register versus a standard APB read.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
UDRDY[1:0] ODSWRDY ODRDY

Reserved
rc_w1 rc_w1 r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
VOS
BRE EWUP BRR PVDO SBF WUF
Res RDY Reserved Reserved.
r rw rw r r r r

Bits 19:18 UDRDY[1:0]: Under-drive ready flag
These bits are set by hardware when MCU entered stop Under-drive mode and exited.
When the under-drive mode is enabled, these bits are not set as long as the MCU has not
entered stop mode yet. They are cleared by programming them to 1.
00: Under-drive is disabled
01: Reserved
10: Reserved
11:Under-drive mode is activated in Stop mode.
Bit 17 ODSWRDY: Over-drive mode switching ready
0: Over-drive mode is not active.
1: Over-drive mode is active on digital area on 1.2 V domain
Bit 16 ODRDY: Over-drive mode ready
0: Over-drive mode not ready.
1: Over-drive mode ready
Bit 14 VOSRDY: Regulator voltage scaling output selection ready bit
0: Not ready
1: Ready
Bit 9 BRE: Backup regulator enable
When set, the Backup regulator (used to maintain backup SRAM content in Standby and
VBAT modes) is enabled. If BRE is reset, the backup regulator is switched off. The backup
SRAM can still be used but its content will be lost in the Standby and VBAT modes. Once set,
the application must wait that the Backup Regulator Ready flag (BRR) is set to indicate that
the data written into the RAM will be maintained in the Standby and VBAT modes.
0: Backup regulator disabled
1: Backup regulator enabled
Note: This bit is not reset when the device wakes up from Standby mode, by a system reset,
or by a power reset.
RM0090 Rev 19 147/1751

149
Bit 8 EWUP: Enable WKUP pin

0: WKUP pin is used for general purpose I/O. An event on the WKUP pin does not wakeup
the device from Standby mode.
1: WKUP pin is used for wakeup from Standby mode and forced in input pull down
configuration (rising edge on WKUP pin wakes-up the system from Standby mode).
Note: This bit is reset by a system reset.
Bit 3 BRR: Backup regulator ready
Set by hardware to indicate that the Backup Regulator is ready.
0: Backup Regulator not ready
1: Backup Regulator ready
Note: This bit is not reset when the device wakes up from Standby mode or by a system reset
or power reset.
Bit 2 PVDO: PVD output
This bit is set and cleared by hardware. It is valid only if PVD is enabled by the PVDE bit.
0: VDD is higher than the PVD threshold selected with the PLS[2:0] bits.
1: VDD is lower than the PVD threshold selected with the PLS[2:0] bits.
Note: The PVD is stopped by Standby mode. For this reason, this bit is equal to 0 after
Standby or reset until the PVDE bit is set.
Bit 1 SBF: Standby flag
This bit is set by hardware and cleared only by a POR/PDR (power-on reset/power-down
reset) or by setting the CSBF bit in the PWR power control register (PWR_CR) for
0: Device has not been in Standby mode
1: Device has been in Standby mode
Bit 0 WUF: Wakeup flag
This bit is set by hardware and cleared either by a system reset or by setting the CWUF bit in
the PWR_CR register.
0: No wakeup event occurred
1: A wakeup event was received from the WKUP pin or from the RTC alarm (Alarm A or
Alarm B), RTC Tamper event, RTC TimeStamp event or RTC Wakeup).
Note: An additional wakeup event is detected if the WKUP pin is enabled (by setting the
EWUP bit) when the WKUP pin level is already high.
148/1751 RM0090 Rev 19

5.6 PWR register map

The following table summarizes the PWR registers.
Table 31. PWR - register map and reset values for STM32F405xx/07xx and STM32F415xx/17xx
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
CWUF
PDDS
PVDE
FPDS
CSBF
LPDS
VOS
DBP
PWR_CR PLS[2:0]
Reset value 1 0 0 0 0 0 0 0 0 0 0
VOSRDY
EWUP
PVDO
WUF
BRR
BRE
SBF
PWR_CSR
Table 32. PWR - register map and reset values for STM32F42xxx and STM32F43xxx
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
UDEN[1:0]
ODSWEN
ADCDC1
VOS[1:0]
MRUDS
LPUDS
Reserved
CWUF
ODEN
PDDS
PVDE
CSBF
FPDS
LPDS
DBP
PWR_CR PLS[2:0]
0x000 Reserved
Reset value 1 1 1 1 1 1 0 1 1 0 0 0 0 0 0 0 0 0 0
UDRDY[1:0]
ODSWRDY
VOSRDY
ODRDY
EWUP
Reserved
PVDO
WUF
BRR
BRE
SBF
PWR_CSR
Reset value 0 0 0 0 0 0 0 0 0 0 0
Refer to Section 2.3: Memory mapfor the register boundary addresses.
RM0090 Rev 19 149/1751

149
Reset and clock control for STM32F42xxx and STM32F43xxx (RCC) RM0090

STM32F42xxx and STM32F43xxx (RCC)
6.1 Reset
There are three types of reset, defined as system Reset, power Reset and backup domain
Reset.
6.1.1 System reset

A system reset sets all registers to their reset values except the reset flags in the clock
controller CSR register and the registers in the Backup domain (see Figure 4).
A system reset is generated when one of the following events occurs:
1. A low level on the NRST pin (external reset)
2. Window watchdog end of count condition (WWDG reset)
3. Independent watchdog end of count condition (IWDG reset)
4. A software reset (SW reset) (see Software reset)
5. Low-power management reset (see Low-power management reset)
Software reset
The reset source can be identified by checking the reset flags in the RCC clock control &
status register (RCC_CSR).
The SYSRESETREQ bit in Cortex®-M4 with FPU Application Interrupt and Reset Control
Register must be set to force a software reset on the device. Refer to the Cortex®-M4 with
FPU technical reference manual for more details.
Low-power management reset

There are two ways of generating a low-power management reset:
1. Reset generated when entering the Standby mode:
This type of reset is enabled by resetting the nRST_STDBY bit in the user option bytes.
In this case, whenever a Standby mode entry sequence is successfully executed, the
device is reset instead of entering the Standby mode.
2. Reset when entering the Stop mode:
This type of reset is enabled by resetting the nRST_STOP bit in the user option bytes.
In this case, whenever a Stop mode entry sequence is successfully executed, the
device is reset instead of entering the Stop mode.
6.1.2 Power reset

A power reset is generated when one of the following events occurs:
1. Power-on/power-down reset (POR/PDR reset) or brownout (BOR) reset
2. When exiting the Standby mode
A power reset sets all registers to their reset values except the Backup domain (see
Figure 4)
150/1751 RM0090 Rev 19

RM0090 Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)
These sources act on the NRST pin and it is always kept low during the delay phase. The
RESET service routine vector is fixed at address 0x0000_0004 in the memory map.
The system reset signal provided to the device is output on the NRST pin. The pulse
generator guarantees a minimum reset pulse duration of 20 µs for each internal reset
source. In case of an external reset, the reset pulse is generated while the NRST pin is
asserted low.
Figure 15. Simplified diagram of the reset circuit
VDD/VDDA
RPU
External System reset
reset Filter
NRST
WWDG reset
Pulse
IWDG reset
generator
Power reset
(min 20 μs)
Software reset
ai16095c
The Backup domain has two specific resets that affect only the Backup domain (see
Figure 4).
6.1.3 Backup domain reset

The backup domain reset sets all RTC registers and the RCC_BDCR register to their reset
values. The BKPSRAM is not affected by this reset. The only way of resetting the
BKPSRAM is through the Flash interface by requesting a protection level change from 1 to
0.
A backup domain reset is generated when one of the following events occurs:
1. Software reset, triggered by setting the BDRST bit in the RCC Backup domain control
register (RCC_BDCR).
2. VDD or VBAT power on, if both supplies have previously been powered off.
6.2 Clocks
Three different clock sources can be used to drive the system clock (SYSCLK):
• HSI oscillator clock
• HSE oscillator clock
• Main PLL (PLL) clock
The devices have the two following secondary clock sources:
• 32 kHz low-speed internal RC (LSI RC) which drives the independent watchdog and,
optionally, the RTC used for Auto-wakeup from the Stop/Standby mode.
• 32.768 kHz low-speed external crystal (LSE crystal) which optionally drives the RTC
clock (RTCCLK)
RM0090 Rev 19 151/1751

212
Each clock source can be switched on or off independently when it is not used, to optimize
power consumption.
Figure 16. Clock tree

Watchdog IWDGCLK
LSI enable To Independent
LSI RC watchdog
32 kHz
RTCSEL[1:0]
RTC
OSC32_IN RTCCLK
LSE OSC LSE enable to RTC
32.768 kHz
OSC32_OUT
SYSCLK
MCO2 /1 to 5
HSE_RTC
LSE
Peripheral Ethernet
MCO1 /1 to 5 /2 to 31 clock enable PTP clock
180 MHz max. HCLK

to AHB bus, core,
16 MHz HSI Clock
memory and DMA
HSI RC enable
SW /8 to Cortex System timer
HSI
HSE AHB FCLK Cortex
PRESC free-running clock
PLLCLK /1,2,..512
SYSCLK Peripheral clock

OSC_OUT HSE 180 MHz
4-26 MHz APBx enable APBx peripheral clocks
HSE OSC max PRESC
OSC_IN /1,2,4,8,16
Peripheral clock
if (APBx presc = enable APBx timer clocks
1x1 else x2
/M
VCO /P Peripheral
PLL48CK 48 MHz
/Q clock enable
xN clocks
PLL /R
VCO /P
/Q
I2SSRC
xN
PLLI2SCLK Peripheral
/R clock enable I2S clocks
PLLI2S Ext. clock
I2S_CKIN
Peripheral SAI1_A clock
PLLSAICLK clock enable
DIV
Peripheral SAI1_B clock
VCO /P
clock enable
PLLSAICLK
/Q DIV
xN Peripheral LCD-TFT
PLLLCDCLK clock enable clock
/R DIV
PLLSAI
ETH_MII_TX_CLK_MII Peripheral
clock enable MACTXCLK
PHY Ethernet
25 to 50 MHz
/2,20 MII_RMII_SEL in SYSCFG_PMC to Ethernet MAC
Peripheral
clock enable MACRXCLK
ETH_MII_RX_CLK_MI
I Peripheral
clock enable MACRMIICLK
Peripheral
clock enable USBHS
USB2.0 PHY OTG_HS_ULPI_CK
ULPI clock
24 to 60 MHz
MS30434V2
1. For full details about the internal and external clock source characteristics, refer to the Electrical characteristics section in
2. When TIMPRE bit of the RCC_DCKCFGR register is reset, if APBx prescaler is 1, then TIMxCLK = PCLKx, otherwise
TIMxCLK = 2x PCLKx.
3. When TIMPRE bit in the RCC_DCKCFGR register is set, if APBx prescaler is 1,2 or 4, then TIMxCLK = HCLK, otherwise
TIMxCLK = 4x PCLKx.
152/1751 RM0090 Rev 19

The clock controller provides a high degree of flexibility to the application in the choice of the
external crystal or the oscillator to run the core and peripherals at the highest frequency
and, guarantee the appropriate frequency for peripherals that need a specific clock like
Ethernet, USB OTG FS and HS, I2S, SAI, LTDC, and SDIO.
Several prescalers are used to configure the AHB frequency, the high-speed APB (APB2)
and the low-speed APB (APB1) domains. The maximum frequency of the AHB domain is
180 MHz. The maximum allowed frequency of the high-speed APB2 domain is 90 MHz. The
maximum allowed frequency of the low-speed APB1 domain is 45 MHz
All peripheral clocks are derived from the system clock (SYSCLK) except for:
• The USB OTG FS clock (48 MHz), the random analog generator (RNG) clock
(≤ 48 MHz) and the SDIO clock (≤ 48 MHz) which are coming from a specific output of
PLL (PLL48CLK)
• I2S clock
To achieve high-quality audio performance, the I2S clock can be derived either from a
specific PLL (PLLI2S) or from an external clock mapped on the I2S_CKIN pin. For
more information about I2S clock frequency and precision, refer to Section 28.4.4:
Clock generator.
• SAI1 clock
The SAI1 clock is generated from a specific PLL (PLLSAI or PLLI2S) or from an
external clock mapped on the I2S_CKIN pin.
The PLLSAI can be used as clock source for SAI1 peripheral in case the PLLI2S is
programmed to achieve another audio sampling frequency (49.152 MHz or
11.2896 MHz), and the application requires both frequencies at the same time.
• LTDC clock
The LTDC clock is generated from a specific PLL (PLLSAI).
• The USB OTG HS (60 MHz) clock which is provided from the external PHY
• The Ethernet MAC clocks (TX, RX and RMII) which are provided from the external
PHY. For further information on the Ethernet configuration, please refer to
Section 33.4.4: MII/RMII selection in the Ethernet peripheral description. When the
Ethernet is used, the AHB clock frequency must be at least 25 MHz.
The RCC feeds the external clock of the Cortex System Timer (SysTick) with the AHB clock
(HCLK) divided by 8. The SysTick can work either with this clock or with the Cortex clock
(HCLK), configurable in the SysTick control and status register.
The timer clock frequencies for STM32F42xxx and STM32F43xxx are automatically set by
hardware. There are two cases depending on the value of TIMPRE bit in RCC_CFGR
register:
• If TIMPRE bit in RCC_DKCFGR register is reset:
If the APB prescaler is configured to a division factor of 1, the timer clock frequencies
(TIMxCLK) are set to PCLKx. Otherwise, the timer clock frequencies are twice the
frequency of the APB domain to which the timers are connected: TIMxCLK = 2xPCLKx.
• If TIMPRE bit in RCC_DKCFGR register is set:
If the APB prescaler is configured to a division factor of 1, 2 or 4, the timer clock
frequencies (TIMxCLK) are set to HCLK. Otherwise, the timer clock frequencies is four
times the frequency of the APB domain to which the timers are connected: TIMxCLK =
4xPCLKx.
RM0090 Rev 19 153/1751

212
FCLK acts as Cortex®-M4 with FPU free-running clock. For more details, refer to the
Cortex®-M4 with FPU technical reference manual.
6.2.1 HSE clock

The high speed external clock signal (HSE) can be generated from two possible clock
sources:
• HSE external crystal/ceramic resonator
• HSE external user clock
The resonator and the load capacitors have to be placed as close as possible to the
oscillator pins in order to minimize output distortion and startup stabilization time. The
loading capacitance values must be adjusted according to the selected oscillator.
Figure 17. HSE/ LSE clock sources

Hardware configuration
OSC_OUT
External clock
(HI-Z)
External
source
OSC_IN OSC_OUT
Crystal/ceramic
resonators
CL1 CL2
Load
capacitors
External source (HSE bypass)

In this mode, an external clock source must be provided. You select this mode by setting the
HSEBYP and HSEON bits in the RCC clock control register (RCC_CR). The external clock
signal (square, sinus or triangle) with ~50% duty cycle has to drive the OSC_IN pin while the
OSC_OUT pin should be left HI-Z. See Figure 17.
External crystal/ceramic resonator (HSE crystal)

The HSE has the advantage of producing a very accurate rate on the main clock.
The associated hardware configuration is shown in Figure 17. Refer to the electrical
characteristics section of the datasheet for more details.
The HSERDY flag in the RCC clock control register (RCC_CR) indicates if the high-speed
external oscillator is stable or not. At startup, the clock is not released until this bit is set by
hardware. An interrupt can be generated if enabled in the RCC clock interrupt register
(RCC_CIR).
154/1751 RM0090 Rev 19

The HSE Crystal can be switched on and off using the HSEON bit in the RCC clock control
register (RCC_CR).
6.2.2 HSI clock

The HSI clock signal is generated from an internal 16 MHz RC oscillator and can be used
directly as a system clock, or used as PLL input.
The HSI RC oscillator has the advantage of providing a clock source at low cost (no external
components). It also has a faster startup time than the HSE crystal oscillator however, even
with calibration the frequency is less accurate than an external crystal oscillator or ceramic
resonator.
Calibration
RC oscillator frequencies can vary from one chip to another due to manufacturing process
variations, this is why each device is factory calibrated by ST for 1% accuracy at TA= 25 °C.
After reset, the factory calibration value is loaded in the HSICAL[7:0] bits in the RCC clock
control register (RCC_CR).
If the application is subject to voltage or temperature variations this may affect the RC
oscillator speed. You can trim the HSI frequency in the application using the HSITRIM[4:0]
bits in the RCC clock control register (RCC_CR).
The HSIRDY flag in the RCC clock control register (RCC_CR) indicates if the HSI RC is
stable or not. At startup, the HSI RC output clock is not released until this bit is set by
hardware.
The HSI RC can be switched on and off using the HSION bit in the RCC clock control
register (RCC_CR).
The HSI signal can also be used as a backup source (Auxiliary clock) if the HSE crystal
oscillator fails. Refer to Section 6.2.7: Clock security system (CSS) on page 157.
6.2.3 PLL configuration

The STM32F4xx devices feature three PLLs:
• A main PLL (PLL) clocked by the HSE or HSI oscillator and featuring two different
output clocks:
– The first output is used to generate the high speed system clock (up to 180 MHz)
– The second output is used to generate the clock for the USB OTG FS (48 MHz),
the random analog generator (≤ 48 MHz) and the SDIO (≤ 48 MHz).
• Two dedicated PLLs (PLLI2S and PLLSAI) used to generate an accurate clock to
achieve high-quality audio performance on the I2S and SAI1 interfaces. PLLSAI is also
used for the LCD-TFT clock.
Since the main-PLL configuration parameters cannot be changed once PLL is enabled, it is
recommended to configure PLL before enabling it (selection of the HSI or HSE oscillator as
PLL clock source, and configuration of division factors M, N, P, and Q).
The PLLI2S and PLLSAI use the same input clock as PLL (PLLM[5:0] and PLLSRC bits are
common to both PLLs). However, the PLLI2S and PLLSAI have dedicated enable/disable
and division factors (N and R) configuration bits. Once the PLLI2S and PLLSAI are enabled,
the configuration parameters cannot be changed.
RM0090 Rev 19 155/1751

212
The three PLLs are disabled by hardware when entering Stop and Standby modes, or when
an HSE failure occurs when HSE or PLL (clocked by HSE) are used as system clock. RCC
PLL configuration register (RCC_PLLCFGR),RCC clock configuration register
(RCC_CFGR), and RCC Dedicated Clock Configuration Register (RCC_DCKCFGR) can be
used to configure PLL, PLLI2S, and PLLSAI.
6.2.4 LSE clock

The LSE clock is generated from a 32.768 kHz low-speed external crystal or ceramic
resonator. It has the advantage providing a low-power but highly accurate clock source to
the real-time clock peripheral (RTC) for clock/calendar or other timing functions.
The LSE oscillator is switched on and off using the LSEON bit in RCC Backup domain
control register (RCC_BDCR).
The LSERDY flag in the RCC Backup domain control register (RCC_BDCR) indicates if the
LSE crystal is stable or not. At startup, the LSE crystal output clock signal is not released
until this bit is set by hardware. An interrupt can be generated if enabled in the RCC clock
interrupt register (RCC_CIR).
External source (LSE bypass)

In this mode, an external clock source must be provided. It must have a frequency up to
1 MHz. You select this mode by setting the LSEBYP and LSEON bits in the RCC Backup
domain control register (RCC_BDCR). The external clock signal (square, sinus or triangle)
with ~50% duty cycle has to drive the OSC32_IN pin while the OSC32_OUT pin should be
left HI-Z. See Figure 17.
6.2.5 LSI clock

The LSI RC acts as an low-power clock source that can be kept running in Stop and
Standby mode for the independent watchdog (IWDG) and Auto-wakeup unit (AWU). The
clock frequency is around 32 kHz. For more details, refer to the electrical characteristics
section of the datasheets.
The LSI RC can be switched on and off using the LSION bit in the RCC clock control &
The LSIRDY flag in the RCC clock control & status register (RCC_CSR) indicates if the low-
speed internal oscillator is stable or not. At startup, the clock is not released until this bit is
set by hardware. An interrupt can be generated if enabled in the RCC clock interrupt register
(RCC_CIR).
6.2.6 System clock (SYSCLK) selection

After a system reset, the HSI oscillator is selected as the system clock. When a clock source
is used directly or through PLL as the system clock, it is not possible to stop it.
A switch from one clock source to another occurs only if the target clock source is ready
(clock stable after startup delay or PLL locked). If a clock source that is not yet ready is
selected, the switch occurs when the clock source is ready. Status bits in the RCC clock
control register (RCC_CR) indicate which clock(s) is (are) ready and which clock is currently
used as the system clock.
156/1751 RM0090 Rev 19

6.2.7 Clock security system (CSS)

The clock security system can be activated by software. In this case, the clock detector is
enabled after the HSE oscillator startup delay, and disabled when this oscillator is stopped.
If a failure is detected on the HSE clock, this oscillator is automatically disabled, a clock
failure event is sent to the break inputs of advanced-control timers TIM1 and TIM8, and an
interrupt is generated to inform the software about the failure (clock security system
interrupt CSSI), allowing the MCU to perform rescue operations. The CSSI is linked to the
Cortex®-M4 with FPU NMI (non-maskable interrupt) exception vector.
Note: When the CSS is enabled, if the HSE clock happens to fail, the CSS generates an interrupt,
which causes the automatic generation of an NMI. The NMI is executed indefinitely unless
the CSS interrupt pending bit is cleared. As a consequence, the application has to clear the
CSS interrupt in the NMI ISR by setting the CSSC bit in the Clock interrupt register
(RCC_CIR).
If the HSE oscillator is used directly or indirectly as the system clock (indirectly meaning that
it is directly used as PLL input clock, and that PLL clock is the system clock) and a failure is
detected, then the system clock switches to the HSI oscillator and the HSE oscillator is
disabled.
If the HSE oscillator clock was the clock source of PLL used as the system clock when the
failure occurred, PLL is also disabled. In this case, if the PLLI2S was enabled, it is also
disabled when the HSE fails.
6.2.8 RTC/AWU clock

Once the RTCCLK clock source has been selected, the only possible way of modifying the
selection is to reset the power domain.
The RTCCLK clock source can be either the HSE 1 MHz (HSE divided by a programmable
prescaler), the LSE or the LSI clock. This is selected by programming the RTCSEL[1:0] bits
in the RCC Backup domain control register (RCC_BDCR) and the RTCPRE[4:0] bits in RCC
clock configuration register (RCC_CFGR). This selection cannot be modified without
resetting the Backup domain.
If the LSE is selected as the RTC clock, the RTC operates normally if the backup or the
system supply disappears. If the LSI is selected as the AWU clock, the AWU state is not
guaranteed if the system supply disappears. If the HSE oscillator divided by a value
between 2 and 31 is used as the RTC clock, the RTC state is not guaranteed if the backup
or the system supply disappears.
The LSE clock is in the Backup domain, whereas the HSE and LSI clocks are not. As a
consequence:
• If LSE is selected as the RTC clock:
– The RTC continues to work even if the VDD supply is switched off, provided the
VBAT supply is maintained.
• If LSI is selected as the Auto-wakeup unit (AWU) clock:
– The AWU state is not guaranteed if the VDD supply is powered off. Refer to
Section 6.2.5: LSI clock on page 156 for more details on LSI calibration.
• If the HSE clock is used as the RTC clock:
– The RTC state is not guaranteed if the VDD supply is powered off or if the internal
voltage regulator is powered off (removing power from the 1.2 V domain).
RM0090 Rev 19 157/1751

212
Note: To read the RTC calendar register when the APB1 clock frequency is less than seven times
the RTC clock frequency (fAPB1 < 7xfRTCLCK), the software must read the calendar time and
date registers twice. The data are correct if the second read access to RTC_TR gives the
same result than the first one. Otherwise a third read access must be performed.
6.2.9 Watchdog clock

If the independent watchdog (IWDG) is started by either hardware option or software
access, the LSI oscillator is forced ON and cannot be disabled. After the LSI oscillator
temporization, the clock is provided to the IWDG.
6.2.10 Clock-out capability

Two microcontroller clock output (MCO) pins are available:
• MCO1
You can output four different clock sources onto the MCO1 pin (PA8) using the
configurable prescaler (from 1 to 5):
– HSI clock
– LSE clock
– HSE clock
– PLL clock
The desired clock source is selected using the MCO1PRE[2:0] and MCO1[1:0] bits in
the RCC clock configuration register (RCC_CFGR).
• MCO2
You can output four different clock sources onto the MCO2 pin (PC9) using the
– HSE clock
– PLL clock
– System clock (SYSCLK)
– PLLI2S clock
The desired clock source is selected using the MCO2PRE[2:0] and MCO2 bits in the
RCC clock configuration register (RCC_CFGR).
For the different MCO pins, the corresponding GPIO port has to be programmed in alternate
function mode.
The selected clock to output onto MCO must not exceed 100 MHz (the maximum I/O
speed).
6.2.11 Internal/external clock measurement using TIM5/TIM11

It is possible to indirectly measure the frequencies of all on-board clock source generators
by means of the input capture of TIM5 channel4 and TIM11 channel1 as shown in Figure 18
and Figure 19.
Internal/external clock measurement using TIM5 channel4

TIM5 has an input multiplexer which allows choosing whether the input capture is triggered
by the I/O or by an internal clock. This selection is performed through the TI4_RMP [1:0] bits
in the TIM5_OR register.
158/1751 RM0090 Rev 19

The primary purpose of having the LSE connected to the channel4 input capture is to be
able to precisely measure the HSI (this requires to have the HSI used as the system clock
source). The number of HSI clock counts between consecutive edges of the LSE signal
provides a measurement of the internal clock period. Taking advantage of the high precision
of LSE crystals (typically a few tens of ppm) we can determine the internal clock frequency
with the same resolution, and trim the source to compensate for manufacturing-process
and/or temperature- and voltage-related frequency deviations.
The HSI oscillator has dedicated, user-accessible calibration bits for this purpose.
The basic concept consists in providing a relative measurement (e.g. HSI/LSE ratio): the
precision is therefore tightly linked to the ratio between the two clock sources. The greater
the ratio, the better the measurement.
It is also possible to measure the LSI frequency: this is useful for applications that do not
have a crystal. The ultralow-power LSI oscillator has a large manufacturing process
deviation: by measuring it versus the HSI clock source, it is possible to determine its
frequency with the precision of the HSI. The measured value can be used to have more
accurate RTC time base timeouts (when LSI is used as the RTC clock source) and/or an
IWDG timeout with an acceptable accuracy.
Use the following procedure to measure the LSI frequency:
1. Enable the TIM5 timer and configure channel4 in Input capture mode.
2. This bit is set the TI4_RMP bits in the TIM5_OR register to 0x01 to connect the LSI
clock internally to TIM5 channel4 input capture for calibration purposes.
3. Measure the LSI clock frequency using the TIM5 capture/compare 4 event or interrupt.
4. Use the measured LSI frequency to update the prescaler of the RTC depending on the
desired time base and/or to compute the IWDG timeout.
Figure 18. Frequency measurement with TIM5 in Input capture mode
TIM5
TI4_RMP[1:0]
GPIO
RTC_WakeUp_IT TI4
LSE
LSI
ai17741V2

by the I/O or by an internal clock. This selection is performed through TI1_RMP [1:0] bits in
the TIM11_OR register. The HSE_RTC clock (HSE divided by a programmable prescaler) is
connected to channel 1 input capture to have a rough indication of the external crystal
frequency. This requires that the HSI is the system clock source. This can be useful for
instance to ensure compliance with the IEC 60730/IEC 61335 standards which require to be
able to determine harmonic or subharmonic frequencies (–50/+100% deviations).
RM0090 Rev 19 159/1751

212
TIM11
TI1_RMP[1:0]
GPIO
TI1
HSE_RTC(1 MHz)
ai18433
160/1751 RM0090 Rev 19

6.3 RCC registers

Refer to Section 1.1: List of abbreviations for registers for a list of abbreviations used in
register descriptions.
6.3.1 RCC clock control register (RCC_CR)

Reset value: 0x0000 XX83 where X is undefined.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PLLSAI PLLSAI PLLI2S PLLI2S PLLRD CSS HSE HSE HSE
PLLON
Reserved RDY ON RDY ON Y Reserved ON BYP RDY ON
r rw r rw r rw rw rw r rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
HSI
HSICAL[7:0] HSITRIM[4:0] HSION
Res. RDY
r r r r r r r r rw rw rw rw rw r rw

Bit 29 PLLSAIRDY: PLLSAI clock ready flag
Set by hardware to indicate that the PLLSAI is locked.
0: PLLSAI unlocked
1: PLLSAI locked
Bit 28 PLLSAION: PLLSAI enable
Set and cleared by software to enable PLLSAI.
Cleared by hardware when entering Stop or Standby mode.
0: PLLSAI OFF
1: PLLSAI ON
Bit 27 PLLI2SRDY: PLLI2S clock ready flag
Set by hardware to indicate that the PLLI2S is locked.
0: PLLI2S unlocked
1: PLLI2S locked
Bit 26 PLLI2SON: PLLI2S enable
Set and cleared by software to enable PLLI2S.
0: PLLI2S OFF
1: PLLI2S ON
Bit 25 PLLRDY: Main PLL (PLL) clock ready flag
Set by hardware to indicate that PLL is locked.
0: PLL unlocked
1: PLL locked
Bit 24 PLLON: Main PLL (PLL) enable
Set and cleared by software to enable PLL.
Cleared by hardware when entering Stop or Standby mode. This bit cannot be reset if PLL
clock is used as the system clock.
0: PLL OFF
1: PLL ON
RM0090 Rev 19 161/1751

212

Bit 19 CSSON: Clock security system enable
Set and cleared by software to enable the clock security system. When CSSON is set, the
clock detector is enabled by hardware when the HSE oscillator is ready, and disabled by
hardware if an oscillator failure is detected.
0: Clock security system OFF (Clock detector OFF)
1: Clock security system ON (Clock detector ON if HSE oscillator is stable, OFF if not)
Bit 18 HSEBYP: HSE clock bypass
Set and cleared by software to bypass the oscillator with an external clock. The external
clock must be enabled with the HSEON bit, to be used by the device.
The HSEBYP bit can be written only if the HSE oscillator is disabled.
0: HSE oscillator not bypassed
1: HSE oscillator bypassed with an external clock
Bit 17 HSERDY: HSE clock ready flag
Set by hardware to indicate that the HSE oscillator is stable. After the HSEON bit is cleared,
HSERDY goes low after 6 HSE oscillator clock cycles.
0: HSE oscillator not ready
1: HSE oscillator ready
Bit 16 HSEON: HSE clock enable
Set and cleared by software.
Cleared by hardware to stop the HSE oscillator when entering Stop or Standby mode. This
bit cannot be reset if the HSE oscillator is used directly or indirectly as the system clock.
0: HSE oscillator OFF
1: HSE oscillator ON
Bits 15:8 HSICAL[7:0]: Internal high-speed clock calibration
These bits are initialized automatically at startup.
Bits 7:3 HSITRIM[4:0]: Internal high-speed clock trimming
These bits provide an additional user-programmable trimming value that is added to the
HSICAL[7:0] bits. It can be programmed to adjust to variations in voltage and temperature
that influence the frequency of the internal HSI RC.
Bit 1 HSIRDY: Internal high-speed clock ready flag
Set by hardware to indicate that the HSI oscillator is stable. After the HSION bit is cleared,
HSIRDY goes low after 6 HSI clock cycles.
0: HSI oscillator not ready
1: HSI oscillator ready
Bit 0 HSION: Internal high-speed clock enable
Set by hardware to force the HSI oscillator ON when leaving the Stop or Standby mode or in
case of a failure of the HSE oscillator used directly or indirectly as the system clock. This bit
cannot be cleared if the HSI is used directly or indirectly as the system clock.
0: HSI oscillator OFF
1: HSI oscillator ON
162/1751 RM0090 Rev 19

6.3.2 RCC PLL configuration register (RCC_PLLCFGR)

Access: no wait state, word, half-word and byte access.
This register is used to configure the PLL clock outputs according to the formulas:
• f(VCO clock) = f(PLL clock input) × (PLLN / PLLM)
• f(PLL general clock output) = f(VCO clock) / PLLP
• f(USB OTG FS, SDIO, RNG clock output) = f(VCO clock) / PLLQ
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PLLSR
PLLQ3 PLLQ2 PLLQ1 PLLQ0 Reserv PLLP1 PLLP0
Reserved C Reserved
ed
rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserv PLLN PLLM5 PLLM4 PLLM3 PLLM2 PLLM1 PLLM0

ed rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 27:24 PLLQ: Main PLL (PLL) division factor for USB OTG FS, SDIO and random number generator
clocks
Set and cleared by software to control the frequency of USB OTG FS clock, the random
number generator clock and the SDIO clock. These bits should be written only if PLL is
disabled.
Caution: The USB OTG FS requires a 48 MHz clock to work correctly. The SDIO and the
random number generator need a frequency lower than or equal to 48 MHz to work
correctly.
USB OTG FS clock frequency = VCO frequency / PLLQ with 2 ≤ PLLQ ≤ 15
0000: PLLQ = 0, wrong configuration
0010: PLLQ = 2
0011: PLLQ = 3
0100: PLLQ = 4
...
1111: PLLQ = 15
Bit 22 PLLSRC: Main PLL(PLL) and audio PLL (PLLI2S) entry clock source
Set and cleared by software to select PLL and PLLI2S clock source. This bit can be written
only when PLL and PLLI2S are disabled.
0: HSI clock selected as PLL and PLLI2S clock entry
1: HSE oscillator clock selected as PLL and PLLI2S clock entry
RM0090 Rev 19 163/1751

212
Bits 17:16 PLLP: Main PLL (PLL) division factor for main system clock
Set and cleared by software to control the frequency of the general PLL output clock. These
bits can be written only if PLL is disabled.
Caution: The software has to set these bits correctly not to exceed 180 MHz on this domain.
PLL output clock frequency = VCO frequency / PLLP with PLLP = 2, 4, 6, or 8
00: PLLP = 2
01: PLLP = 4
10: PLLP = 6
11: PLLP = 8
Bits 14:6 PLLN: Main PLL (PLL) multiplication factor for VCO
Set and cleared by software to control the multiplication factor of the VCO. These bits can
be written only when PLL is disabled. Only half-word and word accesses are allowed to
write these bits.
Caution: The software has to set these bits correctly to ensure that the VCO output
frequency is between 100 and 432 MHz.
VCO output frequency = VCO input frequency × PLLN with 50 ≤ PLLN ≤ 432
000000000: PLLN = 0, wrong configuration
...
000110010: PLLN = 50
...
001100011: PLLN = 99
001100100: PLLN = 100
...
110110000: PLLN = 432
...
Note: Multiplication factors ranging from 50 and 99 are possible for VCO input frequency
higher than 1 MHz. However care must be taken that the minimum VCO output
frequency respects the value specified above.
Bits 5:0 PLLM: Division factor for the main PLL (PLL) and audio PLL (PLLI2S) input clock
Set and cleared by software to divide the PLL and PLLI2S input clock before the VCO.
These bits can be written only when the PLL and PLLI2S are disabled.
Caution: The software has to set these bits correctly to ensure that the VCO input frequency
ranges from 1 to 2 MHz. It is recommended to select a frequency of 2 MHz to limit
PLL jitter.
VCO input frequency = PLL input clock frequency / PLLM with 2 ≤ PLLM ≤ 63
000000: PLLM = 0, wrong configuration
000010: PLLM = 2
000011: PLLM = 3
000100: PLLM = 4
...
111110: PLLM = 62
111111: PLLM = 63
164/1751 RM0090 Rev 19

6.3.3 RCC clock configuration register (RCC_CFGR)

Access: 0 ≤ wait state ≤ 2, word, half-word and byte access
1 or 2 wait states inserted only if the access occurs during a clock source switch.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
I2SSC
MCO2 MCO2 PRE[2:0] MCO1 PRE[2:0] MCO1 RTCPRE[4:0]
R
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PPRE2[2:0] PPRE1[2:0] HPRE[3:0] SWS1 SWS0 SW1 SW0
Reserved
rw rw rw rw rw rw rw rw rw rw r r rw rw
Bits 31:30 MCO2[1:0]: Microcontroller clock output 2

Set and cleared by software. Clock source selection may generate glitches on MCO2. It is
highly recommended to configure these bits only after reset before enabling the external
oscillators and the PLLs.
00: System clock (SYSCLK) selected
01: PLLI2S clock selected
10: HSE oscillator clock selected
11: PLL clock selected
Bits 27:29 MCO2PRE: MCO2 prescaler
Set and cleared by software to configure the prescaler of the MCO2. Modification of this
prescaler may generate glitches on MCO2. It is highly recommended to change this
prescaler only after reset before enabling the external oscillators and the PLLs.
0xx: no division
100: division by 2
101: division by 3
110: division by 4
111: division by 5
prescaler only after reset before enabling the external oscillators and the PLL.
0xx: no division
100: division by 2
101: division by 3
110: division by 4
111: division by 5
Bit 23 I2SSRC: I2S clock selection
Set and cleared by software. This bit allows to select the I2S clock source between the
PLLI2S clock and the external clock. It is highly recommended to change this bit only after
reset and before enabling the I2S module.
0: PLLI2S clock used as I2S clock source
1: External clock mapped on the I2S_CKIN pin used as I2S clock source
RM0090 Rev 19 165/1751

212
Bits 22:21 MCO1: Microcontroller clock output 1

oscillators and PLL.
00: HSI clock selected
01: LSE oscillator selected
Bits 20:16 RTCPRE: HSE division factor for RTC clock
Set and cleared by software to divide the HSE clock input clock to generate a 1 MHz clock
for RTC.
Caution: The software has to set these bits correctly to ensure that the clock supplied to the
RTC is 1 MHz. These bits must be configured if needed before selecting the RTC
clock source.
00000: no clock
00001: no clock
00010: HSE/2
00011: HSE/3
00100: HSE/4
...
11110: HSE/30
11111: HSE/31
Bits 15:13 PPRE2: APB high-speed prescaler (APB2)
Set and cleared by software to control APB high-speed clock division factor.
The clocks are divided with the new prescaler factor from 1 to 16 AHB cycles after
PPRE2 write.
0xx: AHB clock not divided
100: AHB clock divided by 2
Bits 12:10 PPRE1: APB Low speed prescaler (APB1)
Set and cleared by software to control APB low-speed clock division factor.
PPRE1 write.
166/1751 RM0090 Rev 19

Bits 7:4 HPRE: AHB prescaler

Set and cleared by software to control AHB clock division factor.
Caution: The clocks are divided with the new prescaler factor from 1 to 16 AHB cycles after
HPRE write.
Caution: The AHB clock frequency must be at least 25 MHz when the Ethernet is used.
0xxx: system clock not divided
1000: system clock divided by 2
Bits 3:2 SWS: System clock switch status
Set and cleared by hardware to indicate which clock source is used as the system clock.
00: HSI oscillator used as the system clock
01: HSE oscillator used as the system clock
10: PLL used as the system clock
11: not applicable
Bits 1:0 SW: System clock switch
Set and cleared by software to select the system clock source.
Set by hardware to force the HSI selection when leaving the Stop or Standby mode or in
case of failure of the HSE oscillator used directly or indirectly as the system clock.
00: HSI oscillator selected as system clock
01: HSE oscillator selected as system clock
10: PLL selected as system clock
11: not allowed
6.3.4 RCC clock interrupt register (RCC_CIR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PLLSAI PLLI2S PLL HSE HSI LSE LSI
CSSC
Reserved RDYC RDYC RDYC RDYC RDYC RDYC RDYC
w w w w w w w w
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PLLSAI PLLI2S PLL HSE HSI LSE LSI PLLSAI PLLI2S PLL HSE HSI LSE LSI
Reserv CSSF
RDYIE RDYIE RDYIE RDYIE RDYIE RDYIE RDYIE RDYF RDYF RDYF RDYF RDYF RDYF RDYF
ed
rw rw rw rw rw rw rw r r r r r r r r
RM0090 Rev 19 167/1751

212

Bit 23 CSSC: Clock security system interrupt clear
This bit is set by software to clear the CSSF flag.
0: No effect
1: Clear CSSF flag
Bit 22 PLLSAIRDYC: PLLSAI Ready Interrupt Clear
This bit is set by software to clear PLLSAIRDYF flag. It is reset by hardware when the
PLLSAIRDYF is cleared.
0: PLLSAIRDYF not cleared
1: PLLSAIRDYF cleared
Bit 21 PLLI2SRDYC: PLLI2S ready interrupt clear
This bit is set by software to clear the PLLI2SRDYF flag.
0: No effect
1: PLLI2SRDYF cleared
Bit 20 PLLRDYC: Main PLL(PLL) ready interrupt clear
This bit is set by software to clear the PLLRDYF flag.
0: No effect
1: PLLRDYF cleared
Bit 19 HSERDYC: HSE ready interrupt clear
This bit is set by software to clear the HSERDYF flag.
0: No effect
1: HSERDYF cleared
Bit 18 HSIRDYC: HSI ready interrupt clear
This bit is set software to clear the HSIRDYF flag.
0: No effect
1: HSIRDYF cleared
Bit 17 LSERDYC: LSE ready interrupt clear
This bit is set by software to clear the LSERDYF flag.
0: No effect
1: LSERDYF cleared
Bit 16 LSIRDYC: LSI ready interrupt clear
This bit is set by software to clear the LSIRDYF flag.
0: No effect
1: LSIRDYF cleared
Bit 14 PLLSAIRDYIE: PLLSAI Ready Interrupt Enable
This bit is set and reset by software to enable/disable interrupt caused by PLLSAI lock.
0: PLLSAI lock interrupt disabled
1: PLLSAI lock interrupt enabled
Bit 13 PLLI2SRDYIE: PLLI2S ready interrupt enable
This bit is set and cleared by software to enable/disable interrupt caused by PLLI2S lock.
0: PLLI2S lock interrupt disabled
1: PLLI2S lock interrupt enabled
168/1751 RM0090 Rev 19

Bit 12 PLLRDYIE: Main PLL (PLL) ready interrupt enable

This bit is set and cleared by software to enable/disable interrupt caused by PLL lock.
0: PLL lock interrupt disabled
1: PLL lock interrupt enabled
Bit 11 HSERDYIE: HSE ready interrupt enable
This bit is set and cleared by software to enable/disable interrupt caused by the HSE
oscillator stabilization.
0: HSE ready interrupt disabled
1: HSE ready interrupt enabled
Bit 10 HSIRDYIE: HSI ready interrupt enable
This bit is set and cleared by software to enable/disable interrupt caused by the HSI
0: HSI ready interrupt disabled
1: HSI ready interrupt enabled
Bit 9 LSERDYIE: LSE ready interrupt enable
This bit is set and cleared by software to enable/disable interrupt caused by the LSE
0: LSE ready interrupt disabled
1: LSE ready interrupt enabled
Bit 8 LSIRDYIE: LSI ready interrupt enable
This bit is set and cleared by software to enable/disable interrupt caused by LSI oscillator
stabilization.
0: LSI ready interrupt disabled
1: LSI ready interrupt enabled
Bit 7 CSSF: Clock security system interrupt flag
This bit is set by hardware when a failure is detected in the HSE oscillator.
It is cleared by software by setting the CSSC bit.
0: No clock security interrupt caused by HSE clock failure
1: Clock security interrupt caused by HSE clock failure
Bit 6 PLLSAIRDYF: PLLSAI Ready Interrupt flag
This bit is set by hardware when the PLLSAI is locked and PLLSAIRDYDIE is set.
It is cleared by software by setting the PLLSAIRDYC bit.
0: No clock ready interrupt caused by PLLSAI lock
1: Clock ready interrupt caused by PLLSAI lock
Bit 5 PLLI2SRDYF: PLLI2S ready interrupt flag
This bit is set by hardware when the PLLI2S is locked and PLLI2SRDYDIE is set.
It is cleared by software by setting the PLLRI2SDYC bit.
0: No clock ready interrupt caused by PLLI2S lock
1: Clock ready interrupt caused by PLLI2S lock
Bit 4 PLLRDYF: Main PLL (PLL) ready interrupt flag
This bit is set by hardware when PLL is locked and PLLRDYDIE is set.
It is cleared by software setting the PLLRDYC bit.
0: No clock ready interrupt caused by PLL lock
1: Clock ready interrupt caused by PLL lock
RM0090 Rev 19 169/1751

212
Bit 3 HSERDYF: HSE ready interrupt flag

This bit is set by hardware when External High Speed clock becomes stable and
HSERDYDIE is set.
It is cleared by software by setting the HSERDYC bit.
0: No clock ready interrupt caused by the HSE oscillator
1: Clock ready interrupt caused by the HSE oscillator
Bit 2 HSIRDYF: HSI ready interrupt flag
This bit is set by hardware when the Internal High Speed clock becomes stable and
HSIRDYDIE is set.
It is cleared by software by setting the HSIRDYC bit.
0: No clock ready interrupt caused by the HSI oscillator
1: Clock ready interrupt caused by the HSI oscillator
Bit 1 LSERDYF: LSE ready interrupt flag
This bit is set by hardware when the External Low Speed clock becomes stable and
LSERDYDIE is set.
It is cleared by software by setting the LSERDYC bit.
0: No clock ready interrupt caused by the LSE oscillator
1: Clock ready interrupt caused by the LSE oscillator
Bit 0 LSIRDYF: LSI ready interrupt flag
This bit is set by hardware when the internal low speed clock becomes stable and
LSIRDYDIE is set.
It is cleared by software by setting the LSIRDYC bit.
0: No clock ready interrupt caused by the LSI oscillator
1: Clock ready interrupt caused by the LSI oscillator
6.3.5 RCC AHB1 peripheral reset register (RCC_AHB1RSTR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
OTGH
ETHMAC DMA2D DMA2 DMA1
S
Reserved Reserved RST Res. RST RST RST Reserved
RST
rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CRCR GPIOK GPIOJ GPIOI GPIOH GPIOGG GPIOF GPIOE GPIOD GPIOC GPIOB GPIOA
Reserved ST Res. RST RST RST RST RST RST RST RST RST RST RST

Bit 29 OTGHSRST: USB OTG HS module reset
0: does not reset the USB OTG HS module
1: resets the USB OTG HS module
170/1751 RM0090 Rev 19

Bit 25 ETHMACRST: Ethernet MAC reset

0: does not reset Ethernet MAC
1: resets Ethernet MAC
Bit 23 DMA2DRST: DMA2D reset
This bit is set and reset by software.
0: does not reset DMA2D
1: resets DMA2D
Bit 22 DMA2RST: DMA2 reset
0: does not reset DMA2
1: resets DMA2
1: resets DMA2
Bit 12 CRCRST: CRC reset
0: does not reset CRC
1: resets CRC
Bit 10 GPIOKRST: IO port K reset
0: does not reset IO port K
1: resets IO port K
Bit 9 GPIOJRST: IO port J reset
0: does not reset IO port J
1: resets IO port J
Bit 8 GPIOIRST: IO port I reset
0: does not reset IO port I
1: resets IO port I
Bit 7 GPIOHRST: IO port H reset
0: does not reset IO port H
1: resets IO port H
Bit 6 GPIOGRST: IO port G reset
0: does not reset IO port G
1: resets IO port G
RM0090 Rev 19 171/1751

212
Bit 5 GPIOFRST: IO port F reset

0: does not reset IO port F
1: resets IO port F
Bit 4 GPIOERST: IO port E reset
0: does not reset IO port E
1: resets IO port E
Bit 3 GPIODRST: IO port D reset
0: does not reset IO port D
1: resets IO port D
Bit 2 GPIOCRST: IO port C reset
0: does not reset IO port C
1: resets IO port C
Bit 1 GPIOBRST: IO port B reset
0: does not reset IO port B
1:resets IO port B
Bit 0 GPIOARST: IO port A reset
0: does not reset IO port A
1: resets IO port A
172/1751 RM0090 Rev 19


31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OTGFS RNG HASH CRYP DCMI
Reserved RST RST RST RST Reserved RST
rw rw rw rw rw

Bit 7 OTGFSRST: USB OTG FS module reset
0: does not reset the USB OTG FS module
1: resets the USB OTG FS module
Bit 6 RNGRST: Random number generator module reset
0: does not reset the random number generator module
1: resets the random number generator module
Bit 5 HASHRST: Hash module reset
0: does not reset the HASH module
1: resets the HASH module
Bit 4 CRYPRST: Cryptographic module reset
0: does not reset the cryptographic module
1: resets the cryptographic module
Bit 0 DCMIRST: Camera interface reset
0: does not reset the Camera interface
1: resets the Camera interface
RM0090 Rev 19 173/1751

212

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FMCRST
Reserved
rw

Bit 0 FMCRST: Flexible memory controller module reset
0: does not reset the FMC module
1: resets the FMC module
6.3.8 RCC APB1 peripheral reset register (RCC_APB1RSTR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
UART8R UART7R PWR CAN2 CAN1 I2C3 I2C2 I2C1 UART5 UART4 UART3 UART2
DACRST
ST ST RST Reser- RST RST Reser- RST RST RST RST RST RST RST Reser-
ved ved ved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SPI3 SPI2 WWDG TIM14 TIM13 TIM12 TIM7 TIM6 TIM5 TIM4 TIM3 TIM2
RST RST Reserved RST Reserved RST RST RST RST RST RST RST RST RST
174/1751 RM0090 Rev 19

Bit 31 UART8RST: UART8 reset

0: does not reset UART8
1: resets UART8
1: resets UART7
Bit 29 DACRST: DAC reset
0: does not reset the DAC interface
1: resets the DAC interface
Bit 28 PWRRST: Power interface reset
0: does not reset the power interface
1: resets the power interface
Bit 26 CAN2RST: CAN2 reset
0: does not reset CAN2
1: resets CAN2
1: resets CAN1
Bit 23 I2C3RST: I2C3 reset
0: does not reset I2C3
1: resets I2C3
1: resets I2C2
1: resets I2C1
1: resets UART5
Bit 19 UART4RST: USART4 reset
1: resets UART4
RM0090 Rev 19 175/1751

212
Bit 18 USART3RST: USART3 reset

0: does not reset USART3
1: resets USART3
1: resets USART2
Bit 15 SPI3RST: SPI3 reset
0: does not reset SPI3
1: resets SPI3
1: resets SPI2
Bit 11 WWDGRST: Window watchdog reset
0: does not reset the window watchdog
1: resets the window watchdog
Bit 8 TIM14RST: TIM14 reset
0: does not reset TIM14
1: resets TIM14
1: resets TIM13
1: resets TIM12
1: resets TIM7
1: resets TIM6
176/1751 RM0090 Rev 19


1: resets TIM5
1: resets TIM4
1: resets TIM3
1: resets TIM2
RM0090 Rev 19 177/1751

212

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
LTD SAI1 SPI6 SPI5 TIM11 TIM10 TIM9
Reserved CRST Reserved RST RST RST Res. RST RST RST
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
USART USART
SYSCF SPI4 SPI1 SDIO ADC TIM8 TIM1
Reser- 6 1
G RST RST RST RST Reserved RST Reserved Reserved RST RST
ved RST RST

Bit 26 LTDCRST: LTDC reset
0: does not reset LCD-TFT
1: resets LCD-TFT
Bit 22 SAI1RST: SAI1 reset
0: does not reset SAI1
1: resets SAI1
1: resets SPI6
1: resets SPI5
1: resets TIM14
1: resets TIM10
1: resets TIM9
178/1751 RM0090 Rev 19


Bit 14 SYSCFGRST: System configuration controller reset
0: does not reset the System configuration controller
1: resets the System configuration controller
1: resets SPI4
1: resets SPI1
Bit 11 SDIORST: SDIO reset
0: does not reset the SDIO module
1: resets the SDIO module
Bit 8 ADCRST: ADC interface reset (common to all ADCs)
0: does not reset the ADC interface
1: resets the ADC interface
1: resets USART6
1: resets USART1
1: resets TIM8
1: resets TIM1
RM0090 Rev 19 179/1751

212
6.3.10 RCC AHB1 peripheral clock register (RCC_AHB1ENR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
OTGH
ETHM ETHM ETHM
S OTGH ETHMA DMA2D DMA2E DMA1E CCMDAT BKPSR
Reser- ULPIE ACPTP ACRXE ACTXE Res.
SEN CEN Res. EN N N ARAMEN AMEN Reserved
ved EN N N
N
rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CRCE GPIOK GPIOJ GPIOIE GPIOH GPIOG GPIOFE GPIOD GPIOC GPIO GPIO
GPIOEEN
Reserved N Res. EN EN N EN EN N EN EN BEN AEN

Bit 30 OTGHSULPIEN: USB OTG HSULPI clock enable
This bit is set and cleared by software. It must be cleared when the OTG_HS is used in FS
mode.
0: USB OTG HS ULPI clock disabled
1: USB OTG HS ULPI clock enabled
Bit 29 OTGHSEN: USB OTG HS clock enable
0: USB OTG HS clock disabled
1: USB OTG HS clock enabled
Bit 28 ETHMACPTPEN: Ethernet PTP clock enable
0: Ethernet PTP clock disabled
1: Ethernet PTP clock enabled
Bit 27 ETHMACRXEN: Ethernet Reception clock enable
0: Ethernet Reception clock disabled
1: Ethernet Reception clock enabled
Bit 26 ETHMACTXEN: Ethernet Transmission clock enable
0: Ethernet Transmission clock disabled
1: Ethernet Transmission clock enabled
Bit 25 ETHMACEN: Ethernet MAC clock enable
0: Ethernet MAC clock disabled
1: Ethernet MAC clock enabled
Bit 23 DMA2DEN: DMA2D clock enable
0: DMA2D clock disabled
1: DMA2D clock enabled
180/1751 RM0090 Rev 19

Bit 22 DMA2EN: DMA2 clock enable

0: DMA2 clock disabled
1: DMA2 clock enabled
Bit 20 CCMDATARAMEN: CCM data RAM clock enable
0: CCM data RAM clock disabled
1: CCM data RAM clock enabled
Bit 18 BKPSRAMEN: Backup SRAM interface clock enable
0: Backup SRAM interface clock disabled
1: Backup SRAM interface clock enabled
Bit 12 CRCEN: CRC clock enable
0: CRC clock disabled
1: CRC clock enabled
Bit 10 GPIOKEN: IO port K clock enable
0: IO port K clock disabled
1: IO port K clock enabled
Bit 9 GPIOJEN: IO port J clock enable
0: IO port J clock disabled
1: IO port J clock enabled
Bit 8 GPIOIEN: IO port I clock enable
0: IO port I clock disabled
1: IO port I clock enabled
Bit 7 GPIOHEN: IO port H clock enable
0: IO port H clock disabled
1: IO port H clock enabled
Bit 6 GPIOGEN: IO port G clock enable
0: IO port G clock disabled
1: IO port G clock enabled
RM0090 Rev 19 181/1751

212
Bit 5 GPIOFEN: IO port F clock enable

0: IO port F clock disabled
1: IO port F clock enabled
Bit 4 GPIOEEN: IO port E clock enable
0: IO port E clock disabled
1: IO port E clock enabled
Bit 3 GPIODEN: IO port D clock enable
0: IO port D clock disabled
1: IO port D clock enabled
Bit 2 GPIOCEN: IO port C clock enable
0: IO port C clock disabled
1: IO port C clock enabled
Bit 1 GPIOBEN: IO port B clock enable
0: IO port B clock disabled
1: IO port B clock enabled
Bit 0 GPIOAEN: IO port A clock enable
0: IO port A clock disabled
1: IO port A clock enabled
6.3.11 RCC AHB2 peripheral clock enable register (RCC_AHB2ENR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserved EN EN EN EN Reserved EN
rw rw rw rw rw

Bit 7 OTGFSEN: USB OTG FS clock enable
0: USB OTG FS clock disabled
1: USB OTG FS clock enabled
Bit 6 RNGEN: Random number generator clock enable
0: Random number generator clock disabled
1: Random number generator clock enabled
182/1751 RM0090 Rev 19

Bit 5 HASHEN: Hash modules clock enable

0: Hash modules clock disabled
1: Hash modules clock enabled
Bit 4 CRYPEN: Cryptographic modules clock enable
0: cryptographic module clock disabled
1: cryptographic module clock enabled
Bit 0 DCMIEN: Camera interface enable
0: Camera interface clock disabled
1: Camera interface clock enabled

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FMCEN
Reserved
rw

Bit 0 FMCEN: Flexible memory controller module clock enable
0: FMC module clock disabled
1: FMC module clock enabled
6.3.13 RCC APB1 peripheral clock enable register (RCC_APB1ENR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
USART USART
UART8 UART7 DAC PWR CAN2 CAN1 I2C3 I2C2 I2C1 UART5 UART4
Reser- Reser- 3 2 Reser-
EN EN EN EN EN EN EN EN EN EN EN
ved ved EN EN ved
rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
EN EN Reserved EN Reserved EN EN EN EN EN EN EN EN EN
RM0090 Rev 19 183/1751

212
Bit 31 UART8EN: UART8 clock enable

0: UART8 clock disabled
1: UART8 clock enabled
Bit 29 DACEN: DAC interface clock enable
0: DAC interface clock disabled
1: DAC interface clock enable
Bit 28 PWREN: Power interface clock enable
0: Power interface clock disabled
1: Power interface clock enable
Bit 26 CAN2EN: CAN 2 clock enable
0: CAN 2 clock disabled
1: CAN 2 clock enabled
Bit 23 I2C3EN: I2C3 clock enable
0: I2C3 clock disabled
1: I2C3 clock enabled
184/1751 RM0090 Rev 19

Bit 18 USART3EN: USART3 clock enable

0: USART3 clock disabled
1: USART3 clock enabled
Bit 15 SPI3EN: SPI3 clock enable
0: SPI3 clock disabled
1: SPI3 clock enabled
Bit 11 WWDGEN: Window watchdog clock enable
0: Window watchdog clock disabled
1: Window watchdog clock enabled
Bit 10:9 Reserved, must be kept at reset value.
Bit 8 TIM14EN: TIM14 clock enable
0: TIM14 clock disabled
1: TIM14 clock enabled
RM0090 Rev 19 185/1751

212

186/1751 RM0090 Rev 19


31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
LTDC TIM11 TIM10 TIM9
SAI1EN SPI6EN SPI5EN
Reserved EN Reserved Res. EN EN EN
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
USART USART
SYSCF SPI4E SPI1 SDIO ADC3 ADC2 ADC1 TIM8 TIM1
Reser- 6 1
G EN N EN EN EN EN EN Reserved Reserved EN EN
ved EN EN

Bit 26 LTDCEN: LTDC clock enable
0: LTDC clock disabled
1: LTDC clock enabled
Bits 27: 23 Reserved, must be kept at reset value.
Bit 22 SAI1EN: SAI1 clock enable
0: SAI1 clock disabled
1: SAI1 clock enabled
RM0090 Rev 19 187/1751

212
Bit 14 SYSCFGEN: System configuration controller clock enable

0: System configuration controller clock disabled
1: System configuration controller clock enabled
Bit 11 SDIOEN: SDIO clock enable
0: SDIO module clock disabled
1: SDIO module clock enabled
Bit 10 ADC3EN: ADC3 clock enable
0: ADC3 clock disabled
1: ADC3 clock enabled
188/1751 RM0090 Rev 19

(RCC_AHB1LPENR)
Reset value: 0x7EEF 97FF
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
OTGHS OTGH ETHPT ETHMA BKPSRA SRAM SRAM
ETHRX ETHTX DMA2D DMA2 DMA1 SRAM3
ULPILPE S P C M 2 1
Res. LPEN LPEN Res. LPEN LPEN LPEN Res. LPEN
N LPEN LPEN LPEN LPEN LPEN LPEN
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
GPIOG GPIO
FLITF CRC GPIOK GPIOIJ GPIOI GPIOH GPIOE GPIOD GPIOC GPIOB GPIOA
G F
LPEN Reserved LPEN Res. LPEN LPEN LPEN LPEN LPEN LPEN LPEN LPEN LPEN
LPEN LPEN

Bit 30 OTGHSULPILPEN: USB OTG HS ULPI clock enable during Sleep mode
This bit is set and cleared by software. It must be cleared when the OTG_HS is used in FS
mode.
0: USB OTG HS ULPI clock disabled during Sleep mode
1: USB OTG HS ULPI clock enabled during Sleep mode
Bit 29 OTGHSLPEN: USB OTG HS clock enable during Sleep mode
0: USB OTG HS clock disabled during Sleep mode
1: USB OTG HS clock enabled during Sleep mode
Bit 28 ETHMACPTPLPEN: Ethernet PTP clock enable during Sleep mode
0: Ethernet PTP clock disabled during Sleep mode
1: Ethernet PTP clock enabled during Sleep mode
Bit 27 ETHMACRXLPEN: Ethernet reception clock enable during Sleep mode
0: Ethernet reception clock disabled during Sleep mode
1: Ethernet reception clock enabled during Sleep mode
Bit 26 ETHMACTXLPEN: Ethernet transmission clock enable during Sleep mode
0: Ethernet transmission clock disabled during sleep mode
1: Ethernet transmission clock enabled during sleep mode
Bit 25 ETHMACLPEN: Ethernet MAC clock enable during Sleep mode
0: Ethernet MAC clock disabled during Sleep mode
1: Ethernet MAC clock enabled during Sleep mode
RM0090 Rev 19 189/1751

212
Bit 23 DMA2DLPEN: DMA2D clock enable during Sleep mode

0: DMA2D clock disabled during Sleep mode
1: DMA2D clock enabled during Sleep mode
Bit 22 DMA2LPEN: DMA2 clock enable during Sleep mode
0: DMA2 clock disabled during Sleep mode
1: DMA2 clock enabled during Sleep mode
Bit 19 SRAM3LPEN: SRAM3 interface clock enable during Sleep mode
0: SRAM3 interface clock disabled during Sleep mode
1: SRAM3 interface clock enabled during Sleep mode
Bit 18 BKPSRAMLPEN: Backup SRAM interface clock enable during Sleep mode
0: Backup SRAM interface clock disabled during Sleep mode
1: Backup SRAM interface clock enabled during Sleep mode
Bit 15 FLITFLPEN: Flash interface clock enable during Sleep mode
0: Flash interface clock disabled during Sleep mode
1: Flash interface clock enabled during Sleep mode
Bit 12 CRCLPEN: CRC clock enable during Sleep mode
0: CRC clock disabled during Sleep mode
1: CRC clock enabled during Sleep mode
Bit 10 GPIOKLPEN: IO port K clock enable during Sleep mode
0: IO port K clock disabled during Sleep mode
1: IO port K clock enabled during Sleep mode
190/1751 RM0090 Rev 19

Bit 9 GPIOJLPEN: IO port J clock enable during Sleep mode

0: IO port J clock disabled during Sleep mode
1: IO port J clock enabled during Sleep mode
Bit 8 GPIOILPEN: IO port I clock enable during Sleep mode
0: IO port I clock disabled during Sleep mode
1: IO port I clock enabled during Sleep mode
Bit 7 GPIOHLPEN: IO port H clock enable during Sleep mode
0: IO port H clock disabled during Sleep mode
1: IO port H clock enabled during Sleep mode
Bits 6 GPIOGLPEN: IO port G clock enable during Sleep mode
0: IO port G clock disabled during Sleep mode
1: IO port G clock enabled during Sleep mode
Bit 5 GPIOFLPEN: IO port F clock enable during Sleep mode
0: IO port F clock disabled during Sleep mode
1: IO port F clock enabled during Sleep mode
Bit 4 GPIOELPEN: IO port E clock enable during Sleep mode
0: IO port E clock disabled during Sleep mode
1: IO port E clock enabled during Sleep mode
Bit 3 GPIODLPEN: IO port D clock enable during Sleep mode
0: IO port D clock disabled during Sleep mode
1: IO port D clock enabled during Sleep mode
Bit 2 GPIOCLPEN: IO port C clock enable during Sleep mode
0: IO port C clock disabled during Sleep mode
1: IO port C clock enabled during Sleep mode
Bit 1 GPIOBLPEN: IO port B clock enable during Sleep mode
0: IO port B clock disabled during Sleep mode
1: IO port B clock enabled during Sleep mode
Bit 0 GPIOALPEN: IO port A clock enable during sleep mode
0: IO port A clock disabled during Sleep mode
1: IO port A clock enabled during Sleep mode
RM0090 Rev 19 191/1751

212
(RCC_AHB2LPENR)
Reset value: 0x0000 00F1
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserved LPEN LPEN LPEN LPEN Reserved LPEN
rw rw rw rw rw

Bit 7 OTGFSLPEN: USB OTG FS clock enable during Sleep mode
0: USB OTG FS clock disabled during Sleep mode
1: USB OTG FS clock enabled during Sleep mode
Bit 6 RNGLPEN: Random number generator clock enable during Sleep mode
0: Random number generator clock disabled during Sleep mode
1: Random number generator clock enabled during Sleep mode
Bit 5 HASHLPEN: Hash modules clock enable during Sleep mode
0: Hash modules clock disabled during Sleep mode
1: Hash modules clock enabled during Sleep mode
Bit 4 CRYPLPEN: Cryptography modules clock enable during Sleep mode
0: cryptography modules clock disabled during Sleep mode
1: cryptography modules clock enabled during Sleep mode
Bit 0 DCMILPEN: Camera interface enable during Sleep mode
0: Camera interface clock disabled during Sleep mode
1: Camera interface clock enabled during Sleep mode
192/1751 RM0090 Rev 19

(RCC_AHB3LPENR)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FMC
Reserved LPEN
rw
Bits 31:1Reserved, must be kept at reset value.

FMCLPEN: Flexible memory controller module clock enable during Sleep mode
Bit 0
0: FMC module clock disabled during Sleep mode
1: FMC module clock enabled during Sleep mode
(RCC_APB1LPENR)
Reset value: 0xF6FE C9FF
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
USART USART
UART8 UART7 DAC PWR RESER CAN2 CAN1 I2C3 I2C2 I2C1 UART5 UART4
Reser- 3 2 Reser-
LPEN LPEN LPEN LPEN VED LPEN LPEN LPEN LPEN LPEN LPEN LPEN
ved LPEN LPEN ved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
LPEN LPEN Reserved LPEN Reserved LPEN LPEN LPEN LPEN LPEN LPEN LPEN LPEN LPEN
RM0090 Rev 19 193/1751

212
Bit 31 UART8LPEN: UART8 clock enable during Sleep mode

0: UART8 clock disabled during Sleep mode
1: UART8 clock enabled during Sleep mode
Bit 29 DACLPEN: DAC interface clock enable during Sleep mode
0: DAC interface clock disabled during Sleep mode
1: DAC interface clock enabled during Sleep mode
Bit 28 PWRLPEN: Power interface clock enable during Sleep mode
0: Power interface clock disabled during Sleep mode
1: Power interface clock enabled during Sleep mode
Bit 26 CAN2LPEN: CAN 2 clock enable during Sleep mode
0: CAN 2 clock disabled during sleep mode
1: CAN 2 clock enabled during sleep mode
0: CAN 1 clock disabled during Sleep mode
1: CAN 1 clock enabled during Sleep mode
Bit 23 I2C3LPEN: I2C3 clock enable during Sleep mode
0: I2C3 clock disabled during Sleep mode
1: I2C3 clock enabled during Sleep mode
194/1751 RM0090 Rev 19

Bit 18 USART3LPEN: USART3 clock enable during Sleep mode

0: USART3 clock disabled during Sleep mode
1: USART3 clock enabled during Sleep mode
Bit 15 SPI3LPEN: SPI3 clock enable during Sleep mode
0: SPI3 clock disabled during Sleep mode
1: SPI3 clock enabled during Sleep mode
Bit 11 WWDGLPEN: Window watchdog clock enable during Sleep mode
0: Window watchdog clock disabled during sleep mode
1: Window watchdog clock enabled during sleep mode
Bit 8 TIM14LPEN: TIM14 clock enable during Sleep mode
0: TIM14 clock disabled during Sleep mode
1: TIM14 clock enabled during Sleep mode
RM0090 Rev 19 195/1751

212

196/1751 RM0090 Rev 19

6.3.19 RCC APB2 peripheral clock enabled in low power mode register
(RCC_APB2LPENR)
Reset value: 0x0x0477 7F33
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
LTDC SAI1 SPI6 SPI5 TIM11 TIM10 TIM9
LPEN LPEN LPEN LPEN Reser- LPEN LPEN LPEN
Reserved Reserved
ved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SYSC USART USART
SPI4 SPI1 SDIO ADC3 ADC2 ADC1 TIM8 TIM1
Reser- FG 6 1
LPEN LPEN LPEN LPEN LPEN LPEN Reserved Reserved LPEN LPEN
ved LPEN LPEN LPEN

Bit 26 LTDCLPEN: LTDC clock enable during Sleep mode
0: LTDC clock disabled during Sleep mode
1: LTDC clock enabled during Sleep mode
Bit 22 SAI1LPEN: SAI1 clock enable during Sleep mode
0: SAI1 clock disabled during Sleep mode
1: SAI1 clock enabled during Sleep mode
RM0090 Rev 19 197/1751

212
Bit 16 TIM9LPEN: TIM9 clock enable during sleep mode

Bit 14 SYSCFGLPEN: System configuration controller clock enable during Sleep mode
0: System configuration controller clock disabled during Sleep mode
1: System configuration controller clock enabled during Sleep mode
Bit 11 SDIOLPEN: SDIO clock enable during Sleep mode
0: SDIO module clock disabled during Sleep mode
1: SDIO module clock enabled during Sleep mode
Bit 10 ADC3LPEN: ADC 3 clock enable during Sleep mode
0: ADC 3 clock disabled during Sleep mode
Bit 9 ADC2LPEN: ADC2 clock enable during Sleep mode
0: ADC2 clock disabled during Sleep mode
198/1751 RM0090 Rev 19


6.3.20 RCC Backup domain control register (RCC_BDCR)

Reset value: 0x0000 0000, reset by Backup domain reset.
Wait states are inserted in case of successive accesses to this register.
The LSEON, LSEBYP, RTCSEL and RTCEN bits in the RCC Backup domain control
register (RCC_BDCR) are in the Backup domain. As a result, after Reset, these bits are
write-protected and the DBP bit in the PWR power control register (PWR_CR) for
STM32F42xxx and STM32F43xxx has to be set before these can be modified. Refer to
Section 6.1.1: System reset on page 150 for further information. These bits are only reset
after a Backup domain Reset (see Section 6.1.3: Backup domain reset). Any internal or
external Reset will not have any effect on these bits.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
BDRST
Reserved
rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
LSEBY LSERD
RTCEN RTCSEL[1:0] LSEON
Reserved Reserved P Y
rw rw rw rw r rw

Bit 16 BDRST: Backup domain software reset
0: Reset not activated
1: Resets the entire Backup domain
Note: The BKPSRAM is not affected by this reset, the only way of resetting the BKPSRAM is
through the Flash interface when a protection level change from level 1 to level 0 is
requested.
Bit 15 RTCEN: RTC clock enable
0: RTC clock disabled
1: RTC clock enabled
RM0090 Rev 19 199/1751

212
Bits 9:8 RTCSEL[1:0]: RTC clock source selection

These bits are set by software to select the clock source for the RTC. Once the RTC clock
source has been selected, it cannot be changed anymore unless the Backup domain is
reset. The BDRST bit can be used to reset them.
00: No clock
01: LSE oscillator clock used as the RTC clock
10: LSI oscillator clock used as the RTC clock
11: HSE oscillator clock divided by a programmable prescaler (selection through the
RTCPRE[4:0] bits in the RCC clock configuration register (RCC_CFGR)) used as the RTC
clock
Bit 2 LSEBYP: External low-speed oscillator bypass
This bit is set and cleared by software to bypass the oscillator. This bit can be written only
when the LSE clock is disabled.
0: LSE oscillator not bypassed
1: LSE oscillator bypassed
Bit 1 LSERDY: External low-speed oscillator ready
This bit is set and cleared by hardware to indicate when the external 32 kHz oscillator is
stable. After the LSEON bit is cleared, LSERDY goes low after 6 external low-speed
oscillator clock cycles.
0: LSE clock not ready
1: LSE clock ready
Bit 0 LSEON: External low-speed oscillator enable
0: LSE clock OFF
1: LSE clock ON
6.3.21 RCC clock control & status register (RCC_CSR)

Reset value: 0x0E00 0000, reset by system reset, except reset flags by power reset only.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
LPWR WWDG IWDG SFT POR PIN BORRS
RMVF
RSTF RSTF RSTF RSTF RSTF RSTF TF Reserved
r r r r r r r rt_w
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
LSIRDY LSION
Reserved
r rw
200/1751 RM0090 Rev 19

Bit 31 LPWRRSTF: Low-power reset flag

This bit is set by hardware when a Low-power management reset occurs.
Cleared by writing to the RMVF bit.
0: No Low-power management reset occurred
1: Low-power management reset occurred
For further information on Low-power management reset, refer to Low-power management
reset.
Bit 30 WWDGRSTF: Window watchdog reset flag
This bit is set by hardware when a window watchdog reset occurs.
0: No window watchdog reset occurred
1: Window watchdog reset occurred
Bit 29 IWDGRSTF: Independent watchdog reset flag
This bit is set by hardware when an independent watchdog reset from VDD domain occurs.
0: No watchdog reset occurred
1: Watchdog reset occurred
Bit 28 SFTRSTF: Software reset flag
This bit is set by hardware when a software reset occurs.
0: No software reset occurred
1: Software reset occurred
Bit 27 PORRSTF: POR/PDR reset flag
This bit is set by hardware when a POR/PDR reset occurs.
0: No POR/PDR reset occurred
1: POR/PDR reset occurred
Bit 26 PINRSTF: PIN reset flag
This bit is set by hardware when a reset from the NRST pin occurs.
0: No reset from NRST pin occurred
1: Reset from NRST pin occurred
Bit 25 BORRSTF: BOR reset flag
Cleared by software by writing the RMVF bit.
This bit is set by hardware when a POR/PDR or BOR reset occurs.
0: No POR/PDR or BOR reset occurred
1: POR/PDR or BOR reset occurred
Bit 24 RMVF: Remove reset flag
This bit is set by software to clear the reset flags.
0: No effect
1: Clear the reset flags
RM0090 Rev 19 201/1751

212

Bit 1 LSIRDY: Internal low-speed oscillator ready
This bit is set and cleared by hardware to indicate when the internal RC 40 kHz oscillator is
stable. After the LSION bit is cleared, LSIRDY goes low after 3 LSI clock cycles.
0: LSI RC oscillator not ready
1: LSI RC oscillator ready
Bit 0 LSION: Internal low-speed oscillator enable
0: LSI RC oscillator OFF
1: LSI RC oscillator ON
6.3.22 RCC spread spectrum clock generation register (RCC_SSCGR)

The spread spectrum clock generation is available only for the main PLL.
The RCC_SSCGR register must be written either before the main PLL is enabled or after
the main PLL disabled.
Note: For full details about PLL spread spectrum clock generation (SSCG) characteristics, refer to
the “Electrical characteristics” section in your device datasheet.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
SPR
SSCG
EAD INCSTEP
EN Reserved
SEL
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
INCSTEP MODPER
Bit 31 SSCGEN: Spread spectrum modulation enable

0: Spread spectrum modulation DISABLE. (To write after clearing CR[24]=PLLON bit)
1: Spread spectrum modulation ENABLE. (To write before setting CR[24]=PLLON bit)
Bit 30 SPREADSEL: Spread Select
To write before to set CR[24]=PLLON bit.
0: Center spread
1: Down spread
202/1751 RM0090 Rev 19


Bits 27:13 INCSTEP: Incrementation step
These bits are set and cleared by software. To write before setting CR[24]=PLLON bit.
Configuration input for modulation profile amplitude.
Bits 12:0 MODPER: Modulation period
These bits are set and cleared by software. To write before setting CR[24]=PLLON bit.
Configuration input for modulation profile period.
6.3.23 RCC PLLI2S configuration register (RCC_PLLI2SCFGR)

This register is used to configure the PLLI2S clock outputs according to the formulas:
f(VCO clock) = f(PLLI2S clock input) × (PLLI2SN / PLLM)
f(PLL I2S clock output) = f(VCO clock) / PLLI2SR
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PLLI2S PLLI2S PLLI2S
Reserv PLLI2SQ
R2 R1 R0 Reserved
ed
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PLLI2SN PLLI2SN PLLI2SN PLLI2SN PLLI2SN PLLI2SN PLLI2SN PLLI2SN PLLI2SN
Reserv 8 7 6 5 4 3 2 1 0 Reserved
ed
RM0090 Rev 19 203/1751

212

Bits 30:28 PLLI2SR: PLLI2S division factor for I2S clocks
These bits are set and cleared by software to control the I2S clock frequency. These bits
should be written only if the PLLI2S is disabled. The factor must be chosen in accordance
with the prescaler values inside the I2S peripherals, to reach 0.3% error when using
standard crystals and 0% error with audio crystals. For more information about I2S clock
frequency and precision, refer to Section 28.4.4: Clock generator in the I2S chapter.
Caution: The I2Ss requires a frequency lower than or equal to 192 MHz to work correctly.
I2S clock frequency = VCO frequency / PLLR with 2 ≤ PLLR ≤ 7
000: PLLR = 0, wrong configuration
010: PLLR = 2
...
111: PLLR = 7
Bits 27:24 PLLI2SQ: PLLI2S division factor for SAI1 clock
These bits are set and cleared by software to control the SAI1 clock frequency.
They should be written when the PLLI2S is disabled.
SAI1 clock frequency = VCO frequency / PLLI2SQ with 2 <= PLLI2SIQ <= 15
0000: PLLI2SQ = 0, wrong configuration
0001: PLLI2SQ = 1, wrong configuration
0010: PLLI2SQ = 2
0011: PLLI2SQ = 3
0100: PLLI2SQ = 4
0101: PLLI2SQ = 5
...
1111: PLLI2SQ = 15
204/1751 RM0090 Rev 19


Bits 14:6 PLLI2SN: PLLI2S multiplication factor for VCO
These bits are set and cleared by software to control the multiplication factor of the VCO.
These bits can be written only when PLLI2S is disabled. Only half-word and word accesses
are allowed to write these bits.
VCO output frequency = VCO input frequency × PLLI2SN with 50 ≤ PLLI2SN ≤ 432
000000000: PLLI2SN = 0, wrong configuration
...
000110010: PLLI2SN = 50
...
001100011: PLLI2SN = 99
001100100: PLLI2SN = 100
001100101: PLLI2SN = 101
001100110: PLLI2SN = 102
...
110110000: PLLI2SN = 432
...
RM0090 Rev 19 205/1751

212
6.3.24 RCC PLL configuration register (RCC_PLLSAICFGR)

This register is used to configure the PLLSAI clock outputs according to the formulas:
• f(VCO clock) = f(PLLSAI clock input) × (PLLSAIN / PLLM)
• f(PLLSAI1 clock output) = f(VCO clock) / PLLSAIQ
• f(PLL LCD clock output) = f(VCO clock) / PLLSAIR
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PLLSAIR PLLSAIQ
Reserved Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PLLSAIN
Reserved Reserved

Bits 30:28 PLLSAIR: PLLSAI division factor for LCD clock
Set and reset by software to control the LCD clock frequency.
These bits should be written when the PLLSAI is disabled.
LCD clock frequency = VCO frequency / PLLSAIR with 2 ≤ PLLSAIR ≤ 7
000: PLLSAIR = 0, wrong configuration
001: PLLSAIR = 1, wrong configuration
010: PLLSAIR = 2
...
111: PLLSAIR = 7
Bits 27:24 PLLSAIQ: PLLSAI division factor for SAI1 clock
Set and reset by software to control the frequency of SAI1 clock.
These bits should be written when the PLLSAI is disabled.
SAI1 clock frequency = VCO frequency / PLLSAIQ with 2 ≤ PLLSAIQ ≤15
0000: PLLSAIQ = 0, wrong configuration
0001: PLLSAIQ = 1, wrong configuration
...
0010: PLLSAIQ = 2
0011: PLLSAIQ = 3
0100: PLLSAIQ = 4
0101: PLLSAIQ = 5
...
1111: PLLSAIQ = 15
206/1751 RM0090 Rev 19


Bits 14:6 PLLSAIN: PLLSAI division factor for VCO
These bits can be written only when PLLSAI is disabled. Only half-word and word accesses
Caution: The software has to set these bits correctly to ensure that the VCO output frequency
is between 100 and 432 MHz.
VCO output frequency = VCO input frequency × PLLISAIN with 50 ≤ PLLISAIN ≤ 432
000000000: PLLISAIN = 0, wrong configuration
...
000110010: PLLISAIN = 50
...
001100011: PLLISAIN = 99
001100100: PLLISAIN = 100
001100101: PLLISAIN = 101
001100110: PLLISAIN = 102
...
110110000: PLLISAIN = 432
...
Bits 5:0 Reserved, must be kept at reset value
6.3.25 RCC Dedicated Clock Configuration Register (RCC_DCKCFGR)

The RCC_DCKCFGR register allows to configure the timer clock prescalers and the PLLSAI
and PLLI2S output clock dividers for SAI1 and LTDC peripherals according to the following
formula:
f(PLLSAIDIVQ clock output) = f(PLLSAI_Q) / PLLSAIDIVQ
f(PLLSAIDIVR clock output) = f(PLLSAI_R) / PLLSAIDIVR
f(PLLI2SDIVQ clock output) = f(PLLI2S_Q) / PLLI2SDIVQ
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
TIMPRE SAI1BSRC SAI1ASRC PLLSAIDIVR
Reserved Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PLLSAIDIVQ PLLS2DIVQ
Reserved Reserved
RM0090 Rev 19 207/1751

212

Bit 24 TIMPRE: Timers clocks prescalers selection
This bit is set and reset by software to control the clock frequency of all the timers connected
to APB1 and APB2 domain.
0: If the APB prescaler (PPRE1, PPRE2 in the RCC_CFGR register) is configured to a
division factor of 1, TIMxCLK = PCLKx. Otherwise, the timer clock frequencies are set to
twice to the frequency of the APB domain to which the timers are connected:
TIMxCLK = 2xPCLKx.
1:If the APB prescaler (PPRE1, PPRE2 in the RCC_CFGR register) is configured to a
division factor of 1, 2 or 4, TIMxCLK = HCLK. Otherwise, the timer clock frequencies are set
to four times to the frequency of the APB domain to which the timers are connected:
TIMxCLK = 4xPCLKx.
Bits 23:22 SAI1BSRC: SAI1-B clock source selection
These bits are set and cleared by software to control the SAI1-B clock frequency.
They should be written when the PLLSAI and PLLI2S are disabled.
00: SAI1-B clock frequency = f(PLLSAI_Q) / PLLSAIDIVQ
01: SAI1-B clock frequency = f(PLLI2S_Q) / PLLI2SDIVQ
10: SAI1-B clock frequency = Alternate function input frequency
11: wrong configuration
Bits 21:20 SAI1ASRC: SAI1-A clock source selection
These bits are set and cleared by software to control the SAI1-A clock frequency.
They should be written when the PLLSAI and PLLI2S are disabled.
00: SAI1-A clock frequency = f(PLLSAI_Q) / PLLSAIDIVQ
01: SAI1-A clock frequency = f(PLLI2S_Q) / PLLI2SDIVQ
10: SAI1-A clock frequency = Alternate function input frequency
11: wrong configuration
Bits 17:16 PLLSAIDIVR: division factor for LCD_CLK
These bits are set and cleared by software to control the frequency of LCD_CLK.
They should be written only if PLLSAI is disabled.
LCD_CLK frequency = f(PLLSAI_R) / PLLSAIDIVR with 2 ≤ PLLSAIDIVR ≤ 16
00: PLLSAIDIVR = /2
01: PLLSAIDIVR = /4
10: PLLSAIDIVR = /8
11: PLLSAIDIVR = /16
208/1751 RM0090 Rev 19

Bits 12:8 PLLSAIDIVQ: PLLSAI division factor for SAI1 clock

These bits are set and reset by software to control the SAI1 clock frequency.
They should be written only if PLLSAI is disabled.
SAI1 clock frequency = f(PLLSAI_Q) / PLLSAIDIVQ with 1 ≤ PLLSAIDIVQ ≤ 31
00000: PLLSAIDIVQ = /1
...
Bits 4:0 PLLI2SDIVQ: PLLI2S division factor for SAI1 clock
These bits are set and reset by software to control the SAI1 clock frequency.
They should be written only if PLLI2S is disabled.
SAI1 clock frequency = f(PLLI2S_Q) / PLLI2SDIVQ with 1 <= PLLI2SDIVQ <= 31
00000: PLLI2SDIVQ = /1
...
11111: PLLI2SDIVQ = /32
RM0090 Rev 19 209/1751

212
0x00
0x38
0x34
0x30
0x28
0x24
0x20
0x18
0x14
0x10
0x08
0x04
0x0C
0x3C
0x2C
0x1C
Addr.
offset
6.3.26
210/1751
R
TR
TR
TR
TR
TR
RCC_
RCC_
RCC_
name
RCC_CR
Reserved
Reserved
Reserved
Reserved
RCC_CIR
AHB3ENR
AHB2ENR
AHB1ENR
Register
RCC_CFGR
RCC_PLLCFG
RCC_APB2RS
RCC_APB1RS
RCC_AHB3RS
RCC_AHB2RS
RCC_AHB1RS
Reserved UART8RST MCO2 1 31
Reserved Reserved
OTGHSULPIEN UART7RST MCO2 0 30
OTGHSEN DACRST OTGHSRST MCO2PRE2 PLL SAIRDY 29
Reserved
ETHMACPTPEN PWRRST MCO2PRE1 PLL SAION
Reserved
28
ETHMACRXEN Reserved Reserved MCO2PRE0 PLLQ 3 PLL I2SRDY 27
Reserved
ETHMACTXEN LTDCRST CAN2RST MCO1PRE2 PLLQ 2 PLL I2SON 26
RCC register map
ETHMACEN CAN1RST ETHMACRST MCO1PRE1 PLLQ 1 PLL RDY 25

Reserved Reserved Reserved Reserved MCO1PRE0 PLLQ 0 PLL ON 24
DMA2DEN I2C3RST DMA2DRST CSSC I2SSRC Reserved 23
DMA2EN SAI1RST I2C2RST DMA2RST PLLSAIRDYC MCO1 1 PLLSRC 22
DMA1EN SPI6RST I2C1RST DMA1RST PLLI2SRDYC MCO1 0 21
Reserved
CCMDATARAMEN SPI5RST UART5RST PLLRDYC RTCPRE 4 20

Reserved Reserved UART4RST HSERDYC RTCPRE 3 CSSON 19
Reserved
Reserved
Reserved
BKPSRAMEN TIM11RST UART3RST HSIRDYC RTCPRE 2 HSEBYP 18

TIM10RST UART2RST LSERDYC RTCPRE 1 PLLP 1 HSERDY 17
RM0090 Rev 19
TIM9RST Reserved LSIRDYC RTCPRE 0 PLLP 0 HSEON 16
Reserved
Table 33 gives the register map and reset values.
Reserved Reserved SPI3RST Reserved PPRE2 2 Reserved HSICAL 7
Reserved
Reserved
15
Reserved
Reserved
Reserved
Reserved
SYSCFGRST SPI2RST PLLSAIRDYIE PPRE2 1 PLLN 8 HSICAL 6 14
SP45RST PLLI2SRDYIE PPRE2 0 PLLN 7 HSICAL 5 13
Reserved
CRCEN SPI1RST CRCRST PLLRDYIE PPRE1 2 PLLN 6 HSICAL 4 12
PPRE1 1
Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)
Reserved SDIORST WWDGRST Reserved HSERDYIE PLLN 5 HSICAL 3 11

GPIOKEN GPIOKRST HSIRDYIE PPRE1 0 PLLN 4 HSICAL 2 10
Reserved Reserved
GPIOJEN GPIOJRST LSERDYIE PLLN 3 HSICAL 1 9
Reserved
GPIOIEN ADCRST TIM14RST GPIOIRST LSIRDYIE PLLN 2 HSICAL 0 8
OTGFSEN GPIOHEN TIM13RST OTGFSRS GPIOHRST CSSF HPRE 3 PLLN 1 HSITRIM 4 7
Reserved
RNGEN GPIOGEN TIM12RST RNGRST GPIOGRST PLLSAIRDYF HPRE 2 PLLN 0 HSITRIM 3 6
HASHEN GPIOFEN USART6RST TIM7RST HSAHRST GPIOFRST PLLI2SRDYF HPRE 1 PLLM 5 HSITRIM 2 5
CRYPEN GPIOEEN USART1RST TIM6RST CRYPRST GPIOERST PLLRDYF HPRE 0 PLLM 4 HSITRIM 1 4
Table 33. RCC register map and reset values for STM32F42xxx and STM32F43xxx
GPIODEN TIM5RST GPIODRST HSERDYF SWS 1 PLLM 3 HSITRIM 0 3

Reserved
Reserved GPIOCEN TIM4RST Reserved GPIOCRST HSIRDYF SWS 0 PLLM 2 Reserved 2
GPIOBEN TIM8RST TIM3RST GPIOBRST LSERDYF SW 1 PLLM 1 HSIRDY 1
FMCEN DCMIEN GPIOAEN TIM1RST TIM2RST FMCRS DCMIRST GPIOARST LSIRDYF SW 0 PLLM 0 HSION 0
RM0090
0x84
0x74
0x70
0x80
0x78
0x68
0x64
0x60
0x58
0x54
0x50
0x48
0x44
0x40
0x7C
0x6C
0x5C
0x4C
Addr.
offset
RM0090
FGR
ENR
ENR
ENR
ENR
ENR
RCC_
RCC_
name
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
APB2ENR
APB1ENR
RCC_CSR
Register
RCC_BDCR
RCC_SSCGR
RCC_APB2LP
RCC_APB1LP
RCC_AHB3LP
RCC_AHB2LP
RCC_AHB1LP
RCC_PLLI2SC
Reserved SSCGEN LPWRRSTF UART8LPEN Reserved UART8EN 31
SPREADSEL WWDGRSTF UART7LPEN OTGHSULPILPEN UART7EN 30
WDGRSTF DACLPEN OTGHSLPEN DACEN 29
Reserved
SFTRSTF PWRLPEN ETHMACPTPLPEN PWREN
Reserved
Reserved
PLLI2SRx
28
PORRSTF Reserved ETHMACRXLPEN Reserved 27
PADRSTF LTDCLPEN CAN2LPEN ETHMACTXLPEN LTDCEN CAN2EN 26
BORRSTF CAN1LPEN ETHMACLPEN CAN1EN 25
PLLI2SQ
RMVF Reserved Reserved Reserved Reserved Reserved 24
I2C3LPEN DMA2DLPEN I2C3EN
Reserved
23
SAI1LPEN I2C2LPEN DMA2LPEN SAI1EN I2C2EN 22
SPI6LPEN I2C1LPEN DMA1LPEN SPI6EN I2C1EN 21
SPI5LPEN UART5LPEN Reserved SPI5EN UART5EN 20
Reserved UART4LPEN SRAM3LPEN Reserved UART4EN
INCSTEP
19
Reserved
TIM11LPEN USART3LPEN BKPSRAMLPEN TIM11EN USART3EN
Reserved
18
TIM10LPEN USART2LPEN SRAM2LPEN TIM10EN USART2EN 17
RM0090 Rev 19
BDRST TIM9LPEN Reserved SRAM1LPEN TIM9EN Reserved 16
RTCEN Reserved SPI3LPEN FLITFLPEN Reserved SPI3EN
Reserved
15
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
SYSCFGLPEN SPI2LPEN SYSCFGEN SPI2EN 14

Reserved
SPI4LPEN SPI4EN 13
Reserved Reserved
SPI1LPEN CRCLPEN SPI1EN 12
Reserved
SDIOLPEN WWDGLPEN Reserved SDIOEN WWDGEN
Reserved
11
ADC3LPEN GPIOKLPEN ADC3EN 10
Reserved Reserved
RTCSEL 1 ADC2LPEN GPIOJLPEN ADC2EN
PLLI2SNx
9
RTCSEL 0 ADC1LPEN TIM14LPEN GPIOILPEN ADC1EN TIM14EN 8
TIM13LPEN OTGFSLPEN GPIOHLPEN TIM13EN 7
Reserved Reserved
TIM12LPEN RNGLPEN GPIOGLPEN TIM12EN 6
USART6LPEN TIM7LPEN HASHLPEN GPIOFLPEN USART6EN TIM7EN
MODPER
5
USART1LPEN TIM6LPEN CRYPLPEN GPIOELPEN USART1EN TIM6EN
Reserved
4
TIM5LPEN GPIODLPEN TIM5EN 3
Reserved Reserved
LSEBYP TIM4LPEN Reserved GPIOCLPEN TIM4EN 2
Reserved
LSIRDY LSERDY TIM8LPEN TIM3LPEN GPIOBLPEN TIM8EN TIM3EN 1
Table 33. RCC register map and reset values for STM32F42xxx and STM32F43xxx (continued)
LSION LSEON TIM1LPEN TIM2LPEN FMCLPEN DCMILPEN GPIOALPEN TIM1EN TIM2EN 0

Reset and clock control for STM32F42xxx and STM32F43xxx (RCC)
211/1751
212
Table 33. RCC register map and reset values for STM32F42xxx and STM32F43xxx (continued)
Addr. Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
offset name
Reserved
Reserved
PLLSAIQ
PLLSAIR
PLLSAIN
RCC_PLLSAI
0x88
CFGR
PLLSAIDIVR
SAI1BSCR
SAI1ASCR
Reserved
TIMPRE
RCC_DCKCF
0x8C Reserved Reserved PLLSAIDIVQ Reserved PLLI2SDIVQ
GR
Refer to Section 2.3: Memory map for the register boundary addresses.
212/1751 RM0090 Rev 19

RM0090 Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)

STM32F405xx/07xx and STM32F415xx/17xx(RCC)
7.1 Reset
There are three types of reset, defined as system Reset, power Reset and backup domain
Reset.
7.1.1 System reset

A system reset sets all registers to their reset values except the reset flags in the clock
controller CSR register and the registers in the Backup domain (see Figure 4).
A system reset is generated when one of the following events occurs:
1. A low level on the NRST pin (external reset)
2. Window watchdog end of count condition (WWDG reset)
3. Independent watchdog end of count condition (IWDG reset)
4. A software reset (SW reset) (see Software reset)
5. Low-power management reset (see Low-power management reset)
Software reset
The reset source can be identified by checking the reset flags in the RCC clock control &
The SYSRESETREQ bit in Cortex®-M4 with FPU Application Interrupt and Reset Control
Register must be set to force a software reset on the device. Refer to the Cortex®-M4 with
FPU technical reference manual for more details.
RM0090 Rev 19 213/1751

266
Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC) RM0090

There are two ways of generating a low-power management reset:
1. Reset generated when entering the Standby mode:
This type of reset is enabled by resetting the nRST_STDBY bit in the user option bytes.
In this case, whenever a Standby mode entry sequence is successfully executed, the
device is reset instead of entering the Standby mode.
2. Reset when entering the Stop mode:
This type of reset is enabled by resetting the nRST_STOP bit in the user option bytes.
In this case, whenever a Stop mode entry sequence is successfully executed, the
device is reset instead of entering the Stop mode.
7.1.2 Power reset

A power reset is generated when one of the following events occurs:
1. Power-on/power-down reset (POR/PDR reset) or brownout (BOR) reset
2. When exiting the Standby mode
A power reset sets all registers to their reset values except the Backup domain (see
Figure 4)
These sources act on the NRST pin and it is always kept low during the delay phase. The
RESET service routine vector is fixed at address 0x0000_0004 in the memory map.
The system reset signal provided to the device is output on the NRST pin. The pulse
generator guarantees a minimum reset pulse duration of 20 µs for each internal reset
source. In case of an external reset, the reset pulse is generated while the NRST pin is
asserted low.
Figure 20. Simplified diagram of the reset circuit
VDD/VDDA
RPU
External System reset
reset Filter
NRST
WWDG reset
Pulse
IWDG reset
generator
Power reset
(min 20 μs)
Software reset
ai16095c
The Backup domain has two specific resets that affect only the Backup domain (see
Figure 4).
7.1.3 Backup domain reset

The backup domain reset sets all RTC registers and the RCC_BDCR register to their reset
values. The BKPSRAM is not affected by this reset. The only way of resetting the
BKPSRAM is through the Flash interface by requesting a protection level change from 1 to
0.
214/1751 RM0090 Rev 19

A backup domain reset is generated when one of the following events occurs:
1. Software reset, triggered by setting the BDRST bit in the RCC Backup domain control
register (RCC_BDCR).
2. VDD or VBAT power on, if both supplies have previously been powered off.
7.2 Clocks
Three different clock sources can be used to drive the system clock (SYSCLK):
• HSI oscillator clock
• HSE oscillator clock
• Main PLL (PLL) clock
The devices have the two following secondary clock sources:
• 32 kHz low-speed internal RC (LSI RC) which drives the independent watchdog and,
optionally, the RTC used for Auto-wakeup from the Stop/Standby mode.
• 32.768 kHz low-speed external crystal (LSE crystal) which optionally drives the RTC
clock (RTCCLK)
Each clock source can be switched on or off independently when it is not used, to optimize
power consumption.
RM0090 Rev 19 215/1751

266
Figure 21. Clock tree
Watchdog
enable IWDGCLK
LSI RC LSI to independent
32 kHz watchdog
RTC S E L[1:0]
RTC RTCCLK
OSC32_IN enable to RTC
LSE OS C LSE
32.768 kHz
OSC32_OUT
SYSCLK
MCO2 /1 to 5
HSE_RTC
LSE
MCO1 /1 to 5 Peripheral
/2 to 31 clock enable Ethernet
PTP clock
168 MHz max. HCLK

to AHB bus, core,
16 MHz Clock memory and DMA
HSI Enable
HS I R C /8 to Cortex System
timer
SW
HSI FCLK Cortex
free-running clock
HSE AHB
PRESC
PLLCLK / 1,2,..512
SYSCLK
168 MHz Peripheral
APBx clock enable APBx
max PRESC peripheral
/ 1,2,4,8,16 Peripheral clocks
clock enable APBx timer
if (APBx presc = 1x1
else x2 clocks
OSC _OUT HSE
4-26 MHz
HS E O S C
OS C _IN
/M
VCO /P
Peripheral 48 MHz
PLL48CK clock enable
/Q clocks
xN
/R
PLL
VCO /P
/Q I2SSRC
xN
PLLI2SCLK Peripheral I2S clocks
/R clock enable
PLLI2S
I2S_CKIN Ext. clock
ETH_MII_TX_CLK_MII Peripheral
PHY Ethernet
clock enable MACTXCLK
25 to 50 MHz
/2,20 MII_RMII_SEL in SYSCFG_PMC to Ethernet MAC
Peripheral
clock enable MACRXCLK
ETH_MII_RX_
CLK_MII Peripheral
clock enable MACRMIICLK Peripheral
OTG_HS_ULPI_CK clock enable USBHS
USB2.0 PHY ULPI clock
24 to 60 MHz
ai16088d
1. For full details about the internal and external clock source characteristics, refer to the Electrical characteristics section in
216/1751 RM0090 Rev 19

The clock controller provides a high degree of flexibility to the application in the choice of the
external crystal or the oscillator to run the core and peripherals at the highest frequency
and, guarantee the appropriate frequency for peripherals that need a specific clock like
Ethernet, USB OTG FS and HS, I2S and SDIO.
Several prescalers are used to configure the AHB frequency, the high-speed APB (APB2)
and the low-speed APB (APB1) domains. The maximum frequency of the AHB domain is
168 MHz. The maximum allowed frequency of the high-speed APB2 domain is 84 MHz. The
maximum allowed frequency of the low-speed APB1 domain is 42 MHz
All peripheral clocks are derived from the system clock (SYSCLK) except for:
• The USB OTG FS clock (48 MHz), the random analog generator (RNG) clock
(≤ 48 MHz) and the SDIO clock (≤ 48 MHz) which are coming from a specific output of
PLL (PLL48CLK)
• The I2S clock
To achieve high-quality audio performance, the I2S clock can be derived either from a
specific PLL (PLLI2S) or from an external clock mapped on the I2S_CKIN pin. For
more information about I2S clock frequency and precision, refer to Section 28.4.4:
Clock generator.
• The USB OTG HS (60 MHz) clock which is provided from the external PHY
• The Ethernet MAC clocks (TX, RX and RMII) which are provided from the external
PHY. For further information on the Ethernet configuration, please refer to
Section 33.4.4: MII/RMII selection in the Ethernet peripheral description. When the
Ethernet is used, the AHB clock frequency must be at least 25 MHz.
The RCC feeds the external clock of the Cortex System Timer (SysTick) with the AHB clock
(HCLK) divided by 8. The SysTick can work either with this clock or with the Cortex clock
(HCLK), configurable in the SysTick control and status register.
The timer clock frequencies are automatically set by hardware. There are two cases:
1. If the APB prescaler is 1, the timer clock frequencies are set to the same frequency as
that of the APB domain to which the timers are connected.
2. Otherwise, they are set to twice (×2) the frequency of the APB domain to which the
timers are connected.
FCLK acts as Cortex®-M4 with FPU free-running clock. For more details, refer to the
Cortex®-M4 with FPU technical reference manual.
7.2.1 HSE clock

The high speed external clock signal (HSE) can be generated from two possible clock
sources:
• HSE external crystal/ceramic resonator
• HSE external user clock
The resonator and the load capacitors have to be placed as close as possible to the
oscillator pins in order to minimize output distortion and startup stabilization time. The
loading capacitance values must be adjusted according to the selected oscillator.
RM0090 Rev 19 217/1751

266
Figure 22. HSE/ LSE clock sources

Hardware configuration
OSC_OUT
External clock
(HI-Z)
External
source
OSC_IN OSC_OUT
Crystal/ceramic
resonators
CL1 CL2
Load
capacitors
External source (HSE bypass)

In this mode, an external clock source must be provided. You select this mode by setting the
HSEBYP and HSEON bits in the RCC clock control register (RCC_CR). The external clock
signal (square, sinus or triangle) with ~50% duty cycle has to drive the OSC_IN pin while the
OSC_OUT pin should be left HI-Z. See Figure 22.
External crystal/ceramic resonator (HSE crystal)

The HSE has the advantage of producing a very accurate rate on the main clock.
The associated hardware configuration is shown in Figure 22. Refer to the electrical
characteristics section of the datasheet for more details.
The HSERDY flag in the RCC clock control register (RCC_CR) indicates if the high-speed
external oscillator is stable or not. At startup, the clock is not released until this bit is set by
hardware. An interrupt can be generated if enabled in the RCC clock interrupt register
(RCC_CIR).
The HSE Crystal can be switched on and off using the HSEON bit in the RCC clock control
register (RCC_CR).
7.2.2 HSI clock

The HSI clock signal is generated from an internal 16 MHz RC oscillator and can be used
directly as a system clock, or used as PLL input.
The HSI RC oscillator has the advantage of providing a clock source at low cost (no external
components). It also has a faster startup time than the HSE crystal oscillator however, even
with calibration the frequency is less accurate than an external crystal oscillator or ceramic
resonator.
218/1751 RM0090 Rev 19

Calibration
RC oscillator frequencies can vary from one chip to another due to manufacturing process
variations, this is why each device is factory calibrated by ST for 1% accuracy at TA= 25 °C.
After reset, the factory calibration value is loaded in the HSICAL[7:0] bits in the RCC clock
control register (RCC_CR).
If the application is subject to voltage or temperature variations this may affect the RC
oscillator speed. You can trim the HSI frequency in the application using the HSITRIM[4:0]
bits in the RCC clock control register (RCC_CR).
The HSIRDY flag in the RCC clock control register (RCC_CR) indicates if the HSI RC is
stable or not. At startup, the HSI RC output clock is not released until this bit is set by
hardware.
The HSI RC can be switched on and off using the HSION bit in the RCC clock control
register (RCC_CR).
The HSI signal can also be used as a backup source (Auxiliary clock) if the HSE crystal
oscillator fails. Refer to Section 7.2.7: Clock security system (CSS) on page 220.
7.2.3 PLL configuration

The STM32F4xx devices feature two PLLs:
• A main PLL (PLL) clocked by the HSE or HSI oscillator and featuring two different
output clocks:
– The first output is used to generate the high speed system clock (up to 168 MHz)
– The second output is used to generate the clock for the USB OTG FS (48 MHz),
the random analog generator (≤48 MHz) and the SDIO (≤ 48 MHz).
• A dedicated PLL (PLLI2S) used to generate an accurate clock to achieve high-quality
audio performance on the I2S interface.
Since the main-PLL configuration parameters cannot be changed once PLL is enabled, it is
recommended to configure PLL before enabling it (selection of the HSI or HSE oscillator as
PLL clock source, and configuration of division factors M, N, P, and Q).
The PLLI2S uses the same input clock as PLL (PLLM[5:0] and PLLSRC bits are common to
both PLLs). However, the PLLI2S has dedicated enable/disable and division factors (N and
R) configuration bits. Once the PLLI2S is enabled, the configuration parameters cannot be
changed.
The two PLLs are disabled by hardware when entering Stop and Standby modes, or when
an HSE failure occurs when HSE or PLL (clocked by HSE) are used as system clock. RCC
PLL configuration register (RCC_PLLCFGR) and RCC clock configuration register
(RCC_CFGR) can be used to configure PLL and PLLI2S, respectively.
7.2.4 LSE clock

The LSE clock is generated from a 32.768 kHz low-speed external crystal or ceramic
resonator. It has the advantage providing a low-power but highly accurate clock source to
the real-time clock peripheral (RTC) for clock/calendar or other timing functions.
The LSE oscillator is switched on and off using the LSEON bit in RCC Backup domain
control register (RCC_BDCR).
RM0090 Rev 19 219/1751

266
The LSERDY flag in the RCC Backup domain control register (RCC_BDCR) indicates if the
LSE crystal is stable or not. At startup, the LSE crystal output clock signal is not released
until this bit is set by hardware. An interrupt can be generated if enabled in the RCC clock
interrupt register (RCC_CIR).
External source (LSE bypass)

In this mode, an external clock source must be provided. It must have a frequency up to
1 MHz. You select this mode by setting the LSEBYP and LSEON bits in the RCC Backup
domain control register (RCC_BDCR). The external clock signal (square, sinus or triangle)
with ~50% duty cycle has to drive the OSC32_IN pin while the OSC32_OUT pin should be
left HI-Z. See Figure 22.
7.2.5 LSI clock

The LSI RC acts as an low-power clock source that can be kept running in Stop and
Standby mode for the independent watchdog (IWDG) and Auto-wakeup unit (AWU). The
clock frequency is around 32 kHz. For more details, refer to the electrical characteristics
section of the datasheets.
The LSI RC can be switched on and off using the LSION bit in the RCC clock control &
The LSIRDY flag in the RCC clock control & status register (RCC_CSR) indicates if the low-
speed internal oscillator is stable or not. At startup, the clock is not released until this bit is
set by hardware. An interrupt can be generated if enabled in the RCC clock interrupt register
(RCC_CIR).
7.2.6 System clock (SYSCLK) selection

After a system reset, the HSI oscillator is selected as the system clock. When a clock source
is used directly or through PLL as the system clock, it is not possible to stop it.
A switch from one clock source to another occurs only if the target clock source is ready
(clock stable after startup delay or PLL locked). If a clock source that is not yet ready is
selected, the switch occurs when the clock source is ready. Status bits in the RCC clock
control register (RCC_CR) indicate which clock(s) is (are) ready and which clock is currently
used as the system clock.
7.2.7 Clock security system (CSS)

The clock security system can be activated by software. In this case, the clock detector is
enabled after the HSE oscillator startup delay, and disabled when this oscillator is stopped.
If a failure is detected on the HSE clock, this oscillator is automatically disabled, a clock
failure event is sent to the break inputs of advanced-control timers TIM1 and TIM8, and an
interrupt is generated to inform the software about the failure (clock security system
interrupt CSSI), allowing the MCU to perform rescue operations. The CSSI is linked to the
Cortex®-M4 with FPU NMI (non-maskable interrupt) exception vector.
Note: When the CSS is enabled, if the HSE clock happens to fail, the CSS generates an interrupt,
which causes the automatic generation of an NMI. The NMI is executed indefinitely unless
the CSS interrupt pending bit is cleared. As a consequence, the application has to clear the
CSS interrupt in the NMI ISR by setting the CSSC bit in the Clock interrupt register
(RCC_CIR).
220/1751 RM0090 Rev 19

If the HSE oscillator is used directly or indirectly as the system clock (indirectly meaning that
it is directly used as PLL input clock, and that PLL clock is the system clock) and a failure is
detected, then the system clock switches to the HSI oscillator and the HSE oscillator is
disabled.
If the HSE oscillator clock was the clock source of PLL used as the system clock when the
failure occurred, PLL is also disabled. In this case, if the PLLI2S was enabled, it is also
disabled when the HSE fails.
7.2.8 RTC/AWU clock

Once the RTCCLK clock source has been selected, the only possible way of modifying the
selection is to reset the power domain.
The RTCCLK clock source can be either the HSE 1 MHz (HSE divided by a programmable
prescaler), the LSE or the LSI clock. This is selected by programming the RTCSEL[1:0] bits
in the RCC Backup domain control register (RCC_BDCR) and the RTCPRE[4:0] bits in RCC
clock configuration register (RCC_CFGR). This selection cannot be modified without
resetting the Backup domain.
If the LSE is selected as the RTC clock, the RTC operates normally if the backup or the
system supply disappears. If the LSI is selected as the AWU clock, the AWU state is not
guaranteed if the system supply disappears. If the HSE oscillator divided by a value
between 2 and 31 is used as the RTC clock, the RTC state is not guaranteed if the backup
or the system supply disappears.
The LSE clock is in the Backup domain, whereas the HSE and LSI clocks are not. As a
consequence:
• If LSE is selected as the RTC clock:
– The RTC continues to work even if the VDD supply is switched off, provided the
VBAT supply is maintained.
• If LSI is selected as the Auto-wakeup unit (AWU) clock:
– The AWU state is not guaranteed if the VDD supply is powered off. Refer to
Section 7.2.5: LSI clock on page 220 for more details on LSI calibration.
• If the HSE clock is used as the RTC clock:
– The RTC state is not guaranteed if the VDD supply is powered off or if the internal
voltage regulator is powered off (removing power from the 1.2 V domain).
Note: To read the RTC calendar register when the APB1 clock frequency is less than seven times
the RTC clock frequency (fAPB1 < 7xfRTCLCK), the software must read the calendar time and
date registers twice. The data are correct if the second read access to RTC_TR gives the
same result than the first one. Otherwise a third read access must be performed.
7.2.9 Watchdog clock

If the independent watchdog (IWDG) is started by either hardware option or software
access, the LSI oscillator is forced ON and cannot be disabled. After the LSI oscillator
temporization, the clock is provided to the IWDG.
RM0090 Rev 19 221/1751

266
7.2.10 Clock-out capability

Two microcontroller clock output (MCO) pins are available:
• MCO1
You can output four different clock sources onto the MCO1 pin (PA8) using the
– HSI clock
– LSE clock
– HSE clock
– PLL clock
The desired clock source is selected using the MCO1PRE[2:0] and MCO1[1:0] bits in
the RCC clock configuration register (RCC_CFGR).
• MCO2
You can output four different clock sources onto the MCO2 pin (PC9) using the
– HSE clock
– PLL clock
– System clock (SYSCLK)
– PLLI2S clock
The desired clock source is selected using the MCO2PRE[2:0] and MCO2 bits in the
RCC clock configuration register (RCC_CFGR).
For the different MCO pins, the corresponding GPIO port has to be programmed in alternate
function mode.
The selected clock to output onto MCO must not exceed 100 MHz (the maximum I/O
speed).
7.2.11 Internal/external clock measurement using TIM5/TIM11

It is possible to indirectly measure the frequencies of all on-board clock source generators
by means of the input capture of TIM5 channel4 and TIM11 channel1 as shown in Figure 23
and Figure 24.

by the I/O or by an internal clock. This selection is performed through the TI4_RMP [1:0] bits
in the TIM5_OR register.
The primary purpose of having the LSE connected to the channel4 input capture is to be
able to precisely measure the HSI (this requires to have the HSI used as the system clock
source). The number of HSI clock counts between consecutive edges of the LSE signal
provides a measurement of the internal clock period. Taking advantage of the high precision
of LSE crystals (typically a few tens of ppm) we can determine the internal clock frequency
with the same resolution, and trim the source to compensate for manufacturing-process
and/or temperature- and voltage-related frequency deviations.
The HSI oscillator has dedicated, user-accessible calibration bits for this purpose.
222/1751 RM0090 Rev 19

The basic concept consists in providing a relative measurement (e.g. HSI/LSE ratio): the
precision is therefore tightly linked to the ratio between the two clock sources. The greater
the ratio, the better the measurement.
It is also possible to measure the LSI frequency: this is useful for applications that do not
have a crystal. The ultralow-power LSI oscillator has a large manufacturing process
deviation: by measuring it versus the HSI clock source, it is possible to determine its
frequency with the precision of the HSI. The measured value can be used to have more
accurate RTC time base timeouts (when LSI is used as the RTC clock source) and/or an
IWDG timeout with an acceptable accuracy.
Use the following procedure to measure the LSI frequency:
1. Enable the TIM5 timer and configure channel4 in Input capture mode.
2. Set the TI4_RMP bits in the TIM5_OR register to 0x01 to connect the LSI clock
internally to TIM5 channel4 input capture for calibration purposes.
3. Measure the LSI clock frequency using the TIM5 capture/compare 4 event or interrupt.
4. Use the measured LSI frequency to update the prescaler of the RTC depending on the
desired time base and/or to compute the IWDG timeout.
TIM5
TI4_RMP[1:0]
GPIO
RTC_WakeUp_IT TI4
LSE
LSI
ai17741V2

by the I/O or by an internal clock. This selection is performed through TI1_RMP [1:0] bits in
the TIM11_OR register. The HSE_RTC clock (HSE divided by a programmable prescaler) is
connected to channel 1 input capture to have a rough indication of the external crystal
frequency. This requires that the HSI is the system clock source. This can be useful for
instance to ensure compliance with the IEC 60730/IEC 61335 standards which require to be
able to determine harmonic or subharmonic frequencies (–50/+100% deviations).
TIM11
TI1_RMP[1:0]
GPIO
TI1
HSE_RTC(1 MHz)
ai18433
RM0090 Rev 19 223/1751

266
7.3 RCC registers

7.3.1 RCC clock control register (RCC_CR)

Reset value: 0x0000 XX83 where X is undefined.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PLLI2S PLLI2S PLLRD CSS HSE HSE HSE
PLLON
Reserved RDY ON Y Reserved ON BYP RDY ON
r rw r rw rw rw r rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
HSI
HSICAL[7:0] HSITRIM[4:0] HSION
Res. RDY
r r r r r r r r rw rw rw rw rw r rw

Bit 27 PLLI2SRDY: PLLI2S clock ready flag
Set by hardware to indicate that the PLLI2S is locked.
0: PLLI2S unlocked
1: PLLI2S locked
Bit 26 PLLI2SON: PLLI2S enable
Set and cleared by software to enable PLLI2S.
0: PLLI2S OFF
1: PLLI2S ON
Bit 25 PLLRDY: Main PLL (PLL) clock ready flag
Set by hardware to indicate that PLL is locked.
0: PLL unlocked
1: PLL locked
Bit 24 PLLON: Main PLL (PLL) enable
Set and cleared by software to enable PLL.
Cleared by hardware when entering Stop or Standby mode. This bit cannot be reset if PLL
clock is used as the system clock.
0: PLL OFF
1: PLL ON
Bit 19 CSSON: Clock security system enable
Set and cleared by software to enable the clock security system. When CSSON is set, the
clock detector is enabled by hardware when the HSE oscillator is ready, and disabled by
hardware if an oscillator failure is detected.
0: Clock security system OFF (Clock detector OFF)
1: Clock security system ON (Clock detector ON if HSE oscillator is stable, OFF if not)
224/1751 RM0090 Rev 19

Bit 18 HSEBYP: HSE clock bypass

Set and cleared by software to bypass the oscillator with an external clock. The external
clock must be enabled with the HSEON bit, to be used by the device.
The HSEBYP bit can be written only if the HSE oscillator is disabled.
0: HSE oscillator not bypassed
1: HSE oscillator bypassed with an external clock
Bit 17 HSERDY: HSE clock ready flag
Set by hardware to indicate that the HSE oscillator is stable. After the HSEON bit is cleared,
HSERDY goes low after 6 HSE oscillator clock cycles.
0: HSE oscillator not ready
1: HSE oscillator ready
Bit 16 HSEON: HSE clock enable
Cleared by hardware to stop the HSE oscillator when entering Stop or Standby mode. This
bit cannot be reset if the HSE oscillator is used directly or indirectly as the system clock.
0: HSE oscillator OFF
1: HSE oscillator ON
Bits 15:8 HSICAL[7:0]: Internal high-speed clock calibration
These bits are initialized automatically at startup.
Bits 7:3 HSITRIM[4:0]: Internal high-speed clock trimming
These bits provide an additional user-programmable trimming value that is added to the
HSICAL[7:0] bits. It can be programmed to adjust to variations in voltage and temperature
that influence the frequency of the internal HSI RC.
Bit 1 HSIRDY: Internal high-speed clock ready flag
Set by hardware to indicate that the HSI oscillator is stable. After the HSION bit is cleared,
HSIRDY goes low after 6 HSI clock cycles.
0: HSI oscillator not ready
1: HSI oscillator ready
Bit 0 HSION: Internal high-speed clock enable
Set by hardware to force the HSI oscillator ON when leaving the Stop or Standby mode or in
case of a failure of the HSE oscillator used directly or indirectly as the system clock. This bit
cannot be cleared if the HSI is used directly or indirectly as the system clock.
0: HSI oscillator OFF
1: HSI oscillator ON
RM0090 Rev 19 225/1751

266
7.3.2 RCC PLL configuration register (RCC_PLLCFGR)

This register is used to configure the PLL clock outputs according to the formulas:
• f(VCO clock) = f(PLL clock input) × (PLLN / PLLM)
• f(PLL general clock output) = f(VCO clock) / PLLP
• f(USB OTG FS, SDIO, RNG clock output) = f(VCO clock) / PLLQ
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PLLSR
PLLQ3 PLLQ2 PLLQ1 PLLQ0 Reserv PLLP1 PLLP0
Reserved C Reserved
ed
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserv PLLN PLLM5 PLLM4 PLLM3 PLLM2 PLLM1 PLLM0

ed rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 27:24 PLLQ: Main PLL (PLL) division factor for USB OTG FS, SDIO and random number generator
clocks
Set and cleared by software to control the frequency of USB OTG FS clock, the random
number generator clock and the SDIO clock. These bits should be written only if PLL is
disabled.
Caution: The USB OTG FS requires a 48 MHz clock to work correctly. The SDIO and the
random number generator need a frequency lower than or equal to 48 MHz to work
correctly.
USB OTG FS clock frequency = VCO frequency / PLLQ with 2 ≤ PLLQ ≤ 15
0010: PLLQ = 2
0011: PLLQ = 3
0100: PLLQ = 4
...
1111: PLLQ = 15
Bit 22 PLLSRC: Main PLL(PLL) and audio PLL (PLLI2S) entry clock source
Set and cleared by software to select PLL and PLLI2S clock source. This bit can be written
only when PLL and PLLI2S are disabled.
0: HSI clock selected as PLL and PLLI2S clock entry
1: HSE oscillator clock selected as PLL and PLLI2S clock entry
226/1751 RM0090 Rev 19

Bits 17:16 PLLP: Main PLL (PLL) division factor for main system clock
Set and cleared by software to control the frequency of the general PLL output clock. These
bits can be written only if PLL is disabled.
PLL output clock frequency = VCO frequency / PLLP with PLLP = 2, 4, 6, or 8
00: PLLP = 2
01: PLLP = 4
10: PLLP = 6
11: PLLP = 8
Bits 14:6 PLLN: Main PLL (PLL) multiplication factor for VCO
Set and cleared by software to control the multiplication factor of the VCO. These bits can
be written only when PLL is disabled. Only half-word and word accesses are allowed to
write these bits.
VCO output frequency = VCO input frequency × PLLN with 50 ≤ PLLN ≤ 432
...
000110010: PLLN = 50
...
001100011: PLLN = 99
001100100: PLLN = 100
...
110110000: PLLN = 432
...
Bits 5:0 PLLM: Division factor for the main PLL (PLL) and audio PLL (PLLI2S) input clock
Set and cleared by software to divide the PLL and PLLI2S input clock before the VCO.
These bits can be written only when the PLL and PLLI2S are disabled.
Caution: The software has to set these bits correctly to ensure that the VCO input frequency
ranges from 1 to 2 MHz. It is recommended to select a frequency of 2 MHz to limit
PLL jitter.
VCO input frequency = PLL input clock frequency / PLLM with 2 ≤ PLLM ≤ 63
000010: PLLM = 2
000011: PLLM = 3
000100: PLLM = 4
...
111110: PLLM = 62
111111: PLLM = 63
RM0090 Rev 19 227/1751

266
7.3.3 RCC clock configuration register (RCC_CFGR)

1 or 2 wait states inserted only if the access occurs during a clock source switch.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
I2SSC
MCO2 MCO2 PRE[2:0] MCO1 PRE[2:0] MCO1 RTCPRE[4:0]
R
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PPRE2[2:0] PPRE1[2:0] HPRE[3:0] SWS1 SWS0 SW1 SW0
Reserved
rw rw rw rw rw rw rw rw rw rw r r rw rw
Bits 31:30 MCO2[1:0]: Microcontroller clock output 2

oscillators and the PLLs.
00: System clock (SYSCLK) selected
01: PLLI2S clock selected
prescaler only after reset before enabling the external oscillators and the PLLs.
0xx: no division
100: division by 2
101: division by 3
110: division by 4
111: division by 5
prescaler only after reset before enabling the external oscillators and the PLL.
0xx: no division
100: division by 2
101: division by 3
110: division by 4
111: division by 5
Bit 23 I2SSRC: I2S clock selection
Set and cleared by software. This bit allows to select the I2S clock source between the
PLLI2S clock and the external clock. It is highly recommended to change this bit only after
reset and before enabling the I2S module.
0: PLLI2S clock used as I2S clock source
1: External clock mapped on the I2S_CKIN pin used as I2S clock source
228/1751 RM0090 Rev 19

Bits 22:21 MCO1: Microcontroller clock output 1

oscillators and PLL.
00: HSI clock selected
01: LSE oscillator selected
Bits 20:16 RTCPRE: HSE division factor for RTC clock
Set and cleared by software to divide the HSE clock input clock to generate a 1 MHz clock
for RTC.
Caution: The software has to set these bits correctly to ensure that the clock supplied to the
RTC is 1 MHz. These bits must be configured if needed before selecting the RTC
clock source.
00000: no clock
00001: no clock
00010: HSE/2
00011: HSE/3
00100: HSE/4
...
11110: HSE/30
11111: HSE/31
Bits 15:13 PPRE2: APB high-speed prescaler (APB2)
Set and cleared by software to control APB high-speed clock division factor.
PPRE2 write.
Bits 12:10 PPRE1: APB Low speed prescaler (APB1)
Set and cleared by software to control APB low-speed clock division factor.
PPRE1 write.
RM0090 Rev 19 229/1751

266
Bits 7:4 HPRE: AHB prescaler

Set and cleared by software to control AHB clock division factor.
Caution: The clocks are divided with the new prescaler factor from 1 to 16 AHB cycles after
HPRE write.
Caution: The AHB clock frequency must be at least 25 MHz when the Ethernet is used.
0xxx: system clock not divided
Bits 3:2 SWS: System clock switch status
Set and cleared by hardware to indicate which clock source is used as the system clock.
00: HSI oscillator used as the system clock
01: HSE oscillator used as the system clock
10: PLL used as the system clock
11: not applicable
Bits 1:0 SW: System clock switch
Set and cleared by software to select the system clock source.
Set by hardware to force the HSI selection when leaving the Stop or Standby mode or in
case of failure of the HSE oscillator used directly or indirectly as the system clock.
00: HSI oscillator selected as system clock
01: HSE oscillator selected as system clock
10: PLL selected as system clock
11: not allowed
7.3.4 RCC clock interrupt register (RCC_CIR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PLLI2S PLL HSE HSI LSE LSI
CSSC Reserv RDYC
Reserved RDYC RDYC RDYC RDYC RDYC
ed
w w w w w w w
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PLLI2S PLL HSE HSI LSE LSI PLLI2S PLL HSE HSI LSE LSI
CSSF Reserv
Reserved RDYIE RDYIE RDYIE RDYIE RDYIE RDYIE RDYF RDYF RDYF RDYF RDYF RDYF
ed
rw rw rw rw rw rw r r r r r r r
230/1751 RM0090 Rev 19


Bit 23 CSSC: Clock security system interrupt clear
This bit is set by software to clear the CSSF flag.
0: No effect
1: Clear CSSF flag
Bits 22 Reserved, must be kept at reset value.
Bit 21 PLLI2SRDYC: PLLI2S ready interrupt clear
This bit is set by software to clear the PLLI2SRDYF flag.
0: No effect
1: PLLI2SRDYF cleared
Bit 20 PLLRDYC: Main PLL(PLL) ready interrupt clear
This bit is set by software to clear the PLLRDYF flag.
0: No effect
1: PLLRDYF cleared
Bit 19 HSERDYC: HSE ready interrupt clear
This bit is set by software to clear the HSERDYF flag.
0: No effect
1: HSERDYF cleared
Bit 18 HSIRDYC: HSI ready interrupt clear
This bit is set software to clear the HSIRDYF flag.
0: No effect
1: HSIRDYF cleared
Bit 17 LSERDYC: LSE ready interrupt clear
This bit is set by software to clear the LSERDYF flag.
0: No effect
1: LSERDYF cleared
Bit 16 LSIRDYC: LSI ready interrupt clear
This bit is set by software to clear the LSIRDYF flag.
0: No effect
1: LSIRDYF cleared
Bit 13 PLLI2SRDYIE: PLLI2S ready interrupt enable
Set and cleared by software to enable/disable interrupt caused by PLLI2S lock.
0: PLLI2S lock interrupt disabled
1: PLLI2S lock interrupt enabled
Bit 12 PLLRDYIE: Main PLL (PLL) ready interrupt enable
Set and cleared by software to enable/disable interrupt caused by PLL lock.
0: PLL lock interrupt disabled
1: PLL lock interrupt enabled
Bit 11 HSERDYIE: HSE ready interrupt enable
Set and cleared by software to enable/disable interrupt caused by the HSE oscillator
stabilization.
0: HSE ready interrupt disabled
1: HSE ready interrupt enabled
RM0090 Rev 19 231/1751

266
Bit 10 HSIRDYIE: HSI ready interrupt enable

Set and cleared by software to enable/disable interrupt caused by the HSI oscillator
stabilization.
0: HSI ready interrupt disabled
1: HSI ready interrupt enabled
Bit 9 LSERDYIE: LSE ready interrupt enable
Set and cleared by software to enable/disable interrupt caused by the LSE oscillator
stabilization.
0: LSE ready interrupt disabled
1: LSE ready interrupt enabled
Bit 8 LSIRDYIE: LSI ready interrupt enable
Set and cleared by software to enable/disable interrupt caused by LSI oscillator
stabilization.
0: LSI ready interrupt disabled
1: LSI ready interrupt enabled
Bit 7 CSSF: Clock security system interrupt flag
Set by hardware when a failure is detected in the HSE oscillator.
Cleared by software setting the CSSC bit.
0: No clock security interrupt caused by HSE clock failure
1: Clock security interrupt caused by HSE clock failure
Bit 5 PLLI2SRDYF: PLLI2S ready interrupt flag
Set by hardware when the PLLI2S locks and PLLI2SRDYDIE is set.
Cleared by software setting the PLLRI2SDYC bit.
0: No clock ready interrupt caused by PLLI2S lock
1: Clock ready interrupt caused by PLLI2S lock
Bit 4 PLLRDYF: Main PLL (PLL) ready interrupt flag
Set by hardware when PLL locks and PLLRDYDIE is set.
Cleared by software setting the PLLRDYC bit.
0: No clock ready interrupt caused by PLL lock
1: Clock ready interrupt caused by PLL lock
Bit 3 HSERDYF: HSE ready interrupt flag
Set by hardware when External High Speed clock becomes stable and HSERDYDIE is set.
Cleared by software setting the HSERDYC bit.
0: No clock ready interrupt caused by the HSE oscillator
1: Clock ready interrupt caused by the HSE oscillator
232/1751 RM0090 Rev 19

Bit 2 HSIRDYF: HSI ready interrupt flag

Set by hardware when the Internal High Speed clock becomes stable and HSIRDYDIE is
set.
Cleared by software setting the HSIRDYC bit.
0: No clock ready interrupt caused by the HSI oscillator
1: Clock ready interrupt caused by the HSI oscillator
Bit 1 LSERDYF: LSE ready interrupt flag
Set by hardware when the External Low Speed clock becomes stable and LSERDYDIE is
set.
Cleared by software setting the LSERDYC bit.
0: No clock ready interrupt caused by the LSE oscillator
1: Clock ready interrupt caused by the LSE oscillator
Bit 0 LSIRDYF: LSI ready interrupt flag
Set by hardware when the internal low speed clock becomes stable and LSIRDYDIE is set.
Cleared by software setting the LSIRDYC bit.
0: No clock ready interrupt caused by the LSI oscillator
1: Clock ready interrupt caused by the LSI oscillator

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
OTGH
ETHMAC DMA2 DMA1
S
Reserved Reserved RST Reserved RST RST Reserved
RST
rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CRCR GPIOI GPIOH GPIOGG GPIOF GPIOE GPIOD GPIOC GPIOB GPIOA
Reserved ST Reserved RST RST RST RST RST RST RST RST RST

Bit 29 OTGHSRST: USB OTG HS module reset
0: does not reset the USB OTG HS module
1: resets the USB OTG HS module
Bit 25 ETHMACRST: Ethernet MAC reset
0: does not reset Ethernet MAC
1: resets Ethernet MAC
RM0090 Rev 19 233/1751

266

1: resets DMA2
1: resets DMA1
Bit 12 CRCRST: CRC reset
0: does not reset CRC
1: resets CRC
Bit 8 GPIOIRST: IO port I reset
0: does not reset IO port I
1: resets IO port I
Bit 7 GPIOHRST: IO port H reset
0: does not reset IO port H
1: resets IO port H
Bits 6 GPIOGRST: IO port G reset
0: does not reset IO port G
1: resets IO port G
Bit 5 GPIOFRST: IO port F reset
0: does not reset IO port F
1: resets IO port F
Bit 4 GPIOERST: IO port E reset
0: does not reset IO port E
1: resets IO port E
Bit 3 GPIODRST: IO port D reset
0: does not reset IO port D
1: resets IO port D
234/1751 RM0090 Rev 19

Bit 2 GPIOCRST: IO port C reset

0: does not reset IO port C
1: resets IO port C
Bit 1 GPIOBRST: IO port B reset
0: does not reset IO port B
1:resets IO port B
Bit 0 GPIOARST: IO port A reset
0: does not reset IO port A
1: resets IO port A
RM0090 Rev 19 235/1751

266

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserved RST RST RST RST Reserved RST
rw rw rw rw rw

Bit 7 OTGFSRST: USB OTG FS module reset
0: does not reset the USB OTG FS module
1: resets the USB OTG FS module
Bit 6 RNGRST: Random number generator module reset
0: does not reset the random number generator module
1: resets the random number generator module
Bit 5 HASHRST: Hash module reset
0: does not reset the HASH module
1: resets the HASH module
Bit 4 CRYPRST: Cryptographic module reset
0: does not reset the cryptographic module
1: resets the cryptographic module
Bit 0 DCMIRST: Camera interface reset
0: does not reset the Camera interface
1: resets the Camera interface
236/1751 RM0090 Rev 19


31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FSMCRST
Reserved
rw

Bit 0 FSMCRST: Flexible static memory controller module reset
0: does not reset the FSMC module
1: resets the FSMC module

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PWR CAN2 CAN1 I2C3 I2C2 I2C1 UART5 UART4 UART3 UART2
DACRST
RST Reser- RST RST Reser- RST RST RST RST RST RST RST Reser-
Reserved
ved ved ved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RST RST RST RST RST RST RST RST RST RST RST RST
Reserved Reserved

Bit 29 DACRST: DAC reset
0: does not reset the DAC interface
1: resets the DAC interface
Bit 28 PWRRST: Power interface reset
0: does not reset the power interface
1: resets the power interface
RM0090 Rev 19 237/1751

266

1: resets CAN2
1: resets CAN1
1: resets I2C3
1: resets I2C2
1: resets I2C1
1: resets UART5
Bit 19 UART4RST: USART4 reset
1: resets UART4
1: resets USART3
1: resets USART2
1: resets SPI3
1: resets SPI2
238/1751 RM0090 Rev 19


Bit 11 WWDGRST: Window watchdog reset
0: does not reset the window watchdog
1: resets the window watchdog
1: resets TIM14
1: resets TIM13
1: resets TIM12
1: resets TIM7
1: resets TIM6
1: resets TIM5
1: resets TIM4
1: resets TIM3
1: resets TIM2
RM0090 Rev 19 239/1751

266

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
TIM11 TIM10 TIM9
Reserved RST RST RST
rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
USART USART
SYSCF SPI1 SDIO ADC TIM8 TIM1
Reser- 6 1
G RST Reser- RST RST Reserved RST Reserved Reserved RST RST
ved ved RST RST

1: resets TIM14
1: resets TIM10
1: resets TIM9
Bit 14 SYSCFGRST: System configuration controller reset
0: does not reset the System configuration controller
1: resets the System configuration controller
1: resets SPI1
Bit 11 SDIORST: SDIO reset
0: does not reset the SDIO module
1: resets the SDIO module
240/1751 RM0090 Rev 19

Bit 8 ADCRST: ADC interface reset (common to all ADCs)

0: does not reset the ADC interface
1: resets the ADC interface
1: resets USART6
1: resets USART1
1: resets TIM8
1: resets TIM1
RM0090 Rev 19 241/1751

266

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
OTGH
ETHM ETHM ETHM
S OTGH ETHMA DMA2E DMA1E CCMDAT BKPSR
Reser- ULPIE ACPTP ACRXE ACTXE Res.
SEN CEN Reserved N N ARAMEN AMEN Reserved
ved EN N N
N
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CRCE GPIOIE GPIOH GPIOG GPIOFE GPIOD GPIOC GPIO GPIO
GPIOEEN
Reserved N Reserved N EN EN N EN EN BEN AEN

Bit 30 OTGHSULPIEN: USB OTG HSULPI clock enable
Set and cleared by software. This bit must be cleared when the OTG_HS is used in FS
mode.
0: USB OTG HS ULPI clock disabled
1: USB OTG HS ULPI clock enabled
Bit 29 OTGHSEN: USB OTG HS clock enable
0: USB OTG HS clock disabled
1: USB OTG HS clock enabled
Bit 28 ETHMACPTPEN: Ethernet PTP clock enable
0: Ethernet PTP clock disabled
1: Ethernet PTP clock enabled
Bit 27 ETHMACRXEN: Ethernet Reception clock enable
0: Ethernet Reception clock disabled
1: Ethernet Reception clock enabled
Bit 26 ETHMACTXEN: Ethernet Transmission clock enable
0: Ethernet Transmission clock disabled
1: Ethernet Transmission clock enabled
Bit 25 ETHMACEN: Ethernet MAC clock enable
0: Ethernet MAC clock disabled
1: Ethernet MAC clock enabled
242/1751 RM0090 Rev 19


Bit 20 CCMDATARAMEN: CCM data RAM clock enable
0: CCM data RAM clock disabled
1: CCM data RAM clock enabled
Bit 18 BKPSRAMEN: Backup SRAM interface clock enable
0: Backup SRAM interface clock disabled
1: Backup SRAM interface clock enabled
Bit 12 CRCEN: CRC clock enable
0: CRC clock disabled
1: CRC clock enabled
Bit 8 GPIOIEN: IO port I clock enable
0: IO port I clock disabled
1: IO port I clock enabled
Bit 7 GPIOHEN: IO port H clock enable
0: IO port H clock disabled
1: IO port H clock enabled
Bit 6 GPIOGEN: IO port G clock enable
0: IO port G clock disabled
1: IO port G clock enabled
Bit 5 GPIOFEN: IO port F clock enable
0: IO port F clock disabled
1: IO port F clock enabled
Bit 4 GPIOEEN: IO port E clock enable
0: IO port E clock disabled
1: IO port E clock enabled
Bit 3 GPIODEN: IO port D clock enable
0: IO port D clock disabled
1: IO port D clock enabled
RM0090 Rev 19 243/1751

266
Bit 2 GPIOCEN: IO port C clock enable

0: IO port C clock disabled
1: IO port C clock enabled
Bit 1 GPIOBEN: IO port B clock enable
0: IO port B clock disabled
1: IO port B clock enabled
Bit 0 GPIOAEN: IO port A clock enable
0: IO port A clock disabled
1: IO port A clock enabled

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserved EN EN EN EN Reserved EN
rw rw rw rw rw

Bit 7 OTGFSEN: USB OTG FS clock enable
0: USB OTG FS clock disabled
1: USB OTG FS clock enabled
Bit 6 RNGEN: Random number generator clock enable
0: Random number generator clock disabled
1: Random number generator clock enabled
Bit 5 HASHEN: Hash modules clock enable
0: Hash modules clock disabled
1: Hash modules clock enabled
244/1751 RM0090 Rev 19

Bit 4 CRYPEN: Cryptographic modules clock enable

0: cryptographic module clock disabled
1: cryptographic module clock enabled
Bit 0 DCMIEN: Camera interface enable
0: Camera interface clock disabled
1: Camera interface clock enabled

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FSMCEN
Reserved
rw

Bit 0 FSMCEN: Flexible static memory controller module clock enable
0: FSMC module clock disabled
1: FSMC module clock enabled
7.3.13 RCC APB1 peripheral clock enable register

(RCC_APB1ENR)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
USART USART
DAC PWR CAN2 CAN1 I2C3 I2C2 I2C1 UART5 UART4
Reser- Reser- 3 2 Reser-
Reserved EN EN EN EN EN EN EN EN EN
ved ved EN EN ved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
EN EN Reserved EN Reserved EN EN EN EN EN EN EN EN EN
RM0090 Rev 19 245/1751

266

Bit 29 DACEN: DAC interface clock enable
0: DAC interface clock disabled
1: DAC interface clock enable
Bit 28 PWREN: Power interface clock enable
0: Power interface clock disabled
1: Power interface clock enable
246/1751 RM0090 Rev 19


Bit 11 WWDGEN: Window watchdog clock enable
0: Window watchdog clock disabled
1: Window watchdog clock enabled
RM0090 Rev 19 247/1751

266


31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
TIM11 TIM10 TIM9
Reserved EN EN EN
rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
USART USART
SYSCF SPI1 SDIO ADC3 ADC2 ADC1 TIM8 TIM1
Reser- 6 1
G EN Reser- EN EN EN EN EN Reserved Reserved EN EN
ved ved EN EN

248/1751 RM0090 Rev 19

Bit 14 SYSCFGEN: System configuration controller clock enable

0: System configuration controller clock disabled
1: System configuration controller clock enabled
Bit 11 SDIOEN: SDIO clock enable
0: SDIO module clock disabled
1: SDIO module clock enabled
RM0090 Rev 19 249/1751

266
(RCC_AHB1LPENR)
Reset value: 0x7E67 91FF
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
OTGHS OTGH ETHPT ETHMA BKPSRA SRAM SRAM
ETHRX ETHTX DMA2 DMA1
Reser ULPILPE S P C M 2 1
LPEN LPEN Reserved LPEN LPEN Reserved
-ved N LPEN LPEN LPEN LPEN LPEN LPEN
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
GPIOG GPIO
FLITF CRC GPIOI GPIOH GPIOE GPIOD GPIOC GPIOB GPIOA
G F
LPEN Reserved LPEN Reserved LPEN LPEN LPEN LPEN LPEN LPEN LPEN
LPEN LPEN

Bit 30 OTGHSULPILPEN: USB OTG HS ULPI clock enable during Sleep mode
Set and cleared by software. This bit must be cleared when the OTG_HS is used in FS
mode.
0: USB OTG HS ULPI clock disabled during Sleep mode
1: USB OTG HS ULPI clock enabled during Sleep mode
Bit 29 OTGHSLPEN: USB OTG HS clock enable during Sleep mode
0: USB OTG HS clock disabled during Sleep mode
1: USB OTG HS clock enabled during Sleep mode
Bit 28 ETHMACPTPLPEN: Ethernet PTP clock enable during Sleep mode
0: Ethernet PTP clock disabled during Sleep mode
1: Ethernet PTP clock enabled during Sleep mode
Bit 27 ETHMACRXLPEN: Ethernet reception clock enable during Sleep mode
0: Ethernet reception clock disabled during Sleep mode
1: Ethernet reception clock enabled during Sleep mode
Bit 26 ETHMACTXLPEN: Ethernet transmission clock enable during Sleep mode
0: Ethernet transmission clock disabled during sleep mode
1: Ethernet transmission clock enabled during sleep mode
Bit 25 ETHMACLPEN: Ethernet MAC clock enable during Sleep mode
0: Ethernet MAC clock disabled during Sleep mode
1: Ethernet MAC clock enabled during Sleep mode
250/1751 RM0090 Rev 19


Bit 18 BKPSRAMLPEN: Backup SRAM interface clock enable during Sleep mode
0: Backup SRAM interface clock disabled during Sleep mode
1: Backup SRAM interface clock enabled during Sleep mode
Bit 17 SRAM2LPEN: SRAM 2 interface clock enable during Sleep mode
0: SRAM 2 interface clock disabled during Sleep mode
1: SRAM 2 interface clock enabled during Sleep mode
Bit 16 SRAM1LPEN: SRAM 1interface clock enable during Sleep mode
0: SRAM 1 interface clock disabled during Sleep mode
1: SRAM 1 interface clock enabled during Sleep mode
Bit 15 FLITFLPEN: Flash interface clock enable during Sleep mode
0: Flash interface clock disabled during Sleep mode
1: Flash interface clock enabled during Sleep mode
Bit 12 CRCLPEN: CRC clock enable during Sleep mode
0: CRC clock disabled during Sleep mode
1: CRC clock enabled during Sleep mode
Bit 8 GPIOILPEN: IO port I clock enable during Sleep mode
0: IO port I clock disabled during Sleep mode
1: IO port I clock enabled during Sleep mode
Bit 7 GPIOHLPEN: IO port H clock enable during Sleep mode
0: IO port H clock disabled during Sleep mode
1: IO port H clock enabled during Sleep mode
Bits 6 GPIOGLPEN: IO port G clock enable during Sleep mode
0: IO port G clock disabled during Sleep mode
1: IO port G clock enabled during Sleep mode
RM0090 Rev 19 251/1751

266
Bit 5 GPIOFLPEN: IO port F clock enable during Sleep mode

0: IO port F clock disabled during Sleep mode
1: IO port F clock enabled during Sleep mode
Bit 4 GPIOELPEN: IO port E clock enable during Sleep mode
0: IO port E clock disabled during Sleep mode
1: IO port E clock enabled during Sleep mode
Bit 3 GPIODLPEN: IO port D clock enable during Sleep mode
0: IO port D clock disabled during Sleep mode
1: IO port D clock enabled during Sleep mode
Bit 2 GPIOCLPEN: IO port C clock enable during Sleep mode
0: IO port C clock disabled during Sleep mode
1: IO port C clock enabled during Sleep mode
Bit 1 GPIOBLPEN: IO port B clock enable during Sleep mode
0: IO port B clock disabled during Sleep mode
1: IO port B clock enabled during Sleep mode
Bit 0 GPIOALPEN: IO port A clock enable during sleep mode
0: IO port A clock disabled during Sleep mode
1: IO port A clock enabled during Sleep mode
(RCC_AHB2LPENR)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserved LPEN LPEN LPEN LPEN Reserved LPEN
rw rw rw rw rw
252/1751 RM0090 Rev 19


Bit 7 OTGFSLPEN: USB OTG FS clock enable during Sleep mode
0: USB OTG FS clock disabled during Sleep mode
1: USB OTG FS clock enabled during Sleep mode
Bit 6 RNGLPEN: Random number generator clock enable during Sleep mode
0: Random number generator clock disabled during Sleep mode
1: Random number generator clock enabled during Sleep mode
Bit 5 HASHLPEN: Hash modules clock enable during Sleep mode
0: Hash modules clock disabled during Sleep mode
1: Hash modules clock enabled during Sleep mode
Bit 4 CRYPLPEN: Cryptography modules clock enable during Sleep mode
0: cryptography modules clock disabled during Sleep mode
1: cryptography modules clock enabled during Sleep mode
Bit 0 DCMILPEN: Camera interface enable during Sleep mode
0: Camera interface clock disabled during Sleep mode
1: Camera interface clock enabled during Sleep mode
(RCC_AHB3LPENR)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FSMC
Reserved LPEN
rw
Bits 31:1Reserved, must be kept at reset value.

FSMCLPEN: Flexible static memory controller module clock enable during Sleep mode
Bit 0
0: FSMC module clock disabled during Sleep mode
1: FSMC module clock enabled during Sleep mode
RM0090 Rev 19 253/1751

266
(RCC_APB1LPENR)
Reset value: 0x36FE C9FF
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
USART USART
DAC PWR RESER CAN2 CAN1 I2C3 I2C2 I2C1 UART5 UART4
Reser- 3 2 Reser-
Reserved LPEN LPEN VED LPEN LPEN LPEN LPEN LPEN LPEN LPEN
ved LPEN LPEN ved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
LPEN LPEN Reserved LPEN Reserved LPEN LPEN LPEN LPEN LPEN LPEN LPEN LPEN LPEN

Bit 29 DACLPEN: DAC interface clock enable during Sleep mode
0: DAC interface clock disabled during Sleep mode
1: DAC interface clock enabled during Sleep mode
Bit 28 PWRLPEN: Power interface clock enable during Sleep mode
0: Power interface clock disabled during Sleep mode
1: Power interface clock enabled during Sleep mode
0: CAN 2 clock disabled during sleep mode
1: CAN 2 clock enabled during sleep mode
0: CAN 1 clock disabled during Sleep mode
1: CAN 1 clock enabled during Sleep mode
254/1751 RM0090 Rev 19


Bit 11 WWDGLPEN: Window watchdog clock enable during Sleep mode
0: Window watchdog clock disabled during sleep mode
1: Window watchdog clock enabled during sleep mode
RM0090 Rev 19 255/1751

266

256/1751 RM0090 Rev 19

7.3.19 RCC APB2 peripheral clock enabled in low power mode

register (RCC_APB2LPENR)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
TIM11 TIM10 TIM9
Reserved LPEN LPEN LPEN
rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SYSC USART USART
SPI1 SDIO ADC3 ADC2 ADC1 TIM8 TIM1
Reser- FG Reser- 6 1
LPEN LPEN LPEN LPEN LPEN Reserved Reserved LPEN LPEN
ved LPEN ved LPEN LPEN

Bit 16 TIM9LPEN: TIM9 clock enable during sleep mode
Bit 14 SYSCFGLPEN: System configuration controller clock enable during Sleep mode
0: System configuration controller clock disabled during Sleep mode
1: System configuration controller clock enabled during Sleep mode
Bit 11 SDIOLPEN: SDIO clock enable during Sleep mode
0: SDIO module clock disabled during Sleep mode
1: SDIO module clock enabled during Sleep mode
RM0090 Rev 19 257/1751

266
Bit 10 ADC3LPEN: ADC 3 clock enable during Sleep mode

258/1751 RM0090 Rev 19

7.3.20 RCC Backup domain control register (RCC_BDCR)

Reset value: 0x0000 0000, reset by Backup domain reset.
The LSEON, LSEBYP, RTCSEL and RTCEN bits in the RCC Backup domain control
register (RCC_BDCR) are in the Backup domain. As a result, after Reset, these bits are
write-protected and the DBP bit in the PWR power control register (PWR_CR) for
STM32F405xx/07xx and STM32F415xx/17xx has to be set before these can be modified.
Refer to Section 7.1.1: System reset on page 213 for further information. These bits are only
reset after a Backup domain Reset (see Section 7.1.3: Backup domain reset). Any internal
or external Reset will not have any effect on these bits.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
BDRST
Reserved
rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
LSEBY LSERD
RTCEN RTCSEL[1:0] LSEON
Reserved Reserved P Y
rw rw rw rw r rw

Bit 16 BDRST: Backup domain software reset
0: Reset not activated
1: Resets the entire Backup domain
Note: The BKPSRAM is not affected by this reset, the only way of resetting the BKPSRAM is
through the Flash interface when a protection level change from level 1 to level 0 is
requested.
Bit 15 RTCEN: RTC clock enable
0: RTC clock disabled
1: RTC clock enabled
Bits 9:8 RTCSEL[1:0]: RTC clock source selection
Set by software to select the clock source for the RTC. Once the RTC clock source has been
selected, it cannot be changed anymore unless the Backup domain is reset. The BDRST bit
can be used to reset them.
00: No clock
01: LSE oscillator clock used as the RTC clock
10: LSI oscillator clock used as the RTC clock
11: HSE oscillator clock divided by a programmable prescaler (selection through the
RTCPRE[4:0] bits in the RCC clock configuration register (RCC_CFGR)) used as the RTC
clock
RM0090 Rev 19 259/1751

266
Bit 2 LSEBYP: External low-speed oscillator bypass

Set and cleared by software to bypass the oscillator. This bit can be written only when the
LSE clock is disabled.
0: LSE oscillator not bypassed
1: LSE oscillator bypassed
Bit 1 LSERDY: External low-speed oscillator ready
Set and cleared by hardware to indicate when the external 32 kHz oscillator is stable. After
the LSEON bit is cleared, LSERDY goes low after 6 external low-speed oscillator clock
cycles.
0: LSE clock not ready
1: LSE clock ready
Bit 0 LSEON: External low-speed oscillator enable
0: LSE clock OFF
1: LSE clock ON
7.3.21 RCC clock control & status register (RCC_CSR)

Reset value: 0x0E00 0000, reset by system reset, except reset flags by power reset only.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
LPWR WWDG IWDG SFT POR PIN BORRS
RMVF
RSTF RSTF RSTF RSTF RSTF RSTF TF Reserved
r r r r r r r rt_w
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
LSIRDY LSION
Reserved
r rw
Bit 31 LPWRRSTF: Low-power reset flag

Set by hardware when a Low-power management reset occurs.
0: No Low-power management reset occurred
1: Low-power management reset occurred
For further information on Low-power management reset, refer to Low-power management
reset.
Bit 30 WWDGRSTF: Window watchdog reset flag
Set by hardware when a window watchdog reset occurs.
0: No window watchdog reset occurred
1: Window watchdog reset occurred
Bit 29 IWDGRSTF: Independent watchdog reset flag
Set by hardware when an independent watchdog reset from VDD domain occurs.
0: No watchdog reset occurred
1: Watchdog reset occurred
260/1751 RM0090 Rev 19

Bit 28 SFTRSTF: Software reset flag

Set by hardware when a software reset occurs.
0: No software reset occurred
1: Software reset occurred
Bit 27 PORRSTF: POR/PDR reset flag
Set by hardware when a POR/PDR reset occurs.
0: No POR/PDR reset occurred
1: POR/PDR reset occurred
Bit 26 PINRSTF: PIN reset flag
Set by hardware when a reset from the NRST pin occurs.
0: No reset from NRST pin occurred
1: Reset from NRST pin occurred
Bit 25 BORRSTF: BOR reset flag
Cleared by software by writing the RMVF bit.
Set by hardware when a POR/PDR or BOR reset occurs.
0: No POR/PDR or BOR reset occurred
1: POR/PDR or BOR reset occurred
Bit 24 RMVF: Remove reset flag
Set by software to clear the reset flags.
0: No effect
1: Clear the reset flags
Bit 1 LSIRDY: Internal low-speed oscillator ready
Set and cleared by hardware to indicate when the internal RC 40 kHz oscillator is stable.
After the LSION bit is cleared, LSIRDY goes low after 3 LSI clock cycles.
0: LSI RC oscillator not ready
1: LSI RC oscillator ready
Bit 0 LSION: Internal low-speed oscillator enable
0: LSI RC oscillator OFF
1: LSI RC oscillator ON
RM0090 Rev 19 261/1751

266
7.3.22 RCC spread spectrum clock generation register (RCC_SSCGR)

The spread spectrum clock generation is available only for the main PLL.
The RCC_SSCGR register must be written either before the main PLL is enabled or after
the main PLL disabled.
Note: For full details about PLL spread spectrum clock generation (SSCG) characteristics, refer to
the “Electrical characteristics” section in your device datasheet.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
SPR
SSCG
EAD INCSTEP
EN Reserved
SEL
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
INCSTEP MODPER
Bit 31 SSCGEN: Spread spectrum modulation enable

0: Spread spectrum modulation DISABLE. (To write after clearing CR[24]=PLLON bit)
1: Spread spectrum modulation ENABLE. (To write before setting CR[24]=PLLON bit)
Bit 30 SPREADSEL: Spread Select
To write before to set CR[24]=PLLON bit.
0: Center spread
1: Down spread
Bits 27:13 INCSTEP: Incrementation step
Set and cleared by software. To write before setting CR[24]=PLLON bit.
Configuration input for modulation profile amplitude.
Bits 12:0 MODPER: Modulation period
Set and cleared by software. To write before setting CR[24]=PLLON bit.
Configuration input for modulation profile period.
262/1751 RM0090 Rev 19

7.3.23 RCC PLLI2S configuration register (RCC_PLLI2SCFGR)

This register is used to configure the PLLI2S clock outputs according to the formulas:
• f(VCO clock) = f(PLLI2S clock input) × (PLLI2SN / PLLM)
• f(PLL I2S clock output) = f(VCO clock) / PLLI2SR
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PLLI2S PLLI2S PLLI2S
Reserv R2 R1 R0 Reserved
ed
rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PLLI2SN PLLI2SN PLLI2SN PLLI2SN PLLI2SN PLLI2SN PLLI2SN PLLI2SN PLLI2SN
Reserv 8 7 6 5 4 3 2 1 0 Reserved
ed

Bits 30:28 PLLI2SR: PLLI2S division factor for I2S clocks
Set and cleared by software to control the I2S clock frequency. These bits should be written
only if the PLLI2S is disabled. The factor must be chosen in accordance with the prescaler
values inside the I2S peripherals, to reach 0.3% error when using standard crystals and 0%
error with audio crystals. For more information about I2S clock frequency and precision,
refer to Section 28.4.4: Clock generator in the I2S chapter.
Caution: The I2Ss requires a frequency lower than or equal to 192 MHz to work correctly.
I2S clock frequency = VCO frequency / PLLR with 2 ≤ PLLR ≤ 7
010: PLLR = 2
...
111: PLLR = 7
RM0090 Rev 19 263/1751

266

Bits 14:6 PLLI2SN: PLLI2S multiplication factor for VCO
These bits can be written only when PLLI2S is disabled. Only half-word and word accesses
VCO output frequency = VCO input frequency × PLLI2SN with 50 ≤ PLLI2SN ≤ 432
...
000110010: PLLI2SN = 50
...
001100011: PLLI2SN = 99
001100100: PLLI2SN = 100
001100101: PLLI2SN = 101
001100110: PLLI2SN = 102
...
110110000: PLLI2SN = 432
...
264/1751 RM0090 Rev 19

0x00
0x38
0x34
0x30
0x28
0x24
0x20
0x18
0x14
0x10
0x08
0x04
0x0C
0x3C
0x2C
0x1C
Addr.
offset
7.3.24
RM0090
RCC_
RCC_
RCC_
RCC_
RCC_
RCC_
RCC_
RCC_
RCC_
name
RCC_CR
Reserved
Reserved
Reserved
Reserved
RCC_CIR
PLLCFGR
AHB3ENR
AHB2ENR
AHB1ENR
Register
APB2RSTR
APB1RSTR
AHB3RSTR
AHB2RSTR
AHB1RSTR
RCC_CFGR
Reserved MCO2 1 31
Reserved Reserved
OTGHSULPIEN MCO2 0 30
OTGHSEN DACRST OTGHSRST MCO2PRE2 29
Reserved
Reserved
ETHMACPTPEN PWRRST MCO2PRE1 28
ETHMACRXEN Reserved Reserved MCO2PRE0 PLLQ 3 PLL I2SRDY 27
Reserved
ETHMACTXEN CAN2RST MCO1PRE2 PLLQ 2 PLL I2SON 26
RCC register map
ETHMACEN CAN1RST ETHMACRST MCO1PRE1 PLLQ 1 PLL RDY 25

Reserved MCO1PRE0 PLLQ 0 PLL ON
Reserved
24
Reserved Reserved
I2C3RST CSSC I2SSRC Reserved 23
DMA2EN I2C2RST DMA2RST Reserved MCO1 1 PLLSRC 22
DMA1EN I2C1RST DMA1RST PLLI2SRDYC MCO1 0 21
Reserved
CCMDATARAMEN UART5RST PLLRDYC RTCPRE 4 20

Reserved UART4RST HSERDYC RTCPRE 3 CSSON 19
Reserved
Reserved
Reserved
BKPSRAMEN TIM11RST UART3RST HSIRDYC RTCPRE 2 HSEBYP 18

TIM10RST UART2RST LSERDYC RTCPRE 1 PLLP 1 HSERDY 17
RM0090 Rev 19
TIM9RST Reserved LSIRDYC RTCPRE 0 PLLP 0 HSEON 16
Reserved
Table 34 gives the register map and reset values.
Reserved Reserved SPI3RST PPRE2 2 Reserved HSICAL 7
Reserved
Reserved
15
Reserved
Reserved
Reserved
Reserved
Reserved
SYSCFGRST SPI2RST PPRE2 1 PLLN 8 HSICAL 6 14
Reserved PLLI2SRDYIE PPRE2 0 PLLN 7 HSICAL 5 13
Reserved
CRCEN SPI1RST CRCRST PLLRDYIE PPRE1 2 PLLN 6 HSICAL 4 12
SDIORST WWDGRST HSERDYIE PPRE1 1 PLLN 5 HSICAL 3 11
Table 34. RCC register map and reset values
Reserved Reserved HSIRDYIE PPRE1 0 PLLN 4 HSICAL 2 10

Reserved Reserved
LSERDYIE PLLN 3 HSICAL 1 9
Reserved
GPIOIEN ADCRST TIM14RST GPIOIRST LSIRDYIE PLLN 2 HSICAL 0 8
OTGFSEN GPIOHEN TIM13RST OTGFSRS GPIOHRST CSSF HPRE 3 PLLN 1 HSITRIM 4 7
Reserved
RNGEN GPIOGEN TIM12RST RNGRST GPIOGRST Reserved HPRE 2 PLLN 0 HSITRIM 3 6
HASHEN GPIOFEN USART6RST TIM7RST HSAHRST GPIOFRST PLLI2SRDYF HPRE 1 PLLM 5 HSITRIM 2 5
CRYPEN GPIOEEN USART1RST TIM6RST CRYPRST GPIOERST PLLRDYF HPRE 0 PLLM 4 HSITRIM 1 4
GPIODEN TIM5RST GPIODRST HSERDYF SWS 1 PLLM 3 HSITRIM 0 3
Reserved
Reserved GPIOCEN TIM4RST Reserved GPIOCRST HSIRDYF SWS 0 PLLM 2 Reserved 2
GPIOBEN TIM8RST TIM3RST GPIOBRST LSERDYF SW 1 PLLM 1 HSIRDY 1
FSMCEN DCMIEN GPIOAEN TIM1RST TIM2RST FSMCRST DCMIRST GPIOARST LSIRDYF SW 0 PLLM 0 HSION 0
Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)
265/1751
266
0x84
0x74
0x70
0x80
0x78
0x68
0x64
0x60
0x58
0x54
0x50
0x48
0x44
0x40
0x7C
0x6C
0x5C
0x4C
Addr.
offset
266/1751
FGR
ENR
ENR
ENR
ENR
ENR
RCC_
RCC_
name
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
APB2ENR
APB1ENR
RCC_CSR
Register
RCC_BDCR
RCC_SSCGR
RCC_APB2LP
RCC_APB1LP
RCC_AHB3LP
RCC_AHB2LP
RCC_AHB1LP
RCC_PLLI2SC
Reserved SSCGEN LPWRRSTF Reserved 31
Reserved Reserved
SPREADSEL WWDGRSTF OTGHSULPILPEN 30
WDGRSTF DACLPEN OTGHSLPEN DACEN 29
Reserved
SFTRSTF PWRLPEN ETHMACPTPLPEN PWREN
PLLI2SRx
28
PORRSTF Reserved ETHMACRXLPEN Reserved 27
PADRSTF CAN2LPEN ETHMACTXLPEN CAN2EN 26
BORRSTF CAN1LPEN ETHMACLPEN CAN1EN 25
RMVF Reserved Reserved
Reserved
Reserved
24
Reserved
I2C3LPEN I2C3EN
Reserved
23
I2C2LPEN DMA2LPEN I2C2EN 22
I2C1LPEN DMA1LPEN I2C1EN 21
UART5LPEN UART5EN
Reserved
20
Reserved
UART4LPEN UART4EN
INCSTEP
19
Reserved
TIM11LPEN USART3LPEN BKPSRAMLPEN TIM11EN USART3EN 18
TIM10LPEN USART2LPEN SRAM2LPEN TIM10EN USART2EN 17
RM0090 Rev 19
BDRST TIM9LPEN Reserved SRAM1LPEN TIM9EN Reserved 16
RTCEN Reserved SPI3LPEN FLITFLPEN Reserved SPI3EN
Reserved
15
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
SYSCFGLPEN SPI2LPEN SYSCFGEN SPI2EN 14

Reserved
Reserved Reserved 13
Reserved Reserved
SPI1LPEN CRCLPEN SPI1EN 12
Reserved
SDIOLPEN WWDGLPEN SDIOEN WWDGEN
Reserved
11
ADC3LPEN Reserved ADC3EN 10
Reserved Reserved
RTCSEL 1 ADC2LPEN ADC2EN
PLLI2SNx
9
RTCSEL 0 ADC1LPEN TIM14LPEN GPIOILPEN ADC1EN TIM14EN 8
Table 34. RCC register map and reset values (continued)
Reset and clock control for STM32F405xx/07xx and STM32F415xx/17xx(RCC)
TIM13LPEN OTGFSLPEN GPIOHLPEN TIM13EN 7
Reserved Reserved
TIM12LPEN RNGLPEN GPIOGLPEN TIM12EN 6
USART6LPEN TIM7LPEN HASHLPEN GPIOFLPEN USART6EN TIM7EN
MODPER
5
USART1LPEN TIM6LPEN CRYPLPEN GPIOELPEN USART1EN TIM6EN
Reserved
4
TIM5LPEN GPIODLPEN TIM5EN 3
Reserved Reserved
LSEBYP TIM4LPEN Reserved GPIOCLPEN TIM4EN 2
Reserved
LSIRDY LSERDY TIM8LPEN TIM3LPEN GPIOBLPEN TIM8EN TIM3EN 1
LSION LSEON TIM1LPEN TIM2LPEN FSMCLPEN DCMILPEN GPIOALPEN TIM1EN TIM2EN 0
RM0090
RM0090 General-purpose I/Os (GPIO)
8 General-purpose I/Os (GPIO)
8.1 GPIO introduction

Each general-purpose I/O port has four 32-bit configuration registers (GPIOx_MODER,
GPIOx_OTYPER, GPIOx_OSPEEDR and GPIOx_PUPDR), two 32-bit data registers
(GPIOx_IDR and GPIOx_ODR), a 32-bit set/reset register (GPIOx_BSRR), a 32-bit locking
register (GPIOx_LCKR) and two 32-bit alternate function selection register (GPIOx_AFRH
and GPIOx_AFRL).
8.2 GPIO main features

• Up to 16 I/Os under control
• Output states: push-pull or open drain + pull-up/down
• Output data from output data register (GPIOx_ODR) or peripheral (alternate function
output)
• Speed selection for each I/O
• Input states: floating, pull-up/down, analog
• Input data to input data register (GPIOx_IDR) or peripheral (alternate function input)
• Bit set and reset register (GPIOx_BSRR) for bitwise write access to GPIOx_ODR
• Locking mechanism (GPIOx_LCKR) provided to freeze the I/O configuration
• Analog function
• Alternate function input/output selection registers (at most 16 AFs per I/O)
• Fast toggle capable of changing every two clock cycles
• Highly flexible pin multiplexing allows the use of I/O pins as GPIOs or as one of several
peripheral functions
8.3 GPIO functional description

Subject to the specific hardware characteristics of each I/O port listed in the datasheet, each
port bit of the general-purpose I/O (GPIO) ports can be individually configured by software in
several modes:
• Input floating
• Input pull-up
• Input-pull-down
• Analog
• Output open-drain with pull-up or pull-down capability
• Output push-pull with pull-up or pull-down capability
• Alternate function push-pull with pull-up or pull-down capability
• Alternate function open-drain with pull-up or pull-down capability
RM0090 Rev 19 267/1751

288
General-purpose I/Os (GPIO) RM0090
Each I/O port bit is freely programmable, however the I/O port registers have to be
accessed as 32-bit words, half-words or bytes. The purpose of the GPIOx_BSRR register is
to allow atomic read/modify accesses to any of the GPIO registers. In this way, there is no
risk of an IRQ occurring between the read and the modify access.
Figure 25 shows the basic structure of a 5 V tolerant I/O port bit. Table 39 gives the possible
port bit configurations.
Figure 25. Basic structure of a five-volt tolerant I/O port bit
To on-chip Analog
peripheral
Alternate function input
on/off
Input data register
Read
VDD VDD_FT (1)
TTL Schmitt
Bit set/reset registers
Protection
trigger on/off
Pull diode
up
Input driver I/O pin
Write
Output data register
Output driver VDD on/off Pull Protection

down
diode
P-MOS
Output VSS
VSS
control
Read/write N-MOS
From on-chip VSS

Push-pull,
peripheral Alternate function output open-drain or
disabled
Analog
ai15939b
1. VDD_FT is a potential specific to five-volt tolerant I/Os and different from VDD.
Table 35. Port bit configuration table(1)

MODER(i) OSPEEDR(i) PUPDR(i)
OTYPER(i) I/O configuration
[1:0] [B:A] [1:0]
0 0 0 GP output PP
0 0 1 GP output PP + PU
0 1 0 GP output PP + PD
0 SPEED 1 1 Reserved
01
1 [B:A] 0 0 GP output OD
1 0 1 GP output OD + PU
1 1 0 GP output OD + PD
1 1 1 Reserved (GP output OD)
268/1751 RM0090 Rev 19

Table 35. Port bit configuration table(1) (continued)

MODER(i) OSPEEDR(i) PUPDR(i)
OTYPER(i) I/O configuration
[1:0] [B:A] [1:0]
0 0 0 AF PP
0 0 1 AF PP + PU
0 1 0 AF PP + PD
0 SPEED 1 1 Reserved
10
1 [B:A] 0 0 AF OD
1 0 1 AF OD + PU
1 1 0 AF OD + PD
1 1 1 Reserved
x x x 0 0 Input Floating
x x x 0 1 Input PU
00
x x x 1 0 Input PD
x x x 1 1 Reserved (input floating)
x x x 0 0 Input/output Analog
x x x 0 1
11
x x x 1 0 Reserved
x x x 1 1
1. GP = general-purpose, PP = push-pull, PU = pull-up, PD = pull-down, OD = open-drain, AF = alternate
function.
8.3.1 General-purpose I/O (GPIO)

During and just after reset, the alternate functions are not active and the I/O ports are
configured in input floating mode.
The debug pins are in AF pull-up/pull-down after reset:
• PA15: JTDI in pull-up
• PA14: JTCK/SWCLK in pull-down
• PA13: JTMS/SWDAT in pull-up
• PB4: NJTRST in pull-up
• PB3: JTDO in floating state
When the pin is configured as output, the value written to the output data register
(GPIOx_ODR) is output on the I/O pin. It is possible to use the output driver in push-pull
mode or open-drain mode (only the N-MOS is activated when 0 is output).
The input data register (GPIOx_IDR) captures the data present on the I/O pin at every AHB1
clock cycle.
All GPIO pins have weak internal pull-up and pull-down resistors, which can be activated or
not depending on the value in the GPIOx_PUPDR register.
RM0090 Rev 19 269/1751

288
8.3.2 I/O pin multiplexer and mapping

The microcontroller I/O pins are connected to onboard peripherals/modules through a
multiplexer that allows only one peripheral’s alternate function (AF) connected to an I/O pin
at a time. In this way, there can be no conflict between peripherals sharing the same I/O pin.
Each I/O pin has a multiplexer with sixteen alternate function inputs (AF0 to AF15) that can
be configured through the GPIOx_AFRL (for pin 0 to 7) and GPIOx_AFRH (for pin 8 to 15)
registers:
• After reset all I/Os are connected to the system’s alternate function 0 (AF0)
• The peripherals’ alternate functions are mapped from AF1 to AF13
• Cortex®-M4 with FPU EVENTOUT is mapped on AF15
This structure is shown in Figure 26 below.
In addition to this flexible I/O multiplexing architecture, each peripheral has alternate
functions mapped onto different I/O pins to optimize the number of peripherals available in
smaller packages.
To use an I/O in a given configuration, proceed as follows:
• System function
Connect the I/O to AF0 and configure it depending on the function used:
– JTAG/SWD, after each device reset these pins are assigned as dedicated pins
immediately usable by the debugger host (not controlled by the GPIO controller)
– RTC_REFIN: this pin should be configured in Input floating mode
– MCO1 and MCO2: these pins have to be configured in alternate function mode.
Note: The user can disable some or all of the JTAG/SWD pins and so release the associated pins
for GPIO usage (released pins highlighted in gray in the table).
For more details refer to Section 7.2.10: Clock-out capability and Section 6.2.10: Clock-out
capability.
270/1751 RM0090 Rev 19

Table 36. Flexible SWJ-DP pin assignment

SWJ I/O pin assigned
Available debug ports PA13 / PA14 /

PA15 / PB3 / PB4/
JTMS/ JTCK/
JTDI JTDO NJTRST
SWDIO SWCLK
Full SWJ (JTAG-DP + SW-DP) - Reset state X X X X X

Full SWJ (JTAG-DP + SW-DP) but without
X X X X
NJTRST
JTAG-DP Disabled and SW-DP Enabled X X
JTAG-DP Disabled and SW-DP Disabled Released
• GPIO
Configure the desired I/O as output or input in the GPIOx_MODER register.
• Peripheral alternate function
For the ADC and DAC, configure the desired I/O as analog in the GPIOx_MODER
register.
For other peripherals:
– Configure the desired I/O as an alternate function in the GPIOx_MODER register
– Select the type, pull-up/pull-down and output speed via the GPIOx_OTYPER,
GPIOx_PUPDR and GPIOx_OSPEEDR registers, respectively
– Connect the I/O to the desired AFx in the GPIOx_AFRL or GPIOx_AFRH register
• EVENTOUT
Configure the I/O pin used to output the Cortex®-M4 with FPU EVENTOUT signal by
connecting it to AF15
Note: EVENTOUT is not mapped onto the following I/O pins: PC13, PC14, PC15, PH0, PH1 and
PI8.
Refer to the “Alternate function mapping” table in the datasheets for the detailed mapping of
the system and peripherals’ alternate function I/O pins.
RM0090 Rev 19 271/1751

288
Figure 26. Selecting an alternate function on STM32F405xx/07xx and

STM32F415xx/17xx
For pins 0 to 7, the GPIOx_AFRL[31:0] register selects the dedicated alternate function
AF0 (system)
AF1 (TIM1/TIM2)
AF2 (TIM3..5)
AF3 (TIM8..11)
AF4 (I2C1..3)
AF5 (SPI1/SPI2)
AF6 (SPI3) Pin x (x = 0..7)
AF7 (USART1..3)
1
AF8 (USART4..6)
AF9 (CAN1/CAN2, TIM12..14)
AF10 (OTG_FS, OTG_HS)
AF11 (ETH)
AF12 (FSMC, SDIO, OTG_HS(1))
AF13 (DCMI)
AF14
AF15 (EVENTOUT)
AFRL[31:0]
For pins 8 to 15, the GPIOx_AFRH[31:0] register selects the dedicated alternate function
AF0 (system)
AF1 (TIM1/TIM2)
AF2 (TIM3..5)
AF3 (TIM8..11)
AF4 (I2C1..3)
AF5 (SPI1/SPI2)
AF6 (SPI3) Pin x (x = 8..15)
AF7 (USART1..3)
AF8 (USART4..6) 1
AF11 (ETH)
AF12 (FSMC, SDIO, OTG_HS(1))
AF13 (DCMI)
AF14
AF15 (EVENTOUT)
AFRH[31:0] ai17538
1. Configured in FS.
272/1751 RM0090 Rev 19

Figure 27. Selecting an alternate function on STM32F42xxx and STM32F43xxx
For pins 0 to 7, the GPIOx_AFRL[31:0] register selects the dedicated alternate function
AF0 (system)
AF1 (TIM1/TIM2)
AF2 (TIM3..5)
AF3 (TIM8..11)
AF4 (I2C1..3)
AF5 (SPI1/2/3/4/5/6)
AF6 (SPI2/3/SAI1) Pin x (x = 0..7)
AF7 (USART1..3)
1
AF8 (USART4..8)
AF9 (CAN1/CAN2, LTDC, TIM12..14)
AF11 (ETH)
AF12 (FMC, SDIO, OTG_HS(1))
AF13 (DCMI)
AF14 (LTDC)
AF15 (EVENTOUT)
AFRL[31:0]
For pins 8 to 15, the GPIOx_AFRH[31:0] register selects the dedicated alternate function
AF0 (system)
AF1 (TIM1/TIM2)
AF2 (TIM3..5)
AF3 (TIM8..11)
AF4 (I2C1..3)
AF5 (SPI1/2/4/5/6)
AF6 (SPI2/3/SAI1) Pin x (x = 8..15)
AF7 (USART1..3)
AF8 (USART4..8) 1
AF11 (ETH)
AF12 (FMC, SDIO, OTG_HS(1))
AF13 (DCMI)
AF14 (LTDC)
AF15 (EVENTOUT)
AFRH[31:0]
MS60208V1
1. Configured in FS.
RM0090 Rev 19 273/1751

288
8.3.3 I/O port control registers

Each of the GPIOs has four 32-bit memory-mapped control registers (GPIOx_MODER,
GPIOx_OTYPER, GPIOx_OSPEEDR, GPIOx_PUPDR) to configure up to 16 I/Os.
The GPIOx_MODER register is used to select the I/O direction (input, output, AF, analog).
The GPIOx_OTYPER and GPIOx_OSPEEDR registers are used to select the output type
(push-pull or open-drain) and speed (the I/O speed pins are directly connected to the
corresponding GPIOx_OSPEEDR register bits whatever the I/O direction). The
GPIOx_PUPDR register is used to select the pull-up/pull-down whatever the I/O direction.
8.3.4 I/O port data registers

Each GPIO has two 16-bit memory-mapped data registers: input and output data registers
(GPIOx_IDR and GPIOx_ODR). GPIOx_ODR stores the data to be output, it is read/write
accessible. The data input through the I/O are stored into the input data register
(GPIOx_IDR), a read-only register.
See Section 8.4.5: GPIO port input data register (GPIOx_IDR) (x = A..I/J/K) and
Section 8.4.6: GPIO port output data register (GPIOx_ODR) (x = A..I/J/K) for the register
descriptions.
8.3.5 I/O data bitwise handling

The bit set reset register (GPIOx_BSRR) is a 32-bit register which allows the application to
set and reset each individual bit in the output data register (GPIOx_ODR). The bit set reset
register has twice the size of GPIOx_ODR.
To each bit in GPIOx_ODR, correspond two control bits in GPIOx_BSRR: BSRR(i) and
BSRR(i+SIZE). When written to 1, bit BSRR(i) sets the corresponding ODR(i) bit. When
written to 1, bit BSRR(i+SIZE) resets the ODR(i) corresponding bit.
Writing any bit to 0 in GPIOx_BSRR does not have any effect on the corresponding bit in
GPIOx_ODR. If there is an attempt to both set and reset a bit in GPIOx_BSRR, the set
action takes priority.
Using the GPIOx_BSRR register to change the values of individual bits in GPIOx_ODR is a
“one-shot” effect that does not lock the GPIOx_ODR bits. The GPIOx_ODR bits can always
be accessed directly. The GPIOx_BSRR register provides a way of performing atomic
bitwise handling.
There is no need for the software to disable interrupts when programming the GPIOx_ODR
at bit level: it is possible to modify one or more bits in a single atomic AHB1 write access.
8.3.6 GPIO locking mechanism

It is possible to freeze the GPIO control registers by applying a specific write sequence to
the GPIOx_LCKR register. The frozen registers are GPIOx_MODER, GPIOx_OTYPER,
GPIOx_OSPEEDR, GPIOx_PUPDR, GPIOx_AFRL and GPIOx_AFRH.
To write the GPIOx_LCKR register, a specific write / read sequence has to be applied. When
the right LOCK sequence is applied to bit 16 in this register, the value of LCKR[15:0] is used
to lock the configuration of the I/Os (during the write sequence the LCKR[15:0] value must
be the same). When the LOCK sequence has been applied to a port bit, the value of the port
bit can no longer be modified until the next MCU or peripheral reset. Each GPIOx_LCKR bit
274/1751 RM0090 Rev 19

freezes the corresponding bit in the control registers (GPIOx_MODER, GPIOx_OTYPER,

GPIOx_OSPEEDR, GPIOx_PUPDR, GPIOx_AFRL and GPIOx_AFRH).
The LOCK sequence (refer to Section 8.4.8: GPIO port configuration lock register
(GPIOx_LCKR) (x = A..I/J/K)) can only be performed using a word (32-bit long) access to
the GPIOx_LCKR register due to the fact that GPIOx_LCKR bit 16 has to be set at the same
time as the [15:0] bits.
For more details refer to LCKR register description in Section 8.4.8: GPIO port configuration
lock register (GPIOx_LCKR) (x = A..I/J/K).
8.3.7 I/O alternate function input/output

Two registers are provided to select one out of the sixteen alternate function inputs/outputs
available for each I/O. With these registers, you can connect an alternate function to some
other pin as required by your application.
This means that a number of possible peripheral functions are multiplexed on each GPIO
using the GPIOx_AFRL and GPIOx_AFRH alternate function registers. The application can
thus select any one of the possible functions for each I/O. The AF selection signal being
common to the alternate function input and alternate function output, a single channel is
selected for the alternate function input/output of one I/O.
To know which functions are multiplexed on each GPIO pin, refer to the datasheets.
Note: The application is allowed to select one of the possible peripheral functions for each I/O at a
time.
8.3.8 External interrupt/wakeup lines

All ports have external interrupt capability. To use external interrupt lines, the port must be
configured in input mode, refer to Section 12.2: External interrupt/event controller (EXTI)
and Section 12.2.3: Wakeup event management.
8.3.9 Input configuration

When the I/O port is programmed as Input:
• the output buffer is disabled
• the Schmitt trigger input is activated
• the pull-up and pull-down resistors are activated depending on the value in the
GPIOx_PUPDR register
• The data present on the I/O pin are sampled into the input data register every AHB1
clock cycle
• A read access to the input data register provides the I/O State
Figure 28 shows the input configuration of the I/O port bit.
RM0090 Rev 19 275/1751

288
Figure 28. Input floating/pull up/pull down configurations
Input data register

on
Read
VDD VDD
TTL Schmitt on/off
trigger protection
pull diode
Write up
Output data register input driver I/O pin
on/off
output driver
pull protection
down diode
VSS VSS
Read/write
ai15940b
8.3.10 Output configuration

When the I/O port is programmed as output:
• The output buffer is enabled:
– Open drain mode: A “0” in the Output register activates the N-MOS whereas a “1”
in the Output register leaves the port in Hi-Z (the P-MOS is never activated)
– Push-pull mode: A “0” in the Output register activates the N-MOS whereas a “1” in
the Output register activates the P-MOS
• The Schmitt trigger input is activated
• The weak pull-up and pull-down resistors are activated or not depending on the value
in the GPIOx_PUPDR register
clock cycle
• A read access to the input data register gets the I/O state
• A read access to the output data register gets the last written value
Figure 29 shows the output configuration of the I/O port bit.
276/1751 RM0090 Rev 19

Figure 29. Output configuration
Input data register

on
Read
TTL Schmitt VDD VDD
Bit set/reset registers trigger

on/off
Write protection
Input driver diode

pull
up
I/O pin
Output driver VDD on/off
P-MOS pull protection

Output down diode
Read/write control VSS
N-MOS VSS
Push-pull or
VSS Open-drain
ai15941b
8.3.11 Alternate function configuration

When the I/O port is programmed as alternate function:
• The output buffer can be configured as open-drain or push-pull
• The output buffer is driven by the signal coming from the peripheral (transmitter enable
and data)
• The Schmitt trigger input is activated
• The weak pull-up and pull-down resistors are activated or not depending on the value
in the GPIOx_PUPDR register
clock cycle
• A read access to the input data register gets the I/O state
Figure 30 shows the Alternate function configuration of the I/O port bit.
Figure 30. Alternate function configuration
To on-chip Alternate function input

peripheral
Input data register
on
Read
VDD VDD
TTL Schmitt on/off
trigger protection
Pull diode
Input driver up
Write
I/O pin
Output driver VDD on/off
Pull protection
P-MOS down diode
Output
control VSS VSS
N-MOS
Read/write
VSS push-pull or
open-drain
From on-chip
peripheral Alternate function output
ai15942b
RM0090 Rev 19 277/1751

288
8.3.12 Analog configuration

When the I/O port is programmed as analog configuration:
• The output buffer is disabled
• The Schmitt trigger input is deactivated, providing zero consumption for every analog
value of the I/O pin. The output of the Schmitt trigger is forced to a constant value (0).
• The weak pull-up and pull-down resistors are disabled
• Read access to the input data register gets the value “0”
Note: In the analog configuration, the I/O pins cannot be 5 Volt tolerant.
Figure 31 shows the high-impedance, analog-input configuration of the I/O port bit.
Figure 31. High impedance-analog configuration
Analog
To on-chip
peripheral
Input data register
Read off
0
VDD
TTL Schmitt
trigger protection
Write diode
Input driver
I/O pin
protection
diode
Read/write VSS
From on-chip Analog

peripheral
ai15943
8.3.13 Using the OSC32_IN/OSC32_OUT pins as GPIO PC14/PC15

port pins
The LSE oscillator pins OSC32_IN and OSC32_OUT can be used as general-purpose
PC14 and PC15 I/Os, respectively, when the LSE oscillator is off. The PC14 and PC15 I/Os
are only configured as LSE oscillator pins OSC32_IN and OSC32_OUT when the LSE
oscillator is ON. This is done by setting the LSEON bit in the RCC_BDCR register. The LSE
has priority over the GPIO function.
Note: The PC14/PC15 GPIO functionality is lost when the 1.2 V domain is powered off (by the
device entering the standby mode) or when the backup domain is supplied by VBAT (VDD no
more supplied). In this case the I/Os are set in analog input mode.
8.3.14 Using the OSC_IN/OSC_OUT pins as GPIO PH0/PH1 port pins

The HSE oscillator pins OSC_IN/OSC_OUT can be used as general-purpose PH0/PH1
I/Os, respectively, when the HSE oscillator is OFF. (after reset, the HSE oscillator is off). The
PH0/PH1 I/Os are only configured as OSC_IN/OSC_OUT HSE oscillator pins when the
HSE oscillator is ON. This is done by setting the HSEON bit in the RCC_CR register. The
HSE has priority over the GPIO function.
278/1751 RM0090 Rev 19

8.3.15 Selection of RTC_AF1 and RTC_AF2 alternate functions

The STM32F4xx feature two GPIO pins RTC_AF1 and RTC_AF2 that can be used for the
detection of a tamper or time stamp event, or RTC_ALARM, or RTC_CALIB RTC outputs.
• The RTC_AF1 (PC13) can be used for the following purposes:
RTC_ALARM output: this output can be RTC Alarm A, RTC Alarm B or RTC Wakeup
depending on the OSEL[1:0] bits in the RTC_CR register
• RTC_CALIB output: this feature is enabled by setting the COE[23] in the RTC_CR
register
• RTC_TAMP1: tamper event detection
• RTC_TS: time stamp event detection
The RTC_AF2 (PI8) can be used for the following purposes:
• RTC_TS: time stamp event detection
The selection of the corresponding pin is performed through the RTC_TAFCR register as
follows:
• TAMP1INSEL is used to select which pin is used as the RTC_TAMP1 tamper input
• TSINSEL is used to select which pin is used as the RTC_TS time stamp input
• ALARMOUTTYPE is used to select whether the RTC_ALARM is output in push-pull or
open-drain mode
The output mechanism follows the priority order listed in Table 37 and Table 38.
Table 37. RTC_AF1 pin(1)

TSINSEL
Pin Time TAMP1INSEL ALARMOUTTYPE
RTC_ALARM RTC_CALIB Tamper TIMESTAMP
configuration stamp TAMPER1 RTC_ALARM
enabled enabled enabled pin
and function enabled pin selection configuration
selection
Alarm out Don’t Don’t

1 Don’t care Don’t care Don’t care 0
output OD care care
Alarm out Don’t Don’t
1 Don’t care Don’t care Don’t care 1
output PP care care
Calibration Don’t Don’t
0 1 Don’t care Don’t care Don’t care
out output PP care care
TAMPER1
0 0 1 0 0 Don’t care Don’t care
input floating
TIMESTAMP
and
0 0 1 1 0 0 Don’t care
TAMPER1
input floating
TIMESTAMP
0 0 0 1 Don’t care 0 Don’t care
input floating
Standard
0 0 0 0 Don’t care Don’t care Don’t care
GPIO
1. OD: open drain; PP: push-pull.
RM0090 Rev 19 279/1751

288
Table 38. RTC_AF2 pin

TSINSEL
Time TAMP1INSEL ALARMOUTTYPE
Tamper TIMESTAMP
Pin configuration and function stamp TAMPER1 RTC_ALARM
enabled pin
enabled pin selection configuration
selection
TAMPER1 input floating 1 0 1 Don’t care Don’t care

TIMESTAMP and TAMPER1 input
1 1 1 1 Don’t care
floating
TIMESTAMP input floating 0 1 Don’t care 1 Don’t care
Standard GPIO 0 0 Don’t care Don’t care Don’t care
280/1751 RM0090 Rev 19

8.4 GPIO registers

This section gives a detailed description of the GPIO registers.
For a summary of register bits, register address offsets and reset values, refer to Table 39.
The GPIO registers can be accessed by byte (8 bits), half-words (16 bits) or words (32 bits).
8.4.1 GPIO port mode register (GPIOx_MODER) (x = A..I/J/K)

Reset values:
• 0xA800 0000 for port A
• 0x0000 0280 for port B
• 0x0000 0000 for other ports
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
MODER15[1:0] MODER14[1:0] MODER13[1:0] MODER12[1:0] MODER11[1:0] MODER10[1:0] MODER9[1:0] MODER8[1:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MODER7[1:0] MODER6[1:0] MODER5[1:0] MODER4[1:0] MODER3[1:0] MODER2[1:0] MODER1[1:0] MODER0[1:0]
Bits 2y:2y+1 MODERy[1:0]: Port x configuration bits (y = 0..15)

These bits are written by software to configure the I/O direction mode.
00: Input (reset state)
01: General purpose output mode
10: Alternate function mode
11: Analog mode
8.4.2 GPIO port output type register (GPIOx_OTYPER)

(x = A..I/J/K)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OT15 OT14 OT13 OT12 OT11 OT10 OT9 OT8 OT7 OT6 OT5 OT4 OT3 OT2 OT1 OT0

Bits 15:0 OTy: Port x configuration bits (y = 0..15)
These bits are written by software to configure the output type of the I/O port.
0: Output push-pull (reset state)
1: Output open-drain
RM0090 Rev 19 281/1751

288
8.4.3 GPIO port output speed register (GPIOx_OSPEEDR)

(x = A..I/J/K)
Reset values:
• 0x0C00 0000 for port A
• 0x0000 00C0 for port B
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
OSPEEDR15 OSPEEDR14 OSPEEDR13 OSPEEDR12 OSPEEDR11 OSPEEDR10 OSPEEDR9 OSPEEDR8
[1:0] [1:0] [1:0] [1:0] [1:0] [1:0] [1:0] [1:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OSPEEDR1 OSPEEDR0
OSPEEDR7[1:0] OSPEEDR6[1:0] OSPEEDR5[1:0] OSPEEDR4[1:0] OSPEEDR3[1:0] OSPEEDR2[1:0]
[1:0] 1:0]
Bits 2y:2y+1 OSPEEDRy[1:0]: Port x configuration bits (y = 0..15)

These bits are written by software to configure the I/O output speed.
00: Low speed
01: Medium speed
10: High speed
11: Very high speed
Note: Refer to the product datasheets for the values of OSPEEDRy bits versus VDD
range and external load.
8.4.4 GPIO port pull-up/pull-down register (GPIOx_PUPDR)

(x = A..I/J/K)
Reset values:
• 0x6400 0000 for port A
• 0x0000 0100 for port B
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PUPDR15[1:0] PUPDR14[1:0] PUPDR13[1:0] PUPDR12[1:0] PUPDR11[1:0] PUPDR10[1:0] PUPDR9[1:0] PUPDR8[1:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PUPDR7[1:0] PUPDR6[1:0] PUPDR5[1:0] PUPDR4[1:0] PUPDR3[1:0] PUPDR2[1:0] PUPDR1[1:0] PUPDR0[1:0]
282/1751 RM0090 Rev 19

Bits 2y:2y+1 PUPDRy[1:0]: Port x configuration bits (y = 0..15)

These bits are written by software to configure the I/O pull-up or pull-down
00: No pull-up, pull-down
01: Pull-up
10: Pull-down
11: Reserved
8.4.5 GPIO port input data register (GPIOx_IDR) (x = A..I/J/K)

Reset value: 0x0000 XXXX (where X means undefined)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
IDR15 IDR14 IDR13 IDR12 IDR11 IDR10 IDR9 IDR8 IDR7 IDR6 IDR5 IDR4 IDR3 IDR2 IDR1 IDR0
r r r r r r r r r r r r r r r r

Bits 15:0 IDRy: Port input data (y = 0..15)
These bits are read-only and can be accessed in word mode only. They contain the input
value of the corresponding I/O port.
8.4.6 GPIO port output data register (GPIOx_ODR) (x = A..I/J/K)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ODR15 ODR14 ODR13 ODR12 ODR11 ODR10 ODR9 ODR8 ODR7 ODR6 ODR5 ODR4 ODR3 ODR2 ODR1 ODR0

Bits 15:0 ODRy: Port output data (y = 0..15)
These bits can be read and written by software.
Note: For atomic bit set/reset, the ODR bits can be individually set and reset by writing to the
GPIOx_BSRR register (x = A..I/J/K).
RM0090 Rev 19 283/1751

288
8.4.7 GPIO port bit set/reset register (GPIOx_BSRR) (x = A..I/J/K)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
BR15 BR14 BR13 BR12 BR11 BR10 BR9 BR8 BR7 BR6 BR5 BR4 BR3 BR2 BR1 BR0
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
BS15 BS14 BS13 BS12 BS11 BS10 BS9 BS8 BS7 BS6 BS5 BS4 BS3 BS2 BS1 BS0
Bits 31:16 BRy: Port x reset bit y (y = 0..15)

These bits are write-only and can be accessed in word, half-word or byte mode. A read to
these bits returns the value 0x0000.
0: No action on the corresponding ODRx bit
1: Resets the corresponding ODRx bit
Note: If both BSx and BRx are set, BSx has priority.
Bits 15:0 BSy: Port x set bit y (y= 0..15)
These bits are write-only and can be accessed in word, half-word or byte mode. A read to
these bits returns the value 0x0000.
0: No action on the corresponding ODRx bit
1: Sets the corresponding ODRx bit
8.4.8 GPIO port configuration lock register (GPIOx_LCKR)

(x = A..I/J/K)
This register is used to lock the configuration of the port bits when a correct write sequence
is applied to bit 16 (LCKK). The value of bits [15:0] is used to lock the configuration of the
GPIO. During the write sequence, the value of LCKR[15:0] must not change. When the
LOCK sequence has been applied on a port bit, the value of this port bit can no longer be
modified until the next MCU or peripheral reset.
Note: A specific write sequence is used to write to the GPIOx_LCKR register. Only word access
(32-bit long) is allowed during this write sequence.
Each lock bit freezes a specific configuration register (control and alternate function
registers).
Access: 32-bit word only, read/write register
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
LCKK
Reserved
rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
LCK15 LCK14 LCK13 LCK12 LCK11 LCK10 LCK9 LCK8 LCK7 LCK6 LCK5 LCK4 LCK3 LCK2 LCK1 LCK0
284/1751 RM0090 Rev 19


Bit 16 LCKK[16]: Lock key
This bit can be read any time. It can only be modified using the lock key write sequence.
0: Port configuration lock key not active
1: Port configuration lock key active. The GPIOx_LCKR register is locked until an MCU reset
or a peripheral reset occurs.
LOCK key write sequence:

WR LCKR[16] = ‘1’ + LCKR[15:0]
WR LCKR[16] = ‘0’ + LCKR[15:0]
WR LCKR[16] = ‘1’ + LCKR[15:0]
RD LCKR
RD LCKR[16] = ‘1’ (this read operation is optional but it confirms that the lock is active)
Note: During the LOCK key write sequence, the value of LCK[15:0] must not change.
Any error in the lock sequence aborts the lock.
After the first lock sequence on any bit of the port, any read access on the LCKK bit
returns ‘1’ until the next CPU reset.
Bits 15:0 LCKy: Port x lock bit y (y= 0..15)
These bits are read/write but can only be written when the LCKK bit is ‘0.
0: Port configuration not locked
1: Port configuration locked
8.4.9 GPIO alternate function low register (GPIOx_AFRL) (x = A..I/J/K)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
AFRL7[3:0] AFRL6[3:0] AFRL5[3:0] AFRL4[3:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
AFRL3[3:0] AFRL2[3:0] AFRL1[3:0] AFRL0[3:0]
Bits 31:0 AFRLy: Alternate function selection for port x bit y (y = 0..7)
These bits are written by software to configure alternate function I/Os
AFRLy selection:
0000: AF0 1000: AF8
0001: AF1 1001: AF9
0010: AF2 1010: AF10
0011: AF3 1011: AF11
0100: AF4 1100: AF12
0101: AF5 1101: AF13
0110: AF6 1110: AF14
0111: AF7 1111: AF15
RM0090 Rev 19 285/1751

288
8.4.10 GPIO alternate function high register (GPIOx_AFRH)

(x = A..I/J)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
AFRH15[3:0] AFRH14[3:0] AFRH13[3:0] AFRH12[3:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
AFRH11[3:0] AFRH10[3:0] AFRH9[3:0] AFRH8[3:0]
Bits 31:0 AFRHy: Alternate function selection for port x bit y (y = 8..15)
These bits are written by software to configure alternate function I/Os
AFRHy selection:
0000: AF0 1000: AF8
0001: AF1 1001: AF9
0010: AF2 1010: AF10
0011: AF3 1011: AF11
0100: AF4 1100: AF12
0101: AF5 1101: AF13
0110: AF6 1110: AF14
0111: AF7 1111: AF15
286/1751 RM0090 Rev 19

0x08
0x08
0x04
0x00
0x00
0x00
0x0C
0x0C
Offset
8.4.11
RM0090
B)
A..I/J/K)
C..I/J/K)
GPIOx_
GPIOx_
MODER
MODER
GPIOB_
GPIOB_
GPIOA_
OTYPER
(where x =
(where x =
(where x =
OSPEEDR
OSPEEDR
Register
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
A..I/J/K except
GPIOB_PUPDR
GPIOA_PUPDR
GPIOx_MODER
0
0
0
0
0
0
1
PUPDR15[1:0] PUPDR15[1:0] OSPEEDR15[1:0] OSPEEDR15[1:0] MODER15[1:0] MODER15[1:0] MODER15[1:0] 31
0
1
0
0
0
0
0
30
0
1
0
0
0
0
1
0
0
0
0
0
0
0
28
0
0
0
0
0
0
1
0
1
0
0
0
0
0
26
GPIO register map
0
0
0
0
0
0
0 25
PUPDR12[1:0] PUPDR12[1:0] OSPEEDR12[1:0] OSPEEDR12[1:0] MODER12[1:0] MODER12[1:0] MODER12[1:0]
0
0
0
0
0
0
0
24
0
0
0
0
0
0
0
PUPDR11[1:0] PUPDR11[1:0] OSPEEDR11[1:0] OSPEEDR11[1:0] 23
Reserved
MODER11[1:0] MODER11[1:0] MODER11[1:0]
0
0
0
0
0
0
0
22
0
0
0
0
0
0
0
0
0
0
0
0
0
0
20
0
0
0
0
0
0
0
0
0
0
0
0
0
0
18
0
0
0
0
0
0
0
RM0090 Rev 19
0
0
0
0
0
0
0
16
0
0
0
0
0
0
0
0
OT15 15
0
0
0
0
0
0
0
0
OT14 14
0
0
0
0
0
0
0
0
OT13 13
0
0
0
0
0
0
0
0
OT12 12
0
0
0
0
0
0
0
0
OT11
Table 39. GPIO register map and reset values
0
0
0
0
0
0
0
0 OT10 10
0
0
0
0
0
1
0
0
OT9 9
The following table gives the GPIO register map and the reset values.
1
0
0
0
0
0
0
OT8 8
0
0
1
0
0
1
0
OT7 7
0
0
1
0
0
0
0
OT6 6
0
0
0
0
0
0
0
OT5 5
0
0
0
0
0
0
0
OT4 4
0
0
0
0
0
0
0
OT3 3
0
0
0
0
0
0
0
OT2 2
0
0
0
0
0
0
0
OT1 1
0
0
0
0
0
0
0
OT0
287/1751
General-purpose I/Os (GPIO)
288
Table 39. GPIO register map and reset values (continued)
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
PUPDR15[1:0]
PUPDR14[1:0]
PUPDR13[1:0]
PUPDR12[1:0]
PUPDR10[1:0]
PUPDR11[1:0]
PUPDR9[1:0]
PUPDR8[1:0]
PUPDR7[1:0]
PUPDR6[1:0]
PUPDR5[1:0]
PUPDR4[1:0]
PUPDR3[1:0]
PUPDR2[1:0]
PUPDR1[1:0]
PUPDR0[1:0]
GPIOx_PUPDR
(where x =
0x0C C..I/J/K)
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
GPIOx_IDR
IDR15
IDR14
IDR13
IDR12
IDR10
IDR11
IDR9
IDR8
IDR7
IDR6
IDR5
IDR4
IDR3
IDR2
IDR1
IDR0
(where x =
0x10 A..I/J/K) Reserved
Reset value x x x x x x x x x x x x x x x x
GPIOx_ODR
ODR15
ODR14
ODR13
ODR12
ODR10
ODR11
ODR9
ODR8
ODR7
ODR6
ODR5
ODR4
ODR3
ODR2
ODR1
ODR0
(where x =
0x14 A..I/J/K) Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
GPIOx_BSRR
BR15
BR14
BR13
BR12
BR10
BS15
BS14
BS13
BS12
BS10
BR11
BS11
BR9
BR8
BR7
BR6
BR5
BR4
BR3
BR2
BR1
BR0
BS9
BS8
BS7
BS6
BS5
BS4
BS3
BS2
BS1
BS0
(where x =
0x18 A..I/J/K)
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
GPIOx_LCKR
LCK15
LCK14
LCK13
LCK12
LCK10
LCK11
LCKK
LCK9
LCK8
LCK7
LCK6
LCK5
LCK4
LCK3
LCK2
LCK1
LCK0
(where x =
0x1C A..I/J/K) Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
GPIOx_AFRL
(where x = AFRL7[3:0] AFRL6[3:0] AFRL5[3:0] AFRL4[3:0] AFRL3[3:0] AFRL2[3:0] AFRL1[3:0] AFRL0[3:0]
0x20 A..I/J/K)
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
GPIOx_AFRH
(where x = AFRH15[3:0] AFRH14[3:0] AFRH13[3:0] AFRH12[3:0] AFRH11[3:0] AFRH10[3:0] AFRH9[3:0] AFRH8[3:0]
0x24 A..I/J)
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
288/1751 RM0090 Rev 19

RM0090 System configuration controller (SYSCFG)
9 System configuration controller (SYSCFG)
The system configuration controller is mainly used to remap the memory accessible in the
code area, select the Ethernet PHY interface and manage the external interrupt line
connection to the GPIOs.
9.1 I/O compensation cell

By default the I/O compensation cell is not used. However when the I/O output buffer speed
is configured in 50 MHz or 100 MHz mode, it is recommended to use the compensation cell
for slew rate control on I/O tf(IO)out)/tr(IO)out commutation to reduce the I/O noise on power
supply.
When the compensation cell is enabled, a READY flag is set to indicate that the
compensation cell is ready and can be used. The I/O compensation cell can be used only
when the supply voltage ranges from 2.4 to 3.6 V.
9.2 SYSCFG registers for STM32F405xx/07xx and

STM32F415xx/17xx
9.2.1 SYSCFG memory remap register (SYSCFG_MEMRMP)

This register is used for specific configurations on memory remap:
• Two bits are used to configure the type of memory accessible at address 0x0000 0000.
These bits are used to select the physical remap by software and so, bypass the BOOT
pins.
• After reset these bits take the value selected by the BOOT pins. When booting from
main Flash memory with BOOT pins set to 10 [(BOOT1,BOOT0) = (1,0)] this register
takes the value 0x00.
When the FSMC is remapped at address 0x0000 0000, only the first two regions of Bank 1
memory controller (Bank1 NOR/PSRAM 1 and NOR/PSRAM 2) can be remapped. In remap
mode, the CPU can access the external memory via ICode bus instead of System bus
which boosts up the performance.
Reset value: 0x0000 000X (X is the memory mode selected by the BOOT pins)
)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MEM_MODE
Reserved
rw rw
RM0090 Rev 19 289/1751

301
System configuration controller (SYSCFG) RM0090

Bits 1:0 MEM_MODE: Memory mapping selection
Set and cleared by software. This bit controls the memory internal mapping at
address 0x0000 0000. After reset these bits take the value selected by the Boot
pins (except for FSMC).
00: Main Flash memory mapped at 0x0000 0000
01: System Flash memory mapped at 0x0000 0000
10: FSMC Bank1 (NOR/PSRAM 1 and 2) mapped at 0x0000 0000
11: Embedded SRAM (SRAM1) mapped at 0x0000 0000
Note: Refer to Section 2.3: Memory map for details about the memory mapping at
address 0x0000 0000.
9.2.2 SYSCFG peripheral mode configuration register (SYSCFG_PMC)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
MII_RMII
Reserved _SEL Reserved
rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserved

Bit 23 MII_RMII_SEL: Ethernet PHY interface selection
Set and Cleared by software.These bits control the PHY interface for the
Ethernet MAC.
0: MII interface is selected
1: RMII PHY interface is selected
Note: This configuration must be done while the MAC is under reset and before
enabling the MAC clocks.
290/1751 RM0090 Rev 19


(SYSCFG_EXTICR1)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
EXTI3[3:0] EXTI2[3:0] EXTI1[3:0] EXTI0[3:0]

Bits 15:0 EXTIx[3:0]: EXTI x configuration (x = 0 to 3)
These bits are written by software to select the source input for the EXTIx
external interrupt.
0000: PA[x] pin
0001: PB[x] pin
0010: PC[x] pin
0011: PD[x] pin
0100: PE[x] pin
0101: PF[x] pin
0110: PG[x] pin
0111: PH[x] pin
1000: PI[x] pin

(SYSCFG_EXTICR2)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RM0090 Rev 19 291/1751

301

external interrupt.
0000: PA[x] pin
0001: PB[x] pin
0010: PC[x] pin
0011: PD[x] pin
0100: PE[x] pin
0101: PF[x] pin
0110: PG[x] pin
0111: PH[x] pin
1000: PI[x] pin

(SYSCFG_EXTICR3)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

These bits are written by software to select the source input for the EXTIx external
interrupt.
0000: PA[x] pin
0001: PB[x] pin
0010: PC[x] pin
0011: PD[x] pin
0100: PE[x] pin
0101: PF[x] pin
0110: PG[x] pin
0111: PH[x] pin
1000: PI[x] pin
292/1751 RM0090 Rev 19


(SYSCFG_EXTICR4)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

interrupt.
0000: PA[x] pin
0001: PB[x] pin
0010: PC[x] pin
0011: PD[x] pin
0100: PE[x] pin
0101: PF[x] pin
0110: PG[x] pin
0111: PH[x] pin
Note: PI[15:12] are not used.
9.2.7 Compensation cell control register (SYSCFG_CMPCR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
READY CMP_PD
Reserved Reserved
r rw

Bit 8 READY: Compensation cell ready flag
0: I/O compensation cell not ready
1: O compensation cell ready
Bit 0 CMP_PD: Compensation cell power-down
0: I/O compensation cell power-down mode
1: I/O compensation cell enabled
RM0090 Rev 19 293/1751

301
9.2.8 SYSCFG register maps for

The following table gives the SYSCFG register map and the reset values.
Table 40. SYSCFG register map and reset values (STM32F405xx/07xx and STM32F415xx/17xx)
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
MEM_MODE
SYSCFG_
0x00 MEMRMP Reserved
Reset value x x
MII_RMII_SEL
SYSCFG_PMC
Reset value 0
SYSCFG_EXTICR1 EXTI3[3:0] EXTI2[3:0] EXTI1[3:0] EXTI0[3:0]
0x08 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x0C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x10 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x14 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CMP_PD
READY
SYSCFG_CMPCR
Reset value 0 0
9.3 SYSCFG registers for STM32F42xxx and STM32F43xxx
9.3.1 SYSCFG memory remap register (SYSCFG_MEMRMP)

This register is used for specific configurations on memory remap:
• Three bits are used to configure the type of memory accessible at address
0x0000 0000. These bits are used to select the physical remap by software and so,
bypass the BOOT pins.
• After reset these bits take the value selected by the BOOT pins. When booting from
main Flash memory with BOOT pins set to 10 [(BOOT1,BOOT0) = (1,0)] this register
takes the value 0x00.
• Other bits are used to swap FMC SDRAM Bank 1/2 with FMC Bank 3/4 and configure
the Flash Bank 1/2 mapping
294/1751 RM0090 Rev 19

There are two possible FMC remap at address 0x0000 0000:

• FMC Bank 1 (NOR/PSRAM 1 and 2) remap:
Only the first two regions of Bank 1 memory controller (Bank1 NOR/PSRAM 1 and
NOR/PSRAM 2) can be remapped.
• FMC SDRAM Bank 1 remap.
In remap mode at address 0x0000 0000, the CPU can access the external memory via
ICode bus instead of System bus which boosts up the performance.

Reset value: 0x0000 000X (X is the memory mode selected by the BOOT pins)
Note: Booting from NOR Flash memory or SDRAM is not allowed. The regions can only be
mapped at 0x0000 0000 through software remap.
)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FB_
SWP_FMC
MODE
MEM_MODE[2:0]
Reserved Res. Reserved
rw rw rw rw rw rw

Bits 11:10 SWP_FMC: FMC memory mapping swap
Set and cleared by software. These bits are used to swap the FMC SDRAM
Bank 1/2 and FMC Bank 3/4 (SDRAM Bank 1/2 and NAND Bank 2/PCCARD
Bank) in order to enable the code execution from SDRAM Banks without a
physical remapping at 0x0000 0000 address.
00: No FMC memory mapping swap
01: SDRAM banks and NAND Bank 2/PCCARD mapping are swapped. SDRAM
Bank 1 and 2 are mapped at NAND Bank 2 (0x8000 0000) and PCCARD Bank
(0x9000 0000) address, respectively. NAND Bank 2 and PCCARD Bank are
mapped at 0xC000 0000 and 0xD000 0000, respectively.
10: Reserved
11: Reserved
RM0090 Rev 19 295/1751

301
Bit 8 FB_MODE: Flash Bank mode selection

Set and cleared by software. This bit controls the Flash Bank 1/2 mapping.
0: Flash Bank 1 is mapped at 0x0800 0000 (and aliased at 0x0000 0000) and
Flash Bank 2 is mapped at 0x0810 0000 (and aliased at 0x0010 0000)
1: Flash Bank 2 is mapped at 0x0800 0000 (and aliased at 0x0000 0000) and
Flash Bank 1 is mapped at 0x0810 0000 (and aliased at 0x0010 0000)
Bits 2:0 MEM_MODE: Memory mapping selection
Set and cleared by software. This bit controls the memory internal mapping at
address 0x0000 0000. After reset these bits take the value selected by the Boot
pins (except for FMC).
000: Main Flash memory mapped at 0x0000 0000
001: System Flash memory mapped at 0x0000 0000
010: FMC Bank1 (NOR/PSRAM 1 and 2) mapped at 0x0000 0000
011: Embedded SRAM (SRAM1) mapped at 0x0000 0000
100: FMC/SDRAM Bank 1 mapped at 0x0000 0000
Other configurations are reserved
Note: Refer to Section 2.3: Memory map for details about the memory mapping at
address 0x0000 0000.
9.3.2 SYSCFG peripheral mode configuration register (SYSCFG_PMC)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
MII_RMII
ADCxDC2
Reserved _SEL Reserved
rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserved

Bit 23 MII_RMII_SEL: Ethernet PHY interface selection
Set and Cleared by software.These bits control the PHY interface for the
Ethernet MAC.
0: MII interface is selected
1: RMII PHY interface is selected
Note: This configuration must be done while the MAC is under reset and before
enabling the MAC clocks.
296/1751 RM0090 Rev 19


Bits 18:16 ADCxDC2:
0: No effect.
1: Refer to AN4073 on how to use this bit.
Note: These bits can be set only if the following conditions are met:
- ADC clock higher or equal to 30 MHz.
- Only one ADCxDC2 bit must be selected if ADC conversions do not start
at the same time and the sampling times differ.
- These bits must not be set when the ADCDC1 bit is set in PWR_CR
register.

(SYSCFG_EXTICR1)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

external interrupt.
0000: PA[x] pin
0001: PB[x] pin
0010: PC[x] pin
0011: PD[x] pin
0100: PE[x] pin
0101: PF[x] pin
0110: PG[x] pin
0111: PH[x] pin
1000: PI[x] pin
1001: PJ[x] pin
1010: PK[x] pin
RM0090 Rev 19 297/1751

301

(SYSCFG_EXTICR2)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

external interrupt.
0000: PA[x] pin
0001: PB[x] pin
0010: PC[x] pin
0011: PD[x] pin
0100: PE[x] pin
0101: PF[x] pin
0110: PG[x] pin
0111: PH[x] pin
1000: PI[x] pin
1001: PJ[x] pin
1010: PK[x] pin

(SYSCFG_EXTICR3)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
298/1751 RM0090 Rev 19


external interrupt.
0000: PA[x] pin
0001: PB[x] pin
0010: PC[x] pin
0011: PD[x] pin
0100: PE[x] pin
0101: PF[x] pin
0110: PG[x] pin
0111: PH[x] pin
1000: PI[x] pin
1001: PJ[x] pin
Note: PK[11:8] are not used.

(SYSCFG_EXTICR4)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

interrupt.
0000: PA[x] pin
0001: PB[x] pin
0010: PC[x] pin
0011: PD[x] pin
0100: PE[x] pin
0101: PF[x] pin
0110: PG[x] pin
0111: PH[x] pin
1000: PI[x] pin
1001: PJ[x] pin
Note: PK[15:12] are not used.
RM0090 Rev 19 299/1751

301
9.3.7 Compensation cell control register (SYSCFG_CMPCR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
READY CMP_PD
Reserved Reserved
r rw

Bit 8 READY: Compensation cell ready flag
0: I/O compensation cell not ready
1: O compensation cell ready
Bit 0 CMP_PD: Compensation cell power-down
0: I/O compensation cell power-down mode
1: I/O compensation cell enabled
300/1751 RM0090 Rev 19

9.3.8 SYSCFG register maps for STM32F42xxx and STM32F43xxx

The following table gives the SYSCFG register map and the reset values.
Table 41. SYSCFG register map and reset values (STM32F42xxx and STM32F43xxx)
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
SWP_FMC
FB_MODE
Reserved
SYSCFG_ MEM_
0x00 MEMRMP Reserved Reserved MODE
Reset value MII_RMII_SEL 0 0 0 x x x
ADC3DC2
ADC2DC2
ADC1DC2
SYSCFG_PMC
Reset value 0 0 0 0
0x08 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x0C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x10 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x14 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CMP_PD
READY
SYSCFG_CMPCR
Reset value 0 0
RM0090 Rev 19 301/1751

301
DMA controller (DMA) RM0090
10 DMA controller (DMA)
10.1 DMA introduction

Direct memory access (DMA) is used in order to provide high-speed data transfer between
peripherals and memory and between memory and memory. Data can be quickly moved by
DMA without any CPU action. This keeps CPU resources free for other operations.
The DMA controller combines a powerful dual AHB master bus architecture with
independent FIFO to optimize the bandwidth of the system, based on a complex bus matrix
architecture.
The two DMA controllers have 16 streams in total (8 for each controller), each dedicated to
managing memory access requests from one or more peripherals. Each stream can have
up to 8 channels (requests) in total. And each has an arbiter for handling the priority
between DMA requests.
10.2 DMA main features

The main DMA features are:
• Dual AHB master bus architecture, one dedicated to memory accesses and one
dedicated to peripheral accesses
• AHB slave programming interface supporting only 32-bit accesses
• 8 streams for each DMA controller, up to 8 channels (requests) per stream
• Four-word depth 32 first-in, first-out memory buffers (FIFOs) per stream, that can be
used in FIFO mode or direct mode:
– FIFO mode: with threshold level software selectable between 1/4, 1/2 or 3/4 of the
FIFO size
– Direct mode
Each DMA request immediately initiates a transfer from/to the memory. When it is
configured in direct mode (FIFO disabled), to transfer data in memory-to-
peripheral mode, the DMA preloads only one data from the memory to the internal
302/1751 RM0090 Rev 19

RM0090 DMA controller (DMA)
FIFO to ensure an immediate data transfer as soon as a DMA request is triggered

by a peripheral.
• Each stream can be configured by hardware to be:
– a regular channel that supports peripheral-to-memory, memory-to-peripheral and
memory-to-memory transfers
– a double buffer channel that also supports double buffering on the memory side
• Each of the 8 streams are connected to dedicated hardware DMA channels (requests)
• Priorities between DMA stream requests are software-programmable (4 levels
consisting of very high, high, medium, low) or hardware in case of equality (request 0
has priority over request 1, etc.)
• Each stream also supports software trigger for memory-to-memory transfers (only
available for the DMA2 controller)
• Each stream request can be selected among up to 8 possible channel requests. This
selection is software-configurable and allows several peripherals to initiate DMA
requests
• The number of data items to be transferred can be managed either by the DMA
controller or by the peripheral:
– DMA flow controller: the number of data items to be transferred is software-
programmable from 1 to 65535
– Peripheral flow controller: the number of data items to be transferred is unknown
and controlled by the source or the destination peripheral that signals the end of
the transfer by hardware
• Independent source and destination transfer width (byte, half-word, word): when the
data widths of the source and destination are not equal, the DMA automatically
packs/unpacks the necessary transfers to optimize the bandwidth. This feature is only
available in FIFO mode
• Incrementing or non-incrementing addressing for source and destination
• Supports incremental burst transfers of 4, 8 or 16 beats. The size of the burst is
software-configurable, usually equal to half the FIFO size of the peripheral
• Each stream supports circular buffer management
• 5 event flags (DMA Half Transfer, DMA Transfer complete, DMA Transfer Error, DMA
FIFO Error, Direct Mode Error) logically ORed together in a single interrupt request for
each stream
RM0090 Rev 19 303/1751

338
10.3 DMA functional description
10.3.1 General description

Figure 32 shows the block diagram of a DMA.
Figure 32. DMA block diagram
DMA controller
AHB master
REQ_STR0_CH0
REQ_STR0_CH1 Memory port
REQ_STR0_CH7
STREAM 1
STREAM 2
STREAM 3
STREAM 5
STREAM 6
STREAM 0
STREAM 4
STREAM 7
REQ_STR1_CH0
REQ_STR1_CH1
REQ_STREAM0
REQ_STREAM1
REQ_STR1_CH7 REQ_STREAM2
REQ_STREAM3
FIFO
FIFO
FIFO
FIFO
FIFO
FIFO
FIFO
FIFO
REQ_STREAM4 Arbiter
REQ_STREAM5
REQ_STREAM6
STREAM 0
STREAM 1
STREAM 2
STREAM 3
STREAM 4
STREAM 5
STREAM 6
STREAM 7
REQ_STREAM7
REQ_STR7_CH0
REQ_STR7_CH1
AHB master
REQ_STR7_CH7 Peripheral port
Channel
selection
AHB slave
programming Programming port
interface
ai15945
The DMA controller performs direct memory transfer: as an AHB master, it can take the
control of the AHB bus matrix to initiate AHB transactions.
It can carry out the following transactions:
• peripheral-to-memory
• memory-to-peripheral
• memory-to-memory
The DMA controller provides two AHB master ports: the AHB memory port, intended to be
connected to memories and the AHB peripheral port, intended to be connected to
peripherals. However, to allow memory-to-memory transfers, the AHB peripheral port must
also have access to the memories.
The AHB slave port is used to program the DMA controller (it supports only 32-bit
accesses).
304/1751 RM0090 Rev 19

See Figure 33 and Figure 34 for the implementation of the system of two DMA controllers.
Figure 33. System implementation of the two DMA controllers

DCODE
Bus matrix
(AHB Flash
ICODE
multilayer) memory
112 KB SRAM
16 KB SRAM
AHB1 peripherals
DMA controller 2
AHB memory
port
peripherals
AHB-APB APB2
APB2
To AHB2
bridge2
Arbiter
FIFO
AHB slave
peripherals
(dual AHB)
AHB periph
port
AHB-APB APB1
APB1
bridge1
peripherals
(dual AHB)
DMA request
MAPPING
AHB memory
AHB2 peripherals
port
peripherals
Arbiter
To AHB2
AHB slave
FIFO
External memory
AHB periph
controller (FSMC)
port
DMA controller 1
MS19927V2
1. The DMA1 controller AHB peripheral port is not connected to the bus matrix like DMA2 controller. As a result, only DMA2
streams are able to perform memory-to-memory transfers.
RM0090 Rev 19 305/1751

338
Figure 34. System implementation of the two DMA controllers (STM32F42xxx and
STM32F43xxx)
Bus Matrix
(AHB multilayer) DCODE
Flash
ICODE
memory
112 KB SRAM
16 KB SRAM
64 KB SRAM
AHB1 peripherals
DMA controller 2 AHB memory
port
peripherals
AHB-APB APB2
APB2
To AHB2
bridge2
Arbiter
FIFO
AHB slave
peripherals
(dual AHB)
AHB periph
port
AHB-APB APB1
APB1
bridge1
peripherals
(dual AHB)
DMA request
MAPPING
AHB memory
AHB2 peripherals
port
peripherals
Arbiter
To AHB2
AHB slave
FIFO
External memory
AHB periph
controller (FMC)
port
DMA controller 1
MS30438V1
1. The DMA1 controller AHB peripheral port is not connected to the bus matrix like in the case of the DMA2
controller, thus only DMA2 streams are able to perform memory-to-memory transfers.
10.3.2 DMA transactions

A DMA transaction consists of a sequence of a given number of data transfers. The number
of data items to be transferred and their width (8-bit, 16-bit or 32-bit) are software-
programmable.
Each DMA transfer consists of three operations:
• A loading from the peripheral data register or a location in memory, addressed through
the DMA_SxPAR or DMA_SxM0AR register
• A storage of the data loaded to the peripheral data register or a location in memory
addressed through the DMA_SxPAR or DMA_SxM0AR register
• A post-decrement of the DMA_SxNDTR register, which contains the number of
transactions that still have to be performed
306/1751 RM0090 Rev 19

After an event, the peripheral sends a request signal to the DMA controller. The DMA
controller serves the request depending on the channel priorities. As soon as the DMA
controller accesses the peripheral, an Acknowledge signal is sent to the peripheral by the
DMA controller. The peripheral releases its request as soon as it gets the Acknowledge
signal from the DMA controller. Once the request has been deasserted by the peripheral,
the DMA controller releases the Acknowledge signal. If there are more requests, the
peripheral can initiate the next transaction.
10.3.3 Channel selection

Each stream is associated with a DMA request that can be selected out of 8 possible
channel requests. The selection is controlled by the CHSEL[2:0] bits in the DMA_SxCR
register.
Figure 35. Channel selection

REQ_STRx_CH7
REQ_STRx_CH6
REQ_STRx_CH5
REQ_STREAMx
REQ_STRx_CH4
REQ_STRx_CH3
REQ_STRx_CH2
REQ_STRx_CH1
REQ_STRx_CH0
31 27 25 0
DMA_SxCR CHSEL[2:0]
ai15947b
The 8 requests from the peripherals (TIM, ADC, SPI, I2C, etc.) are independently connected
to each channel and their connection depends on the product implementation.
See the following table(s) for examples of DMA request mappings.
Table 42. DMA1 request mapping

Peripheral
Stream 0 Stream 1 Stream 2 Stream 3 Stream 4 Stream 5 Stream 6 Stream 7
requests
Channel 0 SPI3_RX - SPI3_RX SPI2_RX SPI2_TX SPI3_TX - SPI3_TX
Channel 1 I2C1_RX - TIM7_UP - TIM7_UP I2C1_RX I2C1_TX I2C1_TX
I2S3_EXT_ I2S2_EXT_ I2S3_EXT_

Channel 2 TIM4_CH1 - TIM4_CH2 TIM4_UP TIM4_CH3
RX TX TX
I2S3_EXT_ TIM2_UP I2S2_EXT_ TIM2_CH2 TIM2_UP

Channel 3 I2C3_RX I2C3_TX TIM2_CH1
RX TIM2_CH3 RX TIM2_CH4 TIM2_CH4
Channel 4 UART5_RX USART3_RX UART4_RX USART3_TX UART4_TX USART2_RX USART2_TX UART5_TX
(1) TIM3_CH4 TIM3_CH1

Channel 5 UART8_TX UART7_TX(1) UART7_RX(1) TIM3_CH2 UART8_RX(1) TIM3_CH3
TIM3_UP TIM3_TRIG
RM0090 Rev 19 307/1751

338
Table 42. DMA1 request mapping (continued)

Peripheral
requests
TIM5_CH3 TIM5_CH4 TIM5_CH4

Channel 6 TIM5_CH1 TIM5_CH2 - TIM5_UP -
TIM5_UP TIM5_TRIG TIM5_TRIG
Channel 7 - TIM6_UP I2C2_RX I2C2_RX USART3_TX DAC1 DAC2 I2C2_TX
1. These requests are available on STM32F42xxx and STM32F43xxx only.
Table 43. DMA2 request mapping

Peripheral
requests
TIM8_CH1 TIM1_CH1
Channel 0 ADC1 SAI1_A(1) TIM8_CH2 SAI1_A(1) ADC1 SAI1_B(1) TIM1_CH2 -
TIM8_CH3 TIM1_CH3
Channel 1 - DCMI ADC2 ADC2 SAI1_B(1) SPI6_TX(1) SPI6_RX(1) DCMI
Channel 2 ADC3 ADC3 - SPI5_RX(1) SPI5_TX(1) CRYP_OUT CRYP_IN HASH_IN
Channel 3 SPI1_RX - SPI1_RX SPI1_TX - SPI1_TX - -

Channel 4 SPI4_RX(1) SPI4_TX(1) USART1_RX SDIO - USART1_RX SDIO USART1_TX
Channel 5 - USART6_RX USART6_RX SPI4_RX(1) SPI4_TX(1) - USART6_TX USART6_TX
TIM1_CH4
Channel 6 TIM1_TRIG TIM1_CH1 TIM1_CH2 TIM1_CH1 TIM1_TRIG TIM1_UP TIM1_CH3 -
TIM1_COM
TIM8_CH4
Channel 7 - TIM8_UP TIM8_CH1 TIM8_CH2 TIM8_CH3 SPI5_RX(1) SPI5_TX(1) TIM8_TRIG
TIM8_COM
1. These requests are available on STM32F42xxx and STM32F43xxx.
10.3.4 Arbiter
An arbiter manages the 8 DMA stream requests based on their priority for each of the two
AHB master ports (memory and peripheral ports) and launches the peripheral/memory
access sequences.
Priorities are managed in two stages:
• Software: each stream priority can be configured in the DMA_SxCR register. There are
four levels:
– Very high priority
– High priority
– Medium priority
– Low priority
• Hardware: If two requests have the same software priority level, the stream with the
lower number takes priority over the stream with the higher number. For example,
Stream 2 takes priority over Stream 4.
308/1751 RM0090 Rev 19

10.3.5 DMA streams

Each of the 8 DMA controller streams provides a unidirectional transfer link between a
source and a destination.
Each stream can be configured to perform:
• Regular type transactions: memory-to-peripherals, peripherals-to-memory or memory-
to-memory transfers
• Double-buffer type transactions: double buffer transfers using two memory pointers for
the memory (while the DMA is reading/writing from/to a buffer, the application can
write/read to/from the other buffer).
The amount of data to be transferred (up to 65535) is programmable and related to the
source width of the peripheral that requests the DMA transfer connected to the peripheral
AHB port. The register that contains the amount of data items to be transferred is
decremented after each transaction.
10.3.6 Source, destination and transfer modes

Both source and destination transfers can address peripherals and memories in the entire
4 GB area, at addresses comprised between 0x0000 0000 and 0xFFFF FFFF.
The direction is configured using the DIR[1:0] bits in the DMA_SxCR register and offers
three possibilities: memory-to-peripheral, peripheral-to-memory or memory-to-memory
transfers. Table 44 describes the corresponding source and destination addresses.
Table 44. Source and destination address

Bits DIR[1:0] of the
Direction Source address Destination address
DMA_SxCR register
00 Peripheral-to-memory DMA_SxPAR DMA_SxM0AR

01 Memory-to-peripheral DMA_SxM0AR DMA_SxPAR
10 Memory-to-memory DMA_SxPAR DMA_SxM0AR
11 reserved - -
When the data width (programmed in the PSIZE or MSIZE bits in the DMA_SxCR register)
is a half-word or a word, respectively, the peripheral or memory address written into the
DMA_SxPAR or DMA_SxM0AR/M1AR registers has to be aligned on a word or half-word
address boundary, respectively.
Peripheral-to-memory mode
Figure 36 describes this mode.
When this mode is enabled (by setting the bit EN in the DMA_SxCR register), each time a
peripheral request occurs, the stream initiates a transfer from the source to fill the FIFO.
When the threshold level of the FIFO is reached, the contents of the FIFO are drained and
stored into the destination.
The transfer stops once the DMA_SxNDTR register reaches zero, when the peripheral
requests the end of transfers (in case of a peripheral flow controller) or when the EN bit in
the DMA_SxCR register is cleared by software.
RM0090 Rev 19 309/1751

338
In direct mode (when the DMDIS value in the DMA_SxFCR register is ‘0’), the threshold
level of the FIFO is not used: after each single data transfer from the peripheral to the FIFO,
the corresponding data are immediately drained and stored into the destination.
The stream has access to the AHB source or destination port only if the arbitration of the
corresponding stream is won. This arbitration is performed using the priority defined for
each stream using the PL[1:0] bits in the DMA_SxCR register.
Figure 36. Peripheral-to-memory mode
DMA controller DMA_SxM0AR
DMA_SxM1AR(1)
AHB memory Memory bus

port
Memory
destination
FIFO
REQ_STREAMx Arbiter level
FIFO
AHB peripheral Peripheral bus

port
Peripheral
DMA_SxPAR source
Peripheral DMA request

MS47543V1
1. For double-buffer mode.
Memory-to-peripheral mode
Figure 37 describes this mode.
When this mode is enabled (by setting the EN bit in the DMA_SxCR register), the stream
immediately initiates transfers from the source to entirely fill the FIFO.
Each time a peripheral request occurs, the contents of the FIFO are drained and stored into
the destination. When the level of the FIFO is lower than or equal to the predefined
threshold level, the FIFO is fully reloaded with data from the memory.
The transfer stops once the DMA_SxNDTR register reaches zero, when the peripheral
requests the end of transfers (in case of a peripheral flow controller) or when the EN bit in
the DMA_SxCR register is cleared by software.
In direct mode (when the DMDIS value in the DMA_SxFCR register is '0'), the threshold
level of the FIFO is not used. Once the stream is enabled, the DMA preloads the first data to
transfer into an internal FIFO. As soon as the peripheral requests a data transfer, the DMA
transfers the preloaded value into the configured destination. It then reloads again the
310/1751 RM0090 Rev 19

empty internal FIFO with the next data to be transfer. The preloaded data size corresponds
to the value of the PSIZE bitfield in the DMA_SxCR register.
Figure 37. Memory-to-peripheral mode
DMA_SxM1AR(1)
Memory bus
AHB memory
port
Memory
source
FIFO
Arbiter
REQ_STREAMx level FIFO

port
Peripheral
DMA_SxPAR destination
Peripheral DMA request
ai15949
Memory-to-memory mode
The DMA channels can also work without being triggered by a request from a peripheral.
This is the memory-to-memory mode, described in Figure 38.
When the stream is enabled by setting the Enable bit (EN) in the DMA_SxCR register, the
stream immediately starts to fill the FIFO up to the threshold level. When the threshold level
is reached, the FIFO contents are drained and stored into the destination.
The transfer stops once the DMA_SxNDTR register reaches zero or when the EN bit in the
DMA_SxCR register is cleared by software.
Note: When memory-to-memory mode is used, the Circular and direct modes are not allowed.
Only the DMA2 controller is able to perform memory-to-memory transfers.
RM0090 Rev 19 311/1751

338
Figure 38. Memory-to-memory mode
DMA_SxM1AR(1)
AHB memory Memory bus

port
Memory 2
destination
Arbiter FIFO
Stream enable level FIFO

port
Memory 1
DMA_SxPAR source
ai15950
10.3.7 Pointer incrementation

Peripheral and memory pointers can optionally be automatically post-incremented or kept
constant after each transfer depending on the PINC and MINC bits in the DMA_SxCR
register.
Disabling the Increment mode is useful when the peripheral source or destination data are
accessed through a single register.
If the Increment mode is enabled, the address of the next transfer will be the address of the
previous one incremented by 1 (for bytes), 2 (for half-words) or 4 (for words) depending on
the data width programmed in the PSIZE or MSIZE bits in the DMA_SxCR register.
In order to optimize the packing operation, it is possible to fix the increment offset size for
the peripheral address whatever the size of the data transferred on the AHB peripheral port.
The PINCOS bit in the DMA_SxCR register is used to align the increment offset size with
the data size on the peripheral AHB port, or on a 32-bit address (the address is then
incremented by 4). The PINCOS bit has an impact on the AHB peripheral port only.
If PINCOS bit is set, the address of the next transfer is the address of the previous one
incremented by 4 (automatically aligned on a 32-bit address) whatever the PSIZE value.
The AHB memory port, however, is not impacted by this operation.
312/1751 RM0090 Rev 19

10.3.8 Circular mode

The Circular mode is available to handle circular buffers and continuous data flows (e.g.
ADC scan mode). This feature can be enabled using the CIRC bit in the DMA_SxCR
register.
When the circular mode is activated, the number of data items to be transferred is
automatically reloaded with the initial value programmed during the stream configuration
phase, and the DMA requests continue to be served.
Note: In the circular mode, it is mandatory to respect the following rule in case of a burst mode
configured for memory:
DMA_SxNDTR = Multiple of ((Mburst beat) × (Msize)/(Psize)), where:
– (Mburst beat) = 4, 8 or 16 (depending on the MBURST bits in the DMA_SxCR
register)
– ((Msize)/(Psize)) = 1, 2, 4, 1/2 or 1/4 (Msize and Psize represent the MSIZE and
PSIZE bits in the DMA_SxCR register. They are byte dependent)
– DMA_SxNDTR = Number of data items to transfer on the AHB peripheral port
For example: Mburst beat = 8 (INCR8), MSIZE = ‘00’ (byte) and PSIZE = ‘01’ (half-word), in
this case: DMA_SxNDTR must be a multiple of (8 × 1/2 = 4).
If this formula is not respected, the DMA behavior and data integrity are not guaranteed.
NDTR must also be a multiple of the Peripheral burst size multiplied by the peripheral data
size, otherwise this could result in a bad DMA behavior.
10.3.9 Double buffer mode

This mode is available for all the DMA1 and DMA2 streams.
The Double buffer mode is enabled by setting the DBM bit in the DMA_SxCR register.
A double-buffer stream works as a regular (single buffer) stream with the difference that it
has two memory pointers. When the Double buffer mode is enabled, the Circular mode is
automatically enabled (CIRC bit in DMA_SxCR is don’t care) and at each end of transaction,
the memory pointers are swapped.
In this mode, the DMA controller swaps from one memory target to another at each end of
transaction. This allows the software to process one memory area while the second memory
area is being filled/used by the DMA transfer. The double-buffer stream can work in both
directions (the memory can be either the source or the destination) as described in
Table 45: Source and destination address registers in Double buffer mode (DBM=1).
Note: In Double buffer mode, it is possible to update the base address for the AHB memory port
on-the-fly (DMA_SxM0AR or DMA_SxM1AR) when the stream is enabled, by respecting the
following conditions:
• When the CT bit is ‘0’ in the DMA_SxCR register, the DMA_SxM1AR register can be
written. Attempting to write to this register while CT = '1' sets an error flag (TEIF) and
the stream is automatically disabled.
• When the CT bit is ‘1’ in the DMA_SxCR register, the DMA_SxM0AR register can be
written. Attempting to write to this register while CT = '0', sets an error flag (TEIF) and
the stream is automatically disabled.
To avoid any error condition, it is advised to change the base address as soon as the TCIF
flag is asserted because, at this point, the targeted memory must have changed from
RM0090 Rev 19 313/1751

338
memory 0 to 1 (or from 1 to 0) depending on the value of CT in the DMA_SxCR register in

accordance with one of the two above conditions.
For all the other modes (except the Double buffer mode), the memory address registers are
write-protected as soon as the stream is enabled.
Table 45. Source and destination address registers in Double buffer mode (DBM=1)
Bits DIR[1:0] of the
Direction Source address Destination address
DMA_SxCR register
DMA_SxM0AR /
00 Peripheral-to-memory DMA_SxPAR
DMA_SxM1AR
DMA_SxM0AR /
01 Memory-to-peripheral DMA_SxPAR
DMA_SxM1AR
10 Not allowed(1)
11 Reserved - -
1. When the Double buffer mode is enabled, the Circular mode is automatically enabled. Since the memory-
to-memory mode is not compatible with the Circular mode, when the Double buffer mode is enabled, it is
not allowed to configure the memory-to-memory mode.
10.3.10 Programmable data width, packing/unpacking, endianess

The number of data items to be transferred has to be programmed into DMA_SxNDTR
(number of data items to transfer bit, NDT) before enabling the stream (except when the
flow controller is the peripheral, PFCTRL bit in DMA_SxCR is set).
When using the internal FIFO, the data widths of the source and destination data are
programmable through the PSIZE and MSIZE bits in the DMA_SxCR register (can be 8-,
16- or 32-bit).
When PSIZE and MSIZE are not equal:
• The data width of the number of data items to transfer, configured in the DMA_SxNDTR
register is equal to the width of the peripheral bus (configured by the PSIZE bits in the
DMA_SxCR register). For instance, in case of peripheral-to-memory, memory-to-
peripheral or memory-to-memory transfers and if the PSIZE[1:0] bits are configured for
half-word, the number of bytes to be transferred is equal to 2 × NDT.
• The DMA controller only copes with little-endian addressing for both source and
destination. This is described in Table 46: Packing/unpacking & endian behavior (bit
PINC = MINC = 1).
This packing/unpacking procedure may present a risk of data corruption when the operation
is interrupted before the data are completely packed/unpacked. So, to ensure data
coherence, the stream may be configured to generate burst transfers: in this case, each
group of transfers belonging to a burst are indivisible (refer to Section 10.3.11: Single and
burst transfers).
In direct mode (DMDIS = 0 in the DMA_SxFCR register), the packing/unpacking of data is
not possible. In this case, it is not allowed to have different source and destination transfer
data widths: both are equal and defined by the PSIZE bits in the DMA_SxCR MSIZE bits are
don’t care).
314/1751 RM0090 Rev 19

Table 46. Packing/unpacking & endian behavior (bit PINC = MINC = 1)

Number - Peripheral port address / byte lane
AHB Peripher
AHB of data Memory Memory port
memory al
peripheral items to transfer address / byte
port transfer
port width transfer - number lane PINCOS = 1 PINCOS = 0
width number
(NDT)
1 0x0 / B0[7:0] 1 0x0 / B0[7:0] 0x0 / B0[7:0]

2 0x1 / B1[7:0] 2 0x4 / B1[7:0] 0x1 / B1[7:0]
8 8 4 -
3 0x2 / B2[7:0] 3 0x8 / B2[7:0] 0x2 / B2[7:0]
4 0x3 / B3[7:0] 4 0xC / B3[7:0] 0x3 / B3[7:0]
1 0x0 / B0[7:0] 1 0x0 / B1|B0[15:0] 0x0 / B1|B0[15:0]

2 0x1 / B1[7:0]
8 16 2 -
3 0x2 / B2[7:0] 2 0x4 / B3|B2[15:0] 0x2 / B3|B2[15:0]
4 0x3 / B3[7:0]
1 0x0 / B0[7:0] 1 0x0 / B3|B2|B1|B0[31:0] 0x0 / B3|B2|B1|B0[31:0]

2 0x1 / B1[7:0]
8 32 1 -
3 0x2 / B2[7:0]
4 0x3 / B3[7:0]
1 0x0 / B1|B0[15:0] 1 0x0 / B0[7:0] 0x0 / B0[7:0]

2 0x4 / B1[7:0] 0x1 / B1[7:0]
16 8 4 -
2 0x2 / B3|B2[15:0] 3 0x8 / B2[7:0] 0x2 / B2[7:0]
4 0xC / B3[7:0] 0x3 / B3[7:0]
1 0x0 / B1|B0[15:0] 1 0x0 / B1|B0[15:0] 0x0 / B1|B0[15:0]

16 16 2 -
2 0x2 / B1|B0[15:0] 2 0x4 / B3|B2[15:0] 0x2 / B3|B2[15:0]
1 0x0 / B1|B0[15:0] 1 0x0 / B3|B2|B1|B0[31:0] 0x0 / B3|B2|B1|B0[31:0]

16 32 1 -
2 0x2 / B3|B2[15:0]
1 0x0 / B3|B2|B1|B0[31:0] 1 0x0 / B0[7:0] 0x0 / B0[7:0]

2 0x4 / B1[7:0] 0x1 / B1[7:0]
32 8 4 -
3 0x8 / B2[7:0] 0x2 / B2[7:0]
4 0xC / B3[7:0] 0x3 / B3[7:0]
1 0x0 /B3|B2|B1|B0[31:0] 1 0x0 / B1|B0[15:0] 0x0 / B1|B0[15:0]

32 16 2 -
2 0x4 / B3|B2[15:0] 0x2 / B3|B2[15:0]
32 32 1 -1 0x0 /B3|B2|B1|B0 [31:0] 1 0x0 /B3|B2|B1|B0 [31:0] 0x0 / B3|B2|B1|B0[31:0]
Note: Peripheral port may be the source or the destination (it could also be the memory source in
the case of memory-to-memory transfer).
PSIZE, MSIZE and NDT[15:0] have to be configured so as to ensure that the last transfer
will not be incomplete. This can occur when the data width of the peripheral port (PSIZE
bits) is lower than the data width of the memory port (MSIZE bits). This constraint is
summarized in Table 47.
RM0090 Rev 19 315/1751

338
Table 47. Restriction on NDT versus PSIZE and MSIZE

PSIZE[1:0] of DMA_SxCR MSIZE[1:0] of DMA_SxCR NDT[15:0] of DMA_SxNDTR
00 (8-bit) 01 (16-bit) must be a multiple of 2

10.3.11 Single and burst transfers

The DMA controller can generate single transfers or incremental burst transfers of 4, 8 or 16
beats.
The size of the burst is configured by software independently for the two AHB ports by using
the MBURST[1:0] and PBURST[1:0] bits in the DMA_SxCR register.
The burst size indicates the number of beats in the burst, not the number of bytes
transferred.
To ensure data coherence, each group of transfers that form a burst are indivisible: AHB
transfers are locked and the arbiter of the AHB bus matrix does not degrant the DMA master
during the sequence of the burst transfer.
Depending on the single or burst configuration, each DMA request initiates a different
number of transfers on the AHB peripheral port:
• When the AHB peripheral port is configured for single transfers, each DMA request
generates a data transfer of a byte, half-word or word depending on the PSIZE[1:0] bits
in the DMA_SxCR register
• When the AHB peripheral port is configured for burst transfers, each DMA request
generates 4,8 or 16 beats of byte, half word or word transfers depending on the
PBURST[1:0] and PSIZE[1:0] bits in the DMA_SxCR register.
The same as above has to be considered for the AHB memory port considering the
MBURST and MSIZE bits.
In direct mode, the stream can only generate single transfers and the MBURST[1:0] and
PBURST[1:0] bits are forced by hardware.
The address pointers (DMA_SxPAR or DMA_SxM0AR registers) must be chosen so as to
ensure that all transfers within a burst block are aligned on the address boundary equal to
the size of the transfer.
The burst configuration has to be selected in order to respect the AHB protocol, where
bursts must not cross the 1 KB address boundary because the minimum address space that
can be allocated to a single slave is 1 KB. This means that the 1 KB address boundary
should not be crossed by a burst block transfer, otherwise an AHB error would be
generated, that is not reported by the DMA registers.
316/1751 RM0090 Rev 19

10.3.12 FIFO
FIFO structure
The FIFO is used to temporarily store data coming from the source before transmitting them
to the destination.
Each stream has an independent 4-word FIFO and the threshold level is software-
configurable between 1/4, 1/2, 3/4 or full.
To enable the use of the FIFO threshold level, the direct mode must be disabled by setting
the DMDIS bit in the DMA_SxFCR register.
The structure of the FIFO differs depending on the source and destination data widths, and
is described in Figure 39: FIFO structure.
Figure 39. FIFO structure
4 words
Empty 1/4 1/2 3/4 Full

byte lane 3 B15 B 11 B7 B3
Source: byte Destination: word
byte lane 2 B14 B10 B6 B2
B15 B14 B13 B12 B11 B10 B9 B8 B7 B6 B5 B4 B3 B2 B1 B0 W3, W2, W1, W0
byte lane 0 W3 B12 W2

B8
W1
B4
W0
B0
4 words

Source: byte Destination: half-word
byte lane 2 H7 B14 H5 B10 H3 B6 H1 B2
B15 B14 B13 B12 B11 B10 B9 B8 B7 B6 B5 B4 B3 B2 B1 B0 H7, H6, H5, H4, H3, H2, H1, H0
4 words

byte lane 3
Source: half-word H7 H5 H3 H1 Destination: word
byte lane 2
H7 H6 H5 H4 H3 H2 H1 H0 W3, W2, W1, W0
byte lane 1
H6 H4 H2 H0
byte lane 0 W3 W2 W1 W0
4-words

Source: half-word Destination: byte
H7 H6 H5 H4 H3 H2 H1 H0 B15 B14 B13 B12 B11 B10 B9 B8
byte lane 1 B13 B9 B5 B1 B7 B6 B5 B4 B3 B2 B1 B0
ai15951
RM0090 Rev 19 317/1751

338
FIFO threshold and burst configuration

Caution is required when choosing the FIFO threshold (bits FTH[1:0] of the DMA_SxFCR
register) and the size of the memory burst (MBURST[1:0] of the DMA_SxCR register): The
content pointed by the FIFO threshold must exactly match to an integer number of memory
burst transfers. If this is not in the case, a FIFO error (flag FEIFx of the DMA_HISR or
DMA_LISR register) will be generated when the stream is enabled, then the stream will be
automatically disabled. The allowed and forbidden configurations are described in the
Table 48: FIFO threshold configurations.
Table 48. FIFO threshold configurations

MSIZE FIFO level MBURST = INCR4 MBURST = INCR8 MBURST = INCR16
1/4 1 burst of 4 beats forbidden

1/2 2 bursts of 4 beats 1 burst of 8 beats forbidden
Byte
3/4 3 bursts of 4 beats forbidden
Full 4 bursts of 4 beats 2 bursts of 8 beats 1 burst of 16 beats
1/4 forbidden
1/2 1 burst of 4 beats forbidden
Half-word
3/4 forbidden
Full 2 bursts of 4 beats 1 burst of 8 beats
forbidden
1/4
1/2 forbidden
Word forbidden
3/4
Full 1 burst of 4 beats
In all cases, the burst size multiplied by the data size must not exceed the FIFO size (data
size can be: 1 (byte), 2 (half-word) or 4 (word)).
Incomplete Burst transfer at the end of a DMA transfer may happen if one of the following
conditions occurs:
• For the AHB peripheral port configuration: the total number of data items (set in the
DMA_SxNDTR register) is not a multiple of the burst size multiplied by the data size
• For the AHB memory port configuration: the number of remaining data items in the
FIFO to be transferred to the memory is not a multiple of the burst size multiplied by the
data size
In such cases, the remaining data to be transferred will be managed in single mode by the
DMA, even if a burst transaction was requested during the DMA stream configuration.
Note: When burst transfers are requested on the peripheral AHB port and the FIFO is used
(DMDIS = 1 in the DMA_SxCR register), it is mandatory to respect the following rule to
avoid permanent underrun or overrun conditions, depending on the DMA stream direction:
If (PBURST × PSIZE) = FIFO_SIZE (4 words), FIFO_Threshold = 3/4 is forbidden with
PSIZE = 1, 2 or 4 and PBURST = 4, 8 or 16.
This rule ensures that enough FIFO space at a time will be free to serve the request from
the peripheral.
318/1751 RM0090 Rev 19

FIFO flush
The FIFO can be flushed when the stream is disabled by resetting the EN bit in the
DMA_SxCR register and when the stream is configured to manage peripheral-to-memory or
memory-to-memory transfers: If some data are still present in the FIFO when the stream is
disabled, the DMA controller continues transferring the remaining data to the destination
(even though stream is effectively disabled). When this flush is completed, the transfer
complete status bit (TCIFx) in the DMA_LISR or DMA_HISR register is set.
The remaining data counter DMA_SxNDTR keeps the value in this case to indicate how
many data items are currently available in the destination memory.
Note that during the FIFO flush operation, if the number of remaining data items in the FIFO
to be transferred to memory (in bytes) is less than the memory data width (for example 2
bytes in FIFO while MSIZE is configured to word), data will be sent with the data width set in
the MSIZE bit in the DMA_SxCR register. This means that memory will be written with an
undesired value. The software may read the DMA_SxNDTR register to determine the
memory area that contains the good data (start address and last address).
If the number of remaining data items in the FIFO is lower than a burst size (if the MBURST
bits in DMA_SxCR register are set to configure the stream to manage burst on the AHB
memory port), single transactions will be generated to complete the FIFO flush.
Direct mode
By default, the FIFO operates in direct mode (DMDIS bit in the DMA_SxFCR is reset) and
the FIFO threshold level is not used. This mode is useful when the system requires an
immediate and single transfer to or from the memory after each DMA request.
When the DMA is configured in direct mode (FIFO disabled), to transfer data in memory-to-
peripheral mode, the DMA preloads one data from the memory to the internal FIFO to
ensure an immediate data transfer as soon as a DMA request is triggered by a peripheral.
To avoid saturating the FIFO, it is recommended to configure the corresponding stream with
a high priority.
This mode is restricted to transfers where:
• The source and destination transfer widths are equal and both defined by the
PSIZE[1:0] bits in DMA_SxCR (MSIZE[1:0] bits are don’t care)
• Burst transfers are not possible (PBURST[1:0] and MBURST[1:0] bits in DMA_SxCR
are don’t care)
Direct mode must not be used when implementing memory-to-memory transfers.
10.3.13 DMA transfer completion

Different events can generate an end of transfer by setting the TCIFx bit in the DMA_LISR
or DMA_HISR status register:
• In DMA flow controller mode:
– The DMA_SxNDTR counter has reached zero in the memory-to-peripheral mode
– The stream is disabled before the end of transfer (by clearing the EN bit in the
DMA_SxCR register) and (when transfers are peripheral-to-memory or memory-
RM0090 Rev 19 319/1751

338
to-memory) all the remaining data have been flushed from the FIFO into the
memory
• In Peripheral flow controller mode:
– The last external burst or single request has been generated from the peripheral
and (when the DMA is operating in peripheral-to-memory mode) the remaining
data have been transferred from the FIFO into the memory
– The stream is disabled by software, and (when the DMA is operating in peripheral-
to-memory mode) the remaining data have been transferred from the FIFO into
the memory
Note: The transfer completion is dependent on the remaining data in FIFO to be transferred into
memory only in the case of peripheral-to-memory mode. This condition is not applicable in
memory-to-peripheral mode.
If the stream is configured in noncircular mode, after the end of the transfer (that is when the
number of data to be transferred reaches zero), the DMA is stopped (EN bit in DMA_SxCR
register is cleared by Hardware) and no DMA request is served unless the software
reprograms the stream and re-enables it (by setting the EN bit in the DMA_SxCR register).
10.3.14 DMA transfer suspension

At any time, a DMA transfer can be suspended to be restarted later on or to be definitively
disabled before the end of the DMA transfer.
There are two cases:
• The stream disables the transfer with no later-on restart from the point where it was
stopped. There is no particular action to do, except to clear the EN bit in the
DMA_SxCR register to disable the stream. The stream may take time to be disabled
(ongoing transfer is completed first). The transfer complete interrupt flag (TCIF in the
DMA_LISR or DMA_HISR register) is set in order to indicate the end of transfer. The
value of the EN bit in DMA_SxCR is now ‘0’ to confirm the stream interruption. The
DMA_SxNDTR register contains the number of remaining data items at the moment
when the stream was stopped so that the software can determine how many data items
have been transferred before the stream was interrupted.
• The stream suspends the transfer before the number of remaining data items to be
transferred in the DMA_SxNDTR register reaches 0. The aim is to restart the transfer
later by re-enabling the stream. In order to restart from the point where the transfer was
stopped, the software has to read the DMA_SxNDTR register after disabling the stream
by writing the EN bit in DMA_SxCR register (and then checking that it is at ‘0’) to know
the number of data items already collected. Then:
– The peripheral and/or memory addresses have to be updated in order to adjust
the address pointers
– The SxNDTR register has to be updated with the remaining number of data items
to be transferred (the value read when the stream was disabled)
– The stream may then be re-enabled to restart the transfer from the point it was
stopped
Note: Note that a Transfer complete interrupt flag (TCIF in DMA_LISR or DMA_HISR) is set to
indicate the end of transfer due to the stream interruption.
320/1751 RM0090 Rev 19

10.3.15 Flow controller

The entity that controls the number of data to be transferred is known as the flow controller.
This flow controller is configured independently for each stream using the PFCTRL bit in the
DMA_SxCR register.
The flow controller can be:
• The DMA controller: in this case, the number of data items to be transferred is
programmed by software into the DMA_SxNDTR register before the DMA stream is
enabled.
• The peripheral source or destination: this is the case when the number of data items to
be transferred is unknown. The peripheral indicates by hardware to the DMA controller
when the last data are being transferred. This feature is only supported for peripherals
which are able to signal the end of the transfer, that is:
– SDIO
When the peripheral flow controller is used for a given stream, the value written into the
DMA_SxNDTR has no effect on the DMA transfer. Actually, whatever the value written, it will
be forced by hardware to 0xFFFF as soon as the stream is enabled, to respect the following
schemes:
• Anticipated stream interruption: EN bit in DMA_SxCR register is reset to 0 by the
software to stop the stream before the last data hardware signal (single or burst) is sent
by the peripheral. In such a case, the stream is switched off and the FIFO flush is
triggered in the case of a peripheral-to-memory DMA transfer. The TCIFx flag of the
corresponding stream is set in the status register to indicate the DMA completion. To
know the number of data items transferred during the DMA transfer, read the
DMA_SxNDTR register and apply the following formula:
– Number_of_data_transferred = 0xFFFF – DMA_SxNDTR
• Normal stream interruption due to the reception of a last data hardware signal: the
stream is automatically interrupted when the peripheral requests the last transfer
(single or burst) and when this transfer is complete. the TCIFx flag of the corresponding
stream is set in the status register to indicate the DMA transfer completion. To know the
number of data items transferred, read the DMA_SxNDTR register and apply the same
formula as above.
• The DMA_SxNDTR register reaches 0: the TCIFx flag of the corresponding stream is
set in the status register to indicate the forced DMA transfer completion. The stream is
automatically switched off even though the last data hardware signal (single or burst)
has not been yet asserted. The already transferred data will not be lost. This means
that a maximum of 65535 data items can be managed by the DMA in a single
transaction, even in peripheral flow control mode.
Note: When configured in memory-to-memory mode, the DMA is always the flow controller and
the PFCTRL bit is forced to 0 by hardware.
The Circular mode is forbidden in the peripheral flow controller mode.
RM0090 Rev 19 321/1751

338
10.3.16 Summary of the possible DMA configurations

Table 49 summarizes the different possible DMA configurations.
Table 49. Possible DMA configurations

DMA transfer Flow Circular Transfer Direct Double
Source Destination
mode controller mode type mode buffer mode
single possible
DMA possible possible
Peripheral-to- AHB AHB burst forbidden
memory peripheral port memory port single possible
Peripheral forbidden forbidden
burst forbidden
single possible
DMA possible possible
Memory-to- AHB AHB burst forbidden
peripheral memory port peripheral port single possible
Peripheral forbidden forbidden
burst forbidden
Memory-to- AHB AHB single

DMA only forbidden forbidden forbidden
memory peripheral port memory port burst
10.3.17 Stream configuration procedure

The following sequence should be followed to configure a DMA stream x (where x is the
stream number):
1. If the stream is enabled, disable it by resetting the EN bit in the DMA_SxCR register,
then read this bit in order to confirm that there is no ongoing stream operation. Writing
this bit to 0 is not immediately effective since it is actually written to 0 once all the
current transfers have finished. When the EN bit is read as 0, this means that the
stream is ready to be configured. It is therefore necessary to wait for the EN bit to be
cleared before starting any stream configuration. All the stream dedicated bits set in the
status register (DMA_LISR and DMA_HISR) from the previous data block DMA
transfer should be cleared before the stream can be re-enabled.
2. Set the peripheral port register address in the DMA_SxPAR register. The data will be
moved from/ to this address to/ from the peripheral port after the peripheral event.
3. Set the memory address in the DMA_SxMA0R register (and in the DMA_SxMA1R
register in the case of a double buffer mode). The data will be written to or read from
this memory after the peripheral event.
4. Configure the total number of data items to be transferred in the DMA_SxNDTR
register. After each peripheral event or each beat of the burst, this value is
decremented.
5. Select the DMA channel (request) using CHSEL[2:0] in the DMA_SxCR register.
6. If the peripheral is intended to be the flow controller and if it supports this feature, set
the PFCTRL bit in the DMA_SxCR register.
7. Configure the stream priority using the PL[1:0] bits in the DMA_SxCR register.
8. Configure the FIFO usage (enable or disable, threshold in transmission and reception)
9. Configure the data transfer direction, peripheral and memory incremented/fixed mode,
single or burst transactions, peripheral and memory data widths, Circular mode,
322/1751 RM0090 Rev 19

Double buffer mode and interrupts after half and/or full transfer, and/or errors in the
DMA_SxCR register.
10. Activate the stream by setting the EN bit in the DMA_SxCR register.
As soon as the stream is enabled, it can serve any DMA request from the peripheral
connected to the stream.
Once half the data have been transferred on the AHB destination port, the half-transfer flag
(HTIF) is set and an interrupt is generated if the half-transfer interrupt enable bit (HTIE) is
set. At the end of the transfer, the transfer complete flag (TCIF) is set and an interrupt is
generated if the transfer complete interrupt enable bit (TCIE) is set.
Warning: To switch off a peripheral connected to a DMA stream

request, it is mandatory to, first, switch off the DMA stream to
which the peripheral is connected, then to wait for EN bit = 0.
Only then can the peripheral be safely disabled.
10.3.18 Error management

The DMA controller can detect the following errors:
• Transfer error: the transfer error interrupt flag (TEIFx) is set when:
– A bus error occurs during a DMA read or a write access
– A write access is requested by software on a memory address register in Double
buffer mode whereas the stream is enabled and the current target memory is the
one impacted by the write into the memory address register (refer to
Section 10.3.9: Double buffer mode)
• FIFO error: the FIFO error interrupt flag (FEIFx) is set if:
– A FIFO underrun condition is detected
– A FIFO overrun condition is detected (no detection in memory-to-memory mode
because requests and transfers are internally managed by the DMA)
– The stream is enabled while the FIFO threshold level is not compatible with the
size of the memory burst (refer to Table 48: FIFO threshold configurations)
• Direct mode error: the direct mode error interrupt flag (DMEIFx) can only be set in the
peripheral-to-memory mode while operating in direct mode and when the MINC bit in
the DMA_SxCR register is cleared. This flag is set when a DMA request occurs while
the previous data have not yet been fully transferred into the memory (because the
memory bus was not granted). In this case, the flag indicates that 2 data items were be
transferred successively to the same destination address, which could be an issue if
the destination is not able to manage this situation
In direct mode, the FIFO error flag can also be set under the following conditions:
• In the peripheral-to-memory mode, the FIFO can be saturated (overrun) if the memory
bus is not granted for several peripheral requests
• In the memory-to-peripheral mode, an underrun condition may occur if the memory bus
has not been granted before a peripheral request occurs
If the TEIFx or the FEIFx flag is set due to incompatibility between burst size and FIFO
threshold level, the faulty stream is automatically disabled through a hardware clear of its
EN bit in the corresponding stream configuration register (DMA_SxCR).
RM0090 Rev 19 323/1751

338
If the DMEIFx or the FEIFx flag is set due to an overrun or underrun condition, the faulty
stream is not automatically disabled and it is up to the software to disable or not the stream
by resetting the EN bit in the DMA_SxCR register. This is because there is no data loss
when this kind of errors occur.
When the stream's error interrupt flag (TEIF, FEIF, DMEIF) in the DMA_LISR or DMA_HISR
register is set, an interrupt is generated if the corresponding interrupt enable bit (TEIE,
FEIE, DMIE) in the DMA_SxCR or DMA_SxFCR register is set.
Note: When a FIFO overrun or underrun condition occurs, the data are not lost because the
peripheral request is not acknowledged by the stream until the overrun or underrun
condition is cleared. If this acknowledge takes too much time, the peripheral itself may
detect an overrun or underrun condition of its internal buffer and data might be lost.
10.4 DMA interrupts

For each DMA stream, an interrupt can be produced on the following events:
• Half-transfer reached
• Transfer complete
• Transfer error
• Fifo error (overrun, underrun or FIFO level error)
• Direct mode error
Separate interrupt enable control bits are available for flexibility as shown in Table 50.
Table 50. DMA interrupt requests

Half-transfer HTIF HTIE

Transfer complete TCIF TCIE
Transfer error TEIF TEIE
FIFO overrun/underrun FEIF FEIE
Direct mode error DMEIF DMEIE
Note: Before setting an Enable control bit to ‘1’, the corresponding event flag should be cleared,
otherwise an interrupt is immediately generated.
324/1751 RM0090 Rev 19

10.5 DMA registers

The DMA registers have to be accessed by words (32 bits).
10.5.1 DMA low interrupt status register (DMA_LISR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved TCIF3 HTIF3 TEIF3 DMEIF3 Reserv FEIF3 TCIF2 HTIF2 TEIF2 DMEIF2 Reserv FEIF2
r r r r r r r r ed r r r r r ed r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserved TCIF1 HTIF1 TEIF1 DMEIF1 Reserv FEIF1 TCIF0 HTIF0 TEIF0 DMEIF0 Reserv FEIF0
r r r r r r r r ed r r r r r ed r
Bits 31:28, 15:12 Reserved, must be kept at reset value.

Bits 27, 21, 11, 5 TCIFx: Stream x transfer complete interrupt flag (x = 3..0)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_LIFCR register.
0: No transfer complete event on stream x
1: A transfer complete event occurred on stream x
Bits 26, 20, 10, 4 HTIFx: Stream x half transfer interrupt flag (x=3..0)
DMA_LIFCR register.
0: No half transfer event on stream x
1: A half transfer event occurred on stream x
Bits 25, 19, 9, 3 TEIFx: Stream x transfer error interrupt flag (x=3..0)
DMA_LIFCR register.
0: No transfer error on stream x
1: A transfer error occurred on stream x
Bits 24, 18, 8, 2 DMEIFx: Stream x direct mode error interrupt flag (x=3..0)
DMA_LIFCR register.
0: No Direct Mode Error on stream x
1: A Direct Mode Error occurred on stream x
Bits 23, 17, 7, 1 Reserved, must be kept at reset value.
Bits 22, 16, 6, 0 FEIFx: Stream x FIFO error interrupt flag (x=3..0)
DMA_LIFCR register.
0: No FIFO Error event on stream x
1: A FIFO Error event occurred on stream x
RM0090 Rev 19 325/1751

338
10.5.2 DMA high interrupt status register (DMA_HISR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
TCIF7 HTIF7 TEIF7 DMEIF7 Reserv FEIF7 TCIF6 HTIF6 TEIF6 DMEIF6 Reserv FEIF6
Reserved
ed ed
r r r r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TCIF5 HTIF5 TEIF5 DMEIF5 Reserv FEIF5 TCIF4 HTIF4 TEIF4 DMEIF4 Reserv FEIF4
Reserved
ed ed
r r r r r r r r r r

Bits 27, 21, 11, 5 TCIFx: Stream x transfer complete interrupt flag (x=7..4)
DMA_HIFCR register.
0: No transfer complete event on stream x
1: A transfer complete event occurred on stream x
Bits 26, 20, 10, 4 HTIFx: Stream x half transfer interrupt flag (x=7..4)
DMA_HIFCR register.
0: No half transfer event on stream x
1: A half transfer event occurred on stream x
Bits 25, 19, 9, 3 TEIFx: Stream x transfer error interrupt flag (x=7..4)
DMA_HIFCR register.
0: No transfer error on stream x
1: A transfer error occurred on stream x
Bits 24, 18, 8, 2 DMEIFx: Stream x direct mode error interrupt flag (x=7..4)
DMA_HIFCR register.
0: No Direct mode error on stream x
1: A Direct mode error occurred on stream x
Bits 22, 16, 6, 0 FEIFx: Stream x FIFO error interrupt flag (x=7..4)
DMA_HIFCR register.
0: No FIFO error event on stream x
1: A FIFO error event occurred on stream x
326/1751 RM0090 Rev 19

10.5.3 DMA low interrupt flag clear register (DMA_LIFCR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CTCIF3 CHTIF3 CTEIF3 CDMEIF3 CFEIF3 CTCIF2 CHTIF2 CTEIF2 CDMEIF2 CFEIF2
Reserved Reserved Reserved
w w w w w w w w w w
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
w w w w w w w w w w

Bits 27, 21, 11, 5 CTCIFx: Stream x clear transfer complete interrupt flag (x = 3..0)
Writing 1 to this bit clears the corresponding TCIFx flag in the DMA_LISR register
Bits 26, 20, 10, 4 CHTIFx: Stream x clear half transfer interrupt flag (x = 3..0)
Writing 1 to this bit clears the corresponding HTIFx flag in the DMA_LISR register
Bits 25, 19, 9, 3 CTEIFx: Stream x clear transfer error interrupt flag (x = 3..0)
Writing 1 to this bit clears the corresponding TEIFx flag in the DMA_LISR register
Bits 24, 18, 8, 2 CDMEIFx: Stream x clear direct mode error interrupt flag (x = 3..0)
Writing 1 to this bit clears the corresponding DMEIFx flag in the DMA_LISR register
Bits 22, 16, 6, 0 CFEIFx: Stream x clear FIFO error interrupt flag (x = 3..0)
Writing 1 to this bit clears the corresponding CFEIFx flag in the DMA_LISR register
10.5.4 DMA high interrupt flag clear register (DMA_HIFCR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
w w w w w w w w w w
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
w w w w w w w w w w

Bits 27, 21, 11, 5 CTCIFx: Stream x clear transfer complete interrupt flag (x = 7..4)
Writing 1 to this bit clears the corresponding TCIFx flag in the DMA_HISR register
Bits 26, 20, 10, 4 CHTIFx: Stream x clear half transfer interrupt flag (x = 7..4)
Writing 1 to this bit clears the corresponding HTIFx flag in the DMA_HISR register
Bits 25, 19, 9, 3 CTEIFx: Stream x clear transfer error interrupt flag (x = 7..4)
Writing 1 to this bit clears the corresponding TEIFx flag in the DMA_HISR register
RM0090 Rev 19 327/1751

338
Bits 24, 18, 8, 2 CDMEIFx: Stream x clear direct mode error interrupt flag (x = 7..4)
Writing 1 to this bit clears the corresponding DMEIFx flag in the DMA_HISR register
Bits 22, 16, 6, 0 CFEIFx: Stream x clear FIFO error interrupt flag (x = 7..4)
Writing 1 to this bit clears the corresponding CFEIFx flag in the DMA_HISR register
10.5.5 DMA stream x configuration register (DMA_SxCR) (x = 0..7)

This register is used to configure the concerned stream.
Address offset: 0x10 + 0x18 × stream number
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CHSEL[2:0] MBURST [1:0] PBURST[1:0] Reser- CT DBM PL[1:0]
Reserved
rw rw rw rw rw rw rw ved rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PINCOS MSIZE[1:0] PSIZE[1:0] MINC PINC CIRC DIR[1:0] PFCTRL TCIE HTIE TEIE DMEIE EN

Bits 27:25 CHSEL[2:0]: Channel selection
These bits are set and cleared by software.
000: channel 0 selected
These bits are protected and can be written only if EN is ‘0’
Bits 24:23 MBURST: Memory burst transfer configuration
00: single transfer
01: INCR4 (incremental burst of 4 beats)
In direct mode, these bits are forced to 0x0 by hardware as soon as bit EN= '1'.
Bits 22:21 PBURST[1:0]: Peripheral burst transfer configuration
00: single transfer
In direct mode, these bits are forced to 0x0 by hardware.
328/1751 RM0090 Rev 19

Bit 19 CT: Current target (only in double buffer mode)

This bits is set and cleared by hardware. It can also be written by software.
0: The current target memory is Memory 0 (addressed by the DMA_SxM0AR pointer)
1: The current target memory is Memory 1 (addressed by the DMA_SxM1AR pointer)
This bit can be written only if EN is ‘0’ to indicate the target memory area of the first transfer.
Once the stream is enabled, this bit operates as a status flag indicating which memory area
is the current target.
Bit 18 DBM: Double buffer mode
This bits is set and cleared by software.
0: No buffer switching at the end of transfer
1: Memory target switched at the end of the DMA transfer
This bit is protected and can be written only if EN is ‘0’.
Bits 17:16 PL[1:0]: Priority level
00: Low
01: Medium
10: High
11: Very high
These bits are protected and can be written only if EN is ‘0’.
Bit 15 PINCOS: Peripheral increment offset size
This bit is set and cleared by software
0: The offset size for the peripheral address calculation is linked to the PSIZE
1: The offset size for the peripheral address calculation is fixed to 4 (32-bit alignment).
This bit has no meaning if bit PINC = '0'.
This bit is protected and can be written only if EN = '0'.
This bit is forced low by hardware when the stream is enabled (bit EN = '1') if the direct
mode is selected or if PBURST are different from “00”.
Bits 14:13 MSIZE[1:0]: Memory data size
00: byte (8-bit)
01: half-word (16-bit)
10: word (32-bit)
11: reserved
In direct mode, MSIZE is forced by hardware to the same value as PSIZE as soon as bit EN
= '1'.
Bits 12:11 PSIZE[1:0]: Peripheral data size
00: Byte (8-bit)
01: Half-word (16-bit)
10: Word (32-bit)
11: reserved
Bit 10 MINC: Memory increment mode
0: Memory address pointer is fixed
1: Memory address pointer is incremented after each data transfer (increment is done
according to MSIZE)
RM0090 Rev 19 329/1751

338
Bit 9 PINC: Peripheral increment mode

0: Peripheral address pointer is fixed
1: Peripheral address pointer is incremented after each data transfer (increment is done
according to PSIZE)
Bit 8 CIRC: Circular mode
This bit is set and cleared by software and can be cleared by hardware.
0: Circular mode disabled
1: Circular mode enabled
When the peripheral is the flow controller (bit PFCTRL=1) and the stream is enabled (bit
EN=1), then this bit is automatically forced by hardware to 0.
It is automatically forced by hardware to 1 if the DBM bit is set, as soon as the stream is
enabled (bit EN ='1').
Bits 7:6 DIR[1:0]: Data transfer direction
00: Peripheral-to-memory
01: Memory-to-peripheral
10: Memory-to-memory
11: reserved
Bit 5 PFCTRL: Peripheral flow controller
0: The DMA is the flow controller
1: The peripheral is the flow controller
When the memory-to-memory mode is selected (bits DIR[1:0]=10), then this bit is
automatically forced to 0 by hardware.
Bit 4 TCIE: Transfer complete interrupt enable
0: TC interrupt disabled
1: TC interrupt enabled
Bit 3 HTIE: Half transfer interrupt enable
0: HT interrupt disabled
1: HT interrupt enabled
Bit 2 TEIE: Transfer error interrupt enable
0: TE interrupt disabled
1: TE interrupt enabled
Bit 1 DMEIE: Direct mode error interrupt enable
0: DME interrupt disabled
1: DME interrupt enabled
330/1751 RM0090 Rev 19

Bit 0 EN: Stream enable / flag stream ready when read low
0: Stream disabled
1: Stream enabled
This bit may be cleared by hardware:
– on a DMA end of transfer (stream ready to be configured)
– if a transfer error occurs on the AHB master buses
– when the FIFO threshold on memory AHB port is not compatible with the size of the
burst
When this bit is read as 0, the software is allowed to program the Configuration and FIFO
bits registers. It is forbidden to write these registers when the EN bit is read as 1.
Note: Before setting EN bit to '1' to start a new transfer, the event flags corresponding to the
stream in DMA_LISR or DMA_HISR register must be cleared.
10.5.6 DMA stream x number of data register (DMA_SxNDTR) (x = 0..7)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
NDT[15:0]

Bits 15:0 NDT[15:0]: Number of data items to transfer
Number of data items to be transferred (0 up to 65535). This register can be written only
when the stream is disabled. When the stream is enabled, this register is read-only,
indicating the remaining data items to be transmitted. This register decrements after each
DMA transfer.
Once the transfer has completed, this register can either stay at zero (when the stream is in
normal mode) or be reloaded automatically with the previously programmed value in the
following cases:
– when the stream is configured in Circular mode.
– when the stream is enabled again by setting EN bit to '1'
If the value of this register is zero, no transaction can be served even if the stream is
enabled.
RM0090 Rev 19 331/1751

338
10.5.7 DMA stream x peripheral address register (DMA_SxPAR) (x = 0..7)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PAR[31:16]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PAR[15:0]
Bits 31:0 PAR[31:0]: Peripheral address

Base address of the peripheral data register from/to which the data will be read/written.
These bits are write-protected and can be written only when bit EN = '0' in the DMA_SxCR register.
10.5.8 DMA stream x memory 0 address register (DMA_SxM0AR) (x = 0..7)

Address offset: 0x1C + 0x18 × stream number
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
M0A[31:16]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
M0A[15:0]
Bits 31:0 M0A[31:0]: Memory 0 address

Base address of Memory area 0 from/to which the data will be read/written.
These bits are write-protected. They can be written only if:
– the stream is disabled (bit EN= '0' in the DMA_SxCR register) or
– the stream is enabled (EN=’1’ in DMA_SxCR register) and bit CT = '1' in the
DMA_SxCR register (in Double buffer mode).
10.5.9 DMA stream x memory 1 address register (DMA_SxM1AR) (x = 0..7)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
M1A[31:16]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
M1A[15:0]
332/1751 RM0090 Rev 19

Bits 31:0 M1A[31:0]: Memory 1 address (used in case of Double buffer mode)
Base address of Memory area 1 from/to which the data will be read/written.
This register is used only for the Double buffer mode.
These bits are write-protected. They can be written only if:
– the stream is disabled (bit EN= '0' in the DMA_SxCR register) or
– the stream is enabled (EN=’1’ in DMA_SxCR register) and bit CT = '0' in the
DMA_SxCR register.
10.5.10 DMA stream x FIFO control register (DMA_SxFCR) (x = 0..7)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FEIE Reser FS[2:0] DMDIS FTH[1:0]
Reserved
rw ved r r r rw rw rw

Bit 7 FEIE: FIFO error interrupt enable
0: FE interrupt disabled
1: FE interrupt enabled
RM0090 Rev 19 333/1751

338
Bits 5:3 FS[2:0]: FIFO status

These bits are read-only.
000: 0 < fifo_level < 1/4
001: 1/4 ≤ fifo_level < 1/2
010: 1/2 ≤ fifo_level < 3/4
011: 3/4 ≤ fifo_level < full
100: FIFO is empty
101: FIFO is full
others: no meaning
These bits are not relevant in the direct mode (DMDIS bit is zero).
Bit 2 DMDIS: Direct mode disable
This bit is set and cleared by software. It can be set by hardware.
0: Direct mode enabled
1: Direct mode disabled
This bit is set by hardware if the memory-to-memory mode is selected (DIR bit in
DMA_SxCR are “10”) and the EN bit in the DMA_SxCR register is ‘1’ because the direct
mode is not allowed in the memory-to-memory configuration.
Bits 1:0 FTH[1:0]: FIFO threshold selection
00: 1/4 full FIFO
01: 1/2 full FIFO
10: 3/4 full FIFO
11: full FIFO
These bits are not used in the direct mode when the DMIS value is zero.
334/1751 RM0090 Rev 19

10.5.11 DMA register map

Table 51 summarizes the DMA registers.
Table 51. DMA register map and reset values

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
DMEIF3
DMEIF2
DMEIF1
DMEIF0
TCIF3
HTIF3
TCIF2
HTIF2
TCIF1
HTIF1
TCIF0
HTIF0
TEIF3
FEIF3
TEIF2
FEIF2
TEIF1
FEIF1
TEIF0
FEIF0
Reserved
Reserved
Reserved
Reserved
DMA_LISR
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMEIF7
DMEIF6
DMEIF5
DMEIF4
TCIF7
HTIF7
TCIF6
HTIF6
TCIF5
HTIF5
TCIF4
HTIF4
TEIF7
FEIF7
TEIF6
FEIF6
TEIF5
FEIF5
TEIF4
FEIF4
Reserved
Reserved
Reserved
Reserved
DMA_HISR
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CDMEIF3
CDMEIF2
CDMEIF1
CDMEIF0
Reserved Reserved
Reserved Reserved
Reserved Reserved
CTCIF3
CHTIF3
CTCIF2
CHTIF2
CTCIF1
CHTIF1
CTCIF0
CHTIF0
CFEIF3
CTEIF2
CFEIF2
CTEIF1
CFEIF1
CTEIF0
CFEIF0
TEIF3
DMA_LIFCR
Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CDMEIF7
CDMEIF6
CDMEIF5
CDMEIF4
CTCIF7
CHTIF7
CTCIF6
CHTIF6
CTCIF5
CHTIF5
CTCIF4
CHTIF4
CTEIF7
CFEIF7
CTEIF6
CFEIF6
CTEIF5
CFEIF5
CTEIF4
CFEIF4
Reserved
DMA_HIFCR - - -
Reset value 0 0 0 0 0 0 0 0 0 - 0 0 0 0 0 - 0 0 0 0 0 - 0
MBURST[1:0]
PBURST[1:0]
CHSEL[2:0]
MSIZE[1:0]
PSIZE[1:0]
PFCTRL
PINCOS
DIR[1:0]
PL[1:0]
DMEIE
Reserved
MINC
CIRC
PINC
HTIE
TCIE
DBM
TEIE
DMA_S0CR
EN
CT
0x0010 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA_S0NDTR NDT[15:.]
0x0014 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA_S0PAR PA[31:0]
0x0018
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA_S0M0AR M0A[31:0]
0x001C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x0020
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMDIS
FTH
Reserved
FEIE
DMA_S0FCR FS[2:0]
0x0024 Reserved [1:0]
PBURST[1:0]
MBURST[1:]
MSIZE[1:0]
PSIZE[1:0]
PFCTRL
PINCOS
DIR[1:0]
CHSEL
PL[1:0]
DMEIE
Reserved
MINC
CIRC
PINC
TCIE
HTIE
DBM
TEIE
[2:0]
DMA_S1CR
EN
CT
0x0028 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x002C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RM0090 Rev 19 335/1751

338
Table 51. DMA register map and reset values (continued)

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
DMA_S1PAR PA[31:0]
0x0030
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x0034
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x0038
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMDIS
FTH
Reserved
FEIE
DMA_S1FCR FS[2:0]
0x003C Reserved [1:0]
MBURST[1:0]
PBURST[1:0]
MSIZE[1:0]
PSIZE[1:0]
PFCTRL
PINCOS
CHSEL
PL[1:0]
DMEIE
MINC
Reserved
CIRC
PINC
HTIE
TCIE
TEIE
DBM
[2:0]
[1:0]
DIR
DMA_S2CR
EN
CT
0x0040 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x0044 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA_S2PAR PA[31:0]
0x0048
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x004C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x0050
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMDIS
FTH
Reserved
FEIE
DMA_S2FCR FS[2:0]
MBURST[1:0]
PBURST[1:0]
CHSEL[2:0]
MSIZE[1:0]
PSIZE[1:0]
PFCTRL
PINCOS
DIR[1:0]
PL[1:0]
DMEIE
Reserved
MINC
CIRC
PINC
TCIE
HTIE
DBM
TEIE
DMA_S3CR
EN
CT
0x0058 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x005C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA_S3PAR PA[31:0]
0x0060
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x0064
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
336/1751 RM0090 Rev 19


Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
0x0068
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMDIS
FTH
Reserved
FEIE
DMA_S3FCR FS[2:0]
MBURST[1:0]
PBURST[1:0]
CHSEL[2:0]
MSIZE[1:0]
PSIZE[1:0]
PFCTRL
PINCOS
PL[1:0]
DMEIE
Reserved
MINC
CIRC
PINC
TCIE
HTIE
DBM
TEIE
[1:0]
DIR
DMA_S4CR
EN
CT
0x0070 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x0074 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA_S4PAR PA[31:0]
0x0078
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x007C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x0080
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMDIS
FTH
Reserved
FEIE
DMA_S4FCR FS[2:0]
MBURST[1:0]
PBURST[1:0]
CHSEL[2:0]
MSIZE[1:0]
PSIZE[1:0]
PFCTRL
PINCOS
DIR[1:0]
PL[1:0]
DMEIE
Reserved
MINC
CIRC
PINC
TCIE
HTIE
DBM
TEIE
DMA_S5CR
EN
CT
0x0088 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x008C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA_S5PAR PA[31:0]
0x0090
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x0094
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x0098
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMDIS
FTH
Reserved
FEIE
DMA_S5FCR FS[2:0]
MBURST[1:0]
PBURST[1:0]
CHSEL[2:0]
MSIZE[1:0]
PSIZE[1:0]
PFCTRL
PINCOS
DIR[1:0]
PL[1:0]
DMEIE
Reserved
MINC
CIRC
PINC
HTIE
TCIE
DBM
TEIE
DMA_S6CR
EN
CT
0x00A0 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x00A4 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA_S6PAR PA[31:0]
0x00A8
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RM0090 Rev 19 337/1751

338

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
0x00AC
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x00B0
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMDIS
FTH
Reserved
FEIE
DMA_S6FCR FS[2:0]
0x00B4 Reserved [1:0]
MBURST[1:0]
PBURST[1:0]
CHSEL[2:0]
MSIZE[1:0]
PSIZE[1:0]
PFCTRL
PINCOS
DIR[1:0]
PL[1:0]
DMEIE
MINC
CIRC
PINC
Reserved
TCIE
HTIE
DBM
TEIE
DMA_S7CR
EN
CT
0x00B8 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 - 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x00BC Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA_S7PAR PA[31:0]
0x00C0
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x00C4
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x00C8
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMDIS
FTH
Reserved
FEIE
DMA_S7FCR FS[2:0]
0x00CC Reserved [1:0]
338/1751 RM0090 Rev 19

RM0090 Chrom-Art Accelerator™ controller (DMA2D)
11 Chrom-Art Accelerator™ controller (DMA2D)
11.1 DMA2D introduction

The Chrom-Art Accelerator™ (DMA2D) is a specialized DMA dedicated to image
manipulation. It can perform the following operations:
• Filling a part or the whole of a destination image with a specific color
• Copying a part or the whole of a source image into a part or the whole of a destination
image
• Copying a part or the whole of a source image into a part or the whole of a destination
image with a pixel format conversion
• Blending a part and/or two complete source images with different pixel format and copy
the result into a part or the whole of a destination image with a different color format.
All the classical color coding schemes are supported from 4-bit up to 32-bit per pixel with
indexed or direct color mode. The DMA2D has its own dedicated memories for CLUTs (color
look-up tables).
RM0090 Rev 19 339/1751

370
Chrom-Art Accelerator™ controller (DMA2D) RM0090
11.2 DMA2D main features

The main DMA2D features are:
• Single AHB master bus architecture.
• AHB slave programming interface supporting 8/16/32-bit accesses (except for CLUT
accesses which are 32-bit).
• User programmable working area size
• User programmable offset for sources and destination areas
• User programmable sources and destination addresses on the whole memory space
• Up to 2 sources with blending operation
• Alpha value can be modified (source value, fixed value or modulated value)
• User programmable source and destination color format
• Up to 11 color formats supported from 4-bit up to 32-bit per pixel with indirect or direct
color coding
• 2 internal memories for CLUT storage in indirect color mode
• Automatic CLUT loading or CLUT programming via the CPU
• User programmable CLUT size
• Internal timer to control AHB bandwidth
• 4 operating modes: register-to-memory, memory-to-memory, memory-to-memory with
pixel format conversion, and memory-to-memory with pixel format conversion and
blending
• Area filling with a fixed color
• Copy from an area to another
• Copy with pixel format conversion between source and destination images
• Copy from two sources with independent color format and blending
• Abort and suspend of DMA2D operations
• Watermark interrupt on a user programmable destination line
• Interrupt generation on bus error or access conflict
• Interrupt generation on process completion
11.3 DMA2D functional description

The DMA2D controller performs direct memory transfer. As an AHB master, it can take the
control of the AHB bus matrix to initiate AHB transactions.
The DMA2D can operate in the following modes:
• Register-to-memory
• Memory-to-memory
• Memory-to-memory with Pixel Format Conversion
• Memory-to-memory with Pixel Format Conversion and Blending
The AHB slave port is used to program the DMA2D controller.
The block diagram of the DMA2D is shown in Figure 40: DMA2D block diagram.
340/1751 RM0090 Rev 19

Figure 40. DMA2D block diagram
AHB MASTER
FG PFC
α mode
Color mode 8-bit α

α 32
FG 32 α 8 X
Expander
FIFO
32 RGB BLENDER OUT PFC
Color
CLUT itf 32 α Color mode
32
Red Converter OUT
256x32-bit FIFO
Green 32/24/16
RAM 32
Blue
BG PFC
α mode
Color mode 8-bit α

α 32
BG 32 α 8 X
Expander
FIFO
32 RGB
CLUT itf
256x32-bit
RAM AHB SLAVE
MS30439V1
11.3.2 DMA2D control

The DMA2D controller is configured through the DMA2D Control Register (DMA2D_CR)
which allows selecting:
The user application can perform the following operations:
• Select the operating mode
• Enable/disable the DMA2D interrupt
• Start/suspend/abort ongoing data transfers
11.3.3 DMA2D foreground and background FIFOs

The DMA2D foreground (FG) FG FIFO and background (BG) FIFO fetch the input data to
be copied and/or processed.
The FIFOs fetch the pixels according to the color format defined in their respective pixel
format converter (PFC).
RM0090 Rev 19 341/1751

370
They are programmed through a set of control registers:

• DMA2D foreground memory address register (DMA2D_FGMAR)
• DMA2D foreground offset register (DMA2D_FGOR)
• DMA2D background memory address register (DMA2D_BGMAR)
• DMA2D background offset register (DMA2D_BGBOR)
• DMA2D number of lines register (number of lines and pixel per lines) (DMA2D_NLR)
When the DMA2D operates in register-to-memory mode, none of the FIFOs is activated.
When the DMA2D operates in memory-to-memory mode (no pixel format conversion nor
blending operation), only the FG FIFO is activated and acts as a buffer.
When the DMA2D operates in memory-to-memory operation with pixel format conversion
(no blending operation), the BG FIFO is not activated.
11.3.4 DMA2D foreground and background pixel format converter (PFC)

DMA2D foreground pixel format converter (PFC) and background pixel format converter
perform the pixel format conversion to generate a 32-bit per pixel value. The PFC can also
modify the alpha channel.
The first stage of the converter converts the color format. The original color format of the
foreground pixel and background pixels are configured through the CM[3:0] bits of the
DMA2D_FGPFCCR and DMA2D_BGPFCCR, respectively.
The supported input formats are given in Table 52: Supported color mode in input.
Table 52. Supported color mode in input

CM[3:0] Color mode
0000 ARGB8888
0001 RGB888
0010 RGB565
0011 ARGB1555
0100 ARGB4444
0101 L8
0110 AL44
0111 AL88
1000 L4
1001 A8
1010 A4
342/1751 RM0090 Rev 19

The color format are coded as follows:

• Alpha value field: transparency
0xFF value corresponds to an opaque pixel and 0x00 to a transparent one.
• R field for Red
• G field for Green
• B field for Blue
• L field: luminance
This field is the index to a CLUT to retrieve the three/four RGB/ARGB components.
If the original format was direct color mode, then the extension to 8-bit per channel is
performed by copying the MSBs into the LSBs. This ensures a perfect linearity of the
conversion.
If the original format does not include an alpha channel, the alpha value is automatically set
to 0xFF (opaque).
If the original format is indirect color mode, a CLUT is required and each pixel format
converter is associated with a 256 entry 32-bit CLUT.
For the specific alpha mode A4 and A8, no color information is stored nor indexed. The color
to be used for the image generation is fixed and is defined in the DMA2D_FGCOLR for
foreground pixels and in the DMA2D_BGCOLR register for background pixels.
The order of the fields in the system memory is defined in Table 53: Data order in memory.
Table 53. Data order in memory

Color Mode @+3 @+2 @+1 @+0
ARGB8888 A0[7:0] R0[7:0] G0[7:0] B0[7:0]

B1[7:0] R0[7:0] G0[7:0] B0[7:0]
RGB888 G2[7:0] B2[7:0] R1[7:0] G1[7:0]
R3[7:0] G3[7:0] B3[7:0] R2[7:0]
RGB565 R1[4:0]G1[5:3] G1[2:0]B1[4:0] R0[4:0]G0[5:3] G0[2:0]B0[4:0]
ARGB1555 A1[0]R1[4:0]G1[4:3] G1[2:0]B1[4:0] A0[0]R0[4:0]G0[4:3] G0[2:0]B0[4:0]
ARGB4444 A1[3:0]R1[3:0] G1[3:0]B1[3:0] A0[3:0]R0[3:0] G0[3:0]B0[3:0]
L8 L3[7:0] L2[7:0] L1[7:0] L0[7:0]
AL44 A3[3:0]L3[3:0] A2[3:0]L2[3:0] A1[3:0]L1[3:0] A0[3:0]L0[3:0]
AL88 A1[7:0] L1[7:0] A0[7:0] L0[7:0]
L4 L7[3:0]L6[3:0] L5[3:0]L4[3:0] L3[3:0]L2[3:0] L1[3:0]L0[3:0]
A8 A3[7:0] A2[7:0] A1[7:0] A0[7:0]
A4 A7[3:0]A6[3:0] A5[3:0]A4[3:0] A3[3:0]A2[3:0] A1[3:0]A0[3:0]
The 24-bit RGB888 aligned on 32-bit is supported through the ARGB8888 mode.
Once the 32-bit value is generated, the alpha channel can be modified according to the
AM[1:0] field of the DMA2D_FGPFCCR/DMA2D_BGPFCCR registers as shown in
Table 54: Alpha mode configuration.
RM0090 Rev 19 343/1751

370
The alpha channel can be:

• kept as it is (no modification),
• replaced by the ALPHA[7:0] value of DMA2D_FGPFCCR/DMA2D_BGPFCCR,
• or replaced by the original alpha value multiplied by the ALPHA[7:0] value of
DMA2D_FGPFCCR/DMA2D_BGPFCCR divided by 255.
Table 54. Alpha mode configuration

AM[1:0] Alpha mode
00 No modification
01 Replaced by value in DMA2D_xxPFCCR
10 Replaced by original value multiplied by the value in DMA2D_xxPFCCR / 255
11 Reserved
11.3.5 DMA2D foreground and background CLUT interface

The CLUT interface manages the CLUT memory access and the automatic loading of the
CLUT.
Three kinds of accesses are possible:
• CLUT read by the PFC during pixel format conversion operation
• CLUT accessed through the AHB slave port when the CPU is reading or writing data
into the CLUT
• CLUT written through the AHB master port when an automatic loading of the CLUT is
performed
The CLUT memory loading can be done in two different ways:
• Automatic loading
The following sequence should be followed to load the CLUT:
a) Program the CLUT address into the DMA2D_FGCMAR register (foreground
CLUT) or DMA2D_BGCMAR register (background CLUT)
b) Program the CLUT size in the CS[7:0] field of the DMA2D_FGPFCCR register
(foreground CLUT) or DMA2D_BGPFCCR register (background CLUT).
c) Set the START bit of the DMA2D_FGPFCCR register (foreground CLUT) or
DMA2D_BGPFCCR register (background CLUT) to start the transfer. During this
automatic loading process, the CLUT is not accessible by the CPU. If a conflict
occurs, a CLUT access error interrupt is raised assuming CAEIE is set to ‘1’ in
DMA2D_CR.
• Manual loading
The application has to program the CLUT manually through the DMA2D AHB slave
port to which the local CLUT memory is mapped.The foreground CLUT is located at
address offset 0x0400 and the background CLUT at address offset 0x0800.
The CLUT format can be 24 or 32 bits. It is configured through the CCM bit of the
DMA2D_FGPFCCR register (foreground CLUT) or DMA2D_BGPFCCR register
(background CLUT) as shown in Table 55: Supported CLUT color mode.
344/1751 RM0090 Rev 19

Table 55. Supported CLUT color mode

CCM CLUT color mode
0 32-bit ARGB8888
1 24-bit RGB888
The way the CLUT data are organized in the system memory is specified in Table 56: CLUT
data order in memory.
Table 56. CLUT data order in memory

CLUT Color Mode @+3 @+2 @+1 @+0
ARGB8888 A0[7:0] R0[7:0] G0[7:0] B0[7:0]
B1[7:0] R0[7:0] G0[7:0] B0[7:0]
RGB888 G2[7:0] B2[7:0] R1[7:0] G1[7:0]
R3[7:0] G3[7:0] B3[7:0] R2[7:0]
11.3.6 DMA2D blender

The DMA2D blender blends the source pixels by pair to compute the resulting pixel.
The blending is performed according to the following equation:
αFG . αBG
with αMult =
255
αOUT = αFG + αBG - αMult
CFG.αFG + CBG.αBG - CBG.αMult

COUT = with C = R or G or B
αOUT
Division is rounded to the nearest lower integer
No configuration register is required by the blender. The blender usage depends on the
DMA2D operating mode defined in MODE[1:0] field of the DMA2D_CR register.
11.3.7 DMA2D output PFC

The output PFC performs the pixel format conversion from 32 bits to the output format
defined in the CM[2:0] field of the DMA2D output pixel format converter configuration
register (DMA2D_OPFCCR).
The supported output formats are given in Table 57: Supported color mode in output
RM0090 Rev 19 345/1751

370
Table 57. Supported color mode in output

CM[2:0] Color mode
000 ARGB8888
001 RGB888
010 RGB565
011 ARGB1555
100 ARGB4444
11.3.8 DMA2D output FIFO

The output FIFO programs the pixels according to the color format defined in the output
PFC.
The destination area is defined through a set of control registers:
• DMA2D output memory address register (DMA2D_OMAR)
• DMA2D output offset register (DMA2D_OOR)
• DMA2D number of lines register (number of lines and pixel per lines) (DMA2D_NLR)
If the DMA2D operates in register-to-memory mode, the configured output rectangle is filled
by the color specified in the DMA2D output color register (DMA2D_OCOLR) which contains
a fixed 32-bit, 24-bit or 16-bit value. The format is selected by the CM[2:0] field of the
DMA2D_OPFCCR register.
The data are stored into the memory in the order defined in Table 58: Data order in memory
Table 58. Data order in memory

Color Mode @+3 @+2 @+1 @+0
ARGB8888 A0[7:0] R0[7:0] G0[7:0] B0[7:0]

B1[7:0] R0[7:0] G0[7:0] B0[7:0]
RGB888 G2[7:0] B2[7:0] R1[7:0] G1[7:0]
R3[7:0] G3[7:0] B3[7:0] R2[7:0]
RGB565 R1[4:0]G1[5:3] G1[2:0]B1[4:0] R0[4:0]G0[5:3] G0[2:0]B0[4:0]
ARGB1555 A1[0]R1[4:0]G1[4:3] G1[2:0]B1[4:0] A0[0]R0[4:0]G0[4:3] G0[2:0]B0[4:0]
ARGB4444 A1[3:0]R1[3:0] G1[3:0]B1[3:0] A0[3:0]R0[3:0] G0[3:0]B0[3:0]
The RGB888 aligned on 32-bit is supported through the ARGB8888 mode.
11.3.9 DMA2D AHB master port timer

An 8-bit timer is embedded into the AHB master port to provide an optional limitation of the
bandwidth on the crossbar.
This timer is clocked by the AHB clock and counts a dead time between two consecutive
accesses. This limits the bandwidth usage.
346/1751 RM0090 Rev 19

The timer enabling and the dead time value are configured through the AHB master port
timer configuration register (DMA2D_AMPTCR).
11.3.10 DMA2D transactions

DMA2D transactions consist of a sequence of a given number of data transfers. The
number of data and the width can be programmed by software.
Each DMA2D data transfer is composed of up to 4 steps:
1. Data loading from the memory location pointed by the DMA2D_FGMAR register and
pixel format conversion as defined in DMA2D_FGCR.
2. Data loading from a memory location pointed by the DMA2D_BGMAR register and
pixel format conversion as defined in DMA2D_BGCR.
3. Blending of all retrieved pixels according to the alpha channels resulting of the PFC
operation on alpha values.
4. Pixel format conversion of the resulting pixels according to the DMA2D_OCR register
and programming of the data to the memory location addressed through the
DMA2D_OMAR register.
11.3.11 DMA2D configuration

Both source and destination data transfers can target peripherals and memories in the
whole 4 Gbyte memory area, at addresses ranging between 0x0000 0000 and
0xFFFF FFFF.
The DMA2D can operate in any of the four following modes selected through MODE[1:0]
bits of the DMA2D_CR register:
• Register-to-memory
• Memory-to-memory
• Memory-to-memory with PFC
• Memory-to-memory with PFC and blending
Register-to-memory
The register-to-memory mode is used to fill a user defined area with a predefined color.
The color format is set in the DMA2D_OPFCCR.
The DMA2D does not perform any data fetching from any source. It just writes the color
defined in the DMA2D_OCOLR register to the area located at the address pointed by the
DMA2D_OMAR and defined in the DMA2D_NLR and DMA2D_OOR.
Memory-to-memory
In memory-to-memory mode, the DMA2D does not perform any graphical data
transformation. The foreground input FIFO acts as a buffer and the data are transferred
from the source memory location defined in DMA2D_FGMAR to the destination memory
location pointed by DMA2D_OMAR.
The color mode programmed in the CM[3:0] bits of the DMA2D_FGPFCCR register defines
the number of bits per pixel for both input and output.
RM0090 Rev 19 347/1751

370
The size of the area to be transferred is defined by the DMA2D_NLR and DMA2D_FGOR
registers for the source, and by DMA2D_NLR and DMA2D_OOR registers for the
destination.
Memory-to-memory with PFC

In this mode, the DMA2D performs a pixel format conversion of the source data and stores
them in the destination memory location.
The size of the areas to be transferred are defined by the DMA2D_NLR and DMA2D_FGOR
registers for the source, and by DMA2D_NLR and DMA2D_OOR registers for the
destination.
Data are fetched from the location defined in the DMA2D_FGMAR register and processed
by the foreground PFC. The original pixel format is configured through the
DMA2D_FGPFCCR register.
If the original pixel format is direct color mode, then the color channels are all expanded to 8
bits.
If the pixel format is indirect color mode, the associated CLUT has to be loaded into the
CLUT memory.
The CLUT loading can be done automatically by following the sequence below:
1. Set the CLUT address into the DMA2D_FGCMAR.
2. Set the CLUT size in the CS[7:0] bits of the DMA2D_FGPFCCR register.
3. Set the CLUT format (24 or 32 bits) in the CCM bit of the DMA2D_FGPFCCR register.
4. Start the CLUT loading by setting the START bit of the DMA2D_FGPFCCR register.
Once the CLUT loading is complete, the CTCIF flag of the DMA2D_IFR register is raised,
and an interrupt is generated if the CTCIE bit is set in DMA2D_CR. The automatic CLUT
loading process can not work in parallel with classical DMA2D transfers.
The CLUT can also be filled by the CPU or by any other master through the APB port. The
access to the CLUT is not possible when a DMA2D transfer is ongoing and uses the CLUT
(indirect color format).
In parallel to the color conversion process, the alpha value can be added or changed
depending on the value programmed in the DMA2D_FGPFCCR register. If the original
image does not have an alpha channel, a default alpha value of 0xFF is automatically added
to obtain a fully opaque pixel. The alpha value can be modified according to the AM[1:0] bits
of the DMA2D_FGPFCCR register:
• It can be unchanged.
• It can be replaced by the value defined in the ALPHA[7:0] value of the
DMA2D_FGPFCCR register.
• It can be replaced by the original value multiplied by the ALPHA[7:0] value of the
DMA2D_FGPFCCR register divided by 255.
The resulting 32-bit data are encoded by the OUT PFC into the format specified by the
CM[2:0] field of the DMA2D_OPFCCR register. The output pixel format cannot be the
indirect mode since no CLUT generation process is supported.
The processed data are written into the destination memory location pointed by
DMA2D_OMAR.
348/1751 RM0090 Rev 19

Memory-to-memory with PFC and blending

In this mode, 2 sources are fetched in the foreground FIFO and background FIFO from the
memory locations defined by DMA2D_FGMAR and DMA2D_BGMAR.
The two pixel format converters have to be configured as described in the memory-to-
memory mode. Their configurations can be different as each pixel format converter are
independent and have their own CLUT memory.
Once each pixel has been converted into 32 bits by their respective PFCs, they are blended
according to the equation below:
αFG . αBG
with αMult =
255
αOUT = αFG + αBG - αMult
CFG.αFG + CBG.αBG - CBG.αMult

COUT = with C = R or G or B
αOUT
Division are rounded to the nearest lower integer
The resulting 32-bit pixel value is encoded by the output PFC according to the specified
output format, and the data are written into the destination memory location pointed by
DMA2D_OMAR.
Configuration error detection

The DMA2D checks that the configuration is correct before any transfer. The configuration
error interrupt flag is set by hardware when a wrong configuration is detected when a new
transfer/automatic loading starts. An interrupt is then generated if the CEIE bit of the
DMA2D_CR is set.
RM0090 Rev 19 349/1751

370
The wrong configurations that can be detected are listed below:

• Foreground CLUT automatic loading: MA bits of DMA2D_FGCMAR not aligned with
CCM of DMA2D_FGPFCCR.
• Background CLUT automatic loading: MA of DMA2D_BGCMAR not aligned with CCM
of DMA2D_BGPFCCR
• Memory transfer (except in register-to-memory mode): MA of DMA2D_FGMAR not
aligned with CM of DMA2D_FGPFCCR
• Memory transfer (except in register-to-memory mode): CM in DMA2D_FGPFCCR are
invalid
• Memory transfer (except in register-to-memory mode): PL bits of DMA2D_NLR odd
while CM of DMA2D_FGPFCCR is A4 or L4
• Memory transfer (except in register-to-memory mode): LO bits in DMA2D_FGOR odd
while CM of DMA2D_FGPFCCR is A4 or L4
• Memory transfer (only in blending mode): MA bits in DMA2D_BGMAR are not aligned
with the CM of DMA2D_BGPFCCR
• Memory transfer: CM of DMA2D_BGPFCCR invalid (only in blending mode)
• Memory transfer (only in blending mode): PL bits of DMA2D_NLR odd while CM of
DMA2D_BGPFCCR is A4 or L4
• Memory transfer (only in blending mode): LO bits of DMA2D_BGOR odd while CM of
DMA2D_BGPFCCR is A4 or L4
• Memory transfer (except in memory to memory mode): MA bits in DMA2D_OMAR are
not aligned with CM bits in DMA2D_OPFCCR.
• Memory transfer (except in memory to memory mode): CM bits in DMA2D_OPFCCR
invalid
• Memory transfer: NL bits in DMA2D_NLR = 0
• Memory transfer: PL bits in DMA2D_NLR = 0
11.3.12 DMA2D transfer control (start, suspend, abort and completion)

Once the DMA2D is configured, the transfer can be launched by setting the START bit of the
DMA2D_CR register. Once the transfer is completed, the START bit is automatically reset
and the TCIF flag of the DMA2D_ISR register is raised. An interrupt can be generated if the
TCIE bit of the DMA2D_CR is set.
The user application can suspend the DMA2D at any time by setting the SUSP bit of the
DMA2D_CR register. The transaction can then be aborted by setting the ABORT bit of the
DMA2D_CR register or can be restarted by resetting the SUSP bit of the DMA2D_CR
register.
The user application can abort at any time an ongoing transaction by setting the ABORT bit
of the DMA2D_CR register. In this case, the TCIF flag is not raised.
Automatic CLUT transfers can also be aborted or suspended by using the ABORT or the
SUSP bit of the DMA2D_CR register.
350/1751 RM0090 Rev 19

11.3.13 Watermark
A watermark can be programmed to generate an interrupt when the last pixel of a given line
has been written to the destination memory area.
The line number is defined in the LW[15:0] field of the DMA2D_LWR register.
When the last pixel of this line has been transferred, the TWIF flag of the DMA2D_ISR
register is raised and an interrupt is generated if the TWIE bit of the DMA2D_CR is set.

Two kind of errors can be triggered:
• AHB master port errors signalled by the TEIF flag of the DMA2D_ISR register.
• Conflicts caused by CLUT access (CPU trying to access the CLUT while a CLUT
loading or a DMA2D transfer is ongoing) signalled by the CAEIF flag of the
DMA2D_ISR register.
Both flags are associated to their own interrupt enable flag in the DMA2D_CR register to
generate an interrupt if need be (TEIE and CAEIE).
11.3.15 AHB dead time

To limit the AHB bandwidth usage, a dead time between two consecutive AHB accesses
can be programmed.
This feature can be enabled by setting the EN bit in the DMA2D_AMTCR register.
The dead time value is stored in the DT[7:0] field of the DMA2D_AMTCR register. This
value represents the guaranteed minimum number of cycles between two consecutive
transactions on the AHB bus.
The update of the dead time value while the DMA2D is running will be taken into account for
the next AHB transfer.
11.4 DMA2D interrupts

An interrupt can be generated on the following events:
• Configuration error
• CLUT transfer complete
• CLUT access error
• Transfer watermark reached
• Transfer complete
• Transfer error
Separate interrupt enable bits are available for flexibility.
Table 59. DMA2D interrupt requests

Configuration error CEIF CEIE

CLUT transfer complete CTCIF CTCIE
RM0090 Rev 19 351/1751

370
Table 59. DMA2D interrupt requests (continued)

CLUT access error CAEIF CAEIE

Transfer watermark TWF TWIE
Transfer complete TCIF TCIE
Transfer error TEIF TEIE
11.5 DMA2D registers
11.5.1 DMA2D control register (DMA2D_CR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
MODE
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CEIE CTCIE CAEIE TWIE TCIE TEIE ABORT SUSP START

Reserved Reserved
rw rw rw rw rw rw rs rw rs

Bits 17:16 MODE: DMA2D mode
This bit is set and cleared by software. It cannot be modified while a transfer is ongoing.
00: Memory-to-memory (FG fetch only)
01: Memory-to-memory with PFC (FG fetch only with FG PFC active)
10: Memory-to-memory with blending (FG and BG fetch with PFC and blending)
11: Register-to-memory (no FG nor BG, only output stage active)
Bit 13 CEIE: Configuration Error Interrupt Enable
0: CE interrupt disable
1: CE interrupt enable
Bit 12 CTCIE: CLUT transfer complete interrupt enable
0: CTC interrupt disable
1: CTC interrupt enable
Bit 11 CAEIE: CLUT access error interrupt enable
0: CAE interrupt disable
1: CAE interrupt enable
352/1751 RM0090 Rev 19

Bit 10 TWIE: Transfer watermark interrupt enable

0: TW interrupt disable
1: TW interrupt enable
Bit 9 TCIE: Transfer complete interrupt enable
0: TC interrupt disable
1: TC interrupt enable
Bit 8 TEIE: Transfer error interrupt enable
0: TE interrupt disable
1: TE interrupt enable
Bit 2 ABORT: Abort
This bit can be used to abort the current transfer. This bit is set by software and is
automatically reset by hardware when the START bit is reset.
0: No transfer abort requested
1: Transfer abort requested
Bit 1 SUSP: Suspend
This bit can be used to suspend the current transfer. This bit is set and reset by
software. It is automatically reset by hardware when the START bit is reset.
0: Transfer not suspended
1: Transfer suspended
Bit 0 START: Start
This bit can be used to launch the DMA2D according to the parameters loaded in the
various configuration registers. This bit is automatically reset by the following events:
– At the end of the transfer
– When the data transfer is aborted by the user application by setting the ABORT
bit in DMA2D_CR
– When a data transfer error occurs
– When the data transfer has not started due to a configuration error or another
transfer operation already ongoing (automatic CLUT loading).
RM0090 Rev 19 353/1751

370
11.5.2 DMA2D Interrupt Status Register (DMA2D_ISR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CEIF CTCIF CAEIF TWIF TCIF TEIF

Reserved
r r r r r r

Bit 5 CEIF: Configuration error interrupt flag
This bit is set when the START bit of DMA2D_CR, DMA2DFGPFCCR or
DMA2D_BGPFCCR is set and a wrong configuration has been programmed.
Bit 4 CTCIF: CLUT transfer complete interrupt flag
This bit is set when the CLUT copy from a system memory area to the internal DMA2D
memory is complete.
Bit 3 CAEIF: CLUT access error interrupt flag
This bit is set when the CPU accesses the CLUT while the CLUT is being automatically
copied from a system memory to the internal DMA2D.
Bit 2 TWIF: Transfer watermark interrupt flag
This bit is set when the last pixel of the watermarked line has been transferred.
Bit 1 TCIF: Transfer complete interrupt flag
This bit is set when a DMA2D transfer operation is complete (data transfer only).
Bit 0 TEIF: Transfer error interrupt flag
This bit is set when an error occurs during a DMA transfer (data transfer or automatic
CLUT loading).
354/1751 RM0090 Rev 19

11.5.3 DMA2D interrupt flag clear register (DMA2D_IFCR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCEIF CCTCIF CAECIF CTWIF CTCIF CTEIF

Reserved

Bit 5 CCEIF: Clear configuration error interrupt flag
Programming this bit to 1 clears the CEIF flag in the DMA2D_ISR register
Bit 4 CCTCIF: Clear CLUT transfer complete interrupt flag
Programming this bit to 1 clears the CTCIF flag in the DMA2D_ISR register
Bit 3 CAECIF: Clear CLUT access error interrupt flag
Programming this bit to 1 clears the CAEIF flag in the DMA2D_ISR register
Bit 2 CTWIF: Clear transfer watermark interrupt flag
Programming this bit to 1 clears the TWIF flag in the DMA2D_ISR register
Bit 1 CTCIF: Clear transfer complete interrupt flag
Programming this bit to 1 clears the TCIF flag in the DMA2D_ISR register
Bit 0 CTEIF: Clear Transfer error interrupt flag
Programming this bit to 1 clears the TEIF flag in the DMA2D_ISR register
RM0090 Rev 19 355/1751

370
11.5.4 DMA2D foreground memory address register (DMA2D_FGMAR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
MA[31:16]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MA[15:0]
Bits 31:0 MA[31: 0]: Memory address

Address of the data used for the foreground image. This register can only be written
when data transfers are disabled. Once the data transfer has started, this register is
read-only.
The address alignment must match the image format selected e.g. a 32-bit per pixel
format must be 32-bit aligned, a 16-bit per pixel format must be 16-bit aligned and a 4-
bit per pixel format must be 8-bit aligned.
11.5.5 DMA2D foreground offset register (DMA2D_FGOR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
LO[13:0]
Reserved

Bits 13:0 LO[13: 0]: Line offset
Line offset used for the foreground expressed in pixel. This value is used to generate
the address. It is added at the end of each line to determine the starting address of the
next line.
These bits can only be written when data transfers are disabled. Once a data transfer
has started, they become read-only.
If the image format is 4-bit per pixel, the line offset must be even.
356/1751 RM0090 Rev 19

11.5.6 DMA2D background memory address register (DMA2D_BGMAR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
MA[31:16]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MA[15:0]
Bits 31: 0 MA[31: 0]: Memory address

Address of the data used for the background image. This register can only be written
when data transfers are disabled. Once a data transfer has started, this register is read-
only.
format must be 32-bit aligned, a 16-bit per pixel format must be 16-bit aligned and a 4-
bit per pixel format must be 8-bit aligned.
11.5.7 DMA2D background offset register (DMA2D_BGOR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
LO[13:0]
Reserved

Bits 13:0 LO[13: 0]: Line offset
Line offset used for the background image (expressed in pixel). This value is used for
the address generation. It is added at the end of each line to determine the starting
address of the next line.
These bits can only be written when data transfers are disabled. Once data transfer has
started, they become read-only.
If the image format is 4-bit per pixel, the line offset must be even.
RM0090 Rev 19 357/1751

370
11.5.8 DMA2D foreground PFC control register (DMA2D_FGPFCCR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
ALPHA[7:0] AM[1:0]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CS[7:0] START CCM CM[3:0]

Reserved
rw rw rw rw rw rw rw rw rs rw rw rw rw rw
Bits 31:24 ALPHA[7: 0]: Alpha value

These bits define a fixed alpha channel value which can replace the original alpha value
or be multiplied by the original alpha value according to the alpha mode selected
through the AM[1:0] bits.
These bits can only be written when data transfers are disabled. Once a transfer has
started, they become read-only.
Bits 17:16 AM[1: 0]: Alpha mode
These bits select the alpha channel value to be used for the foreground image. They
can only be written data the transfer are disabled. Once the transfer has started, they
become read-only.
00: No modification of the foreground image alpha channel value
01: Replace original foreground image alpha channel value by ALPHA[7: 0]
10: Replace original foreground image alpha channel value by ALPHA[7:0] multiplied
with original alpha channel value
other configurations are meaningless
Bits 15:8 CS[7: 0]: CLUT size
These bits define the size of the CLUT used for the foreground image. Once the CLUT
transfer has started, this field is read-only.
The number of CLUT entries is equal to CS[7:0] + 1.
358/1751 RM0090 Rev 19

Bit 5 START: Start

This bit can be set to start the automatic loading of the CLUT. It is automatically reset:
– at the end of the transfer
– when the transfer is aborted by the user application by setting the ABORT bit in
DMA2D_CR
– when a transfer error occurs
– when the transfer has not started due to a configuration error or another
transfer operation already ongoing (data transfer or automatic background
CLUT transfer).
Bit 4 CCM: CLUT color mode
This bit defines the color format of the CLUT. It can only be written when the transfer is
disabled. Once the CLUT transfer has started, this bit is read-only.
0: ARGB8888
1: RGB888
others: meaningless
Bits 3:0 CM[3: 0]: Color mode
These bits defines the color format of the foreground image. They can only be written
when data transfers are disabled. Once the transfer has started, they are read-only.
0000: ARGB8888
0001: RGB888
0010: RGB565
0011: ARGB1555
0100: ARGB4444
0101: L8
0110: AL44
0111: AL88
1000: L4
1001: A8
1010: A4
others: meaningless
RM0090 Rev 19 359/1751

370
11.5.9 DMA2D foreground color register (DMA2D_FGCOLR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
RED[7:0]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
GREEN[7:0] BLUE[7:0]

Bits 23:16 RED[7: 0]: Red Value
These bits defines the red value for the A4 or A8 mode of the foreground image. They
can only be written when data transfers are disabled. Once the transfer has started,
they are read-only.
Bits 15:8 GREEN[7: 0]: Green Value
These bits defines the green value for the A4 or A8 mode of the foreground image. They
They are read-only.
Bits 7:0 BLUE[7: 0]: Blue Value
These bits defines the blue value for the A4 or A8 mode of the foreground image. They
They are read-only.
360/1751 RM0090 Rev 19

11.5.10 DMA2D background PFC control register (DMA2D_BGPFCCR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
ALPHA[7:0] AM[1:0]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CS[7:0] START CCM CM[3:0]

Reserved
rw rw rw rw rw rw rw rw rs rw rw rw rw rw
Bits 31:24 ALPHA[7: 0]: Alpha value

These bits define a fixed alpha channel value which can replace the original alpha value
or be multiplied with the original alpha value according to the alpha mode selected with
bits AM[1: 0]. These bits can only be written when data transfers are disabled. Once the
transfer has started, they are read-only.
Bits 17:16 AM[1: 0]: Alpha mode
These bits define which alpha channel value to be used for the background image.
These bits can only be written when data transfers are disabled. Once the transfer has
started, they are read-only.
00: No modification of the foreground image alpha channel value
01: Replace original background image alpha channel value by ALPHA[7: 0]
10: Replace original background image alpha channel value by ALPHA[7:0] multiplied
with original alpha channel value
others: meaningless
Bits 15:8 CS[7: 0]: CLUT size
These bits define the size of the CLUT used for the BG. Once the CLUT transfer has
started, this field is read-only.
The number of CLUT entries is equal to CS[7:0] + 1.
RM0090 Rev 19 361/1751

370
Bit 5 START: Start

This bit is set to start the automatic loading of the CLUT. This bit is automatically reset:
– at the end of the transfer
– when the transfer is aborted by the user application by setting the ABORT bit in
the DMA2D_CR
– when a transfer error occurs
– when the transfer has not started due to a configuration error or another
transfer operation already on going (data transfer or automatic foreground
CLUT transfer).
Bit 4 CCM: CLUT Color mode
These bits define the color format of the CLUT. This register can only be written when
the transfer is disabled. Once the CLUT transfer has started, this bit is read-only.
0: ARGB8888
1: RGB888
others: meaningless
Bits 3:0 CM[3: 0]: Color mode
These bits define the color format of the foreground image. These bits can only be
written when data transfers are disabled. Once the transfer has started, they are read-
only.
0000: ARGB8888
0001: RGB888
0010: RGB565
0011: ARGB1555
0100: ARGB4444
0101: L8
0110: AL44
0111: AL88
1000: L4
1001: A8
1010: A4
others: meaningless
362/1751 RM0090 Rev 19

11.5.11 DMA2D background color register (DMA2D_BGCOLR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
RED[7:0]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

These bits define the red value for the A4 or A8 mode of the background. These bits
they are read-only.
These bits define the green value for the A4 or A8 mode of the background. These bits
they are read-only.
These bits define the blue value for the A4 or A8 mode of the background. These bits
they are read-only.
11.5.12 DMA2D foreground CLUT memory address register

(DMA2D_FGCMAR)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
MA[31:16]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MA[15:0]
Bits 31: 0 MA[31: 0]: Memory Address

Address of the data used for the CLUT address dedicated to the foreground image. This
register can only be written when no transfer is ongoing. Once the CLUT transfer has
started, this register is read-only.
If the foreground CLUT format is 32-bit, the address must be 32-bit aligned.
RM0090 Rev 19 363/1751

370
11.5.13 DMA2D background CLUT memory address register

(DMA2D_BGCMAR)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
MA[31:16]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MA[15:0]
Bits 31: 0 MA[31: 0]: Memory address

Address of the data used for the CLUT address dedicated to the background image.
This register can only be written when no transfer is on going. Once the CLUT transfer
has started, this register is read-only.
If the background CLUT format is 32-bit, the address must be 32-bit aligned.
11.5.14 DMA2D output PFC control register (DMA2D_OPFCCR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CM[2:0]
Reserved
rw rw rw
Bits 31: 3 Reserved, must be kept at reset value

Bits 2: 0 CM[2: 0]: Color mode
These bits define the color format of the output image. These bits can only be written
000: ARGB8888
001: RGB888
010: RGB565
011: ARGB1555
100: ARGB4444
others: meaningless
364/1751 RM0090 Rev 19

11.5.15 DMA2D output color register (DMA2D_OCOLR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
ALPHA[7:0] RED[7:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RED[4:0] GREEN[5:0] BLUE[4:0]
A RED[4:0] GREEN[4:0] BLUE[4:0]
ALPHA[3:0] RED[3:0] GREEN[3:0] BLUE[3:0]
Bits 31:24 ALPHA[7: 0]: Alpha Channel Value

These bits define the alpha channel of the output color. These bits can only be written
These bits define the red value of the output image. These bits can only be written when
data transfers are disabled. Once the transfer has started, they are read-only.
These bits define the green value of the output image. These bits can only be written
These bits define the blue value of the output image. These bits can only be written
RM0090 Rev 19 365/1751

370
11.5.16 DMA2D output memory address register (DMA2D_OMAR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
MA[31:16]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MA[15:0]
Bits 31: 0 MA[31: 0]: Memory Address

Address of the data used for the output FIFO. These bits can only be written when data
transfers are disabled. Once the transfer has started, they are read-only.
format must be 32-bit aligned and a 16-bit per pixel format must be 16-bit aligned.
366/1751 RM0090 Rev 19

11.5.17 DMA2D output offset register (DMA2D_OOR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
LO[13:0]
Reserved

Bits 13:0 LO[13: 0]: Line Offset
Line offset used for the output (expressed in pixels). This value is used for the address
generation. It is added at the end of each line to determine the starting address of the
next line. These bits can only be written when data transfers are disabled. Once the
transfer has started, they are read-only.
11.5.18 DMA2D number of line register (DMA2D_NLR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PL[13:0]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
NL[15:0]

Bits 29:16 PL[13: 0]: Pixel per lines
Number of pixels per lines of the area to be transferred. These bits can only be written
If any of the input image format is 4-bit per pixel, pixel per lines must be even.
Bits 15:0 NL[15: 0]: Number of lines
Number of lines of the area to be transferred. These bits can only be written when data
transfers are disabled. Once the transfer has started, they are read-only.
RM0090 Rev 19 367/1751

370
11.5.19 DMA2D line watermark register (DMA2D_LWR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
LW[15:0]

Bits 15:0 LW[15:0]: Line watermark
These bits allow to configure the line watermark for interrupt generation.
An interrupt is raised when the last pixel of the watermarked line has been transferred.
These bits can only be written when data transfers are disabled. Once the transfer has
started, they are read-only.
11.5.20 DMA2D AHB master timer configuration register (DMA2D_AMTCR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DT[7:0] EN
Reserved
Bits 31:16 Reserved

Bits 15:8 DT[7: 0]: Dead Time
Dead time value in the AHB clock cycle inserted between two consecutive accesses on
the AHB master port. These bits represent the minimum guaranteed number of cycles
between two consecutive AHB accesses.
Bits 7:1 Reserved
Bit 0 EN: Enable
Enables the dead time functionality.
368/1751 RM0090 Rev 19

11.5.21 DMA2D register map

The following table summarizes the DMA2D registers. Refer to Section 2.3: Memory map
for the DMA2D register base address.
Table 60. DMA2D register map and reset values
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
MODE[1:0]
ABORT
START
CTCIE
CAEIE
SUSP
TWIE
CEIE
TCIE
TEIE
DMA2D_CR
Reset value 0 0 0 0 0 0 0 0 0 0 0
CTCIF
CAEIF
TWIF
CEIF
TCIF
TEIF
DMA2D_ISR
0x0004 Reserved
Reset value 0 0 0 0 0 0
CAECIF
CCTCIF
CTWIF
CCEIF
CTCIF
CTEIF
DMA2D_IFCR
0x0008 Reserved
DMA2D_FGMAR MA[31:0]
0x000C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA2D_FGOR LO[13:0]
0x0010 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA2D_BGMAR MA[31:0]
0x0014
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA2D_BGOR LO[13:0]
0x0018 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0
AM[1:0]
START
CCM
DMA2D_FGPFCCR ALPHA[7:0] CS[7:0] CM[3:0]
0x001C Reserved Res
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA2D_FGCOLR APLHA[7:0] RED[7:0] GREEN[7:0] BLUE[7:0]
0x0020
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
AM[1:0]
START
CCM
DMA2D_BGPFCCR ALPHA[7:0] CS[7:0] CM[3:0]

0x0024 Reserved Res
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA2D_BGCOLR APLHA[7:0] RED[7:0] GREEN[7:0] BLUE[7:0]
0x0028
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA2D_FGCMAR MA[31:0]
0x002C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA2D_BGCMAR MA[31:0]
0x0030
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA2D_OPFCCR CM[2:0]
0x0034 Reserved
Reset value 0 0 0
APLHA[7:0] RED[7:0] GREEN[7:0] BLUE[7:0]
Reserved RED[4:0] GREEN[6:0] BLUE[4:0]
DMA2D_OCOLR
0x0038 Reserved A RED[4:0] GREEN[4:0] BLUE[4:0]
Reserved ALPHA[3:0] RED[3:0] GREEN[3:0] BLUE[3:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA2D_OMAR MA[31:0]
0x003C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RM0090 Rev 19 369/1751

370
Table 60. DMA2D register map and reset values (continued)
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
DMA2D_OOR LO[13:0]
0x0040 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA2D_NLR PL[13:0] NL[15:0]
0x0044 Res
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA2D_LWR LW[15:0]
0x0048 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMA2D_AMTCR DT[7:0]
EN
Reset value 0 0 0 0 0 0 0 0 0
0x0050-
- Reserved
Ox03FF
0x0400- DMA2D_FGCLUT APLHA[7:0][255:0] RED[7:0][255:0] GREEN[7:0][255:0] BLUE[7:0][255:0]

0x07FF Reset value X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X
0x0800- DMA2D_BGCLUT APLHA[7:0][255:0] RED[7:0][255:0] GREEN[7:0][255:0] BLUE[7:0][255:0]

0x0BFF Reset value X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X
370/1751 RM0090 Rev 19

RM0090 Interrupts and events
12 Interrupts and events
This Section applies to the whole STM32F4xx family, unless otherwise specified.
12.1 Nested vectored interrupt controller (NVIC)
12.1.1 NVIC features

The nested vector interrupt controller NVIC includes the following features:
• 82 maskable interrupt channels for STM32F405xx/07xx and STM32F415xx/17xx, and
up to 91 maskable interrupt channels for STM32F42xxx and STM32F43xxx (not
including the 16 interrupt lines of Cortex®-M4 with FPU)
• 16 programmable priority levels (4 bits of interrupt priority are used)
• low-latency exception and interrupt handling
• power management control
• implementation of system control registers
The NVIC and the processor core interface are closely coupled, which enables low latency
interrupt processing and efficient processing of late arriving interrupts.
All interrupts including the core exceptions are managed by the NVIC. For more information
on exceptions and NVIC programming, refer to programming manual PM0214.
12.1.2 SysTick calibration value register

The SysTick calibration value is fixed to 18750, which gives a reference time base of 1 ms
with the SysTick clock set to 18.75 MHz (HCLK/8, with HCLK set to 150 MHz).
12.1.3 Interrupt and exception vectors

See Table 62 and Table 62, for the vector table for the STM32F405xx/07xx and
STM32F415xx/17xx and STM32F42xxx and STM32F43xxx devices.
12.2 External interrupt/event controller (EXTI)

The external interrupt/event controller consists of up to 23 edge detectors for generating
event/interrupt requests. Each input line can be independently configured to select the type
(interrupt or event) and the corresponding trigger event (rising or falling or both). Each line
can also masked independently. A pending register maintains the status line of the interrupt
requests.
The grey rows in the following tables describe the vectors without specific position.
RM0090 Rev 19 371/1751

387
Interrupts and events RM0090
Table 61. Vector table for STM32F405xx/07xx and STM32F415xx/17xx

Position
Priority
Type of
Acronym Description Address
priority
- - - - Reserved 0x0000 0000

- -3 fixed Reset Reset 0x0000 0004
Non maskable interrupt. The RCC
- -2 fixed NMI Clock Security System (CSS) is linked 0x0000 0008
to the NMI vector.
- -1 fixed HardFault All class of fault 0x0000 000C
- 0 settable MemManage Memory management 0x0000 0010
- 1 settable BusFault Pre-fetch fault, memory access fault 0x0000 0014
- 2 settable UsageFault Undefined instruction or illegal state 0x0000 0018
0x0000 001C - 0x0000
- - - - Reserved
002B
System service call via SWI
- 3 settable SVCall 0x0000 002C
instruction
- 4 settable Debug Monitor Debug Monitor 0x0000 0030
- - - - Reserved 0x0000 0034
- 5 settable PendSV Pendable request for system service 0x0000 0038
- 6 settable SysTick System tick timer 0x0000 003C
0 7 settable WWDG Window Watchdog interrupt 0x0000 0040
PVD through EXTI line detection
1 8 settable PVD 0x0000 0044
interrupt
Tamper and TimeStamp interrupts
2 9 settable TAMP_STAMP 0x0000 0048
through the EXTI line
RTC Wakeup interrupt through the
3 10 settable RTC_WKUP 0x0000 004C
EXTI line
4 11 settable FLASH Flash global interrupt 0x0000 0050
5 12 settable RCC RCC global interrupt 0x0000 0054
6 13 settable EXTI0 EXTI Line0 interrupt 0x0000 0058
7 14 settable EXTI1 EXTI Line1 interrupt 0x0000 005C
11 18 settable DMA1_Stream0 DMA1 Stream0 global interrupt 0x0000 006C
372/1751 RM0090 Rev 19

Table 61. Vector table for STM32F405xx/07xx and STM32F415xx/17xx (continued)

Position
Priority
Type of
priority
12 19 settable DMA1_Stream1 DMA1 Stream1 global interrupt 0x0000 0070

ADC1, ADC2 and ADC3 global
18 25 settable ADC 0x0000 0088
interrupts
19 26 settable CAN1_TX CAN1 TX interrupts 0x0000 008C
20 27 settable CAN1_RX0 CAN1 RX0 interrupts 0x0000 0090
21 28 settable CAN1_RX1 CAN1 RX1 interrupt 0x0000 0094
22 29 settable CAN1_SCE CAN1 SCE interrupt 0x0000 0098
23 30 settable EXTI9_5 EXTI Line[9:5] interrupts 0x0000 009C
TIM1 Break interrupt and TIM9 global
24 31 settable TIM1_BRK_TIM9 0x0000 00A0
interrupt
TIM1 Update interrupt and TIM10
25 32 settable TIM1_UP_TIM10 0x0000 00A4
global interrupt
TIM1 Trigger and Commutation
26 33 settable TIM1_TRG_COM_TIM11 0x0000 00A8
interrupts and TIM11 global interrupt
27 34 settable TIM1_CC TIM1 Capture Compare interrupt 0x0000 00AC
28 35 settable TIM2 TIM2 global interrupt 0x0000 00B0
31 38 settable I2C1_EV I2C1 event interrupt 0x0000 00BC
32 39 settable I2C1_ER I2C1 error interrupt 0x0000 00C0
2
33 40 settable I2C2_EV I C2 event interrupt 0x0000 00C4
2C2
34 41 settable I2C2_ER I error interrupt 0x0000 00C8
35 42 settable SPI1 SPI1 global interrupt 0x0000 00CC
36 43 settable SPI2 SPI2 global interrupt 0x0000 00D0
37 44 settable USART1 USART1 global interrupt 0x0000 00D4
39 46 settable USART3 USART3 global interrupt 0x0000 00DC
RM0090 Rev 19 373/1751

387

Position
Priority
Type of
priority
40 47 settable EXTI15_10 EXTI Line[15:10] interrupts 0x0000 00E0

RTC Alarms (A and B) through EXTI
41 48 settable RTC_Alarm 0x0000 00E4
line interrupt
USB On-The-Go FS Wakeup through
42 49 settable OTG_FS_WKUP 0x0000 00E8
EXTI line interrupt
TIM8 Break interrupt and TIM12
43 50 settable TIM8_BRK_TIM12 0x0000 00EC
global interrupt
TIM8 Update interrupt and TIM13
44 51 settable TIM8_UP_TIM13 0x0000 00F0
global interrupt
TIM8 Trigger and Commutation
45 52 settable TIM8_TRG_COM_TIM14 0x0000 00F4
interrupts and TIM14 global interrupt
46 53 settable TIM8_CC TIM8 Capture Compare interrupt 0x0000 00F8
47 54 settable DMA1_Stream7 DMA1 Stream7 global interrupt 0x0000 00FC
48 55 settable FSMC FSMC global interrupt 0x0000 0100
49 56 settable SDIO SDIO global interrupt 0x0000 0104
50 57 settable TIM5 TIM5 global interrupt 0x0000 0108
51 58 settable SPI3 SPI3 global interrupt 0x0000 010C
52 59 settable UART4 UART4 global interrupt 0x0000 0110
TIM6 global interrupt,
54 61 settable TIM6_DAC DAC1 and DAC2 underrun error 0x0000 0118
interrupts
55 62 settable TIM7 TIM7 global interrupt 0x0000 011C
61 68 settable ETH Ethernet global interrupt 0x0000 0134
Ethernet Wakeup through EXTI line
62 69 settable ETH_WKUP 0x0000 0138
interrupt
374/1751 RM0090 Rev 19


Position
Priority
Type of
priority

67 74 settable OTG_FS USB On The Go FS global interrupt 0x0000 014C
71 78 settable USART6 USART6 global interrupt 0x0000 015C
72 79 settable I2C3_EV I2C3 event interrupt 0x0000 0160
2C3
73 80 settable I2C3_ER I error interrupt 0x0000 0164
USB On The Go HS End Point 1 Out
74 81 settable OTG_HS_EP1_OUT 0x0000 0168
global interrupt
USB On The Go HS End Point 1 In
75 82 settable OTG_HS_EP1_IN 0x0000 016C
global interrupt
USB On The Go HS Wakeup through
76 83 settable OTG_HS_WKUP 0x0000 0170
EXTI interrupt
77 84 settable OTG_HS USB On The Go HS global interrupt 0x0000 0174
78 85 settable DCMI DCMI global interrupt 0x0000 0178
79 86 settable CRYP CRYP crypto global interrupt 0x0000 017C
80 87 settable HASH_RNG Hash and Rng global interrupt 0x0000 0180
81 88 settable FPU FPU global interrupt 0x0000 0184
Table 62. Vector table for STM32F42xxx and STM32F43xxx

Position
Priority
Type of
priority
- - - - Reserved 0x0000 0000

- -3 fixed Reset Reset 0x0000 0004
Non maskable interrupt, Clock Security
- -2 fixed NMI 0x0000 0008
System
RM0090 Rev 19 375/1751

387
Table 62. Vector table for STM32F42xxx and STM32F43xxx (continued)

Position
Priority
Type of
priority
- -1 fixed HardFault All class of fault 0x0000 000C

- 0 settable MemManage Memory management 0x0000 0010
- 1 settable BusFault Pre-fetch fault, memory access fault 0x0000 0014
- 2 settable UsageFault Undefined instruction or illegal state 0x0000 0018
0x0000 001C -
- - - - Reserved
0x0000 002B
- 3 settable SVCall System Service call via SWI instruction 0x0000 002C
- 4 settable Debug Monitor Debug Monitor 0x0000 0030
- - - Reserved 0x0000 0034
- 5 settable PendSV Pendable request for system service 0x0000 0038
- 6 settable Systick System tick timer 0x0000 003C
0 7 settable WWDG Window Watchdog interrupt 0x0000 0040
1 8 settable PVD PVD through EXTI line detection interrupt 0x0000 0044
Tamper and TimeStamp interrupts through
2 9 settable TAMP_STAMP 0x0000 0048
the EXTI line
RTC Wakeup interrupt through the EXTI
3 10 settable RTC_WKUP 0x0000 004C
line
4 11 settable FLASH Flash global interrupt 0x0000 0050
5 12 settable RCC RCC global interrupt 0x0000 0054
7 14 settable EXTI1 EXTI Line1 interrupt 0x0000 005C
376/1751 RM0090 Rev 19


Position
Priority
Type of
priority
18 25 settable ADC ADC1, ADC2 and ADC3 global interrupts 0x0000 0088
23 30 settable EXTI9_5 EXTI Line[9:5] interrupts 0x0000 009C
24 31 settable TIM1_BRK_TIM9 0x0000 00A0
interrupt
TIM1 Update interrupt and TIM10 global
25 32 settable TIM1_UP_TIM10 0x0000 00A4
interrupt
TIM1 Trigger and Commutation interrupts
26 33 settable TIM1_TRG_COM_TIM11 0x0000 00A8
and TIM11 global interrupt
27 34 settable TIM1_CC TIM1 Capture Compare interrupt 0x0000 00AC
2C1
31 38 settable I2C1_EV I event interrupt 0x0000 00BC
32 39 settable I2C1_ER I2C1 error interrupt 0x0000 00C0
33 40 settable I2C2_EV I2C2 event interrupt 0x0000 00C4
2
34 41 settable I2C2_ER I C2 error interrupt 0x0000 00C8
35 42 settable SPI1 SPI1 global interrupt 0x0000 00CC
36 43 settable SPI2 SPI2 global interrupt 0x0000 00D0
39 46 settable USART3 USART3 global interrupt 0x0000 00DC
40 47 settable EXTI15_10 EXTI Line[15:10] interrupts 0x0000 00E0
RTC Alarms (A and B) through EXTI line
41 48 settable RTC_Alarm 0x0000 00E4
interrupt
USB On-The-Go FS Wakeup through EXTI
42 49 settable OTG_FS_WKUP 0x0000 00E8
line interrupt
43 50 settable TIM8_BRK_TIM12 0x0000 00EC
interrupt
RM0090 Rev 19 377/1751

387

Position
Priority
Type of
priority
TIM8 Update interrupt and TIM13 global

44 51 settable TIM8_UP_TIM13 0x0000 00F0
interrupt
TIM8_TRG_COM_TIM1 TIM8 Trigger and Commutation interrupts
45 52 settable 0x0000 00F4
4 and TIM14 global interrupt
46 53 settable TIM8_CC TIM8 Capture Compare interrupt 0x0000 00F8
47 54 settable DMA1_Stream7 DMA1 Stream7 global interrupt 0x0000 00FC
48 55 settable FSMC FSMC global interrupt 0x0000 0100
49 56 settable SDIO SDIO global interrupt 0x0000 0104
50 57 settable TIM5 TIM5 global interrupt 0x0000 0108
51 58 settable SPI3 SPI3 global interrupt 0x0000 010C
TIM6 global interrupt,
54 61 settable TIM6_DAC 0x0000 0118
DAC1 and DAC2 underrun error interrupts
55 62 settable TIM7 TIM7 global interrupt 0x0000 011C
61 68 settable ETH Ethernet global interrupt 0x0000 0134
Ethernet Wakeup through EXTI line
62 69 settable ETH_WKUP 0x0000 0138
interrupt
67 74 settable OTG_FS USB On The Go FS global interrupt 0x0000 014C
71 78 settable USART6 USART6 global interrupt 0x0000 015C
378/1751 RM0090 Rev 19


Position
Priority
Type of
priority
72 79 settable I2C3_EV I2C3 event interrupt 0x0000 0160

73 80 settable I2C3_ER I2 C3 error interrupt 0x0000 0164
USB On The Go HS End Point 1 Out global
74 81 settable OTG_HS_EP1_OUT 0x0000 0168
interrupt
USB On The Go HS End Point 1 In global
75 82 settable OTG_HS_EP1_IN 0x0000 016C
interrupt
USB On The Go HS Wakeup through EXTI
76 83 settable OTG_HS_WKUP 0x0000 0170
interrupt
77 84 settable OTG_HS USB On The Go HS global interrupt 0x0000 0174
78 85 settable DCMI DCMI global interrupt 0x0000 0178
79 86 settable CRYP CRYP crypto global interrupt 0x0000 017C
80 87 settable HASH_RNG Hash and Rng global interrupt 0x0000 0180
81 88 Settable FPU FPU global interrupt 0x0000 0184
82 89 settable UART7 UART 7 global interrupt 0x0000 0188
83 90 settable UART8 UART 8 global interrupt 0x0000 018C
84 91 settable SPI4 SPI 4 global interrupt 0x0000 0190
87 94 settable SAI1 SAI1 global interrupt 0x0000 019C
88 95 settable LCD-TFT LTDC global interrupt 0x0000 01A0
89 96 settable LCD-TFT LTDC global Error interrupt 0x0000 01A4
90 97 settable DMA2D DMA2D global interrupt 0x0000 01A8
12.2.1 EXTI main features

The main features of the EXTI controller are the following:
• independent trigger and mask on each interrupt/event line
• dedicated status bit for each interrupt line
• generation of up to 23 software event/interrupt requests
• detection of external signals with a pulse width lower than the APB2 clock period. Refer
to the electrical characteristics section of the STM32F4xx datasheets for details on this
parameter.
RM0090 Rev 19 379/1751

387
12.2.2 EXTI block diagram

Figure 41 shows the block diagram.
Figure 41. External interrupt/event controller block diagram
AMBA APB bus
PCLK2 Peripheral interface
23 23 23 23 23
Pending Interrupt Software Rising Falling

request mask interrupt trigger trigger
register register event selection selection
register register register
To NVIC interrupt 23 23 23 23
controller
23
Pulse Edge detect Input

23 generator 23 23 circuit line
Event
mask
register
MS32662V1
12.2.3 Wakeup event management

The STM32F4xx are able to handle external or internal events in order to wake up the core
(WFE). The wakeup event can be generated either by:
• enabling an interrupt in the peripheral control register but not in the NVIC, and enabling
the SEVONPEND bit in the Cortex®-M4 with FPU System Control register. When the
MCU resumes from WFE, the peripheral interrupt pending bit and the peripheral NVIC
IRQ channel pending bit (in the NVIC interrupt clear pending register) have to be
cleared.
• or configuring an external or internal EXTI line in event mode. When the CPU resumes
from WFE, it is not necessary to clear the peripheral interrupt pending bit or the NVIC
IRQ channel pending bit as the pending bit corresponding to the event line is not set.
To use an external line as a wakeup event, refer to Section 12.2.4: Functional description.
12.2.4 Functional description

To generate the interrupt, the interrupt line should be configured and enabled. This is done
by programming the two trigger registers with the desired edge detection and by enabling
the interrupt request by writing a ‘1’ to the corresponding bit in the interrupt mask register.
When the selected edge occurs on the external interrupt line, an interrupt request is
380/1751 RM0090 Rev 19

generated. The pending bit corresponding to the interrupt line is also set. This request is
reset by writing a ‘1’ in the pending register.
To generate the event, the event line should be configured and enabled. This is done by
programming the two trigger registers with the desired edge detection and by enabling the
event request by writing a ‘1’ to the corresponding bit in the event mask register. When the
selected edge occurs on the event line, an event pulse is generated. The pending bit
corresponding to the event line is not set.
An interrupt/event request can also be generated by software by writing a ‘1’ in the software
interrupt/event register.
Hardware interrupt selection

To configure the 23 lines as interrupt sources, use the following procedure:
• Configure the mask bits of the 23 interrupt lines (EXTI_IMR)
• Configure the Trigger selection bits of the interrupt lines (EXTI_RTSR and EXTI_FTSR)
• Configure the enable and mask bits that control the NVIC IRQ channel mapped to the
external interrupt controller (EXTI) so that an interrupt coming from one of the 23 lines
can be correctly acknowledged.
Hardware event selection

To configure the 23 lines as event sources, use the following procedure:
• Configure the mask bits of the 23 event lines (EXTI_EMR)
• Configure the Trigger selection bits of the event lines (EXTI_RTSR and EXTI_FTSR)
Software interrupt/event selection

The 23 lines can be configured as software interrupt/event lines. The following is the
procedure to generate a software interrupt.
• Configure the mask bits of the 23 interrupt/event lines (EXTI_IMR, EXTI_EMR)
• Set the required bit in the software interrupt register (EXTI_SWIER)
RM0090 Rev 19 381/1751

387
12.2.5 External interrupt/event line mapping

Up to 140 GPIOs (STM32F405xx/07xx and STM32F415xx/17xx), 168 GPIOs
(STM32F42xxx and STM32F43xxx) are connected to the 16 external interrupt/event lines in
the following manner:
Figure 42. External interrupt/event GPIO mapping (STM32F405xx/07xx and

STM32F415xx/17xx)
EXTI0[3:0] bits in the SYSCFG_EXTICR1 register
PA0
PB0
PC0 EXTI0
PD0
PE0
PF0
PG0
PH0
PI0
PA1
PB1
PC1 EXTI1
PD1
PE1
PF1
PG1
PH1
PI1
...
PA15
PB15
PC15
EXTI15
PD15
PE15
PF15
PG15
PH15
ai15897
382/1751 RM0090 Rev 19

Figure 43. External interrupt/event GPIO mapping (STM32F42xxx and STM32F43xxx)
PA0
PB0
PC0
PD0
PE0 EXTI0
PF0
PG0
PH0
PI0
PJ0
PK0
PA1
PB1
PC1
PD1
PE1 EXTI1
PF1
PG1
PH1
PI1
PJ1
PK1
PA15
PB15
PC15
PD15 EXTI15
PE15
PF15
PG15
PH15
PJ15
MS30440V1
The seven other EXTI lines are connected as follows:

• EXTI line 16 is connected to the PVD output
• EXTI line 17 is connected to the RTC Alarm event
• EXTI line 18 is connected to the USB OTG FS Wakeup event
• EXTI line 19 is connected to the Ethernet Wakeup event
• EXTI line 20 is connected to the USB OTG HS (configured in FS) Wakeup event
• EXTI line 21 is connected to the RTC Tamper and TimeStamp events
• EXTI line 22 is connected to the RTC Wakeup event
RM0090 Rev 19 383/1751

387
12.3 EXTI registers

12.3.1 Interrupt mask register (EXTI_IMR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
MR22 MR21 MR20 MR19 MR18 MR17 MR16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MR15 MR14 MR13 MR12 MR11 MR10 MR9 MR8 MR7 MR6 MR5 MR4 MR3 MR2 MR1 MR0

Bits 22:0 MRx: Interrupt mask on line x
0: Interrupt request from line x is masked
1: Interrupt request from line x is not masked
12.3.2 Event mask register (EXTI_EMR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
MR22 MR21 MR20 MR19 MR18 MR17 MR16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MR15 MR14 MR13 MR12 MR11 MR10 MR9 MR8 MR7 MR6 MR5 MR4 MR3 MR2 MR1 MR0

Bits 22:0 MRx: Event mask on line x
0: Event request from line x is masked
1: Event request from line x is not masked
384/1751 RM0090 Rev 19

12.3.3 Rising trigger selection register (EXTI_RTSR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
TR22 TR21 TR20 TR19 TR18 TR17 TR16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TR15 TR14 TR13 TR12 TR11 TR10 TR9 TR8 TR7 TR6 TR5 TR4 TR3 TR2 TR1 TR0

Bits 22:0 TRx: Rising trigger event configuration bit of line x
0: Rising trigger disabled (for Event and Interrupt) for input line
1: Rising trigger enabled (for Event and Interrupt) for input line
Note: The external wakeup lines are edge triggered, no glitch must be generated on these lines.
If a rising edge occurs on the external interrupt line while writing to the EXTI_RTSR register,
the pending bit is be set.
Rising and falling edge triggers can be set for the same interrupt line. In this configuration,
both generate a trigger condition.
12.3.4 Falling trigger selection register (EXTI_FTSR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
TR22 TR21 TR20 TR19 TR18 TR17 TR16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TR15 TR14 TR13 TR12 TR11 TR10 TR9 TR8 TR7 TR6 TR5 TR4 TR3 TR2 TR1 TR0

Bits 22:0 TRx: Falling trigger event configuration bit of line x
0: Falling trigger disabled (for Event and Interrupt) for input line
1: Falling trigger enabled (for Event and Interrupt) for input line.
Note: The external wakeup lines are edge triggered, no glitch must be generated on these lines.
If a falling edge occurs on the external interrupt line while writing to the EXTI_FTSR register,
the pending bit is not set.
Rising and falling edge triggers can be set for the same interrupt line. In this configuration,
both generate a trigger condition.
RM0090 Rev 19 385/1751

387
12.3.5 Software interrupt event register (EXTI_SWIER)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
SWIER SWIER SWIER SWIER SWIER SWIER SWIER
Reserved 22 21 20 19 18 17 16
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SWIER SWIER SWIER SWIER SWIER SWIER SWIER SWIER SWIER SWIER SWIER SWIER SWIER SWIER SWIER SWIER
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Bits 22:0 SWIERx: Software Interrupt on line x
If interrupt are enabled on line x in the EXTI_IMR register, writing '1' to SWIERx bit when it is
set at '0' sets the corresponding pending bit in the EXTI_PR register, thus resulting in an
interrupt request generation.
This bit is cleared by clearing the corresponding bit in EXTI_PR (by writing a 1 to the bit).
12.3.6 Pending register (EXTI_PR)

Reset value: undefined
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PR22 PR21 PR20 PR19 PR18 PR17 PR16
Reserved
rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PR15 PR14 PR13 PR12 PR11 PR10 PR9 PR8 PR7 PR6 PR5 PR4 PR3 PR2 PR1 PR0
rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1

Bits 22:0 PRx: Pending bit
0: No trigger request occurred
1: selected trigger request occurred
This bit is set when the selected edge event arrives on the external interrupt line.
This bit is cleared by programming it to ‘1’.
386/1751 RM0090 Rev 19

12.3.7 EXTI register map

Table 64 gives the EXTI register map and the reset values.
Table 63. External interrupt/event controller register map and reset values
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
EXTI_IMR MR[22:0]
0x00 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
EXTI_EMR MR[22:0]
0x04 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
EXTI_RTSR TR[22:0]
0x08 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
EXTI_FTSR TR[22:0]
0x0C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
EXTI_SWIER SWIER[22:0]
0x10 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
EXTI_PR PR[22:0]
0x14 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RM0090 Rev 19 387/1751

387
Analog-to-digital converter (ADC) RM0090
13 Analog-to-digital converter (ADC)
13.1 ADC introduction

The 12-bit ADC is a successive approximation analog-to-digital converter. It has up to 19
multiplexed channels allowing it to measure signals from 16 external sources, two internal
sources, and the VBAT channel. The A/D conversion of the channels can be performed in
single, continuous, scan or discontinuous mode. The result of the ADC is stored into a left-
or right-aligned 16-bit data register.
The analog watchdog feature allows the application to detect if the input voltage goes
beyond the user-defined, higher or lower thresholds.
13.2 ADC main features

• 12-bit, 10-bit, 8-bit or 6-bit configurable resolution
• Interrupt generation at the end of conversion, end of injected conversion, and in case of
analog watchdog or overrun events
• Single and continuous conversion modes
• Scan mode for automatic conversion of channel 0 to channel ‘n’
• Data alignment with in-built data coherency
• Channel-wise programmable sampling time
• External trigger option with configurable polarity for both regular and injected
conversions
• Discontinuous mode
• Dual/Triple mode (on devices with 2 ADCs or more)
• Configurable DMA data storage in Dual/Triple ADC mode
• Configurable delay between conversions in Dual/Triple interleaved mode
• ADC conversion type (refer to the datasheets)
• ADC supply requirements: 2.4 V to 3.6 V at full speed and down to 1.8 V at slower
speed
• ADC input range: VREF– ≤ VIN ≤ VREF+
• DMA request generation during regular channel conversion
Figure 44 shows the block diagram of the ADC.
Note: VREF–, if available (depending on package), must be tied to VSSA.
388/1751 RM0090 Rev 19

RM0090 Analog-to-digital converter (ADC)
13.3 ADC functional description

Figure 44 shows a single ADC block diagram and Table 65 gives the ADC pin description.
Figure 44. Single ADC block diagram
Flags Interrupt
enable bits
DMA overrun
OVR OVRIE
End of conversion
EOC EOCIE
End of injected conversion ADC Interrupt to NVIC
JEOC JEOCIE
Analog watchdog event
AWD AWDIE
Analog watchdog
Compare result
Higher threshold (12 bits)

Lower threshold (12 bits)
Address/data bus
Injected data registers
V REF+ (4 x 16 bits)
V REF-
Regular data register
V DDA (16 bits)
V SSA
Analog DMA request
mux
ADCx_IN0
ADCx_IN1
GPIO up to 4 ADCCLK
Injected
ports channels Analog to digital
up to 16 Regular converter
ADCx_IN15
channels
Temp. sensor
V REFINT
V BAT
From ADC prescaler
JEXTSEL[3:0] bits EXTSEL[3:0] bits
TIM1_CH4 TIM1_CH1
TIM1_TRGO JEXTEN EXTEN TIM1_CH2
TIM2_CH1 [1:0] bits [1:0] bits TIM1_CH3
TIM2_TRGO TIM2_CH2
TIM3_CH2 TIM2_CH3
TIM3_CH4 TIM2_CH4
TIM4_CH1 TIM2_TRGO
TIM4_CH2 TIM3_CH1
TIM4_CH3 TIM3_TRGO
TIM4_TRGO TIM4_CH4
Start trigger Start trigger
TIM5_CH4 TIM5_CH1
(injected group) (regular group)
TIM5_TRGO TIM5_CH2
TIM8_CH2 TIM5_CH3
TIM8_CH3 TIM8_CH1
TIM8_CH4 TIM8_TRGO
EXTI_15
EXTI_11
ai16046
RM0090 Rev 19 389/1751

432
Table 65. ADC pins

Name Signal type Remarks
Input, analog reference The higher/positive reference voltage for the ADC,
VREF+
positive 1.8 V ≤ VREF+ ≤ VDDA
Analog power supply equal to VDD and
VDDA Input, analog supply 2.4 V ≤ VDDA ≤ VDD (3.6 V) for full speed
1.8 V ≤ VDDA ≤ VDD (3.6 V) for reduced speed
Input, analog reference The lower/negative reference voltage for the ADC,
VREF–
negative VREF– = VSSA
Input, analog supply
VSSA Ground for analog power supply equal to VSS
ground
ADCx_IN[15:0] Analog input signals 16 analog input channels
13.3.1 ADC on-off control

The ADC is powered on by setting the ADON bit in the ADC_CR2 register. When the ADON
bit is set for the first time, it wakes up the ADC from the Power-down mode.
Conversion starts when either the SWSTART or the JSWSTART bit is set.
You can stop conversion and put the ADC in power down mode by clearing the ADON bit. In
this mode the ADC consumes almost no power (only a few µA).
13.3.2 ADC clock

The ADC features two clock schemes:
• Clock for the analog circuitry: ADCCLK, common to all ADCs
This clock is generated from the APB2 clock divided by a programmable prescaler that
allows the ADC to work at fPCLK2/2, /4, /6 or /8. Refer to the datasheets for the
maximum value of ADCCLK.
• Clock for the digital interface (used for registers read/write access)
This clock is equal to the APB2 clock. The digital interface clock can be
enabled/disabled individually for each ADC through the RCC APB2 peripheral clock
enable register (RCC_APB2ENR).
13.3.3 Channel selection

There are 16 multiplexed channels. It is possible to organize the conversions in two groups:
regular and injected. A group consists of a sequence of conversions that can be done on
any channel and in any order. For instance, it is possible to implement the conversion
sequence in the following order: ADC_IN3, ADC_IN8, ADC_IN2, ADC_IN2, ADC_IN0,
ADC_IN2, ADC_IN2, ADC_IN15.
• A regular group is composed of up to 16 conversions. The regular channels and their
order in the conversion sequence must be selected in the ADC_SQRx registers. The
total number of conversions in the regular group must be written in the L[3:0] bits in the
ADC_SQR1 register.
• An injected group is composed of up to 4 conversions. The injected channels and
their order in the conversion sequence must be selected in the ADC_JSQR register.
390/1751 RM0090 Rev 19

The total number of conversions in the injected group must be written in the L[1:0] bits
in the ADC_JSQR register.
If the ADC_SQRx or ADC_JSQR registers are modified during a conversion, the current
conversion is reset and a new start pulse is sent to the ADC to convert the newly chosen
group.
Temperature sensor, VREFINT and VBAT internal channels

• For the STM32F40x and STM32F41x devices, the temperature sensor is internally
connected to channel ADC1_IN16.
The internal reference voltage VREFINT is connected to ADC1_IN17.
• For the STM32F42x and STM32F43x devices, the temperature sensor is internally
connected to ADC1_IN18 channel which is shared with VBAT. Only one conversion,
temperature sensor or VBAT, must be selected at a time. When the temperature sensor
and VBAT conversion are set simultaneously, only the VBAT conversion is performed.
The internal reference voltage VREFINT is connected to ADC1_IN17.
The VBAT channel (connected to channel ADC1_IN18) can also be converted as an injected
or regular channel.
Note: The temperature sensor, VREFINT and the VBAT channel are available only on the master
ADC1 peripheral.
13.3.4 Single conversion mode

In Single conversion mode the ADC does one conversion. This mode is started with the
CONT bit at 0 by either:
• setting the SWSTART bit in the ADC_CR2 register (for a regular channel only)
• setting the JSWSTART bit (for an injected channel)
• external trigger (for a regular or injected channel)
Once the conversion of the selected channel is complete:
• If a regular channel was converted:
– The converted data are stored into the 16-bit ADC_DR register
– The EOC (end of conversion) flag is set
– An interrupt is generated if the EOCIE bit is set
• If an injected channel was converted:
– The converted data are stored into the 16-bit ADC_JDR1 register
– The JEOC (end of conversion injected) flag is set
– An interrupt is generated if the JEOCIE bit is set
Then the ADC stops.
RM0090 Rev 19 391/1751

432
13.3.5 Continuous conversion mode

In continuous conversion mode, the ADC starts a new conversion as soon as it finishes one.
This mode is started with the CONT bit at 1 either by external trigger or by setting the
SWSTART bit in the ADC_CR2 register (for regular channels only).
After each conversion:
• If a regular group of channels was converted:
– The last converted data are stored into the 16-bit ADC_DR register
– The EOC (end of conversion) flag is set
– An interrupt is generated if the EOCIE bit is set
Note: Injected channels cannot be converted continuously. The only exception is when an injected
channel is configured to be converted automatically after regular channels in continuous
mode (using JAUTO bit), refer to Auto-injection section).
13.3.6 Timing diagram

As shown in Figure 45, the ADC needs a stabilization time of tSTAB before it starts
converting accurately. After the start of the ADC conversion and after 15 clock cycles, the
EOC flag is set and the 16-bit ADC data register contains the result of the conversion.
Figure 45. Timing diagram
ADC_CLK
ADON
SWSTART/
JSWSTART
Start 1st conversion Start next conversion
ADC conversion Next ADC conversion

ADC
Conversion time
tSTAB
(total conv. time)
EOC
Software clears the EOC bit

ai16047b
13.3.7 Analog watchdog

The AWD analog watchdog status bit is set if the analog voltage converted by the ADC is
below a lower threshold or above a higher threshold. These thresholds are programmed in
the 12 least significant bits of the ADC_HTR and ADC_LTR 16-bit registers. An interrupt can
be enabled by using the AWDIE bit in the ADC_CR1 register.
The threshold value is independent of the alignment selected by the ALIGN bit in the
ADC_CR2 register. The analog voltage is compared to the lower and higher thresholds
before alignment.
Table 66 shows how the ADC_CR1 register should be configured to enable the analog
watchdog on one or more channels.
392/1751 RM0090 Rev 19

Figure 46. Analog watchdog’s guarded area
Analog voltage
Higher threshold HTR

Guarde d area
Lower threshold LT R
ai16048
Table 66. Analog watchdog channel selection

ADC_CR1 register control bits (x = don’t care)
Channels guarded by the analog
watchdog
AWDSGL bit AWDEN bit JAWDEN bit
None x 0 0
All injected channels 0 0 1
All regular channels 0 1 0
All regular and injected channels 0 1 1
Single(1) injected channel 1 0 1
Single(1) regular channel 1 1 0
(1)
Single regular or injected channel 1 1 1
1. Selected by the AWDCH[4:0] bits
13.3.8 Scan mode

This mode is used to scan a group of analog channels.
The Scan mode is selected by setting the SCAN bit in the ADC_CR1 register. Once this bit
has been set, the ADC scans all the channels selected in the ADC_SQRx registers (for
regular channels) or in the ADC_JSQR register (for injected channels). A single conversion
is performed for each channel of the group. After each end of conversion, the next channel
in the group is converted automatically. If the CONT bit is set, regular channel conversion
does not stop at the last selected channel in the group but continues again from the first
selected channel.
If the DMA bit is set, the direct memory access (DMA) controller is used to transfer the data
converted from the regular group of channels (stored in the ADC_DR register) to SRAM
after each regular channel conversion.
The EOC bit is set in the ADC_SR register:
• At the end of each regular group sequence if the EOCS bit is cleared to 0
• At the end of each regular channel conversion if the EOCS bit is set to 1
The data converted from an injected channel are always stored into the ADC_JDRx
registers.
RM0090 Rev 19 393/1751

432
13.3.9 Injected channel management

Triggered injection
To use triggered injection, the JAUTO bit must be cleared in the ADC_CR1 register.
1. Start the conversion of a group of regular channels either by external trigger or by
setting the SWSTART bit in the ADC_CR2 register.
2. If an external injected trigger occurs or if the JSWSTART bit is set during the
conversion of a regular group of channels, the current conversion is reset and the
injected channel sequence switches to Scan-once mode.
3. Then, the regular conversion of the regular group of channels is resumed from the last
interrupted regular conversion.
If a regular event occurs during an injected conversion, the injected conversion is not
interrupted but the regular sequence is executed at the end of the injected sequence.
Figure 47 shows the corresponding timing diagram.
Note: When using triggered injection, one must ensure that the interval between trigger events is
longer than the injection sequence. For instance, if the sequence length is 30 ADC clock
cycles (that is two conversions with a sampling time of 3 clock periods), the minimum
interval between triggers must be 31 ADC clock cycles.
Auto-injection
If the JAUTO bit is set, then the channels in the injected group are automatically converted
after the regular group of channels. This can be used to convert a sequence of up to 20
conversions programmed in the ADC_SQRx and ADC_JSQR registers.
In this mode, external trigger on injected channels must be disabled.
If the CONT bit is also set in addition to the JAUTO bit, regular channels followed by injected
channels are continuously converted.
Note: It is not possible to use both the auto-injected and discontinuous modes simultaneously.
Figure 47. Injected conversion latency
ADCCLK
Injection event
Reset ADC
max latency (1)

SOC
ai16049
1. The maximum latency value can be found in the electrical characteristics of the STM32F40x and
STM32F41x datasheets.
394/1751 RM0090 Rev 19

13.3.10 Discontinuous mode

Regular group
This mode is enabled by setting the DISCEN bit in the ADC_CR1 register. It can be used to
convert a short sequence of n conversions (n ≤ 8) that is part of the sequence of
conversions selected in the ADC_SQRx registers. The value of n is specified by writing to
the DISCNUM[2:0] bits in the ADC_CR1 register.
When an external trigger occurs, it starts the next n conversions selected in the ADC_SQRx
registers until all the conversions in the sequence are done. The total sequence length is
defined by the L[3:0] bits in the ADC_SQR1 register.
Example:
• n = 3, channels to be converted = 0, 1, 2, 3, 6, 7, 9, 10
• 1st trigger: sequence converted 0, 1, 2. An EOC event is generated at each
conversion.
• 2nd trigger: sequence converted 3, 6, 7. An EOC event is generated at each
conversion
• 3rd trigger: sequence converted 9, 10.An EOC event is generated at each conversion
• 4th trigger: sequence converted 0, 1, 2. An EOC event is generated at each conversion
Note: When a regular group is converted in discontinuous mode, no rollover occurs.
When all subgroups are converted, the next trigger starts the conversion of the first
subgroup. In the example above, the 4th trigger reconverts the channels 0, 1 and 2 in the
1st subgroup.
Injected group
This mode is enabled by setting the JDISCEN bit in the ADC_CR1 register. It can be used to
convert the sequence selected in the ADC_JSQR register, channel by channel, after an
external trigger event.
When an external trigger occurs, it starts the next channel conversions selected in the
ADC_JSQR registers until all the conversions in the sequence are done. The total sequence
length is defined by the JL[1:0] bits in the ADC_JSQR register.
Example:
n = 1, channels to be converted = 1, 2, 3
1st trigger: channel 1 converted
2nd trigger: channel 2 converted
3rd trigger: channel 3 converted and JEOC event generated
4th trigger: channel 1
Note: When all injected channels are converted, the next trigger starts the conversion of the first
injected channel. In the example above, the 4th trigger reconverts the 1st injected channel
1.
It is not possible to use both the auto-injected and discontinuous modes simultaneously.
Discontinuous mode must not be set for regular and injected groups at the same time.
Discontinuous mode must be enabled only for the conversion of one group.
RM0090 Rev 19 395/1751

432
13.4 Data alignment

The ALIGN bit in the ADC_CR2 register selects the alignment of the data stored after
conversion. Data can be right- or left-aligned as shown in Figure 48 and Figure 49.
The converted data value from the injected group of channels is decreased by the user-
defined offset written in the ADC_JOFRx registers so the result can be a negative value.
The SEXT bit represents the extended sign value.
For channels in a regular group, no offset is subtracted so only twelve bits are significant.
Figure 48. Right alignment of 12-bit data
Injected group
SEXT SEXT SEXT SEXT D11 D10 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0
Regular group
0 0 0 0 D11 D10 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0
ai16050
Figure 49. Left alignment of 12-bit data
Injected group
SEXT D11 D10 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0 0 0 0
Regular group
D11 D10 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0 0 0 0 0
ai16051
Special case: when left-aligned, the data are aligned on a half-word basis except when the
resolution is set to 6-bit. in that case, the data are aligned on a byte basis as shown in
Figure 50.
Figure 50. Left alignment of 6-bit data
Injected group
SEXT SEXT SEXT SEXT SEXT SEXT SEXT SEXT SEXT D5 D4 D3 D2 D1 D0 0
Regular group
0 0 0 0 0 0 0 0 D5 D4 D3 D2 D1 D0 0 0
ai16052
396/1751 RM0090 Rev 19

13.5 Channel-wise programmable sampling time

The ADC samples the input voltage for a number of ADCCLK cycles that can be modified
using the SMP[2:0] bits in the ADC_SMPR1 and ADC_SMPR2 registers. Each channel can
be sampled with a different sampling time.
The total conversion time is calculated as follows:
Tconv = Sampling time + 12 cycles
Example:
With ADCCLK = 30 MHz and sampling time = 3 cycles:
Tconv = 3 + 12 = 15 cycles = 0.5 µs with APB2 at 60 MHz
13.6 Conversion on external trigger and trigger polarity

Conversion can be triggered by an external event (e.g. timer capture, EXTI line). If the
EXTEN[1:0] control bits (for a regular conversion) or JEXTEN[1:0] bits (for an injected
conversion) are different from “0b00”, then external events are able to trigger a conversion
with the selected polarity. Table 67 provides the correspondence between the EXTEN[1:0]
and JEXTEN[1:0] values and the trigger polarity.
Table 67. Configuring the trigger polarity

Source EXTEN[1:0] / JEXTEN[1:0]
Trigger detection disabled 00

Detection on the rising edge 01
Detection on the falling edge 10
Detection on both the rising and falling edges 11
Note: The polarity of the external trigger can be changed on the fly.
The EXTSEL[3:0] and JEXTSEL[3:0] control bits are used to select which out of 16 possible
events can trigger conversion for the regular and injected groups.
Table 68 gives the possible external trigger for regular conversion.
RM0090 Rev 19 397/1751

432
Table 68. External trigger for regular channels

Source Type EXTSEL[3:0]
TIM1_CH1 event 0000

TIM1_CH2 event 0001
TIM1_CH3 event 0010
TIM2_CH2 event 0011
TIM2_CH3 event 0100
TIM2_CH4 event 0101
TIM2_TRGO event 0110
Internal signal from on-chip
TIM3_CH1 event 0111
timers
TIM4_CH4 event 1001
TIM5_CH1 event 1010
TIM5_CH2 event 1011
TIM5_CH3 event 1100
TIM8_CH1 event 1101
EXTI line11 External pin 1111
Table 69 gives the possible external trigger for injected conversion.
398/1751 RM0090 Rev 19

Table 69. External trigger for injected channels

Source Connection type JEXTSEL[3:0]
TIM1_CH4 event 0000

TIM2_CH1 event 0010
TIM3_CH2 event 0100
TIM3_CH4 event 0101
TIM4_CH1 event 0110
Internal signal from on-chip
TIM4_CH2 event 0111
timers
TIM4_CH3 event 1000
TIM5_CH4 event 1010
TIM8_CH2 event 1100
TIM8_CH3 event 1101
TIM8_CH4 event 1110
Software source trigger events can be generated by setting SWSTART (for regular
conversion) or JSWSTART (for injected conversion) in ADC_CR2.
A regular group conversion can be interrupted by an injected trigger.
Note: The trigger selection can be changed on the fly. However, when the selection changes,
there is a time frame of 1 APB clock cycle during which the trigger detection is disabled.
This is to avoid spurious detection during transitions.
13.7 Fast conversion mode

It is possible to perform faster conversion by reducing the ADC resolution. The RES bits are
used to select the number of bits available in the data register. The minimum conversion
time for each resolution is then as follows:
• 12 bits: 3 + 12 = 15 ADCCLK cycles
RM0090 Rev 19 399/1751

432
13.8 Data management
13.8.1 Using the DMA

Since converted regular channel values are stored into a unique data register, it is useful to
use DMA for conversion of more than one regular channel. This avoids the loss of the data
already stored in the ADC_DR register.
When the DMA mode is enabled (DMA bit set to 1 in the ADC_CR2 register), after each
conversion of a regular channel, a DMA request is generated. This allows the transfer of the
converted data from the ADC_DR register to the destination location selected by the
software.
Despite this, if data are lost (overrun), the OVR bit in the ADC_SR register is set and an
interrupt is generated (if the OVRIE enable bit is set). DMA transfers are then disabled and
DMA requests are no longer accepted. In this case, if a DMA request is made, the regular
conversion in progress is aborted and further regular triggers are ignored. It is then
necessary to clear the OVR flag and the DMAEN bit in the used DMA stream, and to re-
initialize both the DMA and the ADC to have the wanted converted channel data transferred
to the right memory location. Only then can the conversion be resumed and the data
transfer, enabled again. Injected channel conversions are not impacted by overrun errors.
When OVR = 1 in DMA mode, the DMA requests are blocked after the last valid data have
been transferred, which means that all the data transferred to the RAM can be considered
as valid.
At the end of the last DMA transfer (number of transfers configured in the DMA controller’s
DMA_SxNDTR register):
• No new DMA request is issued to the DMA controller if the DDS bit is cleared to 0 in the
ADC_CR2 register (this avoids generating an overrun error). However the DMA bit is
not cleared by hardware. It must be written to 0, then to 1 to start a new transfer.
• Requests can continue to be generated if the DDS bit is set to 1. This allows
configuring the DMA in double-buffer circular mode.
To recover the ADC from OVR state when the DMA is used, follow the steps below:
1. Reinitialize the DMA (adjust destination address and NDTR counter)
2. Clear the ADC OVR bit in ADC_SR register
3. Trigger the ADC to start the conversion.
13.8.2 Managing a sequence of conversions without using the DMA

If the conversions are slow enough, the conversion sequence can be handled by the
software. In this case the EOCS bit must be set in the ADC_CR2 register for the EOC status
bit to be set at the end of each conversion, and not only at the end of the sequence. When
EOCS = 1, overrun detection is automatically enabled. Thus, each time a conversion is
complete, EOC is set and the ADC_DR register can be read. The overrun management is
the same as when the DMA is used.
To recover the ADC from OVR state when the EOCS is set, follow the steps below:
1. Clear the ADC OVR bit in ADC_SR register
2. Trigger the ADC to start the conversion.
400/1751 RM0090 Rev 19

13.8.3 Conversions without DMA and without overrun detection

It may be useful to let the ADC convert one or more channels without reading the data each
time (if there is an analog watchdog for instance). For that, the DMA must be disabled
(DMA = 0) and the EOC bit must be set at the end of a sequence only (EOCS = 0). In this
configuration, overrun detection is disabled.
13.9 Multi ADC mode

In devices with two ADCs or more, the Dual (with two ADCs) and Triple (with three ADCs)
ADC modes can be used (see Figure 51).
In multi ADC mode, the start of conversion is triggered alternately or simultaneously by the
ADC1 master to the ADC2 and ADC3 slaves, depending on the mode selected by the
MULTI[4:0] bits in the ADC_CCR register.
Note: In multi ADC mode, when configuring conversion trigger by an external event, the
application must set trigger by the master only and disable trigger by slaves to prevent
spurious triggers that would start unwanted slave conversions.
The four possible modes below are implemented:
• Injected simultaneous mode
• Regular simultaneous mode
• Interleaved mode
• Alternate trigger mode
It is also possible to use the previous modes combined in the following ways:
• Injected simultaneous mode + Regular simultaneous mode
• Regular simultaneous mode + Alternate trigger mode
Note: In multi ADC mode, the converted data can be read on the multi-mode data register
(ADC_CDR). The status bits can be read in the multi-mode status register (ADC_CSR).
RM0090 Rev 19 401/1751

432
Figure 51. Multi ADC block diagram(1)

(12bits)
(16 bits)

Regular (4 x 16 bits)
channels
Injected ADC3(2) (Slave)

channels

(12
(16bits)
bits)

Regular (4 x 16 bits)
channels
Injected ADC2 (Slave)

channels
Address/data bus
internal triggers
Common regular data register

(32 bits)(3)
Dual/Triple
mode control Common part

(16 bits)

(4 x 16 bits)
ADCx_IN0
ADCx_IN1
GPIO
Ports
Regular
channels
ADCx_IN15
Injected
channels
Temp. sensor
VREFINT
VBAT
EXTI_11 ADC1 (Master)

Start trigger mux
(regular group)
EXTI_15
Start trigger mux
(injected group)
ai16053
1. Although external triggers are present on ADC2 and ADC3 they are not shown in this diagram.
2. In the Dual ADC mode, the ADC3 slave part is not present.
3. In Triple ADC mode, the ADC common data register (ADC_CDR) contains the ADC1, ADC2 and ADC3’s
regular converted data. All 32 register bits are used according to a selected storage order.
In Dual ADC mode, the ADC common data register (ADC_CDR) contains both the ADC1 and ADC2’s
regular converted data. All 32 register bits are used.
402/1751 RM0090 Rev 19

• DMA requests in Multi ADC mode:

In Multi ADC mode the DMA may be configured to transfer converted data in three
different modes. In all cases, the DMA streams to use are those connected to the ADC:
– DMA mode 1: On each DMA request (one data item is available), a half-word
representing an ADC-converted data item is transferred.
In Triple ADC mode, ADC1 data are transferred on the first request, ADC2 data
are transferred on the second request and ADC3 data are transferred on the third
request; the sequence is repeated. So the DMA first transfers ADC1 data followed
by ADC2 data followed by ADC3 data and so on.
DMA mode 1 is used in regular simultaneous triple mode only.
Example:
Regular simultaneous triple mode: 3 consecutive DMA requests are generated
(one for each converted data item)
1st request: ADC_CDR[31:0] = ADC1_DR[15:0]
2nd request: ADC_CDR[31:0] = ADC2_DR[15:0]
3rd request: ADC_CDR[31:0] = ADC3_DR[15:0]
4th request: ADC_CDR[31:0] = ADC1_DR[15:0]
– DMA mode 2: On each DMA request (two data items are available) two half-
words representing two ADC-converted data items are transferred as a word.
In Dual ADC mode, both ADC2 and ADC1 data are transferred on the first request
(ADC2 data take the upper half-word and ADC1 data take the lower half-word) and
so on.
In Triple ADC mode, three DMA requests are generated. On the first request, both
ADC2 and ADC1 data are transferred (ADC2 data take the upper half-word and
ADC1 data take the lower half-word). On the second request, both ADC1 and
ADC3 data are transferred (ADC1 data take the upper half-word and ADC3 data
take the lower half-word).On the third request, both ADC3 and ADC2 data are
transferred (ADC3 data take the upper half-word and ADC2 data take the lower
half-word) and so on.
DMA mode 2 is used in interleaved mode and in regular simultaneous mode (for
Dual ADC mode only).
Example:
a) Interleaved dual mode: a DMA request is generated each time 2 data items are
available:
1st request: ADC_CDR[31:0] = ADC2_DR[15:0] | ADC1_DR[15:0]
2nd request: ADC_CDR[31:0] = ADC2_DR[15:0] | ADC1_DR[15:0]
b) Interleaved triple mode: a DMA request is generated each time 2 data items are
available
3rd request: ADC_CDR[31:0] = ADC3_DR[15:0] | ADC2_DR[15:0]
4th request: ADC_CDR[31:0] = ADC2_DR[15:0] | ADC1_DR[15:0]
– DMA mode 3: This mode is similar to the DMA mode 2. The only differences are
that the on each DMA request (two data items are available) two bytes
RM0090 Rev 19 403/1751

432
representing two ADC converted data items are transferred as a half-word. The
data transfer order is similar to that of the DMA mode 2.
DMA mode 3 is used in interleaved mode in 6-bit and 8-bit resolutions (dual and
triple mode).
Example:
a) Interleaved dual mode: a DMA request is generated each time 2 data items are
available
b) Interleaved triple mode: a DMA request is generated each time 2 data items are
available
3rd request: ADC_CDR[15:0] = ADC3_DR[7:0] | ADC2_DR[7:0]
4th request: ADC_CDR[15:0] = ADC2_DR[7:0] | ADC1_DR[7:0]
Overrun detection: If an overrun is detected on one of the concerned ADCs (ADC1 and
ADC2 in dual and triple modes, ADC3 in triple mode only), the DMA requests are no longer
issued to ensure that all the data transferred to the RAM are valid. It may happen that the
EOC bit corresponding to one ADC remains set because the data register of this ADC
contains valid data.
13.9.1 Injected simultaneous mode

This mode converts an injected group of channels. The external trigger source comes from
the injected group multiplexer of ADC1 (selected by the JEXTSEL[3:0] bits in the ADC1_CR2
register). A simultaneous trigger is provided to ADC2 and ADC3.
Note: Do not convert the same channel on the two/three ADCs (no overlapping sampling times for
the two/three ADCs when converting the same channel).
In simultaneous mode, one must convert sequences with the same length or ensure that the
interval between triggers is longer than the longer of the 2 sequences (Dual ADC mode) /3
sequences (Triple ADC mode). Otherwise, the ADC with the shortest sequence may restart
while the ADC with the longest sequence is completing the previous conversions.
Regular conversions can be performed on one or all ADCs. In that case, they are
independent of each other and are interrupted when an injected event occurs. They are
resumed at the end of the injected conversion group.
404/1751 RM0090 Rev 19

Dual ADC mode

At the end of conversion event on ADC1 or ADC2:
• The converted data are stored into the ADC_JDRx registers of each ADC interface.
• A JEOC interrupt is generated (if enabled on one of the two ADC interfaces) when the
ADC1/ADC2’s injected channels have all been converted.
Figure 52. Injected simultaneous mode on 4 channels: dual ADC mode
ADC1 CH0 CH1 CH2 CH3 ... CH15

Trigger End of conversion on ADC1 and ADC2

Sampling
Conversion
ai16054
Triple ADC mode

At the end of conversion event on ADC1, ADC2 or ADC3:
• The converted data are stored into the ADC_JDRx registers of each ADC interface.
• A JEOC interrupt is generated (if enabled on one of the three ADC interfaces) when the
ADC1/ADC2/ADC3’s injected channels have all been converted.
Figure 53. Injected simultaneous mode on 4 channels: triple ADC mode

Trigger End of conversion on ADC1, ADC2 and ADC3

Sampling
Conversion
ai16055
13.9.2 Regular simultaneous mode

This mode is performed on a regular group of channels. The external trigger source comes
from the regular group multiplexer of ADC1 (selected by the EXTSEL[3:0] bits in the
ADC1_CR2 register). A simultaneous trigger is provided to ADC2 and ADC3.
Note: Do not convert the same channel on the two/three ADCs (no overlapping sampling times for
the two/three ADCs when converting the same channel).
In regular simultaneous mode, one must convert sequences with the same length or ensure
that the interval between triggers is longer than the long conversion time of the 2 sequences
(Dual ADC mode) /3 sequences (Triple ADC mode). Otherwise, the ADC with the shortest
sequence may restart while the ADC with the longest sequence is completing the previous
conversions.
Injected conversions must be disabled.
RM0090 Rev 19 405/1751

432
Dual ADC mode

At the end of conversion event on ADC1 or ADC2:
• A 32-bit DMA transfer request is generated (if DMA[1:0] bits in the ADC_CCR register
are equal to 0b10). This request transfers the ADC2 converted data stored in the upper
half-word of the ADC_CDR 32-bit register to the SRAM and then the ADC1 converted
data stored in the lower half-word of ADC_CDR to the SRAM.
• An EOC interrupt is generated (if enabled on one of the two ADC interfaces) when the
ADC1/ADC2’s regular channels have all been converted.
Figure 54. Regular simultaneous mode on 16 channels: dual ADC mode

Trigger End of conversion on ADC1 and ADC2

Sampling
Conversion
ai16054
Triple ADC mode

At the end of conversion event on ADC1, ADC2 or ADC3:
• Three 32-bit DMA transfer requests are generated (if DMA[1:0] bits in the ADC_CCR
register are equal to 0b01). Three transfers then take place from the ADC_CDR 32-bit
register to SRAM: first the ADC1 converted data, then the ADC2 converted data and
finally the ADC3 converted data. The process is repeated for each new three
conversions.
• An EOC interrupt is generated (if enabled on one of the three ADC interfaces) when the
ADC1/ADC2/ADC3’s regular channels are have all been converted.
Figure 55. Regular simultaneous mode on 16 channels: triple ADC mode

Trigger End of conversion on ADC1, ADC2 and ADC3

Sampling
Conversion
ai16055
406/1751 RM0090 Rev 19

13.9.3 Interleaved mode

This mode can be started only on a regular group (usually one channel). The external
trigger source comes from the regular channel multiplexer of ADC1.
Dual ADC mode

After an external trigger occurs:
• ADC1 starts immediately
• ADC2 starts after a delay of several ADC clock cycles
The minimum delay which separates 2 conversions in interleaved mode is configured in the
DELAY bits in the ADC_CCR register. However, an ADC cannot start a conversion if the
complementary ADC is still sampling its input (only one ADC can sample the input signal at
a given time). In this case, the delay becomes the sampling time + 2 ADC clock cycles. For
instance, if DELAY = 5 clock cycles and the sampling takes 15 clock cycles on both ADCs,
then 17 clock cycles will separate conversions on ADC1 and ADC2).
If the CONT bit is set on both ADC1 and ADC2, the selected regular channels of both ADCs
are continuously converted.
Note: If the conversion sequence is interrupted (for instance when DMA end of transfer occurs),
the multi-ADC sequencer must be reset by configuring it in independent mode first (bits
DUAL[4:0] = 00000) before reprogramming the interleaved mode.
After an EOC interrupt is generated by ADC2 (if enabled through the EOCIE bit) a 32-bit
DMA transfer request is generated (if the DMA[1:0] bits in ADC_CCR are equal to 0b10).
This request first transfers the ADC2 converted data stored in the upper half-word of the
ADC_CDR 32-bit register into SRAM, then the ADC1 converted data stored in the register’s
lower half-word into SRAM.
Figure 56. Interleaved mode on 1 channel in continuous conversion mode: dual ADC
mode
End of conversion on ADC1 Sampling
CH0 ... CH0 Conversion

ADC1
ADC2 CH0 ... CH0
Trigger
End of conversion on ADC2
8 ADCCLK
cycles ai16056
Triple ADC mode

After an external trigger occurs:
• ADC1 starts immediately and
• ADC2 starts after a delay of several ADC clock cycles
• ADC3 starts after a delay of several ADC clock cycles referred to the ADC2 conversion
The minimum delay which separates 2 conversions in interleaved mode is configured in the
DELAY bits in the ADC_CCR register. However, an ADC cannot start a conversion if the
complementary ADC is still sampling its input (only one ADC can sample the input signal at
RM0090 Rev 19 407/1751

432
a given time). In this case, the delay becomes the sampling time + 2 ADC clock cycles. For
instance, if DELAY = 5 clock cycles and the sampling takes 15 clock cycles on the three
ADCs, then 17 clock cycles will separate the conversions on ADC1, ADC2 and ADC3).
If the CONT bit is set on ADC1, ADC2 and ADC3, the selected regular channels of all ADCs
are continuously converted.
Note: If the conversion sequence is interrupted (for instance when DMA end of transfer occurs),
In this mode a DMA request is generated each time 2 data items are available, (if the
DMA[1:0] bits in the ADC_CCR register are equal to 0b10). The request first transfers the
first converted data stored in the lower half-word of the ADC_CDR 32-bit register to SRAM,
then it transfers the second converted data stored in ADC_CDR’s upper half-word to SRAM.
The sequence is the following:
• 1st request: ADC_CDR[31:0] = ADC2_DR[15:0] | ADC1_DR[15:0]
• 2nd request: ADC_CDR[31:0] = ADC1_DR[15:0] | ADC3_DR[15:0]
• 3rd request: ADC_CDR[31:0] = ADC3_DR[15:0] | ADC2_DR[15:0]
• 4th request: ADC_CDR[31:0] = ADC2_DR[15:0] | ADC1_DR[15:0], ...
Figure 57. Interleaved mode on 1 channel in continuous conversion mode: triple ADC
mode
DMA request every 2 conversions
ADC1 CH0 CH0 ... CH0

ADC2 CH0 CH0 ... CH0
ADC3 CH0 CH0 ... CH0 Sampling
Conversion
Trigger
6 ADCCLK
cycles ai16058
13.9.4 Alternate trigger mode

This mode can be started only on an injected group. The source of external trigger comes
from the injected group multiplexer of ADC1.
Note: Regular conversions can be enabled on one or all ADCs. In this case the regular
conversions are independent of each other. A regular conversion is interrupted when the
408/1751 RM0090 Rev 19

ADC has to perform an injected conversion. It is resumed when the injected conversion is
finished.
If the conversion sequence is interrupted (for instance when DMA end of transfer occurs),
The time interval between 2 trigger events must be greater than or equal to 1 ADC clock
period. The minimum time interval between 2 trigger events that start conversions on the
same ADC is the same as in the single ADC mode.
Dual ADC mode

• When the 1st trigger occurs, all injected ADC1 channels in the group are converted
• When the 2nd trigger occurs, all injected ADC2 channels in the group are converted
• and so on
A JEOC interrupt, if enabled, is generated after all injected ADC1 channels in the group
have been converted.
If another external trigger occurs after all injected channels in the group have been
converted then the alternate trigger process restarts by converting the injected ADC1
channels in the group.
Figure 58. Alternate trigger: injected group of each ADC
1st trigger 3rd trigger (n)th trigger

EOC, JEOC EOC, JEOC
on ADC1 on ADC1
ADC1 ...
ADC2
EOC, JEOC EOC, JEOC

on ADC2 on ADC2 Sampling
2nd trigger 4th trigger (n+1)th trigger Conversion
ai16059
If the injected discontinuous mode is enabled for both ADC1 and ADC2:
• When the 1st trigger occurs, the first injected ADC1 channel is converted.
• When the 2nd trigger occurs, the first injected ADC2 channel are converted
• and so on
converted then the alternate trigger process restarts.
RM0090 Rev 19 409/1751

432
Figure 59. Alternate trigger: 4 injected channels (each ADC) in discontinuous mode
1st trigger 3rd trigger 5th trigger 7th trigger
Sampling
JEOC on ADC1
Conversion
ADC1
ADC2
JEOC on ADC2
2nd trigger 4th trigger 6th trigger 8th trigger ai16060
Triple ADC mode

• When the 1st trigger occurs, all injected ADC1 channels in the group are converted.
• When the 2nd trigger occurs, all injected ADC2 channels in the group are converted.
• When the 3rd trigger occurs, all injected ADC3 channels in the group are converted.
• and so on
converted then the alternate trigger process restarts by converting the injected ADC1
channels in the group.
Figure 60. Alternate trigger: injected group of each ADC

1st trigger 4th trigger (n)th trigger
EOC, JEOC EOC, JEOC Sampling
on ADC1 on ADC1 Conversion
ADC1 ...
ADC2
2nd trigger
3rd trigger (n+1)th trigger
5th trigger (n+2)th trigger

EOC, JEOC EOC, JEOC
on ADC2 on ADC3
ai16061
13.9.5 Combined regular/injected simultaneous mode

It is possible to interrupt the simultaneous conversion of a regular group to start the
simultaneous conversion of an injected group.
Note: In combined regular/injected simultaneous mode, one must convert sequences with the
same length or ensure that the interval between triggers is longer than the long conversion
time of the 2 sequences (Dual ADC mode) /3 sequences (Triple ADC mode). Otherwise, the
410/1751 RM0090 Rev 19

ADC with the shortest sequence may restart while the ADC with the longest sequence is
completing the previous conversions.
13.9.6 Combined regular simultaneous + alternate trigger mode

It is possible to interrupt the simultaneous conversion of a regular group to start the alternate
trigger conversion of an injected group. Figure 61 shows the behavior of an alternate trigger
interrupting a simultaneous regular conversion.
The injected alternate conversion is immediately started after the injected event. If regular
conversion is already running, in order to ensure synchronization after the injected
conversion, the regular conversion of all (master/slave) ADCs is stopped and resumed
synchronously at the end of the injected conversion.
Note: In combined regular simultaneous + alternate trigger mode, one must convert sequences
with the same length or ensure that the interval between triggers is longer than the long
conversion time of the 2 sequences (Dual ADC mode) /3 sequences (Triple ADC mode).
Otherwise, the ADC with the shortest sequence may restart while the ADC with the longest
sequence is completing the previous conversions.
If the conversion sequence is interrupted (for instance when DMA end of transfer occurs),
Figure 61. Alternate + regular simultaneous

1st trigger
ADC1 reg CH0 CH1 CH2 CH2 CH3 CH3 CH4

CH0
ADC1 inj
CH0
ADC2 inj
synchro not lost
2nd trigger ai16062
If a trigger occurs during an injected conversion that has interrupted a regular conversion, it
is ignored. Figure 62 shows the behavior in this case (2nd trigger is ignored).
RM0090 Rev 19 411/1751

432
Figure 62. Case of trigger occurring during injected conversion

1st trigger 3rd trigger

CH0 CH0
ADC1 inj
CH0
ADC2 inj
2nd trigger 2nd trigger

ai16063
13.10 Temperature sensor

The temperature sensor can be used to measure the junction temperature (TJ) of the
device.
• On STM32F40x and STM32F41x devices, the temperature sensor is internally
connected to ADC1_IN16 channel which is used to convert the sensor output voltage
to a digital value.
• On STM32F42x and STM32F43x devices, the temperature sensor is internally
connected to the same input channel, ADC1_IN18, as VBAT: ADC1_IN18 is used to
convert the sensor output voltage or VBAT into a digital value. Only one conversion,
temperature sensor or VBAT, must be selected at a time. When the temperature sensor
and the VBAT conversion are set simultaneously, only the VBAT conversion is
performed.
Figure 63 shows the block diagram of the temperature sensor.
When not in use, the sensor can be put in power down mode.
Note: The TSVREFE bit must be set to enable the conversion of both internal channels: the
ADC1_IN16 or ADC1_IN18 (temperature sensor) and the ADC1_IN17 (VREFINT).
Main features
• Supported temperature range: –40 to 125 °C
• Precision: ±1.5 °C
412/1751 RM0090 Rev 19

Figure 63. Temperature sensor and VREFINT channel block diagram

TSVREFE control bit
Temperature V SENSE
sensor ADC1_IN16/
ADC1_IN18(1)
Address/data bus
converted data
ADC1
VREFINT
Internal
power block ADC1_IN17
MS31830V1
1. VSENSE is input to ADC1_IN16 for the STM23F40x and STM32F41x devices and to ADC1_IN18 for the
STM32F42x and STM32F43x devices.
Reading the temperature

To use the sensor:
3. Select ADC1_IN16 or ADC1_IN18 input channel.
4. Select a sampling time greater than the minimum sampling time specified in the
datasheet.
5. Set the TSVREFE bit in the ADC_CCR register to wake up the temperature sensor
from power down mode
6. Start the ADC conversion by setting the SWSTART bit (or by external trigger)
7. Read the resulting VSENSE data in the ADC data register
8. Calculate the temperature using the following formula:
Temperature (in °C) = {(VSENSE – V25) / Avg_Slope} + 25
Where:
– V25 = VSENSE value for 25° C
– Avg_Slope = average slope of the temperature vs. VSENSE curve (given in mV/°C
or µV/°C)
Refer to the datasheet’s electrical characteristics section for the actual values of V25
and Avg_Slope.
Note: The sensor has a startup time after waking from power down mode before it can output
VSENSE at the correct level. The ADC also has a startup time after power-on, so to minimize
the delay, the ADON and TSVREFE bits should be set at the same time.
The temperature sensor output voltage changes linearly with temperature. The offset of this
linear function depends on each chip due to process variation (up to 45 °C from one chip to
another).
RM0090 Rev 19 413/1751

432
The internal temperature sensor is more suited for applications that detect temperature
variations instead of absolute temperatures. If accurate temperature reading is required, an
external temperature sensor should be used.
13.11 Battery charge monitoring

The VBATE bit in the ADC_CCR register is used to switch to the battery voltage. As the
VBAT voltage could be higher than VDDA, to ensure the correct operation of the ADC, the
VBAT pin is internally connected to a bridge divider.
When the VBATE is set, the bridge is automatically enabled to connect:
• VBAT/2 to the ADC1_IN18 input channel, on STM32F40xx and STM32F41xx devices
• VBAT/4 to the ADC1_IN18 input channel, on STM32F42xx and STM32F43xx devices
Note: On STM32F42xx and STM32F43xx devices, VBAT and temperature sensor are connected
to the same ADC internal channel (ADC1_IN18). Only one conversion, either temperature
sensor or VBAT, must be selected at a time. When both conversion are enabled
simultaneously, only the VBAT conversion is performed.
13.12 ADC interrupts

An interrupt can be produced on the end of conversion for regular and injected groups,
when the analog watchdog status bit is set and when the overrun status bit is set. Separate
interrupt enable bits are available for flexibility.
Two other flags are present in the ADC_SR register, but there is no interrupt associated with
them:
• JSTRT (Start of conversion for channels of an injected group)
• STRT (Start of conversion for channels of a regular group)
Table 70. ADC interrupts

End of conversion of a regular group EOC EOCIE

End of conversion of an injected group JEOC JEOCIE
Analog watchdog status bit is set AWD AWDIE
Overrun OVR OVRIE
414/1751 RM0090 Rev 19

13.13 ADC registers

Refer to Section 1.1: List of abbreviations for registers for registers for a list of abbreviations
used in register descriptions.
The peripheral registers must be written at word level (32 bits). Read accesses can be done
by bytes (8 bits), half-words (16 bits) or words (32 bits).
13.13.1 ADC status register (ADC_SR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OVR STRT JSTRT JEOC EOC AWD
Reserved

Bit 5 OVR: Overrun
This bit is set by hardware when data are lost (either in single mode or in dual/triple mode). It
is cleared by software. Overrun detection is enabled only when DMA = 1 or EOCS = 1.
0: No overrun occurred
1: Overrun has occurred
Bit 4 STRT: Regular channel start flag
This bit is set by hardware when regular channel conversion starts. It is cleared by software.
0: No regular channel conversion started
1: Regular channel conversion has started
Bit 3 JSTRT: Injected channel start flag
This bit is set by hardware when injected group conversion starts. It is cleared by software.
0: No injected group conversion started
1: Injected group conversion has started
Bit 2 JEOC: Injected channel end of conversion
This bit is set by hardware at the end of the conversion of all injected channels in the group.
It is cleared by software.
0: Conversion is not complete
1: Conversion complete
Bit 1 EOC: Regular channel end of conversion
This bit is set by hardware at the end of the conversion of a regular group of channels. It is
cleared by software or by reading the ADC_DR register.
0: Conversion not complete (EOCS=0), or sequence of conversions not complete (EOCS=1)
1: Conversion complete (EOCS=0), or sequence of conversions complete (EOCS=1)
Bit 0 AWD: Analog watchdog flag
This bit is set by hardware when the converted voltage crosses the values programmed in
the ADC_LTR and ADC_HTR registers. It is cleared by software.
0: No analog watchdog event occurred
1: Analog watchdog event occurred
RM0090 Rev 19 415/1751

432
13.13.2 ADC control register 1 (ADC_CR1)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
OVRIE RES AWDEN JAWDEN
Reserved Reserved
rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
JDISCE DISC AWDSG
DISCNUM[2:0] JAUTO SCAN JEOCIE AWDIE EOCIE AWDCH[4:0]
N EN L

Bit 26 OVRIE: Overrun interrupt enable
This bit is set and cleared by software to enable/disable the Overrun interrupt.
0: Overrun interrupt disabled
1: Overrun interrupt enabled. An interrupt is generated when the OVR bit is set.
Bits 25:24 RES[1:0]: Resolution
These bits are written by software to select the resolution of the conversion.
00: 12-bit (15 ADCCLK cycles)
Bit 23 AWDEN: Analog watchdog enable on regular channels
0: Analog watchdog disabled on regular channels
1: Analog watchdog enabled on regular channels
Bit 22 JAWDEN: Analog watchdog enable on injected channels
0: Analog watchdog disabled on injected channels
1: Analog watchdog enabled on injected channels
Bits 15:13 DISCNUM[2:0]: Discontinuous mode channel count
These bits are written by software to define the number of regular channels to be converted
in discontinuous mode, after receiving an external trigger.
000: 1 channel
001: 2 channels
...
111: 8 channels
Bit 12 JDISCEN: Discontinuous mode on injected channels
This bit is set and cleared by software to enable/disable discontinuous mode on the injected
channels of a group.
0: Discontinuous mode on injected channels disabled
1: Discontinuous mode on injected channels enabled
416/1751 RM0090 Rev 19

Bit 11 DISCEN: Discontinuous mode on regular channels

This bit is set and cleared by software to enable/disable Discontinuous mode on regular
channels.
0: Discontinuous mode on regular channels disabled
1: Discontinuous mode on regular channels enabled
Bit 10 JAUTO: Automatic injected group conversion
This bit is set and cleared by software to enable/disable automatic injected group conversion
after regular group conversion.
0: Automatic injected group conversion disabled
1: Automatic injected group conversion enabled
Bit 9 AWDSGL: Enable the watchdog on a single channel in scan mode
This bit is set and cleared by software to enable/disable the analog watchdog on the channel
identified by the AWDCH[4:0] bits.
0: Analog watchdog enabled on all channels
1: Analog watchdog enabled on a single channel
Bit 8 SCAN: Scan mode
This bit is set and cleared by software to enable/disable the Scan mode. In Scan mode, the
inputs selected through the ADC_SQRx or ADC_JSQRx registers are converted.
0: Scan mode disabled
1: Scan mode enabled
Note: An EOC interrupt is generated if the EOCIE bit is set:
– At the end of each regular group sequence if the EOCS bit is cleared to 0
– At the end of each regular channel conversion if the EOCS bit is set to 1
Note: A JEOC interrupt is generated only on the end of conversion of the last channel if the
JEOCIE bit is set.
Bit 7 JEOCIE: Interrupt enable for injected channels
This bit is set and cleared by software to enable/disable the end of conversion interrupt for
injected channels.
0: JEOC interrupt disabled
1: JEOC interrupt enabled. An interrupt is generated when the JEOC bit is set.
Bit 6 AWDIE: Analog watchdog interrupt enable
This bit is set and cleared by software to enable/disable the analog watchdog interrupt.
0: Analog watchdog interrupt disabled
1: Analog watchdog interrupt enabled
Bit 5 EOCIE: Interrupt enable for EOC
This bit is set and cleared by software to enable/disable the end of conversion interrupt.
0: EOC interrupt disabled
1: EOC interrupt enabled. An interrupt is generated when the EOC bit is set.
Bits 4:0 AWDCH[4:0]: Analog watchdog channel select bits
These bits are set and cleared by software. They select the input channel to be guarded by
the analog watchdog.
Note: 00000: ADC analog input Channel0
00001: ADC analog input Channel1
...
Other values reserved
RM0090 Rev 19 417/1751

432
13.13.3 ADC control register 2 (ADC_CR2)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
SWST JSWST
EXTEN EXTSEL[3:0] JEXTEN JEXTSEL[3:0]
reserved ART reserved ART
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ALIGN EOCS DDS DMA CONT ADON
reserved Reserved
rw rw rw rw rw rw

Bit 30 SWSTART: Start conversion of regular channels
This bit is set by software to start conversion and cleared by hardware as soon as the
conversion starts.
0: Reset state
1: Starts conversion of regular channels
Note: This bit can be set only when ADON = 1 otherwise no conversion is launched.
Bits 29:28 EXTEN: External trigger enable for regular channels
These bits are set and cleared by software to select the external trigger polarity and enable
the trigger of a regular group.
00: Trigger detection disabled
01: Trigger detection on the rising edge
10: Trigger detection on the falling edge
11: Trigger detection on both the rising and falling edges
Bits 27:24 EXTSEL[3:0]: External event select for regular group
These bits select the external event used to trigger the start of conversion of a regular group:
0000: Timer 1 CC1 event
0110: Timer 2 TRGO event
1111: EXTI line11
418/1751 RM0090 Rev 19

Bit 22 JSWSTART: Start conversion of injected channels

This bit is set by software and cleared by hardware as soon as the conversion starts.
0: Reset state
1: Starts conversion of injected channels
Note: This bit can be set only when ADON = 1 otherwise no conversion is launched.
Bits 21:20 JEXTEN: External trigger enable for injected channels
These bits are set and cleared by software to select the external trigger polarity and enable
the trigger of an injected group.
00: Trigger detection disabled
01: Trigger detection on the rising edge
10: Trigger detection on the falling edge
11: Trigger detection on both the rising and falling edges
Bits 19:16 JEXTSEL[3:0]: External event select for injected group
These bits select the external event used to trigger the start of conversion of an injected
group.
1111: EXTI line15
Bit 11 ALIGN: Data alignment
This bit is set and cleared by software. Refer to Figure 48 and Figure 49.
0: Right alignment
1: Left alignment
Bit 10 EOCS: End of conversion selection
0: The EOC bit is set at the end of each sequence of regular conversions. Overrun detection
is enabled only if DMA=1.
1: The EOC bit is set at the end of each regular conversion. Overrun detection is enabled.
Bit 9 DDS: DMA disable selection (for single ADC mode)
0: No new DMA request is issued after the last transfer (as configured in the DMA controller)
1: DMA requests are issued as long as data are converted and DMA=1
Bit 8 DMA: Direct memory access mode (for single ADC mode)
This bit is set and cleared by software. Refer to the DMA controller chapter for more details.
0: DMA mode disabled
1: DMA mode enabled
RM0090 Rev 19 419/1751

432

Bit 1 CONT: Continuous conversion
This bit is set and cleared by software. If it is set, conversion takes place continuously until it
is cleared.
0: Single conversion mode
1: Continuous conversion mode
Bit 0 ADON: A/D Converter ON / OFF
Note: 0: Disable ADC conversion and go to power down mode
1: Enable ADC
13.13.4 ADC sample time register 1 (ADC_SMPR1)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
SMP18[2:0] SMP17[2:0] SMP16[2:0] SMP15[2:1]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SMP15_0 SMP14[2:0] SMP13[2:0] SMP12[2:0] SMP11[2:0] SMP10[2:0]

Bits 26:0 SMPx[2:0]: Channel x sampling time selection
These bits are written by software to select the sampling time individually for each channel.
During sampling cycles, the channel selection bits must remain unchanged.
Note: 000: 3 cycles
001: 15 cycles
010: 28 cycles
011: 56 cycles
100: 84 cycles
101: 112 cycles
110: 144 cycles
111: 480 cycles
13.13.5 ADC sample time register 2 (ADC_SMPR2)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
SMP9[2:0] SMP8[2:0] SMP7[2:0] SMP6[2:0] SMP5[2:1]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SMP
SMP4[2:0] SMP3[2:0] SMP2[2:0] SMP1[2:0] SMP0[2:0]
5_0
420/1751 RM0090 Rev 19


Bits 29:0 SMPx[2:0]: Channel x sampling time selection
These bits are written by software to select the sampling time individually for each channel.
During sample cycles, the channel selection bits must remain unchanged.
Note: 000: 3 cycles
001: 15 cycles
010: 28 cycles
011: 56 cycles
100: 84 cycles
101: 112 cycles
110: 144 cycles
111: 480 cycles
13.13.6 ADC injected channel data offset register x (ADC_JOFRx) (x=1..4)

Address offset: 0x14-0x20
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
JOFFSETx[11:0]
Reserved

Bits 11:0 JOFFSETx[11:0]: Data offset for injected channel x
These bits are written by software to define the offset to be subtracted from the raw
converted data when converting injected channels. The conversion result can be read from
in the ADC_JDRx registers.
13.13.7 ADC watchdog higher threshold register (ADC_HTR)

Reset value: 0x0000 0FFF
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
HT[11:0]
Reserved

Bits 11:0 HT[11:0]: Analog watchdog higher threshold
These bits are written by software to define the higher threshold for the analog watchdog.
RM0090 Rev 19 421/1751

432
Note: The software can write to these registers when an ADC conversion is ongoing. The
programmed value will be effective when the next conversion is complete. Writing to this
register is performed with a write delay that can create uncertainty on the effective time at
which the new value is programmed.
13.13.8 ADC watchdog lower threshold register (ADC_LTR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
LT[11:0]
Reserved

Bits 11:0 LT[11:0]: Analog watchdog lower threshold
These bits are written by software to define the lower threshold for the analog watchdog.
Note: The software can write to these registers when an ADC conversion is ongoing. The
programmed value will be effective when the next conversion is complete. Writing to this
register is performed with a write delay that can create uncertainty on the effective time at
which the new value is programmed.
13.13.9 ADC regular sequence register 1 (ADC_SQR1)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
L[3:0] SQ16[4:1]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SQ16_0 SQ15[4:0] SQ14[4:0] SQ13[4:0]

Bits 23:20 L[3:0]: Regular channel sequence length
These bits are written by software to define the total number of conversions in the regular
channel conversion sequence.
0000: 1 conversion
0001: 2 conversions
...
1111: 16 conversions
Bits 19:15 SQ16[4:0]: 16th conversion in regular sequence
These bits are written by software with the channel number (0..18) assigned as the 16th in
the conversion sequence.
422/1751 RM0090 Rev 19



31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
SQ12[4:0] SQ11[4:0] SQ10[4:1]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SQ10_0 SQ9[4:0] SQ8[4:0] SQ7[4:0]

These bits are written by software with the channel number (0..18) assigned as the 12th in
the sequence to be converted.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
SQ6[4:0] SQ5[4:0] SQ4[4:1]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SQ4_0 SQ3[4:0] SQ2[4:0] SQ1[4:0]

These bits are written by software with the channel number (0..18) assigned as the 6th in the
sequence to be converted.
RM0090 Rev 19 423/1751

432

Bits 14:10 SQ3[4:0]: 3rd conversion in regular sequence
Bits 9:5 SQ2[4:0]: 2nd conversion in regular sequence
Bits 4:0 SQ1[4:0]: 1st conversion in regular sequence
13.13.12 ADC injected sequence register (ADC_JSQR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
JL[1:0] JSQ4[4:1]
Reserved
rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
JSQ4[0] JSQ3[4:0] JSQ2[4:0] JSQ1[4:0]

Bits 21:20 JL[1:0]: Injected sequence length
These bits are written by software to define the total number of conversions in the injected
channel conversion sequence.
00: 1 conversion
01: 2 conversions
10: 3 conversions
11: 4 conversions
Bits 19:15 JSQ4[4:0]: 4th conversion in injected sequence (when JL[1:0]=3, see note below)
These bits are written by software with the channel number (0..18) assigned as the 4th in the
sequence to be converted.
Bits 14:10 JSQ3[4:0]: 3rd conversion in injected sequence (when JL[1:0]=3, see note below)
Bits 9:5 JSQ2[4:0]: 2nd conversion in injected sequence (when JL[1:0]=3, see note below)
Bits 4:0 JSQ1[4:0]: 1st conversion in injected sequence (when JL[1:0]=3, see note below)
Note: When JL[1:0]=3 (4 injected conversions in the sequencer), the ADC converts the channels
in the following order: JSQ1[4:0], JSQ2[4:0], JSQ3[4:0], and JSQ4[4:0].
When JL=2 (3 injected conversions in the sequencer), the ADC converts the channels in the
following order: JSQ2[4:0], JSQ3[4:0], and JSQ4[4:0].
When JL=1 (2 injected conversions in the sequencer), the ADC converts the channels in
starting from JSQ3[4:0], and then JSQ4[4:0].
When JL=0 (1 injected conversion in the sequencer), the ADC converts only JSQ4[4:0]
channel.
424/1751 RM0090 Rev 19

13.13.13 ADC injected data register x (ADC_JDRx) (x= 1..4)

Address offset: 0x3C - 0x48
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
JDATA[15:0]

Bits 15:0 JDATA[15:0]: Injected data
These bits are read-only. They contain the conversion result from injected channel x. The
data are left -or right-aligned as shown in Figure 48 and Figure 49.
13.13.14 ADC regular data register (ADC_DR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DATA[15:0]

Bits 15:0 DATA[15:0]: Regular data
These bits are read-only. They contain the conversion result from the regular
channels. The data are left- or right-aligned as shown in Figure 48 and
Figure 49.
RM0090 Rev 19 425/1751

432
13.13.15 ADC Common status register (ADC_CSR)

Address offset: 0x00 (this offset address is relative to ADC1 base address + 0x300)
This register provides an image of the status bits of the different ADCs. Nevertheless it is
read-only and does not allow to clear the different status bits. Instead each status bit must
be cleared by writing it to 0 in the corresponding ADC_SR register.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
OVR3 STRT3 JSTRT3 JEOC 3 EOC3 AWD3
Reserved ADC3
r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
JSTRT
OVR2 STRT2 JEOC2 EOC2 AWD2 OVR1 STRT1 JSTRT1 JEOC 1 EOC1 AWD1
2
Reserved Reserved
ADC2 ADC1
r r r r r r r r r r r r

Bit 21 OVR3: Overrun flag of ADC3
This bit is a copy of the OVR bit in the ADC3_SR register.
Bit 20 STRT3: Regular channel Start flag of ADC3
This bit is a copy of the STRT bit in the ADC3_SR register.
Bit 19 JSTRT3: Injected channel Start flag of ADC3
This bit is a copy of the JSTRT bit in the ADC3_SR register.
Bit 18 JEOC3: Injected channel end of conversion of ADC3
This bit is a copy of the JEOC bit in the ADC3_SR register.
Bit 17 EOC3: End of conversion of ADC3
This bit is a copy of the EOC bit in the ADC3_SR register.
Bit 16 AWD3: Analog watchdog flag of ADC3
This bit is a copy of the AWD bit in the ADC3_SR register.
426/1751 RM0090 Rev 19


13.13.16 ADC common control register (ADC_CCR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
TSVREFE VBATE ADCPRE
Reserved Reserved
rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DMA[1:0] DDS DELAY[3:0] MULTI[4:0]
Res. Reserved

Bit 23 TSVREFE: Temperature sensor and VREFINT enable
This bit is set and cleared by software to enable/disable the temperature sensor and the
VREFINT channel.
0: Temperature sensor and VREFINT channel disabled
1: Temperature sensor and VREFINT channel enabled
Note: On STM32F42x and STM32F43x devices, VBATE must be disabled when TSVREFE is
set. If both bits are set, only the VBAT conversion is performed.
Bit 22 VBATE: VBAT enable
This bit is set and cleared by software to enable/disable the VBAT channel.
0: VBAT channel disabled
1: VBAT channel enabled
RM0090 Rev 19 427/1751

432
Bits 17:16 ADCPRE: ADC prescaler

Set and cleared by software to select the frequency of the clock to the ADC. The clock is
common for all the ADCs.
Note: 00: PCLK2 divided by 2
01: PCLK2 divided by 4
Bits 15:14 DMA: Direct memory access mode for multi ADC mode
This bit-field is set and cleared by software. Refer to the DMA controller section for more
details.
00: DMA mode disabled
01: DMA mode 1 enabled (2 / 3 half-words one by one - 1 then 2 then 3)
10: DMA mode 2 enabled (2 / 3 half-words by pairs - 2&1 then 1&3 then 3&2)
11: DMA mode 3 enabled (2 / 3 bytes by pairs - 2&1 then 1&3 then 3&2)
Bit 13 DDS: DMA disable selection (for multi-ADC mode)
0: No new DMA request is issued after the last transfer (as configured in the DMA
controller). DMA bits are not cleared by hardware, however they must have been cleared
and set to the wanted mode by software before new DMA requests can be generated.
1: DMA requests are issued as long as data are converted and DMA = 01, 10 or 11.
428/1751 RM0090 Rev 19

Bit 11:8 DELAY: Delay between 2 sampling phases

Set and cleared by software. These bits are used in dual or triple interleaved modes.
0000: 5 * TADCCLK
0001: 6 * TADCCLK
0010: 7 * TADCCLK
...
1111: 20 * TADCCLK
Bits 4:0 MULTI[4:0]: Multi ADC mode selection
These bits are written by software to select the operating mode.
– All the ADCs independent:
00000: Independent mode
– 00001 to 01001: Dual mode, ADC1 and ADC2 working together, ADC3 is independent
00001: Combined regular simultaneous + injected simultaneous mode
00010: Combined regular simultaneous + alternate trigger mode
00011: Reserved
00101: Injected simultaneous mode only
00110: Regular simultaneous mode only
00111: interleaved mode only
01001: Alternate trigger mode only
– 10001 to 11001: Triple mode: ADC1, 2 and 3 working together
10001: Combined regular simultaneous + injected simultaneous mode
10010: Combined regular simultaneous + alternate trigger mode
10011: Reserved
10101: Injected simultaneous mode only
10110: Regular simultaneous mode only
10111: interleaved mode only
11001: Alternate trigger mode only
All other combinations are reserved and must not be programmed
Note: In multi mode, a change of channel configuration generates an abort that can cause a
loss of synchronization. It is recommended to disable the multi ADC mode before any
configuration change.
RM0090 Rev 19 429/1751

432
13.13.17 ADC common regular data register for dual and triple modes
(ADC_CDR)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DATA2[15:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DATA1[15:0]
Bits 31:16 DATA2[15:0]: 2nd data item of a pair of regular conversions

– In dual mode, these bits contain the regular data of ADC2. Refer to Dual ADC mode.
– In triple mode, these bits contain alternatively the regular data of ADC2, ADC1 and ADC3.
Refer to Triple ADC mode.
Bits 15:0 DATA1[15:0]: 1st data item of a pair of regular conversions
– In dual mode, these bits contain the regular data of ADC1. Refer to Dual ADC mode
– In triple mode, these bits contain alternatively the regular data of ADC1, ADC3 and ADC2.
Refer to Triple ADC mode.
13.13.18 ADC register map

The following table summarizes the ADC registers.
Table 71. ADC global register map

Offset Register
0x000 - 0x04C ADC1

0x050 - 0x0FC Reserved
0x100 - 0x14C ADC2
0x200 - 0x24C ADC3
0x300 - 0x308 Common registers
430/1751 RM0090 Rev 19

Table 72. ADC register map and reset values for each ADC
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
JSTRT
JEOC
STRT
AWD
OVR
EOC
ADC_SR
0x00 Reserved
AWD SGL
JDISCEN
JAWDEN
RES[1:0]
DISCEN
AWDEN
JEOCIE
JAUTO
AWDIE
OVRIE
EOCIE
SCAN
DISC
ADC_CR1 AWDCH[4:0]
0x04 Reserved Reserved NUM [2:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
JEXTEN[1:0]
JSWSTART
EXTEN[1:0]
Re Re
SWSTART
ALIGN
ADON
CONT
EOCS
se se JEXTSEL
DMA
DDS
ADC_CR2 EXTSEL [3:0]
0x08 rv rv [3:0] Reserved Reserved
ed ed
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ADC_SMPR1 Sample time bits SMPx_x
0x0C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ADC_SMPR2 Sample time bits SMPx_x
0x10
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ADC_JOFR1 JOFFSET1[11:0]
0x14 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0
0x18 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0
0x1C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0
0x20 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0
ADC_HTR HT[11:0]
0x24 Reserved
Reset value 1 1 1 1 1 1 1 1 1 1 1 1
ADC_LTR LT[11:0]
0x28 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0
ADC_SQR1 L[3:0] Regular channel sequence SQx_x bits
0x2C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ADC_SQR2 Regular channel sequence SQx_x bits
Reserved Reserved
0x30
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ADC_SQR3 Regular channel sequence SQx_x bits

0x34
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ADC_JSQR JL[1:0] Injected channel sequence JSQx_x bits

0x38 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ADC_JDR1 JDATA[15:0]
0x3C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x40 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x44 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x48 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ADC_DR Regular DATA[15:0]
0x4C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RM0090 Rev 19 431/1751

432
Table 73. ADC register map and reset values (common ADC registers)
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
JSTRT
JSTRT
JSTRT
JEOC
JEOC
JEOC
STRT
STRT
STRT
AWD
AWD
AWD
OVR
EOC
OVR
EOC
OVR
EOC
ADC_CSR
Reserved
Reserved
0x00 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ADC3 ADC2 ADC1
ADCPRE[1:0]
TSVREFE
DMA[1:0]
VBATE
Reserved
DDS
ADC_CCR DELAY [3:0] MULTI [4:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ADC_CDR Regular DATA2[15:0] Regular DATA1[15:0]
0x08
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
432/1751 RM0090 Rev 19

RM0090 Digital-to-analog converter (DAC)
14 Digital-to-analog converter (DAC)
14.1 DAC introduction

The DAC module is a 12-bit, voltage output digital-to-analog converter. The DAC can be
configured in 8- or 12-bit mode and may be used in conjunction with the DMA controller. In
12-bit mode, the data could be left- or right-aligned. The DAC has two output channels, each
with its own converter. In dual DAC channel mode, conversions could be done
independently or simultaneously when both channels are grouped together for synchronous
update operations. An input reference pin, VREF+ (shared with ADC) is available for better
resolution.
14.2 DAC main features

• Two DAC converters: one output channel each
• Left or right data alignment in 12-bit mode
• Synchronized update capability
• Noise-wave generation
• Triangular-wave generation
• Dual DAC channel for independent or simultaneous conversions
• DMA capability for each channel
• DMA underrun error detection
• External triggers for conversion
• Input voltage reference, VREF+
Figure 64 shows the block diagram of a DAC channel and Table 74 gives the pin
description.
RM0090 Rev 19 433/1751

454
Digital-to-analog converter (DAC) RM0090
Figure 64. DAC channel block diagram
DAC control register

TSELx[2:0] bits
SWTR IGx
Trigger selectorx
TIM2_T RGO
TIM4_T RGO DMAENx
TIM5_T RGO
TIM6_T RGO
TIM7_T RGO
TIM8_T RGO
EXTI_9
DM A req ue stx
Control logicx TENx
12-bit
DHRx
LFSRx trianglex MAMPx[3:0] bits
WAVENx[1:0] bits
12-bit
DORx
12-bit
VDDA
Digital-to-analog DAC_ OU Tx
VSSA
converterx
VR EF+
ai14708b
Table 74. DAC pins

Name Signal type Remarks
Input, analog reference The higher/positive reference voltage for the DAC,
VREF+
positive 1.8 V ≤ VREF+ ≤ VDDA
VDDA Input, analog supply Analog power supply
VSSA Input, analog supply ground Ground for analog power supply
DAC_OUTx Analog output signal DAC channelx analog output
Note: Once the DAC channelx is enabled, the corresponding GPIO pin (PA4 or PA5) is
automatically connected to the analog converter output (DAC_OUTx). In order to avoid
parasitic consumption, the PA4 or PA5 pin should first be configured to analog (AIN).
434/1751 RM0090 Rev 19

14.3 DAC functional description
14.3.1 DAC channel enable

Each DAC channel can be powered on by setting its corresponding ENx bit in the DAC_CR
register. The DAC channel is then enabled after a startup time tWAKEUP.
Note: The ENx bit enables the analog DAC Channelx macrocell only. The DAC Channelx digital
interface is enabled even if the ENx bit is reset.
14.3.2 DAC output buffer enable

The DAC integrates two output buffers that can be used to reduce the output impedance,
and to drive external loads directly without having to add an external operational amplifier.
Each DAC channel output buffer can be enabled and disabled using the corresponding
BOFFx bit in the DAC_CR register.
14.3.3 DAC data format

Depending on the selected configuration mode, the data have to be written into the specified
register as described below:
• Single DAC channelx, there are three possibilities:
– 8-bit right alignment: the software has to load data into the DAC_DHR8Rx [7:0]
bits (stored into the DHRx[11:4] bits)
– 12-bit left alignment: the software has to load data into the DAC_DHR12Lx [15:4]
– 12-bit right alignment: the software has to load data into the DAC_DHR12Rx [11:0]
Depending on the loaded DAC_DHRyyyx register, the data written by the user is shifted and
stored into the corresponding DHRx (data holding registerx, which are internal non-memory-
mapped registers). The DHRx register is then loaded into the DORx register either
automatically, by software trigger or by an external event trigger.
RM0090 Rev 19 435/1751

454
Figure 65. Data registers in single DAC channel mode
31 24 15 7 0
8-bit right aligned
12-bit left aligned
12-bit right aligned
ai14710b
• Dual DAC channels, there are three possibilities:
– 8-bit right alignment: data for DAC channel1 to be loaded into the DAC_DHR8RD
[7:0] bits (stored into the DHR1[11:4] bits) and data for DAC channel2 to be loaded
into the DAC_DHR8RD [15:8] bits (stored into the DHR2[11:4] bits)
– 12-bit left alignment: data for DAC channel1 to be loaded into the DAC_DHR12LD
[15:4] bits (stored into the DHR1[11:0] bits) and data for DAC channel2 to be
loaded into the DAC_DHR12LD [31:20] bits (stored into the DHR2[11:0] bits)
– 12-bit right alignment: data for DAC channel1 to be loaded into the
DAC_DHR12RD [11:0] bits (stored into the DHR1[11:0] bits) and data for DAC
channel2 to be loaded into the DAC_DHR12LD [27:16] bits (stored into the
DHR2[11:0] bits)
Depending on the loaded DAC_DHRyyyD register, the data written by the user is shifted
and stored into DHR1 and DHR2 (data holding registers, which are internal non-memory-
mapped registers). The DHR1 and DHR2 registers are then loaded into the DOR1 and
DOR2 registers, respectively, either automatically, by software trigger or by an external
event trigger.
Figure 66. Data registers in dual DAC channel mode

31 3124 24 15 15 77 0 0
8-bit right aligned
8-bit right aligned
12-bit left aligned
12-bit left aligned

ai14709
ai14709b
14.3.4 DAC conversion

The DAC_DORx cannot be written directly and any data transfer to the DAC channelx must
be performed by loading the DAC_DHRx register (write to DAC_DHR8Rx, DAC_DHR12Lx,
DAC_DHR12Rx, DAC_DHR8RD, DAC_DHR12LD or DAC_DHR12LD).
Data stored in the DAC_DHRx register are automatically transferred to the DAC_DORx
register after one APB1 clock cycle, if no hardware trigger is selected (TENx bit in DAC_CR
register is reset). However, when a hardware trigger is selected (TENx bit in DAC_CR
register is set) and a trigger occurs, the transfer is performed three APB1 clock cycles later.
436/1751 RM0090 Rev 19

When DAC_DORx is loaded with the DAC_DHRx contents, the analog output voltage
becomes available after a time tSETTLING that depends on the power supply voltage and the
analog output load.
Figure 67. Timing diagram for conversion with trigger disabled TEN = 0
APB1_CLK
DHR 0x1AC
Output voltage
DOR 0x1AC available on DAC_OUT pin
tSETTLING
ai14711c
14.3.5 DAC output voltage

Digital inputs are converted to output voltages on a linear conversion between 0 and VREF+.
The analog output voltages on each DAC channel pin are determined by the following
equation:
DOR
DACoutput = V REF × --------------
4096
14.3.6 DAC trigger selection

If the TENx control bit is set, conversion can then be triggered by an external event (timer
counter, external interrupt line). The TSELx[2:0] control bits determine which out of 8
possible events will trigger conversion as shown in Table 75.
Table 75. External triggers

Source Type TSEL[2:0]
Timer 6 TRGO event 000

Timer 7 TRGO event Internal signal from on-chip 010
Timer 5 TRGO event timers 011
SWTRIG Software control bit 111
Each time a DAC interface detects a rising edge on the selected timer TRGO output, or on
the selected external interrupt line 9, the last data stored into the DAC_DHRx register are
transferred into the DAC_DORx register. The DAC_DORx register is updated three APB1
cycles after the trigger occurs.
RM0090 Rev 19 437/1751

454
If the software trigger is selected, the conversion starts once the SWTRIG bit is set.
SWTRIG is reset by hardware once the DAC_DORx register has been loaded with the
DAC_DHRx register contents.
Note: TSELx[2:0] bit cannot be changed when the ENx bit is set.
When software trigger is selected, the transfer from the DAC_DHRx register to the
DAC_DORx register takes only one APB1 clock cycle.
14.3.7 DMA request

Each DAC channel has a DMA capability. Two DMA channels are used to service DAC
channel DMA requests.
A DAC DMA request is generated when an external trigger (but not a software trigger)
occurs while the DMAENx bit is set. The value of the DAC_DHRx register is then transferred
into the DAC_DORx register.
In dual mode, if both DMAENx bits are set, two DMA requests are generated. If only one
DMA request is needed, the user should set only the corresponding DMAENx bit. In this
way, the application can manage both DAC channels in dual mode by using one DMA
request and a unique DMA channel.
DMA underrun
The DAC DMA request is not queued so that if a second external trigger arrives before the
acknowledgement for the first external trigger is received (first request), then no new
request is issued and the DMA channelx underrun flag DMAUDRx in the DAC_SR register
is set, reporting the error condition. DMA data transfers are then disabled and no further
DMA request is treated. The DAC channelx continues to convert old data.
The software should clear the DMAUDRx flag by writing “1”, clear the DMAEN bit of the
used DMA stream and re-initialize both DMA and DAC channelx to restart the transfer
correctly. The software should modify the DAC trigger conversion frequency or lighten the
DMA workload to avoid a new DMA underrun. Finally, the DAC conversion could be
resumed by enabling both DMA data transfer and conversion trigger.
For each DAC channelx, an interrupt is also generated if its corresponding DMAUDRIEx bit
in the DAC_CR register is enabled.
14.3.8 Noise generation

In order to generate a variable-amplitude pseudonoise, an LFSR (linear feedback shift
register) is available. DAC noise generation is selected by setting WAVEx[1:0] to “01”. The
preloaded value in LFSR is 0xAAA. This register is updated three APB1 clock cycles after
each trigger event, following a specific calculation algorithm.
438/1751 RM0090 Rev 19

Figure 68. DAC LFSR register calculation algorithm
XOR
X6 X4 X X0
X 12
11 10 9 8 7 6 5 4 3 2 1 0
12
NOR
ai14713c
The LFSR value, that may be masked partially or totally by means of the MAMPx[3:0] bits in
the DAC_CR register, is added up to the DAC_DHRx contents without overflow and this
value is then stored into the DAC_DORx register.
If LFSR is 0x0000, a ‘1 is injected into it (antilock-up mechanism).
It is possible to reset LFSR wave generation by resetting the WAVEx[1:0] bits.
Figure 69. DAC conversion (SW trigger enabled) with LFSR wave generation
APB1_CLK
DHR 0x00
DOR 0xAAA 0xD55
SWTRIG
ai14714b
Note: The DAC trigger must be enabled for noise generation by setting the TENx bit in the
DAC_CR register.
14.3.9 Triangle-wave generation

It is possible to add a small-amplitude triangular waveform on a DC or slowly varying signal.
DAC triangle-wave generation is selected by setting WAVEx[1:0] to “10”. The amplitude is
configured through the MAMPx[3:0] bits in the DAC_CR register. An internal triangle counter
is incremented three APB1 clock cycles after each trigger event. The value of this counter is
then added to the DAC_DHRx register without overflow and the sum is stored into the
DAC_DORx register. The triangle counter is incremented as long as it is less than the
maximum amplitude defined by the MAMPx[3:0] bits. Once the configured amplitude is
reached, the counter is decremented down to 0, then incremented again and so on.
RM0090 Rev 19 439/1751

454
It is possible to reset triangle wave generation by resetting the WAVEx[1:0] bits.
Figure 70. DAC triangle wave generation
MAMPx[3:0] max amplitude

+ DAC_DHRx base value
De
n
tio
c
re
ta
en
m
en
em
ta
cr
tio
In
n
DAC_DHRx base value
0
ai14715c
Figure 71. DAC conversion (SW trigger enabled) with triangle wave generation
APB1_CLK
DHR 0x00
DOR 0xAAA 0xD55
SWTRIG
ai14714b
Note: The DAC trigger must be enabled for noise generation by setting the TENx bit in the
DAC_CR register.
The MAMPx[3:0] bits must be configured before enabling the DAC, otherwise they cannot
be changed.
14.4 Dual DAC channel conversion

To efficiently use the bus bandwidth in applications that require the two DAC channels at the
same time, three dual registers are implemented: DHR8RD, DHR12RD and DHR12LD. A
unique register access is then required to drive both DAC channels at the same time.
Eleven possible conversion modes are possible using the two DAC channels and these dual
registers. All the conversion modes can nevertheless be obtained using separate DHRx
registers if needed.
All modes are described in the paragraphs below.
440/1751 RM0090 Rev 19

14.4.1 Independent trigger without wave generation

To configure the DAC in this conversion mode, the following sequence is required:
• Set the two DAC channel trigger enable bits TEN1 and TEN2
• Configure different trigger sources by setting different values in the TSEL1[2:0] and
TSEL2[2:0] bits
• Load the dual DAC channel data into the desired DHR register (DAC_DHR12RD,
DAC_DHR12LD or DAC_DHR8RD)
When a DAC channel1 trigger arrives, the DHR1 register is transferred into DAC_DOR1
(three APB1 clock cycles later).
When a DAC channel2 trigger arrives, the DHR2 register is transferred into DAC_DOR2
(three APB1 clock cycles later).
14.4.2 Independent trigger with single LFSR generation

TSEL2[2:0] bits
• Configure the two DAC channel WAVEx[1:0] bits as “01” and the same LFSR mask
value in the MAMPx[3:0] bits
• Load the dual DAC channel data into the desired DHR register (DHR12RD, DHR12LD
or DHR8RD)
When a DAC channel1 trigger arrives, the LFSR1 counter, with the same mask, is added to
the DHR1 register and the sum is transferred into DAC_DOR1 (three APB1 clock cycles
later). Then the LFSR1 counter is updated.
When a DAC channel2 trigger arrives, the LFSR2 counter, with the same mask, is added to
later). Then the LFSR2 counter is updated.
14.4.3 Independent trigger with different LFSR generation

TSEL2[2:0] bits
• Configure the two DAC channel WAVEx[1:0] bits as “01” and set different LFSR masks
values in the MAMP1[3:0] and MAMP2[3:0] bits
When a DAC channel1 trigger arrives, the LFSR1 counter, with the mask configured by
MAMP1[3:0], is added to the DHR1 register and the sum is transferred into DAC_DOR1
(three APB1 clock cycles later). Then the LFSR1 counter is updated.
When a DAC channel2 trigger arrives, the LFSR2 counter, with the mask configured by
MAMP2[3:0], is added to the DHR2 register and the sum is transferred into DAC_DOR2
(three APB1 clock cycles later). Then the LFSR2 counter is updated.
RM0090 Rev 19 441/1751

454
14.4.4 Independent trigger with single triangle generation

TSEL2[2:0] bits
• Configure the two DAC channel WAVEx[1:0] bits as “1x” and the same maximum
amplitude value in the MAMPx[3:0] bits
When a DAC channel1 trigger arrives, the DAC channel1 triangle counter, with the same
triangle amplitude, is added to the DHR1 register and the sum is transferred into
DAC_DOR1 (three APB1 clock cycles later). The DAC channel1 triangle counter is then
updated.
When a DAC channel2 trigger arrives, the DAC channel2 triangle counter, with the same
triangle amplitude, is added to the DHR2 register and the sum is transferred into
DAC_DOR2 (three APB1 clock cycles later). The DAC channel2 triangle counter is then
updated.
14.4.5 Independent trigger with different triangle generation

TSEL2[2:0] bits
• Configure the two DAC channel WAVEx[1:0] bits as “1x” and set different maximum
amplitude values in the MAMP1[3:0] and MAMP2[3:0] bits
When a DAC channel1 trigger arrives, the DAC channel1 triangle counter, with a triangle
amplitude configured by MAMP1[3:0], is added to the DHR1 register and the sum is
transferred into DAC_DOR1 (three APB1 clock cycles later). The DAC channel1 triangle
counter is then updated.
When a DAC channel2 trigger arrives, the DAC channel2 triangle counter, with a triangle
amplitude configured by MAMP2[3:0], is added to the DHR2 register and the sum is
transferred into DAC_DOR2 (three APB1 clock cycles later). The DAC channel2 triangle
counter is then updated.
14.4.6 Simultaneous software start

• Load the dual DAC channel data to the desired DHR register (DAC_DHR12RD,
In this configuration, one APB1 clock cycle later, the DHR1 and DHR2 registers are
transferred into DAC_DOR1 and DAC_DOR2, respectively.
442/1751 RM0090 Rev 19

14.4.7 Simultaneous trigger without wave generation

• Configure the same trigger source for both DAC channels by setting the same value in
the TSEL1[2:0] and TSEL2[2:0] bits
• Load the dual DAC channel data to the desired DHR register (DAC_DHR12RD,
When a trigger arrives, the DHR1 and DHR2 registers are transferred into DAC_DOR1 and
DAC_DOR2, respectively (after three APB1 clock cycles).
14.4.8 Simultaneous trigger with single LFSR generation

• Configure the two DAC channel WAVEx[1:0] bits as “01” and the same LFSR mask
value in the MAMPx[3:0] bits
• Load the dual DAC channel data to the desired DHR register (DHR12RD, DHR12LD or
DHR8RD)
When a trigger arrives, the LFSR1 counter, with the same mask, is added to the DHR1
register and the sum is transferred into DAC_DOR1 (three APB1 clock cycles later). The
LFSR1 counter is then updated. At the same time, the LFSR2 counter, with the same mask,
is added to the DHR2 register and the sum is transferred into DAC_DOR2 (three APB1
clock cycles later). The LFSR2 counter is then updated.
14.4.9 Simultaneous trigger with different LFSR generation

• Configure the two DAC channel WAVEx[1:0] bits as “01” and set different LFSR mask
values using the MAMP1[3:0] and MAMP2[3:0] bits
When a trigger arrives, the LFSR1 counter, with the mask configured by MAMP1[3:0], is
added to the DHR1 register and the sum is transferred into DAC_DOR1 (three APB1 clock
cycles later). The LFSR1 counter is then updated.
At the same time, the LFSR2 counter, with the mask configured by MAMP2[3:0], is added to
later). The LFSR2 counter is then updated.
RM0090 Rev 19 443/1751

454
14.4.10 Simultaneous trigger with single triangle generation

• Configure the two DAC channel WAVEx[1:0] bits as “1x” and the same maximum
amplitude value using the MAMPx[3:0] bits
When a trigger arrives, the DAC channel1 triangle counter, with the same triangle
amplitude, is added to the DHR1 register and the sum is transferred into DAC_DOR1 (three
APB1 clock cycles later). The DAC channel1 triangle counter is then updated.
At the same time, the DAC channel2 triangle counter, with the same triangle amplitude, is
added to the DHR2 register and the sum is transferred into DAC_DOR2 (three APB1 clock
cycles later). The DAC channel2 triangle counter is then updated.
14.4.11 Simultaneous trigger with different triangle generation

• Configure the two DAC channel WAVEx[1:0] bits as “1x” and set different maximum
amplitude values in the MAMP1[3:0] and MAMP2[3:0] bits
When a trigger arrives, the DAC channel1 triangle counter, with a triangle amplitude
configured by MAMP1[3:0], is added to the DHR1 register and the sum is transferred into
DAC_DOR1 (three APB1 clock cycles later). Then the DAC channel1 triangle counter is
updated.
At the same time, the DAC channel2 triangle counter, with a triangle amplitude configured
by MAMP2[3:0], is added to the DHR2 register and the sum is transferred into DAC_DOR2
(three APB1 clock cycles later). Then the DAC channel2 triangle counter is updated.
444/1751 RM0090 Rev 19

14.5 DAC registers

The peripheral registers have to be accessed by words (32 bits).
14.5.1 DAC control register (DAC_CR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DMAU DMA
MAMP2[3:0] WAVE2[1:0] TSEL2[2:0] TEN2 BOFF2 EN2
Reserved DRIE2 EN2
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DMAU DMA
MAMP1[3:0] WAVE1[1:0] TSEL1[2:0] TEN1 BOFF1 EN1
Reserved DRIE1 EN1

Bits 29 DMAUDRIE2: DAC channel2 DMA underrun interrupt enable
0: DAC channel2 DMA underrun interrupt disabled
1: DAC channel2 DMA underrun interrupt enabled
Bit 28 DMAEN2: DAC channel2 DMA enable
0: DAC channel2 DMA mode disabled
1: DAC channel2 DMA mode enabled
Bits 27:24 MAMP2[3:0]: DAC channel2 mask/amplitude selector
These bits are written by software to select mask in wave generation mode or amplitude in
triangle generation mode.
0000: Unmask bit0 of LFSR/ triangle amplitude equal to 1
0001: Unmask bits[1:0] of LFSR/ triangle amplitude equal to 3
≥ 1011: Unmask bits[11:0] of LFSR/ triangle amplitude equal to 4095
Bits 23:22 WAVE2[1:0]: DAC channel2 noise/triangle wave generation enable
These bits are set/reset by software.
00: wave generation disabled
01: Noise wave generation enabled
1x: Triangle wave generation enabled
Note: Only used if bit TEN2 = 1 (DAC channel2 trigger enabled)
RM0090 Rev 19 445/1751

454
Bits 21:19 TSEL2[2:0]: DAC channel2 trigger selection

These bits select the external event used to trigger DAC channel2
110: External line9
111: Software trigger
Note: Only used if bit TEN2 = 1 (DAC channel2 trigger enabled).
Bit 18 TEN2: DAC channel2 trigger enable
This bit is set and cleared by software to enable/disable DAC channel2 trigger
0: DAC channel2 trigger disabled and data written into the DAC_DHRx register are
transferred one APB1 clock cycle later to the DAC_DOR2 register
1: DAC channel2 trigger enabled and data from the DAC_DHRx register are transferred
three APB1 clock cycles later to the DAC_DOR2 register
Note: When software trigger is selected, the transfer from the DAC_DHRx register to the
DAC_DOR2 register takes only one APB1 clock cycle.
Bit 17 BOFF2: DAC channel2 output buffer disable
This bit is set and cleared by software to enable/disable DAC channel2 output buffer.
0: DAC channel2 output buffer enabled
1: DAC channel2 output buffer disabled
Bit 16 EN2: DAC channel2 enable
This bit is set and cleared by software to enable/disable DAC channel2.
0: DAC channel2 disabled
1: DAC channel2 enabled
Bit 13 DMAUDRIE1: DAC channel1 DMA Underrun Interrupt enable
0: DAC channel1 DMA Underrun Interrupt disabled
1: DAC channel1 DMA Underrun Interrupt enabled
Bit 12 DMAEN1: DAC channel1 DMA enable
0: DAC channel1 DMA mode disabled
1: DAC channel1 DMA mode enabled
446/1751 RM0090 Rev 19

Bits 11:8 MAMP1[3:0]: DAC channel1 mask/amplitude selector

These bits are written by software to select mask in wave generation mode or amplitude in
triangle generation mode.
0000: Unmask bit0 of LFSR/ triangle amplitude equal to 1
≥ 1011: Unmask bits[11:0] of LFSR/ triangle amplitude equal to 4095
Bits 7:6 WAVE1[1:0]: DAC channel1 noise/triangle wave generation enable
00: wave generation disabled
01: Noise wave generation enabled
1x: Triangle wave generation enabled
Bits 5:3 TSEL1[2:0]: DAC channel1 trigger selection
These bits select the external event used to trigger DAC channel1.
110: External line9
111: Software trigger
Bit 2 TEN1: DAC channel1 trigger enable
This bit is set and cleared by software to enable/disable DAC channel1 trigger.
0: DAC channel1 trigger disabled and data written into the DAC_DHRx register are
transferred one APB1 clock cycle later to the DAC_DOR1 register
1: DAC channel1 trigger enabled and data from the DAC_DHRx register are transferred
three APB1 clock cycles later to the DAC_DOR1 register
Note: When software trigger is selected, the transfer from the DAC_DHRx register to the
DAC_DOR1 register takes only one APB1 clock cycle.
Bit 1 BOFF1: DAC channel1 output buffer disable
This bit is set and cleared by software to enable/disable DAC channel1 output buffer.
0: DAC channel1 output buffer enabled
1: DAC channel1 output buffer disabled
Bit 0 EN1: DAC channel1 enable
This bit is set and cleared by software to enable/disable DAC channel1.
0: DAC channel1 disabled
1: DAC channel1 enabled
RM0090 Rev 19 447/1751

454
14.5.2 DAC software trigger register (DAC_SWTRIGR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SWTRIG2 SWTRIG1
Reserved
w w

Bit 1 SWTRIG2: DAC channel2 software trigger
This bit is set and cleared by software to enable/disable the software trigger.
0: Software trigger disabled
1: Software trigger enabled
Note: This bit is cleared by hardware (one APB1 clock cycle later) once the DAC_DHR2
register value has been loaded into the DAC_DOR2 register.
Bit 0 SWTRIG1: DAC channel1 software trigger
This bit is set and cleared by software to enable/disable the software trigger.
0: Software trigger disabled
1: Software trigger enabled
Note: This bit is cleared by hardware (one APB1 clock cycle later) once the DAC_DHR1
register value has been loaded into the DAC_DOR1 register.

(DAC_DHR12R1)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DACC1DHR[11:0]
Reserved

Bits 11:0 DACC1DHR[11:0]: DAC channel1 12-bit right-aligned data
These bits are written by software which specifies 12-bit data for DAC channel1.
448/1751 RM0090 Rev 19


(DAC_DHR12L1)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DACC1DHR[11:0]
Reserved

Bits 15:4 DACC1DHR[11:0]: DAC channel1 12-bit left-aligned data

(DAC_DHR8R1)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DACC1DHR[7:0]
Reserved

RM0090 Rev 19 449/1751

454

(DAC_DHR12R2)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DACC2DHR[11:0]
Reserved


(DAC_DHR12L2)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DACC2DHR[11:0]
Reserved

These bits are written by software which specify 12-bit data for DAC channel2.

(DAC_DHR8R2)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DACC2DHR[7:0]
Reserved

450/1751 RM0090 Rev 19

14.5.9 Dual DAC 12-bit right-aligned data holding register

(DAC_DHR12RD)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DACC2DHR[11:0]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DACC1DHR[11:0]
Reserved

14.5.10 DUAL DAC 12-bit left aligned data holding register

(DAC_DHR12LD)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DACC2DHR[11:0]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DACC1DHR[11:0]
Reserved

RM0090 Rev 19 451/1751

454
14.5.11 DUAL DAC 8-bit right aligned data holding register

(DAC_DHR8RD)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DACC2DHR[7:0] DACC1DHR[7:0]

14.5.12 DAC channel1 data output register (DAC_DOR1)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DACC1DOR[11:0]
Reserved

Bit 11:0 DACC1DOR[11:0]: DAC channel1 data output
These bits are read-only, they contain data output for DAC channel1.
14.5.13 DAC channel2 data output register (DAC_DOR2)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DACC2DOR[11:0]
Reserved

Bits 11:0 DACC2DOR[11:0]: DAC channel2 data output
These bits are read-only, they contain data output for DAC channel2.
452/1751 RM0090 Rev 19

14.5.14 DAC status register (DAC_SR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DMAUDR2
Reserved Reserved
rc_w1
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DMAUDR1
Reserved Reserved
rc_w1

Bit 29 DMAUDR2: DAC channel2 DMA underrun flag
This bit is set by hardware and cleared by software (by writing it to 1).
0: No DMA underrun error condition occurred for DAC channel2
1: DMA underrun error condition occurred for DAC channel2 (the currently selected trigger is
driving DAC channel2 conversion at a frequency higher than the DMA service capability rate)
Bit 13 DMAUDR1: DAC channel1 DMA underrun flag
This bit is set by hardware and cleared by software (by writing it to 1).
0: No DMA underrun error condition occurred for DAC channel1
1: DMA underrun error condition occurred for DAC channel1 (the currently selected trigger is
driving DAC channel1 conversion at a frequency higher than the DMA service capability rate)
14.5.15 DAC register map

Table 76 summarizes the DAC registers.
Table 76. DAC register map

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
DMAUDRIE2
DMAUDRIE1
Reserved
Reserved
DMAEN2
DMAEN1
BOFF2
BOFF1
WAVE WAVE TSEL1[2

TEN2
TEN1
EN2
EN1
0x00 DAC_CR MAMP2[3:0] TSEL2[2:0] MAMP1[3:0]

2[2:0] 1[2:0] :0]
SWTRIG2
SWTRIG1
DAC_
0x04 Reserved
SWTRIGR
DAC_
0x08 Reserved DACC1DHR[11:0]
DHR12R1
DAC_
0x0C Reserved DACC1DHR[11:0] Reserved
DHR12L1
DAC_
DHR8R1
DAC_
DHR12R2
DAC_
0x18 Reserved DACC2DHR[11:0] Reserved
DHR12L2
RM0090 Rev 19 453/1751

454
Table 76. DAC register map (continued)

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
DAC_
0x1C Reserved DACC2DHR[7:0]
DHR8R2
DAC_
0x20 Reserved DACC2DHR[11:0] Reserved DACC1DHR[11:0]
DHR12RD
DAC_
0x24 DACC2DHR[11:0] Reserved DACC1DHR[11:0] Reserved
DHR12LD
DAC_
0x28 Reserved DACC2DHR[7:0] DACC1DHR[7:0]
DHR8RD
DAC_
0x2C Reserved DACC1DOR[11:0]
DOR1
DAC_
0x30 Reserved DACC2DOR[11:0]
DOR2
DMAUDR2
DMAUDR1
Reserved
0x34 DAC_SR Reserved Reserved
454/1751 RM0090 Rev 19

RM0090 Digital camera interface (DCMI)
15 Digital camera interface (DCMI)
This section applies to all STM32F4xx devices, unless otherwise specified.
15.1 DCMI introduction

The digital camera is a synchronous parallel interface able to receive a high-speed data flow
from an external 8-, 10-, 12- or 14-bit CMOS camera module. It supports different data
formats: YCbCr4:2:2/RGB565 progressive video and compressed data (JPEG).
This interface is for use with black & white cameras, X24 and X5 cameras, and it is
assumed that all pre-processing like resizing is performed in the camera module.
15.2 DCMI main features

• 8-, 10-, 12- or 14-bit parallel interface
• Embedded/external line and frame synchronization
• Continuous or snapshot mode
• Crop feature
• Supports the following data formats:
– 8/10/12/14- bit progressive video: either monochrome or raw bayer
– YCbCr 4:2:2 progressive video
– RGB 565 progressive video
– Compressed data: JPEG
15.3 DCMI pins

Table 77 shows the DCMI pins.
Table 77. DCMI pins

Name Signal type
D[0:13] Data inputs

HSYNC Horizontal synchronization input
VSYNC Vertical synchronization input
PIXCLK Pixel clock input
15.4 DCMI clocks

The digital camera interface uses two clock domains PIXCLK and HCLK. The signals
generated with PIXCLK are sampled on the rising edge of HCLK once they are stable. An
enable signal is generated in the HCLK domain, to indicate that data coming from the
camera are stable and can be sampled. The minimum PIXCLK period must be higher than
2.5 HCLK periods.
RM0090 Rev 19 455/1751

479
Digital camera interface (DCMI) RM0090
15.5 DCMI functional overview

The digital camera interface is a synchronous parallel interface that can receive high-speed
(up to 54 Mbytes/s) data flows. It consists of up to 14 data lines (D13-D0) and a pixel clock
line (PIXCLK). The pixel clock has a programmable polarity, so that data can be captured on
either the rising or the falling edge of the pixel clock.
The data are packed into a 32-bit data register (DCMI_DR) and then transferred through a
general-purpose DMA channel. The image buffer is managed by the DMA, not by the
camera interface.
The data received from the camera can be organized in lines/frames (raw YUB/RGB/Bayer
modes) or can be a sequence of JPEG images. To enable JPEG image reception, the JPEG
bit (bit 3 of DCMI_CR register) must be set.
The data flow is synchronized either by hardware using the optional HSYNC (horizontal
synchronization) and VSYNC (vertical synchronization) signals or by synchronization codes
embedded in the data flow.
Figure 72 shows the DCMI block diagram.
Figure 72. DCMI block diagram
DMA Control/Status
interface register
AHB
interface
FIFO/ Data DCMI_PIXCLK

Synchronizer
Data extraction
formatter
DCMI_D[0:13], DCMI_HSYNC, DCMI_VSYNC
ai15604b
456/1751 RM0090 Rev 19

Figure 73. Top-level block diagram
DCMI_D[0:13]
DCMI_PIXCLK External
HCLK
DCMI_HSYNC interface
DCMI_VSYNC
DCMI
Interrupt DCMI_IT
controller
DMA_REQ
ai15603b
15.5.1 DMA interface

The DMA interface is active when the CAPTURE bit in the DCMI_CR register is set. A DMA
request is generated each time the camera interface receives a complete 32-bit data block
in its register.
15.5.2 DCMI physical interface

The interface is composed of 11/13/15/17 inputs. Only the Slave mode is supported.
The camera interface can capture 8-bit, 10-bit, 12-bit or 14-bit data depending on the
EDM[1:0] bits in the DCMI_CR register. If less than 14 bits are used, the unused data pins
must not be assigned to DCMI interface through GPIO alternate functions.
Table 78. DCMI signals

Signal name Signal description
8 bits D[0..7]
10 bits D[0..9]
Data
12 bits D[0..11]
14 bits D[0..13]
PIXCLK Pixel clock
HSYNC Horizontal synchronization / Data valid
VSYNC Vertical synchronization
The data are synchronous with PIXCLK and change on the rising/falling edge of the pixel
clock depending on the polarity.
The HSYNC signal indicates the start/end of a line.
The VSYNC signal indicates the start/end of a frame
RM0090 Rev 19 457/1751

479
Figure 74. DCMI signal waveforms
DCMI_PIXCLK
DCMI_DR[0:13]
DCMI_HSYNC
DCMI_VSYNC
ai15606b
1. The capture edge of DCMI_PIXCLK is the falling edge, the active state of DCMI_HSYNC and
DCMI_VSYNC is 1.
1. DCMI_HSYNC and DCMI_VSYNC can change states at the same time.
8-bit data
When EDM[1:0] in DCMI_CR are programmed to “00” the interface captures 8 LSB’s at its
input (D[0:7]) and stores them as 8-bit data. The D[13:8] inputs are ignored. In this case, to
capture a 32-bit word, the camera interface takes four pixel clock cycles.
The first captured data byte is placed in the LSB position in the 32-bit word and the 4th
captured data byte is placed in the MSB position in the 32-bit word. Table 79 gives an
example of the positioning of captured data bytes in two 32-bit words.
Table 79. Positioning of captured data bytes in 32-bit words (8-bit width)
Byte address 31:24 23:16 15:8 7:0
0 Dn+3[7:0] Dn+2[7:0] Dn+1[7:0] Dn[7:0]

4 Dn+7[7:0] Dn+6[7:0] Dn+5[7:0] Dn+4[7:0]
10-bit data
When EDM[1:0] in DCMI_CR are programmed to “01”, the camera interface captures 10-bit
data at its input D[0..9] and stores them as the 10 least significant bits of a 16-bit word. The
remaining most significant bits in the DCMI_DR register (bits 11 to 15) are cleared to zero.
So, in this case, a 32-bit data word is made up every two pixel clock cycles.
The first captured data are placed in the LSB position in the 32-bit word and the 2nd
captured data are placed in the MSB position in the 32-bit word as shown in Table 80.
Byte address 31:26 25:16 15:10 9:0
0 0 Dn+1[9:0] 0 Dn[9:0]
4 0 Dn+3[9:0] 0 Dn+2[9:0]
458/1751 RM0090 Rev 19

12-bit data
When EDM[1:0] in DCMI_CR are programmed to “10”, the camera interface captures the
12-bit data at its input D[0..11] and stores them as the 12 least significant bits of a 16-bit
word. The remaining most significant bits are cleared to zero. So, in this case a 32-bit data
word is made up every two pixel clock cycles.
Byte address 31:28 27:16 15:12 11:0
0 0 Dn+1[11:0] 0 Dn[11:0]
4 0 Dn+3[11:0] 0 Dn+2[11:0]
14-bit data
When EDM[1:0] in DCMI_CR are programmed to “11”, the camera interface captures the
14-bit data at its input D[0..13] and stores them as the 14 least significant bits of a 16-bit
word. The remaining most significant bits are cleared to zero. So, in this case a 32-bit data
word is made up every two pixel clock cycles.
Byte address 31:30 29:16 15:14 13:0
0 0 Dn+1[13:0] 0 Dn[13:0]
4 0 Dn+3[13:0] 0 Dn+2[13:0]
15.5.3 Synchronization
The digital camera interface supports embedded or hardware (HSYNC & VSYNC)
synchronization. When embedded synchronization is used, it is up to the digital camera
module to make sure that the 0x00 and 0xFF values are used ONLY for synchronization
(not in data). Embedded synchronization codes are supported only for the 8-bit parallel data
interface width (that is, in the DCMI_CR register, the EDM[1:0] bits should be cleared to
“00”).
For compressed data, the DCMI supports only the hardware synchronization mode. In this
case, VSYNC is used as a start/end of the image, and HSYNC is used as a Data Valid
signal. Figure 75 shows the corresponding timing diagram.
RM0090 Rev 19 459/1751

479
Figure 75. Timing diagram

Padding data
at the end of the JPEG stream
Beginning of JPEG stream Programmable
JPEG packet size
JPEG data
End of JPEG stream

DCMI_HSYNC
DCMI_VSYNC
Packet dispatching depends on the image content.

This results in a variable blanking duration.
JPEG packet data
ai15944b
460/1751 RM0090 Rev 19

Hardware synchronization mode

In hardware synchronisation mode, the two synchronization signals (HSYNC/VSYNC) are
used.
Depending on the camera module/mode, data may be transmitted during horizontal/vertical
synchronisation periods. The HSYNC/VSYNC signals act like blanking signals since all the
data received during HSYNC/VSYNC active periods are ignored.
In order to correctly transfer images into the DMA/RAM buffer, data transfer is synchronized
with the VSYNC signal. When the hardware synchronisation mode is selected, and capture
is enabled (CAPTURE bit set in DCMI_CR), data transfer is synchronized with the
deactivation of the VSYNC signal (next start of frame).
Transfer can then be continuous, with successive frames transferred by DMA to successive
buffers or the same/circular buffer. To allow the DMA management of successive frames, a
VSIF (Vertical synchronization interrupt flag) is activated at the end of each frame.
Embedded data synchronization mode

In this synchronisation mode, the data flow is synchronised using 32-bit codes embedded in
the data flow. These codes use the 0x00/0xFF values that are not used in data anymore.
There are 4 types of codes, all with a 0xFF0000XY format. The embedded synchronization
codes are supported only in 8-bit parallel data width capture (in the DCMI_CR register, the
EDM[1:0] bits should be programmed to “00”). For other data widths, this mode generates
unpredictable results and must not be used.
Note: Camera modules can have 8 such codes (in interleaved mode). For this reason, the
interleaved mode is not supported by the camera interface (otherwise, every other half-
frame would be discarded).
• Mode 2
Four embedded codes signal the following events
– Frame start (FS)
– Frame end (FE)
– Line start (LS)
– Line end (LE)
The XY values in the 0xFF0000XY format of the four codes are programmable (see
Section 15.8.7: DCMI embedded synchronization code register (DCMI_ESCR)).
A 0xFF value programmed as a “frame end” means that all the unused codes are
interpreted as valid frame end codes.
In this mode, once the camera interface has been enabled, the frame capture starts
after the first occurrence of the frame end (FE) code followed by a frame start (FS)
code.
• Mode 1
An alternative coding is the camera mode 1. This mode is ITU656 compatible.
The codes signal another set of events:
– SAV (active line) - line start
– EAV (active line) - line end
– SAV (blanking) - end of line during interframe blanking period
– EAV (blanking) - end of line during interframe blanking period
RM0090 Rev 19 461/1751

479
This mode can be supported by programming the following codes:

• FS ≤ 0xFF
• FE ≤ 0xFF
• LS ≤ SAV (active)
• LE ≤ EAV (active)
An embedded unmask code is also implemented for frame/line start and frame/line end
codes. Using it, it is possible to compare only the selected unmasked bits with the
programmed code. You can therefore select a bit to compare in the embedded code and
detect a frame/line start or frame/line end. This means that there can be different codes for
the frame/line start and frame/line end with the unmasked bit position remaining the same.
Example
FS = 0xA5
Unmask code for FS = 0x10
In this case the frame start code is embedded in the bit 4 of the frame start code.
15.5.4 Capture modes

This interface supports two types of capture: snapshot (single frame) and continuous grab.
Snapshot mode (single frame)

In this mode, a single frame is captured (CM = ‘1’ in the DCMI_CR register). After the
CAPTURE bit is set in DCMI_CR, the interface waits for the detection of a start of frame
before sampling the data. The camera interface is automatically disabled (CAPTURE bit
cleared in DCMI_CR) after receiving the first complete frame. An interrupt is generated
(IT_FRAME) if it is enabled.
In case of an overrun, the frame is lost and the CAPTURE bit is cleared.
Figure 76. Frame capture waveforms in Snapshot mode
DCMI_HSYNC
DCMI_VSYNC
Frame 2
Frame 1 captured not captured
ai15832b
1. Here, the active state of DCMI_HSYNC and DCMI_VSYNC is 1.

462/1751 RM0090 Rev 19

Continuous grab mode

In this mode (CM bit = ‘0’ in DCMI_CR), once the CAPTURE bit has been set in DCMI_CR,
the grabbing process starts on the next VSYNC or embedded frame start depending on the
mode. The process continues until the CAPTURE bit is cleared in DCMI_CR. Once the
CAPTURE bit has been cleared, the grabbing process continues until the end of the current
frame.
Figure 77. Frame capture waveforms in continuous grab mode
DCMI_HSYNC
DCMI_VSYNC
Frame 1 captured Frame 2 captured
ai15833b

In continuous grab mode, you can configure the FCRC bits in DCMI_CR to grab all pictures,
every second picture or one out of four pictures to decrease the frame capture rate.
Note: In the hardware synchronization mode (ESS = ‘0’ in DCMI_CR), the IT_VSYNC interrupt is
generated (if enabled) even when CAPTURE = ‘0’ in DCMI_CR so, to reduce the frame
capture rate even further, the IT_VSYNC interrupt can be used to count the number of
frames between 2 captures in conjunction with the Snapshot mode. This is not allowed by
embedded data synchronization mode.
15.5.5 Crop feature

With the crop feature, the camera interface can select a rectangular window from the
received image. The start (upper left corner) coordinates and size (horizontal dimension in
number of pixel clocks and vertical dimension in number of lines) are specified using two 32-
bit registers (DCMI_CWSTRT and DCMI_CWSIZE). The size of the window is specified in
number of pixel clocks (horizontal dimension) and in number of lines (vertical dimension).
RM0090 Rev 19 463/1751

479
Figure 78. Coordinates and size of the window after cropping
VST bit in DCMI_CSTRT
VLINE bit in DCMI_CSIZE
HOFFCNT bit in DCMI_CSTRT

CAPCNT bit in DCMI_CSIZE
ai15834
These registers specify the coordinates of the starting point of the capture window as a line
number (in the frame, starting from 0) and a number of pixel clocks (on the line, starting from
0), and the size of the window as a line number and a number of pixel clocks. The CAPCNT
value can only be a multiple of 4 (two least significant bits are forced to 0) to allow the
correct transfer of data through the DMA.
If the VSYNC signal goes active before the number of lines is specified in the
DCMI_CWSIZE register, then the capture stops and an IT_FRAME interrupt is generated
when enabled.
Figure 79. Data capture waveforms
DCMI_HSYNC
DCMI_VSYNC
HOFFCNT
CAPCNT
Data not captured in this phase
Data captured in this phase

ai15835

464/1751 RM0090 Rev 19

15.5.6 JPEG format

To allow JPEG image reception, it is necessary to set the JPEG bit in the DCMI_CR register.
JPEG images are not stored as lines and frames, so the VSYNC signal is used to start the
capture while HSYNC serves as a data enable signal. The number of bytes in a line may not
be a multiple of 4, you should therefore be careful when handling this case since a DMA
request is generated each time a complete 32-bit word has been constructed from the
captured data. When an end of frame is detected and the 32-bit word to be transferred has
not been completely received, the remaining data are padded with ‘0s’ and a DMA request
is generated.
The crop feature and embedded synchronization codes cannot be used in the JPEG format.
15.5.7 FIFO
A four-word FIFO is implemented to manage data rate transfers on the AHB. The DCMI
features a simple FIFO controller with a read pointer incremented each time the camera
interface reads from the AHB, and a write pointer incremented each time the camera
interface writes to the FIFO. There is no overrun protection to prevent the data from being
overwritten if the AHB interface does not sustain the data transfer rate.
In case of overrun or errors in the synchronization signals, the FIFO is reset and the DCMI
interface waits for a new start of frame.
15.6 Data format description
15.6.1 Data formats

Three types of data are supported:
• 8-bit progressive video: either monochrome or raw Bayer format
• YCbCr 4:2:2 progressive video
• RGB565 progressive video. A pixel coded in 16 bits (5 bits for blue, 5 bits for red, 6 bits
for green) takes two clock cycles to be transferred.
Compressed data: JPEG
For B&W, YCbCr or RGB data, the maximum input size is 2048 × 2048 pixels. No limit in
JPEG compressed mode.
For monochrome, RGB & YCbCr, the frame buffer is stored in raster mode. 32-bit words are
used. Only the little endian format is supported.
Figure 80. Pixel raster scan order
Word0 Word1 Word2 Pixel row 0
Pixel raster
scan order
(increasing
addresses)
Pixel row n – 1
ai15848
RM0090 Rev 19 465/1751

479
15.6.2 Monochrome format

Characteristics:
• Raster format
• 8 bits per pixel
Table 83 shows how the data are stored.
Table 83. Data storage in monochrome progressive video format

Byte address 31:24 23:16 15:8 7:0
0 n+3 n+2 n+1 n

4 n+7 n+6 n+5 n+4
15.6.3 RGB format

Characteristics:
• Raster format
• RGB
• Interleaved: one buffer: R, G & B interleaved: BRGBRGBRG, etc.
• Optimized for display output
The RGB planar format is compatible with standard OS frame buffer display formats.
Only 16 BPP (bits per pixel): RGB565 (2 pixels per 32-bit word) is supported.
The 24 BPP (palletized format) and grayscale formats are not supported. Pixels are stored
in a raster scan order, that is from top to bottom for pixel rows, and from left to right within a
pixel row. Pixel components are R (red), G (green) and B (blue). All components have the
same spatial resolution (4:4:4 format). A frame is stored in a single part, with the
components interleaved on a pixel basis.
Table 84 shows how the data are stored.
Table 84. Data storage in RGB progressive video format

Byte address 31:27 26:21 20:16 15:11 10:5 4:0
0 Red n + 1 Green n + 1 Blue n + 1 Red n Green n Blue n

4 Red n + 4 Green n + 3 Blue n + 3 Red n + 2 Green n + 2 Blue n + 2
15.6.4 YCbCr format

Characteristics:
• Raster format
• YCbCr 4:2:2
• Interleaved: one Buffer: Y, Cb & Cr interleaved: CbYCrYCbYCr, etc.
Pixel components are Y (luminance or “luma”), Cb and Cr (chrominance or “chroma” blue
and red). Each component is encoded in 8 bits. Luma and chroma are stored together
(interleaved) as shown in Table 85.
466/1751 RM0090 Rev 19

Table 85. Data storage in YCbCr progressive video format

Byte address 31:24 23:16 15:8 7:0
0 Yn+1 Cr n Yn Cb n
4 Yn+3 Cr n + 2 Yn+2 Cb n + 2
15.7 DCMI interrupts

Five interrupts are generated. All interrupts are maskable by software. The global interrupt
(IT_DCMI) is the OR of all the individual interrupts. Table 86 gives the list of all interrupts.
Table 86. DCMI interrupts

Interrupt name Interrupt event
IT_LINE Indicates the end of line

IT_FRAME Indicates the end of frame capture
IT_OVR indicates the overrun of data reception
IT_VSYNC Indicates the synchronization frame
Indicates the detection of an error in the embedded synchronization frame
IT_ERR
detection
IT_DCMI Logic OR of the previous interrupts
15.8 DCMI register description

All DCMI registers have to be accessed as 32-bit words, otherwise a bus error occurs.
15.8.1 DCMI control register 1 (DCMI_CR)

Reset value: 0x0000 0x0000
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CAPTURE
PCKPOL
ENABLE
HSPOL
VSPOL
CROP
JPEG
Reserved
ESS
CM
EDM FCRC
Reserved

Bit 14 ENABLE: DCMI enable
0: DCMI disabled
1: DCMI enabled
Note: The DCMI configuration registers should be programmed correctly before
enabling this Bit
Bit 13: 12 Reserved, must be kept at reset value.
RM0090 Rev 19 467/1751

479
Bits 11:10 EDM[1:0]: Extended data mode

00: Interface captures 8-bit data on every pixel clock
Bits 9:8 FCRC[1:0]: Frame capture rate control
These bits define the frequency of frame capture. They are meaningful only in
Continuous grab mode. They are ignored in snapshot mode.
00: All frames are captured
01: Every alternate frame captured (50% bandwidth reduction)
10: One frame in 4 frames captured (75% bandwidth reduction)
11: reserved
Bit 7 VSPOL: Vertical synchronization polarity
This bit indicates the level on the VSYNC pin when the data are not valid on the
parallel interface.
0: VSYNC active low
1: VSYNC active high
Bit 6 HSPOL: Horizontal synchronization polarity
This bit indicates the level on the HSYNC pin when the data are not valid on the
parallel interface.
0: HSYNC active low
1: HSYNC active high
Bit 5 PCKPOL: Pixel clock polarity
This bit configures the capture edge of the pixel clock
0: Falling edge active.
1: Rising edge active.
Bit 4 ESS: Embedded synchronization select
0: Hardware synchronization data capture (frame/line start/stop) is synchronized
with the HSYNC/VSYNC signals.
1: Embedded synchronization data capture is synchronized with synchronization
codes embedded in the data flow.
Note: Valid only for 8-bit parallel data. HSPOL/VSPOL are ignored when the ESS
bit is set.
This bit is disabled in JPEG mode.
Bit 3 JPEG: JPEG format
0: Uncompressed video format
1: This bit is used for JPEG data transfers. The HSYNC signal is used as data
enable. The crop and embedded synchronization features (ESS bit) cannot be
used in this mode.
Bits 2 CROP: Crop feature
0: The full image is captured. In this case the total number of bytes in an image
frame should be a multiple of 4
1: Only the data inside the window specified by the crop register will be captured.
If the size of the crop window exceeds the picture size, then only the picture size
is captured.
468/1751 RM0090 Rev 19

Bit 1 CM: Capture mode

0: Continuous grab mode - The received data are transferred into the destination
memory through the DMA. The buffer location and mode (linear or circular
buffer) is controlled through the system DMA.
1: Snapshot mode (single frame) - Once activated, the interface waits for the
start of frame and then transfers a single frame through the DMA. At the end of
the frame, the CAPTURE bit is automatically reset.
Bit 0 CAPTURE: Capture enable
0: Capture disabled.
1: Capture enabled.
The camera interface waits for the first start of frame, then a DMA request is
generated to transfer the received data into the destination memory.
In snapshot mode, the CAPTURE bit is automatically cleared at the end of the
1st frame received.
In continuous grab mode, if the software clears this bit while a capture is
ongoing, the bit will be effectively cleared after the frame end.
Note: The DMA controller and all DCMI configuration registers should be
programmed correctly before enabling this bit.
RM0090 Rev 19 469/1751

479
15.8.2 DCMI status register (DCMI_SR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
HSYNC
VSYNC
FNE
Reserved
r r r

Bit 2 FNE: FIFO not empty
This bit gives the status of the FIFO
1: FIFO contains valid data
0: FIFO empty
Bit 1 VSYNC
This bit gives the state of the VSYNC pin with the correct programmed polarity.
When embedded synchronization codes are used, the meaning of this bit is the
following:
0: active frame
1: synchronization between frames
In case of embedded synchronization, this bit is meaningful only if the
CAPTURE bit in DCMI_CR is set.
Bit 0 HSYNC
This bit gives the state of the HSYNC pin with the correct programmed polarity.
When embedded synchronization codes are used, the meaning of this bit is the
following:
0: active line
1: synchronization between lines
In case of embedded synchronization, this bit is meaningful only if the
CAPTURE bit in DCMI_CR is set.
470/1751 RM0090 Rev 19

15.8.3 DCMI raw interrupt status register (DCMI_RIS)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FRAME_RIS
VSYNC_RIS
LINE_RIS
OVR_RIS
ERR_RIS
Reserved
r r r r r
DCMI_RIS gives the raw interrupt status and is accessible in read only. When read, this
register returns the status of the corresponding interrupt before masking with the DCMI_IER
register value.

Bit 4 LINE_RIS: Line raw interrupt status
This bit gets set when the HSYNC signal changes from the inactive state to the
active state. It goes high even if the line is not valid.
In the case of embedded synchronization, this bit is set only if the CAPTURE bit
in DCMI_CR is set.
It is cleared by writing a ‘1’ to the LINE_ISC bit in DCMI_ICR.
Bit 3 VSYNC_RIS: VSYNC raw interrupt status
This bit is set when the VSYNC signal changes from the inactive state to the
active state.
In the case of embedded synchronization, this bit is set only if the CAPTURE bit
is set in DCMI_CR.
It is cleared by writing a ‘1’ to the VSYNC_ISC bit in DCMI_ICR.
Bit 2 ERR_RIS: Synchronization error raw interrupt status
0: No synchronization error detected
1: Embedded synchronization characters are not received in the correct order.
This bit is valid only in the embedded synchronization mode. It is cleared by
writing a ‘1’ to the ERR_ISC bit in DCMI_ICR.
Note: This bit is available only in embedded synchronization mode.
Bit 1 OVR_RIS: Overrun raw interrupt status
0: No data buffer overrun occurred
1: A data buffer overrun occurred and the data FIFO is corrupted.
This bit is cleared by writing a ‘1’ to the OVR_ISC bit in DCMI_ICR.
Bit 0 FRAME_RIS: Capture complete raw interrupt status
0: No new capture
1: A frame has been captured.
This bit is set when a frame or window has been captured.
In case of a cropped window, this bit is set at the end of line of the last line in the
crop. It is set even if the captured frame is empty (e.g. window cropped outside
the frame).
This bit is cleared by writing a ‘1’ to the FRAME_ISC bit in DCMI_ICR.
RM0090 Rev 19 471/1751

479
15.8.4 DCMI interrupt enable register (DCMI_IER)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FRAME_IE
VSYNC_IE
LINE_IE
OVR_IE
ERR_IE
Reserved
rw rw rw rw rw
The DCMI_IER register is used to enable interrupts. When one of the DCMI_IER bits is set,
the corresponding interrupt is enabled. This register is accessible in both read and write.

Bit 4 LINE_IE: Line interrupt enable
0: No interrupt generation when the line is received
1: An Interrupt is generated when a line has been completely received
Bit 3 VSYNC_IE: VSYNC interrupt enable
0: No interrupt generation
1: An interrupt is generated on each VSYNC transition from the inactive to the
active state
The active state of the VSYNC signal is defined by the VSPOL bit.
Bit 2 ERR_IE: Synchronization error interrupt enable
1: An interrupt is generated if the embedded synchronization codes are not
received in the correct order.
Bit 1 OVR_IE: Overrun interrupt enable
1: An interrupt is generated if the DMA was not able to transfer the last data
before new data (32-bit) are received.
Bit 0 FRAME_IE: Capture complete interrupt enable
1: An interrupt is generated at the end of each received frame/crop window (in
crop mode).
472/1751 RM0090 Rev 19

15.8.5 DCMI masked interrupt status register (DCMI_MIS)

This DCMI_MIS register is a read-only register. When read, it returns the current masked
status value (depending on the value in DCMI_IER) of the corresponding interrupt. A bit in
this register is set if the corresponding enable bit in DCMI_IER is set and the corresponding
bit in DCMI_RIS is set.
Reset value: 0x0000
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FRAME_MIS
VSYNC_MIS
LINE_MIS
OVR_MIS
ERR_MIS
Reserved
r r r r r

Bit 4 LINE_MIS: Line masked interrupt status
This bit gives the status of the masked line interrupt
0: No interrupt generation when the line is received
1: An Interrupt is generated when a line has been completely received and the
LINE_IE bit is set in DCMI_IER.
Bit 3 VSYNC_MIS: VSYNC masked interrupt status
This bit gives the status of the masked VSYNC interrupt
0: No interrupt is generated on VSYNC transitions
1: An interrupt is generated on each VSYNC transition from the inactive to the
active state and the VSYNC_IE bit is set in DCMI_IER.
The active state of the VSYNC signal is defined by the VSPOL bit.
Bit 2 ERR_MIS: Synchronization error masked interrupt status
This bit gives the status of the masked synchronization error interrupt
0: No interrupt is generated on a synchronization error
1: An interrupt is generated if the embedded synchronization codes are not
received in the correct order and the ERR_IE bit in DCMI_IER is set.
Bit 1 OVR_MIS: Overrun masked interrupt status
This bit gives the status of the masked overflow interrupt
0: No interrupt is generated on overrun
1: An interrupt is generated if the DMA was not able to transfer the last data
before new data (32-bit) are received and the OVR_IE bit is set in DCMI_IER.
Bit 0 FRAME_MIS: Capture complete masked interrupt status
This bit gives the status of the masked capture complete interrupt
0: No interrupt is generated after a complete capture
1: An interrupt is generated at the end of each received frame/crop window (in
crop mode) and the FRAME_IE bit is set in DCMI_IER.
RM0090 Rev 19 473/1751

479
15.8.6 DCMI interrupt clear register (DCMI_ICR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FRAME_ISC
VSYNC_ISC
LINE_ISC
OVR_ISC
ERR_ISC
Reserved
w w w w w
The DCMI_ICR register is write-only. Writing a ‘1’ into a bit of this register clears the
corresponding bit in the DCMI_RIS and DCMI_MIS registers. Writing a ‘0’ has no effect.

Bit 4 LINE_ISC: line interrupt status clear
Writing a ‘1’ into this bit clears LINE_RIS in the DCMI_RIS register
Bit 3 VSYNC_ISC: Vertical synch interrupt status clear
Writing a ‘1’ into this bit clears the VSYNC_RIS bit in DCMI_RIS
Bit 2 ERR_ISC: Synchronization error interrupt status clear
Writing a ‘1’ into this bit clears the ERR_RIS bit in DCMI_RIS
Bit 1 OVR_ISC: Overrun interrupt status clear
Writing a ‘1’ into this bit clears the OVR_RIS bit in DCMI_RIS
Bits 0 FRAME_ISC: Capture complete interrupt status clear
Writing a ‘1’ into this bit clears the FRAME_RIS bit in DCMI_RIS
474/1751 RM0090 Rev 19

15.8.7 DCMI embedded synchronization code register (DCMI_ESCR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FEC LEC LSC FSC
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
Bits 31:24 FEC: Frame end delimiter code

This byte specifies the code of the frame end delimiter. The code consists of 4
bytes in the form of 0xFF, 0x00, 0x00, FEC.
If FEC is programmed to 0xFF, all the unused codes (0xFF0000XY) are
interpreted as frame end delimiters.
Bits 23:16 LEC: Line end delimiter code
This byte specifies the code of the line end delimiter. The code consists of 4
bytes in the form of 0xFF, 0x00, 0x00, LEC.
Bits 15:8 LSC: Line start delimiter code
This byte specifies the code of the line start delimiter. The code consists of 4
bytes in the form of 0xFF, 0x00, 0x00, LSC.
Bits 7:0 FSC: Frame start delimiter code
This byte specifies the code of the frame start delimiter. The code consists of 4
bytes in the form of 0xFF, 0x00, 0x00, FSC.
If FSC is programmed to 0xFF, no frame start delimiter is detected. But, the 1st
occurrence of LSC after an FEC code will be interpreted as a start of frame
delimiter.
RM0090 Rev 19 475/1751

479
15.8.8 DCMI embedded synchronization unmask register (DCMI_ESUR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FEU LEU LSU FSU
Bits 31:24 FEU: Frame end delimiter unmask

This byte specifies the mask to be applied to the code of the frame end delimiter.
0: The corresponding bit in the FEC byte in DCMI_ESCR is masked while
comparing the frame end delimiter with the received data.
1: The corresponding bit in the FEC byte in DCMI_ESCR is compared while
comparing the frame end delimiter with the received data
Bits 23:16 LEU: Line end delimiter unmask
This byte specifies the mask to be applied to the code of the line end delimiter.
0: The corresponding bit in the LEC byte in DCMI_ESCR is masked while
comparing the line end delimiter with the received data
1: The corresponding bit in the LEC byte in DCMI_ESCR is compared while
comparing the line end delimiter with the received data
Bits 15:8 LSU: Line start delimiter unmask
This byte specifies the mask to be applied to the code of the line start delimiter.
0: The corresponding bit in the LSC byte in DCMI_ESCR is masked while
comparing the line start delimiter with the received data
1: The corresponding bit in the LSC byte in DCMI_ESCR is compared while
comparing the line start delimiter with the received data
Bits 7:0 FSU: Frame start delimiter unmask
This byte specifies the mask to be applied to the code of the frame start
delimiter.
0: The corresponding bit in the FSC byte in DCMI_ESCR is masked while
comparing the frame start delimiter with the received data
1: The corresponding bit in the FSC byte in DCMI_ESCR is compared while
comparing the frame start delimiter with the received data
476/1751 RM0090 Rev 19

15.8.9 DCMI crop window start (DCMI_CWSTRT)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
VST[12:0 Reserv HOFFCNT[13:0]
Reserved
rw rw rw rw rw rw rw rw rw rw rw rw rw ed rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 28:16 VST[12:0]: Vertical start line count
The image capture starts with this line number. Previous line data are ignored.
0x0000 => line 1
0x0001 => line 2
0x0002 => line 3
....
Bit 13:0 HOFFCNT[13:0]: Horizontal offset count
This value gives the number of pixel clocks to count before starting a capture.
15.8.10 DCMI crop window size (DCMI_CWSIZE)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserved
Reserved
VLINE13:0] CAPCNT[13:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bitst 29:16 VLINE[13:0]: Vertical line count
This value gives the number of lines to be captured from the starting point.
0x0000 => 1 line
0x0001 => 2 lines
0x0002 => 3 lines
....
Bits 13:0 CAPCNT[13:0]: Capture count
This value gives the number of pixel clocks to be captured from the starting
point on the same line. It value should corresponds to word-aligned data for
different widths of parallel interfaces.
0x0000 => 1 pixel
0x0001 => 2 pixels
0x0002 => 3 pixels
....
RM0090 Rev 19 477/1751

479
15.8.11 DCMI data register (DCMI_DR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Byte3 Byte2 Byte1 Byte0
r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r r
Bits 31:24 Data byte 3

The digital camera Interface packages all the received data in 32-bit format before
requesting a DMA transfer. A 4-word deep FIFO is available to leave enough time for DMA
transfers and avoid DMA overrun conditions.
15.8.12 DCMI register map

Table 87 summarizes the DCMI registers.
Table 87. DCMI register map and reset values

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
CAPTURE
PCKPOL
ENABLE
HSPOL
VSPOL
CROP
FCR
JPEG
Reserved
ESS
EDM
CM
DCMI_CR
0x00 Reserved C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0
HSYNC
VSYNC
FNE
DCMI_SR
0x04 Reserved
Reset value 0 0 0
FRAME_RIS
VSYNC_RIS
LINE_RIS
OVR_RIS
ERR_RIS
DCMI_RIS
0x08 Reserved
Reset value 0 0 0 0 0
FRAME_IE
VSYNC_IE
LINE_IE
OVR_IE
ERR_IE
DCMI_IER
0x0C Reserved
FRAME_MIS
VSYNC_MIS
LINE_MIS
OVR_MIS
ERR_MIS
DCMI_MIS
0x10 Reserved
478/1751 RM0090 Rev 19

Table 87. DCMI register map and reset values (continued)

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
FRAME_ISC
VSYNC_ISC
LINE_ISC
OVR_ISC
ERR_ISC
DCMI_ICR
0x14 Reserved
DCMI_ESCR FEC LEC LSC FSC

0x18
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DCMI_ESUR FEU LEU LSU FSU

0x1C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DCMI_CWSTR
Reserved
Reserve VST[12:0 HOFFCNT[13:0]
0x20 T
d
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DCMI_CWSIZ
Reserved
Reserved
VLINE13:0] CAPCNT[13:0]
0x24 E
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DCMI_DR Byte3 Byte2 Byte1 Byte0

0x28
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RM0090 Rev 19 479/1751

479
LCD-TFT controller (LTDC) RM0090
16 LCD-TFT controller (LTDC)
This section applies only to STM32F429xx/439xx devices.
16.1 Introduction
The LCD-TFT (Liquid Crystal Display - Thin Film Transistor) display controller provides a
parallel digital RGB (Red, Green, Blue) and signals for horizontal, vertical synchronisation,
Pixel Clock and Data Enable as output to interface directly to a variety of LCD and TFT
panels.
16.2 LTDC main features

• 24-bit RGB Parallel Pixel Output; 8 bits-per-pixel (RGB888)
• 2 display layers with dedicated FIFO (64x32-bit)
• Color Look-Up Table (CLUT) up to 256 color (256x24-bit) per layer
• Supports up to XGA (1024x768) resolution
• Programmable timings for different display panels
• Programmable Background color
• Programmable polarity for HSync, VSync and Data Enable
• Up to 8 Input color formats selectable per layer
– ARGB8888
– RGB888
– RGB565
– ARGB1555
– ARGB4444
– L8 (8-bit Luminance or CLUT)
– AL44 (4-bit alpha + 4-bit luminance)
– AL88 (8-bit alpha + 8-bit luminance)
• Pseudo-random dithering output for low bits per channel
– Dither width 2-bits for Red, Green, Blue
• Flexible blending between two layers using alpha value (per pixel or constant)
• Color Keying (transparency color)
• Programmable Window position and size
• Supports thin film transistor (TFT) color displays
• AHB master interface with burst of 16 words
• Up to 4 programmable interrupt events
480/1751 RM0090 Rev 19

RM0090 LCD-TFT controller (LTDC)
16.3 LTDC functional description
16.3.1 LTDC block diagram

The block diagram of the LTDC is shown in Figure 81: LTDC block diagram.
Figure 81. LTDC block diagram

Pixel clock domain
AHB clock domain
Layer1
PFC LCD_HSYNC
FIFO
AHB Blending Dithering LCD_VSYNC
interface unit unit LCD_DE
Layer1 LCD-TFT
PFC LCD_CLK
FIFO panel
LCD_R[7:0]
LCD_G[7:0]
APB2 clock domain
Configuration LCD_B[7:0]
Timing
and status
generator
registers
Interrupts
MSv19675V1
Layer FIFO: One FIFO 64x32 bit per layer.

PFC: Pixel Format Convertor performing the pixel format conversion from the selected input
pixel format of a layer to words.
AHB interface: For data transfer from memories to the FIFO.
Blending, Dithering unit and Timings Generator: Refer to Section 16.4.1 and Section 16.4.2.
16.3.2 LTDC reset and clocks

The LCD-TFT controller peripheral uses 3 clock domains:
• The AHB clock domain (HCLK) is used for data transfer from the memories to the
Layer FIFO and frame buffer configuration register
• The APB2 clock domain (PCLK2) is used for global configuration register and interrupt
registers
• The Pixel Clock domain (LCD_CLK) is used to generate LCD-TFT interface signals,
pixel data generation and layer configuration. The LCD_CLK output should be
configured following the panel requirements. The LCD_CLK is configured through the
PLLSAI (refer to RCC section).
Table 88 summarizes the clock domain for each register.
RM0090 Rev 19 481/1751

514
Table 88. LTDC registers versus clock domain

LTDC registers Clock domain
LTDC_LxCR
LTDC_LxCFBAR
HCLK
LTDC_LxCFBLR
LTDC_LxCFBLNR
LTDC_SRCR
LTDC_IER
PCLK2
LTDC_ISR
LTDC_ICR
LTDC_SSCR
LTDC_BPCR
LTDC_AWCR
LTDC_TWCR
LTDC_GCR
LTDC_BCCR
LTDC_LIPCR
LTDC_CPSR
LTDC_CDSR Pixel Clock (LCD_CLK)
LTDC_LxWHPCR
LTDC_LxWVPCR
LTDC_LxCKCR
LTDC_LxPFCR
LTDC_LxCACR
LTDC_LxDCCR
LTDC_LxBFCR
LTDC_LxCLUTWR
Care must be taken when accessing the LTDC registers since the APB2 bus is stalling when
the following operations are ongoing:
• Register write access and update for 6 xPCKL2 period + 5x LCD_CLK period (5x
HCLK period for register on AHB clock domain)
• Register read access for 7xPCKL2 period + 5x LCD_CLK period (5x HCLK period for
register on AHB clock domain).
For registers on PCLK2 clock domain, APB2 bus is stalling during the register write access
for 6 xPCKL2 period and 7xPCKL2 period for read access.
The LCD controller can be reset by setting the corresponding bit in the RCC_APB2RSTR
register. It resets the three clock domains.
482/1751 RM0090 Rev 19

16.3.3 LCD-TFT pins and signal interface

The Table below summarizes the LTDC signal interface:
Table 89. LCD-TFT pins and signal interface

LCD-TFT
I/O Description
signals
LCD_CLK O Clock Output

LCD_HSYNC O Horizontal Synchronization
LCD_VSYNC O Vertical Synchronization
LCD_DE O Not Data Enable
LCD_R[7:0] O Data: 8-bit Red data
LCD_G[7:0] O Data: 8-bit Green data
LCD_B[7:0] O Data: 8-bit Blue data
The LCD-TFT controller pins must be configured by the user application. The unused pins
can be used for other purposes.
For LTDC outputs up to 24-bit (RGB888), if less than 8bpp are used to output for example
RGB565 or RGB666 to interface on 16b-bit or 18-bit displays, the RGB display data lines
must be connected to the MSB of the LCD-TFT controller RGB data lines. As an example, in
the case of an LCD-TFT controller interfacing with a RGB565 16-bit display, the LCD display
R[4:0], G[5:0] and B[4:0] data lines pins must be connected to LCD-TFT controller
LCD_R[7:3], LCD_G[7:2] and LCD_B[7:3].
16.4 LTDC programmable parameters

The LCD-TFT controller provides flexible configurable parameters. It can be enabled or
disabled through the LTDC_GCR register.
16.4.1 LTDC Global configuration parameters

Synchronous Timings:
Figure 82 presents the configurable timing parameters generated by the Synchronous
Timings Generator block presented in the block diagram Figure 81. It generates the
Horizontal and Vertical Synchronization timings panel signals, the Pixel Clock and the Data
Enable signals.
RM0090 Rev 19 483/1751

514
Figure 82. LCD-TFT Synchronous timings
Total width
HBP HFP
HSYNC
width
Active width
VSYNC width
VBP
Data1, Line1
Total height
Active display area

Active height
Data(n), Line(n)
VFP
MSv19674V1
Note: The HBP and HFP are respectively the Horizontal back porch and front porch period.
The VBP and the VFP are respectively the Vertical back porch and front porch period.
The LCD-TFT programmable synchronous timings are:
– HSYNC and VSYNC Width: Horizontal and Vertical Synchronization width
configured by programming a value of HSYNC Width - 1 and VSYNC Width - 1 in
the LTDC_SSCR register.
– HBP and VBP: Horizontal and Vertical Synchronization back porch width
configured by programming the accumulated value HSYNC Width + HBP - 1 and
the accumulated value VSYNC Width + VBP - 1 in the LTDC_BPCR register.
– Active Width and Active Height: The Active Width and Active Height are
configured by programming the accumulated value HSYNC Width + HBP +
Active Width - 1 and the accumulated value VSYNC Width + VBP + Active
Height - 1 in the LTDC_AWCR register (only up to 1024x768 is supported).
– Total Width: The Total width is configured by programming the accumulated value
HSYNC Width + HBP + Active Width + HFP - 1 in the LTDC_TWCR register. The
HFP is the Horizontal front porch period.
– Total Height: The Total Height is configured by programming the accumulated
value VSYNC Height + VBP + Active Height + VFP - 1 in the LTDC_TWCR
register. The VFP is the Vertical front porch period.
Note: When the LTDC is enabled, the timings generated start with X/Y=0/0 position as the first
horizontal synchronization pixel in the vertical synchronization area and following the back
porch, active data display area and the front porch.
484/1751 RM0090 Rev 19

When the LTDC is disabled, the timing generator block is reset to X=Total Width - 1,
Y=Total Height - 1 and held the last pixel before the vertical synchronization phase and the
FIFO are flushed. Therefore only blanking data is output continuously.
Example of Synchronous timings configuration

TFT-LCD timings (should be extracted from Panel datasheet):
• Horizontal and Vertical Synchronization width: 0x8 pixels and 0x4 lines
• Horizontal and Vertical back porch: 0x7 pixels and 0x2 lines
• Active Width and Active Height: 0x280 pixels, 0x1E0 lines (640x480)
• Horizontal front porch: 0x6 pixels
• Vertical front porch: 0x2 lines
The programmed values in the LTDC Timings registers will be:
• LTDC_SSCR register: to be programmed to 0x00070003. (HSW[11:0] is 0x7 and
VSH[10:0] is 0x3)
• LTDC_BPCR register: to be programmed to 0x000E0005. (AHBP[11:0] is 0xE(0x8 +
0x6) and AVBP[10:0] is 0x5(0x4 + 0x1))
• LTDC_AWCR register: to be programmed to 0x028E01E5. (AAW[11:0] is 0x28E(0x8
+0x7 +0x27F) and AAH[10:0] is 0x1E5(0x4 +0x2 + 0x1DF)
• LTDC_TWCR register: to be programmed to 0x00000294. (TOTALW[11:0] is
0x294(0x8 +0x7 +0x280 + 0x5)
• LTDC_THCR register: to be programmed to 0x000001E7. (TOTALH[10:0] is
0x1E7(0x4 +0x2 + 0x1E0 + 1)
Programmable polarity
The Horizontal and Vertical Synchronization, Data Enable and Pixel Clock output signals
polarity can be programmed to active high or active low through the LTDC_GCR register.
Background Color
A constant background color (RGB888) can programmed through the LTDC_BCCR
register. It is used for blending with the bottom layer.
Dithering
The Dithering pseudo-random technique using an LFSR is used to add a small random
value (threshold) to each pixel color channel (R, G or B) value, thus rounding up the MSB in
some cases when displaying a 24-bit data on 18-bit display. Thus the Dithering technique is
used to round data which is different from one frame to the other.
The Dither pseudo-random technique is the same as comparing LSBs against a threshold
value and adding a 1 to the MSB part only, if the LSB part is >= the threshold. The LSBs are
typically dropped once dithering was applied.
The width of the added pseudo-random value is 2 bits for each color channel; 2 bits for Red,
2 bits for Green and 2 bits for Blue.
Once the LCD-TFT controller is enabled, the LFSR starts running with the first active pixel
and it is kept running even during blanking periods and when dithering is switched off. If the
LTDC is disabled, the LFSR is reset.
The Dithering can be switched On and Off on the fly through the LTDC_GCR register.
RM0090 Rev 19 485/1751

514
Reload Shadow registers

Some configuration registers are shadowed. The shadow registers values can be reloaded
immediately to the active registers when writing to these registers or at the beginning of the
vertical blanking period following the configuration in the LTDC_SRCR register. If the
immediate reload configuration is selected, the reload should be only activated when all new
registers have been written.
The shadow registers should not be modified again before the reload has been done.
Reading from the shadow registers returns the actual active value. The new written value
can only be read after the reload has taken place.
A register reload interrupt can be generated if enabled in the LTDC_IER register.
The shadowed registers are all the Layer 1 and Layer 2 registers except the
LTDC_LxCLUTWR register.
Interrupt generation event

Refer to Section 16.5: LTDC interrupts for interrupt configuration.
16.4.2 Layer programmable parameters

Up to two layers can be enabled, disabled and configured separately. The layer display
order is fixed and it is bottom up. If two layers are enabled, the Layer2 is the top displayed
window.
Windowing
Every layer can be positioned and resized and it must be inside the Active Display area.
The window position and size are configured through the top-left and bottom-right X/Y
positions and the Internal timing generator which includes the synchronous, back porch size
and the active data area. Refer to LTDC_LxWHPCR and LTDC_WVPCR registers.
The programmable layer position and size defines the first/last visible pixel of a line and the
first/last visible line in the window. It allows to display either the full image frame or only a
part of the image frame. Refer to Figure 83
• The first and the last visible pixel in the layer are set by configuring the WHSTPOS[11:0]
and WHSPPOS[11:0] in the LTDC_LxWHPCR register.
• The first and the last visible lines in the layer are set by configuring the WVSTPOS[10:0]
and WVSPPOS[10:0] in the LTDC_LxWVPCR register.
486/1751 RM0090 Rev 19

Figure 83. Layer window programmable parameters:

Active data area
WVSTPOS bits in
LTDC_LxWVPCR
WVSPPOS bits in
WHSTPOS bits in
LTDC_LxWVPCR
LTDC_LxWHPCR Window
WHSPPOS bits in
LTDC_LxWHPCR
MSv19676V3
Pixel input Format

The programmable pixel format is used for the data stored in the frame buffer of a layer.
Up to 8 input pixel formats can be configured for every layer through the LTDC_LxPFCR
register
The pixel data is read from the frame buffer and then transformed to the internal 8888
(ARGB) format as follows:
• Components which have a width of less than 8 bits get expanded to 8 bits by bit
replication. The selected bit range is concatenated multiple times until it is longer than
8 bits. Of the resulting vector, the 8 MSB bits are chosen. Example: 5 bits of an
RGB565 red channel become (bit positions): 43210432 (the 3 LSBs are filled with the 3
MSBs of the 5 bits)
The figure below describes the pixel data mapping depending on the selected format.
Table 90. Pixel Data mapping versus Color Format

ARGB8888
@+3 @+2 @+1 @

Ax[7:0] Rx[7:0] Gx[7:0] Bx[7:0]
@+7 @+6 @+5 @+4
Ax+1[7:0] Rx+1[7:0] Gx+1[7:0] Bx+1[7:0]
RGB888
@+3 @+2 @+1 @

Bx+1[7:0] Rx[7:0] Gx[7:0] Bx[7:0]
@+7 @+6 @+5 @+4
Gx+2[7:0] Bx+2[7:0] Rx+1[7:0] Gx+1[7:0]
RGB565
@+3 @+2 @+1 @

Rx+1[4:0] Gx+1[5:3] Gx+1[2:0] Bx+1[4:0] Rx[4:0] Gx[5:3] Gx[2:0] Bx[4:0]
RM0090 Rev 19 487/1751

514
Table 90. Pixel Data mapping versus Color Format (continued)

ARGB8888
@+7 @+6 @+5 @+4

Rx+3[4:0] Gx+3[5:3] Gx+3[2:0] Bx+3[4:0] Rx+2[4:0] Gx+2[5:3] Gx+2[2:0] Bx+2[4:0]
ARGB1555
@+3
@+2 @+1 @
Ax+1[0]Rx+1[4:0]
Gx+1[2:0] Bx+1[4:0] Ax[0] Rx[4:0] Gx[4:3] Gx[2:0] Bx[4:0]
Gx+1[4:3]
@+7 @+5
@+6 @+4
Ax+3[0]Rx+3[4:0] Ax+2[0]Rx+2[4:0]Gx+2[4:
Gx+3[2:0] Bx+3[4:0] Gx+2[2:0] Bx+2[4:0]
Gx+3[4:3] 3]
ARGB4444
@+3 @+2 @+1 @

Ax+1[3:0]Rx+1[3:0] Gx+1[3:0] Bx+1[3:0] Ax[3:0] Rx[3:0] Gx[3:0] Bx[3:0]
@+7 @+6 @+5 @+4
Ax+3[3:0]Rx+3[3:0] Gx+3[3:0] Bx+3[3:0] Ax+2[3:0]Rx+2[3:0] Gx+2[3:0] Bx+2[3:0]
L8
@+3 @+2 @+1 @

Lx+3[7:0] Lx+2[7:0] Lx+1[7:0] Lx[7:0]
@+7 @+6 @+5 @+4
Lx+7[7:0] Lx+6[7:0] Lx+5[7:0] Lx+4[7:0]
AL44
@+3 @+2 @+1 @

Ax+3[3:0] Lx+3[3:0] Ax+2[3:0] Lx+2[3:0] Ax+1[3:0] Lx+1[3:0] Ax[3:0] Lx[3:0]
@+7 @+6 @+5 @+4
Ax+7[3:0] Lx+7[3:0] Ax+6[3:0] Lx+6[3:0] Ax+5[3:0] Lx+5[3:0] Ax+4[3:0] Lx+4[3:0]
AL88
@+3 @+2 @+1 @

Ax+1[7:0] Lx+1[7:0] Ax[7:0] Lx[7:0]
@+7 @+6 @+5 @+4
Ax+3[7:0] Lx+3[7:0] Ax+2[7:0] Lx+2[7:0]
Color Look-Up Table (CLUT)

The CLUT can be enabled at run-time for every layer through the LTDC_LxCR register and
it is only useful in case of indexed color when using the L8, AL44 and AL88 input pixel
format.
First, the CLUT has to be loaded with the R, G and B values that will replace the original R,
G, B values of that pixel (indexed color). Each color (RGB value) has its own address which
is the position within the CLUT.
488/1751 RM0090 Rev 19

The R, G and B values and their own respective address are programmed through the
LTDC_LxCLUTWR register.
• In case of L8 and AL88 input pixel format, the CLUT has to be loaded by 256 colors. The
address of each color is configured in the CLUTADD bits in the LTDC_LxCLUTWR
register.
• In case of AL44 input pixel format, the CLUT has to be only loaded by 16 colors. The
address of each color must be filled by replicating the 4-bit L channel to 8-bit as follows:
– L0 (indexed color 0), at address 0x00
– L1, at address 0x11
– L2, at address 0x22
– .....
– L15, at address 0xFF
Color Frame Buffer Address

Every Layer has a start address for the color frame buffer configured through the
LTDC_LxCFBAR register.
When a layer is enabled, the data is fetched from the Color Frame Buffer.
Color Frame Buffer Length

Every layer has a total line length setting for the color frame buffer in bytes and a number of
lines in the frame buffer configurable in the LTDC_LxCFBLR and LTDC_LxCFBLNR
register respectively.
The line length and the number of lines settings are used to stop the prefetching of data to
the layer FIFO at the end of the frame buffer.
• If it is set to less bytes than required, a FIFO underrun interrupt is generated if it has
been previously enabled.
• If it is set to more bytes than actually required, the useless data read from the FIFO is
discarded. The useless data is not displayed.
Color Frame Buffer Pitch

Every layer has a configurable pitch for the color frame buffer, which is the distance between
the start of one line and the beginning of the next line in bytes. It is configured through the
LTDC_LxCFBLR register.
Layer Blending
The blending is always active and the two layers can be blended following the blending
factors configured through the LTDC_LxBFCR register.
The blending order is fixed and it is bottom up. If two layers are enabled, first the Layer1 is
blended with the Background color, then the Layer2 is blended with the result of blended
color of Layer1 and the background. Refer to Figure 84.
RM0090 Rev 19 489/1751

514
Figure 84. Blending two layers with background
Layer 2
Layer 1 Layer 2 Layer 2 +
BG Layer 1 + BG Layer 1 + BG
MS19677V1
Default color
Every layer can have a default color in the format ARGB which is used outside the defined
layer window or when a layer is disabled.
The default color is configured through the LTDC_LxDCCR register.
The blending is always performed between the two layers even when a layer is disabled. To
avoid displaying the default color when a layer is disabled, keep the blending factors of this
layer in the LTDC_LxBFCR register to their reset value.
Color Keying
A color key (RGB) can be configured to be representative for a transparent pixel.
If the Color Keying is enabled, the current pixels (after format conversion and before
blending) are compared to the color key. If they match for the programmed RGB value, all
channels (ARGB) of that pixel are set to 0.
The Color Key value can be configured and used at run-time to replace the pixel RGB value.
The Color Keying is enabled through the LTDC_LxCKCR register.
16.5 LTDC interrupts

The LTDC provides four maskable interrupts logically ORed to two interrupt vectors.
The interrupt sources can be enabled or disabled separately through the LTDC_IER
register. Setting the appropriate mask bit to 1 enables the corresponding interrupt.
The two interrupts are generated on the following events:
• Line interrupt: generated when a programmed line is reached. The line interrupt
position is programmed in the LTDC_LIPCR register
• Register Reload interrupt: generated when the shadow registers reload was performed
during the vertical blanking period
• FIFO Underrun interrupt: generated when a pixel is requested from an empty layer
FIFO
• Transfer Error interrupt: generated when an AHB bus error occurs during data transfer
Those interrupts events are connected to the NVIC controller as described in the figure
below.
490/1751 RM0090 Rev 19

Figure 85. Interrupt events
Line
LTDC global interrupt
Register reload
FIFO underrun
LTDC global error interrupt
Transfer error
MS19678V1
Table 91. LTDC interrupt requests

Interrupt event Event flag Enable Control bit
Line LIF LIE

Register Reload RRIF RRIEN
FIFO Underrun FUDERRIF FUDERRIE
Transfer Error TERRIF TERRIE
RM0090 Rev 19 491/1751

514
16.6 LTDC programming procedure

• Enable the LTDC clock in the RCC register
• Configure the required Pixel clock following the panel datasheet
• Configure the Synchronous timings: VSYNC, HSYNC, Vertical and Horizontal back
porch, active data area and the front porch timings following the panel datasheet as
described in the Section 16.4.1: LTDC Global configuration parameters
• Configure the synchronous signals and clock polarity in the LTDC_GCR register
• If needed, configure the background color in the LTDC_BCCR register
• Configure the needed interrupts in the LTDC_IER and LTDC_LIPCR register
• Configure the Layer1/2 parameters by programming:
– The Layer window horizontal and vertical position in the LTDC_LxWHPCR and
LTDC_WVPCR registers. The layer window must be in the active data area.
– The pixel input format in the LTDC_LxPFCR register
– The color frame buffer start address in the LTDC_LxCFBAR register
– The line length and pitch of the color frame buffer in the LTDC_LxCFBLR register
– The number of lines of the color frame buffer in the LTDC_LxCFBLNR register
– if needed, load the CLUT with the RGB values and its address in the
LTDC_LxCLUTWR register
– If needed, configure the default color and the blending factors respectively in the
LTDC_LxDCCR and LTDC_LxBFCR registers
• Enable Layer1/2 and if needed the CLUT in the LTDC_LxCR register
• If needed, dithering and color keying can be enabled respectively in the LTDC_GCR
and LTDC_LxCKCR registers. It can be also enabled on the fly.
• Reload the shadow registers to active register through the LTDC_SRCR register.
• Enable the LCD-TFT controller in the LTDC_GCR register.
• All layer parameters can be modified on the fly except the CLUT. The new configuration
has to be either reloaded immediately or during vertical blanking period by configuring
the LTDC_SRCR register.
Note: All layer’s registers are shadowed. Once a register is written, it should not be modified again
before the reload has been done. Thus, a new write to the same register will override the
previous configuration if not yet reloaded.
492/1751 RM0090 Rev 19

16.7 LTDC registers
16.7.1 LTDC Synchronization Size Configuration Register (LTDC_SSCR)

This register defines the number of Horizontal Synchronization pixels minus 1 and the
number of Vertical Synchronization lines minus 1. Refer to Figure 82 and Section 16.4:
LTDC programmable parameters for an example of configuration.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
HSW[11:0]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
VSH[10:0]
Reserved

Bits 27:16 HSW[11:0]: Horizontal Synchronization Width (in units of pixel clock period)
These bits define the number of Horizontal Synchronization pixel minus 1.
Bits 10:0 VSH[10:0]: Vertical Synchronization Height (in units of horizontal scan line)
These bits define the vertical Synchronization height minus 1. It represents the
number
of horizontal synchronization lines.
16.7.2 LTDC Back Porch Configuration Register (LTDC_BPCR)

This register defines the accumulated number of Horizontal Synchronization and back porch
pixels minus 1 (HSYNC Width + HBP- 1) and the accumulated number of Vertical
Synchronization and back porch lines minus 1 (VSYNC Height + VBP - 1). Refer to
Figure 82 and Section 16.4: LTDC programmable parameters for an example of
configuration.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
AHBP[11:0]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
AVBP[10:0]
Reserved
RM0090 Rev 19 493/1751

514

Bits 27:16 AHBP[11:0]: Accumulated Horizontal back porch (in units of pixel clock period)
These bits define the Accumulated Horizontal back porch width which includes the
Horizontal Synchronization and Horizontal back porch pixels minus 1.
The Horizontal back porch is the period between Horizontal Synchronization going
inactive and the start of the active display part of the next scan line.
Bits 10:0 AVBP[10:0]: Accumulated Vertical back porch (in units of horizontal scan line)
These bits define the accumulated Vertical back porch width which includes the Vertical
Synchronization and Vertical back porch lines minus 1.
The Vertical back porch is the number of horizontal scan lines at a start of frame to the
start of the first active scan line of the next frame.
16.7.3 LTDC Active Width Configuration Register (LTDC_AWCR)

This register defines the accumulated number of Horizontal Synchronization, back porch
and Active pixels minus 1 (HSYNC width + HBP + Active Width - 1) and the accumulated
number of Vertical Synchronization, back porch lines and Active lines minus 1 (VSYNC
Height+ BVBP + Active Height - 1). Refer to Figure 82 and Section 16.4: LTDC
programmable parameters for an example of configuration.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
AAW[11:0]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
AAH[10:0]
Reserved

Bits 27:16 AAW[11:0]: Accumulated Active Width (in units of pixel clock period)
These bits define the Accumulated Active Width which includes the Horizontal
Synchronization, Horizontal back porch and Active pixels minus 1.
The Active Width is the number of pixels in active display area of the panel scan line. The
maximum Active Width supported is 0x400.
Bits 10:0 AAH[10:0]: Accumulated Active Height (in units of horizontal scan line)
These bits define the Accumulated Height which includes the Vertical Synchronization,
Vertical back porch and the Active Height lines minus 1. The Active Height is the number
of active lines in the panel. The maximum Active Height supported is 0x300.
494/1751 RM0090 Rev 19

16.7.4 LTDC Total Width Configuration Register (LTDC_TWCR)

This register defines the accumulated number of Horizontal Synchronization, back porch,
Active and front porch pixels minus 1 (HSYNC Width + HBP + Active Width + HFP - 1) and
the accumulated number of Vertical Synchronization, back porch lines, Active and Front
lines minus 1 (VSYNC Height+ BVBP + Active Height + VFP - 1). Refer to Figure 82 and
Section 16.4: LTDC programmable parameters for an example of configuration.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
TOTALW[11:0]
Reserved
16 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TOTALH[10:0]
Reserved

Bits 27:16 TOTALW[11:0]: Total Width (in units of pixel clock period)
These bits defines the accumulated Total Width which includes the Horizontal
Synchronization, Horizontal back porch, Active Width and Horizontal front porch pixels
minus 1.
Bits 10:0 TOTALH[10:0]: Total Height (in units of horizontal scan line)
These bits defines the accumulated Height which includes the Vertical Synchronization,
Vertical back porch, the Active Height and Vertical front porch Height lines minus 1.
16.7.5 LTDC Global Control Register (LTDC_GCR)

This register defines the global configuration of the LCD-TFT controller.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
HSPOL VSPOL DEPOL PCPOL DEN

Reserved
rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserve DRW[2:0] Reser DGW[2:0] Reser DBW[2:0] LTDCEN

Reserved
d ved ved
r r r r r r r r r rw
RM0090 Rev 19 495/1751

514
Bit 31 HSPOL: Horizontal Synchronization Polarity

0: Horizontal Synchronization polarity is active low
1: Horizontal Synchronization polarity is active high
Bit 30 VSPOL: Vertical Synchronization Polarity
0: Vertical Synchronization is active low
1: Vertical Synchronization is active high
Bit 29 DEPOL: Data Enable Polarity
0: Data Enable polarity is active low
1: Data Enable polarity is active high
Bit 28 PCPOL: Pixel Clock Polarity
0: input pixel clock
1: inverted input pixel clock
Bit 16 DEN: Dither Enable
0: Dither disable
1: Dither enable
Bit 15 Reserved, must be kept at reset value
Bits 14:12 DRW[2:0]: Dither Red Width
These bits return the Dither Red Bits
Bits 10:8 DGW[2:0]: Dither Green Width
These bits return the Dither Green Bits
Bits 6:4 DBW[2:0]: Dither Blue Width
These bits return the Dither Blue Bits
Bit 0 LTDCEN: LCD-TFT controller enable bit
0: LTDC disable
1: LTDC enable
496/1751 RM0090 Rev 19

16.7.6 LTDC Shadow Reload Configuration Register (LTDC_SRCR)

This register allows to reload either immediately or during the vertical blanking period, the
shadow registers values to the active registers. The shadow registers are all Layer1 and
Layer2 registers except the LTDC_L1CLUTWR and the LTDC_L2CLUTWR.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
VBR IMR
Reserved
rw rw

Bit 1 VBR: Vertical Blanking Reload
This bit is set by software and cleared only by hardware after reload. (it cannot be cleared
through register write once it is set)
0: No effect
1: The shadow registers are reloaded during the vertical blanking period (at the
beginning of the first line after the Active Display Area)
Bit 0 IMR: Immediate Reload
This bit is set by software and cleared only by hardware after reload.
0: No effect
1: The shadow registers are reloaded immediately
Note: The shadow registers read back the active values. Until the reload has been done, the 'old'
value will be read.
16.7.7 LTDC Background Color Configuration Register (LTDC_BCCR)

This register defines the background color (RGB888).
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
BCRED[7:0]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
BCGREEN[7:0] BCBLUE[7:0]
RM0090 Rev 19 497/1751

514

Bits 23:16 BCRED[7:0]: Background Color Red value
These bits configure the background red value
Bits 15:8 BCGREEN[7:0]: Background Color Green value
These bits configure the background green value
Bits 7:0 BCBLUE[7:0]: Background Color Blue value
These bits configure the background blue value
16.7.8 LTDC Interrupt Enable Register (LTDC_IER)

This register determines which status flags generate an interrupt request by setting the
corresponding bit to 1.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RRIE TERRIE FUIE LIE

Reserved
rw rw rw rw

Bit 3 RRIE: Register Reload interrupt enable
0: Register Reload interrupt disable
1: Register Reload interrupt enable
Bit 2 TERRIE: Transfer Error Interrupt Enable
0: Transfer Error interrupt disable
1: Transfer Error interrupt enable
Bit 1 FUIE: FIFO Underrun Interrupt Enable
0: FIFO Underrun interrupt disable
1: FIFO Underrun Interrupt enable
Bit 0 LIE: Line Interrupt Enable
0: Line interrupt disable
1: Line Interrupt enable
498/1751 RM0090 Rev 19

16.7.9 LTDC Interrupt Status Register (LTDC_ISR)

This register returns the interrupt status flag
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RRIF TERRIF FUIF LIF

Reserved
r r r r

Bit 3 RRIF: Register Reload Interrupt Flag
0: No Register Reload interrupt generated
1: Register Reload interrupt generated when a vertical blanking reload occurs (and the
first line after the active area is reached)
Bit 2 TERRIF: Transfer Error interrupt flag
0: No Transfer Error interrupt generated
1: Transfer Error interrupt generated when a Bus error occurs
Bit 1 FUIF: FIFO Underrun Interrupt flag
0: NO FIFO Underrun interrupt generated.
1: A FIFO underrun interrupt is generated, if one of the layer FIFOs is empty and pixel
data is read from the FIFO
Bit 0 LIF: Line Interrupt flag
0: No Line interrupt generated
1: A Line interrupt is generated, when a programmed line is reached
16.7.10 LTDC Interrupt Clear Register (LTDC_ICR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CRRIF CTERRIF CFUIF CLIF

Reserved
w w w w
RM0090 Rev 19 499/1751

514

Bit 3 CRRIF: Clears Register Reload Interrupt Flag
0: No effect
1: Clears the RRIF flag in the LTDC_ISR register
Bit 2 CTERRIF: Clears the Transfer Error Interrupt Flag
0: No effect
1: Clears the TERRIF flag in the LTDC_ISR register.
Bit 1 CFUIF: Clears the FIFO Underrun Interrupt flag
0: No effect
1: Clears the FUDERRIF flag in the LTDC_ISR register.
Bit 0 CLIF: Clears the Line Interrupt Flag
0: No effect
1: Clears the LIF flag in the LTDC_ISR register.
16.7.11 LTDC Line Interrupt Position Configuration Register (LTDC_LIPCR)

This register defines the position of the line interrupt. The line value to be programmed
depends on the timings parameters. Refer to Figure 82.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
LIPOS[10:0]
Reserved

Bits 10:0 LIPOS[10:0]: Line Interrupt Position
These bits configure the line interrupt position
16.7.12 LTDC Current Position Status Register (LTDC_CPSR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CXPOS[15:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CYPOS{15;0]
500/1751 RM0090 Rev 19

Bits 31:16: CXPOS[15:0]: Current X Position

These bits return the current X position
Bits 15:0 CYPOS[15:0]: Current Y Position
These bits return the current Y position
16.7.13 LTDC Current Display Status Register (LTDC_CDSR)

This register returns the status of the current display phase which is controlled by the
HSYNC, VSYNC, and Horizontal/Vertical DE signals.
Example: if the current display phase is the vertical synchronization, the VSYNCS bit is set
(active high). If the current display phase is the horizontal synchronization, the HSYNCS bit
is active high.
Reset value: 0x0000 000F
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
HSYNC VSYNC
HDES VDES
Reserved S S
r r r r

Bit 3 HSYNCS: Horizontal Synchronization display Status
0: Active low
1: Active high
Bit 2 VSYNCS: Vertical Synchronization display Status
0: Active low
1: Active high
Bit 1 HDES: Horizontal Data Enable display Status
0: Active low
1: Active high
Bit 0 VDES: Vertical Data Enable display Status
0: Active low
1: Active high
Note: The returned status does not depend on the configured polarity in the LTDC_GCR register,
instead it returns the current active display phase.
RM0090 Rev 19 501/1751

514
16.7.14 LTDC Layerx Control Register (LTDC_LxCR) (where x=1..2)

Address offset: 0x84 + 0x80 x (Layerx -1), Layerx = 1 or 2
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CLUTEN COLKEN LEN

Reserved Reserved
rw rw rw

Bit 4 CLUTEN: Color Look-Up Table Enable
0: Color Look-Up Table disable
1: Color Look-Up Table enable
The CLUT is only meaningful for L8, AL44 and AL88 pixel format. Refer to Color Look-Up
Table (CLUT) on page 488
Bit 1 COLKEN: Color Keying Enable
0: Color Keying disable
1: Color Keying enable
Bit 0 LEN: Layer Enable
0: Layer disable
1: Layer enable
502/1751 RM0090 Rev 19

16.7.15 LTDC Layerx Window Horizontal Position Configuration Register

(LTDC_LxWHPCR) (where x=1..2)
This register defines the Horizontal Position (first and last pixel) of the layer 1 or 2 window.
The first visible pixel of a line is the programmed value of AHBP[10:0] bits + 1 in the
LTDC_BPCR register.
The last visible pixel of a line is the programmed value of AAW[10:0] bits in the
LTDC_AWCR register.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
WHSPPOS[11:0]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
WHSTPOS[11:0]
Reserved

Bits 27:16 WHSPPOS[11:0]: Window Horizontal Stop Position
These bits configure the last visible pixel of a line of the layer window.
The following condition must be respected:
WHSPPOS[11:0] ≥ AHBP[10:0] bits + 1 (programmed in LTDC_BPCR register)
Bits 11:0 WHSTPOS[11:0]: Window Horizontal Start Position
These bits configure the first visible pixel of a line of the layer window.
WHSTPOS[11:0] must be ≤ AAW[10:0] bits (programmed in the LTDC_AWCR
register).
Example:
The LTDC_BPCR register is configured to 0x000E0005(AHBP[11:0] is 0xE) and the
LTDC_AWCR register is configured to 0x028E01E5(AAW[11:0] is 0x28E). To configure the
horizontal position of a window size of 630x460, with horizontal start offset of 5 pixels in the
Active data area.
1. Layer window first pixel: WHSTPOS[11:0] should be programmed to 0x14 (0xE+1+0x5)
2. Layer window last pixel: WHSPPOS[11:0] should be programmed to 0x28A
RM0090 Rev 19 503/1751

514
16.7.16 LTDC Layerx Window Vertical Position Configuration Register

(LTDC_LxWVPCR) (where x=1..2)
This register defines the vertical position (first and last line) of the layer1 or 2 window.
The first visible line of a frame is the programmed value of AVBP[10:0] bits + 1 in the
register LTDC_BPCR register.
The last visible line of a frame is the programmed value of AAH[10:0] bits in the
LTDC_AWCR register.
Address offset: 0x8C + 0x80 x (Layerx -1), Layerx = 1 or 2
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
WVSPPOS[10:0]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
WVSTPOS[10:0]
Reserved

Bits 26:16 WVSPPOS[10:0]: Window Vertical Stop Position
These bits configures the last visible line of the layer window.
WHSPPOS[11:0] must be ≥ AVBP[10:0] bits + 1 (programmed in LTDC_BPCR
register)
Bits 10:0 WVSTPOS[10:0]: Window Vertical Start Position
These bits configure the first visible line of the layer window.
WHSTPOS[11:0] must be ≤ AAH[10:0] bits (programmed in the LTDC_AWCR register)
Example:
The LTDC_BPCR register is configured to 0x000E0005 (AVBP[10:0] is 0x5) and the
LTDC_AWCR register is configured to 0x028E01E5 (AAH[10:0] is 0x1E5). To configure the
vertical position of a window size of 630x460, with vertical start offset of 8 lines in the Active
data area:
1. Layer window first line: WVSTPOS[10:0] should be programmed to 0xE (0x5 + 1 + 0x8)
2. Layer window last line: WVSPPOS[10:0] should be programmed to 0x1DA
504/1751 RM0090 Rev 19

16.7.17 LTDC Layerx Color Keying Configuration Register (LTDC_LxCKCR)

(where x=1..2)
This register defines the color key value (RGB), which is used by the Color Keying.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CKRED[7:0]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CKGREEN[7:0] CKBLUE[7:0]

Bits 23:16 CKRED[7:0]: Color Key Red value
Bits 15:8 CKGREEN[7:0]: Color Key Green value
Bits 7:0 CKBLUE[7:0]: Color Key Blue value
16.7.18 LTDC Layerx Pixel Format Configuration Register (LTDC_LxPFCR)

(where x=1..2)
This register defines the pixel format which is used for the stored data in the frame buffer of
a layer. The pixel data is read from the frame buffer and then transformed to the internal
format 8888 (ARGB).
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PF[2:0]
Reserved
rw rw rw
RM0090 Rev 19 505/1751

514

Bits 2:0 PF[2:0]: Pixel Format
These bits configures the Pixel format
000: ARGB8888
001: RGB888
010: RGB565
011: ARGB1555
100: ARGB4444
101: L8 (8-Bit Luminance)
110: AL44 (4-Bit Alpha, 4-Bit Luminance)
111: AL88 (8-Bit Alpha, 8-Bit Luminance)
16.7.19 LTDC Layerx Constant Alpha Configuration Register (LTDC_LxCACR)

(where x=1..2)
This register defines the constant alpha value (divided by 255 by Hardware), which is used
in the alpha blending. Refer to LTDC_LxBFCR register.
Reset value: (Layerx -1) 0x0000 00FF
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CONSTA[7:0]
Reserved

Bits 7:0 CONSTA[7:0]: Constant Alpha
These bits configure the Constant Alpha used for blending. The Constant Alpha is divided
by 255 by hardware.
Example: if the programmed Constant Alpha is 0xFF, the Constant Alpha value is
255/255=1
16.7.20 LTDC Layerx Default Color Configuration Register (LTDC_LxDCCR)

(where x=1..2)
This register defines the default color of a layer in the format ARGB. The default color is
used outside the defined layer window or when a layer is disabled. The reset value of
0x00000000 defines a transparent black color.
Address offset: 0x9C + 0x80 x (Layerx -1), Layerx = 1 or 2
506/1751 RM0090 Rev 19

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DCALPHA[7:0] DCRED[7:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DCGREEN[7:0] DCBLUE[7:0]
Bits 31:24 DCALPHA[7:0]: Default Color Alpha

These bits configure the default alpha value
Bits 23:16 DCRED[7:0]: Default Color Red
These bits configure the default red value
Bits 15:8 DCGREEN[7:0]: Default Color Green
These bits configure the default green value
Bits 7:0 DCBLUE[7:0]: Default Color Blue
These bits configure the default blue value
RM0090 Rev 19 507/1751

514
16.7.21 LTDC Layerx Blending Factors Configuration Register (LTDC_LxBFCR)

(where x=1..2)
This register defines the blending factors F1 and F2.
The general blending formula is: BC = BF1 x C + BF2 x Cs
• BC = Blended color
• BF1 = Blend Factor 1
• C = Current layer color
• BF2 = Blend Factor 2
• Cs = subjacent layers blended color
Address offset: 0xA0 + 0x80 x (Layerx -1), Layerx = 1 or 2
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
BF1[2:0] BF2[2:0]
Reserved Reserved
rw rw rw rw rw rw

Bits 10:8 BF1[2:0]: Blending Factor 1
These bits select the blending factor F1
000: Reserved
001: Reserved
010: Reserved
011: Reserved
100: Constant Alpha
101: Reserved
110: Pixel Alpha x Constant Alpha
111:Reserved
Bits 2:0 BF2[2:0]: Blending Factor 2
These bits select the blending factor F2
000: Reserved
001: Reserved
010: Reserved
011: Reserved
100: Reserved
101: 1 - Constant Alpha
110: Reserved
111: 1 - (Pixel Alpha x Constant Alpha)
508/1751 RM0090 Rev 19

Note: The Constant Alpha value, is the programmed value in the LxCACR register divided by 255
by hardware.
Example: Only layer1 is enabled, BF1 configured to Constant Alpha
BF2 configured to 1 - Constant Alpha
Constant Alpha: The Constant Alpha programmed in the LxCACR register is 240 (0xF0).
Thus, the Constant Alpha value is 240/255 = 0.94
C: Current Layer Color is 128
Cs: Background color is 48
Layer1 is blended with the background color.
BC = Constant Alpha x C + (1 - Constant Alpha) x Cs = 0.94 x 128 + (1- 0.94) x 48 = 123.
16.7.22 LTDC Layerx Color Frame Buffer Address Register (LTDC_LxCFBAR)

(where x=1..2)
This register defines the color frame buffer start address which has to point to the address
where the pixel data of the top left pixel of a layer is stored in the frame buffer.
Address offset: 0xAC + 0x80 x (Layerx -1), Layerx = 1 or 2
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CFBADD[31:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CFBADD[31:0]
Bits 31:0 CFBADD[31:0]: Color Frame Buffer Start Address

These bits defines the color frame buffer start address.
16.7.23 LTDC Layerx Color Frame Buffer Length Register (LTDC_LxCFBLR)

(where x=1..2)
This register defines the color frame buffer line length and pitch.
Address offset: 0xB0 + 0x80 x (Layerx -1), Layerx = 1 or 2
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CFBP[12:0]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CFBLL[12:0]
Reserved
RM0090 Rev 19 509/1751

514
Bits 31:29 Reserved, must be kept at reset valuer

Bits 28:16 CFBP[12:0]: Color Frame Buffer Pitch in bytes
These bits define the pitch which is the increment from the start of one line of pixels to the
start of the next line in bytes.
Bits 12:0 CFBLL[12:0]: Color Frame Buffer Line Length
These bits define the length of one line of pixels in bytes + 3.
The line length is computed as follows: Active high width x number of bytes per pixel + 3.
Example:
• A frame buffer having the format RGB565 (2 bytes per pixel) and a width of 256 pixels
(total number of bytes per line is 256x2=512 bytes), where pitch = line length requires a
value of 0x02000203 to be written into this register.
• A frame buffer having the format RGB888 (3 bytes per pixel) and a width of 320 pixels
(total number of bytes per line is 320x3=960), where pitch = line length requires a value
of 0x03C003C3 to be written into this register.
16.7.24 LTDC Layerx ColorFrame Buffer Line Number Register

(LTDC_LxCFBLNR) (where x=1..2)
This register defines the number of lines in the color frame buffer.
Address offset: 0xB4 + 0x80 x (Layerx -1), Layerx = 1 or 2
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CFBLNBR[10:0]
Reserved

Bits 10:0 CFBLNBR[10:0]: Frame Buffer Line Number
These bits define the number of lines in the frame buffer which corresponds to the Active
high width.
Note: The number of lines and line length settings define how much data is fetched per frame for
every layer. If it is configured to less bytes than required, a FIFO underrun interrupt will be
generated if enabled.
The start address and pitch settings on the other hand define the correct start of every line in
memory.
510/1751 RM0090 Rev 19

16.7.25 LTDC Layerx CLUT Write Register (LTDC_LxCLUTWR)

(where x=1..2)
This register defines the CLUT address and the RGB value.
Address offset: 0xC4 + 0x80 x (Layerx -1), Layerx = 1 or 2
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CLUTADD[7:0] RED[7:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Bits 31:24 CLUTADD[7:0]: CLUT Address

These bits configure the CLUT address (color position within the CLUT) of each RGB
value
Bits 23:16 RED[7:0]: Red value
These bits configure the red value
Bits 15:8 GREEN[7:0]: Green value
These bits configure the green value
Bits 7:0 BLUE[7:0]: Blue value
These bits configure the blue value
Note: The CLUT write register should only be configured during blanking period or if the layer is
disabled. The CLUT can be enabled or disabled in the LTDC_LxCR register.
The CLUT is only meaningful for L8, AL44 and AL88 pixel format.
RM0090 Rev 19 511/1751

514
16.7.26 LTDC register map

The following table summarizes the LTDC registers. Refer to the register boundary
addresses table for the LTDC register base address.
Table 92. LTDC register map and reset values
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
LTDC_SSCR HSW[9:0] VSH[10:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
LTDC_BPCR AHBP[11:0] AVBP[10:0]

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
LTDC_AWCR AAV[11:0] AAH[10:0]

Reset value 0 9 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
LTDC_TWCR TOTALW[11:0] TOTALH[10:0]

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DGW[2:0]
DRW[2:0]
DBW[2:0]
LTDCEN
HSPOL
DEPOL
PCPOL
VSPOL
Reserved
Reserved
Reserved
Reserved
DEN
LTDC_GCR
0x0018 Reserve
Reset value 0 0 0 0 0 0 1 0 0 1 0 0 1 0 0
VBR
IMR
LTDC_SRCR
0x0024 Reserved
Reset value 0 0
LTDC_BCCR BC[23:0]
0x002C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
TERRIE
RRIE
FUIE
LIE
LTDC_IER
0x0034 Reserved
Reset value 0 0 0 0
TERRIF
RRIF
FUIF
LTDC_ISR
LIF
0x0038 Reserved
Reset value 0 0 0 0
CTERRIF
CRRIF
CFUIF
CLIF
LTDC_ICR
0x003C Reserved
Reset value 0 0 0 0
LTDC_LIPCR LIPOS[10:0]
0x0040 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0
LTDC_CPSR CXPOS[15:0] CYPOS[15:0]

0x0044
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HSYNCS
VSYNCS
HDES
VDES
LTDC_CDSR
0x0048 Reserved
Reset value 1 1 1 1
COLKEN
CLUTEN
Reserved
LEN
LTDC_L1CR
0x0084 Reserved
Reset value 0 0 0
LTDC_L1WHPCR WHSPPOS[11:0] WHSTPOS[11:0]

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
512/1751 RM0090 Rev 19

Table 92. LTDC register map and reset values (continued)
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
LTDC_L1WVPCR WVSPPOS[10:0] WVSTPOS[10:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
LTDC_L1CKCR CKRED[7:0] CKGREEN[7:0] CKBLUE[7:0]

0x0090 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
LTDC_L1PFCR PF[2:0]
0x0094 Reserved
Reset value 0 0 0
LTDC_L1CACR CONSTA[7:0]
0x0098 Reserved
Reset value 1 1 1 1 1 1 1 1
LTDC_L1DCCR DCALPHA[7:0] DCRED[7:0] DCGREEN[7:0] DCBLUE[7:0]

0x009C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
LTDC_L1BFCR BF1[2:0] BF2[2:0]

0x00A0 Reserved Reserved
LTDC_L1CFBAR CFBADD[31:0]
0x00AC
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
LTDC_L1CFBLR CFBP[12:0] CFBLL[12:0]

0x00B0 Reserved Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
LTDC_L1CFBLNR CFBLNBR[10:0]
0x00B4 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0
LTDC_L1CLUTWR CLUTADD[7:0] RED[7:0] GREEN[7:0] BLUE[7:0]

0x00C4
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
COLKEN
CLUTEN
Reserved
LEN
LTDC_L2CR
0x0104 Reserved
Reset value 0 0 0
LTDC_L2WHPCR WHSPPOS[11:0] WHSTPOS[11:0]

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
LTDC_L2WVPCR WVSPPOS[10:0] WVSTPOS[10:0]

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
LTDC_L2CKCR CKRED[7:0] CKGREEN[7:0] CKBLUE[7:0]

0x0110 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
LTDC_L2PFCR PF[2:0]
0x0114 Reserved
Reset value 0 0 0
LTDC_L2CACR CONSTA[7:0]
0x0118 Reserved
Reset value 1 1 1 1 1 1 1 1
LTDC_L2DCCR DCALPHA[7:0] DCRED[7:0] DCGREEN[7:0] DCBLUE[7:0]

0x011C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
LTDC_L2BFCR BF1[2:0] BF2[2:0]

LTDC_L2CFBAR CFBADD[31:0]
0x012C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
LTDC_L2CFBLR CFBP[12:0] CFBLL[12:0]

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RM0090 Rev 19 513/1751

514
Table 92. LTDC register map and reset values (continued)
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
LTDC_L2CFBLNR CFBLNBR[10:0]
0x0134 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0
LTDC_L2CLUTWR CLUTADD[7:0] RED[7:0] GREEN[7:0] BLUE[7:0]

0x0144
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
514/1751 RM0090 Rev 19

RM0090 Advanced-control timers (TIM1 and TIM8)
17 Advanced-control timers (TIM1 and TIM8)
17.1 TIM1 and TIM8 introduction

The advanced-control timers (TIM1 and TIM8) consist of a 16-bit auto-reload counter driven
by a programmable prescaler.
It may be used for a variety of purposes, including measuring the pulse lengths of input
signals (input capture) or generating output waveforms (output compare, PWM,
complementary PWM with dead-time insertion).
Pulse lengths and waveform periods can be modulated from a few microseconds to several
milliseconds using the timer prescaler and the RCC clock controller prescalers.
The advanced-control (TIM1 and TIM8) and general-purpose (TIMx) timers are completely
independent, and do not share any resources. They can be synchronized together as
described in Section 17.3.20.
RM0090 Rev 19 515/1751

588
Advanced-control timers (TIM1 and TIM8) RM0090
17.2 TIM1 and TIM8 main features

TIM1 and TIM8 timer features include:
• 16-bit up, down, up/down auto-reload counter.
• 16-bit programmable prescaler allowing dividing (also “on the fly”) the counter clock
frequency either by any factor between 1 and 65536.
• Up to 4 independent channels for:
– Input capture
– Output compare
– PWM generation (Edge and Center-aligned Mode)
– One-pulse mode output
• Complementary outputs with programmable dead-time
• Synchronization circuit to control the timer with external signals and to interconnect
several timers together.
• Repetition counter to update the timer registers only after a given number of cycles of
the counter.
• Break input to put the timer’s output signals in reset state or in a known state.
• Interrupt/DMA generation on the following events:
– Update: counter overflow/underflow, counter initialization (by software or
internal/external trigger)
– Trigger event (counter start, stop, initialization or count by internal/external trigger)
– Input capture
– Output compare
– Break input
• Supports incremental (quadrature) encoder and hall-sensor circuitry for positioning
purposes
• Trigger input for external clock or cycle-by-cycle current management
516/1751 RM0090 Rev 19

Figure 86. Advanced-control timer block diagram
Internal clock (CK_INT)

CK_TIM18 from RCC
TIMx_ETR
ETR Trigger
Polarity selection, ETRF controller
Edge detector and Prescaler ETRP
TRGO To other timers
Input filter To DAC and ADC
ITR0
TGI
ITR1
ITR2 TRGI Slave mode
TRC controller
ITR3
TIF_ED
Reset,
UI
Enable,
Up/Down,
TI1FP1 Count REP Register
Encoder
interface U
TIMx_CH1
TI2FP2
OC1
TIMx_CH1N
AutoReload
U
Register
OC1N
Repetition counter
CK_PSC PSC CK_CNT CNT
TIMx_CH2
(prescaler) (counter) DTG[7:0] registers
TIMx_CH1
CC1I CC1I
TI1 TI1FP1
OC2
Input filter & IC1 IC1PS Capture/Compare OC1REF Output
Prescaler DTG
TIMx_CH2N
Edge detector 1 Register control
TI1FP2
U
TRC CC2I CC2I
TIMx_CH2
TI2FP1
TI2 Input filter & IC2 IC2PS Capture/Compare OC2REF
Prescaler DTG Output
TI2FP2 OC2N
U
TIMx_CH3N TIMx_CH3
TRC CC3I CC3I
TIMx_CH3
TI3FP3 OC3
TI3 Input filter & IC3 IC3PS Capture/Compare OC3REF
Prescaler DTG Output
TI3FP4 U
OC3N
TRC CC4I CC4I
TIMx_CH4
TI4FP3
TI4 Input filter & IC4 IC4PS Capture/Compare OC4REF Output
Prescaler control
Edge detector 4 Register
TI4FP4
U
TRC
OC4
TIMx_CH4
TIMx_BKIN
BRK BI
Polarity selection
Clock failure event from clock controller

CSS (Clock Security System)
Interrupt & DMA output
Event
MS39906V3
RM0090 Rev 19 517/1751

588
17.3 TIM1 and TIM8 functional description
17.3.1 Time-base unit

The main block of the programmable advanced-control timer is a 16-bit counter with its
related auto-reload register. The counter can count up, down or both up and down. The
counter clock can be divided by a prescaler.
The counter, the auto-reload register and the prescaler register can be written or read by
software. This is true even when the counter is running.
The time-base unit includes:
• Counter register (TIMx_CNT)
• Prescaler register (TIMx_PSC)
• Auto-reload register (TIMx_ARR)
• Repetition counter register (TIMx_RCR)
The auto-reload register is preloaded. Writing to or reading from the auto-reload register
accesses the preload register. The content of the preload register are transferred into the
shadow register permanently or at each update event (UEV), depending on the auto-reload
preload enable bit (ARPE) in TIMx_CR1 register. The update event is sent when the counter
reaches the overflow (or underflow when downcounting) and if the UDIS bit equals 0 in the
TIMx_CR1 register. It can also be generated by software. The generation of the update
event is described in detailed for each configuration.
The counter is clocked by the prescaler output CK_CNT, which is enabled only when the
counter enable bit (CEN) in TIMx_CR1 register is set (refer also to the slave mode controller
description to get more details on counter enabling).
Note that the counter starts counting 1 clock cycle after setting the CEN bit in the TIMx_CR1
register.
Prescaler description
The prescaler can divide the counter clock frequency by any factor between 1 and 65536. It
is based on a 16-bit counter controlled through a 16-bit register (in the TIMx_PSC register).
It can be changed on the fly as this control register is buffered. The new prescaler ratio is
taken into account at the next update event.
Figure 87 and Figure 88 give some examples of the counter behavior when the prescaler
ratio is changed on the fly:
518/1751 RM0090 Rev 19

Figure 87. Counter timing diagram with prescaler division change from 1 to 2
CK_PSC
CEN
Timerclock = CK_CNT
Counter register F7 F8 F9 FA FB FC 00 01 02 03
Update event (UEV)
Prescaler control register 0 1
Write a new value in TIMx_PSC
Prescaler buffer 0 1
Prescaler counter 0 0 1 0 1 0 1 0 1
MS31076V2
CK_PSC
CEN
Timerclock = CK_CNT
F7 F8 F9 FA FB FC 00 01
Counter register
Update event (UEV)
0 3
Prescaler control register
0 3
Prescaler buffer
0 0 1 2 3 0 1 2 3
Prescaler counter
MS31077V2
RM0090 Rev 19 519/1751

588
17.3.2 Counter modes

Upcounting mode
In upcounting mode, the counter counts from 0 to the auto-reload value (content of the
TIMx_ARR register), then restarts from 0 and generates a counter overflow event.
If the repetition counter is used, the update event (UEV) is generated after upcounting is
repeated for the number of times programmed in the repetition counter register plus one
(TIMx_RCR+1). Else the update event is generated at each counter overflow.
Setting the UG bit in the TIMx_EGR register (by software or by using the slave mode
controller) also generates an update event.
The UEV event can be disabled by software by setting the UDIS bit in the TIMx_CR1
register. This is to avoid updating the shadow registers while writing new values in the
preload registers. Then no update event occurs until the UDIS bit has been written to 0.
However, the counter restarts from 0, as well as the counter of the prescaler (but the
prescale rate does not change). In addition, if the URS bit (update request selection) in
TIMx_CR1 register is set, setting the UG bit generates an update event UEV but without
setting the UIF flag (thus no interrupt or DMA request is sent). This is to avoid generating
both update and capture interrupts when clearing the counter on the capture event.
When an update event occurs, all the registers are updated and the update flag (UIF bit in
TIMx_SR register) is set (depending on the URS bit):
• The repetition counter is reloaded with the content of TIMx_RCR register,
• The auto-reload shadow register is updated with the preload value (TIMx_ARR),
• The buffer of the prescaler is reloaded with the preload value (content of the TIMx_PSC
register).
The following figures show some examples of the counter behavior for different clock
frequencies when TIMx_ARR=0x36.
Figure 89. Counter timing diagram, internal clock divided by 1
CK_PSC
CNT_EN
Timerclock = CK_CNT
Counter register 31 32 33 34 35 36 00 01 02 03 04 05 06 07
Counter overflow
Update event (UEV)
Update interrupt flag (UIF)

MS31078V3
520/1751 RM0090 Rev 19

CK_PSC
CNT_EN
Timerclock = CK_CNT
Counter register 0034 0035 0036 0000 0001 0002 0003
Counter overflow
Update event (UEV)

MS31079V3
CK_PSC
CNT_EN
Timerclock = CK_CNT
Counter register 0035 0036 0000 0001
Counter overflow
Update event (UEV)

MS31080V3
Figure 92. Counter timing diagram, internal clock divided by N
CK_PSC
Timerclock = CK_CNT
Counter register 1F 20 00
Counter overflow
Update event (UEV)

MS31081V3
RM0090 Rev 19 521/1751

588
preloaded)
CK_PSC
CEN
Timerclock = CK_CNT
Counter overflow
Update event (UEV)
Auto-reload preload register FF 36
Write a new value in TIMx_ARR

MS31082V3
preloaded)
CK_PSC
CEN
Timerclock = CK_CNT
Counter register F0 F1 F2 F3 F4 F5 00 01 02 03 04 05 06 07
Counter overflow
Update event (UEV)
Update interrupt flag

(UIF)
Auto-reload preload
F5 36
register
Auto-reload shadow
register F5 36
Write a new value in TIMx_ARR MS31083V2
522/1751 RM0090 Rev 19

Downcounting mode
In downcounting mode, the counter counts from the auto-reload value (content of the
TIMx_ARR register) down to 0, then restarts from the auto-reload value and generates a
counter underflow event.
If the repetition counter is used, the update event (UEV) is generated after downcounting is
repeated for the number of times programmed in the repetition counter register plus one
(TIMx_RCR+1). Else the update event is generated at each counter underflow.
controller) also generates an update event.
The UEV update event can be disabled by software by setting the UDIS bit in TIMx_CR1
preload registers. Then no update event occurs until UDIS bit has been written to 0.
However, the counter restarts from the current auto-reload value, whereas the counter of the
prescaler restarts from 0 (but the prescale rate doesn’t change).
In addition, if the URS bit (update request selection) in TIMx_CR1 register is set, setting the
UG bit generates an update event UEV but without setting the UIF flag (thus no interrupt or
DMA request is sent). This is to avoid generating both update and capture interrupts when
clearing the counter on the capture event.
• The repetition counter is reloaded with the content of TIMx_RCR register
register)
• The auto-reload active register is updated with the preload value (content of the
TIMx_ARR register). Note that the auto-reload is updated before the counter is
reloaded, so that the next period is the expected one
RM0090 Rev 19 523/1751

588
CK_PSC
CNT_EN
Timerclock = CK_CNT
Counter register 05 04 03 02 01 00 36 35 34 33 32 31 30 2F
Counter underflow
(cnt_udf)
Update event (UEV)

(UIF)
MS31184V1
CK_PSC
CNT_EN
Timerclock = CK_CNT
Counter underflow
Update event (UEV)

(UIF)
MS31185V1
524/1751 RM0090 Rev 19

CK_PSC
CNT_EN
Timerclock = CK_CNT
Counter underflow
Update event (UEV)

MS40510V1
CK_PSC
Timerclock = CK_CNT
Counter register 20 1F 00 36
Counter underflow
Update event (UEV)

(UIF)
MS31187V1
RM0090 Rev 19 525/1751

588
Figure 99. Counter timing diagram, update event when repetition counter is not used
CK_PSC
CEN
Timerclock = CK_CNT
Counter underflow
Update event (UEV)

(UIF)
Auto-reload preload
register FF 36

MS31188V1
Center-aligned mode (up/down counting)

In center-aligned mode, the counter counts from 0 to the auto-reload value (content of the
TIMx_ARR register) – 1, generates a counter overflow event, then counts from the auto-
reload value down to 1 and generates a counter underflow event. Then it restarts counting
from 0.
Center-aligned mode is active when the CMS bits in TIMx_CR1 register are not equal to
'00'. The Output compare interrupt flag of channels configured in output is set when: the
counter counts down (Center aligned mode 1, CMS = "01"), the counter counts up (Center
aligned mode 2, CMS = "10") the counter counts up and down (Center aligned mode 3,
CMS = "11").
In this mode, the DIR direction bit in the TIMx_CR1 register cannot be written. It is updated
by hardware and gives the current direction of the counter.
The update event can be generated at each counter overflow and at each counter underflow
or by setting the UG bit in the TIMx_EGR register (by software or by using the slave mode
controller) also generates an update event. In this case, the counter restarts counting from
0, as well as the counter of the prescaler.
The UEV update event can be disabled by software by setting the UDIS bit in the TIMx_CR1
However, the counter continues counting up and down, based on the current auto-reload
value.
UG bit generates an UEV update event but without setting the UIF flag (thus no interrupt or
526/1751 RM0090 Rev 19

• The repetition counter is reloaded with the content of TIMx_RCR register
register)
TIMx_ARR register). Note that if the update source is a counter overflow, the auto-
reload is updated before the counter is reloaded, so that the next period is the expected
one (the counter is loaded with the new value).
frequencies.
Figure 100. Counter timing diagram, internal clock divided by 1, TIMx_ARR = 0x6
CK_PSC
CEN
Timerclock = CK_CNT
Counter underflow
Counter overflow
Update event (UEV)

MS31189V2
1. Here, center-aligned mode 1 is used (for more details refer to Section 17.4: TIM1 and TIM8 registers).
CK_PSC
CNT_EN
Timerclock = CK_CNT
Counter underflow
Update event (UEV)

MS31190V2
RM0090 Rev 19 527/1751

588
Figure 102. Counter timing diagram, internal clock divided by 4, TIMx_ARR=0x36
CK_PSC
CNT_EN
Timerclock = CK_CNT
Counter overflow
Update event (UEV)

MS31191V2
1. Center-aligned mode 2 or 3 is used with an UIF on overflow.
CK_PSC
Timerclock = CK_CNT
Counter underflow
Update event (UEV)

MS31192V2
528/1751 RM0090 Rev 19

Figure 104. Counter timing diagram, update event with ARPE=1 (counter underflow)
CK_PSC
CEN
Timer clock = CK_CNT
Counter overflow
Update event (UEV)
Auto-reload preload register FD 36
Auto-reload active register FD 36
MS31193V3
Figure 105. Counter timing diagram, Update event with ARPE=1 (counter overflow)
CK_PSC
CEN
Counter register F7 F8 F9 FA FB FC 36 35 34 33 32 31 30 2F
Counter overflow
Update event (UEV)
MS31194V2
17.3.3 Repetition counter

Section 17.3.1: Time-base unit describes how the update event (UEV) is generated with
respect to the counter overflows/underflows. It is actually generated only when the repetition
counter has reached zero. This can be useful when generating PWM signals.
This means that data are transferred from the preload registers to the shadow registers
(TIMx_ARR auto-reload register, TIMx_PSC prescaler register, but also TIMx_CCRx
capture/compare registers in compare mode) every N+1 counter overflows or underflows,
where N is the value in the TIMx_RCR repetition counter register.
RM0090 Rev 19 529/1751

588
The repetition counter is decremented:

• At each counter overflow in upcounting mode,
• At each counter underflow in downcounting mode,
• At each counter overflow and at each counter underflow in center-aligned mode.
Although this limits the maximum number of repetition to 128 PWM cycles, it makes it
possible to update the duty cycle twice per PWM period. When refreshing compare
registers only once per PWM period in center-aligned mode, maximum resolution is
2xTck, due to the symmetry of the pattern.
The repetition counter is an auto-reload type; the repetition rate is maintained as defined by
the TIMx_RCR register value (refer to Figure 106). When the update event is generated by
software (by setting the UG bit in TIMx_EGR register) or by hardware through the slave
mode controller, it occurs immediately whatever the value of the repetition counter is and the
repetition counter is reloaded with the content of the TIMx_RCR register.
In center-aligned mode, for odd values of RCR, the update event occurs either on the
overflow or on the underflow depending on when the RCR register was written and when
the counter was started. If the RCR was written before starting the counter, the UEV occurs
on the overflow. If the RCR was written after starting the counter, the UEV occurs on the
underflow. For example for RCR = 3, the UEV is generated on each 4th overflow or
underflow event depending on when RCR was written.
530/1751 RM0090 Rev 19

Figure 106. Update rate examples depending on mode and TIMx_RCR register
settings
Counter-aligned mode Edge-aligned mode
Counter Upcounting Downcounting
TIMx_CNT
TIMx_RCR = 0
UEV
TIMx_RCR = 1
UEV
UEV
TIMx_RCR = 2
TIMx_RCR = 3 UEV
TIMx_RCR = 3
and
UEV
re-synchronization
(by SW) (by SW) (by SW)
UEV Update event: Preload registers transferred to active registers and update interrupt generated
Update Event if the repetition counter underflow occurs when the counter is equal to the auto-reload value.
MSv31195V1
RM0090 Rev 19 531/1751

588
17.3.4 Clock selection

The counter clock can be provided by the following clock sources:
• Internal clock (CK_INT)
• External clock mode1: external input pin
• External clock mode2: external trigger input ETR
• Internal trigger inputs (ITRx): using one timer as prescaler for another timer, for
example, the user can configure Timer 1 to act as a prescaler for Timer 2. Refer to
Using one timer as prescaler for another timer for more details.
Internal clock source (CK_INT)

If the slave mode controller is disabled (SMS=000), then the CEN, DIR (in the TIMx_CR1
register) and UG bits (in the TIMx_EGR register) are actual control bits and can be changed
only by software (except UG which remains cleared automatically). As soon as the CEN bit
is written to 1, the prescaler is clocked by the internal clock CK_INT.
Figure 107 shows the behavior of the control circuit and the upcounter in normal mode,
without prescaler.
Figure 107. Control circuit in normal mode, internal clock divided by 1
Internal clock
CEN=CNT_EN
UG
CNT_INIT
Counter clock = CK_CNT = CK_PSC
MS31085V2
External clock source mode 1

This mode is selected when SMS=111 in the TIMx_SMCR register. The counter can count at
each rising or falling edge on a selected input.
532/1751 RM0090 Rev 19

Figure 108. TI2 external clock connection example
TIMx_SMCR
TS[2:0]
or TI2F or
TI1F or Encoder
ITRx mode
0xx
TI1_ED
100 TRGI External clock
TI1FP1 mode 1 CK_PSC
TI2F_Rising 101
TI2 Edge 0 TI2FP2 ETRF External clock
Filter 110
detector 1 ETRF mode 2
TI2F_Falling 111
CK_INT Internal clock
mode
ICF[3:0] CC2P (internal clock)
TIMx_CCMR1 TIMx_CCER
ECE SMS[2:0]
TIMx_SMCR
MS31196V1
For example, to configure the upcounter to count in response to a rising edge on the TI2
input, use the following procedure:
1. Configure channel 2 to detect rising edges on the TI2 input by writing CC2S = ‘01’ in
the TIMx_CCMR1 register.
2. Configure the input filter duration by writing the IC2F[3:0] bits in the TIMx_CCMR1
register (if no filter is needed, keep IC2F=0000).
3. Select rising edge polarity by writing CC2P=0 and CC2NP=0 in the TIMx_CCER
register.
4. Configure the timer in external clock mode 1 by writing SMS=111 in the TIMx_SMCR
register.
5. Select TI2 as the trigger input source by writing TS=110 in the TIMx_SMCR register.
6. Enable the counter by writing CEN=1 in the TIMx_CR1 register.
Note: The capture prescaler is not used for triggering, so the user does not need to configure it.
When a rising edge occurs on TI2, the counter counts once and the TIF flag is set.
The delay between the rising edge on TI2 and the actual clock of the counter is due to the
resynchronization circuit on TI2 input.
RM0090 Rev 19 533/1751

588
Figure 109. Control circuit in external clock mode 1
TI2
CNT_EN
Counter register 34 35 36
TIF
Write TIF=0
MS31087V2

This mode is selected by writing ECE=1 in the TIMx_SMCR register.
The counter can count at each rising or falling edge on the external trigger input ETR.
Figure 110 gives an overview of the external trigger input block.
Figure 110. External trigger input block
or TI2F or
TI1F or Encoder
mode
TRGI External clock

ETR mode 1 CK_PSC
0
Divider ETRP
ETR pin Filter ETRF External clock
/1, /2, /4, /8
1 fDTS downcounter mode 2
ETP ETPS[1:0] ETF[3:0] mode
(internal clock)
TIMx_SMCR TIMx_SMCR TIMx_SMCR
ECE SMS[2:0]
TIMx_SMCR
MS33116V1
534/1751 RM0090 Rev 19

For example, to configure the upcounter to count each 2 rising edges on ETR, use the
following procedure:
1. As no filter is needed in this example, write ETF[3:0]=0000 in the TIMx_SMCR register.
2. Set the prescaler by writing ETPS[1:0]=01 in the TIMx_SMCR register
3. Select rising edge detection on the ETR pin by writing ETP=0 in the TIMx_SMCR
register
4. Enable external clock mode 2 by writing ECE=1 in the TIMx_SMCR register.
The counter counts once each 2 ETR rising edges.
The delay between the rising edge on ETR and the actual clock of the counter is due to the
resynchronization circuit on the ETRP signal.
f CK_INT
CNT_EN
ETR
ETRP
ETRF
Counter clock =
CK_INT =CK_PSC
MS33111V2
17.3.5 Capture/compare channels

Each Capture/Compare channel is built around a capture/compare register (including a
shadow register), a input stage for capture (with digital filter, multiplexing and prescaler) and
an output stage (with comparator and output control).
Figure 112 to Figure 115 give an overview of one Capture/Compare channel.
The input stage samples the corresponding TIx input to generate a filtered signal TIxF.
Then, an edge detector with polarity selection generates a signal (TIxFPx) which can be
used as trigger input by the slave mode controller or as the capture command. It is
prescaled before the capture register (ICxPS).
RM0090 Rev 19 535/1751

588
Figure 112. Capture/compare channel (example: channel 1 input stage)
TI1F_ED
To the slave mode controller
TI1 TI1F_Rising
Filter TI1F Edge 0 TI1FP1
fDTS downcounter TI1F_Falling 01
detector 1
TI2FP1 IC1 Divider IC1PS
10
/1, /2, /4, /8
ICF[3:0] CC1P/CC1NP TRC
11
TIMx_CCMR1 TIMx_CCER (from slave mode
TI2F_Rising controller)
0
(from channel 2)
CC1S[1:0] ICPS[1:0] CC1E
TI2F_Falling
1
(from channel 2) TIMx_CCMR1 TIMx_CCER
MS33115V1
The output stage generates an intermediate waveform that is then used for reference:
OCxRef (active high). The polarity acts at the end of the chain.
Figure 113. Capture/compare channel 1 main circuit
APB Bus
MCU-peripheral interface
(if 16-bit)
8 8
high
S write CCR1H
low
Read CCR1H S write_in_progress

read_in_progress write CCR1L
Read CCR1L Capture/compare preload register R
R Output CC1S[1]
capture_transfer compare_transfer mode
CC1S[0]
CC1S[1] Input
mode OC1PE
CC1S[0] Capture/compare shadow register OC1PE
UEV
(from time TIMx_CCMR1
Comparator
IC1PS Capture base unit)
CC1E CNT>CCR1
Counter CNT=CCR1
CC1G
TIMx_EGR
MS31089V3
536/1751 RM0090 Rev 19

Figure 114. Output stage of capture/compare channel (channel 1 to 3)

To the master mode
controller
ETRF 0 Output OC1
enable
‘0’ 1
x0 circuit
OC1REF OC1REFC
01
OC1_DT CC1P
CNT>CCR1 Output 11
Output Dead-time TIM1_CCER
mode
CNT=CCR1 selector generator
controller OC1N_DT
11
10 0
(1) Output OC1N
OCxREF ‘0’ 0x enable
OC5REF 1 circuit
CC1NE CC1E TIM1_CCER
OC1CE OC1M[3:0] DTG[7:0] CC1NE CC1E CC1NP MOE OSSI OSSR

TIM1_CCMR1 TIM1_BDTR TIM1_CCER TIM1_CCER TIM1_BDTR
MS35909V1
Figure 115. Output stage of capture/compare channel (channel 4)

To the master
mode controller
ETR 0
Output
OC4
enable
1 circuit
CNT>CCR4 Output
mode CC4P
CNT>CCR4 OC4REF
controller TIM1_CCER
CC4E TIM1_CCER
MOE OSSI TIM1_BDTR

OC1M[3:0]
TIM1_CCMR2 OIS4 TIM1_CR2
MS37370V1
The capture/compare block is made of one preload register and one shadow register. Write
and read always access the preload register.
In capture mode, captures are actually done in the shadow register, which is copied into the
preload register.
In compare mode, the content of the preload register is copied into the shadow register
which is compared to the counter.
RM0090 Rev 19 537/1751

588
17.3.6 Input capture mode

In Input capture mode, the Capture/Compare registers (TIMx_CCRx) are used to latch the
value of the counter after a transition detected by the corresponding ICx signal. When a
capture occurs, the corresponding CCXIF flag (TIMx_SR register) is set and an interrupt or
a DMA request can be sent if they are enabled. If a capture occurs while the CCxIF flag was
already high, then the over-capture flag CCxOF (TIMx_SR register) is set. CCxIF can be
cleared by software by writing it to ‘0’ or by reading the captured data stored in the
TIMx_CCRx register. CCxOF is cleared when written to ‘0’.
The following example shows how to capture the counter value in TIMx_CCR1 when TI1
input rises. To do this, use the following procedure:
• Select the active input: TIMx_CCR1 must be linked to the TI1 input, so write the CC1S
bits to 01 in the TIMx_CCMR1 register. As soon as CC1S becomes different from 00,
the channel is configured in input and the TIMx_CCR1 register becomes read-only.
• Program the needed input filter duration with respect to the signal connected to the
timer (by programming ICxF bits in the TIMx_CCMRx register if the input is a TIx input).
Let’s imagine that, when toggling, the input signal is not stable during at must five
internal clock cycles. We must program a filter duration longer than these five clock
cycles. We can validate a transition on TI1 when 8 consecutive samples with the new
level have been detected (sampled at fDTS frequency). Then write IC1F bits to 0011 in
• Select the edge of the active transition on the TI1 channel by writing CC1P and CC1NP
bits to 0 in the TIMx_CCER register (rising edge in this case).
• Program the input prescaler. In our example, we wish the capture to be performed at
each valid transition, so the prescaler is disabled (write IC1PS bits to ‘00’ in the
TIMx_CCMR1 register).
• Enable capture from the counter into the capture register by setting the CC1E bit in the
TIMx_CCER register.
• If needed, enable the related interrupt request by setting the CC1IE bit in the
TIMx_DIER register, and/or the DMA request by setting the CC1DE bit in the
TIMx_DIER register.
When an input capture occurs:
• The TIMx_CCR1 register gets the value of the counter on the active transition.
• CC1IF flag is set (interrupt flag). CC1OF is also set if at least two consecutive captures
occurred whereas the flag was not cleared.
• An interrupt is generated depending on the CC1IE bit.
• A DMA request is generated depending on the CC1DE bit.
In order to handle the overcapture, it is recommended to read the data before the
overcapture flag. This is to avoid missing an overcapture which could happen after reading
the flag and before reading the data.
Note: IC interrupt and/or DMA requests can be generated by software by setting the
corresponding CCxG bit in the TIMx_EGR register.
538/1751 RM0090 Rev 19

17.3.7 PWM input mode

This mode is a particular case of input capture mode. The procedure is the same except:
• Two ICx signals are mapped on the same TIx input.
• These 2 ICx signals are active on edges with opposite polarity.
• One of the two TIxFP signals is selected as trigger input and the slave mode controller
is configured in reset mode.
For example, user can measure the period (in TIMx_CCR1 register) and the duty cycle (in
TIMx_CCR2 register) of the PWM applied on TI1 using the following procedure (depending
on CK_INT frequency and prescaler value):
• Select the active input for TIMx_CCR1: write the CC1S bits to 01 in the TIMx_CCMR1
register (TI1 selected).
• Select the active polarity for TI1FP1 (used both for capture in TIMx_CCR1 and counter
clear): write the CC1P and CC1NP bits to ‘0’ (active on rising edge).
• Select the active polarity for TI1FP2 (used for capture in TIMx_CCR2): write the CC2P
and CC2NP bits to ‘1’ (active on falling edge).
• Select the valid trigger input: write the TS bits to 101 in the TIMx_SMCR register
(TI1FP1 selected).
• Configure the slave mode controller in reset mode: write the SMS bits to 100 in the
TIMx_SMCR register.
• Enable the captures: write the CC1E and CC2E bits to ‘1’ in the TIMx_CCER register.
Figure 116. PWM input mode timing
17.3.8 Forced output mode

In output mode (CCxS bits = 00 in the TIMx_CCMRx register), each output compare signal
(OCxREF and then OCx/OCxN) can be forced to active or inactive level directly by software,
independently of any comparison between the output compare register and the counter.
To force an output compare signal (OCXREF/OCx) to its active level, the user just needs to
write 101 in the OCxM bits in the corresponding TIMx_CCMRx register. Thus OCXREF is
RM0090 Rev 19 539/1751

588
forced high (OCxREF is always active high) and OCx get opposite value to CCxP polarity
bit.
For example: CCxP=0 (OCx active high) => OCx is forced to high level.
The OCxREF signal can be forced low by writing the OCxM bits to 100 in the TIMx_CCMRx
register.
Anyway, the comparison between the TIMx_CCRx shadow register and the counter is still
performed and allows the flag to be set. Interrupt and DMA requests can be sent
accordingly. This is described in the output compare mode section below.
17.3.9 Output compare mode

This function is used to control an output waveform or indicating when a period of time has
elapsed.
When a match is found between the capture/compare register and the counter, the output
compare function:
• Assigns the corresponding output pin to a programmable value defined by the output
compare mode (OCxM bits in the TIMx_CCMRx register) and the output polarity (CCxP
bit in the TIMx_CCER register). The output pin can keep its level (OCXM=000), be set
active (OCxM=001), be set inactive (OCxM=010) or can toggle (OCxM=011) on match.
• Sets a flag in the interrupt status register (CCxIF bit in the TIMx_SR register).
• Generates an interrupt if the corresponding interrupt mask is set (CCXIE bit in the
TIMx_DIER register).
• Sends a DMA request if the corresponding enable bit is set (CCxDE bit in the
TIMx_DIER register, CCDS bit in the TIMx_CR2 register for the DMA request
selection).
The TIMx_CCRx registers can be programmed with or without preload registers using the
OCxPE bit in the TIMx_CCMRx register.
In output compare mode, the update event UEV has no effect on OCxREF and OCx output.
The timing resolution is one count of the counter. Output compare mode can also be used to
output a single pulse (in One Pulse mode).
Procedure:
1. Select the counter clock (internal, external, prescaler).
2. Write the desired data in the TIMx_ARR and TIMx_CCRx registers.
3. Set the CCxIE bit if an interrupt request is to be generated.
4. Select the output mode. For example:
– Write OCxM = 011 to toggle OCx output pin when CNT matches CCRx
– Write OCxPE = 0 to disable preload register
– Write CCxP = 0 to select active high polarity
– Write CCxE = 1 to enable the output
5. Enable the counter by setting the CEN bit in the TIMx_CR1 register.
The TIMx_CCRx register can be updated at any time by software to control the output
waveform, provided that the preload register is not enabled (OCxPE=’0’, else TIMx_CCRx
shadow register is updated only at the next update event UEV). An example is given in
Figure 117.
540/1751 RM0090 Rev 19

Figure 117. Output compare mode, toggle on OC1.

Write B201h in the CC1R register
TIM1_CNT 0039 003A 003B B200 B201
TIM1_CCR1 003A B201
OC1REF= OC1
Match detected on CCR1

Interrupt generated if enabled
MS31092V2
17.3.10 PWM mode

Pulse Width Modulation mode allows generating a signal with a frequency determined by
the value of the TIMx_ARR register and a duty cycle determined by the value of the
TIMx_CCRx register.
The PWM mode can be selected independently on each channel (one PWM per OCx
output) by writing ‘110’ (PWM mode 1) or ‘111’ (PWM mode 2) in the OCxM bits in the
TIMx_CCMRx register. The corresponding preload register must be enabled by setting the
OCxPE bit in the TIMx_CCMRx register, and eventually the auto-reload preload register (in
upcounting or center-aligned modes) by setting the ARPE bit in the TIMx_CR1 register.
As the preload registers are transferred to the shadow registers only when an update event
occurs, before starting the counter, the user must initialize all the registers by setting the UG
bit in the TIMx_EGR register.
OCx polarity is software programmable using the CCxP bit in the TIMx_CCER register. It
can be programmed as active high or active low. OCx output is enabled by a combination of
the CCxE, CCxNE, MOE, OSSI and OSSR bits (TIMx_CCER and TIMx_BDTR registers).
Refer to the TIMx_CCER register description for more details.
In PWM mode (1 or 2), TIMx_CNT and TIMx_CCRx are always compared to determine
whether TIMx_CCRx ≤ TIMx_CNT or TIMx_CNT ≤ TIMx_CCRx (depending on the direction
of the counter).
The timer is able to generate PWM in edge-aligned mode or center-aligned mode
depending on the CMS bits in the TIMx_CR1 register.
PWM edge-aligned mode

• Upcounting configuration
Upcounting is active when the DIR bit in the TIMx_CR1 register is low. Refer to
Upcounting mode.
In the following example, we consider PWM mode 1. The reference PWM signal
OCxREF is high as long as TIMx_CNT < TIMx_CCRx else it becomes low. If the
RM0090 Rev 19 541/1751

588
compare value in TIMx_CCRx is greater than the auto-reload value (in TIMx_ARR)
then OCxREF is held at ‘1’. If the compare value is 0 then OCxRef is held at ‘0’.
Figure 118 shows some edge-aligned PWM waveforms in an example where
TIMx_ARR=8.
Figure 118. Edge-aligned PWM waveforms (ARR=8)
Counter register 0 1 2 3 4 5 6 7 8 0 1
OCXREF
CCRx=4
CCxIF
OCXREF
CCRx=8
CCxIF
OCXREF ‘1’
CCRx>8
CCxIF
OCXREF ‘0’
CCRx=0
CCxIF
MS31093V1
• Downcounting configuration
Downcounting is active when DIR bit in TIMx_CR1 register is high. Refer to
Downcounting mode
In PWM mode 1, the reference signal OCxRef is low as long as
TIMx_CNT > TIMx_CCRx else it becomes high. If the compare value in TIMx_CCRx is
greater than the auto-reload value in TIMx_ARR, then OCxREF is held at ‘1’. 0% PWM
is not possible in this mode.
PWM center-aligned mode

Center-aligned mode is active when the CMS bits in TIMx_CR1 register are different from
‘00’ (all the remaining configurations having the same effect on the OCxRef/OCx signals).
The compare flag is set when the counter counts up, when it counts down or both when it
counts up and down depending on the CMS bits configuration. The direction bit (DIR) in the
542/1751 RM0090 Rev 19

TIMx_CR1 register is updated by hardware and must not be changed by software. Refer to
Center-aligned mode (up/down counting).
Figure 119 shows some center-aligned PWM waveforms in an example where:
• TIMx_ARR=8,
• PWM mode is the PWM mode 1,
• The flag is set when the counter counts down corresponding to the center-aligned
mode 1 selected for CMS=01 in TIMx_CR1 register.
Figure 119. Center-aligned PWM waveforms (ARR=8)
Counter register 0 1 2 3 4 5 6 7 8 7 6 5 4 3 2 1 0 1
OCxREF
CCRx = 4
CCxIF CMS=01
CMS=10
CMS=11
OCxREF
CCRx = 7
CMS=10 or 11
CCxIF
'1'
OCxREF
CCRx = 8
CCxIF CMS=01
CMS=10
CMS=11
'1'
OCxREF
CCRx > 8
CCxIF CMS=01
CMS=10
CMS=11
'0'
OCxREF
CCRx = 0
CCxIF CMS=01
CMS=10
CMS=11
ai14681b
Hints on using center-aligned mode:

• When starting in center-aligned mode, the current up-down configuration is used. It
means that the counter counts up or down depending on the value written in the DIR bit
RM0090 Rev 19 543/1751

588
in the TIMx_CR1 register. Moreover, the DIR and CMS bits must not be changed at the
same time by the software.
• Writing to the counter while running in center-aligned mode is not recommended as it
can lead to unexpected results. In particular:
– The direction is not updated if the user writes a value in the counter greater than
the auto-reload value (TIMx_CNT>TIMx_ARR). For example, if the counter was
counting up, it will continue to count up.
– The direction is updated if the user writes 0 or write the TIMx_ARR value in the
counter but no Update Event UEV is generated.
• The safest way to use center-aligned mode is to generate an update by software
(setting the UG bit in the TIMx_EGR register) just before starting the counter and not to
write the counter while it is running.
17.3.11 Complementary outputs and dead-time insertion

The advanced-control timers (TIM1 and TIM8) can output two complementary signals and
manage the switching-off and the switching-on instants of the outputs.
This time is generally known as dead-time and it has to be adjust it depending on the
devices connected to the outputs and their characteristics (intrinsic delays of level-shifters,
delays due to power switches...)
User can select the polarity of the outputs (main output OCx or complementary OCxN)
independently for each output. This is done by writing to the CCxP and CCxNP bits in the
TIMx_CCER register.
The complementary signals OCx and OCxN are activated by a combination of several
control bits: the CCxE and CCxNE bits in the TIMx_CCER register and the MOE, OISx,
OISxN, OSSI and OSSR bits in the TIMx_BDTR and TIMx_CR2 registers. Refer to Table 95
for more details. In particular, the dead-time is activated when switching to the IDLE state
(MOE falling down to 0).
Dead-time insertion is enabled by setting both CCxE and CCxNE bits, and the MOE bit if the
break circuit is present. DTG[7:0] bits of the TIMx_BDTR register are used to control the
dead-time generation for all channels. From a reference waveform OCxREF, it generates 2
outputs OCx and OCxN. If OCx and OCxN are active high:
• The OCx output signal is the same as the reference signal except for the rising edge,
which is delayed relative to the reference rising edge.
• The OCxN output signal is the opposite of the reference signal except for the rising
edge, which is delayed relative to the reference falling edge.
If the delay is greater than the width of the active output (OCx or OCxN) then the
corresponding pulse is not generated.
The following figures show the relationships between the output signals of the dead-time
generator and the reference signal OCxREF. (we suppose CCxP=0, CCxNP=0, MOE=1,
CCxE=1 and CCxNE=1 in these examples)
544/1751 RM0090 Rev 19

Figure 120. Complementary output with dead-time insertion.
OCxREF
OCx
delay
OCxN
delay
MS31095V1
Figure 121. Dead-time waveforms with delay greater than the negative pulse.
OCxREF
OCx
delay
OCxN
MS31096V1
Figure 122. Dead-time waveforms with delay greater than the positive pulse.
OCxREF
OCx
OCxN
delay
MS31097V1
The dead-time delay is the same for each of the channels and is programmable with the
DTG bits in the TIMx_BDTR register. Refer to Section 17.4.18: TIM1 and TIM8 break and
dead-time register (TIMx_BDTR) for delay calculation.
Re-directing OCxREF to OCx or OCxN

In output mode (forced, output compare or PWM), OCxREF can be re-directed to the OCx
output or to OCxN output by configuring the CCxE and CCxNE bits in the TIMx_CCER
register.
This allows the user to send a specific waveform (such as PWM or static active level) on
one output while the complementary remains at its inactive level. Other possibilities are to
RM0090 Rev 19 545/1751

588
have both outputs at inactive level or both outputs active and complementary with
dead-time.
Note: When only OCxN is enabled (CCxE=0, CCxNE=1), it is not complemented and becomes
active as soon as OCxREF is high. For example, if CCxNP=0 then OCxN=OCxRef. On the
other hand, when both OCx and OCxN are enabled (CCxE=CCxNE=1) OCx becomes
active when OCxREF is high whereas OCxN is complemented and becomes active when
OCxREF is low.
17.3.12 Using the break function

When using the break function, the output enable signals and inactive levels are modified
according to additional control bits (MOE, OSSI and OSSR bits in the TIMx_BDTR register,
OISx and OISxN bits in the TIMx_CR2 register). In any case, the OCx and OCxN outputs
cannot be set both to active level at a given time. Refer to Table 95 for more details.
The break source can be either the break input pin or a clock failure event, generated by the
Clock Security System (CSS), from the Reset Clock Controller. For further information on
the Clock Security System, refer to Section 7.2.7: Clock security system (CSS).
When exiting from reset, the break circuit is disabled and the MOE bit is low. User can
enable the break function by setting the BKE bit in the TIMx_BDTR register. The break input
polarity can be selected by configuring the BKP bit in the same register. BKE and BKP can
be modified at the same time. When the BKE and BKP bits are written, a delay of 1 APB
clock cycle is applied before the writing is effective. Consequently, it is necessary to wait 1
APB clock period to correctly read back the bit after the write operation.
Because MOE falling edge can be asynchronous, a resynchronization circuit has been
inserted between the actual signal (acting on the outputs) and the synchronous control bit
(accessed in the TIMx_BDTR register). It results in some delays between the asynchronous
and the synchronous signals. In particular, if MOE is written to 1 whereas it was low, a delay
(dummy instruction) must be inserted before reading it correctly. This is because the user
writes an asynchronous signal, but reads a synchronous signal.
When a break occurs (selected level on the break input):
• The MOE bit is cleared asynchronously, putting the outputs in inactive state, idle state
or in reset state (selected by the OSSI bit). This feature functions even if the MCU
oscillator is off.
• Each output channel is driven with the level programmed in the OISx bit in the
TIMx_CR2 register as soon as MOE=0. If OSSI=0 then the timer releases the enable
output else the enable output remains high.
• When complementary outputs are used:
– The outputs are first put in reset state inactive state (depending on the polarity).
This is done asynchronously so that it works even if no clock is provided to the
timer.
– If the timer clock is still present, then the dead-time generator is reactivated in
order to drive the outputs with the level programmed in the OISx and OISxN bits
after a dead-time. Even in this case, OCx and OCxN cannot be driven to their
546/1751 RM0090 Rev 19

active level together. Note that because of the resynchronization on MOE, the
dead-time duration is a bit longer than usual (around 2 ck_tim clock cycles).
– If OSSI=0 then the timer releases the enable outputs else the enable outputs
remain or become high as soon as one of the CCxE or CCxNE bits is high.
• The break status flag (BIF bit in the TIMx_SR register) is set. An interrupt can be
generated if the BIE bit in the TIMx_DIER register is set. A DMA request can be sent if
the BDE bit in the TIMx_DIER register is set.
• If the AOE bit in the TIMx_BDTR register is set, the MOE bit is automatically set again
at the next update event UEV. This can be used to perform a regulation, for instance.
Else, MOE remains low until it is written to ‘1’ again. In this case, it can be used for
security and the break input can be connected to an alarm from power drivers, thermal
sensors or any security components.
Note: The break inputs is acting on level. Thus, the MOE cannot be set while the break input is
active (neither automatically nor by software). In the meantime, the status flag BIF cannot
be cleared.
The break can be generated by the BRK input which has a programmable polarity and an
enable bit BKE in the TIMx_BDTR register.
There are two solutions to generate a break:
• By using the BRK input which has a programmable polarity and an enable bit BKE in
the TIMx_BDTR register
• By software through the BG bit of the TIMx_EGR register.
In addition to the break input and the output management, a write protection has been
implemented inside the break circuit to safeguard the application. It allows freezing the
configuration of several parameters (dead-time duration, OCx/OCxN polarities and state
when disabled, OCxM configurations, break enable and polarity). The user can choose from
three levels of protection selected by the LOCK bits in the TIMx_BDTR register. Refer to
Section 17.4.18: TIM1 and TIM8 break and dead-time register (TIMx_BDTR). The LOCK
bits can be written only once after an MCU reset.
RM0090 Rev 19 547/1751

588
Figure 123 shows an example of behavior of the outputs in response to a break.
Figure 123. Output behavior in response to a break.
BREAK (MOE )
OCxREF
OCx
(OCxN not implemented, CCxP=0, OISx=1)
OCx
OCx
OCx
OCx
OCxN delay delay delay

(CCxE=1, CCxP=0, OISx=0, CCxNE=1, CCxNP=0, OISxN=1)
OCx
OCxN delay delay delay

OCx
OCxN delay
OCx
OCxN delay
OCx
OCxN
(CCxE=1, CCxP=0, CCxNE=0, CCxNP=0, OISx=OISxN=0 or OISx=OISxN=1)
MS31098V1
548/1751 RM0090 Rev 19

17.3.13 Clearing the OCxREF signal on an external event

The OCxREF signal for a given channel can be driven Low by applying a High level to the
ETRF input (OCxCE enable bit of the corresponding TIMx_CCMRx register set to ‘1’). The
OCxREF signal remains Low until the next update event, UEV, occurs.
This function can only be used in output compare and PWM modes, and does not work in
forced mode.
For example, the ETR signal can be connected to the output of a comparator to be used for
current handling. In this case, the ETR must be configured as follow:
1. The External Trigger Prescaler should be kept off: bits ETPS[1:0] of the TIMx_SMCR
register set to ‘00’.
2. The external clock mode 2 must be disabled: bit ECE of the TIMx_SMCR register set to
‘0’.
3. The External Trigger Polarity (ETP) and the External Trigger Filter (ETF) can be
configured according to the user needs.
Figure 124 shows the behavior of the OCxREF signal when the ETRF Input becomes High,
for both values of the enable bit OCxCE. In this example, the timer TIMx is programmed in
PWM mode.
Figure 124. Clearing TIMx OCxREF
(CCRx)
Counter (CNT)
ETRF
OCxREF (OCxCE = ‘0’)
ETRF becomes high ETRF still high

MSv35889V1
Note: In case of a PWM with a 100% duty cycle (if CCRx>ARR), then OCxREF is enabled again at
the next counter overflow.
RM0090 Rev 19 549/1751

588
17.3.14 6-step PWM generation

When complementary outputs are used on a channel, preload bits are available on the
OCxM, CCxE and CCxNE bits. The preload bits are transferred to the shadow bits at the
COM commutation event. The user can thus program in advance the configuration for the
next step and change the configuration of all the channels at the same time. COM can be
generated by software by setting the COM bit in the TIMx_EGR register or by hardware (on
TRGI rising edge).
A flag is set when the COM event occurs (COMIF bit in the TIMx_SR register), which can
generate an interrupt (if the COMIE bit is set in the TIMx_DIER register) or a DMA request
(if the COMDE bit is set in the TIMx_DIER register).
Figure 125 describes the behavior of the OCx and OCxN outputs when a COM event
occurs, in 3 different examples of programmed configurations.
Figure 125. 6-step generation, COM example (OSSR=1)
550/1751 RM0090 Rev 19

17.3.15 One-pulse mode

One-pulse mode (OPM) is a particular case of the previous modes. It allows the counter to
be started in response to a stimulus and to generate a pulse with a programmable length
after a programmable delay.
Starting the counter can be controlled through the slave mode controller. Generating the
waveform can be done in output compare mode or PWM mode. Select One-pulse mode by
setting the OPM bit in the TIMx_CR1 register. This makes the counter stop automatically at
the next update event UEV.
A pulse can be correctly generated only if the compare value is different from the counter
initial value. Before starting (when the timer is waiting for the trigger), the configuration must
be:
• In upcounting: CNT < CCRx ≤ ARR (in particular, 0 < CCRx)
• In downcounting: CNT > CCRx
Figure 126. Example of one pulse mode.
TI2
OC1REF
OC1
TIM1_ARR
Counter
TIM1_CCR1
0 t
tDELAY tPULSE
MS31099V2
For example the user may want to generate a positive pulse on OC1 with a length of tPULSE
and after a delay of tDELAY as soon as a positive edge is detected on the TI2 input pin.
Let’s use TI2FP2 as trigger 1:
• Map TI2FP2 to TI2 by writing CC2S=’01’ in the TIMx_CCMR1 register.
• TI2FP2 must detect a rising edge, write CC2P=’0’ and CC2NP=’0’ in the TIMx_CCER
register.
• Configure TI2FP2 as trigger for the slave mode controller (TRGI) by writing TS=’110’ in
the TIMx_SMCR register.
• TI2FP2 is used to start the counter by writing SMS to ‘110’ in the TIMx_SMCR register
(trigger mode).
RM0090 Rev 19 551/1751

588
The OPM waveform is defined by writing the compare registers (taking into account the
clock frequency and the counter prescaler).
• The tDELAY is defined by the value written in the TIMx_CCR1 register.
• The tPULSE is defined by the difference between the auto-reload value and the compare
value (TIMx_ARR - TIMx_CCR1).
• Let us say the user wants to build a waveform with a transition from ‘0’ to ‘1’ when a
compare match occurs and a transition from ‘1’ to ‘0’ when the counter reaches the
auto-reload value. To do this, enable PWM mode 2 by writing OC1M=111 in the
TIMx_CCMR1 register. The user can optionally enable the preload registers by writing
OC1PE=’1’ in the TIMx_CCMR1 register and ARPE in the TIMx_CR1 register. In this
case the compare value must be written in the TIMx_CCR1 register, the auto-reload
value in the TIMx_ARR register, generate an update by setting the UG bit and wait for
external trigger event on TI2. CC1P is written to ‘0’ in this example.
In our example, the DIR and CMS bits in the TIMx_CR1 register should be low.
The user only wants one pulse (Single mode), so '1’ must be written in the OPM bit in the
TIMx_CR1 register to stop the counter at the next update event (when the counter rolls over
from the auto-reload value back to 0). When OPM bit in the TIMx_CR1 register is set to '0',
so the Repetitive Mode is selected.
Particular case: OCx fast enable:
In One-pulse mode, the edge detection on TIx input set the CEN bit which enables the
counter. Then the comparison between the counter and the compare value makes the
output toggle. But several clock cycles are needed for these operations and it limits the
minimum delay tDELAY min we can get.
If the user wants to output a waveform with the minimum delay, the OCxFE bit in the
TIMx_CCMRx register must be set. Then OCxRef (and OCx) are forced in response to the
stimulus, without taking in account the comparison. Its new level is the same as if a compare
match had occurred. OCxFE acts only if the channel is configured in PWM1 or PWM2
mode.
17.3.16 Encoder interface mode

To select Encoder Interface mode write SMS=‘001’ in the TIMx_SMCR register if the
counter is counting on TI2 edges only, SMS=’010’ if it is counting on TI1 edges only and
SMS=’011’ if it is counting on both TI1 and TI2 edges.
Select the TI1 and TI2 polarity by programming the CC1P and CC2P bits in the TIMx_CCER
register. When needed, the user can program the input filter as well. CC1NP and CC2NP
must be kept low.
The two inputs TI1 and TI2 are used to interface to an incremental encoder. Refer to
Table 93. The counter is clocked by each valid transition on TI1FP1 or TI2FP2 (TI1 and TI2
after input filter and polarity selection, TI1FP1=TI1 if not filtered and not inverted,
TI2FP2=TI2 if not filtered and not inverted) assuming that it is enabled (CEN bit in
TIMx_CR1 register written to ‘1’). The sequence of transitions of the two inputs is evaluated
and generates count pulses as well as the direction signal. Depending on the sequence the
counter counts up or down, the DIR bit in the TIMx_CR1 register is modified by hardware
accordingly. The DIR bit is calculated at each transition on any input (TI1 or TI2), whatever
the counter is counting on TI1 only, TI2 only or both TI1 and TI2.
Encoder interface mode acts simply as an external clock with direction selection. This
means that the counter just counts continuously between 0 and the auto-reload value in the
552/1751 RM0090 Rev 19

TIMx_ARR register (0 to ARR or ARR down to 0 depending on the direction). So user must
configure TIMx_ARR before starting. in the same way, the capture, compare, prescaler,
repetition counter, trigger output features continue to work as normal. Encoder mode and
External clock mode 2 are not compatible and must not be selected together.
In this mode, the counter is modified automatically following the speed and the direction of
the incremental encoder and its content, therefore, always represents the encoder’s
position. The count direction correspond to the rotation direction of the connected sensor.
Table 93 summarizes the possible combinations, assuming TI1 and TI2 do not switch at the
same time.
Table 93. Counting direction versus encoder signals

TI1FP1 signal TI2FP2 signal
Active Level on opposite signal
edge (TI1FP1 for TI2, TI2FP2 for TI1)
Rising Falling Rising Falling
Counting on High Down Up No Count No Count

TI1 only Low Up Down No Count No Count
Counting on High No Count No Count Up Down

TI2 only Low No Count No Count Down Up
Counting on High Down Up Up Down

TI1 and TI2 Low Up Down Down Up
An external incremental encoder can be connected directly to the MCU without external
interface logic. However, comparators are normally be used to convert the encoder’s
differential outputs to digital signals. This greatly increases noise immunity. The third
encoder output which indicate the mechanical zero position, may be connected to an
external interrupt input and trigger a counter reset.
Figure 127 gives an example of counter operation, showing count signal generation and
direction control. It also shows how input jitter is compensated where both edges are
selected. This might occur if the sensor is positioned near to one of the switching points. For
this example we assume that the configuration is the following:
• CC1S=’01’ (TIMx_CCMR1 register, TI1FP1 mapped on TI1).
• CC2S=’01’ (TIMx_CCMR2 register, TI1FP2 mapped on TI2).
• CC1P=’0’, CC1NP=’0’, and IC1F = ‘0000’ (TIMx_CCER register, TI1FP1 non-inverted,
TI1FP1=TI1).
• CC2P=’0’, CC2NP=’0’, and IC2F = ‘0000’ (TIMx_CCER register, TI1FP2 non-inverted,
TI1FP2= TI2).
• SMS=’011’ (TIMx_SMCR register, both inputs are active on both rising and falling
edges).
• CEN=’1’ (TIMx_CR1 register, Counter enabled).
RM0090 Rev 19 553/1751

588
Figure 127. Example of counter operation in encoder interface mode.

forward jitter backward jitter forward
TI1
TI2
Counter
up down up
MS33107V1
Figure 128 gives an example of counter behavior when TI1FP1 polarity is inverted (same
configuration as above except CC1P=’1’).
Figure 128. Example of encoder interface mode with TI1FP1 polarity inverted.
TI1
TI2
Counter
down up down
MS33108V1
The timer, when configured in Encoder Interface mode provides information on the sensor’s
current position.The user can obtain dynamic information (speed, acceleration,
deceleration) by measuring the period between two encoder events using a second timer
configured in capture mode. The output of the encoder which indicates the mechanical zero
can be used for this purpose. Depending on the time between two events, the counter can
also be read at regular times. This can be done by latching the counter value into a third
input capture register if available (then the capture signal must be periodic and can be
generated by another timer). when available, it is also possible to read its value through a
DMA request generated by a real-time clock.
554/1751 RM0090 Rev 19

17.3.17 Timer input XOR function

The TI1S bit in the TIMx_CR2 register, allows the input filter of channel 1 to be connected to
the output of a XOR gate, combining the three input pins TIMx_CH1, TIMx_CH2 and
TIMx_CH3.
The XOR output can be used with all the timer input functions such as trigger or input
capture. An example of this feature used to interface Hall sensors is given in
Section 17.3.18.
17.3.18 Interfacing with Hall sensors

This is done using the advanced-control timers (TIM1 or TIM8) to generate PWM signals to
drive the motor and another timer TIMx (TIM2, TIM3, TIM4 or TIM5) referred to as
“interfacing timer” in Figure 129. The “interfacing timer” captures the 3 timer input pins
(TIMx_CH1, TIMx_CH2, and TIMx_CH3) connected through a XOR to the TI1 input channel
(selected by setting the TI1S bit in the TIMx_CR2 register).
The slave mode controller is configured in reset mode; the slave input is TI1F_ED. Thus,
each time one of the 3 inputs toggles, the counter restarts counting from 0. This creates a
time base triggered by any change on the Hall inputs.
On the “interfacing timer”, capture/compare channel 1 is configured in capture mode,
capture signal is TRC (see Figure 112). The captured value, which corresponds to the time
elapsed between 2 changes on the inputs, gives information about motor speed.
The “interfacing timer” can be used in output mode to generate a pulse which changes the
configuration of the channels of the advanced-control timer (TIM1 or TIM8) (by triggering a
COM event). The TIM1 timer is used to generate PWM signals to drive the motor. To do this,
the interfacing timer channel must be programmed so that a positive pulse is generated
after a programmed delay (in output compare or PWM mode). This pulse is sent to the
advanced-control timer (TIM1 or TIM8) through the TRGO output.
Example: the user wants to change the PWM configuration of the advanced-control timer
TIM1 after a programmed delay each time a change occurs on the Hall inputs connected to
one of the TIMx timers.
• Configure 3 timer inputs ORed to the TI1 input channel by writing the TI1S bit in the
TIMx_CR2 register to ‘1’,
• Program the time base: write the TIMx_ARR to the max value (the counter must be
cleared by the TI1 change. Set the prescaler to get a maximum counter period longer
than the time between 2 changes on the sensors,
• Program channel 1 in capture mode (TRC selected): write the CC1S bits in the
TIMx_CCMR1 register to ‘11’. The user can also program the digital filter if needed,
• Program channel 2 in PWM 2 mode with the desired delay: write the OC2M bits to ‘111’
and the CC2S bits to ‘00’ in the TIMx_CCMR1 register,
• Select OC2REF as trigger output on TRGO: write the MMS bits in the TIMx_CR2
register to ‘101’,
In the advanced-control timer TIM1, the right ITR input must be selected as trigger input, the
timer is programmed to generate PWM signals, the capture/compare control signals are
preloaded (CCPC=1 in the TIMx_CR2 register) and the COM event is controlled by the
trigger input (CCUS=1 in the TIMx_CR2 register). The PWM control bits (CCxE, OCxM) are
written after a COM event for the next step (this can be done in an interrupt subroutine
generated by the rising edge of OC2REF).
RM0090 Rev 19 555/1751

588
Figure 129 describes this example.
Figure 129. Example of Hall sensor interface
TIH1
TIH2
Interfacing timer
TIH3
Counter (CNT)
(CCR2)
CCR1 C7A3 C7A8 C794 C7A5 C7AB C796
TRGO=OC2REF
COM
OC1
OC1N
Advanced-control timers
OC2
OC2N
OC3
OC3N
Write CCxE, CCxNE

and OCxM for next step
ai17335b
556/1751 RM0090 Rev 19

17.3.19 TIMx and external trigger synchronization

The TIMx timer can be synchronized with an external trigger in several modes: Reset mode,
Gated mode and Trigger mode.
Slave mode: Reset mode

The counter and its prescaler can be reinitialized in response to an event on a trigger input.
Moreover, if the URS bit from the TIMx_CR1 register is low, an update event UEV is
generated. Then all the preloaded registers (TIMx_ARR, TIMx_CCRx) are updated.
In the following example, the upcounter is cleared in response to a rising edge on TI1 input:
• Configure the channel 1 to detect rising edges on TI1. Configure the input filter duration
(in this example, we don’t need any filter, so we keep IC1F=0000). The capture
prescaler is not used for triggering, so there’s no need to configure it. The CC1S bits
select the input capture source only, CC1S = 01 in the TIMx_CCMR1 register. Write
CC1P=0 and CC1NP=’0’ in TIMx_CCER register to validate the polarity (and detect
rising edges only).
• Configure the timer in reset mode by writing SMS=100 in TIMx_SMCR register. Select
TI1 as the input source by writing TS=101 in TIMx_SMCR register.
• Start the counter by writing CEN=1 in the TIMx_CR1 register.
The counter starts counting on the internal clock, then behaves normally until TI1 rising
edge. When TI1 rises, the counter is cleared and restarts from 0. In the meantime, the
trigger flag is set (TIF bit in the TIMx_SR register) and an interrupt request, or a DMA
request can be sent if enabled (depending on the TIE and TDE bits in TIMx_DIER register).
The following figure shows this behavior when the auto-reload register TIMx_ARR=0x36.
The delay between the rising edge on TI1 and the actual reset of the counter is due to the
Figure 130. Control circuit in reset mode
TI1
UG
Counter register 30 31 32 33 34 35 36 00 01 02 03 00 01 02 03
TIF
MS31401V3
RM0090 Rev 19 557/1751

588
Slave mode: Gated mode

The counter can be enabled depending on the level of a selected input.
In the following example, the upcounter counts only when TI1 input is low:
• Configure the channel 1 to detect low levels on TI1. Configure the input filter duration
prescaler is not used for triggering, so the user does not need to configure it. The
CC1S bits select the input capture source only, CC1S=01 in TIMx_CCMR1 register.
Write CC1P=1 and CC1NP=’0’ in TIMx_CCER register to validate the polarity (and
detect low level only).
• Configure the timer in gated mode by writing SMS=101 in TIMx_SMCR register. Select
• Enable the counter by writing CEN=1 in the TIMx_CR1 register (in gated mode, the
counter doesn’t start if CEN=0, whatever is the trigger input level).
The counter starts counting on the internal clock as long as TI1 is low and stops as soon as
TI1 becomes high. The TIF flag in the TIMx_SR register is set both when the counter starts
or stops.
The delay between the rising edge on TI1 and the actual stop of the counter is due to the
Figure 131. Control circuit in gated mode
TI1
CNT_EN
Counter clock = ck_cnt = ck_psc
Counter register 30 31 32 33 34 35 36 37 38
TIF
Write TIF=0
MS31402V3
558/1751 RM0090 Rev 19

Slave mode: Trigger mode

The counter can start in response to an event on a selected input.
In the following example, the upcounter starts in response to a rising edge on TI2 input:
are configured to select the input capture source only, CC2S=01 in TIMx_CCMR1
register. Write CC2P=1 and CC2NP=0 in TIMx_CCER register to validate the polarity
(and detect low level only).
• Configure the timer in trigger mode by writing SMS=110 in TIMx_SMCR register. Select
When a rising edge occurs on TI2, the counter starts counting on the internal clock and the
TIF flag is set.
The delay between the rising edge on TI2 and the actual start of the counter is due to the
Figure 132. Control circuit in trigger mode
TI2
CNT_EN
Counter register 34 35 36 37 38
TIF
MS31403V2
Slave mode: external clock mode 2 + trigger mode

The external clock mode 2 can be used in addition to another slave mode (except external
clock mode 1 and encoder mode). In this case, the ETR signal is used as external clock
input, and another input can be selected as trigger input (in reset mode, gated mode or
trigger mode). It is recommended not to select ETR as TRGI through the TS bits of
TIMx_SMCR register.
In the following example, the upcounter is incremented at each rising edge of the ETR
signal as soon as a rising edge of TI1 occurs:
1. Configure the external trigger input circuit by programming the TIMx_SMCR register as
follows:
– ETF = 0000: no filter
– ETPS = 00: prescaler disabled
– ETP = 0: detection of rising edges on ETR and ECE=1 to enable the external clock
mode 2.
2. Configure the channel 1 as follows, to detect rising edges on TI:
– IC1F=0000: no filter.
– The capture prescaler is not used for triggering and does not need to be
configured.
– CC1S=01 in TIMx_CCMR1 register to select only the input capture source
RM0090 Rev 19 559/1751

588
– CC1P=0 and CC1NP=’0’ in TIMx_CCER register to validate the polarity (and

detect rising edge only).
3. Configure the timer in trigger mode by writing SMS=110 in TIMx_SMCR register. Select
A rising edge on TI1 enables the counter and sets the TIF flag. The counter then counts on
ETR rising edges.
The delay between the rising edge of the ETR signal and the actual reset of the counter is
due to the resynchronization circuit on ETRP input.
Figure 133. Control circuit in external clock mode 2 + trigger mode
TI1
CEN/CNT_EN
ETR
TIF
MS33110V1
17.3.20 Timer synchronization

The TIM timers are linked together internally for timer synchronization or chaining. Refer to
Section 18.3.15: Timer synchronization for details.
Note: The clock of the slave timer must be enabled prior to receive events from the master timer,
and must not be changed on-the-fly while triggers are received from the master timer.
17.3.21 Debug mode

When the microcontroller enters debug mode (Cortex®-M4 with FPU core halted), the TIMx
counter either continues to work normally or stops, depending on DBG_TIMx_STOP
configuration bit in DBG module. For more details, refer to Section 38.16.2: Debug support
for timers, watchdog, bxCAN and I2C.
For safety purposes, when the counter is stopped (DBG_TIMx_STOP = 1 in
DBGMCU_APBx_FZ register), the outputs are disabled (as if the MOE bit was reset). The
outputs can either be forced to an inactive state (OSSI bit = 1), or have their control taken
over by the GPIO controller (OSSI bit = 0) to force them to Hi-Z.
560/1751 RM0090 Rev 19

17.4 TIM1 and TIM8 registers

Refer to Section 2.2 for a list of abbreviations used in register descriptions.
The peripheral registers must be written by half-words (16 bits) or words (32 bits). Read
accesses can be done by bytes (8 bits), half-word (16 bits) or words (32 bits).
17.4.1 TIM1 and TIM8 control register 1 (TIMx_CR1)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CKD[1:0] ARPE CMS[1:0] DIR OPM URS UDIS CEN
Reserved

Bits 9:8 CKD[1:0]: Clock division
This bit-field indicates the division ratio between the timer clock (CK_INT) frequency and the
dead-time and sampling clock (tDTS)used by the dead-time generators and the digital filters
(ETR, TIx),
00: tDTS=tCK_INT
01: tDTS=2*tCK_INT
10: tDTS=4*tCK_INT
11: Reserved, do not program this value
Bit 7 ARPE: Auto-reload preload enable
0: TIMx_ARR register is not buffered
1: TIMx_ARR register is buffered
Bits 6:5 CMS[1:0]: Center-aligned mode selection
00: Edge-aligned mode. The counter counts up or down depending on the direction bit
(DIR).
01: Center-aligned mode 1. The counter counts up and down alternatively. Output compare
interrupt flags of channels configured in output (CCxS=00 in TIMx_CCMRx register) are set
only when the counter is counting down.
only when the counter is counting up.
both when the counter is counting up or down.
Note: It is not allowed to switch from edge-aligned mode to center-aligned mode as long as
the counter is enabled (CEN=1)
Bit 4 DIR: Direction
0: Counter used as upcounter
1: Counter used as downcounter
Note: This bit is read only when the timer is configured in Center-aligned mode or Encoder
mode.
Bit 3 OPM: One pulse mode
0: Counter is not stopped at update event
1: Counter stops counting at the next update event (clearing the bit CEN)
RM0090 Rev 19 561/1751

588
Bit 2 URS: Update request source

This bit is set and cleared by software to select the UEV event sources.
0: Any of the following events generate an update interrupt or DMA request if enabled.
These events can be:
– Counter overflow/underflow
– Setting the UG bit
– Update generation through the slave mode controller
1: Only counter overflow/underflow generates an update interrupt or DMA request if
enabled.
Bit 1 UDIS: Update disable
This bit is set and cleared by software to enable/disable UEV event generation.
0: UEV enabled. The Update (UEV) event is generated by one of the following events:
Buffered registers are then loaded with their preload values.
1: UEV disabled. The Update event is not generated, shadow registers keep their value
(ARR, PSC, CCRx). However the counter and the prescaler are reinitialized if the UG bit is
set or if a hardware reset is received from the slave mode controller.
Bit 0 CEN: Counter enable
0: Counter disabled
1: Counter enabled
Note: External clock, gated mode and encoder mode can work only if the CEN bit has been
previously set by software. However trigger mode can set the CEN bit automatically by
hardware.

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OIS4 OIS3N OIS3 OIS2N OIS2 OIS1N OIS1 TI1S MMS[2:0] CCDS CCUS CCPC
Res. Res.

Bit 14 OIS4: Output Idle state 4 (OC4 output)
refer to OIS1 bit
Bit 13 OIS3N: Output Idle state 3 (OC3N output)
refer to OIS1N bit
refer to OIS1 bit
refer to OIS1N bit
refer to OIS1 bit
562/1751 RM0090 Rev 19


0: OC1N=0 after a dead-time when MOE=0
1: OC1N=1 after a dead-time when MOE=0
Note: This bit can not be modified as long as LOCK level 1, 2 or 3 has been programmed
(LOCK bits in TIMx_BDTR register).
0: OC1=0 (after a dead-time if OC1N is implemented) when MOE=0
1: OC1=1 (after a dead-time if OC1N is implemented) when MOE=0
Note: This bit can not be modified as long as LOCK level 1, 2 or 3 has been programmed
Bit 7 TI1S: TI1 selection
0: The TIMx_CH1 pin is connected to TI1 input
1: The TIMx_CH1, CH2 and CH3 pins are connected to the TI1 input (XOR combination)
Bits 6:4 MMS[2:0]: Master mode selection
These bits allow to select the information to be sent in master mode to slave timers for
synchronization (TRGO). The combination is as follows:
000: Reset - the UG bit from the TIMx_EGR register is used as trigger output (TRGO). If the
reset is generated by the trigger input (slave mode controller configured in reset mode) then
the signal on TRGO is delayed compared to the actual reset.
001: Enable - the Counter Enable signal CNT_EN is used as trigger output (TRGO). It is
useful to start several timers at the same time or to control a window in which a slave timer is
enable. The Counter Enable signal is generated by a logic OR between CEN control bit and
the trigger input when configured in gated mode. When the Counter Enable signal is
controlled by the trigger input, there is a delay on TRGO, except if the master/slave mode is
selected (see the MSM bit description in TIMx_SMCR register).
010: Update - The update event is selected as trigger output (TRGO). For instance a master
timer can then be used as a prescaler for a slave timer.
011: Compare Pulse - The trigger output send a positive pulse when the CC1IF flag is to be
set (even if it was already high), as soon as a capture or a compare match occurred.
(TRGO).
100: Compare - OC1REF signal is used as trigger output (TRGO)
Note: The clock of the slave timer and ADC must be enabled prior to receiving events from
the master timer, and must not be changed on-the-fly while triggers are received from
the master timer.
Bit 3 CCDS: Capture/compare DMA selection
0: CCx DMA request sent when CCx event occurs
1: CCx DMA requests sent when update event occurs
RM0090 Rev 19 563/1751

588
Bit 2 CCUS: Capture/compare control update selection

0: When capture/compare control bits are preloaded (CCPC=1), they are updated by setting
the COMG bit only
1: When capture/compare control bits are preloaded (CCPC=1), they are updated by setting
the COMG bit or when an rising edge occurs on TRGI
Note: This bit acts only on channels that have a complementary output.
Bit 0 CCPC: Capture/compare preloaded control
0: CCxE, CCxNE and OCxM bits are not preloaded
1: CCxE, CCxNE and OCxM bits are preloaded, after having been written, they are updated
only when a commutation event (COM) occurs (COMG bit set or rising edge detected on
TRGI, depending on the CCUS bit).
Note: This bit acts only on channels that have a complementary output.
564/1751 RM0090 Rev 19

17.4.3 TIM1 and TIM8 slave mode control register (TIMx_SMCR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ETP ECE ETPS[1:0] ETF[3:0] MSM TS[2:0] Res. SMS[2:0]
rw rw rw rw rw rw rw rw rw rw rw rw Res. rw rw rw
Bit 15 ETP: External trigger polarity

This bit selects whether ETR or ETR is used for trigger operations
0: ETR is non-inverted, active at high level or rising edge.
1: ETR is inverted, active at low level or falling edge.
Bit 14 ECE: External clock enable
This bit enables External clock mode 2.
0: External clock mode 2 disabled
1: External clock mode 2 enabled. The counter is clocked by any active edge on the ETRF
signal.
Note: 1: Setting the ECE bit has the same effect as selecting external clock mode 1 with
TRGI connected to ETRF (SMS=111 and TS=111).
2: It is possible to simultaneously use external clock mode 2 with the following slave
modes: reset mode, gated mode and trigger mode. Nevertheless, TRGI must not be
connected to ETRF in this case (TS bits must not be 111).
3: If external clock mode 1 and external clock mode 2 are enabled at the same time,
the external clock input is ETRF.
Bits 13:12 ETPS[1:0]: External trigger prescaler
External trigger signal ETRP frequency must be at most 1/4 of TIMxCLK frequency. A
prescaler can be enabled to reduce ETRP frequency. It is useful when inputting fast external
clocks.
00: Prescaler OFF
01: ETRP frequency divided by 2
RM0090 Rev 19 565/1751

588
Bits 11:8 ETF[3:0]: External trigger filter

This bit-field then defines the frequency used to sample ETRP signal and the length of the
digital filter applied to ETRP. The digital filter is made of an event counter in which N
consecutive events are needed to validate a transition on the output:
0000: No filter, sampling is done at fDTS
0001: fSAMPLING=fCK_INT, N=2
0100: fSAMPLING=fDTS/2, N=6
Bit 7 MSM: Master/slave mode
0: No action
1: The effect of an event on the trigger input (TRGI) is delayed to allow a perfect
synchronization between the current timer and its slaves (through TRGO). It is useful if we
want to synchronize several timers on a single external event.
Bits 6:4 TS[2:0]: Trigger selection
This bit-field selects the trigger input to be used to synchronize the counter.
000: Internal Trigger 0 (ITR0)
100: TI1 Edge Detector (TI1F_ED)
101: Filtered Timer Input 1 (TI1FP1)
111: External Trigger input (ETRF)
See Table 94 for more details on ITRx meaning for each Timer.
Note: These bits must be changed only when they are not used (e.g. when SMS=000) to
avoid wrong edge detections at the transition.
566/1751 RM0090 Rev 19

Bits 2:0 SMS: Slave mode selection

When external signals are selected the active edge of the trigger signal (TRGI) is linked to
the polarity selected on the external input (see Input Control register and Control register
description.
000: Slave mode disabled - if CEN = ‘1’ then the prescaler is clocked directly by the internal
clock.
001: Encoder mode 1 - Counter counts up/down on TI2FP1 edge depending on TI1FP2
level.
level.
011: Encoder mode 3 - Counter counts up/down on both TI1FP1 and TI2FP2 edges
depending on the level of the other input.
100: Reset Mode - Rising edge of the selected trigger input (TRGI) reinitializes the counter
and generates an update of the registers.
101: Gated Mode - The counter clock is enabled when the trigger input (TRGI) is high. The
counter stops (but is not reset) as soon as the trigger becomes low. Both start and stop of
the counter are controlled.
110: Trigger Mode - The counter starts at a rising edge of the trigger TRGI (but it is not
reset). Only the start of the counter is controlled.
111: External Clock Mode 1 - Rising edges of the selected trigger (TRGI) clock the counter.
Note: The gated mode must not be used if TI1F_ED is selected as the trigger input
(TS=’100’). Indeed, TI1F_ED outputs 1 pulse for each transition on TI1F, whereas the
gated mode checks the level of the trigger signal.
The clock of the slave timer must be enabled prior to receiving events from the master
timer, and must not be changed on-the-fly while triggers are received from the master
timer.
Table 94. TIMx Internal trigger connection

Slave TIM ITR0 (TS = 000) ITR1 (TS = 001) ITR2 (TS = 010) ITR3 (TS = 011)
TIM1 TIM5_TRGO TIM2_TRGO TIM3_TRGO TIM4_TRGO

17.4.4 TIM1 and TIM8 DMA/interrupt enable register (TIMx_DIER)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TDE COMDE CC4DE CC3DE CC2DE CC1DE UDE BIE TIE COMIE CC4IE CC3IE CC2IE CC1IE UIE
Res.
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 14 TDE: Trigger DMA request enable
0: Trigger DMA request disabled
1: Trigger DMA request enabled
Bit 13 COMDE: COM DMA request enable
0: COM DMA request disabled
1: COM DMA request enabled
RM0090 Rev 19 567/1751

588
Bit 12 CC4DE: Capture/Compare 4 DMA request enable

0: CC4 DMA request disabled
1: CC4 DMA request enabled
Bit 8 UDE: Update DMA request enable
0: Update DMA request disabled
1: Update DMA request enabled
Bit 7 BIE: Break interrupt enable
0: Break interrupt disabled
1: Break interrupt enabled
Bit 6 TIE: Trigger interrupt enable
0: Trigger interrupt disabled
1: Trigger interrupt enabled
Bit 5 COMIE: COM interrupt enable
0: COM interrupt disabled
1: COM interrupt enabled
Bit 4 CC4IE: Capture/Compare 4 interrupt enable
0: CC4 interrupt disabled
1: CC4 interrupt enabled
Bit 0 UIE: Update interrupt enable
0: Update interrupt disabled
1: Update interrupt enabled
568/1751 RM0090 Rev 19

17.4.5 TIM1 and TIM8 status register (TIMx_SR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CC4OF CC3OF CC2OF CC1OF Res. BIF TIF COMIF CC4IF CC3IF CC2IF CC1IF UIF
Reserved
rc_w0 rc_w0 rc_w0 rc_w0 Res. rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0

Bit 12 CC4OF: Capture/Compare 4 overcapture flag
refer to CC1OF description
This flag is set by hardware only when the corresponding channel is configured in input
capture mode. It is cleared by software by writing it to ‘0’.
0: No overcapture has been detected.
1: The counter value has been captured in TIMx_CCR1 register while CC1IF flag was
already set
Bit 7 BIF: Break interrupt flag
This flag is set by hardware as soon as the break input goes active. It can be cleared by
software if the break input is not active.
0: No break event occurred.
1: An active level has been detected on the break input.
Bit 6 TIF: Trigger interrupt flag
This flag is set by hardware on trigger event (active edge detected on TRGI input when the
slave mode controller is enabled in all modes but gated mode. It is set when the counter
starts or stops when gated mode is selected. It is cleared by software.
0: No trigger event occurred.
1: Trigger interrupt pending.
Bit 5 COMIF: COM interrupt flag
This flag is set by hardware on COM event (when Capture/compare Control bits - CCxE,
CCxNE, OCxM - have been updated). It is cleared by software.
0: No COM event occurred.
1: COM interrupt pending.
Bit 4 CC4IF: Capture/Compare 4 interrupt flag
refer to CC1IF description
RM0090 Rev 19 569/1751

588

If channel CC1 is configured as output:
This flag is set by hardware when the counter matches the compare value, with some
exception in center-aligned mode (refer to the CMS bits in the TIMx_CR1 register
description). It is cleared by software.
0: No match.
1: The content of the counter TIMx_CNT matches the content of the TIMx_CCR1 register.
When the contents of TIMx_CCR1 are greater than the contents of TIMx_ARR, the CC1IF
bit goes high on the counter overflow (in upcounting and up/down-counting modes) or
underflow (in downcounting mode)
If channel CC1 is configured as input:
This bit is set by hardware on a capture. It is cleared by software or by reading the
TIMx_CCR1 register.
0: No input capture occurred
1: The counter value has been captured in TIMx_CCR1 register (An edge has been
detected on IC1 which matches the selected polarity)
Bit 0 UIF: Update interrupt flag
This bit is set by hardware on an update event. It is cleared by software.
0: No update occurred.
1: Update interrupt pending. This bit is set by hardware when the registers are updated:
– At overflow or underflow regarding the repetition counter value (update if repetition counter
= 0) and if the UDIS=0 in the TIMx_CR1 register.
– When CNT is reinitialized by software using the UG bit in TIMx_EGR register, if URS=0 and
UDIS=0 in the TIMx_CR1 register.
– When CNT is reinitialized by a trigger event (refer to Section 17.4.3: TIM1 and TIM8 slave
mode control register (TIMx_SMCR)), if URS=0 and UDIS=0 in the TIMx_CR1 register.
17.4.6 TIM1 and TIM8 event generation register (TIMx_EGR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
BG TG COMG CC4G CC3G CC2G CC1G UG
Reserved
w w w w w w w w

Bit 7 BG: Break generation
This bit is set by software in order to generate an event, it is automatically cleared by
hardware.
0: No action
1: A break event is generated. MOE bit is cleared and BIF flag is set. Related interrupt or
DMA transfer can occur if enabled.
Bit 6 TG: Trigger generation
hardware.
0: No action
1: The TIF flag is set in TIMx_SR register. Related interrupt or DMA transfer can occur if
enabled.
570/1751 RM0090 Rev 19

Bit 5 COMG: Capture/Compare control update generation

This bit can be set by software, it is automatically cleared by hardware
0: No action
1: When CCPC bit is set, it allows to update CCxE, CCxNE and OCxM bits
Note: This bit acts only on channels having a complementary output.
Bit 4 CC4G: Capture/Compare 4 generation
refer to CC1G description
hardware.
0: No action
1: A capture/compare event is generated on channel 1:
CC1IF flag is set, Corresponding interrupt or DMA request is sent if enabled.
The current value of the counter is captured in TIMx_CCR1 register. The CC1IF flag is set,
the corresponding interrupt or DMA request is sent if enabled. The CC1OF flag is set if the
CC1IF flag was already high.
Bit 0 UG: Update generation
This bit can be set by software, it is automatically cleared by hardware.
0: No action
1: Reinitialize the counter and generates an update of the registers. Note that the prescaler
counter is cleared too (anyway the prescaler ratio is not affected). The counter is cleared if
the center-aligned mode is selected or if DIR=0 (upcounting), else it takes the auto-reload
value (TIMx_ARR) if DIR=1 (downcounting).
RM0090 Rev 19 571/1751

588
17.4.7 TIM1 and TIM8 capture/compare mode register 1 (TIMx_CCMR1)

Reset value: 0x0000
The channels can be used in input (capture mode) or in output (compare mode). The
direction of a channel is defined by configuring the corresponding CCxS bits. All the other
bits of this register have a different function in input and in output mode. For a given bit,
OCxx describes its function when the channel is configured in output, ICxx describes its
function when the channel is configured in input. So the user must take care that the same
bit can have a different meaning for the input stage and for the output stage.
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OC2 OC2 OC2 OC1 OC1 OC1
OC2M[2:0] OC1M[2:0]
CE PE FE CC2S[1:0] CE PE FE CC1S[1:0]
IC2F[3:0] IC2PSC[1:0] IC1F[3:0] IC1PSC[1:0]
Output compare mode:
Bit 15 OC2CE: Output compare 2 clear enable

Bits 14:12 OC2M[2:0]: Output compare 2 mode
Bit 11 OC2PE: Output compare 2 preload enable
Bit 10 OC2FE: Output compare 2 fast enable
Bits 9:8 CC2S[1:0]: Capture/Compare 2 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC2 channel is configured as output
01: CC2 channel is configured as input, IC2 is mapped on TI2
11: CC2 channel is configured as input, IC2 is mapped on TRC. This mode is working only if
an internal trigger input is selected through the TS bit (TIMx_SMCR register)
Note: CC2S bits are writable only when the channel is OFF (CC2E = ‘0’ in TIMx_CCER).
OC1CE: Output compare 1 Clear Enable
0: OC1Ref is not affected by the ETRF Input
1: OC1Ref is cleared as soon as a High level is detected on ETRF input
572/1751 RM0090 Rev 19

Bits 6:4 OC1M: Output compare 1 mode

These bits define the behavior of the output reference signal OC1REF from which OC1 and
OC1N are derived. OC1REF is active high whereas OC1 and OC1N active level depends
on CC1P and CC1NP bits.
000: Frozen - The comparison between the output compare register TIMx_CCR1 and the
counter TIMx_CNT has no effect on the outputs.(this mode is used to generate a timing
base).
001: Set channel 1 to active level on match. OC1REF signal is forced high when the counter
TIMx_CNT matches the capture/compare register 1 (TIMx_CCR1).
010: Set channel 1 to inactive level on match. OC1REF signal is forced low when the
counter TIMx_CNT matches the capture/compare register 1 (TIMx_CCR1).
011: Toggle - OC1REF toggles when TIMx_CNT=TIMx_CCR1.
100: Force inactive level - OC1REF is forced low.
101: Force active level - OC1REF is forced high.
110: PWM mode 1 - In upcounting, channel 1 is active as long as TIMx_CNT<TIMx_CCR1
else inactive. In downcounting, channel 1 is inactive (OC1REF=‘0’) as long as
TIMx_CNT>TIMx_CCR1 else active (OC1REF=’1’).
111: PWM mode 2 - In upcounting, channel 1 is inactive as long as TIMx_CNT<TIMx_CCR1
else active. In downcounting, channel 1 is active as long as TIMx_CNT>TIMx_CCR1 else
inactive.
Note: 1: These bits can not be modified as long as LOCK level 3 has been programmed
(LOCK bits in TIMx_BDTR register) and CC1S=’00’ (the channel is configured in
output).
2: In PWM mode 1 or 2, the OCREF level changes only when the result of the
comparison changes or when the output compare mode switches from “frozen” mode
to “PWM” mode.
3: On channels having a complementary output, this bit field is preloaded. If the CCPC
bit is set in the TIMx_CR2 register then the OC1M active bits take the new value from
the preloaded bits only when a COM event is generated.
0: Preload register on TIMx_CCR1 disabled. TIMx_CCR1 can be written at anytime, the
new value is taken in account immediately.
1: Preload register on TIMx_CCR1 enabled. Read/Write operations access the preload
register. TIMx_CCR1 preload value is loaded in the active register at each update event.
(LOCK bits in TIMx_BDTR register) and CC1S=’00’ (the channel is configured in
output).
2: The PWM mode can be used without validating the preload register only in one
pulse mode (OPM bit set in TIMx_CR1 register). Else the behavior is not guaranteed.
This bit is used to accelerate the effect of an event on the trigger in input on the CC output.
0: CC1 behaves normally depending on counter and CCR1 values even when the trigger is
ON. The minimum delay to activate CC1 output when an edge occurs on the trigger input is
5 clock cycles.
1: An active edge on the trigger input acts like a compare match on CC1 output. Then, OC is
set to the compare level independently from the result of the comparison. Delay to sample
the trigger input and to activate CC1 output is reduced to 3 clock cycles. OCFE acts only if
the channel is configured in PWM1 or PWM2 mode.
RM0090 Rev 19 573/1751

588
Bits 1:0 CC1S: Capture/Compare 1 selection

an internal trigger input is selected through TS bit (TIMx_SMCR register)
Input capture mode
Bits 15:12 IC2F: Input capture 2 filter

Bits 11:10 IC2PSC[1:0]: Input capture 2 prescaler
11: CC2 channel is configured as input, IC2 is mapped on TRC. This mode is working only if an
internal trigger input is selected through TS bit (TIMx_SMCR register)
Bits 7:4 IC1F[3:0]: Input capture 1 filter
This bit-field defines the frequency used to sample TI1 input and the length of the digital filter applied
to TI1. The digital filter is made of an event counter in which N consecutive events are needed to
validate a transition on the output:
Bits 3:2 IC1PSC: Input capture 1 prescaler
This bit-field defines the ratio of the prescaler acting on CC1 input (IC1).
The prescaler is reset as soon as CC1E=’0’ (TIMx_CCER register).
00: no prescaler, capture is done each time an edge is detected on the capture input
01: capture is done once every 2 events
574/1751 RM0090 Rev 19

Bits 1:0 CC1S: Capture/Compare 1 Selection

11: CC1 channel is configured as input, IC1 is mapped on TRC. This mode is working only if an
internal trigger input is selected through TS bit (TIMx_SMCR register)
17.4.8 TIM1 and TIM8 capture/compare mode register 2 (TIMx_CCMR2)

Reset value: 0x0000
Refer to the above CCMR1 register description.
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OC4 OC4 OC4 OC3 OC3 OC3
OC4M[2:0] OC3M[2:0]
CE PE FE CC4S[1:0] CE. PE FE CC3S[1:0]
Output compare mode

RM0090 Rev 19 575/1751

588
Input capture mode

Bits 1:0 CC3S: Capture/compare 3 selection
17.4.9 TIM1 and TIM8 capture/compare enable register (TIMx_CCER)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CC4NP CC4P CC4E CC3NP CC3NE CC3P CC3E CC2NP CC2NE CC2P CC2E CC1NP CC1NE CC1P CC1E
Res.
Bit 15 CC4NP: Capture/Compare 4 complementary output polarity

refer to CC1NP description
Bit 13 CC4P: Capture/Compare 4 output polarity
refer to CC1P description
Bit 12 CC4E: Capture/Compare 4 output enable
refer to CC1E description
Bit 10 CC3NE: Capture/Compare 3 complementary output enable
refer to CC1NE description
576/1751 RM0090 Rev 19


refer to CC1NE description
CC1 channel configured as output:
0: OC1N active high.
1: OC1N active low.
CC1 channel configured as input:
This bit is used in conjunction with CC1P to define the polarity of TI1FP1 and TI2FP1. Refer
to CC1P description.
Note: On channels having a complementary output, this bit is preloaded. If the CCPC bit is
set in the TIMx_CR2 register then the CC1NP active bit takes the new value from the
preloaded bit only when a Commutation event is generated.
Note: This bit is not writable as soon as LOCK level 2 or 3 has been programmed (LOCK bits
in TIMx_BDTR register) and CC1S=”00” (the channel is configured in output).
0: Off - OC1N is not active. OC1N level is then function of MOE, OSSI, OSSR, OIS1, OIS1N
and CC1E bits.
1: On - OC1N signal is output on the corresponding output pin depending on MOE, OSSI,
OSSR, OIS1, OIS1N and CC1E bits.
set in the TIMx_CR2 register then the CC1NE active bit takes the new value from the
RM0090 Rev 19 577/1751

588

0: OC1 active high
1: OC1 active low
CC1NP/CC1P bits select the active polarity of TI1FP1 and TI2FP1 for trigger or capture
operations.
00: non-inverted/rising edge
The circuit is sensitive to TIxFP1 rising edge (capture or trigger operations in reset, external
clock or trigger mode), TIxFP1 is not inverted (trigger operation in gated mode or encoder
mode).
01: inverted/falling edge
The circuit is sensitive to TIxFP1 falling edge (capture or trigger operations in reset, external
clock or trigger mode), TIxFP1 is inverted (trigger operation in gated mode or encoder
mode).
10: reserved, do not use this configuration.
11: non-inverted/both edges
The circuit is sensitive to both TIxFP1 rising and falling edges (capture or trigger operations
in reset, external clock or trigger mode), TIxFP1 is not inverted (trigger operation in gated
mode). This configuration must not be used in encoder mode.
set in the TIMx_CR2 register then the CC1P active bit takes the new value from the
Note: This bit is not writable as soon as LOCK level 2 or 3 has been programmed (LOCK bits
in TIMx_BDTR register).
0: Off - OC1 is not active. OC1 level is then function of MOE, OSSI, OSSR, OIS1, OIS1N
and CC1NE bits.
1: On - OC1 signal is output on the corresponding output pin depending on MOE, OSSI,
OSSR, OIS1, OIS1N and CC1NE bits.
This bit determines if a capture of the counter value can actually be done into the input
capture/compare register 1 (TIMx_CCR1) or not.
1: Capture enabled.
set in the TIMx_CR2 register then the CC1E active bit takes the new value from the
578/1751 RM0090 Rev 19

Table 95. Output control bits for complementary OCx and OCxN channels with
break feature
Control bits Output states(1)
MOE OSSI OSSR CCxE CCxNE

OCx output state OCxN output state
bit bit bit bit bit
Output Disabled (not driven by Output Disabled (not driven by the

0 0 0
the timer), OCx=0, OCx_EN=0 timer), OCxN=0, OCxN_EN=0
Output Disabled (not driven by OCxREF + Polarity OCxN=OCxREF
0 0 1
the timer), OCx=0, OCx_EN=0 xor CCxNP, OCxN_EN=1
OCxREF + Polarity Output Disabled (not driven by the
0 1 0 OCx=OCxREF xor CCxP, timer)
OCx_EN=1 OCxN=0, OCxN_EN=0
Complementary to OCREF (not
OCREF + Polarity + dead-time
0 1 1 OCREF) + Polarity + dead-time
OCx_EN=1
OCxN_EN=1
1 X Output Disabled (not driven by Output Disabled (not driven by the

1 0 0 the timer) timer)
OCx=CCxP, OCx_EN=0 OCxN=CCxNP, OCxN_EN=0
Off-State (output enabled with OCxREF + Polarity
1 0 1 inactive state) OCxN=OCxREF xor CCxNP,
OCx=CCxP, OCx_EN=1 OCxN_EN=1
OCxREF + Polarity Off-State (output enabled with
1 1 0 OCx=OCxREF xor CCxP, inactive state)
OCx_EN=1 OCxN=CCxNP, OCxN_EN=1
Complementary to OCREF (not
OCREF + Polarity + dead-time
1 1 1 OCREF) + Polarity + dead-time
OCx_EN=1
OCxN_EN=1
0 X X Output disabled (not driven by the timer anymore). The output state is
0 0 defined by the GPIO controller and can be High, Low or Hi-Z.
0 1 Off-State (output enabled with inactive state)

0 X Asynchronously: OCx=CCxP, OCx_EN=1, OCxN=CCxNP,
1 1 0
OCxN_EN=1
Then if the clock is present: OCx=OISx and OCxN=OISxN after a
1 1 dead-time, assuming that OISx and OISxN do not correspond to OCX
and OCxN both in active state
1. When both outputs of a channel are not used (CCxE = CCxNE = 0), the OISx, OISxN, CCxP and CCxNP bits must be kept
cleared.
Note: The state of the external I/O pins connected to the complementary OCx and OCxN channels
depends on the OCx and OCxN channel state and the GPIO registers.
RM0090 Rev 19 579/1751

588
17.4.10 TIM1 and TIM8 counter (TIMx_CNT)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CNT[15:0]
Bits 15:0 CNT[15:0]: Counter value
17.4.11 TIM1 and TIM8 prescaler (TIMx_PSC)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PSC[15:0]
Bits 15:0 PSC[15:0]: Prescaler value

The counter clock frequency (CK_CNT) is equal to fCK_PSC / (PSC[15:0] + 1).
PSC contains the value to be loaded in the active prescaler register at each update event
(including when the counter is cleared through UG bit of TIMx_EGR register or through
trigger controller when configured in “reset mode”).
17.4.12 TIM1 and TIM8 auto-reload register (TIMx_ARR)

Reset value: 0xFFFF
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ARR[15:0]
Bits 15:0 ARR[15:0]: Auto-reload value

ARR is the value to be loaded in the actual auto-reload register.
Refer to Section 17.3.1: Time-base unit for more details about ARR update and behavior.
The counter is blocked while the auto-reload value is null.
580/1751 RM0090 Rev 19

17.4.13 TIM1 and TIM8 repetition counter register (TIMx_RCR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
REP[7:0]
Reserved

Bits 7:0 REP[7:0]: Repetition counter value
These bits allow the user to set-up the update rate of the compare registers (i.e. periodic
transfers from preload to active registers) when preload registers are enable, as well as the
update interrupt generation rate, if this interrupt is enable.
Each time the REP_CNT related downcounter reaches zero, an update event is generated
and it restarts counting from REP value. As REP_CNT is reloaded with REP value only at
the repetition update event U_RC, any write to the TIMx_RCR register is not taken in
account until the next repetition update event.
It means in PWM mode (REP+1) corresponds to:
– the number of PWM periods in edge-aligned mode
– the number of half PWM period in center-aligned mode.
17.4.14 TIM1 and TIM8 capture/compare register 1 (TIMx_CCR1)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR1[15:0]
rw/ro rw/ro rw/ro rw/ro rw/ro rw/ro rw/ro rw/ro rw/ro rw/ro rw/ro rw/ro rw/ro rw/ro rw/ro rw/ro
Bits 15:0 CCR1[15:0]: Capture/Compare 1 value

CCR1 is the value to be loaded in the actual capture/compare 1 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR1 register
(bit OC1PE). Else the preload value is copied in the active capture/compare 1 register when
an update event occurs.
The active capture/compare register contains the value to be compared to the counter
TIMx_CNT and signaled on OC1 output.
CCR1 is the counter value transferred by the last input capture 1 event (IC1). The
TIMx_CCR1 register is read-only and cannot be programmed.
RM0090 Rev 19 581/1751

588

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR2[15:0]

TIMx_CNT and signalled on OC2 output.

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR3[15:0]
Bits 15:0 CCR3[15:0]: Capture/Compare value

582/1751 RM0090 Rev 19


Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR4[15:0]
Bits 15:0 CCR4[15:0]: Capture/Compare value

17.4.18 TIM1 and TIM8 break and dead-time register (TIMx_BDTR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MOE AOE BKP BKE OSSR OSSI LOCK[1:0] DTG[7:0]
Note: As the bits AOE, BKP, BKE, OSSI, OSSR and DTG[7:0] can be write-locked depending on
the LOCK configuration, it can be necessary to configure all of them during the first write
access to the TIMx_BDTR register.
Bit 15 MOE: Main output enable

This bit is cleared asynchronously by hardware as soon as the break input is active. It is set
by software or automatically depending on the AOE bit. It is acting only on the channels
which are configured in output.
0: OC and OCN outputs are disabled or forced to idle state.
1: OC and OCN outputs are enabled if their respective enable bits are set (CCxE, CCxNE in
TIMx_CCER register).
See OC/OCN enable description for more details (Section 17.4.9: TIM1 and TIM8
capture/compare enable register (TIMx_CCER)).
Bit 14 AOE: Automatic output enable
0: MOE can be set only by software
1: MOE can be set by software or automatically at the next update event (if the break input is
not be active)
Note: This bit can not be modified as long as LOCK level 1 has been programmed (LOCK bits
RM0090 Rev 19 583/1751

588
Bit 13 BKP: Break polarity

0: Break input BRK is active low
1: Break input BRK is active high
Note: This bit can not be modified as long as LOCK level 1 has been programmed (LOCK bits
Note: Any write operation to this bit takes a delay of 1 APB clock cycle to become effective.
Bit 12 BKE: Break enable
0: Break inputs (BRK and CSS clock failure event) disabled
1; Break inputs (BRK and CSS clock failure event) enabled
Note: This bit cannot be modified when LOCK level 1 has been programmed (LOCK bits in
TIMx_BDTR register).
Note: Any write operation to this bit takes a delay of 1 APB clock cycle to become effective.
Bit 11 OSSR: Off-state selection for Run mode
This bit is used when MOE=1 on channels having a complementary output which are
configured as outputs. OSSR is not implemented if no complementary output is implemented
in the timer.
0: When inactive, OC/OCN outputs are disabled (OC/OCN enable output signal=0).
1: When inactive, OC/OCN outputs are enabled with their inactive level as soon as CCxE=1
or CCxNE=1. Then, OC/OCN enable output signal=1
Note: This bit can not be modified as soon as the LOCK level 2 has been programmed (LOCK
bits in TIMx_BDTR register).
Bit 10 OSSI: Off-state selection for Idle mode
This bit is used when MOE=0 on channels configured as outputs.
0: When inactive, OC/OCN outputs are disabled (OC/OCN enable output signal=0).
1: When inactive, OC/OCN outputs are forced first with their idle level as soon as CCxE=1 or
CCxNE=1. OC/OCN enable output signal=1)
Note: This bit can not be modified as soon as the LOCK level 2 has been programmed (LOCK
bits in TIMx_BDTR register).
Bits 9:8 LOCK[1:0]: Lock configuration
These bits offer a write protection against software errors.
00: LOCK OFF - No bit is write protected.
01: LOCK Level 1 = DTG bits in TIMx_BDTR register, OISx and OISxN bits in TIMx_CR2
register and BKE/BKP/AOE bits in TIMx_BDTR register can no longer be written.
10: LOCK Level 2 = LOCK Level 1 + CC Polarity bits (CCxP/CCxNP bits in TIMx_CCER
register, as long as the related channel is configured in output through the CCxS bits) as well
as OSSR and OSSI bits can no longer be written.
11: LOCK Level 3 = LOCK Level 2 + CC Control bits (OCxM and OCxPE bits in
TIMx_CCMRx registers, as long as the related channel is configured in output through the
CCxS bits) can no longer be written.
Note: The LOCK bits can be written only once after the reset. Once the TIMx_BDTR register
has been written, their content is frozen until the next reset.
584/1751 RM0090 Rev 19

Bits 7:0 DTG[7:0]: Dead-time generator setup

This bit-field defines the duration of the dead-time inserted between the complementary
outputs. DT correspond to this duration.
DTG[7:5]=0xx => DT=DTG[7:0]x tdtg with tdtg=tDTS.
DTG[7:5]=10x => DT=(64+DTG[5:0])xtdtg with Tdtg=2xtDTS.
DTG[7:5]=110 => DT=(32+DTG[4:0])xtdtg with Tdtg=8xtDTS.
DTG[7:5]=111 => DT=(32+DTG[4:0])xtdtg with Tdtg=16xtDTS.
Example if TDTS=125ns (8MHz), dead-time possible values are:
0 to 15875 ns by 125 ns steps,
16 us to 31750 ns by 250 ns steps,
32 us to 63us by 1 us steps,
64 us to 126 us by 2 us steps
Note: This bit-field can not be modified as long as LOCK level 1, 2 or 3 has been programmed
17.4.19 TIM1 and TIM8 DMA control register (TIMx_DCR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DBL[4:0] DBA[4:0]
Reserved Reserved

Bits 12:8 DBL[4:0]: DMA burst length
This 5-bit vector defines the number of DMA transfers (the timer detects a burst transfer
when a read or a write access to the TIMx_DMAR register address is performed).
the TIMx_DMAR address)
00000: 1 transfer
00001: 2 transfers
00010: 3 transfers
...
10001: 18 transfers
Bits 4:0 DBA[4:0]: DMA base address
This 5-bits vector defines the base-address for DMA transfers (when read/write access are
done through the TIMx_DMAR address). DBA is defined as an offset starting from the
address of the TIMx_CR1 register.
Example:
00000: TIMx_CR1,
00001: TIMx_CR2,
00010: TIMx_SMCR,
...
Example: Let us consider the following transfer: DBL = 7 transfers and DBA = TIMx_CR1. In
this case the transfer is done to/from 7 registers starting from the TIMx_CR1 address.
RM0090 Rev 19 585/1751

588
17.4.20 TIM1 and TIM8 DMA address for full transfer (TIMx_DMAR)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DMAB[31:16]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DMAB[15:0]
Bits 31:0 DMAB[31:0]: DMA register for burst accesses

A read or write operation to the DMAR register accesses the register located at the address
(TIMx_CR1 address) + (DBA + DMA index) x 4
where TIMx_CR1 address is the address of the control register 1, DBA is the DMA base
address configured in TIMx_DCR register, DMA index is automatically controlled by the
DMA transfer, and ranges from 0 to DBL (DBL configured in TIMx_DCR).
Example of how to use the DMA burst feature

In this example the timer DMA burst feature is used to update the contents of the CCRx
registers (x = 2, 3, 4) with the DMA transferring half words into the CCRx registers.
This is done in the following steps:
1. Configure the corresponding DMA channel as follows:
– DMA channel peripheral address is the DMAR register address
– DMA channel memory address is the address of the buffer in the RAM containing
the data to be transferred by DMA into CCRx registers.
– Number of data to transfer = 3 (See note below).
– Circular mode disabled.
2. Configure the DCR register by configuring the DBA and DBL bit fields as follows:
DBL = 3 transfers, DBA = 0xE.
3. Enable the TIMx update DMA request (set the UDE bit in the DIER register).
4. Enable TIMx
5. Enable the DMA channel
Note: This example is for the case where every CCRx register to be updated once. If every CCRx
register is to be updated twice for example, the number of data to transfer should be 6. Let's
take the example of a buffer in the RAM containing data1, data2, data3, data4, data5 and
data6. The data is transferred to the CCRx registers as follows: on the first update DMA
request, data1 is transferred to CCR2, data2 is transferred to CCR3, data3 is transferred to
CCR4 and on the second update DMA request, data4 is transferred to CCR2, data5 is
transferred to CCR3 and data6 is transferred to CCR4.
586/1751 RM0090 Rev 19

17.4.21 TIM1 and TIM8 register map

TIM1 and TIM8 registers are mapped as 16-bit addressable registers as described in the
table below:
Table 96. TIM1 and TIM8 register map and reset values
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
CKD CMS
ARPE
UDIS
OPM
CEN
URS
DIR
TIMx_CR1
0x00 Reserved [1:0] [1:0]
Reset value 0 0 0 0 0 0 0 0 0 0
OIS3N
OIS2N
OIS1N
CCPC
CCDS
CCUS
Reserved
OIS4
OIS3
OIS2
OIS1
TI1S
TIMx_CR2 MMS[2:0]
0x04 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ETP
Reserved
MSM
ECE
ETP
TIMx_SMCR S ETF[3:0] TS[2:0] SMS[2:0]
0x08 Reserved [1:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
COMDE
CC4DE
CC3DE
CC2DE
CC1DE
COMIE
CC4IE
CC3IE
CC2IE
CC1IE
UDE
TDE
UIE
BIE
TIMx_DIER
TIE
0x0C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CC4OF
CC3OF
CC2OF
CC1OF
COMIF
CC4IF
CC3IF
CC2IF
CC1IF
Reserved
TIMx_SR
UIF
BIF
TIF
0x10 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0
COMG
CC4G
CC3G
CC2G
CC1G
TIMx_EGR
UG
BG
TG
0x14 Reserved
Reset value 0 0 0 0 0 0 0 0
TIMx_CCMR1 CC1
OC2CE
OC1CE
OC2PE
OC1PE
OC2FE
OC1FE
OC2M CC2S OC1M
Output compare S
Reserved [2:0] [1:0] [2:0]
mode [1:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x18
TIMx_CCMR1 IC2 IC1 CC1
CC2S
Input capture IC2F[3:0] PSC IC1F[3:0] PSC S
Reserved [1:0]
mode [1:0] [1:0] [1:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
TIMx_CCMR2 CC3
OC4CE
OC3CE
OC4PE
OC3PE
OC4FE
OC3FE
OC4M CC4S OC3M

Output compare S
Reserved [2:0] [1:0] [2:0]
mode [1:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x1C
CC4S
Reserved [1:0]
mode [1:0] [1:0] [1:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CC4NP
CC3NP
CC3NE
CC2NP
CC2NE
CC1NP
CC1NE
CC4P
CC4E
CC3P
CC3E
CC2P
CC2E
CC1P
CC1E
Reserved
TIMx_CCER
0x20 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
TIMx_CNT CNT[15:0]
0x24 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
TIMx_PSC PSC[15:0]
0x28 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RM0090 Rev 19 587/1751

588
Table 96. TIM1 and TIM8 register map and reset values (continued)
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
TIMx_ARR ARR[15:0]
0x2C Reserved
Reset value 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
TIMx_RCR REP[7:0]
0x30 Reserved
Reset value 0 0 0 0 0 0 0 0
TIMx_CCR1 CCR1[15:0]
0x34 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x38 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x3C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x40 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
OSSR
LOCK
OSSI
MOE
AOE
BKP
BKE
TIMx_BDTR DT[7:0]
0x44 Reserved [1:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
TIMx_DCR DBL[4:0] DBA[4:0]

Reset value 0 0 0 0 0 0 0 0 0 0
TIMx_DMAR DMAB[31:0]
0x4C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
588/1751 RM0090 Rev 19

RM0090 General-purpose timers (TIM2 to TIM5)
18 General-purpose timers (TIM2 to TIM5)
18.1 TIM2 to TIM5 introduction

The general-purpose timers consist of a 16-bit or 32-bit auto-reload counter driven by a
programmable prescaler.
They may be used for a variety of purposes, including measuring the pulse lengths of input
signals (input capture) or generating output waveforms (output compare and PWM).
The timers are completely independent, and do not share any resources. They can be
synchronized together as described in Section 18.3.15.
18.2 TIM2 to TIM5 main features

General-purpose TIMx timer features include:
• 16-bit (TIM3 and TIM4) or 32-bit (TIM2 and TIM5) up, down, up/down auto-reload
counter.
• 16-bit programmable prescaler used to divide (also “on the fly”) the counter clock
frequency by any factor between 1 and 65536.
– Input capture
– Output compare
– PWM generation (Edge- and Center-aligned modes)
several timers.
• Interrupt/DMA generation on the following events:
– Update: counter overflow/underflow, counter initialization (by software or
internal/external trigger)
– Trigger event (counter start, stop, initialization or count by internal/external trigger)
– Input capture
– Output compare
• Supports incremental (quadrature) encoder and hall-sensor circuitry for positioning
purposes
• Trigger input for external clock or cycle-by-cycle current management
RM0090 Rev 19 589/1751

649
General-purpose timers (TIM2 to TIM5) RM0090
Figure 134. General-purpose timer block diagram

TIMxCLK from RCC Trigger
ETRF controller
ETR Polarity selection & edge ETRP TRGO
TIMx_ETR Input filter
detector & prescaler to other timers
to DAC/ADC
ITR0 TGI
ITR1 ITR
Slave
ITR2 TRC TRGI controller Reset, enable, up, count
ITR3
mode
TI1F_ED
TI1FP1 Encoder
TI2FP2 interface
U
Auto-reload register UI
Stop, clear or up/down
U
CK_PSC PSC CK_CNT +/- CNT counter
prescaler
CC1I U CC1I
XOR TI1FP1 OC1REF
TI1 Input filter & IC1 IC1PS Output OC1 TIMx_CH1
TI1FP2 Prescaler Capture/Compare 1 register
edge detector control
TIMx_CH1 TRC
CC2I
U CC2I
TI2FP1
TI2 Input filter & IC2 Output OC2
TIMx_CH2 TI2FP2 Prescaler
IC2PS Capture/Compare 2 register OC2REF TIMx_CH2
TRC
CC3I CC3I
U
TI3FP3 OC3REF
TI3 Input filter & IC3 IC3PS Output OC3 TIMx_CH3
TI3FP4 Prescaler Capture/Compare 3 register
TIMx_CH3 edge detector control
TRC CC4I
U CC4I
TI4FP3
TRC
ETRF
Notes:
Preload registers transferred
Reg
to active registers on U event
according to control bit
Event
Interrupt & DMA output
MS19673V1
590/1751 RM0090 Rev 19

18.3 TIM2 to TIM5 functional description

The main block of the programmable timer is a 16-bit/32-bit counter with its related auto-
reload register. The counter can count up. The counter clock can be divided by a prescaler.
• Prescaler register (TIMx_PSC):
• Auto-Reload register (TIMx_ARR)
reaches the overflow (or underflow when downcounting) and if the UDIS bit equals 0 in the
TIMx_CR1 register. It can also be generated by software. The generation of the update
event is described in detail for each configuration.
Note that the actual counter enable signal CNT_EN is set 1 clock cycle after CEN.
is based on a 16-bit counter controlled through a 16-bit/32-bit register (in the TIMx_PSC
register). It can be changed on the fly as this control register is buffered. The new prescaler
ratio is taken into account at the next update event.
ratio is changed on the fly:
RM0090 Rev 19 591/1751

649
CK_PSC
CNT_EN
Timerclock = CK_CNT
Update event (UEV)
MS35833V1
CK_PSC
CNT_EN
Timerclock = CK_CNT
Counter register F7 F8 F9 FA FB FC 00 01
Update event (UEV)
MS35834V1

Upcounting mode
An Update event can be generated at each counter overflow or by setting the UG bit in the
TIMx_EGR register (by software or by using the slave mode controller).
The UEV event can be disabled by software by setting the UDIS bit in TIMx_CR1 register.
This is to avoid updating the shadow registers while writing new values in the preload
registers. Then no update event occurs until the UDIS bit has been written to 0. However,
the counter restarts from 0, as well as the counter of the prescaler (but the prescale rate
592/1751 RM0090 Rev 19

does not change). In addition, if the URS bit (update request selection) in TIMx_CR1
register is set, setting the UG bit generates an update event UEV but without setting the UIF
flag (thus no interrupt or DMA request is sent). This is to avoid generating both update and
capture interrupts when clearing the counter on the capture event.
register)
• The auto-reload shadow register is updated with the preload value (TIMx_ARR)
CK_INT
CNT_EN
Timerclock = CK_CNT
Counter overflow
Update event (UEV)

MS35836V1
CK_INT
CNT_EN
Timerclock = CK_CNT
Counter overflow
Update event (UEV)

MS35835V1
RM0090 Rev 19 593/1751

649
CK_INT
CNT_EN
Timerclock = CK_CNT
Counter overflow
Update event (UEV)

MSv37301V1
CK_INT
Timerclock = CK_CNT
Counter overflow
Update event (UEV)

MSv37302V1
594/1751 RM0090 Rev 19

Figure 141. Counter timing diagram, Update event when ARPE=0 (TIMx_ARR not
preloaded)
CK_INT
CNT_EN
Timerclock = CK_CNT
Counter overflow
Update event (UEV)
Auto-reload register FF 36

MSv37303V1
Figure 142. Counter timing diagram, Update event when ARPE=1 (TIMx_ARR
preloaded)
CK_PSC
CNT_EN
Timerclock = CK_CNT
Counter overflow
Update event (UEV)
Auto-reload preload register F5 36
Auto-reload shadow register F5 36

MSv37304V1
Downcounting mode
In downcounting mode, the counter counts from the auto-reload value (content of the
TIMx_ARR register) down to 0, then restarts from the auto-reload value and generates a
counter underflow event.
An Update event can be generate at each counter underflow or by setting the UG bit in the
TIMx_EGR register (by software or by using the slave mode controller)
RM0090 Rev 19 595/1751

649
However, the counter restarts from the current auto-reload value, whereas the counter of the
prescaler restarts from 0 (but the prescale rate doesn’t change).
register).
TIMx_ARR register). Note that the auto-reload is updated before the counter is
reloaded, so that the next period is the expected one.
CK_INT
CNT_EN
Timerclock = CK_CNT
Counter underflow (cnt_udf)
Update event (UEV)
MSv37305V1
596/1751 RM0090 Rev 19

CK_INT
CNT_EN
Timerclock = CK_CNT
Counter underflow
Update event (UEV)

MSv37306V1
CK_INT
CNT_EN
Timerclock = CK_CNT
Counter underflow
Update event (UEV)

MS40511V1
CK_INT
Timerclock = CK_CNT
Counter overflow
Update event (UEV)

MS37340V1
RM0090 Rev 19 597/1751

649
Figure 147. Counter timing diagram, Update event
CK_INT
CNT_EN
Timerclock = CK_CNT
Counter overflow
Update event (UEV)

MS37341V1
Center-aligned mode (up/down counting)

In center-aligned mode, the counter counts from 0 to the auto-reload value (content of the
TIMx_ARR register) – 1, generates a counter overflow event, then counts from the auto-
reload value down to 1 and generates a counter underflow event. Then it restarts counting
from 0.
Center-aligned mode is active when the CMS bits in TIMx_CR1 register are not equal to
'00'. The Output compare interrupt flag of channels configured in output is set when: the
counter counts down (Center aligned mode 1, CMS = "01"), the counter counts up (Center
aligned mode 2, CMS = "10") the counter counts up and down (Center aligned mode 3,
CMS = "11").
In this mode, the direction bit (DIR from TIMx_CR1 register) cannot be written. It is updated
by hardware and gives the current direction of the counter.
The update event can be generated at each counter overflow and at each counter underflow
or by setting the UG bit in the TIMx_EGR register (by software or by using the slave mode
controller) also generates an update event. In this case, the counter restarts counting from
0, as well as the counter of the prescaler.
However, the counter continues counting up and down, based on the current auto-reload
value.
DMA request is sent). This is to avoid generating both update and capture interrupt when
598/1751 RM0090 Rev 19

register).
TIMx_ARR register). Note that if the update source is a counter overflow, the auto-
reload is updated before the counter is reloaded, so that the next period is the expected
one (the counter is loaded with the new value).
frequencies.
CK_INT
CNT_EN
Timerclock = CK_CNT
Counter underflow
Counter overflow
Update event (UEV)

MS37342V1
1. Here, center-aligned mode 1 is used, for more details refer to Section 18.4.1: TIMx control register 1 (TIMx_CR1).
CK_INT
CNT_EN
Timerclock = CK_CNT
Counter underflow
Update event (UEV)

MS37343V1
RM0090 Rev 19 599/1751

649
CK_INT
CNT_EN
Timerclock = CK_CNT
Counter overflow (cnt_ovf)
Update event (UEV)

MS37344V1
1. Center-aligned mode 2 or 3 is used with an UIF on overflow.

CK_INT
Timerclock = CK_CNT
Counter underflow
Update event (UEV)

MS37345V1
600/1751 RM0090 Rev 19

Figure 152. Counter timing diagram, Update event with ARPE=1 (counter underflow)
CK_INT
CNT_EN
Timerclock = CK_CNT
Counter underflow
Update event (UEV)

MS37360V1
Figure 153. Counter timing diagram, Update event with ARPE=1 (counter overflow)
CK_INT
CNT_EN
Counter register F7 F8 F9 FA FB FC 36 35 34 33 32 31 30 2F
Counter overflow
Update event (UEV)

MS37361V1
RM0090 Rev 19 601/1751

649

• External clock mode1: external input pin (TIx)
• External clock mode2: external trigger input (ETR) available on TIM2, TIM3 and TIM4
only.
• Internal trigger inputs (ITRx): using one timer as prescaler for another timer, for
example, Timer3 can be configured to act as a prescaler for Timer 2. Refer to Using
one timer as prescaler for another timer for more details.

If the slave mode controller is disabled (SMS=000 in the TIMx_SMCR register), then the
CEN, DIR (in the TIMx_CR1 register) and UG bits (in the TIMx_EGR register) are actual
control bits and can be changed only by software (except UG which remains cleared
automatically). As soon as the CEN bit is written to 1, the prescaler is clocked by the internal
clock CK_INT.
without prescaler.
Internal clock
CEN=CNT_EN
UG
CNT_INIT
MS31085V2

This mode is selected when SMS=111 in the TIMx_SMCR register. The counter can count at
each rising or falling edge on a selected input.
602/1751 RM0090 Rev 19


TIMx_SMCR
TS[2:0]
or TI2F or
TI1F or Encoder
ITRx mode
0xx
TI1_ED
100 TRGI External clock
TI1FP1 mode 1 CK_PSC
TI2F_Rising 101
TI2 Edge 0 TI2FP2 ETRF External clock
Filter 110
detector 1 ETRF mode 2
TI2F_Falling 111
mode
ECE SMS[2:0]
TIMx_SMCR
MS31196V1
1. Configure channel 2 to detect rising edges on the TI2 input by writing CC2S= ‘01 in the
TIMx_CCMR1 register.
register (if no filter is needed, keep IC2F=0000).
Note: The capture prescaler is not used for triggering, so there’s no need to configure it.
3. Select rising edge polarity by writing CC2P=0 and CC2NP=0 in the TIMx_CCER
register.
4. Configure the timer in external clock mode 1 by writing SMS=111 in the TIMx_SMCR
register.
5. Select TI2 as the input source by writing TS=110 in the TIMx_SMCR register.
RM0090 Rev 19 603/1751

649
TI2
CNT_EN
TIF
Write TIF=0
MS31087V2

This mode is selected by writing ECE=1 in the TIMx_SMCR register.
The counter can count at each rising or falling edge on the external trigger input ETR.
Figure 157 gives an overview of the external trigger input block.
Figure 157. External trigger input block
or TI2F or
TI1F or Encoder
mode
TRGI External clock
ETR CK_PSC
mode 1
0
ETR pin Divider ETRP Filter ETRF External clock
1 /1, /2, /4, /8 downcounter mode 2
CK_INT CK_INT Internal clock
ETP ETPS[1:0] ETF[3:0] (internal clock) mode
TIMx_SMCR TIMx_SMCR TIMx_SMCR
ECE SMS[2:0]
TIMx_SMCR
MS37365V1
For example, to configure the upcounter to count each 2 rising edges on ETR, use the
following procedure:
1. As no filter is needed in this example, write ETF[3:0]=0000 in the TIMx_SMCR register.
2. Set the prescaler by writing ETPS[1:0]=01 in the TIMx_SMCR register
3. Select rising edge detection on the ETR pin by writing ETP=0 in the TIMx_SMCR
register
4. Enable external clock mode 2 by writing ECE=1 in the TIMx_SMCR register.
The counter counts once each 2 ETR rising edges.
604/1751 RM0090 Rev 19

The delay between the rising edge on ETR and the actual clock of the counter is due to the
resynchronization circuit on the ETRP signal.
CK_INT
CNT_EN
ETR
ETRP
ETRF
Counter clock = CK_INT =CK_PSC
MS37362V1

Each Capture/Compare channel (see Figure 159) is built around a capture/compare register
(including a shadow register), an input stage for capture (with digital filter, multiplexing and
prescaler) and an output stage (with comparator and output control).
TI1F_ED
to the slave mode controller
TI1F_Rising
TI1 TI1F
filter Edge TI1FP1
01
fDTS downcounter Detector TI1F_Falling
TI2FP1 IC1 divider IC1PS
10
/1, /2, /4, /8
11
controller)
TI2F_rising
(from channel 2)
TI2F_falling
The output stage generates an intermediate waveform which is then used for reference:
RM0090 Rev 19 605/1751

649
APB Bus
(if 16-bit)
8 8
high
S write CCR1H
low
R Output CC1S[1]
CC1S[0]
CC1S[1] Input
mode OC1PE
UEV
TIMx_CCMR1
Comparator (from time
IC1PS Capture
base unit)
CC1E CNT>CCR1
Counter CNT=CCR1
CC1G
TIMx_EGR
MS33144V1

ETRF
To the master mode
controller
0
Output OC1
Enable
1 Circuit
CC1P
TIMx_CNT > TIMx_CCR1
Output mode oc1ref TIMx_CCER
TIMx_CNT = TIMx_CCR1 controller
CC1E
TIMx_CCER
OC1M[2:0]
TIMx_CCMR1 ai17187b
preload register.
606/1751 RM0090 Rev 19


cleared by software by writing it to 0 or by reading the captured data stored in the
TIMx_CCRx register. CCxOF is cleared when written to 0.
• Select the active input: TIMx_CCR1 must be linked to the TI1 input, so write the CC1S
bits to 01 in the TIMx_CCMR1 register. As soon as CC1S becomes different from 00,
the channel is configured in input and the TIMx_CCR1 register becomes read-only.
• Program the needed input filter duration with respect to the signal connected to the
timer (by programming the ICxF bits in the TIMx_CCMRx register if the input is one of
the TIx inputs). Let’s imagine that, when toggling, the input signal is not stable during at
must five internal clock cycles. We must program a filter duration longer than these five
clock cycles. We can validate a transition on TI1 when eight consecutive samples with
the new level have been detected (sampled at fDTS frequency). Then write IC1F bits to
0011 in the TIMx_CCMR1 register.
• Select the edge of the active transition on the TI1 channel by writing the CC1P and
CC1NP bits to 00 in the TIMx_CCER register (rising edge in this case).
• Program the input prescaler. In our example, we wish the capture to be performed at
each valid transition, so the prescaler is disabled (write IC1PS bits to 00 in the
• Enable capture from the counter into the capture register by setting the CC1E bit in the
TIMx_CCER register.
• If needed, enable the related interrupt request by setting the CC1IE bit in the
TIMx_DIER register, and/or the DMA request by setting the CC1DE bit in the
TIMx_DIER register.
• A DMA request is generated depending on the CC1DE bit.
Note: IC interrupt and/or DMA requests can be generated by software by setting the
corresponding CCxG bit in the TIMx_EGR register.
RM0090 Rev 19 607/1751

649
18.3.6 PWM input mode

For example, the user can measure the period (in TIMx_CCR1 register) and the duty cycle
(in TIMx_CCR2 register) of the PWM applied on TI1 using the following procedure
(depending on CK_INT frequency and prescaler value):
• Select the active polarity for TI1FP1 (used both for capture in TIMx_CCR1 and counter
clear): write the CC1P to ‘0’ and the CC1NP bit to ‘0’ (active on rising edge).
• Select the active polarity for TI1FP2 (used for capture in TIMx_CCR2): write the CC2P
bit to ‘1’ and the CC2NP bit to ’0’(active on falling edge).
• Select the valid trigger input: write the TS bits to 101 in the TIMx_SMCR register
(TI1FP1 selected).
• Configure the slave mode controller in reset mode: write the SMS bits to 100 in the
TIMx_SMCR register.
• Enable the captures: write the CC1E and CC2E bits to ‘1 in the TIMx_CCER register.
TI1
TIMx_CNT 0004 0000 0001 0002 0003 0004 0000
TIMx_CCR1 0004
TIMx_CCR2 0002
IC1 capture IC2 capture IC1 capture

pu
l eswid th period
IC2 capture
measurement measurement
reset counter
ai15413
608/1751 RM0090 Rev 19


In output mode (CCxS bits = 00 in the TIMx_CCMRx register), each output compare signal
(OCxREF and then OCx) can be forced to active or inactive level directly by software,
To force an output compare signal (ocxref/OCx) to its active level, the user just needs to
write 101 in the OCxM bits in the corresponding TIMx_CCMRx register. Thus ocxref is
bit.
e.g.: CCxP=0 (OCx active high) => OCx is forced to high level.
ocxref signal can be forced low by writing the OCxM bits to 100 in the TIMx_CCMRx
register.
performed and allows the flag to be set. Interrupt and DMA requests can be sent
accordingly. This is described in the next section.

elapsed.
compare function:
• Assigns the corresponding output pin to a programmable value defined by the output
bit in the TIMx_CCER register). The output pin can keep its level (OCXM=000), be set
active (OCxM=001), be set inactive (OCxM=010) or can toggle (OCxM=011) on match.
• Sets a flag in the interrupt status register (CCxIF bit in the TIMx_SR register).
• Generates an interrupt if the corresponding interrupt mask is set (CCXIE bit in the
• Sends a DMA request if the corresponding enable bit is set (CCxDE bit in the
TIMx_DIER register, CCDS bit in the TIMx_CR2 register for the DMA request
selection).
In output compare mode, the update event UEV has no effect on ocxref and OCx output.
output a single pulse (in One-pulse mode).
Procedure:
3. Set the CCxIE and/or CCxDE bits if an interrupt and/or a DMA request is to be
generated.
4. Select the output mode. For example, the user must write OCxM=011, OCxPE=0,
CCxP=0 and CCxE=1 to toggle OCx output pin when CNT matches CCRx, CCRx
preload is not used, OCx is enabled and active high.
RM0090 Rev 19 609/1751

649
waveform, provided that the preload register is not enabled (OCxPE=0, else TIMx_CCRx
Figure 163.
Figure 163. Output compare mode, toggle on OC1

TIMx_CNT 0039 003A 003B B200 B201
TIMx_CCR1 003A B201
OC1REF = OC1

MS37363V1
18.3.9 PWM mode

Pulse width modulation mode allows generating a signal with a frequency determined by the
value of the TIMx_ARR register and a duty cycle determined by the value of the
TIMx_CCRx register.
output) by writing 110 (PWM mode 1) or ‘111 (PWM mode 2) in the OCxM bits in the
TIMx_CCMRx register. The user must enable the corresponding preload register by setting
the OCxPE bit in the TIMx_CCMRx register, and eventually the auto-reload preload register
by setting the ARPE bit in the TIMx_CR1 register.
occurs, before starting the counter, the user has to initialize all the registers by setting the
UG bit in the TIMx_EGR register.
OCx polarity is software programmable using the CCxP bit in the TIMx_CCER register. It
can be programmed as active high or active low. OCx output is enabled by the CCxE bit in
the TIMx_CCER register. Refer to the TIMx_CCERx register description for more details.
whether TIMx_CCRx≤ TIMx_CNT or TIMx_CNT≤ TIMx_CCRx (depending on the direction
of the counter). However, to comply with the ETRF (OCREF can be cleared by an external
event through the ETR signal until the next PWM period), the OCREF signal is asserted
only:
• When the result of the comparison changes, or
• When the output compare mode (OCxM bits in TIMx_CCMRx register) switches from
the “frozen” configuration (no comparison, OCxM=‘000) to one of the PWM modes
(OCxM=‘110 or ‘111).
This forces the PWM by software while the timer is running.
610/1751 RM0090 Rev 19

The timer is able to generate PWM in edge-aligned mode or center-aligned mode

depending on the CMS bits in the TIMx_CR1 register.
Upcounting configuration
Upcounting is active when the DIR bit in the TIMx_CR1 register is low. Refer to Upcounting
mode.
In the following example, we consider PWM mode 1. The reference PWM signal OCxREF is
high as long as TIMx_CNT <TIMx_CCRx else it becomes low. If the compare value in
TIMx_CCRx is greater than the auto-reload value (in TIMx_ARR) then OCxREF is held at ‘1.
If the compare value is 0 then OCxREF is held at ‘0. Figure 164 shows some edge-aligned
PWM waveforms in an example where TIMx_ARR=8.
OCXREF
CCRx=4
CCxIF
OCXREF
CCRx=8
CCxIF
OCXREF ‘1’
CCRx>8
CCxIF
OCXREF ‘0’
CCRx=0
CCxIF
MS31093V1
Downcounting configuration
Downcounting is active when DIR bit in TIMx_CR1 register is high. Refer to Downcounting
mode.
In PWM mode 1, the reference signal ocxref is low as long as TIMx_CNT>TIMx_CCRx else
it becomes high. If the compare value in TIMx_CCRx is greater than the auto-reload value in
TIMx_ARR, then ocxref is held at ‘1. 0% PWM is not possible in this mode.
PWM center-aligned mode

Center-aligned mode is active when the CMS bits in TIMx_CR1 register are different from
‘00 (all the remaining configurations having the same effect on the ocxref/OCx signals). The
compare flag is set when the counter counts up, when it counts down or both when it counts
RM0090 Rev 19 611/1751

649
up and down depending on the CMS bits configuration. The direction bit (DIR) in the
TIMx_CR1 register is updated by hardware and must not be changed by software. Refer to
Center-aligned mode (up/down counting).
Figure 165 shows some center-aligned PWM waveforms in an example where:
• TIMx_ARR=8,
• PWM mode is the PWM mode 1,
• The flag is set when the counter counts down corresponding to the center-aligned
mode 1 selected for CMS=01 in TIMx_CR1 register.
Figure 165. Center-aligned PWM waveforms (ARR=8)
Counter register 0 1 2 3 4 5 6 7 8 7 6 5 4 3 2 1 0 1
OCxREF
CCRx = 4
CCxIF CMS=01
CMS=10
CMS=11
OCxREF
CCRx = 7
CMS=10 or 11
CCxIF
'1'
OCxREF
CCRx = 8
CCxIF CMS=01
CMS=10
CMS=11
'1'
OCxREF
CCRx > 8
CCxIF CMS=01
CMS=10
CMS=11
'0'
OCxREF
CCRx = 0
CCxIF CMS=01
CMS=10
CMS=11
ai14681b
Hints on using center-aligned mode:

• When starting in center-aligned mode, the current up-down configuration is used. It
means that the counter counts up or down depending on the value written in the DIR bit
612/1751 RM0090 Rev 19

in the TIMx_CR1 register. Moreover, the DIR and CMS bits must not be changed at the
same time by the software.
• Writing to the counter while running in center-aligned mode is not recommended as it
can lead to unexpected results. In particular:
– The direction is not updated if the user writes a value in the counter that is greater
than the auto-reload value (TIMx_CNT>TIMx_ARR). For example, if the counter
was counting up, it continues to count up.
– The direction is updated if the user writes 0 or write the TIMx_ARR value in the
counter but no Update Event UEV is generated.
• The safest way to use center-aligned mode is to generate an update by software
(setting the UG bit in the TIMx_EGR register) just before starting the counter and not to
write the counter while it is running.

be:
• In upcounting: CNT<CCRx≤ ARR (in particular, 0<CCRx),
• In downcounting: CNT>CCRx.
Figure 166. Example of one-pulse mode
TI2
OC1REF
OC1
TIM1_ARR
Counter
TIM1_CCR1
0
tDELAY tPULSE t
MS31099V1
RM0090 Rev 19 613/1751

649
Let’s use TI2FP2 as trigger 1:

• Map TI2FP2 on TI2 by writing CC2S=01 in the TIMx_CCMR1 register.
• TI2FP2 must detect a rising edge, write CC2P=0 and CC2NP=’0’ in the TIMx_CCER
register.
• Configure TI2FP2 as trigger for the slave mode controller (TRGI) by writing TS=110 in
• TI2FP2 is used to start the counter by writing SMS to ‘110 in the TIMx_SMCR register
(trigger mode).
value (TIMx_ARR - TIMx_CCR + 1).
• Let us say user wants to build a waveform with a transition from ‘0 to ‘1 when a
compare match occurs and a transition from ‘1 to ‘0 when the counter reaches the
auto-reload value. To do this enable PWM mode 2 by writing OC1M=111 in the
OC1PE=1 in the TIMx_CCMR1 register and ARPE in the TIMx_CR1 register. In this
case write the compare value in the TIMx_CCR1 register, the auto-reload value in the
TIMx_ARR register, generate an update by setting the UG bit and wait for external
trigger event on TI2. CC1P is written to ‘0 in this example.
In our example, the DIR and CMS bits in the TIMx_CR1 register should be low.
User only wants one pulse (Single mode), so write '1 in the OPM bit in the TIMx_CR1
register to stop the counter at the next update event (when the counter rolls over from the
auto-reload value back to 0). When OPM bit in the TIMx_CR1 register is set to '0', so the
Repetitive mode is selected.
Particular case: OCx fast enable:

To output a waveform with the minimum delay, the user can set the OCxFE bit in the
TIMx_CCMRx register. Then OCxRef (and OCx) is forced in response to the stimulus,
without taking in account the comparison. Its new level is the same as if a compare match
had occurred. OCxFE acts only if the channel is configured in PWM1 or PWM2 mode.
18.3.11 Clearing the OCxREF signal on an external event

The OCxREF signal for a given channel can be driven Low by applying a High level to the
ETRF input (OCxCE enable bit of the corresponding TIMx_CCMRx register set to '1'). The
OCxREF signal remains Low until the next update event, UEV, occurs.
This function can only be used in output compare and PWM modes, and does not work in
forced mode.
For example, the ETR signal can be connected to the output of a comparator to be used for
current handling. In this case, ETR must be configured as follows:
614/1751 RM0090 Rev 19

1. The external trigger prescaler should be kept off: bits ETPS[1:0] in the TIMx_SMCR
register are cleared to 00.
2. The external clock mode 2 must be disabled: bit ECE in the TIM1_SMCR register is
cleared to 0.
3. The external trigger polarity (ETP) and the external trigger filter (ETF) can be
configured according to the application’s needs.
Figure 167 shows the behavior of the OCxREF signal when the ETRF input becomes high,
for both values of the OCxCE enable bit. In this example, the timer TIMx is programmed in
PWM mode.
Figure 167. Clearing TIMx OCxREF
(CCRx)
Counter (CNT)
ETRF

ETRF becomes high ETRF still high
MS37368V1
1. In case of a PWM with a 100% duty cycle (if CCRx>ARR), OCxREF is enabled again at the next counter
overflow.
18.3.12 Encoder interface mode

To select Encoder Interface mode write SMS=‘001 in the TIMx_SMCR register if the counter
is counting on TI2 edges only, SMS=010 if it is counting on TI1 edges only and SMS=011 if
it is counting on both TI1 and TI2 edges.
Select the TI1 and TI2 polarity by programming the CC1P and CC2P bits in the TIMx_CCER
register. When needed, program the input filter as well.
The two inputs TI1 and TI2 are used to interface to an incremental encoder. Refer to
Table 97. The counter is clocked by each valid transition on TI1FP1 or TI2FP2 (TI1 and TI2
after input filter and polarity selection, TI1FP1=TI1 if not filtered and not inverted,
TI2FP2=TI2 if not filtered and not inverted) assuming that it is enabled (CEN bit in
TIMx_CR1 register written to ‘1). The sequence of transitions of the two inputs is evaluated
and generates count pulses as well as the direction signal. Depending on the sequence the
counter counts up or down, the DIR bit in the TIMx_CR1 register is modified by hardware
accordingly. The DIR bit is calculated at each transition on any input (TI1 or TI2), whatever
the counter is counting on TI1 only, TI2 only or both TI1 and TI2.
Encoder interface mode acts simply as an external clock with direction selection. This
means that the counter just counts continuously between 0 and the auto-reload value in the
TIMx_ARR register (0 to ARR or ARR down to 0 depending on the direction). So the user
must configure TIMx_ARR before starting. In the same way, the capture, compare,
prescaler, trigger output features continue to work as normal.
RM0090 Rev 19 615/1751

649
In this mode, the counter is modified automatically following the speed and the direction of
the incremental encoder and its content, therefore, always represents the encoder’s
position. The count direction correspond to the rotation direction of the connected sensor.
The table summarizes the possible combinations, assuming TI1 and TI2 don’t switch at the
same time.
Table 97. Counting direction versus encoder signals

Level on opposite TI1FP1 signal TI2FP2 signal
Active edge signal (TI1FP1 for
TI2, TI2FP2 for TI1) Rising Falling Rising Falling
Counting on High Down Up No Count No Count

TI1 only Low Up Down No Count No Count
Counting on High No Count No Count Up Down

TI2 only Low No Count No Count Down Up
Counting on High Down Up Up Down

TI1 and TI2 Low Up Down Down Up
An external incremental encoder can be connected directly to the MCU without external
interface logic. However, comparators are normally be used to convert the encoder’s
differential outputs to digital signals. This greatly increases noise immunity. The third
encoder output which indicate the mechanical zero position, may be connected to an
external interrupt input and trigger a counter reset.
Figure 168 gives an example of counter operation, showing count signal generation and
direction control. It also shows how input jitter is compensated where both edges are
selected. This might occur if the sensor is positioned near to one of the switching points. For
this example we assume that the configuration is the following:
• CC1S= ‘01’ (TIMx_CCMR1 register, TI1FP1 mapped on TI1)
• CC2S= ‘01’ (TIMx_CCMR2 register, TI2FP2 mapped on TI2)
• CC1P= ‘0’, CC1NP = ‘0’, IC1F =’0000’ (TIMx_CCER register, TI1FP1 noninverted,
TI1FP1=TI1)
• CC2P= ‘0’, CC2NP = ‘0’, IC2F =’0000’ (TIMx_CCER register, TI2FP2 noninverted,
TI2FP2=TI2)
• SMS= ‘011’ (TIMx_SMCR register, both inputs are active on both rising and falling
edges)
• CEN = 1 (TIMx_CR1 register, Counter is enabled)
616/1751 RM0090 Rev 19

Figure 168. Example of counter operation in encoder interface mode

TI1
TI2
Counter
up down up
MS33107V1
Figure 169 gives an example of counter behavior when TI1FP1 polarity is inverted (same
configuration as above except CC1P=1).
Figure 169. Example of encoder interface mode with TI1FP1 polarity inverted
TI1
TI2
Counter
down up down
MS33108V1
The timer, when configured in Encoder Interface mode provides information on the sensor’s
current position. The user can obtain dynamic information (speed, acceleration,
deceleration) by measuring the period between two encoder events using a second timer
configured in capture mode. The output of the encoder which indicates the mechanical zero
can be used for this purpose. Depending on the time between two events, the counter can
also be read at regular times. The user can do this by latching the counter value into a third
input capture register if available (then the capture signal must be periodic and can be
generated by another timer). when available, it is also possible to read its value through a
DMA request generated by a Real-Time clock.
RM0090 Rev 19 617/1751

649
18.3.13 Timer input XOR function

The TI1S bit in the TIM_CR2 register, allows the input filter of channel 1 to be connected to
the output of a XOR gate, combining the three input pins TIMx_CH1 to TIMx_CH3.
The XOR output can be used with all the timer input functions such as trigger or input
capture.
18.3.14 Timers and external trigger synchronization

The TIMx Timers can be synchronized with an external trigger in several modes: Reset
mode, Gated mode and Trigger mode.

CC1S bits select the input capture source only, CC1S = 01 in the TIMx_CCMR1
register. Write CC1P=0 and CC1NP=0 in TIMx_CCER register to validate the polarity
(and detect rising edges only).
• Configure the timer in reset mode by writing SMS=100 in TIMx_SMCR register. Select
• Start the counter by writing CEN=1 in the TIMx_CR1 register.
trigger flag is set (TIF bit in the TIMx_SR register) and an interrupt request, or a DMA
request can be sent if enabled (depending on the TIE and TDE bits in TIMx_DIER register).
Figure 170 shows this behavior when the auto-reload register TIMx_ARR=0x36. The delay
between the rising edge on TI1 and the actual reset of the counter is due to the
TI1
UG

TIF
MS37384V1
618/1751 RM0090 Rev 19


• Configure the channel 1 to detect low levels on TI1. Configure the input filter duration
CC1S bits select the input capture source only, CC1S=01 in TIMx_CCMR1 register.
Write CC1P=1 in TIMx_CCER register to validate the polarity (and detect low level
only).
• Configure the timer in gated mode by writing SMS=101 in TIMx_SMCR register. Select
• Enable the counter by writing CEN=1 in the TIMx_CR1 register (in gated mode, the
counter doesn’t start if CEN=0, whatever is the trigger input level).
or stops.
TI1
CNT_EN
TIF
Write TIF=0
MS40512V1
1. The configuration “CCxP=CCxNP=1” (detection of both rising and falling edges) does not have any effect
in gated mode because gated mode acts on a level and not on an edge.

prescaler is not used for triggering, so the user does not need to configure it. CC2S bits
are selecting the input capture source only, CC2S=01 in TIMx_CCMR1 register. Write
CC2P=1 in TIMx_CCER register to validate the polarity (and detect low level only).
• Configure the timer in trigger mode by writing SMS=110 in TIMx_SMCR register. Select
RM0090 Rev 19 619/1751

649
TIF flag is set.

TI2
CNT_EN

TIF
MS37386V1
Slave mode: External Clock mode 2 + trigger mode

The external clock mode 2 can be used in addition to another slave mode (except external
clock mode 1 and encoder mode). In this case, the ETR signal is used as external clock
input, and another input can be selected as trigger input when operating in reset mode,
gated mode or trigger mode. It is recommended not to select ETR as TRGI through the TS
bits of TIMx_SMCR register.
In the following example, the upcounter is incremented at each rising edge of the ETR
signal as soon as a rising edge of TI1 occurs:
1. Configure the external trigger input circuit by programming the TIMx_SMCR register as
follows:
– ETF = 0000: no filter
– ETPS = 00: prescaler disabled
– ETP = 0: detection of rising edges on ETR and ECE=1 to enable the external clock
mode 2.
2. Configure the channel 1 as follows, to detect rising edges on TI:
– IC1F = 0000: no filter.
– The capture prescaler is not used for triggering and does not need to be
configured.
– CC1S = 01 in TIMx_CCMR1 register to select only the input capture source
– CC1P = 0 in TIMx_CCER register to validate the polarity (and detect rising edge
only).
3. Configure the timer in trigger mode by writing SMS=110 in TIMx_SMCR register. Select
A rising edge on TI1 enables the counter and sets the TIF flag. The counter then counts on
ETR rising edges.
The delay between the rising edge of the ETR signal and the actual reset of the counter is
due to the resynchronization circuit on ETRP input.
620/1751 RM0090 Rev 19

Figure 173. Control circuit in external clock mode 2 + trigger mode
TI1
CEN/CNT_EN
ETR
TIF
MS33110V1
18.3.15 Timer synchronization

The TIMx timers are linked together internally for timer synchronization or chaining. When
one Timer is configured in Master mode, it can reset, start, stop or clock the counter of
another Timer configured in Slave mode.
Figure 174 presents an overview of the trigger selection and the master mode selection
blocks.
Note: The clock of the slave timer must be enabled prior to receiving events from the master timer,
Using one timer as prescaler for another timer
Figure 174. Master/Slave timer example

TIM1 TIM2
Clock
MMS TS SMS
UEV
Master Slave
TRGO1 ITR0 CK_PSC
mode mode
Prescaler Counter control control Prescaler Counter
Input trigger
selection
MS37387V1
RM0090 Rev 19 621/1751

649
For example, the user can configure Timer 1 to act as a prescaler for Timer 2 (see
Figure 174). To do this:
• Configure Timer 1 in master mode so that it outputs a periodic trigger signal on each
update event UEV. If you write MMS=010 in the TIM1_CR2 register, a rising edge is
output on TRGO1 each time an update event is generated.
• To connect the TRGO1 output of Timer 1 to Timer 2, Timer 2 must be configured in
slave mode using ITR0 as internal trigger. You select this through the TS bits in the
TIM2_SMCR register (writing TS=000).
• Then you put the slave mode controller in external clock mode 1 (write SMS=111 in the
TIM2_SMCR register). This causes Timer 2 to be clocked by the rising edge of the
periodic Timer 1 trigger signal (which correspond to the timer 1 counter overflow).
• Finally both timers must be enabled by setting their respective CEN bits (TIMx_CR1
register).
Note: If OCx is selected on Timer 1 as trigger output (MMS=1xx), its rising edge is used to clock
the counter of timer 2.
Using one timer to enable another timer

In this example, we control the enable of Timer 2 with the output compare 1 of Timer 1.
Refer to Figure 174 for connections. Timer 2 counts on the divided internal clock only when
OC1REF of Timer 1 is high. Both counter clock frequencies are divided by 3 by the
prescaler compared to CK_INT (fCK_CNT = fCK_INT/3).
• Configure Timer 1 master mode to send its Output compare 1 Reference (OC1REF)
signal as trigger output (MMS=100 in the TIM1_CR2 register).
• Configure the Timer 1 OC1REF waveform (TIM1_CCMR1 register).
• Configure Timer 2 to get the input trigger from Timer 1 (TS=000 in the TIM2_SMCR
register).
• Configure Timer 2 in gated mode (SMS=101 in TIM2_SMCR register).
• Enable Timer 2 by writing ‘1 in the CEN bit (TIM2_CR1 register).
• Start Timer 1 by writing ‘1 in the CEN bit (TIM1_CR1 register).
Note: The counter 2 clock is not synchronized with counter 1, this mode only affects the Timer 2
counter enable signal.
Figure 175. Gating timer 2 with OC1REF of timer 1
CK_INT
TIMER1-OC1REF
TIMER1-CNT FC FD FE FF 00 01
TIMER2-CNT 3045 3046 3047 3048
TIMER2-TIF
Write TIF = 0
MS37388V1
In the example in Figure 175, the Timer 2 counter and prescaler are not initialized before
being started. So they start counting from their current value. It is possible to start from a
given value by resetting both timers before starting Timer 1. You can then write any value
622/1751 RM0090 Rev 19

you want in the timer counters. The timers can easily be reset by software using the UG bit
in the TIMx_EGR registers.
In the next example, we synchronize Timer 1 and Timer 2. Timer 1 is the master and starts
from 0. Timer 2 is the slave and starts from 0xE7. The prescaler ratio is the same for both
timers. Timer 2 stops when Timer 1 is disabled by writing ‘0 to the CEN bit in the TIM1_CR1
register:
• Configure Timer 1 master mode to send its Output compare 1 Reference (OC1REF)
signal as trigger output (MMS=100 in the TIM1_CR2 register).
• Configure the Timer 1 OC1REF waveform (TIM1_CCMR1 register).
register).
• Configure Timer 2 in gated mode (SMS=101 in TIM2_SMCR register).
• Reset Timer 1 by writing ‘1 in UG bit (TIM1_EGR register).
• Reset Timer 2 by writing ‘1 in UG bit (TIM2_EGR register).
• Initialize Timer 2 to 0xE7 by writing ‘0xE7’ in the timer 2 counter (TIM2_CNTL).
• Enable Timer 2 by writing ‘1 in the CEN bit (TIM2_CR1 register).
• Stop Timer 1 by writing ‘0 in the CEN bit (TIM1_CR1 register).
Figure 176. Gating timer 2 with Enable of timer 1
CK_INT
TIMER1-CEN=CNT_EN
TIMER1-CNT_INIT
TIMER1-CNT 75 00 01 02
TIMER2-CNT AB 00 E7 E8 E9
TIMER2-CNT_INIT
TIMER2-write CNT
TIMER2-TIF
Write TIF = 0
MS37389V1
RM0090 Rev 19 623/1751

649
Using one timer to start another timer

In this example, we set the enable of Timer 2 with the update event of Timer 1. Refer to
Figure 174 for connections. Timer 2 starts counting from its current value (which can be
nonzero) on the divided internal clock as soon as the update event is generated by Timer 1.
When Timer 2 receives the trigger signal its CEN bit is automatically set and the counter
counts until we write ‘0 to the CEN bit in the TIM2_CR1 register. Both counter clock
frequencies are divided by 3 by the prescaler compared to CK_INT (fCK_CNT = fCK_INT/3).
• Configure Timer 1 master mode to send its Update Event (UEV) as trigger output
(MMS=010 in the TIM1_CR2 register).
• Configure the Timer 1 period (TIM1_ARR registers).
register).
• Configure Timer 2 in trigger mode (SMS=110 in TIM2_SMCR register).
Figure 177. Triggering timer 2 with update of timer 1
CK_INT
TIMER1-UEV
TIMER1-CNT FD FE FF 00 01 02
TIMER2-CNT 45 46 47 48
TIMER2-CEN=CNT_EN
TIMER2-TIF
Write TIF = 0
MS37390V1
As in the previous example, the user can initialize both counters before starting counting.
Figure 178 shows the behavior with the same configuration as in Figure 177 but in trigger
mode instead of gated mode (SMS=110 in the TIM2_SMCR register).
624/1751 RM0090 Rev 19

Figure 178. Triggering timer 2 with Enable of timer 1
CK_INT
TIMER1-CEN=CNT_EN
TIMER1-CNT_INIT
TIMER1-CNT 75 00 01 02
TIMER2-CNT CD 00 E7 E8 E9 EA
TIMER2-CNT_INIT
TIMER2-write CNT
TIMER2-TIF
Write TIF = 0
MS37391V1
Starting 2 timers synchronously in response to an external trigger

In this example, we set the enable of timer 1 when its TI1 input rises, and the enable of
Timer 2 with the enable of Timer 1. Refer to Figure 174 for connections. To ensure the
counters are aligned, Timer 1 must be configured in Master/Slave mode (slave with respect
to TI1, master with respect to Timer 2):
• Configure Timer 1 master mode to send its Enable as trigger output (MMS=001 in the
TIM1_CR2 register).
• Configure Timer 1 slave mode to get the input trigger from TI1 (TS=100 in the
TIM1_SMCR register).
• Configure Timer 1 in trigger mode (SMS=110 in the TIM1_SMCR register).
• Configure the Timer 1 in Master/Slave mode by writing MSM=1 (TIM1_SMCR register).
register).
• Configure Timer 2 in trigger mode (SMS=110 in the TIM2_SMCR register).
When a rising edge occurs on TI1 (Timer 1), both counters starts counting synchronously on
the internal clock and both TIF flags are set.
Note: In this example both timers are initialized before starting (by setting their respective UG
bits). Both counters starts from 0, but you can easily insert an offset between them by
writing any of the counter registers (TIMx_CNT). You can see that the master/slave mode
insert a delay between CNT_EN and CK_PSC on timer 1.
RM0090 Rev 19 625/1751

649
Figure 179. Triggering timer 1 and 2 with timer 1 TI1 input
CK_INT
TIMER1-TI1
TIMER1-CEN=CNT_EN
TIMER1-CK_PSC
TIMER1-CNT 00 01 02 03 04 05 06 07 08 09
TIMER1-TIF
TIMER2-CEN=CNT_EN
TIMER2-CK_PSC
TIMER2-CNT 00 01 02 03 04 05 06 07 08 09
TIMER2-TIF
MS37392V1
18.3.16 Debug mode

When the microcontroller enters debug mode (Cortex®-M4 with FPU core - halted), the
TIMx counter either continues to work normally or stops, depending on DBG_TIMx_STOP
configuration bit in DBGMCU module. For more details, refer to Section 38.16.2: Debug
support for timers, watchdog, bxCAN and I2C.
626/1751 RM0090 Rev 19

18.4 TIM2 to TIM5 registers

The 32-bit peripheral registers have to be written by words (32 bits). All other peripheral
registers have to be written by half-words (16 bits) or words (32 bits). Read accesses can be
done by bytes (8 bits), half-words (16 bits) or words (32 bits).
18.4.1 TIMx control register 1 (TIMx_CR1)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CKD[1:0] ARPE CMS DIR OPM URS UDIS CEN
Reserved

Bits 9:8 CKD: Clock division
This bit-field indicates the division ratio between the timer clock (CK_INT) frequency and
sampling clock used by the digital filters (ETR, TIx),
00: tDTS = tCK_INT
01: tDTS = 2 × tCK_INT
11: Reserved
Bits 6:5 CMS: Center-aligned mode selection
00: Edge-aligned mode. The counter counts up or down depending on the direction bit
(DIR).
only when the counter is counting down.
only when the counter is counting up.
both when the counter is counting up or down.
Note: It is not allowed to switch from edge-aligned mode to center-aligned mode as long as
the counter is enabled (CEN=1)
Bit 4 DIR: Direction
0: Counter used as upcounter
1: Counter used as downcounter
Note: This bit is read only when the timer is configured in Center-aligned mode or Encoder
mode.
Bit 3 OPM: One-pulse mode
1: Counter stops counting at the next update event (clearing the bit CEN)
RM0090 Rev 19 627/1751

649

0: Any of the following events generate an update interrupt or DMA request if enabled.
1: Only counter overflow/underflow generates an update interrupt or DMA request if
enabled.
(ARR, PSC, CCRx). However the counter and the prescaler are reinitialized if the UG bit is
set or if a hardware reset is received from the slave mode controller.
0: Counter disabled
1: Counter enabled
Note: External clock, gated mode and encoder mode can work only if the CEN bit has been
previously set by software. However trigger mode can set the CEN bit automatically by
hardware.
CEN is cleared automatically in one-pulse mode, when an update event occurs.
628/1751 RM0090 Rev 19

18.4.2 TIMx control register 2 (TIMx_CR2)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TI1S MMS[2:0] CCDS
Reserved Reserved
rw rw rw rw rw

Bit 7 TI1S: TI1 selection
0: The TIMx_CH1 pin is connected to TI1 input
1: The TIMx_CH1, CH2 and CH3 pins are connected to the TI1 input (XOR combination)
These bits allow to select the information to be sent in master mode to slave timers for
000: Reset - the UG bit from the TIMx_EGR register is used as trigger output (TRGO). If the
001: Enable - the Counter enable signal, CNT_EN, is used as trigger output (TRGO). It is
useful to start several timers at the same time or to control a window in which a slave timer is
enabled. The Counter Enable signal is generated by a logic OR between CEN control bit
and the trigger input when configured in gated mode.
When the Counter Enable signal is controlled by the trigger input, there is a delay on TRGO,
except if the master/slave mode is selected (see the MSM bit description in TIMx_SMCR
register).
010: Update - The update event is selected as trigger output (TRGO). For instance a master
timer can then be used as a prescaler for a slave timer.
011: Compare Pulse - The trigger output send a positive pulse when the CC1IF flag is to be
set (even if it was already high), as soon as a capture or a compare match occurred.
(TRGO)
the master timer.
Bit 3 CCDS: Capture/compare DMA selection
0: CCx DMA request sent when CCx event occurs
1: CCx DMA requests sent when update event occurs
RM0090 Rev 19 629/1751

649
18.4.3 TIMx slave mode control register (TIMx_SMCR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ETP ECE ETPS[1:0] ETF[3:0] MSM TS[2:0] SMS[2:0]
Res.
Bit 15 ETP: External trigger polarity

This bit selects whether ETR or ETR is used for trigger operations
0: ETR is noninverted, active at high level or rising edge
1: ETR is inverted, active at low level or falling edge
Bit 14 ECE: External clock enable
This bit enables External clock mode 2.
0: External clock mode 2 disabled
1: External clock mode 2 enabled. The counter is clocked by any active edge on the ETRF
signal.
1: Setting the ECE bit has the same effect as selecting external clock mode 1 with TRGI
connected to ETRF (SMS=111 and TS=111).
2: It is possible to simultaneously use external clock mode 2 with the following slave modes:
reset mode, gated mode and trigger mode. Nevertheless, TRGI must not be connected to
ETRF in this case (TS bits must not be 111).
3: If external clock mode 1 and external clock mode 2 are enabled at the same time, the
external clock input is ETRF.
Bits 13:12 ETPS: External trigger prescaler
External trigger signal ETRP frequency must be at most 1/4 of CK_INT frequency. A
prescaler can be enabled to reduce ETRP frequency. It is useful when inputting fast external
clocks.
00: Prescaler OFF
Bits 11:8 ETF[3:0]: External trigger filter
This bit-field then defines the frequency used to sample ETRP signal and the length of the
digital filter applied to ETRP. The digital filter is made of an event counter in which N
consecutive events are needed to validate a transition on the output:
630/1751 RM0090 Rev 19

Bit 7 MSM: Master/Slave mode

0: No action
synchronization between the current timer and its slaves (through TRGO). It is useful if we
want to synchronize several timers on a single external event.
Bits 6:4 TS: Trigger selection
This bit-field selects the trigger input to be used to synchronize the counter.
001: Internal Trigger 1 (ITR1).
111: External Trigger input (ETRF)
See Table 98: TIMx internal trigger connection on page 632 for more details on ITRx
meaning for each Timer.
Note: These bits must be changed only when they are not used (e.g. when SMS=000) to
When external signals are selected the active edge of the trigger signal (TRGI) is linked to
the polarity selected on the external input (see Input Control register and Control register
description.
000: Slave mode disabled - if CEN = ‘1 then the prescaler is clocked directly by the internal
clock.
level.
level.
011: Encoder mode 3 - Counter counts up/down on both TI1FP1 and TI2FP2 edges
depending on the level of the other input.
100: Reset mode - Rising edge of the selected trigger input (TRGI) reinitializes the counter
and generates an update of the registers.
101: Gated mode - The counter clock is enabled when the trigger input (TRGI) is high. The
counter stops (but is not reset) as soon as the trigger becomes low. Both start and stop of
the counter are controlled.
110: Trigger mode - The counter starts at a rising edge of the trigger TRGI (but it is not
reset). Only the start of the counter is controlled.
111: External Clock mode 1 - Rising edges of the selected trigger (TRGI) clock the counter.
Note: The gated mode must not be used if TI1F_ED is selected as the trigger input (TS=100).
Indeed, TI1F_ED outputs 1 pulse for each transition on TI1F, whereas the gated mode
checks the level of the trigger signal.
The clock of the slave timer must be enabled prior to receiving events from the master
timer.
RM0090 Rev 19 631/1751

649
Table 98. TIMx internal trigger connection


18.4.4 TIMx DMA/Interrupt enable register (TIMx_DIER)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TDE CC4DE CC3DE CC2DE CC1DE UDE TIE CC4IE CC3IE CC2IE CC1IE UIE
Res. Res Res. Res

Bit 14 TDE: Trigger DMA request enable
0: Trigger DMA request disabled.
1: Trigger DMA request enabled.
Bit 13 Reserved, always read as 0
0: CC4 DMA request disabled.
1: CC4 DMA request enabled.
0: Update DMA request disabled.
1: Update DMA request enabled.
0: Trigger interrupt disabled.
1: Trigger interrupt enabled.
0: CC4 interrupt disabled.
1: CC4 interrupt enabled.
632/1751 RM0090 Rev 19


18.4.5 TIMx status register (TIMx_SR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CC4OF CC3OF CC2OF CC1OF TIF CC4IF CC3IF CC2IF CC1IF UIF
Reserved Reserved Res
rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0

Bit 10 CC2OF: Capture/compare 2 overcapture flag
capture mode. It is cleared by software by writing it to ‘0.
0: No overcapture has been detected
already set
0: No trigger event occurred
1: Trigger interrupt pending
RM0090 Rev 19 633/1751

649

Bit 1 CC1IF: Capture/compare 1 interrupt flag
This flag is set by hardware when the counter matches the compare value, with some
exception in center-aligned mode (refer to the CMS bits in the TIMx_CR1 register
description). It is cleared by software.
0: No match
When the contents of TIMx_CCR1 are greater than the contents of TIMx_ARR, the CC1IF bit
goes high on the counter overflow (in upcounting and up/down-counting modes) or underflow
(in downcounting mode)
TIMx_CCR1 register.
0: No input capture occurred
1: The counter value has been captured in TIMx_CCR1 register (An edge has been detected
on IC1 which matches the selected polarity)
″ This bit is set by hardware on an update event. It is cleared by software.
″ At overflow or underflow (for TIM2 to TIM5) and if UDIS=0 in the TIMx_CR1 register.
″ When CNT is reinitialized by software using the UG bit in TIMx_EGR register, if URS=0
and UDIS=0 in the TIMx_CR1 register.
When CNT is reinitialized by a trigger event (refer to the synchro control register description),
if URS=0 and UDIS=0 in the TIMx_CR1 register.
634/1751 RM0090 Rev 19

18.4.6 TIMx event generation register (TIMx_EGR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TG CC4G CC3G CC2G CC1G UG
Reserved Res.
w w w w w w

hardware.
0: No action
1: The TIF flag is set in TIMx_SR register. Related interrupt or DMA transfer can occur if
enabled.
Bit 4 CC4G: Capture/compare 4 generation
hardware.
0: No action
CC1IF flag is set, Corresponding interrupt or DMA request is sent if enabled.
the corresponding interrupt or DMA request is sent if enabled. The CC1OF flag is set if the
CC1IF flag was already high.
0: No action
1: Re-initialize the counter and generates an update of the registers. Note that the prescaler
counter is cleared too (anyway the prescaler ratio is not affected). The counter is cleared if
the center-aligned mode is selected or if DIR=0 (upcounting), else it takes the auto-reload
value (TIMx_ARR) if DIR=1 (downcounting).
RM0090 Rev 19 635/1751

649
18.4.7 TIMx capture/compare mode register 1 (TIMx_CCMR1)

Reset value: 0x0000
function when the channel is configured in input. Take care that the same bit can have a
different meaning for the input stage and for the output stage.
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OC2CE OC2M[2:0] OC2PE OC2FE OC1CE OC1M[2:0] OC1PE OC1FE
CC2S[1:0] CC1S[1:0]
Output compare mode

an internal trigger input is selected through the TS bit (TIMx_SMCR register)
Note: CC2S bits are writable only when the channel is OFF (CC2E = 0 in TIMx_CCER).
OC1CE: Output compare 1 Clear Enable
0: OC1Ref is not affected by the ETRF input
1: OC1Ref is cleared as soon as a High level is detected on ETRF input
636/1751 RM0090 Rev 19


OC1N are derived. OC1REF is active high whereas OC1 and OC1N active level depends
on CC1P and CC1NP bits.
base).
011: Toggle - OC1REF toggles when TIMx_CNT=TIMx_CCR1.
else inactive. In downcounting, channel 1 is inactive (OC1REF=‘0) as long as
TIMx_CNT>TIMx_CCR1 else active (OC1REF=1).
else active. In downcounting, channel 1 is active as long as TIMx_CNT>TIMx_CCR1 else
inactive.
Note: In PWM mode 1 or 2, the OCREF level changes only when the result of the
to “PWM” mode.
(LOCK bits in TIMx_BDTR register) and CC1S=00 (the channel is configured in
output).
2: The PWM mode can be used without validating the preload register only in one-
pulse mode (OPM bit set in TIMx_CR1 register). Else the behavior is not guaranteed.
5 clock cycles.
1: An active edge on the trigger input acts like a compare match on CC1 output. Then, OC
is set to the compare level independently from the result of the comparison. Delay to sample
the trigger input and to activate CC1 output is reduced to 3 clock cycles. OCFE acts only if
the channel is configured in PWM1 or PWM2 mode.
00: CC1 channel is configured as output.
01: CC1 channel is configured as input, IC1 is mapped on TI1.
RM0090 Rev 19 637/1751

649
Input capture mode

This bit-field defines the frequency used to sample TI1 input and the length of the digital filter
applied to TI1. The digital filter is made of an event counter in which N consecutive events
are needed to validate a transition on the output:
The prescaler is reset as soon as CC1E=0 (TIMx_CCER register).
638/1751 RM0090 Rev 19

18.4.8 TIMx capture/compare mode register 2 (TIMx_CCMR2)

Reset value: 0x0000
Refer to the above CCMR1 register description.
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OC4CE OC4M[2:0] OC4PE OC4FE OC3CE OC3M[2:0] OC3PE OC3FE
CC4S[1:0] CC3S[1:0]
Output compare mode

RM0090 Rev 19 639/1751

649
Input capture mode

18.4.9 TIMx capture/compare enable register (TIMx_CCER)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CC4NP CC4P CC4E CC3NP CC3P CC3E CC2NP CC2P CC2E CC1NP CC1P CC1E
Res. Res. Res. Res.
Bit 15 CC4NP: Capture/Compare 4 output Polarity.

Refer to CC1NP description
Bit 13 CC4P: Capture/Compare 4 output Polarity.
Bit 12 CC4E: Capture/Compare 4 output enable.
640/1751 RM0090 Rev 19


CC1NP must be kept cleared in this case.
This bit is used in conjunction with CC1P to define TI1FP1/TI2FP1 polarity. refer to CC1P
description.
0: OC1 active high
1: OC1 active low
CC1NP/CC1P bits select TI1FP1 and TI2FP1 polarity for trigger or capture operations.
00: noninverted/rising edge
Circuit is sensitive to TIxFP1 rising edge (capture, trigger in reset, external clock or trigger
mode), TIxFP1 is not inverted (trigger in gated mode, encoder mode).
Circuit is sensitive to TIxFP1 falling edge (capture, trigger in reset, external clock or trigger
mode), TIxFP1 is inverted (trigger in gated mode, encoder mode).
11: noninverted/both edges
Circuit is sensitive to both TIxFP1 rising and falling edges (capture, trigger in reset, external
clock or trigger mode), TIxFP1 is not inverted (trigger in gated mode). This configuration
must not be used for encoder mode.
0: Off - OC1 is not active
1: On - OC1 signal is output on the corresponding output pin
0: Capture disabled
1: Capture enabled
Table 99. Output control bit for standard OCx channels

CCxE bit OCx output state
0 Output Disabled (OCx=0, OCx_EN=0)
1 OCx=OCxREF + Polarity, OCx_EN=1
RM0090 Rev 19 641/1751

649
Note: The state of the external IO pins connected to the standard OCx channels depends on the
OCx channel state and the GPIO registers.
18.4.10 TIMx counter (TIMx_CNT)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CNT[31:16] (depending on timers)
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CNT[15:0]
Bits 31:16 CNT[31:16]: High counter value (on TIM2 and TIM5).
18.4.11 TIMx prescaler (TIMx_PSC)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PSC[15:0]

The counter clock frequency CK_CNT is equal to fCK_PSC / (PSC[15:0] + 1).
18.4.12 TIMx auto-reload register (TIMx_ARR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
ARR[31:16] (depending on timers)
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ARR[15:0]
642/1751 RM0090 Rev 19

Bits 31:16 ARR[31:16]: High auto-reload value (on TIM2 and TIM5).
Refer to the Section 18.3.1: Time-base unit for more details about ARR update and
behavior.
18.4.13 TIMx capture/compare register 1 (TIMx_CCR1)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CCR1[31:16] (depending on timers)
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR1[15:0]
Bits 31:16 CCR1[31:16]: High Capture/Compare 1 value (on TIM2 and TIM5).
Bits 15:0 CCR1[15:0]: Low Capture/Compare 1 value
If channel CC1is configured as input:


31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR2[15:0]
RM0090 Rev 19 643/1751

649
Bits 15:0 CCR2[15:0]: Low Capture/Compare 2 value
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR register (bit
OC2PE). Else the preload value is copied in the active capture/compare 2 register when an
update event occurs.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR3[15:0]
Bits 15:0 CCR3[15:0]: Low Capture/Compare value
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR register (bit

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR4[15:0]
644/1751 RM0090 Rev 19

Bits 15:0 CCR4[15:0]: Low Capture/Compare value
1. if CC4 channel is configured as output (CC4S bits):
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR register
(bit OC4PE). Else the preload value is copied in the active capture/compare 4 register
when an update event occurs.
2. if CC4 channel is configured as input (CC4S bits in TIMx_CCMR4 register):
18.4.17 TIMx DMA control register (TIMx_DCR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DBL[4:0] DBA[4:0]
Reserved Reserved

Bits 12:8 DBL[4:0]: DMA burst length
This 5-bit vector defines the number of DMA transfers (the timer recognizes a burst transfer
when a read or a write access is done to the TIMx_DMAR address).
00000: 1 transfer,
00001: 2 transfers,
00010: 3 transfers,
...
10001: 18 transfers.
Bits 4:0 DBA[4:0]: DMA base address
This 5-bit vector defines the base-address for DMA transfers (when read/write access are
done through the TIMx_DMAR address). DBA is defined as an offset starting from the
address of the TIMx_CR1 register.
Example:
00000: TIMx_CR1,
00001: TIMx_CR2,
00010: TIMx_SMCR,
...
Example: Let us consider the following transfer: DBL = 7 transfers & DBA = TIMx_CR1. In this
case the transfer is done to/from 7 registers starting from the TIMx_CR1 address.
RM0090 Rev 19 645/1751

649
18.4.18 TIMx DMA address for full transfer (TIMx_DMAR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DMAB[15:0]
Bits 15:0 DMAB[15:0]: DMA register for burst accesses

A read or write operation to the DMAR register accesses the register located at the address
(TIMx_CR1 address) + (DBA + DMA index) x 4
where TIMx_CR1 address is the address of the control register 1, DBA is the DMA base
address configured in TIMx_DCR register, DMA index is automatically controlled by the
DMA transfer, and ranges from 0 to DBL (DBL configured in TIMx_DCR).
Example of how to use the DMA burst feature

In this example the timer DMA burst feature is used to update the contents of the CCRx
registers (x = 2, 3, 4) with the DMA transferring half words into the CCRx registers.
This is done in the following steps:
1. Configure the corresponding DMA channel as follows:
– DMA channel peripheral address is the DMAR register address
– DMA channel memory address is the address of the buffer in the RAM containing
the data to be transferred by DMA into CCRx registers.
– Number of data to transfer = 3 (See note below).
– Circular mode disabled.
2. Configure the DCR register by configuring the DBA and DBL bit fields as follows:
DBL = 3 transfers, DBA = 0xE.
3. Enable the TIMx update DMA request (set the UDE bit in the DIER register).
4. Enable TIMx
5. Enable the DMA channel
Note: This example is for the case where every CCRx register to be updated once. If every CCRx
register is to be updated twice for example, the number of data to transfer should be 6. Let's
take the example of a buffer in the RAM containing data1, data2, data3, data4, data5 and
data6. The data is transferred to the CCRx registers as follows: on the first update DMA
request, data1 is transferred to CCR2, data2 is transferred to CCR3, data3 is transferred to
CCR4 and on the second update DMA request, data4 is transferred to CCR2, data5 is
transferred to CCR3 and data6 is transferred to CCR4.
646/1751 RM0090 Rev 19

18.4.19 TIM2 option register (TIM2_OR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ITR1_RMP
Reserved Reserved
rw rw

Bits 11:10 ITR1_RMP: Internal trigger 1 remap
00: TIM8_TRGOUT
01: PTP trigger output is connected to TIM2_ITR1
10: OTG FS SOF is connected to the TIM2_ITR1 input
11: OTG HS SOF is connected to the TIM2_ITR1 input
18.4.20 TIM5 option register (TIM5_OR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TI4_RMP
Reserved Reserved
rw rw

Bits 7:6 TI4_RMP: Timer Input 4 remap
00: TIM5 Channel4 is connected to the GPIO: Refer to the Alternate function mapping table
in the datasheets.
01: the LSI internal clock is connected to the TIM5_CH4 input for calibration purposes
10: the LSE internal clock is connected to the TIM5_CH4 input for calibration purposes
11: the RTC wakeup interrupt is connected to TIM5_CH4 input for calibration purposes.
Wakeup interrupt should be enabled.
RM0090 Rev 19 647/1751

649
18.4.21 TIMx register map

TIMx registers are mapped as described in the table below:
Table 100. TIM2 to TIM5 register map and reset values

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
CKD CMS
ARPE
UDIS
OPM
CEN
URS
DIR
TIMx_CR1
0x00 Reserved [1:0] [1:0]
Reset value 0 0 0 0 0 0 0 0 0 0
CCDS
TI1S
TIMx_CR2 MMS[2:0]
ETPS
Reserved
MSM
ECE
ETP
TIMx_SMCR ETF[3:0] TS[2:0] SMS[2:0]
0x08 Reserved [1:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
COMDE
CC4DE
CC3DE
CC2DE
CC1DE
CC4IE
CC3IE
CC2IE
CC1IE
Reserved
Reserved
UDE
TDE
UIE
TIMx_DIER
TIE
0x0C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0
CC4OF
CC3OF
CC2OF
CC1OF
CC4IF
CC3IF
CC2IF
CC1IF
Reserved
Reserved
TIMx_SR
UIF
TIF
0x10 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0
CC4G
CC3G
CC2G
CC1G
Reserved
TIMx_EGR
UG
TG
0x14 Reserved
TIMx_CCMR1
OC2CE
OC1CE
OC2PE
OC1PE
OC2FE
OC1FE
OC2M CC2S OC1M CC1S
Output Compare
Reserved [2:0] [1:0] [2:0] [1:0]
mode
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x18
TIMx_CCMR1 IC2 IC1
CC2S CC1S
Input Capture IC2F[3:0] PSC IC1F[3:0] PSC
Reserved [1:0] [1:0]
mode [1:0] [1:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
TIMx_CCMR2
OC4CE
OC3CE
OC4PE
OC3PE
OC4FE
OC3FE
OC4M CC4S OC3M CC3S
Output Compare
Reserved [2:0] [1:0] [2:0] [1:0]
mode
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x1C
TIMx_CCMR2 IC4 IC3
CC4S CC3S
Input Capture IC4F[3:0] PSC IC3F[3:0] PSC
mode [1:0] [1:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Reserved
Reserved
Reserved
Reserved
CC4NP
CC3NP
CC2NP
CC1NP
CC4P
CC4E
CC3P
CC3E
CC2P
CC2E
CC1P
CC1E
TIMx_CCER
0x20 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0
CNT[31:16]
TIMx_CNT CNT[15:0]
0x24 (TIM2 and TIM5 only, reserved on the other timers)
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
TIMx_PSC PSC[15:0]
0x28 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
648/1751 RM0090 Rev 19

Table 100. TIM2 to TIM5 register map and reset values (continued)
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
ARR[31:16]
TIMx_ARR ARR[15:0]
0x2C (TIM2 and TIM5 only, reserved on the other timers)
Reset value 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
0x30 Reserved
CCR1[31:16]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CCR2[31:16]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CCR3[31:16]
0x3C (TIM2 and TIM5 only, reserved on the other timers)
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CCR4[31:16]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x44 Reserved
TIMx_DCR DBL[4:0] DBA[4:0]

Reset value 0 0 0 0 0 0 0 0 0 0
TIMx_DMAR DMAB[15:0]
0x4C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ITR1_
TIM2_OR
0x50 Reserved Reserved RMP Reserved
Reset value 0 0
IT4_
TIM5_OR
0x50 Reserved Reserved RMP Reserved
Reset value 0 0
RM0090 Rev 19 649/1751

649
19 General-purpose timers (TIM9 to TIM14)
19.1 TIM9 to TIM14 introduction

The TIM9 to TIM14 general-purpose timers consist of a 16-bit auto-reload counter driven by
a programmable prescaler.
They may be used for a variety of purposes, including measuring the pulse lengths of input
signals (input capture) or generating output waveforms (output compare, PWM).
The TIM9 to TIM14 timers are completely independent, and do not share any resources.
They can be synchronized together as described in Section 19.3.12.
19.2 TIM9 to TIM14 main features
19.2.1 TIM9/TIM12 main features

The features of the TIM9 to TIM14 general-purpose timers include:
• 16-bit auto-reload upcounter
• 16-bit programmable prescaler used to divide the counter clock frequency by any factor
between 1 and 65536 (can be changed “on the fly”)
– Input capture
– Output compare
– PWM generation (edge-aligned mode)
several timers together
• Interrupt generation on the following events:
– Update: counter overflow, counter initialization (by software or internal trigger)
– Trigger event (counter start, stop, initialization or count by internal trigger)
– Input capture
– Output compare
650/1751 RM0090 Rev 19

Figure 180. General-purpose timer block diagram (TIM9 and TIM12)

Trigger
controller
ITR0 TGI
ITR1 ITR Slave
ITR2 TRC TRGI controller Reset, enable, up, count
ITR3 mode
TI1F_ED
TI1FP1
TI2FP2
U
Auto-reload register UI
Stop, Clear
U
CK_PSC PSC CK_CNT +/- CNT counter
prescaler
CC1I U CC1I
TI1FP1 OC1REF OC1
TI1 Input filter & TI1FP2 IC1 IC1PS Capture/Compare 1 register Output TIMx_CH1
TIMx_CH1 Prescaler
TRC
CC2I
U CC2I
TI2FP1
TRC
Notes:
Preload registers transferred
Reg
Event
Interrupt ai17190b
19.2.2 TIM10/TIM11 and TIM13/TIM14 main features

The features of general-purpose timers TIM10/TIM11 and TIM13/TIM14 include:
• 16-bit programmable prescaler used to divide the counter clock frequency by any factor
between 1 and 65536 (can be changed “on the fly”)
• independent channel for:
– Input capture
– Output compare
– PWM generation (edge-aligned mode)
• Interrupt generation on the following events:
– Update: counter overflow, counter initialization (by software)
– Input capture
– Output compare
RM0090 Rev 19 651/1751

695
Figure 181. General-purpose timer block diagram (TIM10/11/13/14)

Enable
Trigger counter
Controller
U Autoreload register UI
Stop, Clear
U
+/-
prescaler counter
CC1I CC1I
TI1 TI1FP1 IC1 U
Input filter & IC1PS OC1REF output OC1
TIMx_CH1 Prescaler Capture/Compare 1 register TIMx_CH1
Notes:
Reg Preload registers transferred

event
interrupt & DMA output
ai17725c
652/1751 RM0090 Rev 19

19.3 TIM9 to TIM14 functional description

The main block of the timer is a 16-bit counter with its related auto-reload register. The
counters counts up.
The counter clock can be divided by a prescaler.
• Auto-reload register (TIMx_ARR)
reaches the overflow and if the UDIS bit equals 0 in the TIMx_CR1 register. It can also be
generated by software. The generation of the update event is described in details for each
configuration.
Note that the counter starts counting 1 clock cycle after setting the CEN bit in the TIMx_CR1
register.
It can be changed on the fly as this control register is buffered. The new prescaler ratio is
taken into account at the next update event.
ratio is changed on the fly.
RM0090 Rev 19 653/1751

695
CK_PSC
CEN
Timerclock = CK_CNT
Update event (UEV)
MS31076V2
CK_PSC
CEN
Timerclock = CK_CNT
F7 F8 F9 FA FB FC 00 01
Counter register
Update event (UEV)
0 3
Prescaler control register
0 3
Prescaler buffer
0 0 1 2 3 0 1 2 3
Prescaler counter
MS31077V2
654/1751 RM0090 Rev 19


Upcounting mode
controller on TIM9 and TIM12) also generates an update event.
However, the counter restarts from 0, as well as the counter of the prescaler (but the
prescale rate does not change). In addition, if the URS bit (update request selection) in
TIMx_CR1 register is set, setting the UG bit generates an update event UEV but without
setting the UIF flag (thus no interrupt is sent). This is to avoid generating both update and
capture interrupts when clearing the counter on the capture event.
• The auto-reload shadow register is updated with the preload value (TIMx_ARR),
register).
CK_PSC
CNT_EN
Timerclock = CK_CNT
Counter overflow
Update event (UEV)

MS31078V3
RM0090 Rev 19 655/1751

695
CK_PSC
CNT_EN
Timerclock = CK_CNT
Counter overflow
Update event (UEV)

MS31079V3
CK_PSC
CNT_EN
Timerclock = CK_CNT
Counter overflow
Update event (UEV)

MS31080V3
CK_PSC
Timerclock = CK_CNT
Counter overflow
Update event (UEV)

MS31081V3
656/1751 RM0090 Rev 19

preloaded)
CK_PSC
CEN
Timerclock = CK_CNT
Counter overflow
Update event (UEV)

MS31082V3
preloaded)
CK_PSC
CEN
Timerclock = CK_CNT
Counter overflow
Update event (UEV)

(UIF)
Auto-reload preload
F5 36
register
Auto-reload shadow
register F5 36
Write a new value in TIMx_ARR MS31083V2
RM0090 Rev 19 657/1751

695

• External clock mode1 (for TIM9 and TIM12): external input pin (TIx)
• Internal trigger inputs (ITRx) (for TIM9 and TIM12): connecting the trigger output from
another timer. Refer to Using one timer as prescaler for another timer for more details.

The internal clock source is the default clock source for TIM10/TIM11 and TIM13/TIM14.
For TIM9 and TIM12, the internal clock source is selected when the slave mode controller is
disabled (SMS=’000’). The CEN bit in the TIMx_CR1 register and the UG bit in the
TIMx_EGR register are then used as control bits and can be changed only by software
(except for UG which remains cleared). As soon as the CEN bit is programmed to 1, the
prescaler is clocked by the internal clock CK_INT.
Figure 190 shows the behavior of the control circuit and of the upcounter in normal mode,
without prescaler.
Internal clock
CEN=CNT_EN
UG
CNT_INIT
MS31085V2
External clock source mode 1(TIM9 and TIM12)

This mode is selected when SMS=’111’ in the TIMx_SMCR register. The counter can count
at each rising or falling edge on a selected input.
658/1751 RM0090 Rev 19


TIMx_SMCR
TS[2:0]
or TI2F or
TI1F or
ITRx
0xx
TI1_ED
100 TRGI External clock CK_PSC
TI1FP1 mode 1
TI2F_Rising 101
TI2 Edge 0 TI2FP2
Filter 110
detector 1
TI2F_Falling
mode
SMS[2:0]
TIMx_SMCR
MS37337V1
1. Configure channel 2 to detect rising edges on the TI2 input by writing CC2S = ‘01’ in
register (if no filter is needed, keep IC2F=’0000’).
3. Select the rising edge polarity by writing CC2P=’0’ and CC2NP=’0’ in the TIMx_CCER
register.
4. Configure the timer in external clock mode 1 by writing SMS=’111’ in the TIMx_SMCR
register.
5. Select TI2 as the trigger input source by writing TS=’110’ in the TIMx_SMCR register.
6. Enable the counter by writing CEN=’1’ in the TIMx_CR1 register.
Note: The capture prescaler is not used for triggering, so no need to configure it.
TI2
CNT_EN
TIF
Write TIF = 0
MS31087V3
RM0090 Rev 19 659/1751

695

Each Capture/Compare channel is built around a capture/compare register (including a
shadow register), a input stage for capture (with digital filter, multiplexing and prescaler) and
an output stage (with comparator and output control).
Figure 193 to Figure 195 give an overview of a capture/compare channel.
TI1F_ED
To the slave mode controller
TI1 TI1F_Rising
Filter TI1F Edge 0 TI1FP1
fDTS downcounter TI1F_Falling 01
detector 1
TI2FP1 IC1 Divider IC1PS
10
/1, /2, /4, /8
11
TI2F_Rising controller)
0
(from channel 2)
TI2F_Falling
1
MS33115V1
The output stage generates an intermediate waveform which is then used for reference:
660/1751 RM0090 Rev 19

APB Bus
(if 16-bit)
8 8
high
S write CCR1H
low
R Output CC1S[1]
CC1S[0]
CC1S[1] Input
mode OC1PE
UEV
(from time TIMx_CCMR1
Comparator
IC1PS Capture base unit)
CC1E CNT>CCR1
Counter CNT=CCR1
CC1G
TIMx_EGR
MS31089V3
To the master
mode controller
CNT > CCR2

Output 0 Output
OC1_REF OC1
CNT = CCR2 mode enable
controller 1 circuit
CC1P
OC2M[2:0] TIMx_CCER CC1E TIMx_CCER
TIMx_CCMR1
ai17720
preload register.

RM0090 Rev 19 661/1751

695
cleared by software by writing it to ‘0’ or by reading the captured data stored in the
TIMx_CCRx register. CCxOF is cleared when the user writes it to ‘0’.
1. Select the active input: TIMx_CCR1 must be linked to the TI1 input, so write the CC1S
bits to ‘01’ in the TIMx_CCMR1 register. As soon as CC1S becomes different from ‘00’,
the channel is configured in input mode and the TIMx_CCR1 register becomes read-
only.
2. Program the needed input filter duration with respect to the signal connected to the
timer (by programming the ICxF bits in the TIMx_CCMRx register if the input is one of
the TIx inputs). Let’s imagine that, when toggling, the input signal is not stable during at
must 5 internal clock cycles. We must program a filter duration longer than these 5
clock cycles. We can validate a transition on TI1 when 8 consecutive samples with the
new level have been detected (sampled at fDTS frequency). Then write IC1F bits to
‘0011’ in the TIMx_CCMR1 register.
3. Select the edge of the active transition on the TI1 channel by programming CC1P and
CC1NP bits to ‘00’ in the TIMx_CCER register (rising edge in this case).
4. Program the input prescaler. In our example, we wish the capture to be performed at
each valid transition, so the prescaler is disabled (write IC1PS bits to ‘00’ in the
5. Enable capture from the counter into the capture register by setting the CC1E bit in the
TIMx_CCER register.
6. If needed, enable the related interrupt request by setting the CC1IE bit in the
TIMx_DIER register.
Note: IC interrupt requests can be generated by software by setting the corresponding CCxG bit in
the TIMx_EGR register.
662/1751 RM0090 Rev 19

19.3.6 PWM input mode (only for TIM9/12)

For example, the user can measure the period (in TIMx_CCR1 register) and the duty cycle
(in TIMx_CCR2 register) of the PWM applied on TI1 using the following procedure
(depending on CK_INT frequency and prescaler value):
1. Select the active input for TIMx_CCR1: write the CC1S bits to ‘01’ in the TIMx_CCMR1
2. Select the active polarity for TI1FP1 (used both for capture in TIMx_CCR1 and counter
clear): program the CC1P and CC1NP bits to ‘00’ (active on rising edge).
3. Select the active input for TIMx_CCR2: write the CC2S bits to ‘10’ in the TIMx_CCMR1
4. Select the active polarity for TI1FP2 (used for capture in TIMx_CCR2): program the
CC2P and CC2NP bits to ‘11’ (active on falling edge).
5. Select the valid trigger input: write the TS bits to ‘101’ in the TIMx_SMCR register
(TI1FP1 selected).
6. Configure the slave mode controller in reset mode: write the SMS bits to ‘100’ in the
TIMx_SMCR register.
7. Enable the captures: write the CC1E and CC2E bits to ‘1’ in the TIMx_CCER register.
TI1
TIMx_CNT 0004 0000 0001 0002 0003 0004 0000
TIMx_CCR1 0004
TIMx_CCR2 0002
IC1 capture IC2 capture IC1 capture

pu
l eswid th period
IC2 capture
measurement measurement
reset counter
ai15413
1. The PWM input mode can be used only with the TIMx_CH1/TIMx_CH2 signals due to the fact that only
TI1FP1 and TI2FP2 are connected to the slave mode controller.
RM0090 Rev 19 663/1751

695

In output mode (CCxS bits = ‘00’ in the TIMx_CCMRx register), each output compare signal
(OCxREF and then OCx) can be forced to active or inactive level directly by software,
To force an output compare signal (OCXREF/OCx) to its active level, the user just needs to
write ‘101’ in the OCxM bits in the corresponding TIMx_CCMRx register. Thus OCXREF is
bit.
For example: CCxP=’0’ (OCx active high) => OCx is forced to high level.
The OCxREF signal can be forced low by writing the OCxM bits to ‘100’ in the
TIMx_CCMRx register.
performed and allows the flag to be set. Interrupt requests can be sent accordingly. This is
described in the output compare mode section below.

elapsed.
compare function:
1. Assigns the corresponding output pin to a programmable value defined by the output
bit in the TIMx_CCER register). The output pin can keep its level (OCXM=’000’), be set
active (OCxM=’001’), be set inactive (OCxM=’010’) or can toggle (OCxM=’011’) on
match.
2. Sets a flag in the interrupt status register (CCxIF bit in the TIMx_SR register).
3. Generates an interrupt if the corresponding interrupt mask is set (CCXIE bit in the
In output compare mode, the update event UEV has no effect on OCxREF and OCx output.
output a single pulse (in One-pulse mode).
Procedure:
3. Set the CCxIE bit if an interrupt request is to be generated.
4. Select the output mode. For example:
– Write OCxM = ‘011’ to toggle OCx output pin when CNT matches CCRx
– Write OCxPE = ‘0’ to disable preload register
– Write CCxP = ‘0’ to select active high polarity
– Write CCxE = ‘1’ to enable the output
664/1751 RM0090 Rev 19

waveform, provided that the preload register is not enabled (OCxPE=’0’, else TIMx_CCRx
Figure 197.
Figure 197. Output compare mode, toggle on OC1.

TIM1_CNT 0039 003A 003B B200 B201
TIM1_CCR1 003A B201
OC1REF= OC1

MS31092V2
19.3.9 PWM mode

Pulse Width Modulation mode allows the user to generate a signal with a frequency
determined by the value of the TIMx_ARR register and a duty cycle determined by the value
of the TIMx_CCRx register.
output) by writing ‘110’ (PWM mode 1) or ‘111’ (PWM mode 2) in the OCxM bits in the
TIMx_CCMRx register. Enable the corresponding preload register by setting the OCxPE bit
in the TIMx_CCMRx register, and eventually the auto-reload preload register by setting the
ARPE bit in the TIMx_CR1 register.
occurs, before starting the counter, the user has to initialize all the registers by setting the
UG bit in the TIMx_EGR register.
The OCx polarity is software programmable using the CCxP bit in the TIMx_CCER register.
It can be programmed as active high or active low. The OCx output is enabled by the CCxE
bit in the TIMx_CCER register. Refer to the TIMx_CCERx register description for more
details.
whether TIMx_CNT ≤ TIMx_CCRx.
The timer is able to generate PWM in edge-aligned mode only since the counter is
upcounting.

In the following example, we consider PWM mode 1. The reference PWM signal OCxREF is
high as long as TIMx_CNT < TIMx_CCRx else it becomes low. If the compare value in
RM0090 Rev 19 665/1751

695
TIMx_CCRx is greater than the auto-reload value (in TIMx_ARR) then OCxREF is held at
‘1’. If the compare value is 0 then OCxRef is held at ‘0’. Figure 198 shows some edge-
aligned PWM waveforms in an example where TIMx_ARR=8.
OCXREF
CCRx=4
CCxIF
OCXREF
CCRx=8
CCxIF
OCXREF ‘1’
CCRx>8
CCxIF
OCXREF ‘0’
CCRx=0
CCxIF
MS31093V1

be as follows:
CNT < CCRx≤ ARR (in particular, 0 < CCRx)
666/1751 RM0090 Rev 19

Figure 199. Example of one pulse mode.
TI2
OC1REF
OC1
TIM1_ARR
Counter
TIM1_CCR1
0
tDELAY tPULSE t
MS31099V1
Use TI2FP2 as trigger 1:
1. Map TI2FP2 to TI2 by writing CC2S=’01’ in the TIMx_CCMR1 register.
2. TI2FP2 must detect a rising edge, write CC2P=’0’ and CC2NP = ‘0’ in the TIMx_CCER
register.
3. Configure TI2FP2 as trigger for the slave mode controller (TRGI) by writing TS=’110’ in
4. TI2FP2 is used to start the counter by writing SMS to ‘110’ in the TIMx_SMCR register
(trigger mode).
value (TIMx_ARR - TIMx_CCR1).
• Let us say the user wants to build a waveform with a transition from ‘0’ to ‘1’ when a
compare match occurs and a transition from ‘1’ to ‘0’ when the counter reaches the
auto-reload value. To do this enable PWM mode 2 by writing OC1M=’111’ in the
OC1PE=’1’ in the TIMx_CCMR1 register and ARPE in the TIMx_CR1 register. In this
case the user has to write the compare value in the TIMx_CCR1 register, the auto-
reload value in the TIMx_ARR register, generate an update by setting the UG bit and
wait for external trigger event on TI2. CC1P is written to ‘0’ in this example.
The user only wants one pulse (Single mode), so write '1 in the OPM bit in the TIMx_CR1
register to stop the counter at the next update event (when the counter rolls over from the
auto-reload value back to 0). When OPM bit in the TIMx_CR1 register is set to '0', so the
Repetitive mode is selected.
RM0090 Rev 19 667/1751

695
Particular case: OCx fast enable

If the user wants to output a waveform with the minimum delay, set the OCxFE bit in the
TIMx_CCMRx register. Then OCxRef (and OCx) are forced in response to the stimulus,
without taking in account the comparison. Its new level is the same as if a compare match
had occurred. OCxFE acts only if the channel is configured in PWM1 or PWM2 mode.
19.3.11 TIM9/12 external trigger synchronization

The TIM9/12 timers can be synchronized with an external trigger in several modes: Reset
mode, Gated mode and Trigger mode.

1. Configure the channel 1 to detect rising edges on TI1. Configure the input filter duration
(in this example, we don’t need any filter, so we keep IC1F=’0000’). The capture
select the input capture source only, CC1S = ‘01’ in the TIMx_CCMR1 register.
Program CC1P and CC1NP to ‘00’ in TIMx_CCER register to validate the polarity (and
detect rising edges only).
2. Configure the timer in reset mode by writing SMS=’100’ in TIMx_SMCR register. Select
TI1 as the input source by writing TS=’101’ in TIMx_SMCR register.
3. Start the counter by writing CEN=’1’ in the TIMx_CR1 register.
trigger flag is set (TIF bit in the TIMx_SR register) and an interrupt request can be sent if
enabled (depending on the TIE bit in TIMx_DIER register).
The following figure shows this behavior when the auto-reload register TIMx_ARR=0x36.
The delay between the rising edge on TI1 and the actual reset of the counter is due to the
668/1751 RM0090 Rev 19

TI1
UG
TIF
MS31401V2

1. Configure the channel 1 to detect low levels on TI1. Configure the input filter duration
select the input capture source only, CC1S=’01’ in TIMx_CCMR1 register. Program
CC1P=’1’ and CC1NP= ‘0’ in TIMx_CCER register to validate the polarity (and detect
low level only).
2. Configure the timer in gated mode by writing SMS=’101’ in TIMx_SMCR register.
Select TI1 as the input source by writing TS=’101’ in TIMx_SMCR register.
3. Enable the counter by writing CEN=’1’ in the TIMx_CR1 register (in gated mode, the
counter doesn’t start if CEN=’0’, whatever is the trigger input level).
or stops.
RM0090 Rev 19 669/1751

695
TI1
cnt_en
TIF
Write TIF=0
MS31402V1

1. Configure the channel 2 to detect rising edges on TI2. Configure the input filter duration
are configured to select the input capture source only, CC2S=’01’ in TIMx_CCMR1
register. Program CC2P=’1’ and CC2NP=’0’ in TIMx_CCER register to validate the
polarity (and detect low level only).
2. Configure the timer in trigger mode by writing SMS=’110’ in TIMx_SMCR register.
Select TI2 as the input source by writing TS=’110’ in TIMx_SMCR register.
TIF flag is set.
TI2
cnt_en
TIF
MS31403V1
670/1751 RM0090 Rev 19

19.3.12 Timer synchronization (TIM9/12)

The TIM timers are linked together internally for timer synchronization or chaining. Refer to
Section 18.3.15: Timer synchronization for details.
Note: The clock of the slave timer must be enabled prior to receive events from the master timer,
19.3.13 Debug mode

When the microcontroller enters debug mode (Cortex®-M4 with FPU core halted), the TIMx
counter either continues to work normally or stops, depending on DBG_TIMx_STOP
RM0090 Rev 19 671/1751

695

The peripheral registers have to be written by half-words (16 bits) or words (32 bits). Read
accesses can be done by bytes (8 bits), half-words (16 bits) or words (32 bits).
19.4.1 TIM9/12 control register 1 (TIMx_CR1)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CKD[1:0] ARPE OPM URS UDIS CEN
Reserved Reserved

sampling clock used by the digital filters (TIx),
00: tDTS = tCK_INT
11: Reserved
0: TIMx_ARR register is not buffered.
1: TIMx_ARR register is buffered.
0: Counter is not stopped on the update event
1: Counter stops counting on the next update event (clearing the CEN bit).
0: Any of the following events generates an update interrupt if enabled:
– Counter overflow
1: Only counter overflow generates an update interrupt if enabled.
This bit is set and cleared by software to enable/disable update event (UEV) generation.
0: UEV enabled. An UEV is generated by one of the following events:
1: UEV disabled. No UEV is generated, shadow registers keep their value (ARR, PSC,
CCRx). The counter and the prescaler are reinitialized if the UG bit is set.
0: Counter disabled
1: Counter enabled
672/1751 RM0090 Rev 19

19.4.2 TIM9/12 slave mode control register (TIMx_SMCR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MSM TS[2:0] SMS[2:0]
Reserved Res.

Bit 7 MSM: Master/Slave mode
0: No action
synchronization between the current timer and its slaves (through TRGO). It is useful in
order to synchronize several timers on a single external event.
Bits 6:4 TS: Trigger selection
This bit field selects the trigger input to be used to synchronize the counter.
111: Reserved.
See Table 101 for more details on the meaning of ITRx for each timer.
Note: These bits must be changed only when they are not used (e.g. when SMS=’000’) to
RM0090 Rev 19 673/1751

695

When external signals are selected, the active edge of the trigger signal (TRGI) is linked to
the polarity selected on the external input (see Input control register and Control register
descriptions.
000: Slave mode disabled - if CEN = 1 then the prescaler is clocked directly by the internal
clock
001: Reserved
010: Reserved
011: Reserved
100: Reset mode - Rising edge of the selected trigger input (TRGI) reinitializes the counter
and generates an update of the registers
101: Gated mode - The counter clock is enabled when the trigger input (TRGI) is high. The
counter stops (but is not reset) as soon as the trigger becomes low. Counter starts and stops
are both controlled
110: Trigger mode - The counter starts on a rising edge of the trigger TRGI (but it is not
reset). Only the start of the counter is controlled
111: External clock mode 1 - Rising edges of the selected trigger (TRGI) clock the counter
Note: The Gated mode must not be used if TI1F_ED is selected as the trigger input
(TS=’100’). Indeed, TI1F_ED outputs 1 pulse for each transition on TI1F, whereas the
Gated mode checks the level of the trigger signal.
Note: The clock of the slave timer must be enabled prior to receive events from the master
timer.
Table 101. TIMx internal trigger connection

TIM9 TIM2_TRGO TIM3_TRGO TIM10_OC TIM11_OC

TIM12 TIM4_TRGO TIM5_TRGO TIM13_OC TIM14_OC
19.4.3 TIM9/12 Interrupt enable register (TIMx_DIER)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TIE CC2IE CC1IE UIE
Reserved Res
rw rw rw rw

0: Trigger interrupt disabled.
1: Trigger interrupt enabled.
674/1751 RM0090 Rev 19


0: Update interrupt disabled.
1: Update interrupt enabled.
RM0090 Rev 19 675/1751

695
19.4.4 TIM9/12 status register (TIMx_SR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CC2OF CC1OF TIF CC2IF CC1IF UIF

Bit 10 CC2OF: Capture/compare 2 overcapture flag
already set
0: No trigger event occurred.
1: Trigger interrupt pending.
This flag is set by hardware when the counter matches the compare value. It is cleared by
software.
0: No match.
When the contents of TIMx_CCR1 are greater than the contents of TIMx_ARR, the CC1IF
bit goes high on the counter overflow.
TIMx_CCR1 register.
0: No input capture occurred.
1: The counter value has been captured in TIMx_CCR1 register (an edge has been detected
on IC1 which matches the selected polarity).
676/1751 RM0090 Rev 19


– At overflow and if UDIS=’0’ in the TIMx_CR1 register.
– When CNT is reinitialized by software using the UG bit in TIMx_EGR register, if URS=’0’ and
UDIS=’0’ in the TIMx_CR1 register.
– When CNT is reinitialized by a trigger event (refer to the synchro control register
description), if URS=’0’ and UDIS=’0’ in the TIMx_CR1 register.
19.4.5 TIM9/12 event generation register (TIMx_EGR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TG CC2G CC1G UG
Reserved Reserved
w w w w

hardware.
0: No action
1: The TIF flag is set in the TIMx_SR register. Related interrupt can occur if enabled
This bit is set by software to generate an event, it is automatically cleared by hardware.
0: No action
the CC1IF flag is set, the corresponding interrupt is sent if enabled.
The current counter value is captured in the TIMx_CCR1 register. The CC1IF flag is set, the
corresponding interrupt is sent if enabled. The CC1OF flag is set if the CC1IF flag was
already high.
0: No action
1: Re-initializes the counter and generates an update of the registers. The prescaler counter
is also cleared and the prescaler ratio is not affected. The counter is cleared.
RM0090 Rev 19 677/1751

695
19.4.6 TIM9/12 capture/compare mode register 1 (TIMx_CCMR1)

Reset value: 0x0000
bits in this register have different functions in input and output modes. For a given bit, OCxx
describes its function when the channel is configured in output mode, ICxx describes its
function when the channel is configured in input mode. Take care that the same bit can have
different meanings for the input stage and the output stage.
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. OC2M[2:0] OC2PE OC2FE Res. OC1M[2:0] OC1PE OC1FE
CC2S[1:0] CC1S[1:0]
Output compare mode

This bitfield defines the direction of the channel (input/output) as well as the used input.
11: CC2 channel is configured as input, IC2 is mapped on TRC. This mode works only if an
internal trigger input is selected through the TS bit (TIMx_SMCR register
Note: The CC2S bits are writable only when the channel is OFF (CC2E = 0 in TIMx_CCER).
678/1751 RM0090 Rev 19


OC1N are derived. OC1REF is active high whereas the active levels of OC1 and OC1N
depend on the CC1P and CC1NP bits, respectively.
base).
001: Set channel 1 to active level on match. The OC1REF signal is forced high when the
TIMx_CNT counter matches the capture/compare register 1 (TIMx_CCR1).
010: Set channel 1 to inactive level on match. The OC1REF signal is forced low when the
TIMx_CNT counter matches the capture/compare register 1 (TIMx_CCR1).
011: Toggle - OC1REF toggles when TIMx_CNT=TIMx_CCR1
100: Force inactive level - OC1REF is forced low
101: Force active level - OC1REF is forced high
else it is inactive. In downcounting, channel 1 is inactive (OC1REF=‘0) as long as
TIMx_CNT>TIMx_CCR1, else it is active (OC1REF=’1’)
else it is active. In downcounting, channel 1 is active as long as TIMx_CNT>TIMx_CCR1
else it is inactive.
Note: In PWM mode 1 or 2, the OCREF level changes only when the result of the
to “PWM” mode.
new value is taken into account immediately
register. TIMx_CCR1 preload value is loaded into the active register at each update event
Note: The PWM mode can be used without validating the preload register only in one-pulse
mode (OPM bit set in the TIMx_CR1 register). Else the behavior is not guaranteed.
0: CC1 behaves normally depending on the counter and CCR1 values even when the
trigger is ON. The minimum delay to activate the CC1 output when an edge occurs on the
trigger input is 5 clock cycles
1: An active edge on the trigger input acts like a compare match on the CC1 output. Then,
OC is set to the compare level independently of the result of the comparison. Delay to
sample the trigger input and to activate CC1 output is reduced to 3 clock cycles. OC1FE
acts only if the channel is configured in PWM1 or PWM2 mode.
internal trigger input is selected through the TS bit (TIMx_SMCR register)
RM0090 Rev 19 679/1751

695
Input capture mode

internal trigger input is selected through the TS bit (TIMx_SMCR register)
This bitfield defines the frequency used to sample the TI1 input and the length of the digital
filter applied to TI1. The digital filter is made of an event counter in which N consecutive
events are needed to validate a transition on the output:
This bitfield defines the ratio of the prescaler acting on the CC1 input (IC1).
680/1751 RM0090 Rev 19

19.4.7 TIM9/12 capture/compare enable register (TIMx_CCER)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CC2NP CC2P CC2E CC1NP CC1P CC1E
Reserved Res. Res.
rw rw rw rw rw rw

Bit 7 CC2NP: Capture/Compare 2 output Polarity
Bit 5 CC2P: Capture/Compare 2 output Polarity
Bit 3 CC1NP: Capture/Compare 1 complementary output Polarity
CC1 channel configured as output: CC1NP must be kept cleared
CC1 channel configured as input: CC1NP is used in conjunction with CC1P to define
TI1FP1/TI2FP1 polarity (refer to CC1P description).
0: OC1 active high.
1: OC1 active low.
CC1NP/CC1P bits select TI1FP1 and TI2FP1 polarity for trigger or capture operations.
Circuit is sensitive to TIxFP1 rising edge (capture, trigger in reset, external clock or trigger
mode), TIxFP1 is not inverted (trigger in gated mode, encoder mode).
Circuit is sensitive to TIxFP1 falling edge (capture, trigger in reset, external clock or trigger
mode), TIxFP1 is inverted (trigger in gated mode, encoder mode).
Note: 11: noninverted/both edges
Circuit is sensitive to both TIxFP1 rising and falling edges (capture, trigger in reset,
external clock or trigger mode), TIxFP1 is not inverted (trigger in gated mode). This
configuration must not be used for encoder mode.
0: Off - OC1 is not active.
1: On - OC1 signal is output on the corresponding output pin.
1: Capture enabled.
RM0090 Rev 19 681/1751

695

0 Output disabled (OCx=’0’, OCx_EN=’0’)

1 OCx=OCxREF + Polarity, OCx_EN=’1’
Note: The states of the external I/O pins connected to the standard OCx channels depend on the
state of the OCx channel and on the GPIO registers.
19.4.8 TIM9/12 counter (TIMx_CNT)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CNT[15:0]
19.4.9 TIM9/12 prescaler (TIMx_PSC)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PSC[15:0]

19.4.10 TIM9/12 auto-reload register (TIMx_ARR)

Reset value: 0xFFFF
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ARR[15:0]

ARR is the value to be loaded into the actual auto-reload register.
Refer to the Section 19.3.1: Time-base unit for more details about ARR update and
behavior.
682/1751 RM0090 Rev 19

19.4.11 TIM9/12 capture/compare register 1 (TIMx_CCR1)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR1[15:0]

CCR1 is the value to be loaded into the actual capture/compare 1 register (preload value).
(OC1PE bit). Else the preload value is copied into the active capture/compare 1 register
The active capture/compare register contains the value to be compared to the TIMx_CNT
counter and signaled on the OC1 output.
19.4.12 TIM9/12 capture/compare register 2 (TIMx_CCR2)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR2[15:0]

CCR2 is the value to be loaded into the actual capture/compare 2 register (preload value).
(OC2PE bit). Else the preload value is copied into the active capture/compare 2 register
The active capture/compare register contains the value to be compared to the TIMx_CNT
counter and signalled on the OC2 output.
RM0090 Rev 19 683/1751

695
19.4.13 TIM9/12 register map

TIM9/12 registers are mapped as 16-bit addressable registers as described below. The
reserved memory areas are highlighted in gray in the table.
Table 103. TIM9/12 register map and reset values

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
CKD
ARPE
UDIS
OPM
CEN
URS
TIMx_CR1
0x00 Reserved [1:0] Reserved
Reserved
MSM
TIMx_SMCR TS[2:0] SMS[2:0]
0x08 Reserved
CC2IE
CC1IE
UIE
TIMx_DIER
TIE
Reset value 0 0 0 0
CC2OF
CC1OF
CC2IF
CC1IF
Reserved
TIMx_SR
UIF
TIF
CC2G
CC1G
TIMx_EGR
UG
TG
Reset value 0 0 0 0
TIMx_CCMR1 OC2PE CC1
OC1PE
OC2FE
OC1FE
OC2M CC2S OC1M
Reserved
Output compare S
Reserved [2:0] [1:0] [2:0]
mode [1:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x18
CC2S
Reserved [1:0]
mode [1:0] [1:0] [1:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x1C Reserved
CC2NP
CC1NP
CC2P
CC2E
CC1P
CC1E
Reserved
0x20
TIMx_CCER
Reserved Reserved
TIMx_CNT CNT[15:0]
0x24 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
TIMx_PSC PSC[15:0]
0x28 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
TIMx_ARR ARR[15:0]
0x2C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x30 Reserved
0x34 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
684/1751 RM0090 Rev 19

Table 103. TIM9/12 register map and reset values (continued)

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
0x38 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x3C to
Reserved
0x4C
RM0090 Rev 19 685/1751

695
19.5 TIM10/11/13/14 registers

19.5.1 TIM10/11/13/14 control register 1 (TIMx_CR1)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CKD[1:0] ARPE OPM URS UDIS CEN
Reserved Reserved

sampling clock used by the digital filters (TIx),
00: tDTS = tCK_INT
11: Reserved
0: Counter is not stopped on the update event
1: Counter stops counting on the next update event (clearing the CEN bit).
This bit is set and cleared by software to select the update interrupt (UEV) sources.
0: Any of the following events generate an UEV if enabled:
1: Only counter overflow generates an UEV if enabled.
This bit is set and cleared by software to enable/disable update interrupt (UEV) event
generation.
0: UEV enabled. An UEV is generated by one of the following events:
– Setting the UG bit.
1: UEV disabled. No UEV is generated, shadow registers keep their value (ARR, PSC,
CCRx). The counter and the prescaler are reinitialized if the UG bit is set.
0: Counter disabled
1: Counter enabled
686/1751 RM0090 Rev 19

19.5.2 TIM10/11/13/14 Interrupt enable register (TIMx_DIER)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CC1IE UIE
Reserved
rw rw

19.5.3 TIM10/11/13/14 status register (TIMx_SR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CC1OF CC1IF UIF
Reserved Reserved
rc_w0 rc_w0 rc_w0

already set
This flag is set by hardware when the counter matches the compare value. It is cleared by
software.
0: No match.
When the contents of TIMx_CCR1 are greater than the contents of TIMx_ARR, the CC1IF bit
goes high on the counter overflow.
TIMx_CCR1 register.
0: No input capture occurred.
1: The counter value has been captured in TIMx_CCR1 register (an edge has been detected
on IC1 which matches the selected polarity).
RM0090 Rev 19 687/1751

695

– At overflow and if UDIS=’0’ in the TIMx_CR1 register.
– When CNT is reinitialized by software using the UG bit in TIMx_EGR register, if
URS=’0’ and UDIS=’0’ in the TIMx_CR1 register.
19.5.4 TIM10/11/13/14 event generation register (TIMx_EGR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CC1G UG
Reserved
w w

hardware.
0: No action
CC1IF flag is set, Corresponding interrupt or is sent if enabled.
the corresponding interrupt is sent if enabled. The CC1OF flag is set if the CC1IF flag was
already high.
0: No action
1: Re-initialize the counter and generates an update of the registers. Note that the prescaler
counter is cleared too (anyway the prescaler ratio is not affected). The counter is cleared.
19.5.5 TIM10/11/13/14 capture/compare mode register 1 (TIMx_CCMR1)

Reset value: 0x0000
function when the channel is configured in input. So take care that the same bit can have a
different meaning for the input stage and for the output stage.
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserved OC1M[2:0] OC1PE OC1FE
CC1S[1:0]
IC1F[3:0] IC1PSC[1:0]
Reserved
688/1751 RM0090 Rev 19

Output compare mode

These bits define the behavior of the output reference signal OC1REF from which OC1 is
derived. OC1REF is active high whereas OC1 active level depends on CC1P bit.
000: Frozen. The comparison between the output compare register TIMx_CCR1 and the
counter TIMx_CNT has no effect on the outputs.
011: Toggle - OC1REF toggles when TIMx_CNT = TIMx_CCR1.
110: PWM mode 1 - Channel 1 is active as long as TIMx_CNT < TIMx_CCR1 else inactive.
111: PWM mode 2 - Channel 1 is inactive as long as TIMx_CNT < TIMx_CCR1 else active.
Note: In PWM mode 1 or 2, the OCREF level changes when the result of the comparison
changes or when the output compare mode switches from frozen to PWM mode.
Note: The PWM mode can be used without validating the preload register only in one pulse
mode (OPM bit set in TIMx_CR1 register). Else the behavior is not guaranteed.
5 clock cycles.
1: An active edge on the trigger input acts like a compare match on CC1 output. OC is then
set to the compare level independently of the result of the comparison. Delay to sample the
trigger input and to activate CC1 output is reduced to 3 clock cycles. OC1FE acts only if the
channel is configured in PWM1 or PWM2 mode.
10:
11:
RM0090 Rev 19 689/1751

695
Input capture mode

This bit-field defines the frequency used to sample TI1 input and the length of the digital filter
applied to TI1. The digital filter is made of an event counter in which N consecutive events
are needed to validate a transition on the output:
10: Reserved
11: Reserved
690/1751 RM0090 Rev 19

19.5.6 TIM10/11/13/14 capture/compare enable register (TIMx_CCER)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CC1NP CC1P CC1E
Reserved Res.
rw rw rw

Bit 3 CC1NP: Capture/Compare 1 complementary output Polarity.
CC1 channel configured as output: CC1NP must be kept cleared.
CC1 channel configured as input: CC1NP bit is used in conjunction with CC1P to define
TI1FP1 polarity (refer to CC1P description).
0: OC1 active high
1: OC1 active low
The CC1P bit selects TI1FP1 and TI2FP1 polarity for trigger or capture operations.
Circuit is sensitive to TI1FP1 rising edge (capture mode), TI1FP1 is not inverted.
Circuit is sensitive to TI1FP1 falling edge (capture mode), TI1FP1 is inverted.
11: noninverted/both edges
Circuit is sensitive to both TI1FP1 rising and falling edges (capture mode), TI1FP1 is not
inverted.
0: Off - OC1 is not active
1: On - OC1 signal is output on the corresponding output pin
0: Capture disabled
1: Capture enabled

0 Output Disabled (OCx=’0’, OCx_EN=’0’)
1 OCx=OCxREF + Polarity, OCx_EN=’1’
Note: The state of the external I/O pins connected to the standard OCx channels depends on the
OCx channel state and the GPIO registers.
RM0090 Rev 19 691/1751

695
19.5.7 TIM10/11/13/14 counter (TIMx_CNT)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CNT[15:0]
19.5.8 TIM10/11/13/14 prescaler (TIMx_PSC)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PSC[15:0]

19.5.9 TIM10/11/13/14 auto-reload register (TIMx_ARR)

Reset value: 0xFFFF
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ARR[15:0]

692/1751 RM0090 Rev 19

19.5.10 TIM10/11/13/14 capture/compare register 1 (TIMx_CCR1)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR1[15:0]

It is loaded permanently if the preload feature is not selected in the TIMx_CCMR1 register (bit

CCR1 is the counter value transferred by the last input capture 1 event (IC1). The TIMx_CCR1
register is read-only and cannot be programmed.
19.5.11 TIM11 option register 1 (TIM11_OR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TI1_RMP[1:0]
Reserved
rw

Bits 1:0 TI1_RMP[1:0]: TIM11 Input 1 remapping capability
00,01,11: TIM11 Channel1 is connected to the GPIO (refer to the Alternate function mapping
table in the datasheets).
10: HSE_RTC clock (HSE divided by programmable prescaler) is connected to the
TIM11_CH1 input for measurement purposes
RM0090 Rev 19 693/1751

695
19.5.12 TIM10/11/13/14 register map

TIMx registers are mapped as 16-bit addressable registers as described in the table below.
Table 105. TIM10/11/13/14 register map and reset values

Offset Register 31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
CKD
ARPE
UDIS
OPM
CEN
URS
TIMx_CR1 Reserve
0x00 Reserved [1:0]
d
TIMx_SMCR
0x08 Reserved
Reset value
CC1IE
UIE
TIMx_DIER
0x0C Reserved
Reset value 0 0
CC1OF
CC1IF
TIMx_SR
UIF
Reset value 0 0 0
CC1G
TIMx_EGR
UG
0x14 Reserved
Reset value 0 0
TIMx_CCMR1
OC1PE
OC1FE
OC1M CC1S
Output compare
mode
0x18
TIMx_CCMR1 IC1
CC1S
Input capture IC1F[3:0] PSC
Reserved [1:0]
mode [1:0]
Reset value 0 0 0 0 0 0 0 0
0x1C Reserved
CC1NP
CC1P
CC1E
Reserved
TIMx_CCER
0x20 Reserved
Reset value 0 0 0
TIMx_CNT CNT[15:0]
0x24 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
TIMx_PSC PSC[15:0]
0x28 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
TIMx_ARR ARR[15:0]
0x2C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x30 Reserved
0x34 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x38 to
Reserved
0x4C
694/1751 RM0090 Rev 19

Table 105. TIM10/11/13/14 register map and reset values (continued)

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
TI1_RMP
TIMx_OR
0x50 Reserved
Reset value 0 0
RM0090 Rev 19 695/1751

695
Basic timers (TIM6 and TIM7) RM0090
20 Basic timers (TIM6 and TIM7)
20.1 TIM6 and TIM7 introduction

The basic timers TIM6 and TIM7 consist of a 16-bit auto-reload counter driven by a
programmable prescaler.
They may be used as generic timers for time-base generation but they are also specifically
used to drive the digital-to-analog converter (DAC). In fact, the timers are internally
connected to the DAC and are able to drive it through their trigger outputs.
The timers are completely independent, and do not share any resources.
20.2 TIM6 and TIM7 main features

Basic timer (TIM6 and TIM7) features include:
• 16-bit programmable prescaler used to divide (also “on the fly”) the counter clock
frequency by any factor between 1 and 65536
• Synchronization circuit to trigger the DAC
• Interrupt/DMA generation on the update event: counter overflow
Figure 203. Basic timer block diagram
Internal clock (CK_INT) Trigger TRGO to DAC

TIMxCLK from RCC controller
Reset, Enable, Count,

Controller
U
Auto-reload Register UI
Stop, Clear or up
U
±
Prescaler COUNTER
Flag Preload registers transferred

to active registers on U event according to control bit
event
interrupt & DMA output ai14749b
696/1751 RM0090 Rev 19

RM0090 Basic timers (TIM6 and TIM7)
20.3 TIM6 and TIM7 functional description

The main block of the programmable timer is a 16-bit upcounter with its related auto-reload
register. The counter clock can be divided by a prescaler.
• Auto-Reload register (TIMx_ARR)
The auto-reload register is preloaded. The preload register is accessed each time an
attempt is made to write or read the auto-reload register. The contents of the preload
register are transferred into the shadow register permanently or at each update event UEV,
depending on the auto-reload preload enable bit (ARPE) in the TIMx_CR1 register. The
update event is sent when the counter reaches the overflow value and if the UDIS bit equals
0 in the TIMx_CR1 register. It can also be generated by software. The generation of the
update event is described in detail for each configuration.
counter enable bit (CEN) in the TIMx_CR1 register is set.
Note that the actual counter enable signal CNT_EN is set 1 clock cycle after CEN.
It can be changed on the fly as the TIMx_PSC control register is buffered. The new
prescaler ratio is taken into account at the next update event.
ratio is changed on the fly.
RM0090 Rev 19 697/1751

707
CK_PSC
CNT_EN
Timerclock = CK_CNT
Update event (UEV)
MS31076V3
CK_PSC
CNT_EN
Timerclock = CK_CNT
Counter register F7 F8 F9 FA FB FC 00 01
Update event (UEV)
MS31077V3
698/1751 RM0090 Rev 19

20.3.2 Counting mode

The counter counts from 0 to the auto-reload value (contents of the TIMx_ARR register),
then restarts from 0 and generates a counter overflow event.
An update event can be generate at each counter overflow or by setting the UG bit in the
TIMx_EGR register (by software or by using the slave mode controller).
register. This avoids updating the shadow registers while writing new values into the preload
registers. In this way, no update event occurs until the UDIS bit has been written to 0,
however, the counter and the prescaler counter both restart from 0 (but the prescale rate
does not change). In addition, if the URS (update request selection) bit in the TIMx_CR1
register is set, setting the UG bit generates an update event UEV, but the UIF flag is not set
(so no interrupt or DMA request is sent).
the TIMx_SR register) is set (depending on the URS bit):
• The buffer of the prescaler is reloaded with the preload value (contents of the
TIMx_PSC register)
• The auto-reload shadow register is updated with the preload value (TIMx_ARR)
frequencies when TIMx_ARR = 0x36.
CK_INT
CNT_EN
Timerclock = CK_CNT
Counter overflow
Update event (UEV)

MS37364V1
RM0090 Rev 19 699/1751

707
CK_INT
CNT_EN
Timerclock = CK_CNT
Counter overflow
Update event (UEV)

MS35835V1
CK_INT
CNT_EN
Timerclock = CK_CNT
Counter overflow
Update event (UEV)

MSv37301V1
CK_INT
Timerclock = CK_CNT
Counter overflow
Update event (UEV)

MSv37302V1
700/1751 RM0090 Rev 19

preloaded)
CK_INT
CNT_EN
Timerclock = CK_CNT
Counter overflow
Update event (UEV)
Auto-reload register FF 36

MSv37303V1
preloaded)
CK_PSC
CNT_EN
Timerclock = CK_CNT
Counter overflow
Update event (UEV)
Auto-reload preload register F5 36
Auto-reload shadow register F5 36

MSv37304V1
20.3.3 Clock source

The counter clock is provided by the Internal clock (CK_INT) source.
The CEN (in the TIMx_CR1 register) and UG bits (in the TIMx_EGR register) are actual
control bits and can be changed only by software (except for UG that remains cleared
automatically). As soon as the CEN bit is written to 1, the prescaler is clocked by the internal
clock CK_INT.
without prescaler.
RM0090 Rev 19 701/1751

707
Internal clock
CEN=CNT_EN
UG
CNT_INIT
MS31085V2
20.3.4 Debug mode

When the microcontroller enters the debug mode (Cortex®-M4 with FPU core - halted), the
TIMx counter either continues to work normally or stops, depending on the
DBG_TIMx_STOP configuration bit in the DBG module. For more details, refer to
Section 38.16.2: Debug support for timers, watchdog, bxCAN and I2C.


Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ARPE OPM URS UDIS CEN
Reserved Reserved
rw rw rw rw rw

0: TIMx_ARR register is not buffered.
1: TIMx_ARR register is buffered.
1: Counter stops counting at the next update event (clearing the CEN bit).
702/1751 RM0090 Rev 19


0: Any of the following events generates an update interrupt or DMA request if enabled.
1: Only counter overflow/underflow generates an update interrupt or DMA request if enabled.
(ARR, PSC). However the counter and the prescaler are reinitialized if the UG bit is set or if
a hardware reset is received from the slave mode controller.
0: Counter disabled
1: Counter enabled
Note: Gated mode can work only if the CEN bit has been previously set by software. However
trigger mode can set the CEN bit automatically by hardware.
RM0090 Rev 19 703/1751

707

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MMS[2:0]
Reserved Reserved
rw rw rw

These bits are used to select the information to be sent in master mode to slave timers for
000: Reset - the UG bit from the TIMx_EGR register is used as a trigger output (TRGO). If
001: Enable - the Counter enable signal, CNT_EN, is used as a trigger output (TRGO). It is
useful to start several timers at the same time or to control a window in which a slave timer
is enabled. The Counter Enable signal is generated by a logic OR between CEN control bit
and the trigger input when configured in gated mode.
When the Counter Enable signal is controlled by the trigger input, there is a delay on TRGO,
except if the master/slave mode is selected (see the MSM bit description in the TIMx_SMCR
register).
010: Update - The update event is selected as a trigger output (TRGO). For instance a
master timer can then be used as a prescaler for a slave timer.
the master timer.
20.4.3 TIM6 and TIM7 DMA/Interrupt enable register (TIMx_DIER)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
UDE UIE
Reserved Reserved
rw rw

0: Update DMA request disabled.
1: Update DMA request enabled.
0: Update interrupt disabled.
1: Update interrupt enabled.
704/1751 RM0090 Rev 19

20.4.4 TIM6 and TIM7 status register (TIMx_SR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
UIF
Reserved
rc_w0

– At overflow or underflow and if UDIS = 0 in the TIMx_CR1 register.
– When CNT is reinitialized by software using the UG bit in the TIMx_EGR register, if URS = 0
and UDIS = 0 in the TIMx_CR1 register.
20.4.5 TIM6 and TIM7 event generation register (TIMx_EGR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
UG
Reserved
w

0: No action.
1: Re-initializes the timer counter and generates an update of the registers. Note that the
prescaler counter is cleared too (but the prescaler ratio is not affected).
20.4.6 TIM6 and TIM7 counter (TIMx_CNT)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CNT[15:0]
RM0090 Rev 19 705/1751

707
20.4.7 TIM6 and TIM7 prescaler (TIMx_PSC)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PSC[15:0]

20.4.8 TIM6 and TIM7 auto-reload register (TIMx_ARR)

Reset value: 0xFFFF
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ARR[15:0]

ARR is the value to be loaded into the actual auto-reload register.
706/1751 RM0090 Rev 19

20.4.9 TIM6 and TIM7 register map

TIMx registers are mapped as 16-bit addressable registers as described in the table below.
Table 106. TIM6 and TIM7 register map and reset values
Offset Register 31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
ARPE
Reserved
UDIS
OPM
CEN
URS
TIMx_CR1
0x00 Reserved
Reserved
TIMx_CR2 MMS[2:0]
0x04 Reserved
Reset value 0 0 0
0x08 Reserved
Reserved
UDE
UIE
TIMx_DIER
0x0C Reserved
Reset value 0 0
TIMx_SR
UIF
0x10 Reserved
Reset value 0
TIMx_EGR
UG
0x14 Reserved
Reset value 0
0x18 Reserved
0x1C Reserved
0x20 Reserved
TIMx_CNT CNT[15:0]
0x24 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
TIMx_PSC PSC[15:0]
0x28 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
TIMx_ARR ARR[15:0]
0x2C Reserved
Reset value 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
RM0090 Rev 19 707/1751

707
Independent watchdog (IWDG) RM0090
21 Independent watchdog (IWDG)
21.1 IWDG introduction

The devices have two embedded watchdog peripherals which offer a combination of high
safety level, timing accuracy and flexibility of use. Both watchdog peripherals (Independent
and Window) serve to detect and resolve malfunctions due to software failure, and to trigger
system reset or an interrupt (window watchdog only) when the counter reaches a given
timeout value.
The independent watchdog (IWDG) is clocked by its own dedicated low-speed clock (LSI)
and thus stays active even if the main clock fails. The window watchdog (WWDG) clock is
prescaled from the APB1 clock and has a configurable time-window that can be
programmed to detect abnormally late or early application behavior.
The IWDG is best suited to applications which require the watchdog to run as a totally
independent process outside the main application, but have lower timing accuracy
constraints. The WWDG is best suited to applications which require the watchdog to react
within an accurate timing window. For further information on the window watchdog, refer to
Section 22 on page 713.
21.2 IWDG main features

• Free-running downcounter
• clocked from an independent RC oscillator (can operate in Standby and Stop modes)
• Reset (if watchdog activated) when the downcounter value of 0x000 is reached
21.3 IWDG functional description

Figure 213 shows the functional blocks of the independent watchdog module.
When the independent watchdog is started by writing the value 0xCCCC in the Key register
(IWDG_KR), the counter starts counting down from the reset value of 0xFFF. When it
reaches the end of count value (0x000) a reset signal is generated (IWDG reset).
Whenever the key value 0xAAAA is written in the IWDG_KR register, the IWDG_RLR value
is reloaded in the counter and the watchdog reset is prevented.
21.3.1 Hardware watchdog

If the “Hardware watchdog” feature is enabled through the device option bits, the watchdog
is automatically enabled at power-on, and will generate a reset unless the Key register is
written by the software before the counter reaches end of count.
21.3.2 Register access protection

Write access to the IWDG_PR and IWDG_RLR registers is protected. To modify them, first
write the code 0x5555 in the IWDG_KR register. A write access to this register with a
708/1751 RM0090 Rev 19

RM0090 Independent watchdog (IWDG)
different value will break the sequence and register access will be protected again. This
implies that it is the case of the reload operation (writing 0xAAAA).
A status register is available to indicate that an update of the prescaler or the down-counter
reload value is on going.
21.3.3 Debug mode

When the microcontroller enters debug mode (Cortex®-M4 with FPU core halted), the IWDG
counter either continues to work normally or stops, depending on DBG_IWDG_STOP
Figure 213. Independent watchdog block diagram
CORE
Prescaler register Status register Reload register Key register
IWDG_PR IWDG_SR IWDG_RLR IWDG_KR
12-bit reload value

LSI 8-bit
(40 kHz) prescaler
12-bit downcounter IWDG reset
VDD voltage domain
MS19944V2
Note: The watchdog function is implemented in the VDD voltage domain, still functional in Stop and
Standby modes.
Table 107. Min/max IWDG timeout period (in ms) at 32 kHz (LSI)(1)
Prescaler divider PR[2:0] bits Min timeout RL[11:0]= 0x000 Max timeout RL[11:0]= 0xFFF
/4 0 0.125 512
/8 1 0.25 1024
/16 2 0.5 2048
/32 3 1 4096
/64 4 2 8192
/128 5 4 16384
/256 6 8 32768
1. These timings are given for a 32 kHz clock but the microcontroller internal RC frequency can vary. Refer to
the LSI oscillator characteristics table in the device datasheet for maximum and minimum values.
RM0090 Rev 19 709/1751

712
21.4 IWDG registers

Refer to Section 2.2 on page 45 for a list of abbreviations used in register descriptions.
The peripheral registers have to be accessed by half-words (16 bits) or words (32 bits).
21.4.1 Key register (IWDG_KR)

Reset value: 0x0000 0000 (reset by Standby mode)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
KEY[15:0]
Reserved

Bits 15:0 KEY[15:0]: Key value (write only, read 0000h)
These bits must be written by software at regular intervals with the key value AAAAh,
otherwise the watchdog generates a reset when the counter reaches 0.
Writing the key value 5555h to enable access to the IWDG_PR and IWDG_RLR registers
(see Section 21.3.2)
Writing the key value CCCCh starts the watchdog (except if the hardware watchdog option is
selected)
21.4.2 Prescaler register (IWDG_PR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PR[2:0]
Reserved
rw rw rw

Bits 2:0 PR[2:0]: Prescaler divider
These bits are write access protected seeSection 21.3.2. They are written by software to
select the prescaler divider feeding the counter clock. PVU bit of IWDG_SR must be reset in
order to be able to change the prescaler divider.
000: divider /4
001: divider /8
010: divider /16
011: divider /32
100: divider /64
101: divider /128
110: divider /256
111: divider /256
Note: Reading this register returns the prescaler value from the VDD voltage domain. This
value may not be up to date/valid if a write operation to this register is ongoing. For this
reason the value read from this register is valid only when the PVU bit in the IWDG_SR
register is reset.
710/1751 RM0090 Rev 19

RM0090 Independent watchdog (IWDG)
21.4.3 Reload register (IWDG_RLR)

Reset value: 0x0000 0FFF (reset by Standby mode)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RL[11:0]
Reserved

Bits11:0 RL[11:0]: Watchdog counter reload value
These bits are write access protected see Section 21.3.2. They are written by software to
define the value to be loaded in the watchdog counter each time the value AAAAh is written
in the IWDG_KR register. The watchdog counter counts down from this value. The timeout
period is a function of this value and the clock prescaler. Refer to Table 107.
The RVU bit in the IWDG_SR register must be reset in order to be able to change the reload
value.
Note: Reading this register returns the reload value from the VDD voltage domain. This value
may not be up to date/valid if a write operation to this register is ongoing on this
register. For this reason the value read from this register is valid only when the RVU bit
in the IWDG_SR register is reset.
21.4.4 Status register (IWDG_SR)

Reset value: 0x0000 0000 (not reset by Standby mode)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RVU PVU
Reserved
r r

Bit 1 RVU: Watchdog counter reload value update
This bit is set by hardware to indicate that an update of the reload value is ongoing. It is reset
by hardware when the reload value update operation is completed in the VDD voltage domain
(takes up to 5 RC 40 kHz cycles).
Reload value can be updated only when RVU bit is reset.
Bit 0 PVU: Watchdog prescaler value update
This bit is set by hardware to indicate that an update of the prescaler value is ongoing. It is
reset by hardware when the prescaler update operation is completed in the VDD voltage
domain (takes up to 5 RC 40 kHz cycles).
Prescaler value can be updated only when PVU bit is reset.
Note: If several reload values or prescaler values are used by application, it is mandatory to wait
until RVU bit is reset before changing the reload value and to wait until PVU bit is reset
before changing the prescaler value. However, after updating the prescaler and/or the
reload value it is not necessary to wait until RVU or PVU is reset before continuing code
execution (even in case of low-power mode entry, the write operation is taken into account
and will complete)
RM0090 Rev 19 711/1751

712
21.4.5 IWDG register map

The following table gives the IWDG register map and reset values.
Table 108. IWDG register map and reset values

Offset Register 31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
IWDG_KR KEY[15:0]
0x00 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
IWDG_PR PR[2:0]
0x04 Reserved
Reset value 0 0 0
IWDG_RLR RL[11:0]
0x08 Reserved
Reset value 1 1 1 1 1 1 1 1 1 1 1 1
RVU
PVU
IWDG_SR
0x0C Reserved
Reset value 0 0
712/1751 RM0090 Rev 19

RM0090 Window watchdog (WWDG)
22 Window watchdog (WWDG)
22.1 WWDG introduction

The window watchdog is used to detect the occurrence of a software fault, usually
generated by external interference or by unforeseen logical conditions, which causes the
application program to abandon its normal sequence. The watchdog circuit generates an
MCU reset on expiry of a programmed time period, unless the program refreshes the
contents of the downcounter before the T6 bit becomes cleared. An MCU reset is also
generated if the 7-bit downcounter value (in the control register) is refreshed before the
downcounter has reached the window register value. This implies that the counter must be
refreshed in a limited window.
22.2 WWDG main features

• Programmable free-running downcounter
• Conditional reset
– Reset (if watchdog activated) when the downcounter value becomes less than
0x40
– Reset (if watchdog activated) if the downcounter is reloaded outside the window
(see Figure 215)
• Early wakeup interrupt (EWI): triggered (if enabled and the watchdog activated) when
the downcounter is equal to 0x40.
22.3 WWDG functional description

If the watchdog is activated (the WDGA bit is set in the WWDG_CR register) and when the
7-bit downcounter (T[6:0] bits) rolls over from 0x40 to 0x3F (T6 becomes cleared), it initiates
a reset. If the software reloads the counter while the counter is greater than the value stored
in the window register, then a reset is generated.
RM0090 Rev 19 713/1751

719
Window watchdog (WWDG) RM0090
Figure 214. Watchdog block diagram

5(6(7
:DWFKGRJFRQILJXUDWLRQUHJLVWHU ::'*B&)5
: : : : : : :
FRPSDUDWRU
ZKHQ
7!:
:ULWH::'*B&5
:DWFKGRJFRQWUROUHJLVWHU ::'*B&5
:'*$ 7 7 7 7 7 7 7
ELWGRZQFRXQWHU &17
3&/.
IURP5&&FORFNFRQWUROOHU
:'*SUHVFDOHU
:'*7%
06Y9
The application program must write in the WWDG_CR register at regular intervals during
normal operation to prevent an MCU reset. This operation must occur only when the counter
value is lower than the window register value. The value to be stored in the WWDG_CR
register must be between 0xFF and 0xC0.
Enabling the watchdog

The watchdog is always disabled after a reset. It is enabled by setting the WDGA bit in the
WWDG_CR register, then it cannot be disabled again except by a reset.
Controlling the downcounter

This downcounter is free-running, counting down even if the watchdog is disabled. When
the watchdog is enabled, the T6 bit must be set to prevent generating an immediate reset.
The T[5:0] bits contain the number of increments which represents the time delay before the
watchdog produces a reset. The timing varies between a minimum and a maximum value
due to the unknown status of the prescaler when writing to the WWDG_CR register (see
Figure 215). The Configuration register (WWDG_CFR) contains the high limit of the window:
To prevent a reset, the downcounter must be reloaded when its value is lower than the
window register value and greater than 0x3F. Figure 215 describes the window watchdog
process.
Note: The T6 bit can be used to generate a software reset (the WDGA bit is set and the T6 bit is
cleared).
Advanced watchdog interrupt feature

The Early Wakeup Interrupt (EWI) can be used if specific safety operations or data logging
must be performed before the actual reset is generated. The EWI interrupt is enabled by
setting the EWI bit in the WWDG_CFR register. When the downcounter reaches the value
0x40, an EWI interrupt is generated and the corresponding interrupt service routine (ISR)
can be used to trigger specific actions (such as communications or data logging), before
resetting the device.
714/1751 RM0090 Rev 19

In some applications, the EWI interrupt can be used to manage a software system check
and/or system recovery/graceful degradation, without generating a WWDG reset. In this
case, the corresponding interrupt service routine (ISR) should reload the WWDG counter to
avoid the WWDG reset, then trigger the required actions.
The EWI interrupt is cleared by writing '0' to the EWIF bit in the WWDG_SR register.
Note: When the EWI interrupt cannot be served, e.g. due to a system lock in a higher priority task,
the WWDG reset will eventually be generated.
22.4 How to program the watchdog timeout
Warning: When writing to the WWDG_CR register, always write 1 in the

T6 bit to avoid generating an immediate reset.
Figure 215. Window watchdog timing diagram

T[6:0] CNT downcounter
W[6:0]
0x3F
Refresh not allowed Refresh allowed Time
T6 bit
RESET
ai17101c
The formula to calculate the WWDG timeout value is given by:

WDGTB[1:0]
t WWDG = t PCLK1 × 4096 × 2 × ( T[5:0] + 1 ) ( ms )
where:
tWWDG: WWDG timeout
tPCLK1: APB1 clock period measured in ms
4096: value corresponding to internal divider
RM0090 Rev 19 715/1751

719
As an example, let us assume APB1 frequency is equal to 24 MHz, WDGTB[1:0] is set to 3

and T[5:0] is set to 63:
3
t WWDG = 1 ⁄ 24000 × 4096 × 2 × ( 63 + 1 ) = 21.85ms
Refer to Table 109 for the minimum and maximum values of the tWWDG.
Table 109. Minimum and maximum timeout values at 30 MHz (fPCLK1)

Min timeout (µs) Max timeout (ms)
Prescaler WDGTB
T[5:0] = 0x00 T[5:0] = 0x3F
1 0 136.53 8.74
2 1 273.07 17.48
4 2 546.13 34.95
8 3 1092.27 69.91
22.5 Debug mode

When the microcontroller enters debug mode (Cortex®-M4 with FPU core halted), the
WWDG counter either continues to work normally or stops, depending on
DBG_WWDG_STOP configuration bit in DBG module. For more details, refer to
Section 38.16.2: Debug support for timers, watchdog, bxCAN and I2C.
716/1751 RM0090 Rev 19

22.6 WWDG registers

22.6.1 Control register (WWDG_CR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
WDGA T[6:0]
Reserved
rs rw

Bit 7 WDGA: Activation bit
This bit is set by software and only cleared by hardware after a reset. When WDGA = 1, the
watchdog can generate a reset.
0: Watchdog disabled
1: Watchdog enabled
Bits 6:0 T[6:0]: 7-bit counter (MSB to LSB)
These bits contain the value of the watchdog counter. It is decremented every (4096 x
2WDGTB[1:0]) PCLK1 cycles. A reset is produced when it rolls over from 0x40 to 0x3F (T6
becomes cleared).
RM0090 Rev 19 717/1751

719
22.6.2 Configuration register (WWDG_CFR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
EWI WDGTB[1:0] W[6:0]
Reserved
rs rw rw

Bit 9 EWI: Early wakeup interrupt
When set, an interrupt occurs whenever the counter reaches the value 0x40. This interrupt is
only cleared by hardware after a reset.
Bits 8:7 WDGTB[1:0]: Timer base
The time base of the prescaler can be modified as follows:
00: CK Counter Clock (PCLK1 div 4096) div 1
Bits 6:0 W[6:0]: 7-bit window value
These bits contain the window value to be compared to the downcounter.
22.6.3 Status register (WWDG_SR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
EWIF
Reserved
rc_w0

Bit 0 EWIF: Early wakeup interrupt flag
This bit is set by hardware when the counter has reached the value 0x40. It must be cleared
by software by writing ‘0’. A write of ‘1’ has no effect. This bit is also set if the interrupt is not
enabled.
718/1751 RM0090 Rev 19

22.6.4 WWDG register map

The following table gives the WWDG register map and reset values.
Table 110. WWDG register map and reset values

Offset Register 31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
WDGA
WWDG_CR T[6:0]
0x00 Reserved
Reset value 0 1 1 1 1 1 1 1
WDGTB1
WDGTB0
EWI
WWDG_CFR W[6:0]
0x04 Reserved
Reset value 0 0 0 1 1 1 1 1 1 1
EWIF
WWDG_SR
0x08 Reserved
Reset value 0
RM0090 Rev 19 719/1751

719
Cryptographic processor (CRYP) RM0090
23 Cryptographic processor (CRYP)
This section applies to STM32F415/417xx and STM32F43xxx devices.
23.1 CRYP introduction

The cryptographic processor can be used to both encipher and decipher data using the
DES, Triple-DES or AES (128, 192, or 256) algorithms. It is a fully compliant implementation
of the following standards:
• The data encryption standard (DES) and Triple-DES (TDES) as defined by Federal
Information Processing Standards Publication (FIPS PUB 46-3, 1999 October 25). It
follows the American National Standards Institute (ANSI) X9.52 standard.
• The advanced encryption standard (AES) as defined by Federal Information
Processing Standards Publication (FIPS PUB 197, 2001 November 26)
The CRYP processor performs data encryption and decryption using DES and TDES
algorithms in Electronic codebook (ECB) or Cipher block chaining (CBC) mode.
The CRYP peripheral is a 32-bit AHB2 peripheral. It supports DMA transfer for incoming and
processed data, and has input and output FIFOs (each 8 words deep).
23.2 CRYP main features

• Suitable for AES, DES and TDES enciphering and deciphering operations
• AES
– Supports the ECB, CBC, CTR, CCM and GCM chaining algorithms (CCM and
GCM are available on STM32F42xxx and STM32F43xxx only)
– Supports 128-, 192- and 256-bit keys
– 4 × 32-bit initialization vectors (IV) used in the CBC, CTR, CCM and GCM modes

(STM32F415/417xx)
Algorithm / Key size ECB CBC CTR
128b 14 14 14
192b 16 16 16
256b 18 18 18

(STM32F43xxx)
GCM CCM
Algorithm /
ECB CBC CTR
Key size
Init Header Payload Tag Init Header Payload Tag
128b 14 14 14 24 10 14 14 12 14 25 14
720/1751 RM0090 Rev 19

RM0090 Cryptographic processor (CRYP)

(STM32F43xxx)
192b 16 16 16 28 10 16 16 14 16 29 16
256b 18 18 18 32 10 18 18 16 18 33 18
• DES/TDES
– Direct implementation of simple DES algorithms (a single key, K1, is used)
– Supports the ECB and CBC chaining algorithms
– Supports 64-, 128- and 192-bit keys (including parity)
– 2 × 32-bit initialization vectors (IV) used in the CBC mode
– 16 HCLK cycles to process one 64-bit block in DES
– 48 HCLK cycles to process one 64-bit block in TDES
• Common to DES/TDES and AES
– IN and OUT FIFO (each with an 8-word depth, a 32-bit width, corresponding to 4
DES blocks or 2 AES blocks)
– Automatic data flow control with support of direct memory access (DMA) (using 2
channels, one for incoming data the other for processed data)
– Data swapping logic to support 1-, 8-, 16- or 32-bit data
RM0090 Rev 19 721/1751

766
23.3 CRYP functional description

The cryptographic processor implements a Triple-DES (TDES, that also supports DES) core
and an AES cryptographic core. Section 23.3.1 and Section 23.3.2 provide details on these
cores.
Since the TDES and the AES algorithms use block ciphers, incomplete input data blocks
have to be padded prior to encryption (extra bits should be appended to the trailing end of
the data string). After decryption, the padding has to be discarded. The hardware does not
manage the padding operation, the software has to handle it.
Figure 216 shows the block diagram of the cryptographic processor.
Figure 216. Block diagram (STM32F415/417xx)
32-bit AHB2 bus
Status
CRYP_SR CRYP_DIN CRYP_DOUT
DMA control register
CRYP_DMACR
Interrupt registers 8 × 32-bit 8 × 32-bit
CRYP_IMSCR IN FIFO OUT FIFO
CRYP_RIS
CRYP_MISR
Control register
swappi ng swappin g
CRYP_CR
Initialization vectors
IV0...IV127
CRYP_IV0...IV1
DES/TDES/AES
Key Processor core
k255...k0
CRYP_K0. ..K3
ai16068b
722/1751 RM0090 Rev 19

Figure 217. Block diagram (STM32F43xxx)
32-bit AHB2 bus
Status
CRYP_SR CRYP_DIN CRYP_DOUT
DMA control register
CRYP_DMACR
Interrupt registers 8 × 32-bit 8 × 32-bit
CRYP_IMSCR IN FIFO OUT FIFO
CRYP_RIS
CRYP_MISR
Control register
swappi ng swappin g
CRYP_CR
Initialization vectors Context swapping

IV0...IV127
CRYP_IV0...IV1 CRYP_CSGCMCCM0..7
DES/TDES/AES
Key Processor core
k255...k0 CRYP_CSGCM0..7
CRYP_K0. ..K3
MS30441V1
23.3.1 DES/TDES cryptographic core

The DES/Triple-DES cryptographic core consists of three components:
• The DES algorithm (DEA)
• Multiple keys (1 for the DES algorithm, 1 to 3 for the TDES algorithm)
• The initialization vector (used in the CBC mode)
The basic processing involved in the TDES is as follows: an input block is read in the DEA
and encrypted using the first key, K1 (K0 is not used in TDES mode). The output is then
decrypted using the second key, K2, and encrypted using the third key, K3. The key
depends on the algorithm which is used:
• DES mode: Key = [K1]
• TDES mode: Key = [K3 K2 K1]
where Kx=[KxR KxL], R = right, L = left
According to the mode implemented, the resultant output block is used to calculate the
ciphertext.
Note that the outputs of the intermediate DEA stages is never revealed outside the
cryptographic boundary.
RM0090 Rev 19 723/1751

766
The TDES allows three different keying options:

• Three independent keys
The first option specifies that all the keys are independent, that is, K1, K2 and K3 are
independent. FIPS PUB 46-3 – 1999 (and ANSI X9.52 – 1998) refers to this option as
the Keying Option 1 and, to the TDES as 3-key TDES.
• Two independent keys
The second option specifies that K1 and K2 are independent and K3 is equal to K1,
that is, K1 and K2 are independent, K3 = K1. FIPS PUB 46-3 – 1999 (and ANSI X9.52
– 1998) refers to this second option as the Keying Option 2 and, to the TDES as 2-key
TDES.
• Three equal keys
The third option specifies that K1, K2 and K3 are equal, that is, K1 = K2 = K3. FIPS
PUB 46-3 – 1999 (and ANSI X9.52 – 1998) refers to the third option as the Keying
Option 3. This “1-key” TDES is equivalent to single DES.
FIPS PUB 46-3 – 1999 (and ANSI X9.52-1998) provides a thorough explanation of the
processing involved in the four operation modes supplied by the TDEA (TDES algorithm):
TDES-ECB encryption, TDES-ECB decryption, TDES-CBC encryption and TDES-CBC
decryption.
This reference manual only gives a brief explanation of each mode.
DES and TDES Electronic codebook (DES/TDES-ECB) mode

• DES/TDES-ECB mode encryption
Figure 218 illustrates the encryption in DES and TDES Electronic codebook
(DES/TDES-ECB) mode. A 64-bit plaintext data block (P) is used after bit/byte/half-
word swapping (refer to Section 23.3.3: Data type on page 739) as the input block (I).
The input block is processed through the DEA in the encrypt state using K1. The output
of this process is fed back directly to the input of the DEA where the DES is performed
in the decrypt state using K2. The output of this process is fed back directly to the input
of the DEA where the DES is performed in the encrypt state using K3. The resultant 64-
bit output block (O) is used, after bit/byte/half-word swapping, as ciphertext (C) and it is
pushed into the OUT FIFO.
• DES/TDES-ECB mode decryption
Figure 219 illustrates the DES/TDES-ECB decryption. A 64-bit ciphertext block (C) is
used, after bit/byte/half-word swapping, as the input block (I). The keying sequence is
reversed compared to that used in the encryption process. The input block is
processed through the DEA in the decrypt state using K3. The output of this process is
fed back directly to the input of the DEA where the DES is performed in the encrypt
state using K2. The new result is directly fed to the input of the DEA where the DES is
performed in the decrypt state using K1. The resultant 64-bit output block (O), after
bit/byte/half-word swapping, produces the plaintext (P).
724/1751 RM0090 Rev 19

Figure 218. DES/TDES-ECB mode encryption
IN FIFO
plaintext P
P, 64 bits
DATATYPE
swapping
64
K1 DEA, encrypt
64
K2 DEA, decrypt
64
K3 DEA, encrypt
O, 64 bits
DATATYPE
swapping
C, 64 bits
OUT FIFO
ciphertext C
ai16069b
1. K: key; C: cipher text; I: input block; O: output block; P: plain text.
Figure 219. DES/TDES-ECB mode decryption
IN FIFO
ciphertext C
C, 64 bits
DATATYPE
swapping
I, 64 bits
64
K3 DEA, decrypt
64
K2 DEA, encrypt
64
K1 DEA, decrypt
O, 64 bits
DATATYPE
swapping
P, 64 bits
OUT FIFO
plaintext P
MS19021V1
RM0090 Rev 19 725/1751

766
DES and TDES Cipher block chaining (DES/TDES-CBC) mode

• DES/TDES-CBC mode encryption
Figure 220 illustrates the DES and Triple-DES Cipher block chaining (DES/TDES-
CBC) mode encryption. This mode begins by dividing a plaintext message into 64-bit
data blocks. In TCBC encryption, the first input block (I1), obtained after bit/byte/half-
word swapping (refer to Section 23.3.3: Data type on page 739), is formed by
exclusive-ORing the first plaintext data block (P1) with a 64-bit initialization vector IV (I1
= IV ⊕ P1). The input block is processed through the DEA in the encrypt state using K1.
The output of this process is fed back directly to the input of the DEA, which performs
the DES in the decrypt state using K2. The output of this process is fed directly to the
input of the DEA, which performs the DES in the encrypt state using K3. The resultant
64-bit output block (O1) is used directly as the ciphertext (C1), that is, C1 = O1. This first
ciphertext block is then exclusive-ORed with the second plaintext data block to produce
the second input block, (I2) = (C1 ⊕ P2). Note that I2 and P2 now refer to the second
block. The second input block is processed through the TDEA to produce the second
ciphertext block. This encryption process continues to “chain” successive cipher and
plaintext blocks together until the last plaintext block in the message is encrypted. If the
message does not consist of an integral number of data blocks, then the final partial
data block should be encrypted in a manner specified for the application.
• DES/TDES-CBC mode decryption
In DES/TDES-CBC decryption (see Figure 221), the first ciphertext block (C1) is used
directly as the input block (I1). The keying sequence is reversed compared to that used
for the encrypt process. The input block is processed through the DEA in the decrypt
state using K3. The output of this process is fed directly to the input of the DEA where
the DES is processed in the encrypt state using K2. This resulting value is directly fed
to the input of the DEA where the DES is processed in the decrypt state using K1. The
resulting output block is exclusive-ORed with the IV (which must be the same as that
used during encryption) to produce the first plaintext block (P1 = O1 ⊕ IV). The second
ciphertext block is then used as the next input block and is processed through the
TDEA. The resulting output block is exclusive-ORed with the first ciphertext block to
produce the second plaintext data block (P2 = O2 ⊕ C1). (Note that P2 and O2 refer to
the second block of data.) The TCBC decryption process continues in this manner until
the last complete ciphertext block has been decrypted. Ciphertext representing a
partial data block must be decrypted in a manner specified for the application.
726/1751 RM0090 Rev 19

Figure 220. DES/TDES-CBC mode encryption
IN FIFO
plaintext P
P, 64 bits
DATATYPE
AHB2 data write swapping
(before CRYP
is enabled) 64 Ps, 64 bits
IV0(L/R) +
I, 64 bits
64
K1 DEA, encrypt
O is written back 64
into IV at the
same time as it K2 DEA, decrypt
is pushed into
the OUT FIFO 64
K3 DEA, encrypt
O, 64 bits
DATATYPE
swapping
C, 64 bits
OUT FIFO
ciphertext C
ai16070b
1. K: key; C: cipher text; I: input block; O: output block; Ps: plain text before swapping (when decoding) or
after swapping (when encoding); P: plain text; IV: initialization vectors.
RM0090 Rev 19 727/1751

766
Figure 221. DES/TDES-CBC mode decryption
IN FIFO
ciphertext C
C, 64 bits
DATATYPE
swapping
I, 64 bits
64
I is written back K3 DEA, decrypt
into IV at the
same time as P 64
is pushed into
the OUT FIFO K2 DEA, encrypt
64
K1 DEA, decrypt
AHB2 data write O, 64 bits

(before CRYP 64
is enabled)
IV0(L/R) +
Ps, 64 bits
DATATYPE
swapping
P, 64 bits
OUT FIFO
plaintext P
MS19022V1
after swapping (when encoding); P: plain text; IV: initialization vectors.
23.3.2 AES cryptographic core

The AES cryptographic core consists of three components:
• The AES algorithm (AEA: advanced encryption algorithm)
• Multiple keys
• Initialization vector(s) or Nonce
The AES utilizes keys of 3 possible lengths: 128, 192 or 256 bits and, depending on the
operation mode used, zero or one 128-bit initialization vector (IV).
The basic processing involved in the AES is as follows: an input block of 128 bits is read
from the input FIFO and sent to the AEA to be encrypted using the key (K0...3). The key
format depends on the key size:
• If Key size = 128: Key = [K3 K2]
• If Key size = 192: Key = [K3 K2 K1]
• If Key size = 256: Key = [K3 K2 K1 K0]
where Kx=[KxR KxL],R=right, L=left
According to the mode implemented, the resultant output block is used to calculate the
ciphertext.
FIPS PUB 197 (November 26, 2001) provides a thorough explanation of the processing
involved in the four operation modes supplied by the AES core: AES-ECB encryption, AES-
728/1751 RM0090 Rev 19

ECB decryption, AES-CBC encryption and AES-CBC decryption.This reference manual

only gives a brief explanation of each mode.
AES Electronic codebook (AES-ECB) mode

• AES-ECB mode encryption
Figure 222 illustrates the AES Electronic codebook (AES-ECB) mode encryption.
In AES-ECB encryption, a 128- bit plaintext data block (P) is used after bit/byte/half-
word swapping (refer to Section 23.3.3: Data type on page 739) as the input block (I).
The input block is processed through the AEA in the encrypt state using the 128, 192 or
256-bit key. The resultant 128-bit output block (O) is used after bit/byte/half-word
swapping as ciphertext (C). It is then pushed into the OUT FIFO.
• AES-ECB mode decryption
Figure 223 illustrates the AES Electronic codebook (AES-ECB) mode encryption.
To perform an AES decryption in the ECB mode, the secret key has to be prepared (it is
necessary to execute the complete key schedule for encryption) by collecting the last
round key, and using it as the first round key for the decryption of the ciphertext. This
preparation function is computed by the AES core. Refer to Section 23.3.6: Procedure
to perform an encryption or a decryption for more details on how to prepare the key.
In AES-ECB decryption, a 128-bit ciphertext block (C) is used after bit/byte/half-word
swapping as the input block (I). The keying sequence is reversed compared to that of
the encryption process. The resultant 128-bit output block (O), after bit/byte or half-
word swapping, produces the plaintext (P).
Figure 222. AES-ECB mode encryption
IN FIFO
plaintext P
P, 128 bits
DATATYPE
swapping
I, 128 bits
128/192
or 256
K 0...3(1) AEA, encrypt
DATATYPE
swapping
C, 128 bits
OUT FIFO
ciphertext C
ai16071b

2. If Key size = 128: Key = [K3 K2].
If Key size = 192: Key = [K3 K2 K1]
If Key size = 256: Key = [K3 K2 K1 K0].
RM0090 Rev 19 729/1751

766
Figure 223. AES-ECB mode decryption
IN FIFO
ciphertext C
C, 128 bits
DATATYPE
swapping
I, 128 bits
128/192
or 256
K 0...3(1) AEA, decrypt
O, 128 bits
DATATYPE
swapping
P, 128 bits
OUT FIFO
plaintext P
MS19023V1

2. If Key size = 128 => Key = [K3 K2].
If Key size = 192 => Key = [K3 K2 K1]
If Key size = 256 => Key = [K3 K2 K1 K0].
AES Cipher block chaining (AES-CBC) mode

• AES-CBC mode encryption
The AES Cipher block chaining (AES-CBC) mode decryption is shown on Figure 224.
In AES-CBC encryption, the first input block (I1) obtained after bit/byte/half-word
swapping (refer to Section 23.3.3: Data type on page 739) is formed by exclusive-
ORing the first plaintext data block (P1) with a 128-bit initialization vector IV (I1 = IV ⊕
P1). The input block is processed through the AEA in the encrypt state using the 128-,
192- or 256-bit key (K0...K3). The resultant 128-bit output block (O1) is used directly as
ciphertext (C1), that is, C1 = O1. This first ciphertext block is then exclusive-ORed with
the second plaintext data block to produce the second input block, (I2) = (C1 ⊕ P2). Note
that I2 and P2 now refer to the second block. The second input block is processed
through the AEA to produce the second ciphertext block. This encryption process
continues to “chain” successive cipher and plaintext blocks together until the last
plaintext block in the message is encrypted. If the message does not consist of an
integral number of data blocks, then the final partial data block should be encrypted in a
manner specified for the application.
In the CBC mode, like in the ECB mode, the secret key must be prepared to perform an
AES decryption. Refer to Section 23.3.6: Procedure to perform an encryption or a
decryption on page 744 for more details on how to prepare the key.
• AES-CBC mode decryption
In AES-CBC decryption (see Figure 225), the first 128-bit ciphertext block (C1) is used
directly as the input block (I1). The input block is processed through the AEA in the
decrypt state using the 128-, 192- or 256-bit key. The resulting output block is
exclusive-ORed with the 128-bit initialization vector IV (which must be the same as that
used during encryption) to produce the first plaintext block (P1 = O1 ⊕ IV). The second
ciphertext block is then used as the next input block and is processed through the AEA.
The resulting output block is exclusive-ORed with the first ciphertext block to produce
the second plaintext data block (P2 = O2 ⊕ C1). (Note that P2 and O2 refer to the second
730/1751 RM0090 Rev 19

block of data.) The AES-CBC decryption process continues in this manner until the last
complete ciphertext block has been decrypted. Ciphertext representing a partial data
block must be decrypted in a manner specified for the application.
Figure 224. AES-CBC mode encryption
IN FIFO
plaintext P
P, 128 bits
DATATYPE
(before CRYP
is enabled) Ps, 128 bits
128
IV=[IV1 IV0](2) +
I, 128 bits
128, 192
or 256
K 0...3(3) AEA, encrypt
O is written
back into IV
at the same time
as it is pushed
into the OUT FIFO
O, 128 bits
DATATYPE
swapping
C, 128 bits
OUT FIFO
ciphertext C
ai16072b
after swapping (when encoding); P: plain text; IV: Initialization vectors.
2. IVx=[IVxR IVxL], R=right, L=left.
3. If Key size = 128 => Key = [K3 K2].
RM0090 Rev 19 731/1751

766
Figure 225. AES-CBC mode decryption
IN FIFO
ciphertext C
C, 128 bits
DATATYPE
swapping
I, 128 bits
128, 192
or 256
K 0...3(3) AEA, decrypt
I is written
back into IV
at the same time
as P is pushed
into the OUT FIFO
AHB2 data write
(before CRYP O, 128 bits
is enabled) 128
IV=[IV1 IV0](2) +
Ps, 128 bits
DATATYPE
swapping
P, 128 bits
OUT FIFO
plaintext P
MS19024V1
after swapping (when encoding); P: plain text; IV: Initialization vectors.
2. IVx=[IVxR IVxL], R=right, L=left.
3. If Key size = 128 => Key = [K3 K2].
AES counter mode (AES-CTR) mode

The AES counter mode uses the AES block as a key stream generator. The generated keys
are then XORed with the plaintext to obtain the cipher. For this reason, it makes no sense to
speak of different CTR encryption/decryption, since the two operations are exactly the
same.
In fact, given:
• Plaintext: P[0], P[1], ..., P[n] (128 bits each)
• A key K to be used (the size does not matter)
• An initial counter block (call it ICB but it has the same functionality as the IV of CBC)
The cipher is computed as follows:
C[i] = enck(iv[i]) xor P[i], where:
iv[0] = ICB and iv[i+1] = func(iv[i]), where func is an update function
applied to the previous iv block; func is basically an increment of one of the fields
composing the iv block.
Given that the ICB for decryption is the same as the one for encryption, the key stream
generated during decryption is the same as the one generated during encryption. Then, the
ciphertext is XORed with the key stream in order to retrieve the original plaintext. The
decryption operation therefore acts exactly in the same way as the encryption operation.
732/1751 RM0090 Rev 19

Figure 226 and Figure 227 illustrate AES-CTR encryption and decryption, respectively.
Figure 226. AES-CTR mode encryption
IN FIFO
plaintext P
P, 128 bits
DATATYPE
(before CRYP
is enabled)
IV0...1(L/R)
Ps, 128 bits
+1 I, 128 bits
128, 192
or 256
K0...3 AEA, encrypt
(I + 1) is written O, 128 bits

back into IV
at same time
+
than C is pushed Cs, 128 bit
in OUT FIFO
DATATYPE
swapping
C, 128 bits
OUT FIFO
ciphertext C
ai16073b
1. K: key; C: cipher text; I: input Block; o: output block; Ps: plain text before swapping (when decoding) or
after swapping (when encoding); Cs: cipher text after swapping (when decoding) or before swapping (when
encoding); P: plain text; IV: Initialization vectors.
RM0090 Rev 19 733/1751

766
Figure 227. AES-CTR mode decryption
IN FIFO
ciphertext P
C, 128 bits
DATATYPE
(before CRYP
is enabled)
IV0...1(L/R)
Cs, 128 bits
+1 I, 128 bits
128, 192
or 256
K0...3 AEA, encrypt
(I + 1) is written O, 128 bits

back into IV
at same time
+
than P is pushed Ps, 128 bits
in OUT FIFO
DATATYPE
swapping
P, 128 bits
OUT FIFO
plaintext C
MS19025V1
1. K: key; C: cipher text; I: input Block; o: output block; Ps: plain text before swapping (when decoding) or
after swapping (when encoding); Cs: cipher text after swapping (when decoding) or before swapping (when
encoding); P: plain text; IV: Initialization vectors.
Figure 228 shows the structure of the IV block as defined by the standard [2]. It is composed
of three distinct fields.
Figure 228. Initial counter block structure for the Counter mode
Nonce 32 bits
Initialization vector (IV) 64 bits
Counter 32 bits
ai16074
• Nonce is a 32-bit, single-use value. A new nonce should be assigned to each different
communication.
• The initialization vector (IV) is a 64-bit value and the standard specifies that the
encryptor must choose IV so as to ensure that a given value is used only once for a
given key
• The counter is a 32-bit big-endian integer that is incremented each time a block has
been encrypted. The initial value of the counter should be set to ‘1’.
The block increments the least significant 32 bits, while it leaves the other (most significant)
96 bits unchanged.
734/1751 RM0090 Rev 19

AES Galois/counter mode (GCM)

The AES Galois/counter mode (GCM) allows encrypting and authenticating the plaintext,
and generating the correspondent ciphertext and tag (also known as message
authentication code or message integrity check). This algorithm is based on AES counter
mode to ensure confidentiality. It uses a multiplier over a fixed finite field to generate the tag.
An initialization vector is required at the beginning of the algorithm.
The message to be processed is split into 2 parts:
• The header (also knows as additional authentication data): data which is authenticated
but no protected (such as information for routing the packet)
• The payload (also knows as plaintext or ciphertext): the message itself which is
authenticated and encrypted.
Note: The header must precede the payload and the two parts cannot be mixed together.
The GCM standard requires to pass, at the end of the message, a specific 128-bit block
composed of the size of the header (64 bits) and the size of the payload (64 bits). During the
computation, the header blocks must be distinguished from the payload blocks.
In GCM mode, four steps are required to perform an encryption/decryption:
RM0090 Rev 19 735/1751

766
1. GCM init phase

During this first step, the HASH key is calculated and saved internally to be used for
processing all the blocks. It is recommended to follow the sequence below:
a) Make sure that the cryptographic processor is disabled by clearing the CRYPEN
bit in the CRYP_CR register.
b) Select the GCM chaining mode by programming ALGOMODE bits to ‘01000’ in
CRYP_CR.
c) Configure GCM_CCMPH bits to ‘00’ in CRYP_CR to start the GCM Init phase.
d) Initialize the key registers (128,192 and 256 bits) in CRYP_KEYRx as well as the
initialization vector (IV).
e) Set CRYPEN bit to ‘1’ to start the calculation of the HASH key.
f) Wait for the CRYPEN bit to be cleared to ‘0’ before moving on to the next phase.
g) Set the CRYPEN bit to ‘1’.
2. GCM header phase
This step must be performed after the GCM Init phase:
h) Set the GCM_CCMPH bits to ‘01’ in CRYP_CR to indicate that the header phase
has started.
i) Write the header data. Three methods can be used:
– Program the data by blocks of 32 bits into the CRYP_DIN register, and use the
IFNF flag to determine if the input FIFO can receive data. The size of the header
must be a multiple of 128 bits (4 words).
– Program the data into the CRYP_DIN register by blocks of 8 words, and use the
IFEM flag to determine if the input FIFO can receive data (IFEM=’1’). The size of
the header must be a multiple of 128 bits (4 words).
– Use the DMA.
j) Once all header data have been supplied, wait until the BUSY bit is cleared in the
CRYP_SR register.
3. GCM payload phase (encryption/decryption)
This step must be performed after the GCM header phase:
k) Configure GCM_CCMPH to ‘10’ in the CRYP_CR register.
l) Select the algorithm direction (encryption or decryption) by using the ALGODIR bit
in CRYP_CR.
m) Program the payload message into the CRYP_DIN register, and use the IFNF flag
to determine if the input FIFO can receive data. Alternatively, the data could be
programmed into the CRYP_DIN register by blocks of 8 words and the IFEM flag
used to determine if the input FIFO can receive data (IFEM=’1’). In parallel, the
736/1751 RM0090 Rev 19

OFNE/OFFU flag of the CRYP_DOUT register can be monitored to check if the

output FIFO is not empty.
n) Repeat the previous step until all payload blocks have been encrypted or
decrypted. Alternatively, DMA could be used.
4. GCM final phase
This step generates the authentication tag:
o) Configure GCM_CCMPH[1:0] to ‘11’ in CRYP_CR.
p) Write the input into the CRYP_DIN register 4 times. The input must contain the
number of bits in the header (64 bits) concatenated with the number of bits in the
payload (64 bits).
q) Wait till the OFNE flag (FIFO output not empty) is set to ‘1’ in the CRYP_SR
register.
r) Read the CRYP_DOUT register 4 times: the output corresponds to the
authentication tag.
s) Disable the cryptographic processor (CRYPEN bit in CRYP_CR = ‘0’)
Note: When a decryption is performed, it is not required to compute the key at the beginning. At
the end of the decryption, the generated tag should be compared with the expected tag
passed with the message. In addition, the ALGODIR bit (algorithm direction) must be set to ‘1’.
No need to disable/enable CRYP processor when moving from header phase to tag phase.
AES Galois message authentication code (GMAC)

The cryptographic processor also supports GMAC to authenticate the plaintext. It uses the
GCM algorithm and a multiplier over a fixed finite field to generate the corresponding tag.
An initialization vector is required at the beginning of the algorithm.
Actually, the GMAC algorithm corresponds to the GCM algorithm applied on a message
composed of the header only. As a consequence, the payload phase is not required.
AES combined cipher machine (CCM)

The CCM algorithm allows encrypting and authenticating the plaintext, as well as generating
the corresponding ciphertext and tag (also known as message authentication code or
message integrity check). This algorithm is based on AES counter mode to ensure
confidentiality. It uses the AES CBC mode to generate a 128-bit tag.
The CCM standard (RFC 3610 Counter with CBC-MAC (CCM) dated September 2003)
defines particular encoding rules for the first authentication block (called B0 in the standard).
In particular, the first block includes flags, a nonce and the payload length expressed in
bytes. The CCM standard specifies another format, called A or counter, for
encryption/decryption. The counter is incremented during the payload phase and its 32 LSB
bits are initialized to ‘1’ during the tag generation (called A0 packet in the CCM standard).
Note: The hardware does not perform the formatting operation of the B0 packet. It should be
handled by the software.
As for the GCM algorithm, the message to be processed is split into 2 parts:
• The header (also knows as additional authentication data): data which is authenticated
but no protected (such as information for routing the packet)
• The payload (also knows as plaintext or ciphertext): the message itself which is
authenticated and encrypted.
RM0090 Rev 19 737/1751

766
Note: The header part must precede the payload and the two parts cannot be mixed together.
In CCM mode, 4 steps are required to perform and encryption or decryption:
1. CCM init phase
In this first step, the B0 packet of the CCM message (1st packet) is programmed into
the CRYP_DIN register. During this phase, the CRYP_DOUT register does not contain
any output data.
The following sequence must be followed:
a) Make sure that the cryptographic processor is disabled by clearing the CRYPEN
bit in the CRYP_CR register.
b) Select the CCM chaining mode by programming the ALGOMODE bits to ‘01001’
in the CRYP_CR register.
c) Configure the GCM_CCMPH bits to ‘00’ in CRYP_CR to start the CCM Init phase.
d) Initialize the key registers (128,192 and 256 bits) in CRYP_KEYRx as well as the
initialization vector (IV).
e) Set the CRYPEN bit to ‘1’ in CRYP_CR.
f) Program the B0 packet into the input data register.
g) Wait for the CRYPEN bit to be cleared before moving on to the next phase.
h) Set CRYPEN to ‘1’.
2. CCM header phase
This step must be performed after the CCM Init phase. The sequence is identical for
encryption and decryption.
During this phase, the CRYP_DOUT register does not contain any output data.
This phase can be skipped if there is no additional authenticated data.
i) Set the GCM_CCMPH bit to ’01’ in CRYP_CR to indicate that the header phase
has started.
j) Three methods can be used:
– Program the header data by blocks of 32 bits into the CRYP_DIN register, and use
the IFNF flag to determine if the input FIFO can receive data. The size of the
header must be a multiple of 128 bits (4 words).
– Program the header data into the CRYP_DIN register by blocks of 8 words, and
use the IFEM flag to determine if the input FIFO can receive data (IFEM=’1’). The
size of the header must be a multiple of 128 bits (4 words).
– Use the DMA.
738/1751 RM0090 Rev 19

Note: The first block B1 must be formatted with the header length. This task should be handled by
software.
k) Once all header data have been supplied, wait until the BUSY flag is cleared.
3. CCM payload phase (encryption/decryption)
This step must be performed after the CCM header phase. During this phase, the
encrypted/decrypted payload is stored in the CRYP_DOUT register.
l) Configure GCM_CCMPH bits to ‘10’ in CRYP_CR.
m) Select the algorithm direction (encryption or decryption) by using the ALGODIR bit
in CRYP_CR.
n) Program the payload message into the CRYP_DIN register, and use the IFNF flag
to determine if the input FIFO can receive data. Alternatively, the data could be
programmed into the CRYP_DIN register by blocks of 8 words and the IFEM flag
used to determine if the input FIFO can receive data (IFEM=’1’). In parallel, the
OFNE/OFFU flag of the CRYP_DOUT register can be monitored to check if the
output FIFO is not empty.
o) Repeat the previous step until all payload blocks have been encrypted or
decrypted. Alternatively, DMA could be used.
4. CCM final phase
This step generates the authentication tag. During this phase, the authentication tag of
the message is generated and stored in the CRYP_DOUT register.
p) Configure GCM_CCMPH[1:0] bits to ‘11’ in CRYP_CR.
q) Load the A0 initialized counter, and program the 128-bit A0 value by writing 4
times 32 bits into the CRYP_DIN register.
r) Wait till the OFNE flag (FIFO output not empty) is set to ‘1’ in the CRYP_SR
register.
s) Read the CRYP_DOUT register 4 times: the output corresponds to the encrypted
authentication tag.
t) Disable the cryptographic processor (CRYPEN bit in CRYP_CR = ‘0’)
Note: The hardware does not perform the formatting of the original B0 and B1 packets and the tag
comparison between encryption and decryption. They have to be handled by software.
The cryptographic processor does not need to be disabled/enabled when moving from the
header phase to the tag phase.
AES cipher message authentication code (CMAC)
The CMAC algorithm allows authenticating the plaintext, and generating the corresponding
tag. The CMAC sequence is identical to the CCM one, except that the payload phase is
skipped.
23.3.3 Data type

Data enter the CRYP processor 32 bits (word) at a time as they are written into the
CRYP_DIN register. The principle of the DES is that streams of data are processed 64 bits
by 64 bits and, for each 64-bit block, the bits are numbered from M1 to M64, with M1 the left-
most bit and M64 the right-most bit of the block. The same principle is used for the AES, but
with a 128-bit block size.
RM0090 Rev 19 739/1751

766
The system memory organization is little-endian: whatever the data type (bit, byte, 16-bit
half-word, 32-bit word) used, the least-significant data occupy the lowest address locations.
A bit, byte, or half-word swapping operation (depending on the kind of data to be encrypted)
therefore has to be performed on the data read from the IN FIFO before they enter the
CRYP processor. The same swapping operation should be performed on the CRYP data
before they are written into the OUT FIFO. For example, the operation would be byte
swapping for an ASCII text stream.
The kind of data to be processed is configured with the DATATYPE bitfield in the CRYP
control register (CRYP_CR).
Table 113. Data types

DATATYPE in System memory data
Swapping performed
CRYP_CR (plaintext or cypher)
Example: TDES block value 0xABCD77206973FE01 is

represented in system memory as:
system memory
00b No swapping TDES block size = 64bit = 2x 32 bit
0xABCD7720 @
0xABCD7720 6973FE01 @+4
0x6973FE01

Half-word (16-bit) system memory

01b
swapping TDES block size = 64bit = 2x 32 bit
0x7720 ABCD @
0xABCD 7720 6973 FE01
0xFE01 6973 @+4

system memory
10b Byte (8-bit) swapping TDES block size = 64bit = 2x 32 bit
0x 20 77 CD AB @
0xAB CD 77 20 69 73 FE 01 @+4
0x 01 FE 73 69
TDES block value 0x4E6F772069732074 is represented in system

memory as:
TDES Bloc size = 64bit = 2x 32 bit system memory
11b Bit swapping 0x4E 6F 77 20 69 73 20 74 0x04 EE F6 72 @
@+4
0x2E 04 CE 96
0000 0100 1110 1110 1111 0110 0111 0010 @
0100 1110 0110 1111 0111 0111 0010 0000
0110 1001 0111 0011 0010 0000 0111 0100
0010 1110 0000 0100 1100 1110 1001 0110 @+4
Figure 229 shows how the 64-bit data block M1...64 is constructed from two consecutive 32-
bit words popped off the IN FIFO by the CRYP processor, according to the DATATYPE
value. The same schematic can easily be extended to form the 128-bit block for the AES
740/1751 RM0090 Rev 19

cryptographic algorithm (for the AES, the block length is four 32-bit words, but swapping
only takes place at word level, so it is identical to the one described here for the TDES).
Note: The same swapping is performed between the IN FIFO and the CRYP data block, and
between the CRYP data block and the OUT FIFO.
Figure 229. 64-bit block construction according to DATATYPE

bit swapping operation DATATYPE = 11b
bit 31 bit 30 bit 2 bit 1 bit 0 second word written into the CRYP_DIN register
IN FIFO bit 31 bit 30 bit 2 bit 1 bit 0 first word written into the CRYP_DIN register
bit string M1 M2 M30 M31 M32 M33 M34 M62 M63 M64
byte swapping operation DATATYPE = 10b

Byte 1,3 Byte 1,2 Byte 1,1 Byte 1,0
bits 7...0 bits 7...0 bits 7...0 bits 7...0 second word written into the CRYP_DIN register
Byte 0,3 Byte 0,2 Byte 0,1 Byte 0,0 first word written into the CRYP_DIN register
IN FIFO
bits 7...0 bits 7...0 bits 7...0 bits 7...0
(bit ordering within byte is unchanged)
bit string Byte 0,0 Byte 0,1 Byte 0,2 Byte 0,3 Byte 1,0 Byte 1,1 Byte 1,2 Byte 1,3
M1...8 M9...16 M17...24 M25...32 M33...40 M41...48 M49...56 M57...64
hald-word swapping operation DATATYPE = 01b

Half-word 1,1 Half-word 1,0
second word written into the CRYP_DIN register
bits 15...0 bits 15...0
Half-word 0,1 Half-word 0,0 first word written into the CRYP_DIN register
IN FIFO
bits 15...0 bits 15...0
bit string Half-word 0,0 Half-word 0,1 Half-word 1,0 Half-word 1,1
M1...16 M17...32 M33...48 M49...64
No swapping operation DATATYPE = 00b

Word 1
second word written into the CRYP_DIN register
bits 31...0
Word 0 first word written into the CRYP_DIN register
IN FIFO
bits 31...0
bit string Word 0 Word 1

M1...32 M33...64
ai16075
RM0090 Rev 19 741/1751

766
23.3.4 Initialization vectors - CRYP_IV0...1(L/R)

Initialization vectors are considered as two 64-bit data items. They therefore do not have the
same data format and representation in system memory as plaintext or cypher data, and
they are not affected by the DATATYPE value.
Initialization vectors are defined by two consecutive 32-bit words, CRYP_IVL (left part,
noted as bits IV1...32) and CRYP_IVR (right part, noted as bits IV33...64).
During the DES or TDES CBC encryption, the CRYP_IV0(L/R) bits are XORed with the 64-
bit data block popped off the IN FIFO after swapping (according to the DATATYPE value),
that is, with the M1...64 bits of the data block. When the output of the DEA3 block is
available, it is copied back into the CRYP_IV0(L/R) vector, and this new content is XORed
with the next 64-bit data block popped off the IN FIFO, and so on.
During the DES or TDES CBC decryption, the CRYP_IV0(L/R) bits are XORed with the 64-
bit data block (that is, with the M1...64 bits) delivered by the TDEA1 block before swapping
(according to the DATATYPE value), and pushed into the OUT FIFO. When the XORed
result is swapped and pushed into the OUT FIFO, the CRYP_IV0(L/R) value is replaced by
the output of the IN FIFO, then the IN FIFO is popped, and a new 64-bit data block can be
processed.
During the AES CBC encryption, the CRYP_IV0...1(L/R) bits are XORed with the 128-bit
data block popped off the IN FIFO after swapping (according to the DATATYPE value).
When the output of the AES core is available, it is copied back into the CRYP_IV0...1(L/R)
vector, and this new content is XORed with the next 128-bit data block popped off the IN
FIFO, and so on.
During the AES CBC decryption, the CRYP_IV0...1(L/R) bits are XORed with the 128-bit
data block delivered by the AES core before swapping (according to the DATATYPE value)
and pushed into the OUT FIFO. When the XORed result is swapped and pushed into the
OUT FIFO, the CRYP_IV0...1(L/R) value is replaced by the output of the IN FIFO, then the
IN FIFO is popped, and a new 128-bit data block can be processed.
During the AES CTR encryption or decryption, the CRYP_IV0...1(L/R) bits are encrypted by
the AES core. Then the result of the encryption is XORed with the 128-bit data block popped
off the IN FIFO after swapping (according to the DATATYPE value). When the XORed result
is swapped and pushed into the OUT FIFO, the counter part of the CRYP_IV0...1(L/R) value
(32 LSB) is incremented.
Any write operation to the CRYP_IV0...1(L/R) registers when bit BUSY = 1b in the
CRYP_SR register is disregarded (CRYP_IV0...1(L/R) register content not modified). Thus,
you must check that bit BUSY = 0b before modifying initialization vectors.
742/1751 RM0090 Rev 19

Figure 230. Initialization vectors use in the TDES-CBC encryption

TDES-CBC en cryption ex ample, DATATYPE = 11b
bit 31 bit 30 bit 2 bit 1 bit 0 second word written into the CRYP_DIN register
bit 31 bit 30 bit 2 bit 1 bit 0 first word written into the CRYP_DIN register
bit string M1 M2 M30 M31 M32 M33 M34 M62 M63 M64
CRYP_IVL CRYP_IVR
31 30 2 1 0 31 30 2 1 0
IV1 IV2 IV30 IV31 IV32 IV33 IV34 IV62 IV63 IV64
I1 I2 I30 I31 I32 I33 I34 I62 I63 I64
DEA Encrypt, K1
DEA Decrypt, K2
DEA Encrypt, K3
CRYP_IVL CRYP_IVR
31 30 2 1 0 31 30 2 1 0
IV1 IV2 IV30 IV31 IV32 IV33 IV34 IV62 IV63 IV64 CRYP result is copied
back to the CRYP_IVL/R
registers after cyphering
OUT FIFO
First word from the OUT FIFO contains the left part of the cyphertext block (O1...32)
Second word from OUT FIFO contains the right part of cyphertext block (O33...64)
ai16076
23.3.5 CRYP busy state

When there is enough data in the input FIFO (at least 2 words for the DES or TDES
algorithm mode, 4 words for the AES algorithm mode) and enough free-space in the output
FIFO (at least 2 (DES/TDES) or 4 (AES) word locations), and when the bit CRYPEN = 1 in
the CRYP_CR register, then the cryptographic processor automatically starts an encryption
or decryption process (according to the value of the ALGODIR bit in the CRYP_CR
register).
This process takes 48 AHB2 clock cycles for the Triple-DES algorithm, 16 AHB2 clock
cycles for the simple DES algorithm, and 14, 16 or 18 AHB2 clock cycles for the AES with
key lengths of 128, 192 or 256 bits, respectively. During the whole process, the BUSY bit in
the CRYP_SR register is set to ‘1’. At the end of the process, two (DES/TDES) or four (AES)
words are written by the CRYP Core into the output FIFO, and the BUSY bit is cleared. In
RM0090 Rev 19 743/1751

766
the CBC, CTR mode, the initialization vectors CRYP_IVx(L/R)R (x = 0..3) are updated as
well.
A write operation to the key registers (CRYP_Kx(L/R)R, x = 0..3), the initialization registers
(CRYP_IVx(L/R)R, x = 0..3), or to bits [9:2] in the CRYP_CR register are ignored when the
cryptographic processor is busy (bit BUSY = 1b in the CRYP_SR register), and the registers
are not modified. It is thus not possible to modify the configuration of the cryptographic
processor while it is processing a block of data. It is however possible to clear the CRYPEN
bit while BUSY = 1, in which case the ongoing DES, TDES or AES processing is completed
and the two/four word results are written into the output FIFO, and then, only then, the
BUSY bit is cleared.
Note: When a block is being processed in the DES or TDES mode, if the output FIFO becomes full
and if the input FIFO contains at least one new block, then the new block is popped off the
input FIFO and the BUSY bit remains high until there is enough space to store this new
block into the output FIFO.
23.3.6 Procedure to perform an encryption or a decryption

Initialization
1. Initialize the peripheral (the order of operations is not important except for the key
preparation for AES-ECB or AES-CBC decryption. The key size and the key value
must be entered before preparing the key and the algorithm must be configured once
the key has been prepared):
a) Configure the key size (128-, 192- or 256-bit, in the AES only) with the KEYSIZE
bits in the CRYP_CR register
b) Write the symmetric key into the CRYP_KxL/R registers (2 to 8 registers to be
written depending on the algorithm)
c) Configure the data type (1-, 8-, 16- or 32-bit), with the DATATYPE bits in the
CRYP_CR register
d) In case of decryption in AES-ECB or AES-CBC, you must prepare the key:
configure the key preparation mode by setting the ALGOMODE bits to ‘111’ in the
CRYP_CR register. Then write the CRYPEN bit to ‘1’: the BUSY bit is set. Wait
until BUSY returns to 0 (CRYPEN is automatically cleared as well): the key is
prepared for decryption
e) Configure the algorithm and chaining (the DES/TDES in ECB/CBC, the AES in
ECB/CBC/CTR/GCM/CCM) with the ALGOMODE bits in the CRYP_CR register
f) Configure the direction (encryption/decryption), with the ALGODIR bit in the
CRYP_CR register
g) Write the initialization vectors into the CRYP_IVxL/R register (in CBC or CTR
modes only)
2. Flush the IN and OUT FIFOs by writing the FFLUSH bit to 1 in the CRYP_CR register
Processing when the DMA is used to transfer the data from/to the memory
1. Configure the DMA controller to transfer the input data from the memory. The transfer
length is the length of the message. As message padding is not managed by the
peripheral, the message length must be an entire number of blocks. The data are
transferred in burst mode. The burst length is 4 words in the AES and 2 or 4 words in
744/1751 RM0090 Rev 19

the DES/TDES. The DMA should be configured to set an interrupt on transfer

completion of the output data to indicate that the processing is finished.
2. Enable the cryptographic processor by writing the CRYPEN bit to 1. Enable the DMA
requests by setting the DIEN and DOEN bits in the CRYP_DMACR register.
3. All the transfers and processing are managed by the DMA and the cryptographic
processor. The DMA interrupt indicates that the processing is complete. Both FIFOs
are normally empty and BUSY = 0.
Processing when the data are transferred by the CPU during interrupts
1. Enable the interrupts by setting the INIM and OUTIM bits in the CRYP_IMSCR register.
2. Enable the cryptographic processor by setting the CRYPEN bit in the CRYP_CR
register.
3. In the interrupt managing the input data: load the input message into the IN FIFO. You
can load 2 or 4 words at a time, or load data until the FIFO is full. When the last word of
the message has been entered into the FIFO, disable the interrupt by clearing the INIM
bit.
4. In the interrupt managing the output data: read the output message from the OUT
FIFO. You can read 1 block (2 or 4 words) at a time or read data until the FIFO is
empty. When the last word has been read, INIM=0, BUSY=0 and both FIFOs are empty
(IFEM=1 and OFNE=0). You can disable the interrupt by clearing the OUTIM bit and,
the peripheral by clearing the CRYPEN bit.
Processing without using the DMA nor interrupts

1. Enable the cryptographic processor by setting the CRYPEN bit in the CRYP_CR
register.
2. Write the first blocks in the input FIFO (2 to 8 words).
3. Repeat the following sequence until the complete message has been processed:
a) Wait for OFNE=1, then read the OUT-FIFO (1 block or until the FIFO is empty)
b) Wait for IFNF=1, then write the IN FIFO (1 block or until the FIFO is full)
4. At the end of the processing, BUSY=0 and both FIFOs are empty (IFEM=1 and
OFNE=0). You can disable the peripheral by clearing the CRYPEN bit.
23.3.7 Context swapping

If a context switching is needed because a new task launched by the OS requires this
resource, the following tasks have to be performed for full context restoration (example
when the DMA is used):
RM0090 Rev 19 745/1751

766
Case of the AES and DES

1. Context saving
a) Stop DMA transfers on the IN FIFO by clearing the DIEN bit in the CRYP_DMACR
register.
b) Wait until both the IN and OUT FIFOs are empty (IFEM=1 and OFNE=0 in the
CRYP_SR register) and the BUSY bit is cleared.
c) Stop DMA transfers on the OUT FIFO by writing the DOEN bit to 0 in the
CRYP_DMACR register and clear the CRYPEN bit.
d) Save the current configuration (bits [9:2] and bits 19 in the CRYP_CR register)
and, if not in ECB mode, the initialization vectors. The key value must already be
available in the memory. When needed, save the DMA status (pointers for IN and
OUT messages, number of remaining bytes, etc.).
Additional bits should be saved when GCM/GMAC or CCM/CMAC algorithms are
used:
– bits [17:16] in the CRYP_CR register
– context swap registers:
CRYP_CSGCMCCM0..7 for GCM/GMAC or CCM/CMAC algorithm
CRYP_CSGCM0..7 for GCM/GMAC algorithm.
2. Configure and execute the other processing.
3. Context restoration
a) Configure the processor as in Section 23.3.6: Procedure to perform an encryption
or a decryption on page 744, Initialization with the saved configuration. For the
AES-ECB or AES-CBC decryption, the key must be prepared again.
b) If needed, reconfigure the DMA controller to transfer the rest of the message.
c) Enable the processor by setting the CRYPEN bit and, the DMA requests by setting
the DIEN and DOEN bits.
Case of the TDES

Context swapping can be done in the TDES in the same way as in the AES. But as the input
FIFO can contain up to 4 unprocessed blocks and as the processing duration per block is
higher, it can be faster in certain cases to interrupt the processing without waiting for the IN
FIFO to be empty.
1. Context saving
a) Stop DMA transfers on the IN FIFO by clearing the DIEN bit in the CRYP_DMACR
register.
b) Disable the processor by clearing the CRYPEN bit (the processing will stop at the
end of the current block).
c) Wait until the OUT FIFO is empty (OFNE=0 in the CRYP_SR register) and the
BUSY bit is cleared.
d) Stop DMA transfers on the OUT FIFO by writing the DOEN bit to 0 in the
CRYP_DMACR register.
e) Save the current configuration (bits [9:2] and bits 19 in the CRYP_CR register)
and, if not in ECB mode, the initialization vectors. The key value must already be
available in the memory. When needed, save the DMA status (pointers for IN and
OUT messages, number of remaining bytes, etc.). Read back the data loaded in
746/1751 RM0090 Rev 19

the IN FIFO that have not been processed and save them in the memory until the
FIFO is empty.
Note: In GCM/GMAC or CCM/CMAC mode, bits [17:16] of the CRYP_CR register should also be
saved.
2. Configure and execute the other processing.
3. Context restoration
a) Configure the processor as in Section 23.3.6: Procedure to perform an encryption
or a decryption on page 744, Initialization with the saved configuration. For the
AES-ECB or AES-CBC decryption, the key must be prepared again.
b) Write the data that were saved during context saving into the IN FIFO.
c) If needed, reconfigure the DMA controller to transfer the rest of the message.
d) Enable the processor by setting the CRYPEN bit and, the DMA requests by setting
the DIEN and DOEN bits.
23.4 CRYP interrupts

There are two individual maskable interrupt sources generated by the CRYP. These two
sources are combined into a single interrupt signal, which is the only interrupt signal from
the CRYP that drives the NVIC (nested vectored interrupt controller). This combined
interrupt, which is an OR function of the individual masked sources, is asserted if any of the
individual interrupts listed below is asserted and enabled.
You can enable or disable the interrupt sources individually by changing the mask bits in the
CRYP_IMSCR register. Setting the appropriate mask bit to ‘1’ enables the interrupt.
The status of the individual interrupt sources can be read either from the CRYP_RISR
register, for raw interrupt status, or from the CRYP_MISR register, for the masked interrupt
status.
Output FIFO service interrupt - OUTMIS

The output FIFO service interrupt is asserted when there is one or more (32-bit word) data
items in the output FIFO. This interrupt is cleared by reading data from the output FIFO until
there is no valid (32-bit) word left (that is, the interrupt follows the state of the OFNE (output
FIFO not empty) flag).
The output FIFO service interrupt OUTMIS is NOT enabled with the CRYP enable bit.
Consequently, disabling the CRYP will not force the OUTMIS signal low if the output FIFO is
not empty.
Input FIFO service interrupt - INMIS

The input FIFO service interrupt is asserted when there are less than four words in the input
FIFO. It is cleared by performing write operations to the input FIFO until it holds four or more
words.
The input FIFO service interrupt INMIS is enabled with the CRYP enable bit. Consequently,
when CRYP is disabled, the INMIS signal is low even if the input FIFO is empty.
RM0090 Rev 19 747/1751

766
Figure 231. CRYP interrupt mapping diagram

CRYPEN IN FIFO Interrupt - INMIS
INRIS INMIS
INIM Global Interrupt
OUTRIS OUTMIS
OUT FIFO Interrupt - OUTMIS
OUTIM ai16077
23.5 CRYP DMA interface

The cryptographic processor provides an interface to connect to the DMA controller. The
DMA operation is controlled through the CRYP DMA control register, CRYP_DMACR.
The burst and single transfer request signals are not mutually exclusive. They can both be
asserted at the same time. For example, when there are 6 words available in the OUT FIFO,
the burst transfer request and the single transfer request are asserted. After a burst transfer
of 4 words, the single transfer request only is asserted to transfer the last 2 available words.
This is useful for situations where the number of words left to be received in the stream is
less than a burst.
Each request signal remains asserted until the relevant DMA clear signal is asserted. After
the request clear signal is deasserted, a request signal can become active again, depending
on the above described conditions. All request signals are deasserted if the CRYP
peripheral is disabled or the DMA enable bit is cleared (DIEN bit for the IN FIFO and DOEN
bit for the OUT FIFO in the CRYP_DMACR register).
Note: The DMA controller must be configured to perform burst of 4 words or less. Otherwise some
data could be lost.
In order to let the DMA controller empty the OUT FIFO before filling up the IN FIFO, the
OUTDMA channel should have a higher priority than the INDMA channel.
23.6 CRYP registers

The cryptographic core is associated with several control and status registers, eight key
registers and four initialization vectors registers.
23.6.1 CRYP control register (CRYP_CR) for STM32F415/417xx

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CRYPEN FFLUSH KEYSIZE DATATYPE ALGOMODE[2:0] ALGODIR Res. Res.
Reserved
rw w rw rw rw rw rw rw rw rw
748/1751 RM0090 Rev 19


Bit 15 CRYPEN: Cryptographic processor enable
0: CRYP processor is disabled
1: CRYP processor is enabled
Note: The CRYPEN bit is automatically cleared by hardware when the key
preparation process ends (ALGOMODE=111b) or GCM_CCM init Phase
Bit 14 FFLUSH: FIFO flush
When CRYPEN = 0, writing this bit to 1 flushes the IN and OUT FIFOs (that is
read and write pointers of the FIFOs are reset. Writing this bit to 0 has no effect.
When CRYPEN = 1, writing this bit to 0 or 1 has no effect.
Reading this bit always returns 0.
Bits 9:8 KEYSIZE[1:0]: Key size selection (AES mode only)
This bitfield defines the bit-length of the key used for the AES cryptographic core.
This bitfield is ‘don’t care’ in the DES or TDES modes.
00: 128 bit key length
11: Reserved, do not use this value
Bits 7:6 DATATYPE[1:0]: Data type selection
This bitfield defines the format of data entered in the CRYP_DIN register (refer to
Section 23.3.3: Data type).
00: 32-bit data. No swapping of each word. First word pushed into the IN FIFO
(or popped off the OUT FIFO) forms bits 1...32 of the data block, the second
word forms bits 33...64.
01: 16-bit data, or half-word. Each word pushed into the IN FIFO (or popped off
the OUT FIFO) is considered as 2 half-words, which are swapped with each
other.
10: 8-bit data, or bytes. Each word pushed into the IN FIFO (or popped off the
OUT FIFO) is considered as 4 bytes, which are swapped with each other.
11: bit data, or bit-string. Each word pushed into the IN FIFO (or popped off the
OUT FIFO) is considered as 32 bits (1st bit of the string at position 0), which are
swapped with each other.
RM0090 Rev 19 749/1751

766
Bits 5:3 ALGOMODE[2:0]: Algorithm mode

000: TDES-ECB (triple-DES Electronic codebook): no feedback between blocks
of data. Initialization vectors (CRYP_IV0(L/R)) are not used, three key vectors
(K1, K2, and K3) are used (K0 is not used).
001: TDES-CBC (triple-DES Cipher block chaining): output block is XORed with
the subsequent input block before its entry into the algorithm. Initialization
vectors (CRYP_IV0L/R) must be initialized, three key vectors (K1, K2, and K3)
are used (K0 is not used).
010: DES-ECB (simple DES Electronic codebook): no feedback between blocks
of data. Initialization vectors (CRYP_IV0L/R) are not used, only one key vector
(K1) is used (K0, K2, K3 are not used).
011: DES-CBC (simple DES Cipher block chaining): output block is XORed with
vectors (CRYP_IV0L/R) must be initialized. Only one key vector (K1) is used
(K0, K2, K3 are not used).
100: AES-ECB (AES Electronic codebook): no feedback between blocks of data.
Initialization vectors (CRYP_IV0L/R...1L/R) are not used. All four key vectors
(K0...K3) are used.
101: AES-CBC (AES Cipher block chaining): output block is XORed with the
subsequent input block before its entry into the algorithm. Initialization vectors
(CRYP_IV0L/R...1L/R) must be initialized. All four key vectors (K0...K3) are
used.
110: AES-CTR (AES counter mode): output block is XORed with the subsequent
input block before its entry into the algorithm. Initialization vectors
used. CTR decryption does not differ from CTR encryption, since the core
always encrypts the current counter block to produce the key stream that will be
XORed with the plaintext or cipher in input. Thus, ALGODIR is don’t care when
ALGOMODE = 110b, and the key must NOT be unrolled (prepared) for
decryption.
111: AES key preparation for decryption mode. Writing this value when CRYPEN
= 1 immediately starts an AES round for key preparation. The secret key must
have previously been loaded into the K0...K3 registers. The BUSY bit in the
CRYP_SR register is set during the key preparation. After key processing, the
resulting key is copied back into the K0...K3 registers, and the BUSY bit is
cleared.
Bit 2 ALGODIR: Algorithm direction
0: Encrypt
1: Decrypt
Note: Writing to the KEYSIZE, DATATYPE, ALGOMODE and ALGODIR bits while BUSY=1 has
no effect. These bits can only be configured when BUSY=0.
The FFLUSH bit has to be set only when BUSY=0. If not, the FIFO is flushed, but the block
being processed may be pushed into the output FIFO just after the flush operation, resulting
in a nonempty FIFO condition.
23.6.2 CRYP control register (CRYP_CR) for STM32F415/417xx

750/1751 RM0090 Rev 19

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
ALGO
MODE GCM_CCMPH
Reserved [3] Res.
rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CRYPEN FFLUSH KEYSIZE DATATYPE ALGOMODE[2:0] ALGODIR
Reserved Reserved
rw w rw rw rw rw rw rw rw rw
Bits 31:20 Reserved, forced by hardware to 0.

Bit 18 Reserved, forced by hardware to 0.
Bits 17:16 GCM_CCMPH[1:0]: no effect if “GCM or CCM algorithm” is not set
00: GCM_CCM init Phase
01: GCM_CCM header phase
10: GCM_CCM payload phase
11: GCM_CCM final phase
Bit 15 CRYPEN: Cryptographic processor enable
0: CRYP processor is disabled
1: CRYP processor is enabled
Note: The CRYPEN bit is automatically cleared by hardware when the key
preparation process ends (ALGOMODE=111b) or GCM_CCM init Phase
Bit 14 FFLUSH: FIFO flush
When CRYPEN = 0, writing this bit to 1 flushes the IN and OUT FIFOs (that is
read and write pointers of the FIFOs are reset. Writing this bit to 0 has no effect.
When CRYPEN = 1, writing this bit to 0 or 1 has no effect.
Reading this bit always returns 0.
Bits 9:8 KEYSIZE[1:0]: Key size selection (AES mode only)
This bitfield defines the bit-length of the key used for the AES cryptographic core.
This bitfield is ‘don’t care’ in the DES or TDES modes.
11: Reserved, do not use this value
Bits 7:6 DATATYPE[1:0]: Data type selection
This bitfield defines the format of data entered in the CRYP_DIN register (refer to
Section 23.3.3: Data type).
00: 32-bit data. No swapping of each word. First word pushed into the IN FIFO
(or popped off the OUT FIFO) forms bits 1...32 of the data block, the second
word forms bits 33...64.
01: 16-bit data, or half-word. Each word pushed into the IN FIFO (or popped off
the OUT FIFO) is considered as 2 half-words, which are swapped with each
other.
10: 8-bit data, or bytes. Each word pushed into the IN FIFO (or popped off the
OUT FIFO) is considered as 4 bytes, which are swapped with each other.
11: bit data, or bit-string. Each word pushed into the IN FIFO (or popped off the
OUT FIFO) is considered as 32 bits (1st bit of the string at position 0), which are
swapped with each other.
RM0090 Rev 19 751/1751

766
Bits 19 and 5:3 ALGOMODE[3:0]: Algorithm mode

0000: TDES-ECB (triple-DES Electronic codebook): no feedback between
blocks of data. Initialization vectors (CRYP_IV0(L/R)) are not used, three key
vectors (K1, K2, and K3) are used (K0 is not used).
0001: TDES-CBC (triple-DES Cipher block chaining): output block is XORed
with the subsequent input block before its entry into the algorithm. Initialization
vectors (CRYP_IV0L/R) must be initialized, three key vectors (K1, K2, and K3)
are used (K0 is not used).
0010: DES-ECB (simple DES Electronic codebook): no feedback between
blocks of data. Initialization vectors (CRYP_IV0L/R) are not used, only one key
vector (K1) is used (K0, K2, K3 are not used).
0011: DES-CBC (simple DES Cipher block chaining): output block is XORed with
vectors (CRYP_IV0L/R) must be initialized. Only one key vector (K1) is used
(K0, K2, K3 are not used).
0100: AES-ECB (AES Electronic codebook): no feedback between blocks of
data. Initialization vectors (CRYP_IV0L/R...1L/R) are not used. All four key
vectors (K0...K3) are used.
0101: AES-CBC (AES Cipher block chaining): output block is XORed with the
used.
0110: AES-CTR (AES Counter mode): output block is XORed with the
used. CTR decryption does not differ from CTR encryption, since the core
always encrypts the current counter block to produce the key stream that will be
XORed with the plaintext or cipher in input. Thus, ALGODIR is don’t care when
ALGOMODE = 110b, and the key must NOT be unrolled (prepared) for
decryption.
0111: AES key preparation for decryption mode. Writing this value when
CRYPEN = 1 immediately starts an AES round for key preparation. The secret
key must have previously been loaded into the K0...K3 registers. The BUSY bit
in the CRYP_SR register is set during the key preparation. After key processing,
the resulting key is copied back into the K0...K3 registers, and the BUSY bit is
cleared.
1000: Galois Counter Mode (GCM). This algorithm mode is also used for the
GMAC algorithm.
1001: Counter with CBC-MAC (CCM). This algorithm mode is also used for the
CMAC algorithm.
Bit 2 ALGODIR: Algorithm direction
0: Encrypt
1: Decrypt
Bits 1:0 Reserved, must be kept to 0.
Note: Writing to the KEYSIZE, DATATYPE, ALGOMODE and ALGODIR bits while BUSY=1 has
no effect. These bits can only be configured when BUSY=0.
The FFLUSH bit has to be set only when BUSY=0. If not, the FIFO is flushed, but the block
being processed may be pushed into the output FIFO just after the flush operation, resulting
in a nonempty FIFO condition.
752/1751 RM0090 Rev 19

23.6.3 CRYP status register (CRYP_SR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
BUSY OFFU OFNE IFNF IFEM
Reserved
r r r r r

Bit 4 BUSY: Busy bit
0: The CRYP Core is not processing any data. The reason is either that:
– the CRYP core is disabled (CRYPEN=0 in the CRYP_CR register) and
the last processing has completed, or
– The CRYP core is waiting for enough data in the input FIFO or enough
free space in the output FIFO (that is in each case at least 2 words in
the DES, 4 words in the AES).
1: The CRYP core is currently processing a block of data or a key preparation
(for AES decryption).
Bit 3 OFFU: Output FIFO full
0: Output FIFO is not full
1: Output FIFO is full
Bit 2 OFNE: Output FIFO not empty
0: Output FIFO is empty
1: Output FIFO is not empty
Bit 1 IFNF: Input FIFO not full
0: Input FIFO is full
1: Input FIFO is not full
Bit 0 IFEM: Input FIFO empty
0: Input FIFO is not empty
1: Input FIFO is empty
RM0090 Rev 19 753/1751

766
23.6.4 CRYP data input register (CRYP_DIN)

The CRYP_DIN register is the data input register. It is 32-bit wide. It is used to enter up to
four 64-bit (TDES) or two 128-bit (AES) plaintext (when encrypting) or ciphertext (when
decrypting) blocks into the input FIFO, one 32-bit word at a time.
The first word written into the FIFO is the MSB of the input block. The LSB of the input block
is written at the end. Disregarding the data swapping, this gives:
• In the DES/TDES modes: a block is a sequence of bits numbered from bit 1 (leftmost
bit) to bit 64 (rightmost bit). Bit 1 corresponds to the MSB (bit 31) of the first word
entered into the FIFO, bit 64 corresponds to the LSB (bit 0) of the second word entered
into the FIFO.
• In the AES mode: a block is a sequence of bits numbered from 0 (leftmost bit) to 127
(rightmost bit). Bit 0 corresponds to the MSB (bit 31) of the first word written into the
FIFO, bit 127 corresponds to the LSB (bit 0) of the 4th word written into the FIFO.
To fit different data sizes, the data written in the CRYP_DIN register can be swapped before
being processed by configuring the DATATYPE bits in the CRYP_CR register. Refer to
Section 23.3.3: Data type on page 739 for more details.
When CRYP_DIN register is written to, the data are pushed into the input FIFO. When at
least two 32-bit words in the DES/TDES mode (or four 32-bit words in the AES mode) have
been pushed into the input FIFO, and when at least 2 words are free in the output FIFO, the
CRYP engine starts an encrypting or decrypting process. This process takes two 32-bit
words in the DES/TDES mode (or four 32-bit words in the AES mode) from the input FIFO
and delivers two 32-bit words (or 4, respectively) to the output FIFO per process round.
When CRYP_DIN register is read:
• If CRYPEN = 0, the FIFO is popped, and then the data present in the Input FIFO are
returned, from the oldest one (first reading) to the newest one (last reading). The IFEM
flag must be checked before each read operation to make sure that the FIFO is not
empty.
• if CRYPEN = 1, an undefined value is returned.
After the CRYP_DIN register has been read once or several times, the FIFO must be
flushed by setting the FFLUSH bit prior to processing new data.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DATAIN
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DATAIN
Bits 31:0 DATAIN: Data input

Read = returns Input FIFO content if CRYPEN = 0, else returns an undefined
value.
Write = Input FIFO is written.
754/1751 RM0090 Rev 19

23.6.5 CRYP data output register (CRYP_DOUT)

The CRYP_DOUT register is the data output register. It is read-only and 32-bit wide. It is
used to retrieve up to four 64-bit (TDES mode) or two 128-bit (AES mode) ciphertext (when
encrypting) or plaintext (when decrypting) blocks from the output FIFO, one 32-bit word at a
time.
Like for the input data, the MSB of the output block is the first word read from the output
FIFO. The LSB of the output block is read at the end. Disregarding data swapping, this
gives:
• In the DES/TDES modes: Bit 1 (leftmost bit) corresponds to the MSB (bit 31) of the first
word read from the FIFO, bit 64 (rightmost bit) corresponds to the LSB (bit 0) of the
second word read from the FIFO.
• In the AES mode: Bit 0 (leftmost bit) corresponds to the MSB (bit 31) of the first word
read from the FIFO, bit 127 (rightmost bit) corresponds to the LSB (bit 0) of the 4th
word read from the FIFO.
To fit different data sizes, the data can be swapped after processing by configuring the
DATATYPE bits in the CRYP_CR register. Refer to Section 23.3.3: Data type on page 739
for more details.
When CRYP_DOUT register is read, the last data entered into the output FIFO (pointed to
by the read pointer) is returned.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DATAOUT
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DATAOUT
Bits 31:0 DATAOUT: Data output

Read = returns output FIFO content.
Write = No effect.
RM0090 Rev 19 755/1751

766
23.6.6 CRYP DMA control register (CRYP_DMACR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DOEN DIEN
Reserved
rw rw

Bit 1 DOEN: DMA output enable
0: DMA for outgoing data transfer is disabled
1: DMA for outgoing data transfer is enabled
Bit 0 DIEN: DMA input enable
0: DMA for incoming data transfer is disabled
1: DMA for incoming data transfer is enabled
23.6.7 CRYP interrupt mask set/clear register (CRYP_IMSCR)

The CRYP_IMSCR register is the interrupt mask set or clear register. It is a read/write
register. On a read operation, this register gives the current value of the mask on the
relevant interrupt. Writing 1 to the particular bit sets the mask, enabling the interrupt to be
read. Writing 0 to this bit clears the corresponding mask. All the bits are cleared to 0 when
the peripheral is reset.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OUTIM INIM
Reserved
rw rw

Bit 1 OUTIM: Output FIFO service interrupt mask
0: Output FIFO service interrupt is masked
1: Output FIFO service interrupt is not masked
Bit 0 INIM: Input FIFO service interrupt mask
0: Input FIFO service interrupt is masked
1: Input FIFO service interrupt is not masked
756/1751 RM0090 Rev 19

23.6.8 CRYP raw interrupt status register (CRYP_RISR)

The CRYP_RISR register is the raw interrupt status register. It is a read-only register. On a
read, this register gives the current raw status of the corresponding interrupt prior to
masking. A write has no effect.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OUTRIS INRIS
Reserved
r r

Bit 1 OUTRIS: Output FIFO service raw interrupt status
Gives the raw interrupt state prior to masking of the output FIFO service
interrupt.
0: Raw interrupt not pending
1: Raw interrupt pending
Bit 0 INRIS: Input FIFO service raw interrupt status
Gives the raw interrupt state prior to masking of the Input FIFO service interrupt.
0: Raw interrupt not pending
1: Raw interrupt pending
23.6.9 CRYP masked interrupt status register (CRYP_MISR)

The CRYP_MISR register is the masked interrupt status register. It is a read-only register.
On a read, this register gives the current masked status of the corresponding interrupt prior
to masking. A write has no effect.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OUTMIS INMIS
Reserved
r r
RM0090 Rev 19 757/1751

766

Bit 1 OUTMIS: Output FIFO service masked interrupt status
Gives the interrupt state after masking of the output FIFO service interrupt.
0: Interrupt not pending
1: Interrupt pending
Bit 0 INMIS: Input FIFO service masked interrupt status
Gives the interrupt state after masking of the input FIFO service interrupt.
0: Interrupt not pending
1: Interrupt pending when CRYPEN = 1
23.6.10 CRYP key registers (CRYP_K0...3(L/R)R)

Address offset: 0x20 to 0x3C
These registers contain the cryptographic keys.
In the TDES mode, keys are 64-bit binary values (number from left to right, that is the
leftmost bit is bit 1), named K1, K2 and K3 (K0 is not used), each key consists of 56
information bits and 8 parity bits. The parity bits are reserved for error detection purposes
and are not used by the current block. Thus, bits 8, 16, 24, 32, 40, 48, 56 and 64 of each 64-
bit key value Kx[1:64] are not used.
In the AES mode, the key is considered as a single 128-, 192- or 256-bit long bit sequence,
k0k1k2...k127/191/255 (k0 being the leftmost bit). The AES key is entered into the registers as
follows:
• for AES-128: k0..k127 corresponds to b127..b0 (b255..b128 are not used),
• for AES-192: k0..k191 corresponds to b191..b0 (b255..b192 are not used),
• for AES-256: k0..k255 corresponds to b255..b0.
In any case b0 is the rightmost bit.
CRYP_K0LR (address offset: 0x20)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
b255 b254 b253 b252 b251 b250 b249 b248 b247 b246 b245 b244 b243 b242 b241 b240
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CRYP_K0RR (address offset: 0x24)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
758/1751 RM0090 Rev 19


31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
k1.1 k1.2 k1.3 k1.4 k1.5 k1.6 k1.7 k1.8 k1.9 k1.10 k1.11 k1.12 k1.13 k1.14 k1.15 k1.16
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CRYP_K1RR (address offset: 0x2C)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CRYP_K2RR (address offset: 0x34)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RM0090 Rev 19 759/1751

766
CRYP_K3RR (address offset: 0x3C)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Note: Write accesses to these registers are disregarded when the cryptographic processor is busy
(bit BUSY = 1 in the CRYP_SR register).
23.6.11 CRYP initialization vector registers (CRYP_IV0...1(L/R)R)

The CRYP_IV0...1(L/R)R are the left-word and right-word registers for the initialization
vector (64 bits for DES/TDES and 128 bits for AES) and are used in the CBC (Cipher block
chaining) and Counter (CTR) modes. After each computation round of the TDES or AES
Core, the CRYP_IV0...1(L/R)R registers are updated as described in Section : DES and
TDES Cipher block chaining (DES/TDES-CBC) mode on page 726, Section : AES Cipher
block chaining (AES-CBC) mode on page 730 and Section : AES counter mode (AES-CTR)
mode on page 732.
IV0 is the leftmost bit whereas IV63 (DES, TDES) or IV127 (AES) are the rightmost bits of
the initialization vector. IV1(L/R)R is used only in the AES.
CRYP_IV0LR (address offset: 0x40)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
IV0 IV1 IV2 IV3 IV4 IV5 IV6 IV7 IV8 IV9 IV10 IV11 IV12 IV13 IV14 IV15
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CRYP_IV0RR (address offset: 0x44)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
760/1751 RM0090 Rev 19

CRYP_IV1LR (address offset: 0x48)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CRYP_IV1RR (address offset: 0x4C)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Note: In DES/3DES modes, only CRYP_IV0(L/R) is used.

Write access to these registers are disregarded when the cryptographic processor is busy
(bit BUSY = 1 in the CRYP_SR register).
RM0090 Rev 19 761/1751

766
23.6.12 CRYP context swap registers (CRYP_CSGCMCCM0..7R and

CRYP_CSGCM0..7R) for STM32F42xxx and STM32F43xxx
Address offset:
• CRYP_CSGCMCCM0..7: 0x050 to 0x06C: used for GCM/GMAC or CCM/CMAC
alogrithm only
• CRYP_CSGCM0..7: 0x070 to 0x08C: used for GCM/GMAC algorithm only
These registers contain the complete internal register states of the CRYP processor when
the GCM/GMAC or CCM/CMAC algorithm is selected. They are useful when a context swap
has to be performed because a high-priority task needs the cryptographic processor while it
is already in use by another task.
When such an event occurs, the CRYP_CSGCMCCM0..7R and CRYP_CSGCM0..7R (in
GCM/GMAC mode) or CRYP_CSGCMCCM0..7R (in CCM/CMAC mode) registers have to
be read and the values retrieved have to be saved in the system memory space. The
cryptographic processor can then be used by the preemptive task, and when the
cryptographic computation is complete, the saved context can be read from memory and
written back into the corresponding context swap registers.
Note: These registers are used only when GCM/GMAC or CCM/CMAC algorithm mode is
selected.
CRYP_CSGCMCCMxR: where x=[7:0]

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CRYP_CSGCMCCMxR
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CRYP_CSGCMCCMxR
CRYP_CSGCMxR: where x=[7:0]

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CRYP_CSGCMxR
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CRYP_CSGCMxR
762/1751 RM0090 Rev 19

23.6.13 CRYP register map
Table 114. CRYP register map and reset values for STM32F415/417xx
Register
Offset name and
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
reset value
ALOMODE[2:0]
DATATYPE
ALGODIR
KEYSIZE
CRYPEN
FFLUSH
0x00 CRYP_CR
Res.
Reserved Reserved
0x00
Reset value 0 0 0 0 0 0 0 0 0 0
OFNE
OFFU
BUSY
IFEM
IFNF
CRYP_SR
0x04 Reserved
CRYP_DIN DATAIN
0x08
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_DOUT DATAOUT
0x0C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0
DOEN
CRYP_DMAC
DIEN
0x10 R Reserved
Reset value 0 0
OUTIM
CRYP_IMSC
INIM
0x14 R Reserved
Reset value 0 0
OUTRIS
INRIS
CRYP_RISR
0x18 Reserved
Reset value 0 1
OUTMIS
CRYP_MISR IN%IS
0x1C Reserved
Reset value 0 0
CRYP_K0LR CRYP_K0LR
0x20
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_K0RR CRYP_K0RR
0x24
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
...
CRYP_K3LR CRYP_K3LR
0x38
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_K3RR CRYP_K3RR
0x3C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_IV0LR CRYP_IV0LR
0x40
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RM0090 Rev 19 763/1751

766
Table 114. CRYP register map and reset values for STM32F415/417xx (continued)
Register
Offset name and
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
reset value
CRYP_IV0RR CRYP_IV0RR
0x44
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x48
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x4C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Table 115. CRYP register map and reset values for STM32F43xxx
Register
Offset name
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
reset value
ALOMODE[2:0]
ALGOMODE[3]
GCM_CCMPH
DATATYPE
ALGODIR
KEYSIZE
CRYPEN
FFLUSH
0x00 CRYP_CR
Res..
Res.
Reserved Reserved
0x00
Reset value 0 0 0 0 0 0 0 0 0 0 0 0
OFNE
BUSY
OFFU
IFEM
IFNF
CRYP_SR
0x04 Reserved
CRYP_DIN DATAIN
0x08
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_DOUT DATAOUT
0x0C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 DOEN
CRYP_DMAC DIEN
0x10 R Reserved
Reset value 0 0
OUTIM
CRYP_IMSC
INIM
0x14 R Reserved
Reset value 0 0
OUTRIS
INRIS
CRYP_RISR
0x18 Reserved
Reset value 0 1
764/1751 RM0090 Rev 19

Table 115. CRYP register map and reset values for STM32F43xxx (continued)
Register
Offset name
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
reset value
OUTMIS
IN%IS
CRYP_MISR
0x1C Reserved
Reset value 0 0
CRYP_K0LR CRYP_K0LR
0x20
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_K0RR CRYP_K0RR
0x24
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
...
...
CRYP_K3LR CRYP_K3LR
0x38
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_K3RR CRYP_K3RR
0x3C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x40
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x44
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x48
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x4C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_
CSGCMCCM CRYP_CSGCMCCM0R
0x50 R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_
0x54 1R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_
0x58 2R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_
0x5C 3R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RM0090 Rev 19 765/1751

766
Table 115. CRYP register map and reset values for STM32F43xxx (continued)
Register
Offset name
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
reset value
CRYP_
0x60 4R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_
0x64 5R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_
0x68 6R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_
0x6C 7R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_CSGC
CRYP_CSGCM0R
0x70 M0R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_CSGC
CRYP_CSGCM1R
0x74 M1R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_CSGC
CRYP_CSGCM2R
0x78 M2R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_CSGC
CRYP_CSGCM3R
0x7C M3R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_CSGC
CRYP_CSGCM4R
0x80 M4R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_CSGC
CRYP_CSGCM5R
0x84 M5R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_CSGC
CRYP_CSGCM6R
0x88 M6R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CRYP_CSGC
CRYP_CSGCM7R
0x8C M7R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
766/1751 RM0090 Rev 19

RM0090 Random number generator (RNG)
24 Random number generator (RNG)
24.1 RNG introduction

The RNG processor is a random number generator, based on a continuous analog noise,
that provides a random 32-bit value to the host when read.
The RNG passed the FIPS PUB 140-2 (2001 October 10) tests with a success ratio of 99%.
24.2 RNG main features

• It delivers 32-bit random numbers, produced by an analog generator
• 40 periods of the RNG_CLK clock signal between two consecutive random numbers
• Monitoring of the RNG entropy to flag abnormal behavior (generation of stable values,
or of a stable sequence of values)
• It can be disabled to reduce power consumption
24.3 RNG functional description

Figure 232 shows the RNG block diagram.
Figure 232. Block diagram
32-bit AHB bus
data register
Control register RNG_DR
RNG _CR
RNG_CLK LFSR
Status register feed a Linear Feedback

Clock checker & Shift Register
RNG _SR fault detector
Analog seed
ai16080
1. For more details about RNG Clock (RNG_CLK) source, please refer to Section 6: Reset and clock control
for STM32F42xxx and STM32F43xxx (RCC) and Section 7: Reset and clock control for
STM32F405xx/07xx and STM32F415xx/17xx(RCC).
The random number generator implements an analog circuit. This circuit generates seeds
that feed a linear feedback shift register (RNG_LFSR) in order to produce 32-bit random
numbers.
RM0090 Rev 19 767/1751

771
Random number generator (RNG) RM0090
The analog circuit is made of several ring oscillators whose outputs are XORed to generate
the seeds. The RNG_LFSR is clocked by a dedicated clock (RNG_CLK) at a constant
frequency, so that the quality of the random number is independent of the HCLK frequency.
The contents of the RNG_LFSR are transferred into the data register (RNG_DR) when a
significant number of seeds have been introduced into the RNG_LFSR.
In parallel, the analog seed and the dedicated RNG_CLK clock are monitored. Status bits (in
the RNG_SR register) indicate when an abnormal sequence occurs on the seed or when
the frequency of the RNG_CLK clock is too low. An interrupt can be generated when an
error is detected.
24.3.1 Operation
To run the RNG, follow the steps below:
1. Enable the interrupt if needed (to do so, set the IE bit in the RNG_CR register). An
interrupt is generated when a random number is ready or when an error occurs.
2. Enable the random number generation by setting the RNGEN bit in the RNG_CR
register. This activates the analog part, the RNG_LFSR and the error detector.
3. At each interrupt, check that no error occurred (the SEIS and CEIS bits should be ‘0’ in
the RNG_SR register) and that a random number is ready (the DRDY bit is ‘1’ in the
RNG_SR register). The contents of the RNG_DR register can then be read.
As required by the FIPS PUB (Federal Information Processing Standard Publication) 140-2,
the first random number generated after setting the RNGEN bit should not be used, but
saved for comparison with the next generated random number. Each subsequent generated
random number has to be compared with the previously generated number. The test fails if
any two compared numbers are equal (continuous random number generator test).

If the CEIS bit is read as ‘1’ (clock error)
In the case of a clock, the RNG is no more able to generate random numbers because the
RNG_CLK clock is not correct. Check that the clock controller is correctly configured to
provide the RNG clock and clear the CEIS bit. The RNG can work when the CECS bit is ‘0’.
The clock error has no impact on the previously generated random numbers, and the
RNG_DR register contents can be used.
If the SEIS bit is read as ‘1’ (seed error)

In the case of a seed error, the generation of random numbers is interrupted for as long as
the SECS bit is ‘1’. If a number is available in the RNG_DR register, it must not be used
because it may not have enough entropy.
What you should do is clear the SEIS bit, then clear and set the RNGEN bit to reinitialize
and restart the RNG.
24.4 RNG registers

The RNG is associated with a control register, a data register and a status register. They
have to be accessed by words (32 bits).
768/1751 RM0090 Rev 19

24.4.1 RNG control register (RNG_CR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
IE RNGEN
Reserved Reserved
rw rw

Bit 3 IE: Interrupt enable
0: RNG Interrupt is disabled
1: RNG Interrupt is enabled. An interrupt is pending as soon as DRDY=1 or SEIS=1 or
CEIS=1 in the RNG_SR register.
Bit 2 RNGEN: Random number generator enable
0: Random number generator is disabled
1: random Number Generator is enabled.
24.4.2 RNG status register (RNG_SR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SEIS CEIS SECS CECS DRDY
Reserved Reserved
rc_w0 rc_w0 r r r

Bit 6 SEIS: Seed error interrupt status
This bit is set at the same time as SECS, it is cleared by writing it to 0.
0: No faulty sequence detected
1: One of the following faulty sequences has been detected:
– More than 64 consecutive bits at the same value (0 or 1)
– More than 32 consecutive alternances of 0 and 1 (0101010101...01)
An interrupt is pending if IE = 1 in the RNG_CR register.
Bit 5 CEIS: Clock error interrupt status
This bit is set at the same time as CECS, it is cleared by writing it to 0.
0: The RNG_CLK clock was correctly detected
1: The RNG_CLK was not correctly detected (fRNG_CLK< fHCLK/16)
An interrupt is pending if IE = 1 in the RNG_CR register.
RM0090 Rev 19 769/1751

771
Random number generator (RNG) RM0090
Bit 2 SECS: Seed error current status

0: No faulty sequence has currently been detected. If the SEIS bit is set, this means that a
faulty sequence was detected and the situation has been recovered.
1: One of the following faulty sequences has been detected:
– More than 64 consecutive bits at the same value (0 or 1)
– More than 32 consecutive alternances of 0 and 1 (0101010101...01)
Bit 1 CECS: Clock error current status
0: The RNG_CLK clock has been correctly detected. If the CEIS bit is set, this means that a
clock error was detected and the situation has been recovered
1: The RNG_CLK was not correctly detected (fRNG_CLK< fHCLK/16).
Bit 0 DRDY: Data ready
0: The RNG_DR register is not yet valid, no random data is available
1: The RNG_DR register contains valid random data
Note: An interrupt is pending if IE = 1 in the RNG_CR register.
Once the RNG_DR register has been read, this bit returns to 0 until a new valid value is
computed.
24.4.3 RNG data register (RNG_DR)

The RNG_DR register is a read-only register that delivers a 32-bit random value when read.
After being read, this register delivers a new random value after a maximum time of 40
periods of the RNG_CLK clock. The software must check that the DRDY bit is set before
reading the RNDATA value.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
RNDATA
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RNDATA
Bits 31:0 RNDATA: Random data

32-bit random data.
770/1751 RM0090 Rev 19

24.4.4 RNG register map

Table 116 gives the RNG register map and reset values.
Table 116. RNG register map and reset map

Register size
Register
Offset name 31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
reset value
9
8
7
6
5
4
3
2
1
0
Reserved
RNGEN
RNG_CR
0x00 Reserved
IE
0x0000000
Reserved
RNG_SR
DRDY
CECS
SECS
CEIS
SEIS
0x04 Reserved
0x0000000
RNG_DR
0x08 RNDATA[31:0]
0x0000000
RM0090 Rev 19 771/1751

771
Hash processor (HASH) RM0090
25 Hash processor (HASH)
This section applies to STM32F415/417xx and STM32F43xxx devices.
25.1 HASH introduction

The hash processor is a fully compliant implementation of the secure hash algorithm
(SHA-1, SHA-224, SHA-256), the MD5 (message-digest algorithm 5) hash algorithm and
the HMAC (keyed-hash message authentication code) algorithm suitable for a variety of
applications. It computes a message digest (160 bits for the SHA-1 algorithm, 256 bits for
the SHA-256 algorithm and 224 bits for the SHA-224 algorithm,128 bits for the MD5
algorithm) for messages of up to (264 – 1) bits, while HMAC algorithms provide a way of
authenticating messages by means of hash functions. HMAC algorithms consist in calling
the SHA-1, SHA-224, SHA-256 or MD5 hash function twice.
25.2 HASH main features

• Suitable for data authentication applications, compliant with:
– FIPS PUB 180-2 (Federal Information Processing Standards Publication 180-2)
– Secure Hash Standard specifications (SHA-1, SHA-224 and SHA-256)
– IETF RFC 1321 (Internet Engineering Task Force Request For Comments number
1321) specifications (MD5)
• Fast computation of SHA-1, SHA-224 and SHA-256, and MD5 (SHA-224 and SHA-256
are available on STM32F43xxx only)
• AHB slave peripheral
• 32-bit data words for input data, supporting word, half-word, byte and bit bit-string
representations, with little-endian data representation only.
• Automatic swapping to comply with the big-endian SHA1, SHA-224 and SHA-256
computation standard with little-endian input bit-string representation
• Automatic padding to complete the input bit string to fit modulo 512 (16 × 32 bits)
message digest computing
• 5× 32-bit words (H0 to H5) on STM32F415/417xx and 8 × 32-bit words (H0 to H7) on
STM32F43xxx for output message digest, reload able to continue interrupted message
digest computation.
• Corresponding 32-bit words of the digest from consecutive message blocks are added
to each other to form the digest of the whole message
• Automatic data flow control with support for direct memory access (DMA)
Note: Padding, as defined in the SHA-1, SHA-224 and SHA-256 algorithm, consists in adding a bit
at bx1 followed by N bits at bx0 to get a total length congruent to 448 modulo 512. After this,
the message is completed with a 64-bit integer which is the binary representation of the
original message length.
For this hash processor, the quanta for entering the message is a 32-bit word, so an
additional information must be provided at the end of the message entry, which is the
number of valid bits in the last 32-bit word entered.
772/1751 RM0090 Rev 19

RM0090 Hash processor (HASH)
25.3 HASH functional description

Figure 1 shows the block diagram of the hash processor.
Figure 233. Block diagram for STM32F415/417xx
32-bit AHB2 bus
IN buffer
Data
register
Control and status HASH_DIN
registers
write into HASH_DIN
or write DCAL bit to 1
Interrupt registers or 1 complete block
transferred by the DMA
HASH_IMR
HASH_SR 16 × 32-bit
IN FIFO
Control register
HASH_CR
Start register
HASH_ST R swa ppin g
IN FIFO full
or DCAL written to 1
Context swapping
context
HASH_CSR0..50 SHA -1 / MD5
Hash / HMAC
Message digest
digest processor core
HASH_H0...H4
ai16081
RM0090 Rev 19 773/1751

797
Figure 234. Block diagram for STM32F43xxx
32-bit AHB2 bus
IN buffer
Data
register
Control and status HASH_DIN
registers
write into HASH_DIN
or write DCAL bit to 1
Interrupt registers or 1 complete block
transferred by the DMA
HASH_IMR
HASH_SR 16 × 32-bit
IN FIFO
Control register
HASH_CR
Start register
HASH_ST R swa ppin g
IN FIFO full
or DCAL written to 1
Context swapping
context
HASH_CSR0..53
SHA-1, SHA-224, SHA-256 and MD5
Message digest Hash / HMAC processor core
digest
HASH_H0...H7
MS30442V1
The FIPS PUB 180-2 standard and the IETF RFC 1321 publication specify the SHA-1, SHA-
224 and SHA-256 and MD5 secure hash algorithms, respectively, for computing a
condensed representation of a message or data file. When a message of any length below
264 bits is provided on input, the SHA-1, SHA-224 and SHA-256 and MD5 produce
respective a 160-bit, 224 bit, 256 bit and 128-bit output string, respectively, called a
message digest. The message digest can then be processed with a digital signature
algorithm in order to generate or verify the signature for the message. Signing the message
digest rather than the message often improves the efficiency of the process because the
message digest is usually much smaller in size than the message. The verifier of a digital
signature has to use the same hash algorithm as the one used by the creator of the digital
signature.
The SHA-1, SHA-224 and SHA-256 and MD5 are qualified as “secure” because it is
computationally infeasible to find a message that corresponds to a given message digest, or
to find two different messages that produce the same message digest. Any change to a
message in transit will, with very high probability, result in a different message digest, and
the signature will fail to verify. For more detail on the SHA-1 or SHA-224 and SHA-256
algorithm, please refer to the FIPS PUB 180-2 (Federal Information Processing Standards
Publication 180-2), 2002 august 1.
The current implementation of this standard works with little-endian input data convention.
For example, the C string “abc” must be represented in memory as the 24-bit hexadecimal
value 0x434241.
A message or data file to be processed by the hash processor should be considered a bit
string. The length of the message is the number of bits in the message (the empty message
774/1751 RM0090 Rev 19

has length 0). You can consider that 32 bits of this bit string forms a 32-bit word. Note that
the FIPS PUB 180-1 standard uses the convention that bit strings grow from left to right, and
bits can be grouped as bytes (8 bits) or words (32 bits) (but some implementations also use
half-words (16 bits), and implicitly, uses the big-endian byte (half-word) ordering. This
convention is mainly important for padding (see Section 1.3.4: Message padding on
page 12).
25.3.1 Duration of the processing

The computation of an intermediate block of a message takes:
• 66 HCLK clock cycles in SHA-1
• 50 HCLK clock cycles in MD5
to which you must add the time needed to load the 16 words of the block into the processor
(at least 16 clock cycles for a 512-bit block).
The time needed to process the last block of a message (or of a key in HMAC) can be
longer. This time depends on the length of the last block and the size of the key (in HMAC
mode). Compared to the processing of an intermediate block, it can be increased by a factor
of:
• 1 to 2.5 for a hash message
• around 2.5 for an HMAC input-key
• 1 to 2.5 for an HMAC message
• around 2.5 for an HMAC output key in case of a short key
• 3.5 to 5 for an HMAC output key in case of a long key
25.3.2 Data type

Data are entered into the hash processor 32 bits (word) at a time, by writing them into the
HASH_DIN register. But the original bit-string can be organized in bytes, half-words or
words, or even be represented as bits. As the system memory organization is little-endian
and SHA1, SHA-224 and SHA-256 computation is big-endian, depending on the way the
original bit string is grouped, a bit, byte, or half-word swapping operation is performed
automatically by the hash processor.
The kind of data to be processed is configured with the DATATYPE bitfield in the HASH
control register (HASH_CR).
RM0090 Rev 19 775/1751

797
Figure 235. Bit, byte and half-word swapping
A-In case of binary data hash, all bits should be swapped as below
Bit swapping operation DATATYPE =bx 11
HASH_DIN bit 31 bit 30 bit 2 bit 1 bit 0 Bits entred with little-Endian format
padding is performed
on this side of the
bit string.
bit string organization in bit 0 bit 1 bit 29 bit 30 bit 31 “1” “0s”
Hash processor: Big-Endian bit-string grows
in this direction
as defined by
FIPSPUB 180-2 std.
B-In case of byte data hash, all bytes should be swapped as below
Byte swapping operation DATATYPE =bx 10
Byte 3 Byte 2 Byte 1 Byte 0 Bytes entred with little-Endian format

HASH_DIN
bits 7...0 bits 7...0 bits 7...0 bits 7...0
Bit string organization in Byte 0 Byte 1 Byte 2 Byte 3

Hash processor: Big-Endian bits 7...0 bits 7...0 bit-string grows
bits 7...0 bits 7...0
in this direction
as defined by
FIPSPUB 180-2 std.
C-In case of half-word hash, all half-word should be swapped as below

Half-word swapping operation DATATYPE =bx 01
Half-word 1 Half-word 0
HASH_DIN half word entred with little-Endian format
bits 15...0 bits 15...0
Bit string organization in Half-word 0 Half-word 1

Hash processor: Big-Endian bits 15...0 bits 15...0
bit-string grows
in this direction
as defined by
FIPS PUB 180-2 std.
ai16082b
776/1751 RM0090 Rev 19

The least significant bit of the message has to be at position 0 (right) in the first word
entered into the hash processor, the 32nd bit of the bit string has to be at position 0 in the
second word entered into the hash processor and so on.
25.3.3 Message digest computing

The HASH sequentially processes blocks of 512 bits when computing the message digest.
Thus, each time 16 × 32-bit words (= 512 bits) have been written by the DMA or the CPU,
into the hash processor, the HASH automatically starts computing the message digest. This
operation is known as a partial digest computation.
The message to be processed is entered into the peripheral by 32-bit words written into the
HASH_DIN register. The current contents of the HASH_DIN register are transferred to the
input FIFO (IN FIFO) each time the register is written with new data. HASH_DIN and the
input FIFO form a FIFO of a 17-word length (named the IN buffer).
The processing of a block can start only once the last value of the block has entered the IN
FIFO. The peripheral must get the information as to whether the HASH_DIN register
contains the last bits of the message or not. Two cases may occur:
• When the DMA is not used:
– In case of a partial digest computation, this is done by writing an additional word
into the HASH_DIN register (actually the first word of the next block). Then the
software must wait until the processor is ready again (when DINIS=1) before
writing new data into HASH_DIN.
– In case of a final digest computation (last block entered), this is done by writing the
DCAL bit to 1.
• When the DMA is used:
The contents of the HASH_DIN register are interpreted automatically with the
information sent by the DMA controller.
– In case of a single DMA transfer: Multiple DMA transfer (MDMAT) bit should be
cleared on STM32F43xxx. When the last block has been transferred to the
HASH_DIN register via DMA channel, DCAL bit will be set to automatically to 1 in
the HASH_STR register in order to launch the final digest calculation.
– In case of a multiple DMA transfer (available only on STM32F43xxx): Multiple
DMA transfer (MDMAT) bit should be set to 1 by software so DCAL bit does not
get set automatically by HW, in this case the final digest calculation for hash and
for each phases for HMAC (for more details about HMAC phases please refer to
HMAC operation section) will not be launched a the end of the DMA transfer
request, allowing the processor to receive a new DMA transfer. During the last
DMA transfer, Multiple DMA transfer (MDMAT) bit should be cleared by software in
order to set automatically DCAL bit at the end of the last bloc and lunch the final
digest.
– The contents of the HASH_DIN register are interpreted automatically with the
information sent by the DMA controller.
This process —data entering + partial digest computation— continues until the last bits of
the original message are written to the HASH_DIN register. As the length (number of bits) of
a message can be any integer value, the last word written into the HASH processor may
have a valid number of bits between 1 and 32. This number of valid bits in the last word,
NBLW, has to be written into the HASH_STR register, so that message padding is correctly
performed before the final message digest computation.
RM0090 Rev 19 777/1751

797
Once this is done, writing into HASH_STR with bit DCAL = 1 starts the processing of the last
entered block of message by the hash processor. This processing consists in:
• Automatically performing the message padding operation: the purpose of this operation
is to make the total length of a padded message a multiple of 512. The HASH
sequentially processes blocks of 512 bits when computing the message digest
• Computing the final message digest
When the DMA is enabled, it provides the information to the hash processor when it is
transferring the last data word. Then the padding and digest computation are performed
automatically as if DCAL had been written to 1.
25.3.4 Message padding

Message padding consists in appending a “1” followed by m “0”s followed by a 64-bit integer
to the end of the original message to produce a padded message block of length 512. The
“1” is added to the last word written into the HASH_DIN register at the bit position defined by
the NBLW bitfield, and the remaining upper bits are cleared (“0”s).
Example: let us assume that the original message is the ASCII binary-coded form of “abc”,
of length L = 24:
byte 0 byte 1 byte 2 byte 3
01100001 01100010 01100011 UUUUUUUU
<-- 1st word written to HASH_DIN -->
NBLW has to be loaded with the value 24: a “1” is appended at bit location 24 in the bit string
(starting counting from left to right in the above bit string), which corresponds to bit 31 in the
HASH_DIN register (little-endian convention):
01100001 01100010 01100011 1UUUUUUU
Since L = 24, the number of bits in the above bit string is 25, and 423 “0”s are appended,
making now 448. This gives (in hexadecimal, big-endian format):
61626380 00000000 00000000 00000000
00000000 00000000 00000000 00000000
00000000 00000000 00000000 00000000
00000000 00000000
The L value, in two-word representation (that is 00000000 00000018) is appended. Hence
the final padded message in hexadecimal:
61626380 00000000 00000000 00000000
00000000 00000000 00000000 00000000
00000000 00000000 00000000 00000000
00000000 00000000 00000000 00000028
If the HASH is programmed to use the little-endian byte input format, the above message
has to be entered by doing the following steps:
1. 0xUU636261 is written into the HASH_DIN register (where ‘U’ means don’t care)
2. 0x18 is written into the HASH_STR register (the number of valid bits in the last word
written into the HASH_DIN register is 24, as the original message length is 24 bits)
3. 0x10 is written into the HASH_STR register to start the message padding and digest
computation. When NBLW ≠ 0x00, the message padding puts a “1” into the HASH_DIN
register at the bit position defined by the NBLW value, and inserts “0”s at bit locations
[31:(NBLW+1)]. When NBLW == 0x00, the message padding inserts one new word with
778/1751 RM0090 Rev 19

value 0x0000 0001. Then an all zero word (0x0000 0000) is added and the message
length in a two-word representation, to get a block of 16 x 32-bit words.
4. The HASH computing is performed, and the message digest is then available in the
HASH_Hx registers (x = 0...4) for the SHA-1 algorithm. For example:
H0 = 0xA9993E36
H1 = 0x4706816A
H2 = 0xBA3E2571
H3 = 0x7850C26C
H4 = 0x9CD0D89D
25.3.5 Hash operation

The hash function (SHA-1, SHA-224, SHA-256 and MD5) is selected when the INIT bit is
written to ‘1’ in the HASH_CR register while the MODE bit is at ‘0’ in HASH_CR. The
algorithm (SHA-1, SHA-224,SHA-256 or MD5) is selected at the same time (that is when the
INIT bit is set) using the ALGO bits.
The message can then be sent by writing it word by word into the HASH_DIN register. When
a block of 512 bits —that is 16 words— has been written, a partial digest computation starts
upon writing the first data of the next block. The hash processor remains busy for 66 cycles
for the SHA-1 algorithm, or 50 cycles for the MD5 algorithm, SHA-224 algorithm and SHA-
256 algorithm.
The process can then be repeated until the last word of the message. If DMA transfers are
used, refer to the Procedure where the data are loaded by DMA section. Otherwise, if the
message length is not an exact multiple of 512 bits, then the HASH_STR register has to be
written to launch the computation of the final digest.
Once computed, the digest can be read from the HASH_H0...HASH_H4 registers (for the
MD5 algorithm, HASH_H4 is not relevant) on STM32F415/417xx, and from the
HASH_H0...HASH_H7 registers on STM32F43xxx where:
HASH_H4..HASH_H7 are not relevant when the MD5 algorithm is selected,
HASH_H5.. HASH_H7 are not relevant when the SHA-1algorithm is selected,
HASH_H7 is not relevant when the SHA-224 algorithm is selected.
25.3.6 HMAC operation

The HMAC algorithm is used for message authentication, by irreversibly binding the
message being processed to a key chosen by the user. For HMAC specifications, refer to
“HMAC: keyed-hashing for message authentication, H. Krawczyk, M. Bellare, R. Canetti,
February 1997.
Basically, the algorithm consists of two nested hash operations:
HMAC(message) = Hash[((key | pad) XOR 0x5C)
| Hash(((key | pad) XOR 0x36) | message)]
where:
• pad is a sequence of zeroes needed to extend the key to the length of the underlying
hash function data block (that is 512 bits for both the SHA-1, SHA224, SHA-256 and
MD5 hash algorithms)
• | represents the concatenation operator
To compute the HMAC, four different phases are required:
RM0090 Rev 19 779/1751

797
1. The block is initialized by writing the INIT bit to ‘1’ with the MODE bit at ‘1’ and the
ALGO bits set to the value corresponding to the desired algorithm. The LKEY bit must
also be set during this phase if the key being used is longer than 64 bytes (in this case,
the HMAC specifications specify that the hash of the key should be used in place of the
real key).
2. The key (to be used for the inner hash function) is then given to the core. This
operation follows the same mechanism as the one used to send the message in the
hash operation (that is, by writing into HASH_DIN and, finally, into HASH_STR).
3. Once the last word has been entered and computation has started, the hash processor
elaborates the key. It is then ready to accept the message text using the same
mechanism as the one used to send the message in the hash operation.
4. After the first hash round, the hash processor returns “ready” to indicate that it is ready
to receive the key to be used for the outer hash function (normally, this key is the same
as the one used for the inner hash function). When the last word of the key is entered
and computation starts, the HMAC result is made available in the
HASH_H0...HASH_H4 registers on STM32F415/417xx and on HASH_H0...HASH_H7
registers on STM32F43xxx.
Note: 1 The computation latency of the HMAC primitive depends on the lengths of the keys and
message. You could the HMAC as two nested underlying hash functions with the same key
length (long or short).
25.3.7 Context swapping

It is possible to interrupt a hash/HMAC process to perform another processing with a higher
priority, and to complete the interrupted process later on, when the higher-priority task is
complete. To do so, the context of the interrupted task must be saved from the hash
registers to memory, and then be restored from memory to the hash registers.
The procedures where the data flow is controlled by software or by DMA are described
below.
780/1751 RM0090 Rev 19

Procedure where the data are loaded by software

The context can be saved only when no block is currently being processed. That is, you
must wait for DINIS = 1 (the last block has been processed and the input FIFO is empty) or
NBW ≠ 0 (the FIFO is not full and no processing is ongoing).
• Context saving:
Store the contents of the following registers into memory:
– HASH_IMR
– HASH_STR
– HASH_CR
– HASH_CSR0 to HASH_CSR50 on STM32F415/417xx, and HASH_CSR0 to
HASH_CSR53 on STM32F43xxx.
• Context restoring:
The context can be restored when the high-priority task is complete. Please follow the
order of the sequence below.
a) Write the following registers with the values saved in memory: HASH_IMR,
HASH_STR and HASH_CR
b) Initialize the hash processor by setting the INIT bit in the HASH_CR register
c) Write the HASH_CSR0 to HASH_CSR50 (STM32F415/417xx), and HASH_CSR0
to HASH_CSR53 (STM32F43xxx) registers with the values saved in memory
You can now restart the processing from the point where it has been interrupted.
Procedure where the data are loaded by DMA

In this case it is not possible to predict if a DMA transfer is in progress or if the process is
ongoing. Thus, you must stop the DMA transfers, then wait until the HASH is ready in order
to interrupt the processing of a message.
• Interrupting a processing:
– Clear the DMAE bit to disable the DMA interface
– Wait until the current DMA transfer is complete (wait for DMAES = 0 in the
HASH_SR register). Note that the block may or not have been totally transferred
to the HASH.
– Disable the corresponding channel in the DMA controller
– Wait until the hash processor is ready (no block is being processed), that is wait
for DINIS = 1
• The context saving and context restoring phases are the same as above (see
Procedure where the data are loaded by software).
Reconfigure the DMA controller so that it transfers the end of the message. You can now
restart the processing from the point where it was interrupted by setting the DMAE bit.
Note: If context swapping does not involve HMAC operations, the HASH_CSR38 to
HASH_CSR50 (STM32F415/417xx) and HASH_CSR38 to HASH_CSR53 (STM32F43xxx)
registers do not have to be saved and restored.
If context swapping occurs between two blocks (the last block was completely processed
and the next block has not yet been pushed into the IN FIFO, NBW = 000 in the HASH_CR
register), the HASH_CSR22 to HASH_CSR37 registers do not have to be saved and
restored.
RM0090 Rev 19 781/1751

797
25.3.8 HASH interrupt

There are two individual maskable interrupt sources generated by the HASH processor.
They are connected to the same interrupt vector.
You can enable or disable the interrupt sources individually by changing the mask bits in the
HASH_IMR register. Setting the appropriate mask bit to 1 enables the interrupt.
The status of the individual interrupt sources can be read from the HASH_SR register.
Figure 236. HASH interrupt mapping diagram

DCIS
DCIM
HASH interrupt to NVIC
DINIS
DINIM
ai16086
25.4 HASH registers

The HASH core is associated with several control and status registers and five message
digest registers.
All these registers are accessible through word accesses only, else an AHB error is
generated.
25.4.1 HASH control register (HASH_CR) for STM32F415/417xx

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
LKEY
Reserved
rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DINNE NBW ALGO[0] MODE DATATYPE DMAE INIT
Reserved Reserved
r r r r r rw rw rw rw rw w

Bit 16 LKEY: Long key selection
This bit selects between short key (≤ 64 bytes) or long key (> 64 bytes) in HMAC
mode
0: Short key (≤ 64 bytes)
1: Long key (> 64 bytes)
Note: This selection is only taken into account when the INIT bit is set and MODE
= 1. Changing this bit during a computation has no effect.
782/1751 RM0090 Rev 19

Bit 12 DINNE: DIN not empty

This bit is set when the HASH_DIN register holds valid data (that is after being
written at least once). It is cleared when either the INIT bit (initialization) or the
DCAL bit (completion of the previous message processing) is written to 1.
0: No data are present in the data input buffer
1: The input buffer contains at least one word of data
Bits 11:8 NBW: Number of words already pushed
This bitfield reflects the number of words in the message that have already been
pushed into the IN FIFO.
NBW increments (+1) when a write access is performed to the HASH_DIN
register while DINNE = 1.
It goes to 0000 when the INIT bit is written to 1 or when a digest calculation
starts (DCAL written to 1 or DMA end of transfer).
″ If the DMA is not used:
0000 and DINNE=0: no word has been pushed into the DIN buffer (the buffer is
empty, both the HASH_DIN register and the IN FIFO are empty)
0000 and DINNE=1: 1 word has been pushed into the DIN buffer (The
HASH_DIN register contains 1 word, the IN FIFO is empty)
0001: 2 words have been pushed into the DIN buffer (the HASH_DIN register
and the IN FIFO contain 1 word each)
...
1111: 16 words have been pushed into the DIN buffer
″ If the DMA is used, NBW is the exact number of words that have been
Bit 7 ALGO[1:0]: Algorithm selection
These bits selects the SHA-1 or the MD5 algorithm:
0: SHA-1 algorithm selected
1: MD5 algorithm selected
Note: This selection is only taken into account when the INIT bit is set. Changing
this bit during a computation has no effect.
Bit 6 MODE: Mode selection
This bit selects the HASH or HMAC mode for the selected algorithm:
0: Hash mode selected
1: HMAC mode selected. LKEY must be set if the key being used is longer than
64 bytes.
Bits 5:4 DATATYPE: Data type selection
Defines the format of the data entered into the HASH_DIN register:
00: 32-bit data. The data written into HASH_DIN are directly used by the HASH
processing, without reordering.
01: 16-bit data, or half-word. The data written into HASH_DIN are considered as
2 half-words, and are swapped before being used by the HASH processing.
10: 8-bit data, or bytes. The data written into HASH_DIN are considered as 4
bytes, and are swapped before being used by the HASH processing.
11: bit data, or bit-string. The data written into HASH_DIN are considered as 32
bits (1st bit of the sting at position 0), and are swapped before being used by the
HASH processing (1st bit of the string at position 31).
RM0090 Rev 19 783/1751

797
Bit 3 DMAE: DMA enable

0: DMA transfers disabled
1: DMA transfers enabled. A DMA request is sent as soon as the HASH core is
ready to receive data.
Note: 1: This bit is cleared by hardware when the DMA asserts the DMA terminal
count signal (while transferring the last data of the message). This bit is not
cleared when the INIT bit is written to 1.
2: If this bit is written to 0 while a DMA transfer has already been requested
to the DMA, DMAE is cleared but the current transfer is not aborted.
Instead, the DMA interface remains internally enabled until the transfer is
complete or INIT is written to 1.
Bit 2 INIT: Initialize message digest calculation
Writing this bit to 1 resets the hash processor core, so that the HASH is ready to
compute the message digest of a new message.
Writing this bit to 0 has no effect.
Reading this bit always return 0.
784/1751 RM0090 Rev 19

25.4.2 HASH control register (HASH_CR) for STM32F43xxx

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
ALGO[1] LKEY
Reserved
rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MDMAT DINNE NBW ALGO[0] MODE DATATYPE DMAE INIT
Reserved Reserved
rw r r r r r rw rw rw rw rw w

Bit 17 Reserved, forced by hardware to 0.
Bit 16 LKEY: Long key selection
This bit selects between short key (≤ 64 bytes) or long key (> 64 bytes) in HMAC
mode
0: Short key (≤ 64 bytes)
1: Long key (> 64 bytes)
Note: This selection is only taken into account when the INIT bit is set and MODE
= 1. Changing this bit during a computation has no effect.
Bit 13 MDMAT: Multiple DMA Transfers
This bit is set when hashing large files when multiple DMA transfers are needed.
0: DCAL is automatically set at the end of a DMA transfer.
1: DCAL is not automatically set at the end of a DMA transfer.
Bit 12 DINNE: DIN not empty
This bit is set when the HASH_DIN register holds valid data (that is after being
written at least once). It is cleared when either the INIT bit (initialization) or the
DCAL bit (completion of the previous message processing) is written to 1.
0: No data are present in the data input buffer
1: The input buffer contains at least one word of data
RM0090 Rev 19 785/1751

797
Bits 11:8 NBW: Number of words already pushed

This bitfield reflects the number of words in the message that have already been
NBW increments (+1) when a write access is performed to the HASH_DIN
register while DINNE = 1.
It goes to 0000 when the INIT bit is written to 1 or when a digest calculation
starts (DCAL written to 1 or DMA end of transfer).
″ If the DMA is not used:
0000 and DINNE=0: no word has been pushed into the DIN buffer (the buffer is
empty, both the HASH_DIN register and the IN FIFO are empty)
0000 and DINNE=1: 1 word has been pushed into the DIN buffer (The
HASH_DIN register contains 1 word, the IN FIFO is empty)
0001: 2 words have been pushed into the DIN buffer (the HASH_DIN register
and the IN FIFO contain 1 word each)
...
1111: 16 words have been pushed into the DIN buffer
″ If the DMA is used, NBW is the exact number of words that have been
Bit 18 and bit 7 ALGO[1:0]: Algorithm selection
These bits selects the SHA-1, SHA-224, SHA256 or the MD5 algorithm:
00: SHA-1 algorithm selected
01: MD5 algorithm selected
10: SHA224 algorithm selected
11: SHA256 algorithm selected
Bit 6 MODE: Mode selection
This bit selects the HASH or HMAC mode for the selected algorithm:
0: Hash mode selected
1: HMAC mode selected. LKEY must be set if the key being used is longer than
64 bytes.
Bits 5:4 DATATYPE: Data type selection
Defines the format of the data entered into the HASH_DIN register:
00: 32-bit data. The data written into HASH_DIN are directly used by the HASH
processing, without reordering.
01: 16-bit data, or half-word. The data written into HASH_DIN are considered as
2 half-words, and are swapped before being used by the HASH processing.
10: 8-bit data, or bytes. The data written into HASH_DIN are considered as 4
bytes, and are swapped before being used by the HASH processing.
11: bit data, or bit-string. The data written into HASH_DIN are considered as 32
bits (1st bit of the sting at position 0), and are swapped before being used by the
HASH processing (1st bit of the string at position 31).
786/1751 RM0090 Rev 19

Bit 3 DMAE: DMA enable

0: DMA transfers disabled
1: DMA transfers enabled. A DMA request is sent as soon as the HASH core is
ready to receive data.
Note: 1: This bit is cleared by hardware when the DMA asserts the DMA terminal
count signal (while transferring the last data of the message). This bit is not
cleared when the INIT bit is written to 1.
2: If this bit is written to 0 while a DMA transfer has already been requested
to the DMA, DMAE is cleared but the current transfer is not aborted.
Instead, the DMA interface remains internally enabled until the transfer is
complete or INIT is written to 1.
Bit 2 INIT: Initialize message digest calculation
Writing this bit to 1 resets the hash processor core, so that the HASH is ready to
compute the message digest of a new message.
Writing this bit to 0 has no effect.
Reading this bit always return 0.
RM0090 Rev 19 787/1751

797
25.4.3 HASH data input register (HASH_DIN)

HASH_DIN is the data input register. It is 32-bit wide. It is used to enter the message by
blocks of 512 bits. When the HASH_DIN register is written to, the value presented on the
AHB databus is ‘pushed’ into the HASH core and the register takes the new value presented
on the AHB databus. The DATATYPE bits must previously have been configured in the
HASH_CR register to get a correct message representation.
When a block of 16 words has been written to the HASH_DIN register, an intermediate
digest calculation is launched:
• by writing new data into the HASH_DIN register (the first word of the next block) if the
DMA is not used (intermediate digest calculation)
• automatically if the DMA is used
When the last block has been written to the HASH_DIN register, the final digest calculation
(including padding) is launched:
• by writing the DCAL bit to 1 in the HASH_STR register (final digest calculation)
• automatically if the DMA is used and MDMAT bit is set to ‘0’.
When a digest calculation (intermediate or final) is in progress, any new write access to the
HASH_DIN register is extended (by wait-state insertion on the AHB bus) until the HASH
calculation completes.
When the HASH_DIN register is read, the last word written in this location is accessed (zero
after reset).
.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DATAIN
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DATAIN
Bits 31:0 DATAIN: Data input

Read = returns the current register content.
Write = the current register content is pushed into the IN FIFO, and the register
takes the new value presented on the AHB databus.
788/1751 RM0090 Rev 19

25.4.4 HASH start register (HASH_STR)

The HASH_STR register has two functions:
• It is used to define the number of valid bits in the last word of the message entered in
the hash processor (that is the number of valid least significant bits in the last data
written into the HASH_DIN register)
• It is used to start the processing of the last block in the message by writing the DCAL
bit to 1
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DCAL NBLW
Reserved Reserved
w rw rw rw rw rw

Bit 8 DCAL: Digest calculation
Writing this bit to 1 starts the message padding, using the previously written
value of NBLW, and starts the calculation of the final message digest with all data
words written to the IN FIFO since the INIT bit was last written to 1.
Reading this bit returns 0.
Note
Bits 4:0 NBLW: Number of valid bits in the last word of the message in the bit string
organization of hash processor
When these bits are written and DCAL is at ‘0’, they take the value on the AHB
databus:
0x00: All 32 bits of the last data written in the bit string organization of hash
processor (after data swapping) are valid.
0x01: Only bit [31] of the last data written in the bit string organization of hash
processor (after data swapping) are valid
0x02: Only bits [31:30] of the last data written in the bit string organization of
hash processor (after data swapping) are valid
0x03: Only bits [31:29] of the last data written in the bit string organization of
hash processor (after data swapping) are valid
...
0x1F: Only bits [0] of the last data written in the bit string organization of hash
processor (after data swapping) are valid
When these bits are written and DCAL is at ‘1’, the bitfield is not changed.
Reading them returns the last value written to NBLW.
Note: These bits must be configured before setting the DCAL bit, else they are not
taken into account. Especially, it is not possible to configure NBLW and set
DCAL at the same time.
RM0090 Rev 19 789/1751

797
25.4.5 HASH digest registers (HASH_HR0..4/5/6/7)

Address offset: 0x0C to 0x1C (STM32F415/417xx), plus 0x310 to 0x32C (STM32F43xxx)
These registers contain the message digest result named as:
1. H0, H1, H2, H3 and H4, respectively, in the SHA1 algorithm description
Note that in this case, the HASH_H5 to HASH_H7 register is not used, and is read as
zero.
2. A, B, C and D, respectively, in the MD5 algorithm description
Note that in this case, the HASH_H4 to HASH_H7 register is not used, and is read as
zero.
3. H0 to H6, respectively, in the SHA224 algorithm description,
Note that in this case, the HASH_H7 register is not used, and is read as zero.
4. H0 to H7, respectively, in the SHA256 algorithm description,
If a read access to one of these registers occurs while the HASH core is calculating an
intermediate digest or a final message digest (that is when the DCAL bit has been written to
1), then the read is stalled until the completion of the HASH calculation.
Note: H0, H1, H2, H3 and H4 mapping are duplicated in two region.
HASH_HR0
Address offset: 0x0C and 0x310
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
H0
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
H0
HASH_HR1
Address offset: 0x10 and 0x314
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
H1
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
H1
HASH_HR2
Address offset: 0x14 and 0x318
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
H2
790/1751 RM0090 Rev 19

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
H2
HASH_HR3
Address offset: 0x18 and 0x31C
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
H3
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
H3
HASH_HR4
Address offset: 0x1C and 0x320
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
H4
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
H4
HASH_HR5
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
H5
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
H5
HASH_HR6
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
H6
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
H6
HASH_HR7
RM0090 Rev 19 791/1751

797
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
H7
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
H7
Note: When starting a digest computation for a new bit stream (by writing the INIT bit to 1), these
registers assume their reset values.
25.4.6 HASH interrupt enable register (HASH_IMR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DCIE DINIE
Reserved
rw rw

Bit 1 DCIE: Digest calculation completion interrupt enable
0: Digest calculation completion interrupt disabled
1: Digest calculation completion interrupt enabled.
Bit 0 DINIE: Data input interrupt enable
0: Data input interrupt disabled
1: Data input interrupt enabled
792/1751 RM0090 Rev 19

25.4.7 HASH status register (HASH_SR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
BUSY DMAS DCIS DINIS
Reserved
r r rc_w0 rc_w0

Bit 3 BUSY: Busy bit
0: No block is currently being processed
1: The hash core is processing a block of data
Bit 2 DMAS: DMA Status
This bit provides information on the DMA interface activity. It is set with DMAE
and cleared when DMAE=0 and no DMA transfer is ongoing. No interrupt is
associated with this bit.
0: DMA interface is disabled (DMAE=0) and no transfer is ongoing
1: DMA interface is enabled (DMAE=1) or a transfer is ongoing
Bit 1 DCIS: Digest calculation completion interrupt status
This bit is set by hardware when a digest becomes ready (the whole message
has been processed). It is cleared by writing it to 0 or by writing the INIT bit to 1
in the HASH_CR register.
0: No digest available in the HASH_Hx registers
1: Digest calculation complete, a digest is available in the HASH_Hx registers.
An interrupt is generated if the DCIE bit is set in the HASH_IMR register.
Bit 0 DINIS: Data input interrupt status
This bit is set by hardware when the input buffer is ready to get a new block (16
locations are free). It is cleared by writing it to 0 or by writing the HASH_DIN
register.
0: Less than 16 locations are free in the input buffer
1: A new block can be entered into the input buffer. An interrupt is generated if
the DINIE bit is set in the HASH_IMR register.
RM0090 Rev 19 793/1751

797
25.4.8 HASH context swap registers (HASH_CSRx)

Address offset: 0x0F8 to 0x1C0
• For HASH_CSR0 register: Reset value is 0x0000 0002.
• For others registers: Reset value is 0x0000 0000 , except for STM32F43xxx devices
where the HASH_CSR2 register reset value is 0x2000 0000
Additional registers are available from 0x1C1 to 0x1CC on STM32F43xxx:
• Reset value: 0x0000 0000.
These registers contain the complete internal register states of the hash processor, and are
useful when a context swap has to be done because a high-priority task has to use the hash
processor while it is already in use by another task.
When such an event occurs, the HASH_CSRx registers have to be read and the read
values have to be saved somewhere in the system memory space. Then the hash
processor can be used by the preemptive task, and when hash computation is finished, the
saved context can be read from memory and written back into these HASH_CSRx registers.
HASH_CSRx
Address offset: 0x0F8 to 0x1C0 on STM32F415/417xx
Address offset: 0x0F8 to 0x1CC on STM32F43xxx
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CSx
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CSx
794/1751 RM0090 Rev 19

25.4.9 HASH register map

Table 9 gives the summary HASH register map and reset values.
Table 117. HASH register map and reset values on STM32F415/417xx

Register name Register size
Offset
reset value 31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
DATATYPE 5
4
3
2
1
0
ALGO[0]
DINNE
MODE
DMAE
Reserved
Reserved
LKEY
INIT
HASH_CR NBW
0x00 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0
HASH_DIN DATAIN
0x04
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DCAL
Reserved
HASH_STR NBLW
0x08 Reserved
HASH_HR0 H0
0x0C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HASH_HR1 H1
0x10
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HASH_HR2 H2
0x14
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HASH_HR3 H3
0x18
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HASH_HR4 H4
0x1C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DINIE
DCIE
HASH_IMR Reserved
0x20
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 DMAS
BUSY
DINIS
DCIS
HASH_SR
0x24 Reserved
Reset value 0 0 0 1
HASH_CSR0 CSR0
0xF8
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0
...
...
HASH_CSR50 CSR50
0x1C0
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Reserved
HASH_HR0 H0
0x310
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HASH_HR1 H1
0x314
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RM0090 Rev 19 795/1751

797
Table 117. HASH register map and reset values on STM32F415/417xx (continued)
Offset
reset value
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
HASH_HR2 H2
0x318
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HASH_HR3 H3
0x31C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HASH_HR4 H4
0x320
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Table 118. HASH register map and reset values on STM32F43xxx

Offset
reset value
31
30
29
28
27
26
25
24
23
22
21
20
19
ALGO[1] 18
17
16
Reserved 15
14
MDMAT 13
12
10
11
9
8
7
6
DATATYPE 5
4
3
2
1
0
ALGO[0]
DINNE
MODE
DMAE
LKEY
Reserved
Reserved
INIT
HASH_CR NBW
0x00 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HASH_DIN DATAIN
0x04
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DCAL
Reserved
HASH_STR NBLW
0x08 Reserved
HASH_HR0 H0
0x0C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HASH_HR1 H1
0x10
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HASH_HR2 H2
0x14
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HASH_HR3 H3
0x18
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HASH_HR4 H4
0x1C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DINIE
DCIE
HASH_IMR Reserved
0x20
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DMAS
BUSY
DINIS
DCIS
HASH_SR
0x24 Reserved
Reset value 0 0 0 1
796/1751 RM0090 Rev 19

Table 118. HASH register map and reset values on STM32F43xxx (continued)
Offset
reset value
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
HASH_CSR0 CSR0
0xF8
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0
...
...
HASH_CSR53 CSR53
0x1CC
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Reserved
HASH_HR0 H0
0x310
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HASH_HR1 H1
0x314
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HASH_HR2 H2
0x318
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HASH_HR3 H3
0x31C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HASH_HR4 H4
0x320
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HASH_HR5 H5
0x324
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HASH_HR6 H6
0x328
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HASH_HR7 H7
0x32C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RM0090 Rev 19 797/1751

797
Real-time clock (RTC) RM0090
26 Real-time clock (RTC)
26.1 Introduction
The real-time clock (RTC) is an independent BCD timer/counter. The RTC provides a time-
of-day clock/calendar, two programmable alarm interrupts, and a periodic programmable
wakeup flag with interrupt capability. The RTC also includes an automatic wakeup unit to
manage low-power modes.
Two 32-bit registers contain the seconds, minutes, hours (12- or 24-hour format), day (day
of week), date (day of month), month, and year, expressed in binary coded decimal format
(BCD). The sub-seconds value is also available in binary format.
Compensations for 28-, 29- (leap year), 30-, and 31-day months are performed
automatically. Daylight saving time compensation can also be performed.
Additional 32-bit registers contain the programmable alarm subseconds, seconds, minutes,
hours, day, and date.
A digital calibration feature is available to compensate for any deviation in crystal oscillator
accuracy.
After backup domain reset, all RTC registers are protected against possible parasitic write
accesses.
As long as the supply voltage remains in the operating range, the RTC never stops,
regardless of the device status (Run mode, low-power mode or under reset).
798/1751 RM0090 Rev 19

RM0090 Real-time clock (RTC)
26.2 RTC main features

The RTC unit main features are the following (see Figure 237: RTC block diagram):
• Calendar with subseconds, seconds, minutes, hours (12 or 24 format), day (day of
week), date (day of month), month, and year.
• Daylight saving compensation programmable by software.
• Two programmable alarms with interrupt function. The alarms can be triggered by any
combination of the calendar fields.
• Automatic wakeup unit generating a periodic flag that triggers an automatic wakeup
interrupt.
• Reference clock detection: a more precise second source clock (50 or 60 Hz) can be
used to enhance the calendar precision.
• Accurate synchronization with an external clock using the subsecond shift feature.
• Maskable interrupts/events:
– Alarm A
– Alarm B
– Wakeup interrupt
– Timestamp
– Tamper detection
• Digital calibration circuit (periodic counter correction)
– 5 ppm accuracy
– 0.95 ppm accuracy, obtained in a calibration window of several seconds
• Timestamp function for event saving (1 event)
• Tamper detection:
– 2 tamper events with configurable filter and internal pull-up.
• 20 backup registers (80 bytes). The backup registers are reset when a tamper
detection event occurs.
• Alternate function output (RTC_OUT) which selects one of the following two outputs:
– RTC_CALIB: 512 Hz or 1 Hz clock output (with an LSE frequency of 32.768 kHz).
This output is enabled by setting the COE bit in the RTC_CR register. It is routed
to the device RTC_AF1 function.
– RTC_ALARM (Alarm A, Alarm B or wakeup).
This output is selected by configuring the OSEL[1:0] bits in the RTC_CR register.
It is routed to the device RTC_AF1 function.
• RTC alternate function inputs:
– RTC_TS: timestamp event detection. It is routed to the device RTC_AF1 and
RTC_AF2 functions.
– RTC_TAMP1: TAMPER1 event detection. It is routed to the device RTC_AF1 and
RTC_AF2 functions.
– RTC_TAMP2: TAMPER2 event detection.
– RTC_REFIN: reference clock input (usually the mains, 50 or 60 Hz).
Refer to Section 8.3.15: Selection of RTC_AF1 and RTC_AF2 alternate functions.
RM0090 Rev 19 799/1751

838
Figure 237. RTC block diagram
RTC_TAMP3 Backup registers

RTC_TAMP2 and RTC tamper TAMPxF
RTC_TAMP1 control register
RTC_TS Time stamp

TSF
registers
RTC_REFIN
LSE (32.768 Hz)

HSE/32
LSI
RTCCLK
ck_apre ck_spre
RTC_CALR RTC_PRER (default 256 Hz RTC_PRER (default 1 Hz)
Smooth Asynchronous Synchronous
calibration 7-bit prescaler 15-bit prescaler
(default = 128) (default = 256) Calendar
Shadow register Shadow registers

RTC_SSR RTC_TR,
RTC_DR
WUCKSEL[1:0]
Prescaler
1, 2, 4, 8, 16
RTC_CALIB
Output RTC_OUT
RTC_ALARM control
RTC_WUTR
16-bit wakeup WUTF
auto reload timer
OSEL[1:0]
Alarm A
RTC_ALRMAR = ALRAF
RTC_ALRMASSR
Alarm B
= ALRBF
RTC_ALRMBR
RTC_ALRMBSSR
MS30216V6
1. On STM32F4xx devices, the RTC_AF1 and RTC_AF2 alternate functions are connected to PC13 and PI8,
respectively.
800/1751 RM0090 Rev 19

26.3 RTC functional description
26.3.1 Clock and prescalers

The RTC clock source (RTCCLK) is selected through the clock controller among the LSE
clock, the LSI oscillator clock, and the HSE clock. For more information on the RTC clock
source configuration, refer to Section 7: Reset and clock control for STM32F405xx/07xx
and STM32F415xx/17xx(RCC) and Section 6: Reset and clock control for STM32F42xxx
and STM32F43xxx (RCC).
A programmable prescaler stage generates a 1 Hz clock which is used to update the
calendar. To minimize power consumption, the prescaler is split into 2 programmable
prescalers (see Figure 237: RTC block diagram):
• A 7-bit asynchronous prescaler configured through the PREDIV_A bits of the
RTC_PRER register.
• A 15-bit synchronous prescaler configured through the PREDIV_S bits of the
RTC_PRER register.
Note: When both prescalers are used, it is recommended to configure the asynchronous prescaler
to a high value to minimize consumption.
The asynchronous prescaler division factor is set to 128, and the synchronous division
factor to 256, to obtain an internal clock frequency of 1 Hz (ck_spre) with an LSE frequency
of 32.768 kHz.
The minimum division factor is 1 and the maximum division factor is 222.
This corresponds to a maximum input frequency of around 4 MHz.
fck_apre is given by the following formula:
f RTCCLK
f CK_APRE = --------------------------------------
-
PREDIV_A + 1
The ck_apre clock is used to clock the binary RTC_SSR subseconds downcounter. When it
reaches 0, RTC_SSR is reloaded with the content of PREDIV_S.
fck_spre is given by the following formula:
f RTCCLK
f CK_SPRE = ----------------------------------------------------------------------------------------------
( PREDIV_S + 1 ) × ( PREDIV_A + 1 )
The ck_spre clock can be used either to update the calendar or as timebase for the 16-bit
wakeup auto-reload timer. To obtain short timeout periods, the 16-bit wakeup auto-reload
timer can also run with the RTCCLK divided by the programmable 4-bit asynchronous
prescaler (see Section 26.3.4: Periodic auto-wakeup for details).
RM0090 Rev 19 801/1751

838
26.3.2 Real-time clock and calendar

The RTC calendar time and date registers are accessed through shadow registers which
are synchronized with PCLK1 (APB1 clock). They can also be accessed directly in order to
avoid waiting for the synchronization duration.
• RTC_SSR for the subseconds
• RTC_TR for the time
• RTC_DR for the date
Every two RTCCLK periods, the current calendar value is copied into the shadow registers,
and the RSF bit of RTC_ISR register is set (see Section 26.6.4). The copy is not performed
in Stop and Standby mode. When exiting these modes, the shadow registers are updated
after up to 2 RTCCLK periods.
When the application reads the calendar registers, it accesses the content of the shadow
registers.t is possible to make a direct access to the calendar registers by setting the
BYPSHAD control bit in the RTC_CR register. By default, this bit is cleared, and the user
accesses the shadow registers.
When reading the RTC_SSR, RTC_TR or RTC_DR registers in BYPSHAD=0 mode, the
frequency of the APB clock (fAPB) must be at least 7 times the frequency of the RTC clock
(fRTCCLK).
The shadow registers are reset by system reset.
26.3.3 Programmable alarms

The RTC unit provides two programmable alarms, Alarm A and Alarm B.
The programmable alarm functions are enabled through the ALRAIE and ALRBIE bits in the
RTC_CR register. The ALRAF and ALRBF flags are set to 1 if the calendar subseconds,
seconds, minutes, hours, date or day match the values programmed in the alarm registers
RTC_ALRMASSR/RTC_ALRMAR and RTC_ALRMBSSR/RTC_ALRMBR, respectively.
Each calendar field can be independently selected through the MSKx bits of the
RTC_ALRMAR and RTC_ALRMBR registers, and through the MASKSSx bits of the
RTC_ALRMASSR and RTC_ALRMBSSR registers. The alarm interrupts are enabled
through the ALRAIE and ALRBIE bits in the RTC_CR register.
Alarm A and Alarm B (if enabled by bits OSEL[1:0] in RTC_CR register) can be routed to the
RTC_ALARM output. RTC_ALARM polarity can be configured through bit POL in the
RTC_CR register.
Caution: If the seconds field is selected (MSK0 bit reset in RTC_ALRMAR or RTC_ALRMBR), the
synchronous prescaler division factor set in the RTC_PRER register must be at least 3 to
ensure correct behavior.
26.3.4 Periodic auto-wakeup

The periodic wakeup flag is generated by a 16-bit programmable auto-reload down-counter.
The wakeup timer range can be extended to 17 bits.
The wakeup function is enabled through the WUTE bit in the RTC_CR register.
802/1751 RM0090 Rev 19

The wakeup timer clock input can be:

• RTC clock (RTCCLK) divided by 2, 4, 8, or 16.
When RTCCLK is LSE(32.768kHz), this allows configuring the wakeup interrupt period
from 122 µs to 32 s, with a resolution down to 61µs.
• ck_spre (usually 1 Hz internal clock)
When ck_spre frequency is 1Hz, this allows achieving a wakeup time from 1 s to
around 36 hours with one-second resolution. This large programmable time range is
divided in 2 parts:
– from 1s to 18 hours when WUCKSEL [2:1] = 10
– and from around 18h to 36h when WUCKSEL[2:1] = 11. In this last case 216 is
added to the 16-bit counter current value.When the initialization sequence is
complete (see Programming the wakeup timer on page 805), the timer starts
counting down.When the wakeup function is enabled, the down-counting remains
active in low-power modes. In addition, when it reaches 0, the WUTF flag is set in
the RTC_ISR register, and the wakeup counter is automatically reloaded with its
reload value (RTC_WUTR register value).
The WUTF flag must then be cleared by software.
When the periodic wakeup interrupt is enabled by setting the WUTIE bit in the RTC_CR2
register, it can exit the device from low-power modes.
The periodic wakeup flag can be routed to the RTC_ALARM output provided it has been
enabled through bits OSEL[1:0] of RTC_CR register. RTC_ALARM polarity can be
configured through the POL bit in the RTC_CR register.
System reset, as well as low-power modes (Sleep, Stop and Standby) have no influence on
the wakeup timer.
26.3.5 RTC initialization and configuration

RTC register access
The RTC registers are 32-bit registers. The APB interface introduces 2 wait-states in RTC
register accesses except on read accesses to calendar shadow registers when
BYPSHAD=0.
RTC register write protection

After system reset, the RTC registers are protected against parasitic write access with the
DBP bit of the PWR power control register (PWR_CR). The DBP bit must be set to enable
RTC registers write access.
After backup domain reset, all the RTC registers are write-protected. Writing to the RTC
registers is enabled by writing a key into the Write Protection register, RTC_WPR.
The following steps are required to unlock the write protection on all the RTC registers
except for RTC_ISR[13:8], RTC_TAFCR, and RTC_BKPxR.
1. Write ‘0xCA’ into the RTC_WPR register.
2. Write ‘0x53’ into the RTC_WPR register.
Writing a wrong key reactivates the write protection.
The protection mechanism is not affected by system reset.
RM0090 Rev 19 803/1751

838
Calendar initialization and configuration

To program the initial time and date calendar values, including the time format and the
prescaler configuration, the following sequence is required:
1. Set INIT bit to 1 in the RTC_ISR register to enter initialization mode. In this mode, the
calendar counter is stopped and its value can be updated.
2. Poll INITF bit of in the RTC_ISR register. The initialization phase mode is entered when
INITF is set to 1. It takes from 1 to 2 RTCCLK clock cycles (due to clock
synchronization).
3. To generate a 1 Hz clock for the calendar counter, program first the synchronous
prescaler factor in RTC_PRER register, and then program the asynchronous prescaler
factor. Even if only one of the two fields needs to be changed, 2 separate write
accesses must be performed to the RTC_PRER register.
4. Load the initial time and date values in the shadow registers (RTC_TR and RTC_DR),
and configure the time format (12 or 24 hours) through the FMT bit in the RTC_CR
register.
5. Exit the initialization mode by clearing the INIT bit. The actual calendar counter value is
then automatically loaded and the counting restarts after 4 RTCCLK clock cycles.
When the initialization sequence is complete, the calendar starts counting.
Note: After a system reset, the application can read the INITS flag in the RTC_ISR register to
check if the calendar has been initialized or not. If this flag equals 0, the calendar has not
been initialized since the year field is set at its backup domain reset default value (0x00).
To read the calendar after initialization, the software must first check that the RSF flag is set
in the RTC_ISR register.
Daylight saving time

The daylight saving time management is performed through bits SUB1H, ADD1H, and BKP
of the RTC_CR register.
Using SUB1H or ADD1H, the software can subtract or add one hour to the calendar in one
single operation without going through the initialization procedure.
In addition, the software can use the BKP bit to memorize this operation.
Programming the alarm

A similar procedure must be followed to program or update the programmable alarm (Alarm
A or Alarm B):
1. Clear ALRAE or ALRBIE in RTC_CR to disable Alarm A or Alarm B.
2. Poll ALRAWF or ALRBWF in RTC_ISR until it is set to make sure the access to alarm
registers is allowed. This takes 1 to 2 RTCCLK clock cycles (due to clock
synchronization).
3. Program the Alarm A or Alarm B registers (RTC_ALRMASSR/RTC_ALRMAR or
RTC_ALRMBSSR/RTC_ALRMBR).
4. Set ALRAE or ALRBIE in the RTC_CR register to enable Alarm A or Alarm B again.
Note: Each change of the RTC_CR register is taken into account after 1 to 2 RTCCLK clock cycles
due to clock synchronization.
804/1751 RM0090 Rev 19

Programming the wakeup timer

The following sequence is required to configure or change the wakeup timer auto-reload
value (WUT[15:0] in RTC_WUTR):
1. Clear WUTE in RTC_CR to disable the wakeup timer.
2. Poll WUTWF until it is set in RTC_ISR to make sure the access to wakeup auto-reload
counter and to WUCKSEL[2:0] bits is allowed. It takes 1 to 2 RTCCLK clock cycles
(due to clock synchronization).
3. Program the wakeup auto-reload value WUT[15:0], and the wakeup clock selection
(WUCKSEL[2:0] bits in RTC_CR).Set WUTE in RTC_CR to enable the timer again.
The wakeup timer restarts down-counting. The WUTWF bit is cleared up to 2 RTCCLK
clocks cycles after WUTE is cleared, due to clock synchronization.
26.3.6 Reading the calendar

When BYPSHAD control bit is cleared in the RTC_CR register
To read the RTC calendar registers (RTC_SSR, RTC_TR and RTC_DR) properly, the APB1
clock frequency (fPCLK1) must be equal to or greater than seven times the fRTCCLK RTC
clock frequency. This ensures a secure behavior of the synchronization mechanism.
If the APB1 clock frequency is less than seven times the RTC clock frequency, the software
must read the calendar time and date registers twice. If the second read of the RTC_TR
gives the same result as the first read, this ensures that the data is correct. Otherwise a third
read access must be done. In any case the APB1 clock frequency must never be lower than
the RTC clock frequency.
The RSF bit is set in RTC_ISR register each time the calendar registers are copied into the
RTC_SSR, RTC_TR and RTC_DR shadow registers. The copy is performed every two
RTCCLK cycles. To ensure consistency between the 3 values, reading either RTC_SSR or
RTC_TR locks the values in the higher-order calendar shadow registers until RTC_DR is
read. In case the software makes read accesses to the calendar in a time interval smaller
than 2 RTCCLK periods: RSF must be cleared by software after the first calendar read, and
then the software must wait until RSF is set before reading again the RTC_SSR, RTC_TR
and RTC_DR registers.
After waking up from low-power mode (Stop or Standby), RSF must be cleared by software.
The software must then wait until it is set again before reading the RTC_SSR, RTC_TR and
RTC_DR registers.
The RSF bit must be cleared after wakeup and not before entering low-power mode.
Note: After a system reset, the software must wait until RSF is set before reading the RTC_SSR,
RTC_TR and RTC_DR registers. Indeed, a system reset resets the shadow registers to
their default values.
After an initialization (refer to Calendar initialization and configuration on page 804): the
software must wait until RSF is set before reading the RTC_SSR, RTC_TR and RTC_DR
registers.
After synchronization (refer to Section 26.3.8: RTC synchronization): the software must wait
until RSF is set before reading the RTC_SSR, RTC_TR and RTC_DR registers.
RM0090 Rev 19 805/1751

838
When the BYPSHAD control bit is set in the RTC_CR register (bypass shadow
registers)
Reading the calendar registers gives the values from the calendar counters directly, thus
eliminating the need to wait for the RSF bit to be set. This is especially useful after exiting
from low-power modes (STOP or Standby), since the shadow registers are not updated
during these modes.
When the BYPSHAD bit is set to 1, the results of the different registers might not be
coherent with each other if an RTCCLK edge occurs between two read accesses to the
registers. Additionally, the value of one of the registers may be incorrect if an RTCCLK edge
occurs during the read operation. The software must read all the registers twice, and then
compare the results to confirm that the data is coherent and correct. Alternatively, the
software can just compare the two results of the least-significant calendar register.
Note: While BYPSHAD=1, instructions which read the calendar registers require one extra APB
cycle to complete.
26.3.7 Resetting the RTC

The calendar shadow registers (RTC_SSR, RTC_TR and RTC_DR) and some bits of the
RTC status register (RTC_ISR) are reset to their default values by all available system reset
sources.
On the contrary, the following registers are resetted to their default values by a backup
domain reset and are not affected by a system reset: the RTC current calendar registers,
the RTC control register (RTC_CR), the prescaler register (RTC_PRER), the RTC
calibration registers (RTC_CALIBR or RTC_CALR), the RTC shift register (RTC_SHIFTR),
the RTC timestamp registers (RTC_TSSSR, RTC_TSTR and RTC_TSDR), the RTC tamper
and alternate function configuration register (RTC_TAFCR), the RTC backup registers
(RTC_BKPxR), the wakeup timer register (RTC_WUTR), the Alarm A and Alarm B registers
(RTC_ALRMASSR/RTC_ALRMAR and RTC_ALRMBSSR/RTC_ALRMBR).
In addition, when the RTC is clocked by the LSE, it goes on running under system reset if
the reset source is different from a backup domain reset one. Refer to section RTC clock of
the Reset and clock controller for details about the list of the RTC clock sources that are not
affected by system reset.
When a backup domain reset occurs, the RTC is stopped and all the RTC registers are set
to their reset values.
26.3.8 RTC synchronization

The RTC can be synchronized to a remote clock with a high degree of precision. After
reading the sub-second field (RTC_SSR or RTC_TSSSR), a calculation can be made of the
precise offset between the times being maintained by the remote clock and the RTC. The
RTC can then be adjusted to eliminate this offset by “shifting” its clock by a fraction of a
second using RTC_SHIFTR.
RTC_SSR contains the value of the synchronous prescaler’s counter. This allows one to
calculate the exact time being maintained by the RTC down to a resolution of
1 / (PREDIV_S + 1) seconds. As a consequence, the resolution can be improved by
increasing the synchronous prescaler value (PREDIV_S[14:0]. The maximum resolution
allowed (30.52 μs with a 32768 Hz clock) is obtained with PREDIV_S set to 0x7FFF.
806/1751 RM0090 Rev 19

However, increasing PREDIV_S means that PREDIV_A must be decreased in order to

maintain the synchronous prescaler’s output at 1 Hz. In this way, the frequency of the
asynchronous prescaler’s output increases, which may increase the RTC dynamic
consumption.
The RTC can be finely adjusted using the RTC shift control register (RTC_SHIFTR). Writing
to RTC_SHIFTR can shift (either delay or advance) the clock by up to a second with a
resolution of 1 / (PREDIV_S + 1) seconds. The shift operation consists of adding the
SUBFS[14:0] value to the synchronous prescaler counter SS[15:0]: this will delay the clock.
If at the same time the ADD1S bit is set, this results in adding one second and at the same
time subtracting a fraction of second, so this will advance the clock.
Caution: Before initiating a shift operation, the user must check that SS[15] = 0 in order to ensure that
no overflow will occur.
As soon as a shift operation is initiated by a write to the RTC_SHIFTR register, the SHPF
flag is set by hardware to indicate that a shift operation is pending. This bit is cleared by
hardware as soon as the shift operation has completed.
Caution: This synchronization feature is not compatible with the reference clock detection feature:
firmware must not write to RTC_SHIFTR when REFCKON=1.
26.3.9 RTC reference clock detection

The RTC calendar update can be synchronized to a reference clock RTC_REFIN, usually
the mains (50 or 60 Hz). The RTC_REFIN reference clock should have a higher precision
than the 32.768 kHz LSE clock. When the RTC_REFIN detection is enabled (REFCKON bit
of RTC_CR set to 1), the calendar is still clocked by the LSE, and RTC_REFIN is used to
compensate for the imprecision of the calendar update frequency (1 Hz).
Each 1 Hz clock edge is compared to the nearest reference clock edge (if one is found
within a given time window). In most cases, the two clock edges are properly aligned. When
the 1 Hz clock becomes misaligned due to the imprecision of the LSE clock, the RTC shifts
the 1 Hz clock a bit so that future 1 Hz clock edges are aligned. Thanks to this mechanism,
the calendar becomes as precise as the reference clock.
The RTC detects if the reference clock source is present by using the 256 Hz clock
(ck_apre) generated from the 32.768 kHz quartz. The detection is performed during a time
window around each of the calendar updates (every 1 s). The window equals 7 ck_apre
periods when detecting the first reference clock edge. A smaller window of 3 ck_apre
periods is used for subsequent calendar updates.
Each time the reference clock is detected in the window, the synchronous prescaler which
outputs the ck_spre clock is forced to reload. This has no effect when the reference clock
and the 1 Hz clock are aligned because the prescaler is being reloaded at the same
moment. When the clocks are not aligned, the reload shifts future 1 Hz clock edges a little
for them to be aligned with the reference clock.
If the reference clock halts (no reference clock edge occurred during the 3 ck_apre window),
the calendar is updated continuously based solely on the LSE clock. The RTC then waits for
the reference clock using a large 7 ck_apre period detection window centered on the
ck_spre edge.
RM0090 Rev 19 807/1751

838
When the reference clock detection is enabled, PREDIV_A and PREDIV_S must be set to
their default values:
• PREDIV_A = 0x007F
• PREDIV_S = 0x00FF
Note: The reference clock detection is not available in Standby mode.
Caution: The reference clock detection feature cannot be used in conjunction with the coarse digital
calibration: RTC_CALIBR must be kept at 0x0000 0000 when REFCKON=1.
26.3.10 RTC coarse digital calibration

Two digital calibration methods are available: coarse and smooth calibration. To perform
coarse calibration refer to Section 26.6.7: RTC calibration register (RTC_CALIBR).
The two calibration methods are not intended to be used together, the application must
select one of the two methods. Coarse calibration is provided for compatibly reasons. To
perform smooth calibration refer to Section 26.3.11: RTC smooth digital calibration and the
Section 26.6.16: RTC calibration register (RTC_CALR)
The coarse digital calibration can be used to compensate crystal inaccuracy by adding
(positive calibration) or masking (negative calibration) clock cycles at the output of the
asynchronous prescaler (ck_apre).
Positive and negative calibration are selected by setting the DCS bit in RTC_CALIBR
register to ‘0’ and ‘1’, respectively.
When positive calibration is enabled (DCS = ‘0’), 2 ck_apre cycles are added every minute
(around 15360 ck_apre cycles) for 2xDC minutes. This causes the calendar to be updated
sooner, thereby adjusting the effective RTC frequency to be a bit higher.
When negative calibration is enabled (DCS = ‘1’), 1 ck_apre cycle is removed every minute
(around 15360 ck_apre cycles) for 2xDC minutes. This causes the calendar to be updated
later, thereby adjusting the effective RTC frequency to be a bit lower.
DC is configured through bits DC[4:0] of RTC_CALIBR register. This number ranges from 0
to 31 corresponding to a time interval (2xDC) ranging from 0 to 62.
The coarse digital calibration can be configured only in initialization mode, and starts when
the INIT bit is cleared. The full calibration cycle lasts 64 minutes. The first 2xDC minutes of
the 64 -minute cycle are modified as just described.
Negative calibration can be performed with a resolution of about 2 ppm while positive
calibration can be performed with a resolution of about 4 ppm. The maximum calibration
ranges from -63 ppm to 126 ppm.
The calibration can be performed either on the LSE or on the HSE clock.
Caution: Digital calibration may not work correctly if PREDIV_A < 6.
Case of RTCCLK=32.768 kHz and PREDIV_A+1=128

The following description assumes that ck_apre frequency is 256 Hz obtained with an LSE
clock nominal frequency of 32.768 kHz, and PREDIV_A set to 127 (default value).
The ck_spre clock frequency is only modified during the first 2xDC minutes of the 64-minute
cycle. For example, when DC equals 1, only the first 2 minutes are modified. This means
that the first 2xDC minutes of each 64-minute cycle have, once per minute, one second
808/1751 RM0090 Rev 19

either shortened by 256 or lengthened by 128 RTCCLK cycles, given that each ck_apre
cycle represents 128 RTCCLK cycles (with PREDIV_A+1=128).
Therefore each calibration step has the effect of adding 512 or subtracting 256 oscillator
cycles for every 125829120 RTCCLK cycles (64min x 60 s/min x 32768 cycles/s). This is
equivalent to +4.069 ppm or -2.035 ppm per calibration step. As a result, the calibration
resolution is +10.5 or -5.27 seconds per month, and the total calibration ranges from +5.45
to -2.72 minutes per month.
In order to measure the clock deviation, a 512 Hz clock is output for calibration.Refer to
Section 26.3.14: Calibration clock output.
26.3.11 RTC smooth digital calibration

RTC frequency can be digitally calibrated with a resolution of about 0.954 ppm with a range
from -487.1 ppm to +488.5 ppm. The correction of the frequency is performed using series
of small adjustments (adding and/or subtracting individual RTCCLK pulses). These
adjustments are fairly well distributed so that the RTC is well calibrated even when observed
over short durations of time.
The smooth digital calibration is performed during a cycle of about 220 RTCCLK pulses, or
32 seconds when the input frequency is 32768 Hz. This cycle is maintained by a 20-bit
counter, cal_cnt[19:0], clocked by RTCCLK.
The smooth calibration register (RTC_CALR) specifies the number of RTCCLK clock cycles
to be masked during the 32-second cycle:
• Setting the bit CALM[0] to 1 causes exactly one pulse to be masked during the 32-
second cycle.
• Setting CALM[1] to 1 causes two additional cycles to be masked
• Setting CALM[2] to 1 causes four additional cycles to be masked
• and so on up to CALM[8] set to 1 which causes 256 clocks to be masked.
Note: CALM[8:0] (RTC_CALRx) specifies the number of RTCCLK pulses to be masked during the
32-second cycle. Setting the bit CALM[0] to ‘1’ causes exactly one pulse to be masked
during the 32-second cycle at the moment when cal_cnt[19:0] is 0x80000; CALM[1]=1
causes two other cycles to be masked (when cal_cnt is 0x40000 and 0xC0000); CALM[2]=1
causes four other cycles to be masked (cal_cnt = 0x20000/0x60000/0xA0000/ 0xE0000);
and so on up to CALM[8]=1 which causes 256 clocks to be masked (cal_cnt = 0xXX800).
While CALM allows the RTC frequency to be reduced by up to 487.1 ppm with fine
resolution, the bit CALP can be used to increase the frequency by 488.5 ppm. Setting CALP
to ‘1’ effectively inserts an extra RTCCLK pulse every 211 RTCCLK cycles, which means
that 512 clocks are added during every 32-second cycle.
Using CALM together with CALP, an offset ranging from -511 to +512 RTCCLK cycles can
be added during the 32-second cycle, which translates to a calibration range of -487.1 ppm
to +488.5 ppm with a resolution of about 0.954 ppm.
The formula to calculate the effective calibrated frequency (FCAL) given the input frequency
(FRTCCLK) is as follows:
FCAL = FRTCCLK x [1 + (CALP x 512 - CALM) / (220 + CALM - CALP x 512)]
Calibration when PREDIV_A<3

The CALP bit can not be set to 1 when the asynchronous prescaler value (PREDIV_A bits in
RTC_PRER register) is less than 3. If CALP was already set to 1 and PREDIV_A bits are
RM0090 Rev 19 809/1751

838
set to a value less than 3, CALP is ignored and the calibration operates as if CALP was
equal to 0.
To perform a calibration with PREDIV_A less than 3, the synchronous prescaler value
(PREDIV_S) should be reduced so that each second is accelerated by 8 RTCCLK clock
cycles, which is equivalent to adding 256 clock cycles every 32 seconds. As a result,
between 255 and 256 clock pulses (corresponding to a calibration range from 243.3 to
244.1 ppm) can effectively be added during each 32-second cycle using only the CALM bits.
With a nominal RTCCLK frequency of 32768 Hz, when PREDIV_A equals 1 (division factor
of 2), PREDIV_S should be set to 16379 rather than 16383 (4 less). The only other
interesting case is when PREDIV_A equals 0, PREDIV_S should be set to 32759 rather
than 32767 (8 less).
If PREDIV_S is reduced in this way, the formula given the effective frequency of the
calibrated input clock is as follows:
FCAL = FRTCCLK x [1 + (256 - CALM) / (220 + CALM - 256)]
In this case, CALM[7:0] equals 0x100 (the midpoint of the CALM range) is the correct
setting if RTCCLK is exactly 32768.00 Hz.
Verifying the RTC calibration

RTC precision is performed by measuring the precise frequency of RTCCLK and calculating
the correct CALM value and CALP values. An optional 1 Hz output is provided to allow
applications to measure and verify the RTC precision.
Measuring the precise frequency of the RTC over a limited interval can result in a
measurement error of up to 2 RTCCLK clock cycles over the measurement period,
depending on how the digital calibration cycle is aligned with the measurement period.
However, this measurement error can be eliminated if the measurement period is the same
length as the calibration cycle period. In this case, the only error observed is the error due to
the resolution of the digital calibration.
• By default, the calibration cycle period is 32 seconds.
Using this mode and measuring the accuracy of the 1 Hz output over exactly 32
seconds guarantees that the measure is within 0.477 ppm (0.5 RTCCLK cycles over 32
seconds, due to the limitation of the calibration resolution).
• CALW16 bit of the RTC_CALR register can be set to 1 to force a 16- second calibration
cycle period.
In this case, the RTC precision can be measured during 16 seconds with a maximum
error of 0.954 ppm (0.5 RTCCLK cycles over 16 seconds). However, since the
calibration resolution is reduced, the long term RTC precision is also reduced to 0.954
ppm: CALM[0] bit is stuck at 0 when CALW16 is set to 1.
• CALW8 bit of the RTC_CALR register can be set to 1 to force a 8- second calibration
cycle period.
In this case, the RTC precision can be measured during 8 seconds with a maximum
error of 1.907 ppm (0.5 RTCCLK cycles over 8s). The long term RTC precision is also
reduced to 1.907 ppm: CALM[1:0] bits are stuck at 00 when CALW8 is set to 1.
Re-calibration on-the-fly
The calibration register (RTC_CALR) can be updated on-the-fly while RTC_ISR/INITF=0, by
using the follow process:
810/1751 RM0090 Rev 19

1. Poll the RTC_ISR/RECALPF (re-calibration pending flag).

2. If it is set to 0, write a new value to RTC_CALR, if necessary. RECALPF is then
automatically set to 1
3. Within three ck_apre cycles after the write operation to RTC_CALR, the new calibration
settings take effect.
26.3.12 Timestamp function

Timestamp is enabled by setting the TSE bit of RTC_CR register to 1.
The calendar is saved in the timestamp registers (RTC_TSSSR, RTC_TSTR, RTC_TSDR)
when a timestamp event is detected on the pin to which the TIMESTAMP alternate function
is mapped. When a timestamp event occurs, the timestamp flag bit (TSF) in RTC_ISR
register is set.
By setting the TSIE bit in the RTC_CR register, an interrupt is generated when a timestamp
event occurs.
If a new timestamp event is detected while the timestamp flag (TSF) is already set, the
timestamp overflow flag (TSOVF) flag is set and the timestamp registers (RTC_TSTR and
RTC_TSDR) maintain the results of the previous event.
Note: TSF is set 2 ck_apre cycles after the timestamp event occurs due to synchronization
process.
There is no delay in the setting of TSOVF. This means that if two timestamp events are
close together, TSOVF can be seen as '1' while TSF is still '0'. As a consequence, it is
recommended to poll TSOVF only after TSF has been set.
Caution: If a timestamp event occurs immediately after the TSF bit is supposed to be cleared, then
both TSF and TSOVF bits are set. To avoid masking a timestamp event occurring at the
same moment, the application must not write ‘0’ into TSF bit unless it has already read it to
‘1’.
Optionally, a tamper event can cause a timestamp to be recorded. See the description of the
TAMPTS control bit in Section 26.6.17: RTC tamper and alternate function configuration
register (RTC_TAFCR). If the timestamp event is on the same pin as a tamper event
configured in filtered mode (TAMPFLT set to a non-zero value), the timestamp on tamper
detection event mode must be selected by setting TAMPTS='1' in RTC_TAFCR register.
TIMESTAMP alternate function

The TIMESTAMP alternate function (RTC_TS) can be mapped either to RTC_AF1 or to
RTC_AF2 depending on the value of the TSINSEL bit in the RTC_TAFCR register (see
Section 26.6.17: RTC tamper and alternate function configuration register (RTC_TAFCR)).
Mapping the timestamp event on RTC_AF2 is not allowed if RTC_AF1 is used as TAMPER
in filtered mode (TAMPFLT set to a non-zero value).
26.3.13 Tamper detection

Two tamper detection inputs are available. They can be configured either for edge detection,
or for level detection with filtering.
RTC backup registers

The backup registers (RTC_BKPxR) are twenty 32-bit registers for storing 80 bytes of user
application data. They are implemented in the backup domain that remains powered-on by
RM0090 Rev 19 811/1751

838
VBAT when the VDD power is switched off. They are not reset by system reset or when the
device wakes up from Standby mode. They are reset by a backup domain reset
The backup registers are reset when a tamper detection event occurs (see Section 26.6.20:
RTC backup registers (RTC_BKPxR) and Tamper detection initialization on page 812.
Tamper detection initialization

Each tamper detection input is associated with the TAMP1F/TAMP2F flags in the RTC_ISR2
register. Each input can be enabled by setting the corresponding TAMP1E/TAMP2E bits to 1
in the RTC_TAFCR register.
A tamper detection event resets all backup registers (RTC_BKPxR).
By setting the TAMPIE bit in the RTC_TAFCR register, an interrupt is generated when a
tamper detection event occurs.
Timestamp on tamper event

With TAMPTS set to ‘1 , any tamper event causes a timestamp to occur. In this case, either
the TSF bit or the TSOVF bit are set in RTC_ISR, in the same manner as if a normal
timestamp event occurs. The affected tamper flag register (TAMP1F, TAMP2F) is set at the
same time that TSF or TSOVF is set.
Edge detection on tamper inputs

If the TAMPFLT bits are “00”, the TAMPER pins generate tamper detection events
(RTC_TAMP[2:1]) when either a rising edge is observed or an falling edge is observed
depending on the corresponding TAMPxTRG bit. The internal pull-up resistors on the
TAMPER inputs are deactivated when edge detection is selected.
Caution: To avoid losing tamper detection events, the signal used for edge detection is logically
ANDed with TAMPxE in order to detect a tamper detection event in case it occurs before the
TAMPERx pin is enabled.
• When TAMPxTRG = 0: if the TAMPERx alternate function is already high before
tamper detection is enabled (TAMPxE bit set to 1), a tamper event is detected as soon
as TAMPERx is enabled, even if there was no rising edge on TAMPERx after TAMPxE
was set.
• When TAMPxTRG = 1: if the TAMPERx alternate function is already low before tamper
detection is enabled, a tamper event is detected as soon as TAMPERx is enabled
(even if there was no falling edge on TAMPERx after TAMPxE was set.
After a tamper event has been detected and cleared, the TAMPERx alternate function
should be disabled and then re-enabled (TAMPxE set to 1) before re-programming the
backup registers (RTC_BKPxR). This prevents the application from writing to the backup
registers while the TAMPERx value still indicates a tamper detection. This is equivalent to a
level detection on the TAMPERx alternate function.
Note: Tamper detection is still active when VDD power is switched off. To avoid unwanted resetting
of the backup registers, the pin to which the TAMPER alternate function is mapped should
be externally tied to the correct level.
Level detection with filtering on tamper inputs

Level detection with filtering is performed by setting TAMPFLT to a non-zero value. A tamper
detection event is generated when either 2, 4, or 8 (depending on TAMPFLT) consecutive
812/1751 RM0090 Rev 19

samples are observed at the level designated by the TAMPxTRG bits

(TAMP1TRG/TAMP2TRG).
The TAMPER inputs are pre-charged through the I/O internal pull-up resistance before its
state is sampled, unless disabled by setting TAMPPUDIS to 1,The duration of the precharge
is determined by the TAMPPRCH bits, allowing for larger capacitances on the tamper
inputs.
The trade-off between tamper detection latency and power consumption through the pull-up
can be optimized by using TAMPFREQ to determine the frequency of the sampling for level
detection.
Note: Refer to the datasheets for the electrical characteristics of the pull-up resistors.
TAMPER alternate function detection

The TAMPER1 alternate function (RTC_TAMP1) can be mapped either to RTC_AF1(PC13)
or RTC_AF2 (PI8) depending on the value of TAMP1INSEL bit in RTC_TAFCR register (see
Section 26.6.17: RTC tamper and alternate function configuration register (RTC_TAFCR)).
TAMPE bit must be cleared when TAMP1INSEL is modified to avoid unwanted setting of
TAMPF.
The TAMPER 2 alternate function corresponds to RTC_TAMP2 pin.
26.3.14 Calibration clock output

When the COE bit is set to 1 in the RTC_CR register, a reference clock is provided on the
RTC_CALIB device output. If the COSEL bit in the RTC_CR register is reset and
PREDIV_A = 0x7F, the RTC_CALIB frequency is fRTCCLK/64. This corresponds to a
calibration output at 512 Hz for an RTCCLK frequency at 32.768 kHz.
The RTC_CALIB output is not impacted by the calibration value programmed in
RTC_CALIBR register. The RTC_CALIB duty cycle is irregular: there is a light jitter on falling
edges. It is therefore recommended to use rising edges.
If COSEL is set and (PREDIV_S+1) is a non-zero multiple of 256 (i.e: PREDIV_S[7:0] =
0xFF), the RTC_CALIB frequency is fRTCCLK/(256 * (PREDIV_A+1)). This corresponds to a
calibration output at 1 Hz for prescaler default values (PREDIV_A = 0x7F, PREDIV_S =
0xFF), with an RTCCLK frequency at 32.768 kHz. The 1 Hz output is affected when a shift
operation is on going and may toggle during the shift operation (SHPF=1).
Calibration alternate function output

When the COE bit in the RTC_CR register is set to 1, the calibration alternate function
(RTC_CALIB) is enabled on RTC_AF1.
Note: When RTC_CALIB or RTC_ALARM is selected, RTC_AF1 is automatically configured in
output alternate function.
26.3.15 Alarm output

Three functions can be selected on Alarm output: ALRAF, ALRBF and WUTF. These
functions reflect the contents of the corresponding flags in the RTC_ISR register.
The OSEL[1:0] control bits in the RTC_CR register are used to activate the alarm alternate
function output (RTC_ALARM) in RTC_AF1, and to select the function which is output on
RTC_ALARM.
RM0090 Rev 19 813/1751

838
The polarity of the output is determined by the POL control bit in RTC_CR so that the
opposite of the selected flag bit is output when POL is set to 1.
Alarm alternate function output

RTC_ALARM can be configured in output open drain or output push-pull using the control
bit ALARMOUTTYPE in the RTC_TAFCR register.
Note: Once RTC_ALARM is enabled, it has priority over RTC_CALIB (COE bit is don't care on
RTC_AF1).
When RTC_CALIB or RTC_ALARM is selected, RTC_AF1 is automatically configured in
output alternate function.
26.4 RTC and low-power modes

Table 119. Effect of low-power modes on RTC
Mode Description
No effect
Sleep
RTC interrupts cause the device to exit the Sleep mode.
The RTC remains active when the RTC clock source is LSE or LSI. RTC alarm, RTC
Stop tamper event, RTC time stamp event, and RTC Wakeup cause the device to exit the Stop
mode.
The RTC remains active when the RTC clock source is LSE or LSI. RTC alarm, RTC
Standby tamper event, RTC time stamp event, and RTC Wakeup cause the device to exit the
Standby mode.
26.5 RTC interrupts

All RTC interrupts are connected to the EXTI controller.
To enable the RTC Alarm interrupt, the following sequence is required:
1. Configure and enable the EXTI Line 17 in interrupt mode and select the rising edge
sensitivity.
2. Configure and enable the RTC_Alarm IRQ channel in the NVIC.
3. Configure the RTC to generate RTC alarms (Alarm A or Alarm B).
To enable the RTC Wakeup interrupt, the following sequence is required:
sensitivity.
2. Configure and enable the RTC_WKUP IRQ channel in the NVIC.
3. Configure the RTC to generate the RTC wakeup timer event.
To enable the RTC Tamper interrupt, the following sequence is required:
sensitivity.
2. Configure and Enable the TAMP_STAMP IRQ channel in the NVIC.
3. Configure the RTC to detect the RTC tamper event.
To enable the RTC TimeStamp interrupt, the following sequence is required:
814/1751 RM0090 Rev 19

sensitivity.
2. Configure and Enable the TAMP_STAMP IRQ channel in the NVIC.
3. Configure the RTC to detect the RTC timestamp event.
Table 120. Interrupt control bits

Enable Exit the Exit the Exit the
Interrupt event Event flag control Sleep Stop Standby
bit mode mode mode
Alarm A ALRAF ALRAIE yes yes(1) yes(1)

Alarm B ALRBF ALRBIE yes yes(1) yes(1)
Wakeup WUTF WUTIE yes yes(1) yes(1)
TimeStamp TSF TSIE yes yes(1) yes(1)
Tamper1 detection TAMP1F TAMPIE yes yes(1) yes(1)
Tamper2 detection(2) TAMP2F TAMPIE yes yes(1) yes(1)
1. Wakeup from STOP and Standby modes is possible only when the RTC clock source is LSE or LSI.
2. If RTC_TAMPER2 pin is present. Refer to device datasheet pinout.
RM0090 Rev 19 815/1751

838
26.6 RTC registers

26.6.1 RTC time register (RTC_TR)

The RTC_TR is the calendar time shadow register. This register must be written in
initialization mode only. Refer to Calendar initialization and configuration on page 804 and
Reading the calendar on page 805.
Backup domain reset value: 0x0000 0000
System reset: 0x0000 0000 when BYPSHAD = 0. Not affected when BYPSHAD = 1.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PM HT[1:0] HU[3:0]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MNT[2:0] MNU[3:0] ST[2:0] SU[3:0]
Res. Res.
Bits 31-24 Reserved

Bit 22 PM: AM/PM notation
0: AM or 24-hour format
1: PM
Bits 21:20 HT[1:0]: Hour tens in BCD format
Bits 19:16 HU[3:0]: Hour units in BCD format
Bits 14:12 MNT[2:0]: Minute tens in BCD format
Bit 11:8 MNU[3:0]: Minute units in BCD format
Bits 6:4 ST[2:0]: Second tens in BCD format
Bits 3:0 SU[3:0]: Second units in BCD format
Note: This register is write protected. The write access procedure is described in RTC register
write protection on page 803.
816/1751 RM0090 Rev 19

26.6.2 RTC date register (RTC_DR)

The RTC_DR is the calendar date shadow register. This register must be written in
initialization mode only. Refer to Calendar initialization and configuration on page 804 and
Reading the calendar on page 805.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
YT[3:0] YU[3:0]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
WDU[2:0] MT MU[3:0] DT[1:0] DU[3:0]
Reserved
Bits 31-24 Reserved

Bits 23:20 YT[3:0]: Year tens in BCD format
Bits 19:16 YU[3:0]: Year units in BCD format
Bits 15:13 WDU[2:0]: Week day units
000: forbidden
001: Monday
...
111: Sunday
Bit 12 MT: Month tens in BCD format
Bits 11:8 MU: Month units in BCD format
Bits 5:4 DT[1:0]: Date tens in BCD format
Bits 3:0 DU[3:0]: Date units in BCD format
RM0090 Rev 19 817/1751

838
26.6.3 RTC control register (RTC_CR)

System reset: not affected
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
COE OSEL[1:0] POL COSEL BKP SUB1H ADD1H
Reserved
rw rw rw rw rw rw w w
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
BYPS
TSIE WUTIE ALRBIE ALRAIE TSE WUTE ALRBE ALRAE DCE FMT REFCKON TSEDGE WUCKSEL[2:0]
HAD

Bit 23 COE: Calibration output enable
This bit enables the RTC_CALIB output
0: Calibration output disabled
1: Calibration output enabled
Bits 22:21 OSEL[1:0]: Output selection
These bits are used to select the flag to be routed to RTC_ALARM output
00: Output disabled
01: Alarm A output enabled
10:Alarm B output enabled
11: Wakeup output enabled
Bit 20 POL: Output polarity
This bit is used to configure the polarity of RTC_ALARM output
0: The pin is high when ALRAF/ALRBF/WUTF is asserted (depending on OSEL[1:0])
1: The pin is low when ALRAF/ALRBF/WUTF is asserted (depending on OSEL[1:0]).
Bit 19 COSEL: Calibration output selection
When COE=1, this bit selects which signal is output on RTC_CALIB.
0: Calibration output is 512 Hz
1: Calibration output is 1 Hz
These frequencies are valid for RTCCLK at 32.768 kHz and prescalers at their default
values (PREDIV_A=127 and PREDIV_S=255). Refer to Section 26.3.14: Calibration clock
output
Bit 18 BKP: Backup
This bit can be written by the user to memorize whether the daylight saving time change has
been performed or not.
Bit 17 SUB1H: Subtract 1 hour (winter time change)
When this bit is set outside initialization mode, 1 hour is subtracted to the calendar time if the
current hour is not 0. This bit is always read as 0.
Setting this bit has no effect when current hour is 0.
0: No effect
1: Subtracts 1 hour to the current time. This can be used for winter time change.
818/1751 RM0090 Rev 19

Bit 16 ADD1H: Add 1 hour (summer time change)

When this bit is set, 1 hour is added to the calendar time. This bit is always read as 0.
0: No effect
1: Adds 1 hour to the current time. This can be used for summer time change outside
initialization mode.
Bit 15 TSIE: Timestamp interrupt enable
0: Timestamp Interrupt disable
1: Timestamp Interrupt enable
Bit 14 WUTIE: Wakeup timer interrupt enable
0: Wakeup timer interrupt disabled
1: Wakeup timer interrupt enabled
Bit 13 ALRBIE: Alarm B interrupt enable
0: Alarm B Interrupt disable
1: Alarm B Interrupt enable
Bit 12 ALRAIE: Alarm A interrupt enable
0: Alarm A interrupt disabled
1: Alarm A interrupt enabled
Bit 11 TSE: Time stamp enable
0: Time stamp disable
1: Time stamp enable
Bit 10 WUTE: Wakeup timer enable
0: Wakeup timer disabled
1: Wakeup timer enabled
Note: When the wakeup timer is disabled, wait for WUTWF=1 before enabling it again.
Bit 9 ALRBE: Alarm B enable
0: Alarm B disabled
1: Alarm B enabled
Bit 8 ALRAE: Alarm A enable
0: Alarm A disabled
1: Alarm A enabled
Bit 7 DCE: Coarse digital calibration enable
0: Digital calibration disabled
1: Digital calibration enabled
PREDIV_A must be 6 or greater
Bit 6 FMT: Hour format
0: 24 hour/day format
1: AM/PM hour format
Bit 5 BYPSHAD: Bypass the shadow registers
0: Calendar values (when reading from RTC_SSR, RTC_TR, and RTC_DR) are taken from
the shadow registers, which are updated once every two RTCCLK cycles.
1: Calendar values (when reading from RTC_SSR, RTC_TR, and RTC_DR) are taken
directly from the calendar counters.
Note: If the frequency of the APB1 clock is less than seven times the frequency of RTCCLK,
BYPSHAD must be set to ‘1’.
RM0090 Rev 19 819/1751

838
Bit 4 REFCKON: Reference clock detection enable (50 or 60 Hz)

0: Reference clock detection disabled
1: Reference clock detection enabled
Note: PREDIV_S must be 0x00FF.
Bit 3 TSEDGE: Timestamp event active edge
0: TIMESTAMP rising edge generates a timestamp event
1: TIMESTAMP falling edge generates a timestamp event
TSE must be reset when TSEDGE is changed to avoid unwanted TSF setting
Bits 2:0 WUCKSEL[2:0]: Wakeup clock selection
000: RTC/16 clock is selected
10x: ck_spre (usually 1 Hz) clock is selected
11x: ck_spre (usually 1 Hz) clock is selected and 216 is added to the WUT counter value
(see note below)
Note: WUT = Wakeup unit counter value. WUT = (0x0000 to 0xFFFF) + 0x10000 added when
WUCKSEL[2:1 = 11].
Bits 7, 6 and 4 of this register can be written in initialization mode only (RTC_ISR/INITF = 1).
Bits 2 to 0 of this register can be written only when RTC_CR WUTE bit = 0 and RTC_ISR
WUTWF bit = 1.
It is recommended not to change the hour during the calendar hour increment as it could
mask the incrementation of the calendar hour.
ADD1H and SUB1H changes are effective in the next second.
To avoid spuriously setting of TSF, TSE must be reset when TSEDGE is changed.
This register is write protected. The write access procedure is described in RTC register
26.6.4 RTC initialization and status register (RTC_ISR)

System reset value: Not affected except INIT, INITF and RSF which are cleared to 0.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
RECAL
PF
Reserved
r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TAMP TAMP WUT ALRB ALRA
Res. TSOVF TSF WUTF ALRBF ALRAF INIT INITF RSF INITS SHPF
2F 1F WF WF WF
rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rw r rc_w0 r r r r r
820/1751 RM0090 Rev 19

Bits 31:17 Reserved

Bit 16 RECALPF: Recalibration pending Flag
The RECALPF status flag is automatically set to ‘1’ when software writes to the RTC_CALR
register, indicating that the RTC_CALR register is blocked. When the new calibration
settings are taken into account, this bit returns to ‘0’. Refer to Section : Re-calibration on-the-
fly.
Bit 14 TAMP2F: TAMPER2 detection flag
This flag is set by hardware when a tamper detection event is detected on tamper input 2.
It is cleared by software writing 0.
Bit 13 TAMP1F: Tamper detection flag
This flag is set by hardware when a tamper detection event is detected.
It is cleared by software writing 0.
Bit 12 TSOVF: Timestamp overflow flag
This flag is set by hardware when a timestamp event occurs while TSF is already set.
This flag is cleared by software by writing 0. It is recommended to check and then clear
TSOVF only after clearing the TSF bit. Otherwise, an overflow might not be noticed if a
timestamp event occurs immediately before the TSF bit is cleared.
Bit 11 TSF: Timestamp flag
This flag is set by hardware when a timestamp event occurs.
This flag is cleared by software by writing 0.
Bit 10 WUTF: Wakeup timer flag
This flag is set by hardware when the wakeup auto-reload counter reaches 0.
This flag must be cleared by software at least 1.5 RTCCLK periods before WUTF is set to 1
again.
Bit 9 ALRBF: Alarm B flag
This flag is set by hardware when the time/date registers (RTC_TR and RTC_DR) match the
Alarm B register (RTC_ALRMBR).
Bit 8 ALRAF: Alarm A flag
This flag is set by hardware when the time/date registers (RTC_TR and RTC_DR) match the
Alarm A register (RTC_ALRMAR).
Bit 7 INIT: Initialization mode
0: Free running mode
1: Initialization mode used to program time and date register (RTC_TR and RTC_DR), and
prescaler register (RTC_PRER). Counters are stopped and start counting from the new
value when INIT is reset.
Bit 6 INITF: Initialization flag
When this bit is set to 1, the RTC is in initialization state, and the time, date and prescaler
registers can be updated.
0: Calendar registers update is not allowed
1: Calendar registers update is allowed.
RM0090 Rev 19 821/1751

838
Bit 5 RSF: Registers synchronization flag

This bit is set by hardware each time the calendar registers are copied into the shadow
registers (RTC_SSRx, RTC_TRx and RTC_DRx). This bit is cleared by hardware in
initialization mode, while a shift operation is pending (SHPF=1), or when in bypass shadow
register mode (BYPSHAD=1). This bit can also be cleared by software.
0: Calendar shadow registers not yet synchronized
1: Calendar shadow registers synchronized
Bit 4 INITS: Initialization status flag
This bit is set by hardware when the calendar year field is different from 0 (backup domain
reset value state).
0: Calendar has not been initialized
1: Calendar has been initialized
Bit 3 SHPF: Shift operation pending
0: No shift operation is pending
1: A shift operation is pending
This flag is set by hardware as soon as a shift operation is initiated by a write to the
RTC_SHIFTR. It is cleared by hardware when the corresponding shift operation has been
executed. Writing to SHPF has no effect.
Bit 2 WUTWF: Wakeup timer write flag
This bit is set by hardware up to 2 RTCCLK cycles after the WUTE bit has been set to 0 in
RTC_CR, and is cleared up to 2 RTCCLK cycles after the WUTE bit has been set to 1. The
wakeup timer values can be changed when WUTE bit is cleared and WUTWF is set.
0: Wakeup timer configuration update not allowed
1: Wakeup timer configuration update allowed
Bit 1 ALRBWF: Alarm B write flag
This bit is set by hardware when Alarm B values can be changed, after the ALRBIE bit has
been set to 0 in RTC_CR.
It is cleared by hardware in initialization mode.
0: Alarm B update not allowed
1: Alarm B update allowed.
Bit 0 ALRAWF: Alarm A write flag
This bit is set by hardware when Alarm A values can be changed, after the ALRAE bit has
been set to 0 in RTC_CR.
It is cleared by hardware in initialization mode.
0: Alarm A update not allowed
1: Alarm A update allowed
Note: The ALRAF, ALRBF, WUTF and TSF bits are cleared 2 APB clock cycles after programming
them to 0.
This register is write protected (except for RTC_ISR[13:8] bits). The write access procedure
is described in RTC register write protection on page 803.
822/1751 RM0090 Rev 19

26.6.5 RTC prescaler register (RTC_PRER)

Backup domain reset value: 0x007F 00FF
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PREDIV_A[6:0]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PREDIV_S[14:0]
Res.
Bits 31:24 Reserved

Bits 22:16 PREDIV_A[6:0]: Asynchronous prescaler factor
This is the asynchronous division factor:
ck_apre frequency = RTCCLK frequency/(PREDIV_A+1)
Bits 14:0 PREDIV_S[14:0]: Synchronous prescaler factor
This is the synchronous division factor:
ck_spre frequency = ck_apre frequency/(PREDIV_S+1)
Note: This register must be written in initialization mode only. The initialization must be performed
in two separate write accesses. Refer to Calendar initialization and configuration on
page 804
26.6.6 RTC wakeup timer register (RTC_WUTR)

Backup domain reset value: 0x0000 FFFF
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
WUT[15:0]
RM0090 Rev 19 823/1751

838
Bits 31:16 Reserved

Bits 15:0 WUT[15:0]: Wakeup auto-reload value bits
When the wakeup timer is enabled (WUTE set to 1), the WUTF flag is set every (WUT[15:0]
+ 1) ck_wut cycles. The ck_wut period is selected through WUCKSEL[2:0] bits of the
RTC_CR register
When WUCKSEL[2] = 1, the wakeup timer becomes 17-bits and WUCKSEL[1] effectively
becomes WUT[16] the most-significant bit to be reloaded into the timer.
Note: The first assertion of WUTF occurs (WUT+1) ck_wut cycles after WUTE is set. Setting
WUT[15:0] to 0x0000 with WUCKSEL[2:0] =011 (RTCCLK/2) is forbidden.
Note: This register can be written only when WUTWF is set to 1 in RTC_ISR.
26.6.7 RTC calibration register (RTC_CALIBR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DCS DC[4:0]
Reserved Reserved
rw rw rw rw rw rw
Bits 31:8 Reserved

Bit 7 DCS: Digital calibration sign
0: Positive calibration: calendar update frequency is increased
1: Negative calibration: calendar update frequency is decreased
Bits 4:0 DC[4:0]: Digital calibration
DCS = 0 (positive calibration)
00000: + 0 ppm
00001: + 4 ppm (rounded value)
..
DCS = 1 (negative calibration)
00000: − 0 ppm
00001: − 2 ppm (rounded value)
..
Refer to Case of RTCCLK=32.768 kHz and PREDIV_A+1=128 on page 808 for the exact
step value.
824/1751 RM0090 Rev 19

Note: This register can be written in initialization mode only (RTC_ISR/INITF = ‘1’).
26.6.8 RTC alarm A register (RTC_ALRMAR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
MSK4 WDSEL DT[1:0] DU[3:0] MSK3 PM HT[1:0] HU[3:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MSK2 MNT[2:0] MNU[3:0] MSK1 ST[2:0] SU[3:0]
Bit 31 MSK4: Alarm A date mask

0: Alarm A set if the date/day match
1: Date/day don’t care in Alarm A comparison
Bit 30 WDSEL: Week day selection
0: DU[3:0] represents the date units
1: DU[3:0] represents the week day. DT[1:0] is don’t care.
Bits 29:28 DT[1:0]: Date tens in BCD format.
Bits 27:24 DU[3:0]: Date units or day in BCD format.
Bit 23 MSK3: Alarm A hours mask
0: Alarm A set if the hours match
1: Hours don’t care in Alarm A comparison
1: PM
Bits 21:20 HT[1:0]: Hour tens in BCD format.
Bits 19:16 HU[3:0]: Hour units in BCD format.
Bit 15 MSK2: Alarm A minutes mask
0: Alarm A set if the minutes match
1: Minutes don’t care in Alarm A comparison
Bits 14:12 MNT[2:0]: Minute tens in BCD format.
Bits 11:8 MNU[3:0]: Minute units in BCD format.
Bit 7 MSK1: Alarm A seconds mask
0: Alarm A set if the seconds match
1: Seconds don’t care in Alarm A comparison
Bits 6:4 ST[2:0]: Second tens in BCD format.
Bits 3:0 SU[3:0]: Second units in BCD format.
RM0090 Rev 19 825/1751

838
Note: This register can be written only when ALRAWF is set to 1 in RTC_ISR, or in initialization
mode.
26.6.9 RTC alarm B register (RTC_ALRMBR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
MSK4 WDSEL DT[1:0] DU[3:0] MSK3 PM HT[1:0] HU[3:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MSK2 MNT[2:0] MNU[3:0] MSK1 ST[2:0] SU[3:0]
Bit 31 MSK4: Alarm B date mask

0: Alarm B set if the date and day match
1: Date and day don’t care in Alarm B comparison
Bit 30 WDSEL: Week day selection
0: DU[3:0] represents the date units
1: DU[3:0] represents the week day. DT[1:0] is don’t care.
Bits 27:24 DU[3:0]: Date units or day in BCD format
Bit 23 MSK3: Alarm B hours mask
0: Alarm B set if the hours match
1: Hours don’t care in Alarm B comparison
1: PM
Bits 21:20 HT[1:0]: Hour tens in BCD format
Bits 19:16 HU[3:0]: Hour units in BCD format
Bit 15 MSK2: Alarm B minutes mask
0: Alarm B set if the minutes match
1: Minutes don’t care in Alarm B comparison
Bits 14:12 MNT[2:0]: Minute tens in BCD format
Bits 11:8 MNU[3:0]: Minute units in BCD format
Bit 7 MSK1: Alarm B seconds mask
0: Alarm B set if the seconds match
1: Seconds don’t care in Alarm B comparison
Bits 6:4 ST[2:0]: Second tens in BCD format
Bits 3:0 SU[3:0]: Second units in BCD format
826/1751 RM0090 Rev 19

Note: This register can be written only when ALRBWF is set to 1 in RTC_ISR, or in initialization
mode.
26.6.10 RTC write protection register (RTC_WPR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
KEY
Reserved
w w w w w w w w

Bits 7:0 KEY: Write protection key
This byte is written by software.
Reading this byte always returns 0x00.
Refer to RTC register write protection for a description of how to unlock RTC register write
protection.
26.6.11 RTC sub second register (RTC_SSR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SS[15:0]
Bits 31:16 Reserved

Bits 15:0 SS: Sub second value
SS[15:0] is the value in the synchronous prescaler’s counter. The fraction of a second is
given by the formula below:
Second fraction = ( PREDIV_S - SS ) / ( PREDIV_S + 1 )
Note: SS can be larger than PREDIV_S only after a shift operation. In that case, the correct
time/date is one second less than as indicated by RTC_TR/RTC_DR.
RM0090 Rev 19 827/1751

838
26.6.12 RTC shift control register (RTC_SHIFTR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
ADD1S Reserved
w r r r r r r r r r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. SUBFS[14:0]
r w w w w w w w w w w w w w w w
Bit 31 ADD1S: Add one second

0: No effect
1: Add one second to the clock/calendar
This bit is write only and is always read as zero. Writing to this bit has no effect when a shift
operation is pending (when SHPF=1, in RTC_ISR).
This function is intended to be used with SUBFS (see description below) in order to
effectively add a fraction of a second to the clock in an atomic operation.
Bits 30:15 Reserved
Bits 14:0 SUBFS: Subtract a fraction of a second
These bits are write only and is always read as zero. Writing to this bit has no effect when a
shift operation is pending (when SHPF=1, in RTC_ISR).
The value which is written to SUBFS is added to the synchronous prescaler’s counter. Since
this counter counts down, this operation effectively subtracts from (delays) the clock by:
Delay (seconds) = SUBFS / ( PREDIV_S + 1 )
A fraction of a second can effectively be added to the clock (advancing the clock) when the
ADD1S function is used in conjunction with SUBFS, effectively advancing the clock by:
Advance (seconds) = ( 1 - ( SUBFS / ( PREDIV_S + 1 ) ) ) .
Note: Writing to SUBFS causes RSF to be cleared. Software can then wait until RSF=1 to be
sure that the shadow registers have been updated with the shifted time.
Refer to Section 26.3.8: RTC synchronization.
write protection on page 803
26.6.13 RTC time stamp time register (RTC_TSTR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PM HT[1:0] HU[3:0]
Reserved
r r r r r r r
828/1751 RM0090 Rev 19

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MNT[2:0] MNU[3:0] ST[2:0] SU[3:0]
Res. Res.
r r r r r r r r r r r r r r

1: PM
Bits 21:20 HT[1:0]: Hour tens in BCD format.
Bits 19:16 HU[3:0]: Hour units in BCD format.
Bits 14:12 MNT[2:0]: Minute tens in BCD format.
Bits 11:8 MNU[3:0]: Minute units in BCD format.
Bits 6:4 ST[2:0]: Second tens in BCD format.
Bits 3:0 SU[3:0]: Second units in BCD format.
Note: The content of this register is valid only when TSF is set to 1 in RTC_ISR. It is cleared when
TSF bit is reset.
26.6.14 RTC time stamp date register (RTC_TSDR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
WDU[1:0] MT MU[3:0] DT[1:0] DU[3:0]
Reserved
r r r r r r r r r r r r r r

Bits 15:13 WDU[1:0]: Week day units
Bit 12 MT: Month tens in BCD format
Bits 11:8 MU[3:0]: Month units in BCD format
Bit 3:0 DU[3:0]: Date units in BCD format
Note: The content of this register is valid only when TSF is set to 1 in RTC_ISR. It is cleared when
TSF bit is reset.
RM0090 Rev 19 829/1751

838
26.6.15 RTC timestamp sub second register (RTC_TSSSR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SS[15:0]
Bits 31:16 Reserved

Bits 15:0 SS: Sub second value
SS[15:0] is the value of the synchronous prescaler’s counter when the timestamp event
occurred.
Note: The content of this register is valid only when RTC_ISR/TSF is set. It is cleared when the
RTC_ISR/TSF bit is reset.
26.6.16 RTC calibration register (RTC_CALR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CALP CALW8 CALW16 Reserved CALM[8:0]
rw rw rw r r r r rw rw rw rw rw rw rw rw rw
830/1751 RM0090 Rev 19

Bits 31:16 Reserved

Bit 15 CALP: Increase frequency of RTC by 488.5 ppm
0: No RTCCLK pulses are added.
1: One RTCCLK pulse is effectively inserted every 211 pulses (frequency increased by
488.5 ppm).
This feature is intended to be used in conjunction with CALM, which lowers the frequency of
the calendar with a fine resolution. if the input frequency is 32768 Hz, the number of
RTCCLK pulses added during a 32-second window is calculated as follows: (512 * CALP) -
CALM.
Refer to Section 26.3.11: RTC smooth digital calibration.
Bit 14 CALW8: Use an 8-second calibration cycle period
When CALW8 is set to ‘1’, the 8-second calibration cycle period is selected.
CALM[1:0] are stuck at “00” when CALW8=’1’.
Bit 13 CALW16: Use a 16-second calibration cycle period
When CALW16 is set to ‘1’, the 16-second calibration cycle period is selected. This bit must
not be set to ‘1’ if CALW8=1.
Note: CALM[0] is stuck at ‘0’ when CALW16=’1’.
Bits 12:9 Reserved
Bits 8:0 CALM[8:0]: Calibration minus
The frequency of the calendar is reduced by masking CALM out of 220 RTCCLK pulses (32
seconds if the input frequency is 32768 Hz). This decreases the frequency of the calendar
with a resolution of 0.9537 ppm.
To increase the frequency of the calendar, this feature should be used in conjunction with
CALP.
See Section 26.3.11: RTC smooth digital calibration on page 809.
RM0090 Rev 19 831/1751

838
26.6.17 RTC tamper and alternate function configuration register

(RTC_TAFCR)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
ALARMOUT TSIN TAMP1
Reserved TYPE SEL INSEL
rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TAMP- TAMP- TAMPT TAMP2 TAMP2 TAMP1 TAMP1
TAMPFLT[1:0] TAMPFREQ[2:0] TAMPIE
PUDIS PRCH[1:0] S Reserved -TRG E TRG E
Bits 31:19 Reserved. Always read as 0.

Bit 18 ALARMOUTTYPE: RTC_ALARM output type
0: RTC_ALARM is an open-drain output
1: RTC_ALARM is a push-pull output
Bit 17 TSINSEL: TIMESTAMP mapping
0: RTC_AF1 used as TIMESTAMP
1: RTC_AF2 used as TIMESTAMP
Bit 16 TAMP1INSEL: TAMPER1 mapping
0: RTC_AF1 used as TAMPER1
1: RTC_AF2 used as TAMPER1
Note: TAMP1E must be reset when TAMP1INSEL is changed to avoid unwanted setting of
TAMP1F.
Bit 15 TAMPPUDIS: TAMPER pull-up disable
This bit determines if each of the tamper pins are pre-charged before each sample.
0: Precharge tamper pins before sampling (enable internal pull-up)
1: Disable precharge of tamper pins
Note:
Bits 14:13 TAMPPRCH[1:0]: Tamper precharge duration
These bit determines the duration of time during which the pull-up/is activated before each
sample. TAMPPRCH is valid for each of the tamper inputs.
0x0: 1 RTCCLK cycle
0x1: 2 RTCCLK cycles
832/1751 RM0090 Rev 19

Bits 12:11 TAMPFLT[1:0]: Tamper filter count

These bits determines the number of consecutive samples at the specified level
(TAMP*TRG) necessary to activate a Tamper event. TAMPFLT is valid for each of the tamper
inputs.
0x0: Tamper is activated on edge of tamper input transitions to the active level (no internal
pull-up on tamper input).
0x1: Tamper is activated after 2 consecutive samples at the active level.
Bits 10:8 TAMPFREQ[2:0]: Tamper sampling frequency
Determines the frequency at which each of the tamper inputs are sampled.
0x0: RTCCLK / 32768 (1 Hz when RTCCLK = 32768 Hz)
Bit 7 TAMPTS: Activate timestamp on tamper detection event
0: Tamper detection event does not cause a timestamp to be saved
1: Save timestamp on tamper detection event
TAMPTS is valid even if TSE=0 in the RTC_CR register.
Bits 6:5 Reserved. Always read as 0.
Bit 4 TAMP2TRG: Active level for tamper 2
if TAMPFLT != 00
0: TAMPER2 staying low triggers a tamper detection event.
1: TAMPER2 staying high triggers a tamper detection event.
if TAMPFLT = 00:
0: TAMPER2 rising edge triggers a tamper detection event.
1: TAMPER2 falling edge triggers a tamper detection event.
Bit 3 TAMP2E: Tamper 2 detection enable
0: Tamper 2 detection disabled
1: Tamper 2 detection enabled
Bit 2 TAMPIE: Tamper interrupt enable
0: Tamper interrupt disabled
1: Tamper interrupt enabled
RM0090 Rev 19 833/1751

838
Bit 1 TAMP1TRG: Active level for tamper 1

if TAMPFLT != 00
0: TAMPER1 staying low triggers a tamper detection event.
1: TAMPER1 staying high triggers a tamper detection event.
if TAMPFLT = 00:
0: TAMPER1 rising edge triggers a tamper detection event.
1: TAMPER1 falling edge triggers a tamper detection event.
Caution: When TAMPFLT = 0, TAMPxE must be reset when TAMPxTRG is changed to avoid
spuriously setting TAMPxF.
Bit 0 TAMP1E: Tamper 1 detection enable
0: Tamper 1 detection disabled
1: Tamper 1 detection enabled
26.6.18 RTC alarm A sub second register (RTC_ALRMASSR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved MASKSS[3:0] Reserved
r r r r rw rw rw rw r r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserved SS[14:0]
r rw rw rw rw rw rw rw rw rw rw rw rw w rw rw
Bits 31:28 Reserved

Bits 27:24 MASKSS[3:0]: Mask the most-significant bits starting at this bit
0: No comparison on sub seconds for Alarm A. The alarm is set when the seconds unit is
incremented (assuming that the rest of the fields match).
1: SS[14:1] are don’t care in Alarm A comparison. Only SS[0] is compared.
2: SS[14:2] are don’t care in Alarm A comparison. Only SS[1:0] are compared.
3: SS[14:3] are don’t care in Alarm A comparison. Only SS[2:0] are compared.
...
12: SS[14:12] are don’t care in Alarm A comparison. SS[11:0] are compared.
13: SS[14:13] are don’t care in Alarm A comparison. SS[12:0] are compared.
14: SS[14] is don’t care in Alarm A comparison. SS[13:0] are compared.
15: All 15 SS bits are compared and must match to activate alarm.
The overflow bits of the synchronous counter (bits 15) is never compared. This bit can be
different from 0 only after a shift operation.
Bits 23:15 Reserved
Bits 14:0 SS[14:0]: Sub seconds value
This value is compared with the contents of the synchronous prescaler’s counter to
determine if Alarm A is to be activated. Only bits 0 up MASKSS-1 are compared.
834/1751 RM0090 Rev 19

Note: This register can be written only when ALRAE is reset in RTC_CR register, or in initialization
mode.
26.6.19 RTC alarm B sub second register (RTC_ALRMBSSR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved MASKSS[3:0] Reserved
r r r r rw rw rw rw r r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserved SS[14:0]
r rw rw rw rw rw rw rw rw rw rw rw rw w rw rw
Bits 31:28 Reserved

Bits 27:24 MASKSS[3:0]: Mask the most-significant bits starting at this bit
0x0: No comparison on sub seconds for Alarm B. The alarm is set when the seconds unit is
incremented (assuming that the rest of the fields match).
0x1: SS[14:1] are don’t care in Alarm B comparison. Only SS[0] is compared.
0x2: SS[14:2] are don’t care in Alarm B comparison. Only SS[1:0] are compared.
0x3: SS[14:3] are don’t care in Alarm B comparison. Only SS[2:0] are compared.
...
0xC: SS[14:12] are don’t care in Alarm B comparison. SS[11:0] are compared.
0xD: SS[14:13] are don’t care in Alarm B comparison. SS[12:0] are compared.
0xE: SS[14] is don’t care in Alarm B comparison. SS[13:0] are compared.
0xF: All 15 SS bits are compared and must match to activate alarm.
The overflow bits of the synchronous counter (bits 15) is never compared. This bit can be
different from 0 only after a shift operation.
Bits 23:15 Reserved
Bits 14:0 SS[14:0]: Sub seconds value
This value is compared with the contents of the synchronous prescaler’s counter to
determine if Alarm B is to be activated. Only bits 0 up to MASKSS-1 are compared.
Note: This register can be written only when ALRBIE is reset in RTC_CR register, or in
This register is write protected.The write access procedure is described in Section : RTC
register write protection
RM0090 Rev 19 835/1751

838
26.6.20 RTC backup registers (RTC_BKPxR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
BKP[31:16]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
BKP[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw w rw rw
Bits 31:0 BKP[31:0]

The application can write or read data to and from these registers.
They are powered-on by VBAT when VDD is switched off, so that they are not reset by
System reset, and their contents remain valid when the device operates in low-power mode.
This register is reset on a tamper detection event, as long as TAMPxF=1.
26.6.21 RTC register map
Table 121. RTC register map and reset values

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
HT
Reserved
Reserved
RTC_TR HU[3:0] MNT[2:0] MNU[3:0] ST[2:0] SU[3:0]
PM
0x00 Reserved [1:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DT
Reserved
RTC_DR YT[3:0] YU[3:0] WDU[2:0] MU[3:0] DU[3:0]

MT
0x04 Reserved [1:0]
Reset value 0 0 1 0 0 0 0 1 0 0 0 0 0 1
REFCKON
BYPSHAD
TSEDGE
ALRBIE
ALRAIE
COSEL
ADD1H
SUB1H
ALRBE
ALRAE
WUTIE
WUTE
OSEL WCKSEL
TSIE
COE
DCE
FMT
POL
BKP
TSE
RTC_CR
0x08 Reserved [1:0] [2:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ALRBWF
ALRAWF
TAMP2F
TAMP1F
WUTWF
TSOVF
ALRBF
ALRAF
WUTF
SHPF
INITS
INITF
RSF
TSF
INIT
RTC_ISR
0x0C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1
Reserved
RTC_PRER PREDIV_A[6:0] PREDIV_S[14:0]

0x10 Reserved
Reset value 1 1 1 1 1 1 1 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1
RTC_WUTR WUT[15:0]
0x14 Reserved
Reset value 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
Reserved
DCS
RTC_CALIBR DC[4:0]
0x18 Reserved
836/1751 RM0090 Rev 19

Table 121. RTC register map and reset values (continued)

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
WDSEL
DT HT
MSK4
MSK3
MSK2
MSK1
RTC_ALRMAR DU[3:0] HU[3:0] MNT[2:0] MNU[3:0] ST[2:0] SU[3:0]
PM
0x1C [1:0] [1:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
WDSEL
DT HT
MSK4
MSK3
MSK2
MSK2
RTC_ALRMBR DU[3:0] HU[3:0] MNT[2:0] MNU[3:0] ST[2:0] SU[3:0]
PM
0x20 [1:0] [1:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RTC_WPR KEY[7:0]
0x24 Reserved
Reset value 0 0 0 0 0 0 0 0
RTC_SSR SS[15:0]
0x28 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ADD1S
RTC_SHIFTR SUBFS[14:0]
0x2C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
MNT[2:0]
HT[1:0]
Reserved
Reserved
RTC_TSTR HU[3:0] MNU[3:0] ST[2:0] SU[3:0]
PM
0x30 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RTC_TSSSR SS[15:0]
0x38 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CALW16
CALW8
CALP
Reserved
RTC_ CALR CALM[8:0]

0x3C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0
ALARMOUTTYPE
TAMPPRCH[1:0]
TAMPFREQ[2:0]
TAMPFLT[1:0]
TAMP1INSEL
TAMP1ETRG
TAMPPUDIS
TAMP2TRG
TSINSEL
TAMPTS
TAMP2E
TAMP1E
TAMPIE
Reserved
RTC_TAFCR
0x40 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RTC_
MASKSS[3:0] SS[14:0]
0x44 ALRMASSR Reserved Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RTC_
MASKSS[3:0] SS[14:0]
0x48 ALRMBSSR Reserved Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RTC_BKP0R BKP[31:0]
0x50 Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

to 0x9C
to
BKP[31:0]
RTC_BKP19R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RM0090 Rev 19 837/1751

838
Caution: In Table 121, the reset value is the value after a backup domain reset. The majority of the
registers are not affected by a system reset. For more information, please refer to
Section 26.3.7: Resetting the RTC.
838/1751 RM0090 Rev 19

RM0090 Inter-integrated circuit (I2C) interface
27 Inter-integrated circuit (I2C) interface
27.1 I2C introduction

I2C (inter-integrated circuit) bus Interface serves as an interface between the microcontroller
and the serial I2C bus. It provides multimaster capability, and controls all I2C bus-specific
sequencing, protocol, arbitration and timing. It supports the standard mode (Sm, up to
100 kHz) and Fm mode (Fm, up to 400 kHz).
It may be used for a variety of purposes, including CRC generation and verification, SMBus
(system management bus) and PMBus (power management bus).
Depending on specific device implementation DMA capability can be available for reduced
CPU overload.
27.2 I2C main features

• Parallel-bus/I2C protocol converter
• Multimaster capability: the same interface can act as Master or Slave
• I2C Master features:
– Clock generation
– Start and Stop generation
• I2C Slave features:
– Programmable I2C Address detection
– Dual Addressing Capability to acknowledge 2 slave addresses
– Stop bit detection
• Generation and detection of 7-bit/10-bit addressing and General Call
• Supports different communication speeds:
– Standard Speed (up to 100 kHz)
– Fast Speed (up to 400 kHz)
• Analog noise filter
• Programmable digital noise filter for STM32F42xxx and STM32F43xxx
• Status flags:
– Transmitter/Receiver mode flag
– End-of-Byte transmission flag
– I2C busy flag
• Error flags:
– Arbitration lost condition for master mode
– Acknowledgment failure after address/ data transmission
– Detection of misplaced start or stop condition
– Overrun/Underrun if clock stretching is disabled
• 2 Interrupt vectors:
RM0090 Rev 19 839/1751

872
Inter-integrated circuit (I2C) interface RM0090
– 1 Interrupt for successful address/ data communication

– 1 Interrupt for error condition
• Optional clock stretching
• 1-byte buffer with DMA capability
• Configurable PEC (packet error checking) generation or verification:
– PEC value can be transmitted as last byte in Tx mode
– PEC error checking for last received byte
• SMBus 2.0 Compatibility:
– 25 ms clock low timeout delay
– 10 ms master cumulative clock low extend time
– 25 ms slave cumulative clock low extend time
– Hardware PEC generation/verification with ACK control
– Address Resolution Protocol (ARP) supported
• PMBus Compatibility
Note: Some of the above features may not be available in certain products. The user should refer
to the product data sheet, to identify the specific features supported by the I2C interface
implementation.
27.3 I2C functional description

In addition to receiving and transmitting data, this interface converts it from serial to parallel
format and vice versa. The interrupts are enabled or disabled by software. The interface is
connected to the I2C bus by a data pin (SDA) and by a clock pin (SCL). It can be connected
with a standard (up to 100 kHz) or fast (up to 400 kHz) I2C bus.
27.3.1 Mode selection

The interface can operate in one of the four following modes:
• Slave transmitter
• Slave receiver
• Master transmitter
• Master receiver
By default, it operates in slave mode. The interface automatically switches from slave to
master, after it generates a START condition and from master to slave, if an arbitration loss
or a Stop generation occurs, allowing multimaster capability.
Communication flow
In Master mode, the I2C interface initiates a data transfer and generates the clock signal. A
serial data transfer always begins with a start condition and ends with a stop condition. Both
start and stop conditions are generated in master mode by software.
In Slave mode, the interface is capable of recognizing its own addresses (7 or 10-bit), and
the General Call address. The General Call address detection may be enabled or disabled
by software.
840/1751 RM0090 Rev 19

Data and addresses are transferred as 8-bit bytes, MSB first. The first byte(s) following the
start condition contain the address (one in 7-bit mode, two in 10-bit mode). The address is
always transmitted in Master mode.
A 9th clock pulse follows the 8 clock cycles of a byte transfer, during which the receiver must
send an acknowledge bit to the transmitter. Refer to Figure 238.
Figure 238. I2C bus protocol
SDA
MSB ACK
SCL
1 2 8 9
Start Stop
condition condition
MS19854V1
Acknowledge may be enabled or disabled by software. The I2C interface addresses (dual
addressing 7-bit/ 10-bit and/or general call address) can be selected by software.
The block diagram of the I2C interface is shown in Figure 239.
RM0090 Rev 19 841/1751

872
Figure 239. I2C block diagram for STM32F40x/41x
Data register
Noise Data
SDA Data shift register
filter control
Comparator PEC calculation
Own address register

Dual address register
Clock
Noise PEC register
SCL control
filter
Clock control
Register (CCR)
Control registers
(CR1&CR2)
Control
Status registers logic
(SR1&SR2)
SMBA
Interrupts DMA requests & ACK

MS30035V1
842/1751 RM0090 Rev 19

Figure 240. I2C block diagram for STM32F42x/43x
Data register
Noise Data
SDA Data shift register
filter control
Comparator PEC calculation
Own address register

Dual address register
Clock
Noise PEC register
SCL control
filter
Clock control
Register (CCR)
Control registers
(CR1&CR2)
Control
Status registers logic
(SR1&SR2)
SMBA
Interrupts DMA requests & ACK

MS30035V1
1. SMBA is an optional signal in SMBus mode. This signal is not applicable if SMBus is disabled.
27.3.2 I2C slave mode

By default the I2C interface operates in Slave mode. To switch from default Slave mode to
Master mode a Start condition generation is needed.
The peripheral input clock must be programmed in the I2C_CR2 register in order to
generate correct timings. The peripheral input clock frequency must be at least:
• 2 MHz in Sm mode
• 4 MHz in Fm mode
As soon as a start condition is detected, the address is received from the SDA line and sent
to the shift register. Then it is compared with the address of the interface (OAR1) and with
OAR2 (if ENDUAL=1) or the General Call address (if ENGC = 1).
Note: In 10-bit addressing mode, the comparison includes the header sequence (11110xx0),
where xx denotes the two most significant bits of the address.
RM0090 Rev 19 843/1751

872
Header or address not matched: the interface ignores it and waits for another Start
condition.
Header matched (10-bit mode only): the interface generates an acknowledge pulse if the
ACK bit is set and waits for the 8-bit slave address.
Address matched: the interface generates in sequence:
• An acknowledge pulse if the ACK bit is set
• The ADDR bit is set by hardware and an interrupt is generated if the ITEVFEN bit is
set.
• If ENDUAL=1, the software has to read the DUALF bit to check which slave address
has been acknowledged.
In 10-bit mode, after receiving the address sequence the slave is always in Receiver mode.
It will enter Transmitter mode on receiving a repeated Start condition followed by the header
sequence with matching address bits and the least significant bit set (11110xx1).
The TRA bit indicates whether the slave is in Receiver or Transmitter mode.
Slave transmitter
Following the address reception and after clearing ADDR, the slave sends bytes from the
DR register to the SDA line via the internal shift register.
The slave stretches SCL low until ADDR is cleared and DR filled with the data to be sent
(see Figure 241 Transfer sequencing EV1 EV3).
When the acknowledge pulse is received:
• The TxE bit is set by hardware with an interrupt if the ITEVFEN and the ITBUFEN bits
are set.
If TxE is set and some data were not written in the I2C_DR register before the end of the
next data transmission, the BTF bit is set and the interface waits until BTF is cleared by a
read to I2C_SR1 followed by a write to the I2C_DR register, stretching SCL low.
844/1751 RM0090 Rev 19

Figure 241. Transfer sequence diagram for slave transmitter

7-bit slave transmitter
S Address A Data1 A Data2 A DataN NA P
.....
EV1 EV3-1 EV3 EV3 EV3 EV3-2
10-bit slave transmitter

S Header A Address A
EV1
Sr Header A Data1 A .... DataN NA P

EV1 EV3_1 EV3 EV3 EV3-2
Legend: S= Start, Sr = Repeated Start, P= Stop, A= Acknowledge, NA= Non-acknowledge,

EVx= Event (with interrupt if ITEVFEN=1)
EV1: ADDR=1, cleared by reading SR1 followed by reading SR2

EV3-1: TxE=1, shift register empty, data register empty, write Data1 in DR.
EV3: TxE=1, shift register not empty, data register empty, cleared by writing DR
EV3-2: AF=1; AF is cleared by writing ‘0’ in AF bit of SR1 register.
ai18209
1. The EV1 and EV3_1 events stretch SCL low until the end of the corresponding software sequence.
2. The EV3 event stretches SCL low if the software sequence is not completed before the end of the next byte
transmission.
Slave receiver
Following the address reception and after clearing ADDR, the slave receives bytes from the
SDA line into the DR register via the internal shift register. After each byte the interface
generates in sequence:
• An acknowledge pulse if the ACK bit is set
• The RxNE bit is set by hardware and an interrupt is generated if the ITEVFEN and
ITBUFEN bit is set.
If RxNE is set and the data in the DR register is not read before the end of the next data
reception, the BTF bit is set and the interface waits until BTF is cleared by a read from the
I2C_DR register, stretching SCL low (see Figure 242 Transfer sequencing).
RM0090 Rev 19 845/1751

872
Figure 242. Transfer sequence diagram for slave receiver

7-bit slave receiver
S Address A Data1 A Data2 A DataN A P
.....
EV1 EV2 EV2 EV2 EV4
10-bit slav e receiver

S Header A Address A Data1 A DataN A P
.....
EV1 EV2 EV2 EV4
Legend: S= Start, Sr = Repeated Start, P= Stop, A= Acknowledge,

EV1: ADDR=1, cleared by reading SR1 followed by reading SR2

EV2: RxNE=1 cleared by reading DR register.
EV4: STOPF=1, cleared by reading SR1 register followed by writing to the CR1 register
ai18208
1. The EV1 event stretches SCL low until the end of the corresponding software sequence.
reception.
3. After checking the SR1 register content, the user should perform the complete clearing sequence for each
flag found set.
Thus, for ADDR and STOPF flags, the following sequence is required inside the I2C interrupt routine:
READ SR1
if (ADDR == 1) {READ SR1; READ SR2}
if (STOPF == 1) {READ SR1; WRITE CR1}
The purpose is to make sure that both ADDR and STOPF flags are cleared if both are found set.
Closing slave communication

After the last data byte is transferred a Stop Condition is generated by the master. The
interface detects this condition and sets:
• The STOPF bit and generates an interrupt if the ITEVFEN bit is set.
The STOPF bit is cleared by a read of the SR1 register followed by a write to the CR1
register (see EV4 in Figure 242).
27.3.3 I2C master mode

In Master mode, the I2C interface initiates a data transfer and generates the clock signal. A
serial data transfer always begins with a Start condition and ends with a Stop condition.
Master mode is selected as soon as the Start condition is generated on the bus with a
START bit.
The following is the required sequence in master mode.
• Program the peripheral input clock in I2C_CR2 register in order to generate correct
timings
• Configure the clock control registers
• Configure the rise time register
• Program the I2C_CR1 register to enable the peripheral
• Set the START bit in the I2C_CR1 register to generate a Start condition
The peripheral input clock frequency must be at least:
• 2 MHz in Sm mode
• 4 MHz in Fm mode
846/1751 RM0090 Rev 19

SCL master clock generation

The CCR bits are used to generate the high and low level of the SCL clock, starting from the
generation of the rising and falling edge (respectively). As a slave may stretch the SCL line,
the peripheral checks the SCL input from the bus at the end of the time programmed in
TRISE bits after rising edge generation.
• If the SCL line is low, it means that a slave is stretching the bus, and the high level
counter stops until the SCL line is detected high. This allows to guarantee the minimum
HIGH period of the SCL clock parameter.
• If the SCL line is high, the high level counter keeps on counting.
Indeed, the feedback loop from the SCL rising edge generation by the peripheral to the SCL
rising edge detection by the peripheral takes time even if no slave stretches the clock. This
loopback duration is linked to the SCL rising time (impacting SCL VIH input detection), plus
delay due to the noise filter present on the SCL input path, plus delay due to internal SCL
input synchronization with APB clock. The maximum time used by the feedback loop is
programmed in the TRISE bits, so that the SCL frequency remains stable whatever the SCL
rising time.
Start condition
Setting the START bit causes the interface to generate a Start condition and to switch to
Master mode (MSL bit set) when the BUSY bit is cleared.
Note: In master mode, setting the START bit causes the interface to generate a ReStart condition
at the end of the current byte transfer.
Once the Start condition is sent:
• The SB bit is set by hardware and an interrupt is generated if the ITEVFEN bit is set.
Then the master waits for a read of the SR1 register followed by a write in the DR register
with the Slave address (see Figure 243 and Figure 244 Transfer sequencing EV5).
Slave address transmission

Then the slave address is sent to the SDA line via the internal shift register.
• In 10-bit addressing mode, sending the header sequence causes the following event:
– The ADD10 bit is set by hardware and an interrupt is generated if the ITEVFEN bit
is set.
Then the master waits for a read of the SR1 register followed by a write in the DR
register with the second address byte (see Figure 243 and Figure 244 Transfer
sequencing).
– The ADDR bit is set by hardware and an interrupt is generated if the ITEVFEN bit
is set.
Then the master waits for a read of the SR1 register followed by a read of the SR2
register (see Figure 243 and Figure 244 Transfer sequencing).
• In 7-bit addressing mode, one address byte is sent.
As soon as the address byte is sent,
– The ADDR bit is set by hardware and an interrupt is generated if the ITEVFEN bit
is set.
Then the master waits for a read of the SR1 register followed by a read of the SR2
register (see Figure 243 and Figure 244 Transfer sequencing).
RM0090 Rev 19 847/1751

872
The master can decide to enter Transmitter or Receiver mode depending on the LSB of the
slave address sent.
• In 7-bit addressing mode,
– To enter Transmitter mode, a master sends the slave address with LSB reset.
– To enter Receiver mode, a master sends the slave address with LSB set.
• In 10-bit addressing mode,
– To enter Transmitter mode, a master sends the header (11110xx0) and then the
slave address, (where xx denotes the two most significant bits of the address).
– To enter Receiver mode, a master sends the header (11110xx0) and then the
slave address. Then it should send a repeated Start condition followed by the
header (11110xx1), (where xx denotes the two most significant bits of the
address).
The TRA bit indicates whether the master is in Receiver or Transmitter mode.
Master transmitter
Following the address transmission and after clearing ADDR, the master sends bytes from
the DR register to the SDA line via the internal shift register.
The master waits until the first data byte is written into I2C_DR (see Figure 243 Transfer
sequencing EV8_1).
When the acknowledge pulse is received, the TxE bit is set by hardware and an interrupt is
generated if the ITEVFEN and ITBUFEN bits are set.
If TxE is set and a data byte was not written in the DR register before the end of the last data
transmission, BTF is set and the interface waits until BTF is cleared by a write to I2C_DR,
stretching SCL low.
Closing the communication
After the last byte is written to the DR register, the STOP bit is set by software to generate a
Stop condition (see Figure 243 Transfer sequencing EV8_2). The interface automatically
goes back to slave mode (MSL bit cleared).
Note: Stop condition should be programmed during EV8_2 event, when either TxE or BTF is set.
848/1751 RM0090 Rev 19

Figure 243. Transfer sequence diagram for master transmitter

7-bit master transmitter
S Address A Data1 A Data2 A DataN A P
.....
EV5 EV6 EV8_1 EV8 EV8 EV8 EV8_2
10-bit master transmitter

S Header A Address A Data1 A DataN A P
.....
EV5 EV9 EV6 EV8_1 EV8 EV8 EV8_2
Legend: S= Start, Sr = Repeated Start, P= Stop, A= Acknowledge,

EVx= Event (with interrupt if ITEVFEN = 1)
EV5: SB=1, cleared by reading SR1 register followed by writing DR register with Address.
EV6: ADDR=1, cleared by reading SR1 register followed by reading SR2.
EV8_1: TxE=1, shift register empty, data register empty, write Data1 in DR.
EV8: TxE=1, shift register not empty,.data register empty, cleared by writing DR register
EV8_2: TxE=1, BTF = 1, Program Stop request. TxE and BTF are cleared by hardware by the Stop condition
EV9: ADD10=1, cleared by reading SR1 register followed by writing DR register.
ai18210
1. The EV5, EV6, EV9, EV8_1 and EV8_2 events stretch SCL low until the end of the corresponding software sequence.
2. The EV8 event stretches SCL low if the software sequence is not complete before the end of the next byte transmission.
Master receiver
Following the address transmission and after clearing ADDR, the I2C interface enters
Master Receiver mode. In this mode the interface receives bytes from the SDA line into the
DR register via the internal shift register. After each byte the interface generates in
sequence:
1. An acknowledge pulse if the ACK bit is set
2. The RxNE bit is set and an interrupt is generated if the ITEVFEN and ITBUFEN bits are
set (see Figure 244 Transfer sequencing EV7).
If the RxNE bit is set and the data in the DR register is not read before the end of the last
data reception, the BTF bit is set by hardware and the interface waits until BTF is cleared by
a read in the DR register, stretching SCL low.
Closing the communication
The master sends a NACK for the last byte received from the slave. After receiving this
NACK, the slave releases the control of the SCL and SDA lines. Then the master can send
a Stop/Restart condition.
1. To generate the nonacknowledge pulse after the last received data byte, the ACK bit
must be cleared just after reading the second last data byte (after second last RxNE
event).
2. In order to generate the Stop/Restart condition, software must set the STOP/START bit
after reading the second last data byte (after the second last RxNE event).
3. In case a single byte has to be received, the Acknowledge disable is made during EV6
(before ADDR flag is cleared) and the STOP condition generation is made after EV6.
After the Stop condition generation, the interface goes automatically back to slave mode
(MSL bit cleared).
RM0090 Rev 19 849/1751

872
Figure 244. Transfer sequence diagram for master receiver

7-bit master receiver
S Address A Data1 A(1) Data2 A DataN NA P
.....
EV5 EV6 EV7 EV7 EV7_1 EV7
10-bit master receiver

S Header A Address A
EV5 EV9 EV6
Sr Header A Data1 A(1) Data2 A DataN NA P

.....
EV5 EV6 EV7 EV7 EV7_1 EV7
Leg end : S= Start, Sr = Repeated Start, P= Stop, A= Acknowledge, NA= Non-acknowledge,

EV5: SB=1, cleared by reading SR1 register followed by writing DR register.
EV6: ADDR=1, cleared by reading SR1 register followed by reading SR2. In 10-bit master receiver mode, this se-
quence should be followed by writing CR2 with START = 1.
In case of the reception of 1 byte, the Acknowledge disable must be performed during EV6 event, i.e. before clearing ADDR flag.
EV7: RxNE=1 cleared by reading DR register.
EV7_1: RxNE=1 cleared by reading DR register, program AC K=0 and STOP request
EV9: ADD10=1, cleared by reading SR1 register followed by writing DR register.
ai17540d
1. If a single byte is received, it is NA.

2. The EV5, EV6 and EV9 events stretch SCL low until the end of the corresponding software sequence.
reception.
4. The EV7_1 software sequence must be completed before the ACK pulse of the current byte transfer.
The procedures described below are recommended if the EV7-1 software sequence is not
completed before the ACK pulse of the current byte transfer.
These procedures must be followed to make sure:
• The ACK bit is set low on time before the end of the last data reception
• The STOP bit is set high after the last data reception without reception of
supplementary data.
For 2-byte reception:
• Wait until ADDR = 1 (SCL stretched low until the ADDR flag is cleared)
• Set ACK low, set POS high
• Clear ADDR flag
• Wait until BTF = 1 (Data 1 in DR, Data2 in shift register, SCL stretched low until a data
1 is read)
• Set STOP high
• Read data 1 and 2
850/1751 RM0090 Rev 19

For N >2 -byte reception, from N-2 data reception

• Wait until BTF = 1 (data N-2 in DR, data N-1 in shift register, SCL stretched low until
data N-2 is read)
• Set ACK low
• Read data N-2
• Wait until BTF = 1 (data N-1 in DR, data N in shift register, SCL stretched low until a
data N-1 is read)
• Set STOP high
• Read data N-1 and N
27.3.4 Error conditions

The following are the error conditions which may cause communication to fail.
Bus error (BERR)

This error occurs when the I2C interface detects an external Stop or Start condition during
an address or a data transfer. In this case:
• the BERR bit is set and an interrupt is generated if the ITERREN bit is set
• in Slave mode: data are discarded and the lines are released by hardware:
– in case of a misplaced Start, the slave considers it is a restart and waits for an
address, or a Stop condition
– in case of a misplaced Stop, the slave behaves like for a Stop condition and the
lines are released by hardware
• In Master mode: the lines are not released and the state of the current transmission is
not affected. It is up to the software to abort or not the current transmission
Acknowledge failure (AF)

This error occurs when the interface detects a nonacknowledge bit. In this case:
• the AF bit is set and an interrupt is generated if the ITERREN bit is set
• a transmitter which receives a NACK must reset the communication:
– If Slave: lines are released by hardware
– If Master: a Stop or repeated Start condition must be generated by software
Arbitration lost (ARLO)

This error occurs when the I2C interface detects an arbitration lost condition. In this case
• the ARLO bit is set by hardware (and an interrupt is generated if the ITERREN bit is
set)
• the I2C Interface goes automatically back to slave mode (the MSL bit is cleared). When
the I2C loses the arbitration, it is not able to acknowledge its slave address in the same
transfer, but it can acknowledge it after a repeated Start from the winning master.
• lines are released by hardware
RM0090 Rev 19 851/1751

872
Overrun/underrun error (OVR)

An overrun error can occur in slave mode when clock stretching is disabled and the I2C
interface is receiving data. The interface has received a byte (RxNE=1) and the data in DR
has not been read, before the next byte is received by the interface. In this case,
• The last received byte is lost.
• In case of Overrun error, software should clear the RxNE bit and the transmitter should
re-transmit the last received byte.
Underrun error can occur in slave mode when clock stretching is disabled and the I2C
interface is transmitting data. The interface has not updated the DR with the next byte
(TxE=1), before the clock comes for the next byte. In this case,
• The same byte in the DR register will be sent again
• The user should make sure that data received on the receiver side during an underrun
error are discarded and that the next bytes are written within the clock low time
specified in the I2C bus standard.
For the first byte to be transmitted, the DR must be written after ADDR is cleared and before
the first SCL rising edge. If not possible, the receiver must discard the first data.
27.3.5 Programmable noise filter

The programmable noise filter is available on STM32F42xxx and STM32F43xxx devices
only.
In Fm mode, the I2C standard requires that spikes are suppressed to a length of 50 ns on
SDA and SCL lines.
An analog noise filter is implemented in the SDA and SCL I/Os. This filter is enabled by
default and can be disabled by setting the ANOFF bit in the I2C_FLTR register.
A digital noise filter can be enabled by configuring the DNF[3:0] bits to a non-zero value.
This suppresses the spikes on SDA and SCL inputs with a length of up to DNF[3:0] *
TPCLK1.
Enabling the digital noise filter increases the SDA hold time by (DNF[3:0] +1)* TPCLK.
To be compliant with the maximum hold time of the I2C-bus specification version 2.1
(Thd:dat), the DNF bits must be programmed using the constraints shown in Table 122, and
assuming that the analog filter is disabled.
Note: DNF[3:0] must only be configured when the I2C is disabled (PE = 0). If the analog filter is
also enabled, the digital filter is added to the analog filter.
Table 122. Maximum DNF[3:0] value to be compliant with Thd:dat(max)

Maximum DNF value
PCLK1 frequency
Sm mode Fm mode
2 <= FPCLK1 <= 5 2 0

5 < FPCLK1 <= 10 12 0
10 < FPCLK1 <= 20 15 1
20 < FPCLK1 <= 30 15 7
852/1751 RM0090 Rev 19

Table 122. Maximum DNF[3:0] value to be compliant with Thd:dat(max) (continued)

Maximum DNF value
PCLK1 frequency
Sm mode Fm mode
30 < FPCLK1 <= 40 15 13

40 < FPCLK1 <= 50 15 15
Note: For each frequency range, the constraint is given based on the worst case which is the
minimum frequency of the range. Greater DNF values can be used if the system can
support maximum hold time violation.
27.3.6 SDA/SCL line control

• If clock stretching is enabled:
– Transmitter mode: If TxE=1 and BTF=1: the interface holds the clock line low
before transmission to wait for the microcontroller to write the byte in the Data
register (both buffer and shift register are empty).
– Receiver mode: If RxNE=1 and BTF=1: the interface holds the clock line low after
reception to wait for the microcontroller to read the byte in the Data register (both
buffer and shift register are full).
• If clock stretching is disabled in Slave mode:
– Overrun Error in case of RxNE=1 and no read of DR has been done before the
next byte is received. The last received byte is lost.
– Underrun Error in case TxE=1 and no write into DR has been done before the next
byte must be transmitted. The same byte will be sent again.
– Write Collision not managed.
27.3.7 SMBus
Introduction
The System Management Bus (SMBus) is a two-wire interface through which various
devices can communicate with each other and with the rest of the system. It is based on I2C
principles of operation. SMBus provides a control bus for system and power management
related tasks. A system may use SMBus to pass messages to and from devices instead of
toggling individual control lines.
The System Management Bus Specification refers to three types of devices. A slave is a
device that is receiving or responding to a command. A master is a device that issues
commands, generates the clocks, and terminates the transfer. A host is a specialized
master that provides the main interface to the system's CPU. A host must be a master-slave
and must support the SMBus host notify protocol. Only one host is allowed in a system.
Similarities between SMBus and I2C

• 2-wire bus protocol (1 Clk, 1 Data) + SMBus Alert line optional
• Master-slave communication, Master provides clock
• Multi master capability
• SMBus data format similar to I2C 7-bit addressing format (Figure 238).
RM0090 Rev 19 853/1751

872
Differences between SMBus and I2C

The following table describes the differences between SMBus and I2C.
Table 123. SMBus vs. I2C

SMBus I2C
Max. speed 100 kHz Max. speed 400 kHz

Min. clock speed 10 kHz No minimum clock speed
35 ms clock low timeout No timeout
Logic levels are fixed Logic levels are VDD dependent
Different address types (reserved, dynamic etc.) 7-bit, 10-bit and general call slave address types
Different bus protocols (quick command, process
No bus protocols
call etc.)
SMBus application usage

With System Management Bus, a device can provide manufacturer information, tell the
system what its model/part number is, save its state for a suspend event, report different
types of errors, accept control parameters, and return its status. SMBus provides a control
bus for system and power management related tasks.
Device identification
Any device that exists on the System Management Bus as a slave has a unique address
called the Slave Address. For the list of reserved slave addresses, refer to the SMBus
specification version. 2.0 (http://smbus.org/).
Bus protocols
The SMBus specification supports up to nine bus protocols. For more details of these
protocols and SMBus address types, refer to SMBus specification version. 2.0. These
protocols should be implemented by the user software.
Address resolution protocol (ARP)

SMBus slave address conflicts can be resolved by dynamically assigning a new unique
address to each slave device. The Address Resolution Protocol (ARP) has the following
attributes:
• Address assignment uses the standard SMBus physical layer arbitration mechanism
• Assigned addresses remain constant while device power is applied; address retention
through device power loss is also allowed
• No additional SMBus packet overhead is incurred after address assignment. (i.e.
subsequent accesses to assigned slave addresses have the same overhead as
accesses to fixed address devices.)
• Any SMBus master can enumerate the bus
Unique device identifier (UDID)

In order to provide a mechanism to isolate each device for the purpose of address
assignment, each device must implement a unique device identifier (UDID).
854/1751 RM0090 Rev 19

For the details on 128-bit UDID and more information on ARP, refer to SMBus specification
version 2.0.
SMBus alert mode

SMBus Alert is an optional signal with an interrupt line for devices that want to trade their
ability to master for a pin. SMBA is a wired-AND signal just as the SCL and SDA signals are.
SMBA is used in conjunction with the SMBus General Call Address. Messages invoked with
the SMBus are two bytes long.
A slave-only device can signal the host through SMBA that it wants to talk by setting ALERT
bit in I2C_CR1 register. The host processes the interrupt and simultaneously accesses all
SMBA devices through the Alert Response Address (known as ARA having a value 0001
100X). Only the device(s) which pulled SMBA low will acknowledge the Alert Response
Address. This status is identified using SMBALERT Status flag in I2C_SR1 register. The
host performs a modified Receive Byte operation. The 7 bit device address provided by the
slave transmit device is placed in the 7 most significant bits of the byte. The eighth bit can
be a zero or one.
If more than one device pulls SMBA low, the highest priority (lowest address) device will win
communication rights via standard arbitration during the slave address transfer. After
acknowledging the slave address the device must disengage its SMBA pull-down. If the
host still sees SMBA low when the message transfer is complete, it knows to read the ARA
again.
A host which does not implement the SMBA signal may periodically access the ARA.
For more details on SMBus Alert mode, refer to SMBus specification version 2.0.
Timeout error
There are differences in the timing specifications between I2C and SMBus.
SMBus defines a clock low timeout, TIMEOUT of 35 ms. Also SMBus specifies TLOW:
SEXT as the cumulative clock low extend time for a slave device. SMBus specifies TLOW:
MEXT as the cumulative clock low extend time for a master device. For more details on
these timeouts, refer to SMBus specification version 2.0.
The status flag Timeout or Tlow Error in I2C_SR1 shows the status of this feature.
How to use the interface in SMBus mode

To switch from I2C mode to SMBus mode, the following sequence should be performed.
• Set the SMBus bit in the I2C_CR1 register
• Configure the SMBTYPE and ENARP bits in the I2C_CR1 register as required for the
application
If you want to configure the device as a master, follow the Start condition generation
procedure in Section 27.3.3. Otherwise, follow the sequence in Section 27.3.2.
The application has to control the various SMBus protocols by software.
• SMB Device Default Address acknowledged if ENARP=1 and SMBTYPE=0
• SMB Host Header acknowledged if ENARP=1 and SMBTYPE=1
• SMB Alert Response Address acknowledged if SMBALERT=1
RM0090 Rev 19 855/1751

872
27.3.8 DMA requests

DMA requests (when enabled) are generated only for data transfer. DMA requests are
generated by Data register becoming empty in transmission and Data register becoming full
in reception. The DMA must be initialized and enabled before the I2C data transfer. The
DMAEN bit must be set in the I2C_CR2 register before the ADDR event. In master mode or
in slave mode when clock stretching is enabled, the DMAEN bit can also be set during the
ADDR event, before clearing the ADDR flag. The DMA request must be served before the
end of the current byte transfer. When the number of data transfers which has been
programmed for the corresponding DMA stream is reached, the DMA controller sends an
End of Transfer EOT signal to the I2C interface and generates a Transfer Complete interrupt
if enabled:
• Master transmitter: In the interrupt routine after the EOT interrupt, disable DMA
requests then wait for a BTF event before programming the Stop condition.
• Master receiver
– When the number of bytes to be received is equal to or greater than two, the DMA
controller sends a hardware signal, EOT_1, corresponding to the last but one data
byte (number_of_bytes – 1). If, in the I2C_CR2 register, the LAST bit is set, I2C
automatically sends a NACK after the next byte following EOT_1. The user can
generate a Stop condition in the DMA Transfer Complete interrupt routine if
enabled.
– When a single byte must be received: the NACK must be programmed during EV6
event, i.e. program ACK=0 when ADDR=1, before clearing ADDR flag. Then the
user can program the STOP condition either after clearing ADDR flag, or in the
DMA Transfer Complete interrupt routine.
Transmission using DMA

DMA mode can be enabled for transmission by setting the DMAEN bit in the I2C_CR2
register. Data will be loaded from a Memory area configured using the DMA peripheral (refer
to the DMA specification) to the I2C_DR register whenever the TxE bit is set. To map a DMA
stream x for I2C transmission (where x is the stream number), perform the following
sequence:
1. Set the I2C_DR register address in the DMA_SxPAR register. The data will be moved
to this address from the memory after each TxE event.
2. Set the memory address in the DMA_SxMA0R register (and in DMA_SxMA1R register
in the case of a bouble buffer mode). The data will be loaded into I2C_DR from this
memory after each TxE event.
3. Configure the total number of bytes to be transferred in the DMA_SxNDTR register.
After each TxE event, this value will be decremented.
4. Configure the DMA stream priority using the PL[0:1] bits in the DMA_SxCR register
5. Set the DIR bit in the DMA_SxCR register and configure interrupts after half transfer or
full transfer depending on application requirements.
When the number of data transfers which has been programmed in the DMA Controller
registers is reached, the DMA controller sends an End of Transfer EOT/ EOT_1 signal to the
I2C interface and the DMA generates an interrupt, if enabled, on the DMA stream interrupt
vector.
Note: Do not enable the ITBUFEN bit in the I2C_CR2 register if DMA is used for transmission.
856/1751 RM0090 Rev 19

Reception using DMA

DMA mode can be enabled for reception by setting the DMAEN bit in the I2C_CR2 register.
Data will be loaded from the I2C_DR register to a Memory area configured using the DMA
peripheral (refer to the DMA specification) whenever a data byte is received. To map a DMA
stream x for I2C reception (where x is the stream number), perform the following sequence:
1. Set the I2C_DR register address in DMA_SxPAR register. The data will be moved from
this address to the memory after each RxNE event.
2. Set the memory address in the DMA_SxMA0R register (and in DMA_SxMA1R register
in the case of a bouble buffer mode). The data will be loaded from the I2C_DR register
to this memory area after each RxNE event.
3. Configure the total number of bytes to be transferred in the DMA_SxNDTR register.
After each RxNE event, this value will be decremented.
4. Configure the stream priority using the PL[0:1] bits in the DMA_SxCR register
5. Reset the DIR bit and configure interrupts in the DMA_SxCR register after half transfer
or full transfer depending on application requirements.
When the number of data transfers which has been programmed in the DMA Controller
registers is reached, the DMA controller sends an End of Transfer EOT/ EOT_1 signal to the
I2C interface and DMA generates an interrupt, if enabled, on the DMA stream interrupt
vector.
Note: Do not enable the ITBUFEN bit in the I2C_CR2 register if DMA is used for reception.
27.3.9 Packet error checking

A PEC calculator has been implemented to improve the reliability of communication. The
PEC is calculated by using the C(x) = x8 + x2 + x + 1 CRC-8 polynomial serially on each bit.
• PEC calculation is enabled by setting the ENPEC bit in the I2C_CR1 register. PEC is a
CRC-8 calculated on all message bytes including addresses and R/W bits.
– In transmission: set the PEC transfer bit in the I2C_CR1 register after the TxE
event corresponding to the last byte. The PEC will be transferred after the last
transmitted byte.
– In reception: set the PEC bit in the I2C_CR1 register after the RxNE event
corresponding to the last byte so that the receiver sends a NACK if the next
received byte is not equal to the internally calculated PEC. In case of Master-
Receiver, a NACK must follow the PEC whatever the check result. The PEC must
RM0090 Rev 19 857/1751

872
be set before the ACK of the CRC reception in slave mode. It must be set when
the ACK is set low in master mode.
• A PECERR error flag/interrupt is also available in the I2C_SR1 register.
• If DMA and PEC calculation are both enabled:-
– In transmission: when the I2C interface receives an EOT signal from the DMA
controller, it automatically sends a PEC after the last byte.
– In reception: when the I2C interface receives an EOT_1 signal from the DMA
controller, it will automatically consider the next byte as a PEC and will check it. A
DMA request is generated after PEC reception.
• To allow intermediate PEC transfers, a control bit is available in the I2C_CR2 register
(LAST bit) to determine if it is really the last DMA transfer or not. If it is the last DMA
request for a master receiver, a NACK is automatically sent after the last received byte.
• PEC calculation is corrupted by an arbitration loss.
27.4 I2C interrupts

The table below gives the list of I2C interrupt requests.
Table 124. I2C Interrupt requests

Start bit sent (Master) SB

Address sent (Master) or Address matched (Slave) ADDR
10-bit header sent (Master) ADD10 ITEVFEN
Stop received (Slave) STOPF
Data byte transfer finished BTF
Receive buffer not empty RxNE
ITEVFEN and ITBUFEN
Transmit buffer empty TxE
Bus error BERR
Arbitration loss (Master) ARLO
Acknowledge failure AF
Overrun/Underrun OVR ITERREN
PEC error PECERR
Timeout/Tlow error TIMEOUT
SMBus Alert SMBALERT
Note: SB, ADDR, ADD10, STOPF, BTF, RxNE and TxE are logically OR-ed on the same interrupt
channel.
BERR, ARLO, AF, OVR, PECERR, TIMEOUT and SMBALERT are logically OR-ed on the
same interrupt channel.
858/1751 RM0090 Rev 19

Figure 245. I2C interrupt mapping diagram

ITEVFEN
SB
ADDR
ADD10
STOPF
it_event
BTF
TxE
ITBUFEN
RxNE
ITERREN
BERR
ARLO
it_error
AF
OVR
PECERR
TIMEOUT
SMBALERT
MS42082V1
RM0090 Rev 19 859/1751

872
27.5 I2C debug mode

When the microcontroller enters the debug mode (Cortex®-M4 with FPU core halted), the
SMBUS timeout either continues to work normally or stops, depending on the
DBG_I2Cx_SMBUS_TIMEOUT configuration bits in the DBG module. For more details,
refer to Section 38.16.2: Debug support for timers, watchdog, bxCAN and I2C.
27.6 I2C registers

27.6.1 I2C Control register 1 (I2C_CR1)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
NO SMB SMBU
SWRST ALERT PEC POS ACK STOP START ENGC ENPEC ENARP PE
Res. STRETCH TYPE Res. S
Bit 15 SWRST: Software reset

When set, the I2C is under reset state. Before resetting this bit, make sure the I2C lines are
released and the bus is free.
0: I2C Peripheral not under reset
1: I2C Peripheral under reset state
Note: This bit can be used to reinitialize the peripheral after an error or a locked state. As an
example, if the BUSY bit is set and remains locked due to a glitch on the bus, the
SWRST bit can be used to exit from this state.
Bit 13 ALERT: SMBus alert
This bit is set and cleared by software, and cleared by hardware when PE=0.
0: Releases SMBA pin high. Alert Response Address Header followed by NACK.
1: Drives SMBA pin low. Alert Response Address Header followed by ACK.
Bit 12 PEC: Packet error checking
This bit is set and cleared by software, and cleared by hardware when PEC is transferred or
by a START or Stop condition or when PE=0.
0: No PEC transfer
1: PEC transfer (in Tx or Rx mode)
Note: PEC calculation is corrupted by an arbitration loss.
Bit 11 POS: Acknowledge/PEC Position (for data reception)
This bit is set and cleared by software and cleared by hardware when PE=0.
0: ACK bit controls the (N)ACK of the current byte being received in the shift register. The
PEC bit indicates that current byte in shift register is a PEC.
1: ACK bit controls the (N)ACK of the next byte which will be received in the shift register.
The PEC bit indicates that the next byte in the shift register is a PEC
Note: The POS bit must be used only in 2-byte reception configuration in master mode. It
must be configured before data reception starts, as described in the 2-byte reception
procedure recommended in Section : Master receiver on page 849.
860/1751 RM0090 Rev 19

Bit 10 ACK: Acknowledge enable

This bit is set and cleared by software and cleared by hardware when PE=0.
0: No acknowledge returned
1: Acknowledge returned after a byte is received (matched address or data)
Bit 9 STOP: Stop generation
The bit is set and cleared by software, cleared by hardware when a Stop condition is
detected, set by hardware when a timeout error is detected.
In Master Mode:
0: No Stop generation.
1: Stop generation after the current byte transfer or after the current Start condition is sent.
In Slave mode:
0: No Stop generation.
1: Release the SCL and SDA lines after the current byte transfer.
Bit 8 START: Start generation
This bit is set and cleared by software and cleared by hardware when start is sent or PE=0.
In Master Mode:
0: No Start generation
1: Repeated start generation
In Slave mode:
0: No Start generation
1: Start generation when the bus is free
Bit 7 NOSTRETCH: Clock stretching disable (Slave mode)
This bit is used to disable clock stretching in slave mode when ADDR or BTF flag is set, until
it is reset by software.
0: Clock stretching enabled
1: Clock stretching disabled
Bit 6 ENGC: General call enable
0: General call disabled. Address 00h is NACKed.
1: General call enabled. Address 00h is ACKed.
Bit 5 ENPEC: PEC enable
0: PEC calculation disabled
1: PEC calculation enabled
Bit 4 ENARP: ARP enable
0: ARP disable
1: ARP enable
SMBus Device default address recognized if SMBTYPE=0
SMBus Host address recognized if SMBTYPE=1
Bit 3 SMBTYPE: SMBus type
0: SMBus Device
1: SMBus Host
RM0090 Rev 19 861/1751

872

Bit 1 SMBUS: SMBus mode
0: I2C mode
1: SMBus mode
Bit 0 PE: Peripheral enable
0: Peripheral disable
1: Peripheral enable
Note: If this bit is reset while a communication is on going, the peripheral is disabled at the
end of the current communication, when back to IDLE state.
All bit resets due to PE=0 occur at the end of the communication.
In master mode, this bit must not be reset before the end of the communication.
Note: When the STOP, START or PEC bit is set, the software must not perform any write access
to I2C_CR1 before this bit is cleared by hardware. Otherwise there is a risk of setting a
second STOP, START or PEC request.
27.6.2 I2C Control register 2 (I2C_CR2)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
LAST DMAEN ITBUFEN ITEVTEN ITERREN FREQ[5:0]

Reserved Reserved

Bit 12 LAST: DMA last transfer
0: Next DMA EOT is not the last transfer
1: Next DMA EOT is the last transfer
Note: This bit is used in master receiver mode to permit the generation of a NACK on the last
received data.
Bit 11 DMAEN: DMA requests enable
0: DMA requests disabled
1: DMA request enabled when TxE=1 or RxNE =1
Bit 10 ITBUFEN: Buffer interrupt enable
0: TxE = 1 or RxNE = 1 does not generate any interrupt.
1: TxE = 1 or RxNE = 1 generates Event Interrupt (whatever the state of DMAEN)
862/1751 RM0090 Rev 19

Bit 9 ITEVTEN: Event interrupt enable

0: Event interrupt disabled
1: Event interrupt enabled
This interrupt is generated when:
– SB = 1 (Master)
– ADDR = 1 (Master/Slave)
– ADD10= 1 (Master)
– STOPF = 1 (Slave)
– BTF = 1 with no TxE or RxNE event
– TxE event to 1 if ITBUFEN = 1
– RxNE event to 1if ITBUFEN = 1
Bit 8 ITERREN: Error interrupt enable
0: Error interrupt disabled
1: Error interrupt enabled
This interrupt is generated when:
– BERR = 1
– ARLO = 1
– AF = 1
– OVR = 1
– PECERR = 1
– TIMEOUT = 1
– SMBALERT = 1
Bits 5:0 FREQ[5:0]: Peripheral clock frequency
The FREQ bits must be configured with the APB clock frequency value (I2C peripheral
connected to APB). The FREQ field is used by the peripheral to generate data setup and
hold times compliant with the I2C specifications. The minimum allowed frequency is 2 MHz,
the maximum frequency is limited by the maximum APB frequency and cannot exceed
50 MHz (peripheral intrinsic maximum limit).
0b000000: Not allowed
0b000001: Not allowed
0b000010: 2 MHz
...
0b110010: 50 MHz
Higher than 0b101010: Not allowed
RM0090 Rev 19 863/1751

872
27.6.3 I2C Own address register 1 (I2C_OAR1)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ADD
ADD[9:8] ADD[7:1] ADD0
MODE Reserved
Bit 15 ADDMODE Addressing mode (slave mode)

0: 7-bit slave address (10-bit address not acknowledged)
1: 10-bit slave address (7-bit address not acknowledged)
Bit 14 Should always be kept at 1 by software.
Bits 9:8 ADD[9:8]: Interface address
7-bit addressing mode: don’t care
10-bit addressing mode: bits9:8 of address
Bits 7:1 ADD[7:1]: Interface address
bits 7:1 of address
Bit 0 ADD0: Interface address
7-bit addressing mode: don’t care
10-bit addressing mode: bit 0 of address
27.6.4 I2C Own address register 2 (I2C_OAR2)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ADD2[7:1] ENDUAL
Reserved

Bits 7:1 ADD2[7:1]: Interface address
bits 7:1 of address in dual addressing mode
Bit 0 ENDUAL: Dual addressing mode enable
0: Only OAR1 is recognized in 7-bit addressing mode
1: Both OAR1 and OAR2 are recognized in 7-bit addressing mode
864/1751 RM0090 Rev 19

27.6.5 I2C Data register (I2C_DR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DR[7:0]
Reserved

Bits 7:0 DR[7:0] 8-bit data register
Byte received or to be transmitted to the bus.
– Transmitter mode: Byte transmission starts automatically when a byte is written in the DR
register. A continuous transmit stream can be maintained if the next data to be transmitted is
put in DR once the transmission is started (TxE=1)
– Receiver mode: Received byte is copied into DR (RxNE=1). A continuous transmit stream
can be maintained if DR is read before the next data byte is received (RxNE=1).
Note: In slave mode, the address is not copied into DR.
Write collision is not managed (DR can be written if TxE=0).
If an ARLO event occurs on ACK pulse, the received byte is not copied into DR and so
cannot be read.
27.6.6 I2C Status register 1 (I2C_SR1)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SMB TIME PEC
OVR AF ARLO BERR TxE RxNE STOPF ADD10 BTF ADDR SB
ALERT OUT Res. ERR Res.
rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 r r r r r r r
Bit 15 SMBALERT: SMBus alert

In SMBus host mode:
0: no SMBALERT
1: SMBALERT event occurred on pin
In SMBus slave mode:
0: no SMBALERT response address header
1: SMBALERT response address header to SMBALERT LOW received
– Cleared by software writing 0, or by hardware when PE=0.
RM0090 Rev 19 865/1751

872
Bit 14 TIMEOUT: Timeout or Tlow error

0: No timeout error
1: SCL remained LOW for 25 ms (Timeout)
or
Master cumulative clock low extend time more than 10 ms (Tlow:mext)
or
Slave cumulative clock low extend time more than 25 ms (Tlow:sext)
– When set in slave mode: slave resets the communication and lines are released by
hardware
– When set in master mode: Stop condition sent by hardware
Note: This functionality is available only in SMBus mode.
Bit 12 PECERR: PEC Error in reception
0: no PEC error: receiver returns ACK after PEC reception (if ACK=1)
1: PEC error: receiver returns NACK after PEC reception (whatever ACK)
Note: When the received CRC is wrong, PECERR is not set in slave mode if the PEC control
bit is not set before the end of the CRC reception. Nevertheless, reading the PEC value
determines whether the received CRC is right or wrong.
Bit 11 OVR: Overrun/Underrun
0: No overrun/underrun
1: Overrun or underrun
– Set by hardware in slave mode when NOSTRETCH=1 and:
– In reception when a new byte is received (including ACK pulse) and the DR register has not
been read yet. New received byte is lost.
– In transmission when a new byte should be sent and the DR register has not been written
yet. The same byte is sent twice.
Note: If the DR write occurs very close to SCL rising edge, the sent data is unspecified and a
hold timing error occurs
Bit 10 AF: Acknowledge failure
0: No acknowledge failure
1: Acknowledge failure
– Set by hardware when no acknowledge is returned.
Bit 9 ARLO: Arbitration lost (master mode)
0: No Arbitration Lost detected
1: Arbitration Lost detected
Set by hardware when the interface loses the arbitration of the bus to another master
After an ARLO event the interface switches back automatically to Slave mode (MSL=0).
Note: In SMBUS, the arbitration on the data in slave mode occurs only during the data phase,
or the acknowledge transmission (not on the address acknowledge).
866/1751 RM0090 Rev 19

Bit 8 BERR: Bus error

0: No misplaced Start or Stop condition
1: Misplaced Start or Stop condition
– Set by hardware when the interface detects an SDA rising or falling edge while SCL is high,
occurring in a non-valid position during a byte transfer.
Bit 7 TxE: Data register empty (transmitters)
0: Data register not empty
1: Data register empty
– Set when DR is empty in transmission. TxE is not set during address phase.
– Cleared by software writing to the DR register or by hardware after a start or a stop condition
or when PE=0.
TxE is not set if either a NACK is received, or if next byte to be transmitted is PEC (PEC=1)
Note: TxE is not cleared by writing the first data being transmitted, or by writing data when
BTF is set, as in both cases the data register is still empty.
Bit 6 RxNE: Data register not empty (receivers)
0: Data register empty
1: Data register not empty
– Set when data register is not empty in receiver mode. RxNE is not set during address phase.
– Cleared by software reading or writing the DR register or by hardware when PE=0.
RxNE is not set in case of ARLO event.
Note: RxNE is not cleared by reading data when BTF is set, as the data register is still full.
Bit 4 STOPF: Stop detection (slave mode)
0: No Stop condition detected
1: Stop condition detected
– Set by hardware when a Stop condition is detected on the bus by the slave after an
acknowledge (if ACK=1).
– Cleared by software reading the SR1 register followed by a write in the CR1 register, or by
hardware when PE=0
Note: The STOPF bit is not set after a NACK reception.
It is recommended to perform the complete clearing sequence (READ SR1 then
WRITE CR1) after the STOPF is set. Refer to Figure 242.
Bit 3 ADD10: 10-bit header sent (Master mode)
0: No ADD10 event occurred.
1: Master has sent first address byte (header).
– Set by hardware when the master has sent the first byte in 10-bit address mode.
– Cleared by software reading the SR1 register followed by a write in the DR register of the
second address byte, or by hardware when PE=0.
Note: ADD10 bit is not set after a NACK reception
RM0090 Rev 19 867/1751

872
Bit 2 BTF: Byte transfer finished

0: Data byte transfer not done
1: Data byte transfer succeeded
– Set by hardware when NOSTRETCH=0 and:
– In reception when a new byte is received (including ACK pulse) and DR has not been read
yet (RxNE=1).
– In transmission when a new byte should be sent and DR has not been written yet (TxE=1).
– Cleared by software by either a read or write in the DR register or by hardware after a start or
a stop condition in transmission or when PE=0.
Note: The BTF bit is not set after a NACK reception
The BTF bit is not set if next byte to be transmitted is the PEC (TRA=1 in I2C_SR2
register and PEC=1 in I2C_CR1 register)
Bit 1 ADDR: Address sent (master mode)/matched (slave mode)
This bit is cleared by software reading SR1 register followed reading SR2, or by hardware
when PE=0.
Address matched (Slave)
0: Address mismatched or not received.
1: Received address matched.
– Set by hardware as soon as the received slave address matched with the OAR registers
content or a general call or a SMBus Device Default Address or SMBus Host or SMBus Alert
is recognized. (when enabled depending on configuration).
Note: In slave mode, it is recommended to perform the complete clearing sequence (READ
SR1 then READ SR2) after ADDR is set. Refer to Figure 242.
Address sent (Master)
0: No end of address transmission
1: End of address transmission
– For 10-bit addressing, the bit is set after the ACK of the 2nd byte.
– For 7-bit addressing, the bit is set after the ACK of the byte.
Note: ADDR is not set after a NACK reception
Bit 0 SB: Start bit (Master mode)
0: No Start condition
1: Start condition generated.
– Set when a Start condition generated.
– Cleared by software by reading the SR1 register followed by writing the DR register, or by
hardware when PE=0
27.6.7 I2C Status register 2 (I2C_SR2)

Reset value: 0x0000
Note: Reading I2C_SR2 after reading I2C_SR1 clears the ADDR flag, even if the ADDR flag was
set after reading I2C_SR1. Consequently, I2C_SR2 must be read only when ADDR is found
set in I2C_SR1 or when the STOPF bit is cleared.
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SMB SMBDE GEN
PEC[7:0] DUALF TRA BUSY MSL
HOST FAULT CALL Res.
r r r r r r r r r r r r r r r
868/1751 RM0090 Rev 19

Bits 15:8 PEC[7:0] Packet error checking register

This register contains the internal PEC when ENPEC=1.
Bit 7 DUALF: Dual flag (Slave mode)
0: Received address matched with OAR1
1: Received address matched with OAR2
– Cleared by hardware after a Stop condition or repeated Start condition, or when PE=0.
Bit 6 SMBHOST: SMBus host header (Slave mode)
0: No SMBus Host address
1: SMBus Host address received when SMBTYPE=1 and ENARP=1.
Bit 5 SMBDEFAULT: SMBus device default address (Slave mode)
0: No SMBus Device Default address
1: SMBus Device Default address received when ENARP=1
Bit 4 GENCALL: General call address (Slave mode)
0: No General Call
1: General Call Address received when ENGC=1
Bit 2 TRA: Transmitter/receiver
0: Data bytes received
1: Data bytes transmitted
This bit is set depending on the R/W bit of the address byte, at the end of total address
phase.
It is also cleared by hardware after detection of Stop condition (STOPF=1), repeated Start
condition, loss of bus arbitration (ARLO=1), or when PE=0.
Bit 1 BUSY: Bus busy
0: No communication on the bus
1: Communication ongoing on the bus
– Set by hardware on detection of SDA or SCL low
– cleared by hardware on detection of a Stop condition.
It indicates a communication in progress on the bus. This information is still updated when
the interface is disabled (PE=0).
Bit 0 MSL: Master/slave
0: Slave Mode
1: Master Mode
– Set by hardware as soon as the interface is in Master mode (SB=1).
– Cleared by hardware after detecting a Stop condition on the bus or a loss of arbitration
(ARLO=1), or by hardware when PE=0.
Note: Reading I2C_SR2 after reading I2C_SR1 clears the ADDR flag, even if the ADDR flag was
set after reading I2C_SR1. Consequently, I2C_SR2 must be read only when ADDR is found
set in I2C_SR1 or when the STOPF bit is cleared.
RM0090 Rev 19 869/1751

872
27.6.8 I2C Clock control register (I2C_CCR)

Reset value: 0x0000
Note: fPCLK1 must be at least 2 MHz to achieve Sm mode I²C frequencies. It must be at least 4
MHz to achieve Fm mode I²C frequencies. It must be a multiple of 10MHz to reach the
400 kHz maximum I²C Fm mode clock.
The CCR register must be configured only when the I2C is disabled (PE = 0).
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
F/S DUTY CCR[11:0]
Reserved
Bit 15 F/S: I2C master mode selection

0: Sm mode I2C
1: Fm mode I2C
Bit 14 DUTY: Fm mode duty cycle
0: Fm mode tlow/thigh = 2
1: Fm mode tlow/thigh = 16/9 (see CCR)
Bits 11:0 CCR[11:0]: Clock control register in Fm/Sm mode (Master mode)
Controls the SCL clock in master mode.
Sm mode or SMBus:
Thigh = CCR * TPCLK1
Tlow = CCR * TPCLK1
Fm mode:
If DUTY = 0:
Thigh = CCR * TPCLK1
Tlow = 2 * CCR * TPCLK1
If DUTY = 1:
Thigh = 9 * CCR * TPCLK1
Tlow = 16 * CCR * TPCLK1
For instance: in Sm mode, to generate a 100 kHz SCL frequency:
If FREQ = 08, TPCLK1 = 125 ns so CCR must be programmed with 0x28
(0x28 <=> 40d x 125 ns = 5000 ns.)
Note: The minimum allowed value is 0x04, except in FAST DUTY mode where the minimum
allowed value is 0x01
thigh = tr(SCL) + tw(SCLH). See device datasheet for the definitions of parameters.
tlow = tf(SCL) + tw(SCLL). See device datasheet for the definitions of parameters.
I2C communication speed, fSCL ~ 1/(thigh + tlow). The real frequency may differ due to
the analog noise filter input delay.
The CCR register must be configured only when the I2C is disabled (PE = 0).
870/1751 RM0090 Rev 19

27.6.9 I2C TRISE register (I2C_TRISE)

Reset value: 0x0002
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TRISE[5:0]
Reserved
rw rw rw rw rw rw

Bits 5:0 TRISE[5:0]: Maximum rise time in Fm/Sm mode (Master mode)
These bits should provide the maximum duration of the SCL feedback loop in master mode.
The purpose is to keep a stable SCL frequency whatever the SCL rising edge duration.
These bits must be programmed with the maximum SCL rise time given in the I2C bus
specification, incremented by 1.
For instance: in Sm mode, the maximum allowed SCL rise time is 1000 ns.
If, in the I2C_CR2 register, the value of FREQ[5:0] bits is equal to 0x08 and TPCLK1 = 125 ns
therefore the TRISE[5:0] bits must be programmed with 09h.
(1000 ns / 125 ns = 8 + 1)
The filter value can also be added to TRISE[5:0].
If the result is not an integer, TRISE[5:0] must be programmed with the integer part, in order
to respect the tHIGH parameter.
Note: TRISE[5:0] must be configured only when the I2C is disabled (PE = 0).
27.6.10 I2C FLTR register (I2C_FLTR)

Reset value: 0x0000
The I2C_FLTR is available on STM32F42xxx and STM32F43xxx only.
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ANOFF DNF[3:0]
Reserved
rw rw rw rw rw

Bit 4 ANOFF: Analog noise filter OFF
0: Analog noise filter enable
1: Analog noise filter disable
Note: ANOFF must be configured only when the I2C is disabled (PE = 0).
Bits 3:0 DNF[3:0]: Digital noise filter
These bits are used to configure the digital noise filter on SDA and SCL inputs. The digital filter
will suppress the spikes with a length of up to DNF[3:0] * TPCLK1.
0000: Digital noise filter disable
0001: Digital noise filter enabled and filtering capability up to 1* TPCLK1.
...
1111: Digital noise filter enabled and filtering capability up to 15* TPCLK1.
Note: DNF[3:0] must be configured only when the I2C is disabled (PE = 0). If the analog filter
is also enabled, the digital filter is added to the analog filter.
RM0090 Rev 19 871/1751

872
27.6.11 I2C register map

The table below provides the I2C register map and reset values.
Table 125. I2C register map and reset values

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
NOSTRETCH
SMBTYPE
SWRST
SMBUS
ENPEC
ENARP
ALERT
START
ENGC
STOP
Reserved
Reserved
POS
ACK
PEC
I2C_CR1
PE
0x00 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ITERREN
ITBUFEN
ITEVTEN
DMAEN
Reserved
LAST
I2C_CR2 FREQ[5:0]
0x04 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0
ADDMODE
ADD0
I2C_OAR1 ADD[9:8] ADD[7:1]
Reset value 0 0 0 0 0 0 0 0 0 0 0
ENDUAL
I2C_OAR2 ADD2[7:1]
0x0C Reserved
Reset value 0 0 0 0 0 0 0 0
I2C_DR DR[7:0]
0x10 Reserved
Reset value 0 0 0 0 0 0 0 0
SMBALERT
TIMEOUT
PECERR
STOPF
ADD10
ADDR
BERR
ARLO
RxNE
Reserved
Reserved
OVR
BTF
TxE
I2C_SR1
SB
AF
0x14 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0
SMBDEFAUL
SMBHOST
GENCALL
DUALF
BUSY
Reserved
MSL
TRA
I2C_SR2 PEC[7:0]
0x18 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DUTY
Reserved
I2C_CCR CCR[11:0]
F/S
0x1C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0
I2C_TRISE TRISE[5:0]
0x20 Reserved
ANOFF
I2C_FLTR DNF[3:0]
0x24 Reserved
Refer to Section 2.3: Memory map for the register boundary addresses table.
872/1751 RM0090 Rev 19

RM0090 Serial peripheral interface (SPI)
28 Serial peripheral interface (SPI)
28.1 SPI introduction

The SPI interface provides two main functions, supporting either the SPI protocol or the I2S
audio protocol. By default, it is the SPI function that is selected. It is possible to switch the
interface from SPI to I2S by software.
The serial peripheral interface (SPI) allows half/ full-duplex, synchronous, serial
communication with external devices. The interface can be configured as the master and in
this case it provides the communication clock (SCK) to the external slave device. The
interface is also capable of operating in multimaster configuration.
It may be used for a variety of purposes, including simplex synchronous transfers on two
lines with a possible bidirectional data line or reliable communication using CRC checking.
The I2S is also a synchronous serial communication interface. It can address four different
audio standards including the I2S Philips standard, the MSB- and LSB-justified standards,
and the PCM standard. It can operate as a slave or a master device in full-duplex mode
(using 4 pins) or in half-duplex mode (using 3 pins). Master clock can be provided by the
interface to an external slave component when the I2S is configured as the communication
master.
Warning: Since some SPI1 and SPI3/I2S3 pins may be mapped onto
some pins used by the JTAG interface (SPI1_NSS onto JTDI,
SPI3_NSS/I2S3_WS onto JTDI and SPI3_SCK/I2S3_CK onto
JTDO), you may either:
– map SPI/I2S onto other pins
– disable the JTAG and use the SWD interface prior to
configuring the pins listed as SPI I/Os (when debugging the
application) or
– disable both JTAG/SWD interfaces (for standalone
applications).
For more information on the configuration of the JTAG/SWD
interface pins, please refer to Section 8.3.2: I/O pin
multiplexer and mapping.
RM0090 Rev 19 873/1751

925
Serial peripheral interface (SPI) RM0090
28.2 SPI and I2S main features
28.2.1 SPI features

• Full-duplex synchronous transfers on three lines
• Simplex synchronous transfers on two lines with or without a bidirectional data line
• 8- or 16-bit transfer frame format selection
• Master or slave operation
• Multimaster mode capability
• 8 master mode baud rate prescalers (fPCLK/2 max.)
• Slave mode frequency (fPCLK/2 max)
• Faster communication for both master and slave
• NSS management by hardware or software for both master and slave: dynamic change
of master/slave operations
• Programmable clock polarity and phase
• Programmable data order with MSB-first or LSB-first shifting
• Dedicated transmission and reception flags with interrupt capability
• SPI bus busy status flag
• SPI TI mode
• Hardware CRC feature for reliable communication:
– CRC value can be transmitted as last byte in Tx mode
– Automatic CRC error checking for last received byte
• Master mode fault, overrun and CRC error flags with interrupt capability
• 1-byte transmission and reception buffer with DMA capability: Tx and Rx requests
874/1751 RM0090 Rev 19

28.2.2 I2S features

• Full duplex communication
• Half-duplex communication (only transmitter or receiver)
• Master or slave operations
• 8-bit programmable linear prescaler to reach accurate audio sample frequencies (from
8 kHz to 192 kHz)
• Data format may be 16-bit, 24-bit or 32-bit
• Packet frame is fixed to 16-bit (16-bit data frame) or 32-bit (16-bit, 24-bit, 32-bit data
frame) by audio channel
• Programmable clock polarity (steady state)
• Underrun flag in slave transmission mode, overrun flag in reception mode (master and
slave), and Frame Error flag in reception and transmission mode (slave only)
• 16-bit register for transmission and reception with one data register for both channel
sides
• Supported I2S protocols:
– I2S Phillps standard
– MSB-justified standard (left-justified)
– LSB-justified standard (right-justified)
– PCM standard (with short and long frame synchronization on 16-bit channel frame
or 16-bit data frame extended to 32-bit channel frame)
• Data direction is always MSB first
• DMA capability for transmission and reception (16-bit wide)
• Master clock may be output to drive an external audio component. Ratio is fixed at
256 × FS (where FS is the audio sampling frequency)
• Both I2S (I2S2 and I2S3) have a dedicated PLL (PLLI2S) to generate an even more
accurate clock.
• I2S (I2S2 and I2S3) clock can be derived from an external clock mapped on the
I2S_CKIN pin.
RM0090 Rev 19 875/1751

925
28.3 SPI functional description

The block diagram of the SPI is shown in Figure 246.
Figure 246. SPI block diagram

Address and data bus
Read
Rx buffer
SPI_CR2
MOSI
TXE RXNE ERR TXDM RXDM
0 0 SSOE
IE IE IE AEN AEN
MISO Shift register

LSB first SPI_SR
MOD CRC
BSY OVR 0 0 TXE RXNE
Tx buffer F ERR
Write
0
Communication control
1
SCK BR[2:0]
Baud rate generator
LSB
SPE BR2 BR1 BR0 MSTR CPOL CPHA
FIRST
SPI_CR1
Master control logic BIDI BIDI
CRCEN
CRC
DFF
RX
ONLY SSM SSI
MODE OE Next
NSS
MS51604V1
Usually, the SPI is connected to external devices through four pins:

• MISO: Master In / Slave Out data. This pin can be used to transmit data in slave mode
and receive data in master mode.
• MOSI: Master Out / Slave In data. This pin can be used to transmit data in master
mode and receive data in slave mode.
• SCK: Serial Clock output for SPI masters and input for SPI slaves.
• NSS: Slave select. This is an optional pin to select a slave device. This pin acts as a
‘chip select’ to let the SPI master communicate with slaves individually and to avoid
contention on the data lines. Slave NSS inputs can be driven by standard IO ports on
the master device. The NSS pin may also be used as an output if enabled (SSOE bit)
and driven low if the SPI is in master configuration. In this manner, all NSS pins from
devices connected to the Master NSS pin see a low level and become slaves when
they are configured in NSS hardware mode. When configured in master mode with
NSS configured as an input (MSTR=1 and SSOE=0) and if NSS is pulled low, the SPI
enters the master mode fault state: the MSTR bit is automatically cleared and the
device is configured in slave mode (refer to Section 28.3.10).
A basic example of interconnections between a single master and a single slave is
illustrated in Figure 247.
876/1751 RM0090 Rev 19

Figure 247. Single master/ single slave application

Master Slave
MSBit LSBit MSBit LSBit
MISO MISO
8-bit shift register 8-bit shift register
MOSI MOSI
SPI clock SCK SCK

generator
NSS(1) NSS(1)
VDD
Not used if NSS is managed

by software
ai14745
1. Here, the NSS pin is configured as an input.

The MOSI pins are connected together and the MISO pins are connected together. In this
way data is transferred serially between master and slave (most significant bit first).
The communication is always initiated by the master. When the master device transmits
data to a slave device via the MOSI pin, the slave device responds via the MISO pin. This
implies full-duplex communication with both data out and data in synchronized with the
same clock signal (which is provided by the master device via the SCK pin).
Slave select (NSS) pin management

Hardware or software slave select management can be set using the SSM bit in the
SPI_CR1 register.
• Software NSS management (SSM = 1)
The slave select information is driven internally by the value of the SSI bit in the
SPI_CR1 register. The external NSS pin remains free for other application uses.
• Hardware NSS management (SSM = 0)
Two configurations are possible depending on the NSS output configuration (SSOE bit
in register SPI_CR2).
– NSS output enabled (SSM = 0, SSOE = 1)
This configuration is used only when the device operates in master mode. The
NSS signal is driven low when the master starts the communication and is kept
low until the SPI is disabled.
– NSS output disabled (SSM = 0, SSOE = 0)
This configuration allows multimaster capability for devices operating in master
mode. For devices set as slave, the NSS pin acts as a classical NSS input: the
slave is selected when NSS is low and deselected when NSS high.
RM0090 Rev 19 877/1751

925
Clock phase and clock polarity

Four possible timing relationships may be chosen by software, using the CPOL and CPHA
bits in the SPI_CR1 register. The CPOL (clock polarity) bit controls the steady state value of
the clock when no data is being transferred. This bit affects both master and slave modes. If
CPOL is reset, the SCK pin has a low-level idle state. If CPOL is set, the SCK pin has a
high-level idle state.
If the CPHA (clock phase) bit is set, the second edge on the SCK pin (falling edge if the
CPOL bit is reset, rising edge if the CPOL bit is set) is the MSBit capture strobe. Data are
latched on the occurrence of the second clock transition. If the CPHA bit is reset, the first
edge on the SCK pin (falling edge if CPOL bit is set, rising edge if CPOL bit is reset) is the
MSBit capture strobe. Data are latched on the occurrence of the first clock transition.
The combination of the CPOL (clock polarity) and CPHA (clock phase) bits selects the data
capture clock edge.
Figure 248, shows an SPI transfer with the four combinations of the CPHA and CPOL bits.
The diagram may be interpreted as a master or slave timing diagram where the SCK pin,
the MISO pin, the MOSI pin are directly connected between the master and the slave
device.
Note: Prior to changing the CPOL/CPHA bits the SPI must be disabled by resetting the SPE bit.
Master and slave must be programmed with the same timing mode.
The idle state of SCK must correspond to the polarity selected in the SPI_CR1 register (by
pulling up SCK if CPOL=1 or pulling down SCK if CPOL=0).
The Data Frame Format (8- or 16-bit) is selected through the DFF bit in SPI_CR1 register,
and determines the data length during transmission/reception.
878/1751 RM0090 Rev 19

Figure 248. Data clock timing diagram
CPHA =1
CPOL = 1
CPOL = 0
MOSI MSBit LSBit
MISO MSBit LSBit
NSS
(to slave)
Capture strobe
CPHA =0
CPOL = 1
CPOL = 0
MOSI MSBit LSBit
MISO MSBit LSBit
NSS
(to slave)
Capture strobe
ai17154d
1. These timings are shown with the LSBFIRST bit reset in the SPI_CR1 register.
Data frame format

Data can be shifted out either MSB-first or LSB-first depending on the value of the
LSBFIRST bit in the SPI_CR1 register.
Each data frame is 8 or 16 bits long depending on the size of the data programmed using
the DFF bit in the SPI_CR1 register. The selected data frame format is applicable for
transmission and/or reception.
RM0090 Rev 19 879/1751

925
28.3.2 Configuring the SPI in slave mode

In the slave configuration, the serial clock is received on the SCK pin from the master
device. The value set in the BR[2:0] bits in the SPI_CR1 register, does not affect the data
transfer rate.
Note: It is recommended to enable the SPI slave before the master sends the clock. If not,
undesired data transmission might occur. The data register of the slave needs to be ready
before the first edge of the communication clock or before the end of the ongoing
communication. It is mandatory to have the polarity of the communication clock set to the
steady state value before the slave and the master are enabled.
Follow the procedure below to configure the SPI in slave mode:
Procedure
1. Set the DFF bit to define 8- or 16-bit data frame format
2. Select the CPOL and CPHA bits to define one of the four relationships between the
data transfer and the serial clock (see Figure 248). For correct data transfer, the CPOL
and CPHA bits must be configured in the same way in the slave device and the master
device. This step is not required when the TI mode is selected through the FRF bit in
the SPI_CR2 register.
3. The frame format (MSB-first or LSB-first depending on the value of the LSBFIRST bit in
the SPI_CR1 register) must be the same as the master device. This step is not
required when TI mode is selected.
4. In Hardware mode (refer to Slave select (NSS) pin management), the NSS pin must be
connected to a low level signal during the complete byte transmit sequence. In NSS
software mode, set the SSM bit and clear the SSI bit in the SPI_CR1 register. This step
is not required when TI mode is selected.
5. Set the FRF bit in the SPI_CR2 register to select the TI mode protocol for serial
communications.
6. Clear the MSTR bit and set the SPE bit (both in the SPI_CR1 register) to assign the
pins to alternate functions.
In this configuration the MOSI pin is a data input and the MISO pin is a data output.
Transmit sequence
The data byte is parallel-loaded into the Tx buffer during a write cycle.
The transmit sequence begins when the slave device receives the clock signal and the most
significant bit of the data on its MOSI pin. The remaining bits (the 7 bits in 8-bit data frame
format, and the 15 bits in 16-bit data frame format) are loaded into the shift-register. The
TXE flag in the SPI_SR register is set on the transfer of data from the Tx Buffer to the shift
register and an interrupt is generated if the TXEIE bit in the SPI_CR2 register is set.
Receive sequence
For the receiver, when data transfer is complete:
• The Data in shift register is transferred to Rx Buffer and the RXNE flag (SPI_SR
register) is set
• An Interrupt is generated if the RXNEIE bit is set in the SPI_CR2 register.
880/1751 RM0090 Rev 19

After the last sampling clock edge the RXNE bit is set, a copy of the data byte received in
the shift register is moved to the Rx buffer. When the SPI_DR register is read, the SPI
peripheral returns this buffered value.
Clearing of the RXNE bit is performed by reading the SPI_DR register.
SPI TI protocol in slave mode

In slave mode, the SPI interface is compatible with the TI protocol. The FRF bit of the
SPI_CR2 register can be used to configure the slave SPI serial communications to be
compliant with this protocol.
The clock polarity and phase are forced to conform to the TI protocol requirements whatever
the values set in the SPI_CR1 register. NSS management is also specific to the TI protocol
which makes the configuration of NSS management through the SPI_CR1 and SPI_CR2
registers (such as SSM, SSI, SSOE) transparent for the user.
In Slave mode (Figure 249: TI mode - Slave mode, single transfer and Figure 250: TI mode
- Slave mode, continuous transfer), the SPI baud rate prescaler is used to control the
moment when the MISO pin state changes to HI-Z. Any baud rate can be used thus allowing
to determine this moment with optimal flexibility. However, the baud rate is generally set to
the external master clock baud rate. The time for the MISO signal to become HI-Z (trelease)
depends on internal resynchronizations and on the baud rate value set in through BR[2:0] of
SPI_CR1 register. It is given by the formula:
t baud_rate t baud_rate
---------------------------- + 4 × t pclk < t release < ---------------------------- + 6 × t pclk
2 2
Note: This feature is not available for Motorola SPI communications (FRF bit set to 0).
To detect TI frame errors in Slave transmitter only mode by using the Error interrupt (ERRIE
= 1), the SPI must be configured in 2-line unidirectional mode by setting BIDIMODE and
BIDIOE to 1 in the SPI_CR1 register. When BIDIMODE is set to 0, OVR is set to 1 because
the data register is never read and error interrupt are always generated, while when
BIDIMODE is set to 1, data are not received and OVR is never set.
Figure 249. TI mode - Slave mode, single transfer
NSS
input trigger sampling trigger sampling trigger sampling
edge edge edge edge edge edge
SCK
input t
Release
MOSI DONTCARE MSBIN LSBIN DONTCARE

input
MISO
output 1 or 0 MSBOUT LSBOUT
ai18434
RM0090 Rev 19 881/1751

925
Figure 250. TI mode - Slave mode, continuous transfer
NSS
input
trigger sampling trigger sampling trigger sampling
SCK
input
MOSI DONTCARE MSBIN LSBIN MSBIN LSBIN DONTCARE

input
MISO LSBOUT MSBOUT LSBOUT

1 or 0 MSBOUT
output
FRAME 1 FRAME 2
ai18435
28.3.3 Configuring the SPI in master mode

In the master configuration, the serial clock is generated on the SCK pin.
Procedure
1. Select the BR[2:0] bits to define the serial clock baud rate (see SPI_CR1 register).
2. Select the CPOL and CPHA bits to define one of the four relationships between the
data transfer and the serial clock (see Figure 248). This step is not required when the
TI mode is selected.
3. Set the DFF bit to define 8- or 16-bit data frame format
4. Configure the LSBFIRST bit in the SPI_CR1 register to define the frame format. This
step is not required when the TI mode is selected.
5. If the NSS pin is required in input mode, in hardware mode, connect the NSS pin to a
high-level signal during the complete byte transmit sequence. In NSS software mode,
set the SSM and SSI bits in the SPI_CR1 register. If the NSS pin is required in output
mode, the SSOE bit only should be set. This step is not required when the TI mode is
selected.
6. Set the FRF bit in SPI_CR2 to select the TI protocol for serial communications.
7. The MSTR and SPE bits must be set (they remain set only if the NSS pin is connected
to a high-level signal).
In this configuration the MOSI pin is a data output and the MISO pin is a data input.
Transmit sequence
The transmit sequence begins when a byte is written in the Tx Buffer.
The data byte is parallel-loaded into the shift register (from the internal bus) during the first
bit transmission and then shifted out serially to the MOSI pin MSB first or LSB first
depending on the LSBFIRST bit in the SPI_CR1 register. The TXE flag is set on the transfer
of data from the Tx Buffer to the shift register and an interrupt is generated if the TXEIE bit in
the SPI_CR2 register is set.
882/1751 RM0090 Rev 19

Receive sequence
For the receiver, when data transfer is complete:
• The data in the shift register is transferred to the RX Buffer and the RXNE flag is set
• An interrupt is generated if the RXNEIE bit is set in the SPI_CR2 register
At the last sampling clock edge the RXNE bit is set, a copy of the data byte received in the
shift register is moved to the Rx buffer. When the SPI_DR register is read, the SPI
peripheral returns this buffered value.
Clearing the RXNE bit is performed by reading the SPI_DR register.
A continuous transmit stream can be maintained if the next data to be transmitted is put in
the Tx buffer once the transmission is started. Note that TXE flag should be ‘1 before any
attempt to write the Tx buffer is made.
Note: When a master is communicating with SPI slaves which need to be de-selected between
transmissions, the NSS pin must be configured as GPIO or another GPIO must be used and
toggled by software.
SPI TI protocol in master mode

In master mode, the SPI interface is compatible with the TI protocol. The FRF bit of the
SPI_CR2 register can be used to configure the master SPI serial communications to be
compliant with this protocol.
The clock polarity and phase are forced to conform to the TI protocol requirements whatever
the values set in the SPI_CR1 register. NSS management is also specific to the TI protocol
which makes the configuration of NSS management through the SPI_CR1 and SPI_CR2
registers (SSM, SSI, SSOE) transparent for the user.
Figure 251: TI mode - master mode, single transfer and Figure 252: TI mode - master mode,
continuous transfer) show the SPI master communication waveforms when the TI mode is
selected in master mode.
Figure 251. TI mode - master mode, single transfer
NSS
output trigger sampling trigger sampling trigger sampling
edge edge edge edge edge edge
SCK
output
MOSI DONTCARE MSBIN LSBIN DONTCARE

input
MISO
output 1 or 0 MSBOUT LSBOUT
ai18436
RM0090 Rev 19 883/1751

925
Figure 252. TI mode - master mode, continuous transfer
NSS
output
trigger sampling trigger sampling trigger sampling
SCK
output
MOSI
DONTCARE MSBOUT LSBOUT MSBOUT LSBOUT
output
MISO LSBIN MSBIN LSBIN DONTCARE

1 or 0 MSBIN
intput
FRAME 1 FRAME 2
ai18437
28.3.4 Configuring the SPI for half-duplex communication

The SPI is capable of operating in half-duplex mode in 2 configurations.
• 1 clock and 1 bidirectional data wire
• 1 clock and 1 data wire (receive-only or transmit-only)
1 clock and 1 bidirectional data wire (BIDIMODE = 1)

This mode is enabled by setting the BIDIMODE bit in the SPI_CR1 register. In this mode
SCK is used for the clock and MOSI in master or MISO in slave mode is used for data
communication. The transfer direction (Input/Output) is selected by the BIDIOE bit in the
SPI_CR1 register. When this bit is 1, the data line is output otherwise it is input.
1 clock and 1 unidirectional data wire (BIDIMODE = 0)

In this mode, the application can use the SPI either in transmit-only mode or in receive-only
mode.
• Transmit-only mode is similar to full-duplex mode (BIDIMODE=0, RXONLY=0): the
data are transmitted on the transmit pin (MOSI in master mode or MISO in slave mode)
and the receive pin (MISO in master mode or MOSI in slave mode) can be used as a
general-purpose IO. In this case, the application just needs to ignore the Rx buffer (if
the data register is read, it does not contain the received value).
• In receive-only mode, the application can disable the SPI output function by setting the
RXONLY bit in the SPI_CR1 register. In this case, it frees the transmit IO pin (MOSI in
master mode or MISO in slave mode), so it can be used for other purposes.
To start the communication in receive-only mode, configure and enable the SPI:
• In master mode, the communication starts immediately and stops when the SPE bit is
cleared and the current reception stops. There is no need to read the BSY flag in this
mode. It is always set when an SPI communication is ongoing.
• In slave mode, the SPI continues to receive as long as the NSS is pulled down (or the
SSI bit is cleared in NSS software mode) and the SCK is running.
884/1751 RM0090 Rev 19

28.3.5 Data transmission and reception procedures

Rx and Tx buffers
In reception, data are received and then stored into an internal Rx buffer while In
transmission, data are first stored into an internal Tx buffer before being transmitted.
A read access of the SPI_DR register returns the Rx buffered value whereas a write access
to the SPI_DR stores the written data into the Tx buffer.
Start sequence in master mode

• In full-duplex (BIDIMODE=0 and RXONLY=0)
– The sequence begins when data are written into the SPI_DR register (Tx buffer).
– The data are then parallel loaded from the Tx buffer into the 8-bit shift register
during the first bit transmission and then shifted out serially to the MOSI pin.
– At the same time, the received data on the MISO pin is shifted in serially to the 8-
bit shift register and then parallel loaded into the SPI_DR register (Rx buffer).
• In unidirectional receive-only mode (BIDIMODE=0 and RXONLY=1)
– The sequence begins as soon as SPE=1
– Only the receiver is activated and the received data on the MISO pin are shifted in
serially to the 8-bit shift register and then parallel loaded into the SPI_DR register
(Rx buffer).
• In bidirectional mode, when transmitting (BIDIMODE=1 and BIDIOE=1)
– The sequence begins when data are written into the SPI_DR register (Tx buffer).
during the first bit transmission and then shifted out serially to the MOSI pin.
– No data are received.
• In bidirectional mode, when receiving (BIDIMODE=1 and BIDIOE=0)
– The sequence begins as soon as SPE=1 and BIDIOE=0.
– The received data on the MOSI pin are shifted in serially to the 8-bit shift register
and then parallel loaded into the SPI_DR register (Rx buffer).
– The transmitter is not activated and no data are shifted out serially to the MOSI
pin.
Start sequence in slave mode

• In full-duplex mode (BIDIMODE=0 and RXONLY=0)
– The sequence begins when the slave device receives the clock signal and the first
bit of the data on its MOSI pin. The 7 remaining bits are loaded into the shift
register.
– At the same time, the data are parallel loaded from the Tx buffer into the 8-bit shift
register during the first bit transmission, and then shifted out serially to the MISO
RM0090 Rev 19 885/1751

925
pin. The software must have written the data to be sent before the SPI master
device initiates the transfer.
• In unidirectional receive-only mode (BIDIMODE=0 and RXONLY=1)
bit of the data on its MOSI pin. The 7 remaining bits are loaded into the shift
register.
– The transmitter is not activated and no data are shifted out serially to the MISO
pin.
• In bidirectional mode, when transmitting (BIDIMODE=1 and BIDIOE=1)
bit in the Tx buffer is transmitted on the MISO pin.
during the first bit transmission and then shifted out serially to the MISO pin. The
software must have written the data to be sent before the SPI master device
initiates the transfer.
– No data are received.
• In bidirectional mode, when receiving (BIDIMODE=1 and BIDIOE=0)
bit of the data on its MISO pin.
– The received data on the MISO pin are shifted in serially to the 8-bit shift register
and then parallel loaded into the SPI_DR register (Rx buffer).
– The transmitter is not activated and no data are shifted out serially to the MISO
pin.
Handling data transmission and reception

The TXE flag (Tx buffer empty) is set when the data are transferred from the Tx buffer to the
shift register. It indicates that the internal Tx buffer is ready to be loaded with the next data.
An interrupt can be generated if the TXEIE bit in the SPI_CR2 register is set. Clearing the
TXE bit is performed by writing to the SPI_DR register.
Note: The software must ensure that the TXE flag is set to 1 before attempting to write to the Tx
buffer. Otherwise, it overwrites the data previously written to the Tx buffer.
The RXNE flag (Rx buffer not empty) is set on the last sampling clock edge, when the data
are transferred from the shift register to the Rx buffer. It indicates that data are ready to be
read from the SPI_DR register. An interrupt can be generated if the RXNEIE bit in the
SPI_CR2 register is set. Clearing the RXNE bit is performed by reading the SPI_DR
register.
For some configurations, the BSY flag can be used during the last data transfer to wait until
the completion of the transfer.
Full-duplex transmit and receive procedure in master or slave mode (BIDIMODE=0 and
RXONLY=0)
The software has to follow this procedure to transmit and receive data (see Figure 253 and
Figure 254):
886/1751 RM0090 Rev 19

1. Enable the SPI by setting the SPE bit to 1.

2. Write the first data item to be transmitted into the SPI_DR register (this clears the TXE
flag).
3. Wait until TXE=1 and write the second data item to be transmitted. Then wait until
RXNE=1 and read the SPI_DR to get the first received data item (this clears the RXNE
bit). Repeat this operation for each data item to be transmitted/received until the n–1
received data.
4. Wait until RXNE=1 and read the last received data.
5. Wait until TXE=1 and then wait until BSY=0 before disabling the SPI.
This procedure can also be implemented using dedicated interrupt subroutines launched at
each rising edges of the RXNE or TXE flag.
Figure 253. TXE/RXNE/BSY behavior in Master / full-duplex mode (BIDIMODE=0 and

RXONLY=0) in case of continuous transfers
Example in Master mode with CPOL=1, CPHA=1
SCK
DATA1 = 0xF1 DATA2 = 0xF2 DATA3 = 0xF3

MISO/MOSI (out) b0 b1 b2 b3 b4 b5 b6 b7 b0 b1 b2 b3 b4 b5 b6 b7 b0 b1 b2 b3 b4 b5 b6 b7
set by hardware set by hardware

TXE flag cleared by software cleared by software set by hardware
Tx buffer 0xF1 0xF2 0xF3

(write SPI_DR)
BSY flag set by hardware reset by hardware
DATA 1 = 0xA1 DATA 2 = 0xA2 DATA 3 = 0xA3

MISO/MOSI (in) b0 b1 b2 b3 b4 b5 b6 b7 b0 b1 b2 b3 b4 b5 b6 b7 b0 b1 b2 b3 b4 b5 b6 b7
set by hardware cleared by software

RXNE flag
Rx buffer 0xA1 0xA2 0xA3

(read SPI_DR)
software software waits software waits software waits software waits software waits
writes 0xF1 until TXE=1 and until RXNE=1 until TXE=1 and until RXNE=1 until RXNE=1
into SPI_DR writes 0xF2 into and reads 0xA1 writes 0xF3 into and reads 0xA2 and reads 0xA3
SPI_DR from SPI_DR SPI_DR from SPI_ DR from SPI_DR
ai17343
RM0090 Rev 19 887/1751

925
Figure 254. TXE/RXNE/BSY behavior in Slave / full-duplex mode (BIDIMODE=0,

RXONLY=0) in case of continuous transfers
Example in Slave mode with CPOL=1, CPHA=1
SCK
DATA 1 = 0xF1 DATA 2 = 0xF2 DATA 3 = 0xF3



(write to SPI_DR)
BSY flag set by cleared by software reset by hardware


RXNE flag

(read from SPI_DR)
software software waits software waits software waits software waits software waits
writes 0xF1 until TXE=1 and until RXNE=1 until TXE=1 and until RXNE=1 until RXNE=1
into SPI_DR writes 0xF2 into and reads 0xA1 writes 0xF3 into and reads 0xA2 and reads 0xA3
SPI_DR from SPI_DR SPI_DR from SPI_ DR from SPI_DR
ai17344
Transmit-only procedure (BIDIMODE=0 RXONLY=0)

In this mode, the procedure can be reduced as described below and the BSY bit can be
used to wait until the completion of the transmission (see Figure 255 and Figure 256).
1. Enable the SPI by setting the SPE bit to 1.
2. Write the first data item to send into the SPI_DR register (this clears the TXE bit).
3. Wait until TXE=1 and write the next data item to be transmitted. Repeat this step for
each data item to be transmitted.
4. After writing the last data item into the SPI_DR register, wait until TXE=1, then wait until
BSY=0, this indicates that the transmission of the last data is complete.
This procedure can be also implemented using dedicated interrupt subroutines launched at
each rising edge of the TXE flag.
Note: During discontinuous communications, there is a 2 APB clock period delay between the
write operation to SPI_DR and the BSY bit setting. As a consequence, in transmit-only
mode, it is mandatory to wait first until TXE is set and then until BSY is cleared after writing
the last data.
After transmitting two data items in transmit-only mode, the OVR flag is set in the SPI_SR
register since the received data are never read.
888/1751 RM0090 Rev 19

Figure 255. TXE/BSY behavior in Master transmit-only mode (BIDIMODE=0 and RXONLY=0)
in case of continuous transfers
Example in Master mode with CPOL=1, CPHA=1
SCK



(write to SPI_DR)
software writes software waits software waits

0xF1 into until TXE=1 and until TXE=1 and
software waits until TXE=1 software waits until BSY=0
SPI_DR writes 0xF2 into writes 0xF3 into
SPI_DR SPI_DR
ai17345
Figure 256. TXE/BSY in Slave transmit-only mode (BIDIMODE=0 and RXONLY=0) in case of
continuous transfers
Example in slave mode with CPOL=1, CPHA=1
SCK



(write to SPI_DR)
software writes software waits software waits

0xF1 into until TXE=1 and until TXE=1 and
software waits until TXE=1 software waits until BSY=0
SPI_DR writes 0xF2 into writes 0xF3 into
SPI_DR SPI_DR ai17346
Bidirectional transmit procedure (BIDIMODE=1 and BIDIOE=1)

In this mode, the procedure is similar to the procedure in Transmit-only mode except that
the BIDIMODE and BIDIOE bits both have to be set in the SPI_CR2 register before enabling
the SPI.
Unidirectional receive-only procedure (BIDIMODE=0 and RXONLY=1)

In this mode, the procedure can be reduced as described below (see Figure 257):
RM0090 Rev 19 889/1751

925
1. Set the RXONLY bit in the SPI_CR1 register.

2. Enable the SPI by setting the SPE bit to 1:
a) In master mode, this immediately activates the generation of the SCK clock, and
data are serially received until the SPI is disabled (SPE=0).
b) In slave mode, data are received when the SPI master device drives NSS low and
generates the SCK clock.
3. Wait until RXNE=1 and read the SPI_DR register to get the received data (this clears
the RXNE bit). Repeat this operation for each data item to be received.
This procedure can also be implemented using dedicated interrupt subroutines launched at
each rising edge of the RXNE flag.
Note: If it is required to disable the SPI after the last transfer, follow the recommendation
described in Section 28.3.8.
Figure 257. RXNE behavior in receive-only mode (BIDIRMODE=0 and RXONLY=1)

in case of continuous transfers
Example with CPOL=1, CPHA=1, RXONLY=1
SCK


RXNE flag

(read from SPI_DR)
software waits until RXNE=1 software waits until RXNE=1 software waits until RXNE=1
and reads 0xA1 from SPI_DR and reads 0xA2 from SPI_DR and reads 0xA3 from SPI_DR
ai17347
Bidirectional receive procedure (BIDIMODE=1 and BIDIOE=0)

In this mode, the procedure is similar to the Receive-only mode procedure except that the
BIDIMODE bit has to be set and the BIDIOE bit cleared in the SPI_CR2 register before
enabling the SPI.
Continuous and discontinuous transfers

When transmitting data in master mode, if the software is fast enough to detect each rising
edge of TXE (or TXE interrupt) and to immediately write to the SPI_DR register before the
ongoing data transfer is complete, the communication is said to be continuous. In this case,
there is no discontinuity in the generation of the SPI clock between each data item and the
BSY bit is never cleared between each data transfer.
On the contrary, if the software is not fast enough, this can lead to some discontinuities in
the communication. In this case, the BSY bit is cleared between each data transmission
(see Figure 258).
In Master receive-only mode (RXONLY=1), the communication is always continuous and
the BSY flag is always read at 1.
890/1751 RM0090 Rev 19

In slave mode, the continuity of the communication is decided by the SPI master device. In
any case, even if the communication is continuous, the BSY flag goes low between each
transfer for a minimum duration of one SPI clock cycle (see Figure 256).
Figure 258. TXE/BSY behavior when transmitting (BIDIRMODE=0 and RXONLY=0)

in case of discontinuous transfers
Example with CPOL=1, CPHA=1
SCK

MOSI (out) b0 b1 b2 b3 b4 b5 b6 b7 b0 b1 b2 b3 b4 b5 b6 b7 b0 b1 b2 b3 b4 b5 b6 b7
TXE flag

(write to SPI_DR)
BSY flag
software writes 0xF1 software waits until TXE=1 but is software waits until TXE=1 but software waits software waits until BSY=0
into SPI_DR late to write 0xF2 into SPI_DR is late to write 0xF3 into until TXE=1
SPI_DR
ai17348
28.3.6 CRC calculation

A CRC calculator has been implemented for communication reliability. Separate CRC
calculators are implemented for transmitted data and received data. The CRC is calculated
using a programmable polynomial serially on each bit. It is calculated on the sampling clock
edge defined by the CPHA and CPOL bits in the SPI_CR1 register.
Note: This SPI offers two kinds of CRC calculation standard which depend directly on the data
frame format selected for the transmission and/or reception: 8-bit data (CR8) and 16-bit data
(CRC16).
CRC calculation is enabled by setting the CRCEN bit in the SPI_CR1 register. This action
resets the CRC registers (SPI_RXCRCR and SPI_TXCRCR). In full duplex or transmitter
only mode, when the transfers are managed by the software (CPU mode), it is necessary to
write the bit CRCNEXT immediately after the last data to be transferred is written to the
SPI_DR. At the end of this last data transfer, the SPI_TXCRCR value is transmitted.
In receive only mode and when the transfers are managed by software (CPU mode), it is
necessary to write the CRCNEXT bit after the second last data has been received. The CRC
is received just after the last data reception and the CRC check is then performed.
At the end of data and CRC transfers, the CRCERR flag in the SPI_SR register is set if
corruption occurs during the transfer.
If data are present in the TX buffer, the CRC value is transmitted only after the transmission
of the data byte. During CRC transmission, the CRC calculator is switched off and the
register value remains unchanged.
SPI communication using the CRC is possible through the following procedure:
RM0090 Rev 19 891/1751

925
1. Program the CPOL, CPHA, LSBFirst, BR, SSM, SSI and MSTR values.
2. Program the polynomial in the SPI_CRCPR register.
3. Enable the CRC calculation by setting the CRCEN bit in the SPI_CR1 register. This
also clears the SPI_RXCRCR and SPI_TXCRCR registers.
4. Enable the SPI by setting the SPE bit in the SPI_CR1 register.
5. Start the communication and sustain the communication until all but one byte or half-
word have been transmitted or received.
– In full duplex or transmitter-only mode, when the transfers are managed by
software, when writing the last byte or half word to the Tx buffer, set the
CRCNEXT bit in the SPI_CR1 register to indicate that the CRC will be transmitted
after the transmission of the last byte.
– In receiver only mode, set the bit CRCNEXT just after the reception of the second
to last data to prepare the SPI to enter in CRC Phase at the end of the reception of
the last data. CRC calculation is frozen during the CRC transfer.
6. After the transfer of the last byte or half word, the SPI enters the CRC transfer and
check phase. In full duplex mode or receiver-only mode, the received CRC is
compared to the SPI_RXCRCR value. If the value does not match, the CRCERR flag in
SPI_SR is set and an interrupt can be generated when the ERRIE bit in the SPI_CR2
register is set.
Note: When the SPI is in slave mode, be careful to enable CRC calculation only when the clock is
stable, that is, when the clock is in the steady state. If not, a wrong CRC calculation may be
done. In fact, the CRC is sensitive to the SCK slave input clock as soon as CRCEN is set,
and this, whatever the value of the SPE bit.
With high bitrate frequencies, be careful when transmitting the CRC. As the number of used
CPU cycles has to be as low as possible in the CRC transfer phase, it is forbidden to call
software functions in the CRC transmission sequence to avoid errors in the last data and
CRC reception. In fact, CRCNEXT bit has to be written before the end of the
transmission/reception of the last data.
For high bit rate frequencies, it is advised to use the DMA mode to avoid the degradation of
the SPI speed performance due to CPU accesses impacting the SPI bandwidth.
When the devices are configured as slaves and the NSS hardware mode is used, the NSS
pin needs to be kept low between the data phase and the CRC phase.
When the SPI is configured in slave mode with the CRC feature enabled, CRC calculation
takes place even if a high level is applied on the NSS pin. This may happen for example in
case of a multislave environment where the communication master addresses slaves
alternately.
Between a slave deselection (high level on NSS) and a new slave selection (low level on
NSS), the CRC value should be cleared on both master and slave sides in order to
resynchronize the master and slave for their respective CRC calculation.
To clear the CRC, follow the procedure below:
1. Disable SPI (SPE = 0)
2. Clear the CRCEN bit
3. Set the CRCEN bit
4. Enable the SPI (SPE = 1)
892/1751 RM0090 Rev 19

28.3.7 Status flags

Four status flags are provided for the application to completely monitor the state of the SPI
bus.
Tx buffer empty flag (TXE)

When it is set, this flag indicates that the Tx buffer is empty and the next data to be
transmitted can be loaded into the buffer. The TXE flag is cleared when writing to the
SPI_DR register.
Rx buffer not empty (RXNE)

When set, this flag indicates that there are valid received data in the Rx buffer. It is cleared
when SPI_DR is read.
BUSY flag
This BSY flag is set and cleared by hardware (writing to this flag has no effect). The BSY
flag indicates the state of the communication layer of the SPI.
When BSY is set, it indicates that the SPI is busy communicating. There is one exception in
master mode / bidirectional receive mode (MSTR=1 and BDM=1 and BDOE=0) where the
BSY flag is kept low during reception.
The BSY flag is useful to detect the end of a transfer if the software wants to disable the SPI
and enter Halt mode (or disable the peripheral clock). This avoids corrupting the last
transfer. For this, the procedure described below must be strictly respected.
The BSY flag is also useful to avoid write collisions in a multimaster system.
The BSY flag is set when a transfer starts, with the exception of master mode / bidirectional
receive mode (MSTR=1 and BDM=1 and BDOE=0).
It is cleared:
• when a transfer is finished (except in master mode if the communication is continuous)
• when the SPI is disabled
• when a master mode fault occurs (MODF=1)
When communication is not continuous, the BSY flag is low between each communication.
When communication is continuous:
• in master mode, the BSY flag is kept high during all the transfers
• in slave mode, the BSY flag goes low for one SPI clock cycle between each transfer
Note: Do not use the BSY flag to handle each data transmission or reception. It is better to use the
TXE and RXNE flags instead.
RM0090 Rev 19 893/1751

925
28.3.8 Disabling the SPI

When a transfer is terminated, the application can stop the communication by disabling the
SPI peripheral. This is done by clearing the SPE bit.
For some configurations, disabling the SPI and entering the Halt mode while a transfer is
ongoing can cause the current transfer to be corrupted and/or the BSY flag might become
unreliable.
To avoid any of those effects, it is recommended to respect the following procedure when
disabling the SPI:
In master or slave full-duplex mode (BIDIMODE=0, RXONLY=0)

1. Wait until RXNE=1 to receive the last data
2. Wait until TXE=1
3. Then wait until BSY=0
4. Disable the SPI (SPE=0) and, eventually, enter the Halt mode (or disable the peripheral
clock)
In master or slave unidirectional transmit-only mode (BIDIMODE=0,

RXONLY=0) or bidirectional transmit mode (BIDIMODE=1, BIDIOE=1)
After the last data is written into the SPI_DR register:
1. Wait until TXE=1
2. Then wait until BSY=0
3. Disable the SPI (SPE=0) and, eventually, enter the Halt mode (or disable the peripheral
clock)
In master unidirectional receive-only mode (MSTR=1, BIDIMODE=0,

RXONLY=1) or bidirectional receive mode (MSTR=1, BIDIMODE=1, BIDIOE=0)
This case must be managed in a particular way to ensure that the SPI does not initiate a
new transfer. The sequence below is valid only for SPI Motorola configuration (FRF bit set to
0):
1. Wait for the second to last occurrence of RXNE=1 (n–1)
2. Then wait for one SPI clock cycle (using a software loop) before disabling the SPI
(SPE=0)
3. Then wait for the last RXNE=1 before entering the Halt mode (or disabling the
peripheral clock)
When the SPI is configured in TI mode (Bit FRF set to 1), the following procedure has to be
respected to avoid generating an undesired pulse on NSS when the SPI is disabled:
1. Wait for the second to last occurrence of RXNE = 1 (n-1).
2. Disable the SPI (SPE = 0) in the following window frame using a software loop:
– After at least one SPI clock cycle,
– Before the beginning of the LSB data transfer.
Note: In master bidirectional receive mode (MSTR=1 and BDM=1 and BDOE=0), the BSY flag is
kept low during transfers.
894/1751 RM0090 Rev 19

In slave receive-only mode (MSTR=0, BIDIMODE=0, RXONLY=1) or

bidirectional receive mode (MSTR=0, BIDIMODE=1, BIDOE=0)
1. You can disable the SPI (write SPE=1) at any time: the current transfer will complete
before the SPI is effectively disabled
2. Then, if you want to enter the Halt mode, you must first wait until BSY = 0 before
entering the Halt mode (or disabling the peripheral clock).
28.3.9 SPI communication using DMA (direct memory addressing)

To operate at its maximum speed, the SPI needs to be fed with the data for transmission
and the data received on the Rx buffer should be read to avoid overrun. To facilitate the
transfers, the SPI features a DMA capability implementing a simple request/acknowledge
protocol.
A DMA access is requested when the enable bit in the SPI_CR2 register is enabled.
Separate requests must be issued to the Tx and Rx buffers (see Figure 259 and
Figure 260):
• In transmission, a DMA request is issued each time TXE is set to 1. The DMA then
writes to the SPI_DR register (this clears the TXE flag).
• In reception, a DMA request is issued each time RXNE is set to 1. The DMA then reads
the SPI_DR register (this clears the RXNE flag).
When the SPI is used only to transmit data, it is possible to enable only the SPI Tx DMA
channel. In this case, the OVR flag is set because the data received are not read.
When the SPI is used only to receive data, it is possible to enable only the SPI Rx DMA
channel.
In transmission mode, when the DMA has written all the data to be transmitted (flag TCIF is
set in the DMA_ISR register), the BSY flag can be monitored to ensure that the SPI
communication is complete. This is required to avoid corrupting the last transmission before
disabling the SPI or entering the Stop mode. The software must first wait until TXE=1 and
then until BSY=0.
Note: During discontinuous communications, there is a 2 APB clock period delay between the
write operation to SPI_DR and the BSY bit setting. As a consequence, it is mandatory to
wait first until TXE=1 and then until BSY=0 after writing the last data.
RM0090 Rev 19 895/1751

925
Figure 259. Transmission using DMA

SCK

TXE flag cleared by DMA write clear by DMA write set by hardware
reset
BSY flag set by hardware by hardware
DMA request ignored by the DMA because

DMA transfer is complete

(write to SPI_DR)
DMA writes to SPI_DR
set by hardware clear by software

DMA TCIF flag
(DMA transfer complete)
software configures the DMA writes DMA writes DMA writes DMA transfer is software waits software waits until BSY=0
DMA SPI Tx channel DATA1 into DATA2 into DATA3 into complete (TCIF=1 in until TXE=1
to send 3 data items SPI_DR SPI_DR SPI_DR DMA_ISR)
and enables the SPI
ai17349
Figure 260. Reception using DMA

SCK

set by hardware clear by DMA read

RXNE flag
DMA request
Rx buffer
0xA1 0xA2 0xA3
(read from SPI_DR)
DMA read from SPI_DR
clear
set by hardware
flag DMA TCIF by software
(DMA transfer complete)
software configures the DMA reads DMA reads DMA reads The DMA transfer is
DMA SPI Rx channel DATA1 from DATA2 from DATA3 from complete (TCIF=1 in
to receive 3 data items SPI_DR SPI_DR SPI_DR DMA_ISR)
and enables the SPI
ai17350
896/1751 RM0090 Rev 19

DMA capability with CRC

When SPI communication is enabled with CRC communication and DMA mode, the
transmission and reception of the CRC at the end of communication are automatic that is
without using the bit CRCNEXT. After the CRC reception, the CRC must be read in the
SPI_DR register to clear the RXNE flag.
At the end of data and CRC transfers, the CRCERR flag in SPI_SR is set if corruption
occurs during the transfer.
28.3.10 Error flags

Master mode fault (MODF)
Master mode fault occurs when the master device has its NSS pin pulled low (in NSS
hardware mode) or SSI bit low (in NSS software mode), this automatically sets the MODF
bit. Master mode fault affects the SPI peripheral in the following ways:
• The MODF bit is set and an SPI interrupt is generated if the ERRIE bit is set.
• The SPE bit is cleared. This blocks all output from the device and disables the SPI
interface.
• The MSTR bit is cleared, thus forcing the device into slave mode.
Use the following software sequence to clear the MODF bit:
1. Make a read or write access to the SPI_SR register while the MODF bit is set.
2. Then write to the SPI_CR1 register.
To avoid any multiple slave conflicts in a system comprising several MCUs, the NSS pin
must be pulled high during the MODF bit clearing sequence. The SPE and MSTR bits can
be restored to their original state after this clearing sequence.
As a security, hardware does not allow the setting of the SPE and MSTR bits while the
MODF bit is set.
In a slave device the MODF bit cannot be set. However, in a multimaster configuration, the
device can be in slave mode with this MODF bit set. In this case, the MODF bit indicates
that there might have been a multimaster conflict for system control. An interrupt routine can
be used to recover cleanly from this state by performing a reset or returning to a default
state.
Overrun condition
An overrun condition occurs when the master device has sent data bytes and the slave
device has not cleared the RXNE bit resulting from the previous data byte transmitted.
When an overrun condition occurs:
• the OVR bit is set and an interrupt is generated if the ERRIE bit is set.
In this case, the receiver buffer contents will not be updated with the newly received data
from the master device. A read from the SPI_DR register returns this byte. All other
subsequently transmitted bytes are lost.
Clearing the OVR bit is done by a read from the SPI_DR register followed by a read access
to the SPI_SR register.
RM0090 Rev 19 897/1751

925
CRC error
This flag is used to verify the validity of the value received when the CRCEN bit in the
SPI_CR1 register is set. The CRCERR flag in the SPI_SR register is set if the value
received in the shift register does not match the receiver SPI_RXCRCR value.
TI mode frame format error

A TI mode frame format error is detected when an NSS pulse occurs during an ongoing
communication when the SPI is acting in slave mode and configured to conform to the TI
mode protocol. When this error occurs, the FRE flag is set in the SPI_SR register. The SPI
is not disabled when an error occurs, the NSS pulse is ignored, and the SPI waits for the
next NSS pulse before starting a new transfer. The data may be corrupted since the error
detection may result in the lost of two data bytes.
The FRE flag is cleared when SPI_SR register is read. If the bit ERRIE is set, an interrupt is
generated on the NSS error detection. In this case, the SPI should be disabled because
data consistency is no more guaranteed and communications should be reinitiated by the
master when the slave SPI is re-enabled.
Figure 261. TI mode frame format error detection
NSS
output
trigger sampling trigger sampling trigger sampling trigger sampling trigger sampling trigger sampling trigger sampling
SCK
output
MOSI
DONTCARE MSBIN LSBIN DONTCARE MSBIN LSBIN
input
MISO LSBOUT MSBOUT LSBOUT

1 or 0 MSBOUT
output
TIFRFE
ai18438
28.3.11 SPI interrupts
Table 126. SPI interrupt requests

Transmit buffer empty flag TXE TXEIE

Receive buffer not empty flag RXNE RXNEIE
Master Mode fault event MODF
Overrun error OVR ERRIE
CRC error flag CRCERR
TI frame format error FRE ERRIE
898/1751 RM0090 Rev 19

28.4 I2S functional description
28.4.1 I2S general description

The block diagram of the I2S is shown in Figure 262.
Figure 262. I2S block diagram
Address and data bus
Tx buffer
CH
BSY OVR MODF CRC UDR TxE RxNE FRE
ERR SIDE
16-bit
MOSI/ SD
MISO/
Shift register
I2S2ext_SD/
I2S3ext_SD(1) LSB first Communication
16-bit control
Rx buffer
NSS/WS
I2SCFG I2SSTD CK DATLEN CH

[1:0] [1:0] POL [1:0] LEN
I2S
MOD I2SE
Master control logic

Bidi Bidi CRC CRC Rx
DFF SSM SSI
mode OE EN Next only
SPI LSB
SPE BR2 BR1 BR0 MSTR CPOL CPHA
baud rate generator First
CK
I2S clock generator

I2S_CK
I2SMOD
MCK
MCKOEODD I2SDIV[7:0] I2SxCLK
MS19909V1
1. I2S2ext_SD and I2S3ext_SD are the extended SD pins that control the I2S full duplex mode.
The SPI could function as an audio I2S interface when the I2S capability is enabled
(by
setting the I2SMOD bit in the SPI_I2SCFGR register). This interface uses almost the same
pins, flags and interrupts as the SPI.
RM0090 Rev 19 899/1751

925
The I2S shares three common pins with the SPI:

• SD: Serial Data (mapped on the MOSI pin) to transmit or receive the two time-
multiplexed data channels (in half-duplex mode only).
• WS: Word Select (mapped on the NSS pin) is the data control signal output in master
mode and input in slave mode.
• CK: Serial Clock (mapped on the SCK pin) is the serial clock output in master mode
and serial clock input in slave mode.
• I2S2ext_SD and I2S3ext_SD: additional pins (mapped on the MISO pin) to control the
I2S full duplex mode.
An additional pin could be used when a master clock output is needed for some external
audio devices:
• MCK: Master Clock (mapped separately) is used, when the I2S is configured in master
mode (and when the MCKOE bit in the SPI_I2SPR register is set), to output this
additional clock generated at a preconfigured frequency rate equal to 256 × FS, where
FS is the audio sampling frequency.
The I2S uses its own clock generator to produce the communication clock when it is set in
master mode. This clock generator is also the source of the master clock output. Two
additional registers are available in I2S mode. One is linked to the clock generator
configuration SPI_I2SPR and the other one is a generic I2S configuration register
SPI_I2SCFGR (audio standard, slave/master mode, data format, packet frame, clock
polarity, etc.).
The SPI_CR1 register and all CRC registers are not used in the I2S mode. Likewise, the
SSOE bit in the SPI_CR2 register and the MODF and CRCERR bits in the SPI_SR are not
used.
The I2S uses the same SPI register for data transfer (SPI_DR) in 16-bit wide mode.
28.4.2 I2S full duplex

To support I2S full duplex mode, two extra I2S instances called extended I2Ss (I2S2_ext,
I2S3_ext) are available in addition to I2S2 and I2S3 (see Figure 263). The first I2S full-
duplex interface is consequently based on I2S2 and I2S2_ext, and the second one on I2S3
and I2S3_ext.
Note: I2S2_ext an I2S3_ext are used only in full-duplex mode.
Figure 263. I2S full duplex block diagram
SPI/I2Sx SPIx_MOSI/I2Sx_SD(in/out)
I2Sx_SCK
I2S_ WS
I2Sx_ext I2Sx_extSD(in/out)
MS19910V1
1. Where x can be 2 or 3.
900/1751 RM0090 Rev 19

I2Sx can operate in master mode. As a result:

• Only I2Sx can output SCK and WS in half duplex mode
• Only I2Sx can deliver SCK and WS to I2S2_ext and I2S3_ext in full duplex mode.
The extended I2Ss (I2Sx_ext) can be used only in full duplex mode. The I2Sx_ext operate
always in slave mode.
Both I2Sx and I2Sx_ext can be configured as transmitters or receivers.
28.4.3 Supported audio protocols

The four-line bus has to handle only audio data generally time-multiplexed on two channels:
the right channel and the left channel. However there is only one 16-bit register for the
transmission and the reception. So, it is up to the software to write into the data register the
adequate value corresponding to the considered channel side, or to read the data from the
data register and to identify the corresponding channel by checking the CHSIDE bit in the
SPI_SR register. Channel Left is always sent first followed by the channel right (CHSIDE
has no meaning for the PCM protocol).
Four data and packet frames are available. Data may be sent with a format of:
• 16-bit data packed in 16-bit frame
When using 16-bit data extended on 32-bit packet, the first 16 bits (MSB) are the significant
bits, the 16-bit LSB is forced to 0 without any need for software action or DMA request (only
one read/write operation).
The 24-bit and 32-bit data frames need two CPU read or write operations to/from the
SPI_DR or two DMA operations if the DMA is preferred for the application. For 24-bit data
frame specifically, the 8 nonsignificant bits are extended to 32 bits with 0-bits (by hardware).
For all data formats and communication standards, the most significant bit is always sent
first (MSB first).
The I2S interface supports four audio standards, configurable using the I2SSTD[1:0] and
PCMSYNC bits in the SPI_I2SCFGR register.
I2S Philips standard

For this standard, the WS signal is used to indicate which channel is being transmitted. It is
activated one CK clock cycle before the first bit (MSB) is available.
RM0090 Rev 19 901/1751

925
Figure 264. I2S Philips protocol waveforms (16/32-bit full accuracy, CPOL = 0)
CK
WS transmission reception
Can be 16-bit or 32-bit
SD MSB LSB MSB
Channel left
Channel
right
MS19591V1
Data are latched on the falling edge of CK (for the transmitter) and are read on the rising
edge (for the receiver). The WS signal is also latched on the falling edge of CK.
Figure 265. I2S Philips standard waveforms (24-bit frame with CPOL = 0)
CK
WS Transmission Reception
24-bit data 8-bit remaining 0 forced

SD
MSB LSB
Channel left 32-bit

Channel right
MS19592V1
This mode needs two write or read operations to/from the SPI_DR.
• In transmission mode:
if 0x8EAA33 has to be sent (24-bit):
Figure 266. Transmitting 0x8EAA33
First write to Data register Second write to Data register

0x8EAA 0x33XX
Only the 8 MSB are sent
to compare the 24 bits
8 LSBs have no meaning
and can be anything
MS19593V1
• In reception mode:
if data 0x8EAA33 is received:
902/1751 RM0090 Rev 19

Figure 267. Receiving 0x8EAA33
First read to Data register Second read to Data register

0x8EAA 0x33XX
Only the 8 MSB are sent
to compare the 24 bits
8 LSBs have no meaning
and can be anything
MS19594V1
Figure 268. I2S Philips standard (16-bit extended to 32-bit packet frame with
CPOL = 0)
CK
16-bit data 16-bit remaining 0 forced

SD
MSB LSB
Channel left 32-bit

Channel right
MS19599V1
When 16-bit data frame extended to 32-bit channel frame is selected during the I2S
configuration phase, only one access to SPI_DR is required. The 16 remaining bits are
forced by hardware to 0x0000 to extend the data to 32-bit format.
If the data to transmit or the received data are 0x76A3 (0x76A30000 extended to 32-bit), the
operation shown in Figure 269 is required.
Figure 269. Example
Only one access to SPIx_DR
0x76A3
MS19595V1
For transmission, each time an MSB is written to SPI_DR, the TXE flag is set and its
interrupt, if allowed, is generated to load SPI_DR with the new value to send. This takes
place even if 0x0000 have not yet been sent because it is done by hardware.
For reception, the RXNE flag is set and its interrupt, if allowed, is generated when the first
16 MSB half-word is received.
In this way, more time is provided between two write or read operations, which prevents
underrun or overrun conditions (depending on the direction of the data transfer).
MSB justified standard

For this standard, the WS signal is generated at the same time as the first data bit, which is
the MSBit.
RM0090 Rev 19 903/1751

925
Figure 270. MSB justified 16-bit or 32-bit full-accuracy length with CPOL = 0
CK
16- or 32 bit data

SD
MSB LSB MSB
Channel left
Channel right
MS30100 V1
Data are latched on the falling edge of CK (for transmitter) and are read on the rising edge
(for the receiver).
Figure 271. MSB justified 24-bit frame length with CPOL = 0
CK
24 bit data 8-bit remaining

SD 0 forced
MSB LSB
Channel left 32-bit
Channel right
MS30101V1
Figure 272. MSB justified 16-bit extended to 32-bit packet frame with CPOL = 0
CK
16-bit data 16-bit remaining

SD 0 forced
MSB LSB
Channel left 32-bit
Channel right
MS30102V1
LSB justified standard

This standard is similar to the MSB justified standard (no difference for the 16-bit and 32-bit
full-accuracy frame formats).
904/1751 RM0090 Rev 19

Figure 273. LSB justified 16-bit or 32-bit full-accuracy with CPOL = 0
CK
WS
Transmission Reception
16- or 32-bit data
SD
MSB LSB MSB
Channel left
Channel right
MS30103V1
Figure 274. LSB justified 24-bit frame length with CPOL = 0
CK
WS Reception
Transmission
SD 0 forced
MSB LSB
Channel left 32-bit

Channel right
MS30104V1
• In transmission mode:
If data 0x3478AE have to be transmitted, two write operations to the SPI_DR register
are required from software or by DMA. The operations are shown below.
Figure 275. Operations required to transmit 0x3478AE
First write to Data register Second write to Data register

conditioned by TXE=1 conditioned by TXE=1
0xXX34 0x78AE
Only the 8 LSB of the

half-word are significant.
A field of 0x00 is forced
instead of the 8 MSBs. MS19596V1
• In reception mode:
If data 0x3478AE are received, two successive read operations from SPI_DR are
required on each RXNE event.
RM0090 Rev 19 905/1751

925
Figure 276. Operations required to receive 0x3478AE

First read from Data register Second read from Data register
conditioned by RXNE=1 conditioned by RXNE=1
0xXX34 0x78AE
Only the 8 LSB of the

half-word are significant.
A field of 0x00 is forced
instead of the 8 MSBs.
MS19597V1
Figure 277. LSB justified 16-bit extended to 32-bit packet frame with CPOL = 0
CK
Reception
WS
Transmission
SD 0 forced
MSB LSB
Channel left 32-bit

Channel right
MS30105V1
When 16-bit data frame extended to 32-bit channel frame is selected during the I2S
configuration phase, Only one access to SPI_DR is required. The 16 remaining bits are
forced by hardware to 0x0000 to extend the data to 32-bit format. In this case it corresponds
to the half-word MSB.
If the data to transmit or the received data are 0x76A3 (0x0000 76A3 extended to 32-bit),
the operation shown in Figure 278 is required.
Figure 278. Example of LSB justified 16-bit extended to 32-bit packet frame
Only one access to the SPIx-DR register
0x76A3
MS19598V1
In transmission mode, when TXE is asserted, the application has to write the data to be
transmitted (in this case 0x76A3). The 0x000 field is transmitted first (extension on 32-bit).
TXE is asserted again as soon as the effective data (0x76A3) is sent on SD.
In reception mode, RXNE is asserted as soon as the significant half-word is received (and
not the 0x0000 field).
In this way, more time is provided between two write or read operations to prevent underrun
or overrun conditions.
906/1751 RM0090 Rev 19

PCM standard
For the PCM standard, there is no need to use channel-side information. The two PCM
modes (short and long frame) are available and configurable using the PCMSYNC bit in
SPI_I2SCFGR.
Figure 279. PCM standard waveforms (16-bit)
CK
WS
short frame
13-bits
WS
long frame
SD MSB LSB MSB
MS30106V1
For long frame synchronization, the WS signal assertion time is fixed 13 bits in master
mode.
For short frame synchronization, the WS synchronization signal is only one cycle long.
Figure 280. PCM standard waveforms (16-bit extended to 32-bit packet frame)
CK
WS
short frame
Up to 13-bits
WS
long frame
16 bits
SD MSB LSB
MS30107V1
Note: For both modes (master and slave) and for both synchronizations (short and long), the
number of bits between two consecutive pieces of data (and so two synchronization signals)
needs to be specified (DATLEN and CHLEN bits in the SPI_I2SCFGR register) even in
slave mode.
28.4.4 Clock generator

The I2S bitrate determines the dataflow on the I2S data line and the I2S clock signal
frequency.
I2S bitrate = number of bits per channel × number of channels × sampling audio frequency
For a 16-bit audio, left and right channel, the I2S bitrate is calculated as follows:
I2S bitrate = 16 × 2 × FS
It will be: I2S bitrate = 32 x 2 x FS if the packet length is 32-bit wide.
RM0090 Rev 19 907/1751

925
Figure 281. Audio sampling frequency definition
16-or 32-bit left 16-or 32-bit

channel right channel
32- or 64-bits
FS
sampling point sampling point
FS : audio sampling frequency
MS30108V1
When the master mode is configured, a specific action needs to be taken to properly
program the linear divider in order to communicate with the desired audio frequency.
Figure 282. I2S clock generator architecture
MCK
I²SxCLK 8-bit linear divider

+ reshaping stage 0
Div2 0
Divider by 4 1 CK
1
MCKOE
I²SMOD
CHLEN
MCKOE ODD I²SDIV[7:0]
MS30109V1
1. Where x could be 2 or 3.
Figure 281 presents the communication clock architecture. To achieve high-quality audio
performance, the I2SxCLK clock source can be either the PLLI2S output (through R division
factor) or an external clock (mapped to I2S_CKIN pin).
The audio sampling frequency can be 192 kHz, 96 kHz, or 48 kHz. In order to reach the
desired frequency, the linear divider needs to be programmed according to the formulas
below:
When the master clock is generated (MCKOE in the SPI_I2SPR register is set):
FS = I2SxCLK / [(16*2)*((2*I2SDIV)+ODD)*8)] when the channel frame is 16-bit wide
FS = I2SxCLK / [(32*2)*((2*I2SDIV)+ODD)*4)] when the channel frame is 32-bit wide
When the master clock is disabled (MCKOE bit cleared):
FS = I2SxCLK / [(16*2)*((2*I2SDIV)+ODD))] when the channel frame is 16-bit wide
FS = I2SxCLK / [(32*2)*((2*I2SDIV)+ODD))] when the channel frame is 32-bit wide
Table 127 provides example precision values for different clock configurations.
Note: Other configurations are possible that allow optimum clock precision.
908/1751 RM0090 Rev 19

Table 127. Audio frequency precision (for PLLM VCO = 1 MHz or 2 MHz)(1)
Master Target fS Data
PLLI2SN PLLI2SR I2SDIV I2SODD Real fS (Hz) Error
clock (Hz) format
16-bit 192 2 187 1 8000 0.0000%

8000
32-bit 192 3 62 1 8000 0.0000%
16-bit 192 3 62 1 16000 0.0000%
16000
32-bit 256 2 62 1 16000 0.0000%
16-bit 256 2 62 1 32000 0.0000%
32000
32-bit 256 5 12 1 32000 0.0000%
16-bit 192 5 12 1 48000 0.0000%
48000
32-bit 384 5 12 1 48000 0.0000%
Disabled
16-bit 384 5 12 1 96000 0.0000%
96000
32-bit 424 3 11 1 96014.49219 0.0151%
16-bit 290 3 68 1 22049.87695 0.0006%
22050
32-bit 302 2 53 1 22050.23438 0.0011%
16-bit 302 2 53 1 44100.46875 0.0011%
44100
32-bit 429 4 19 0 44099.50781 0.0011%
16-bit 424 3 11 1 192028.9844 0.0151%
192000
32-bit 258 3 3 1 191964.2813 0.0186%
8000 don't care 256 5 12 1 8000 0.0000%
16000 don't care 213 2 13 0 16000.60059 0.0038%
32000 don't care 213 2 6 1 32001.20117 0.0038%
Enabled 48000 don't care 258 3 3 1 47991.07031 0.0186%
96000 don't care 344 2 3 1 95982.14063 0.0186%
22050 don't care 429 4 9 1 22049.75391 0.0011%
44100 don't care 271 2 6 0 44108.07422 0.0183%
1. This table gives only example values for different clock configurations. Other configurations allowing optimum clock
precision are possible.
28.4.5 I2S master mode

The I2S can be configured as follows:
• In master mode for transmission or reception (half-duplex mode using I2Sx)
• In master mode transmission and reception (full duplex mode using I2Sx and
I2Sx_ext).
This means that the serial clock is generated on the CK pin as well as the Word Select
signal WS. Master clock (MCK) may be output or not, thanks to the MCKOE bit in the
SPI_I2SPR register.
RM0090 Rev 19 909/1751

925
Procedure
1. Select the I2SDIV[7:0] bits in the SPI_I2SPR register to define the serial clock baud
rate to reach the proper audio sample frequency. The ODD bit in the SPI_I2SPR
register also has to be defined.
2. Select the CKPOL bit to define the steady level for the communication clock. Set the
MCKOE bit in the SPI_I2SPR register if the master clock MCK needs to be provided to
the external DAC/ADC audio component (the I2SDIV and ODD values should be
computed depending on the state of the MCK output, for more details refer to
Section 28.4.4: Clock generator).
3. Set the I2SMOD bit in SPI_I2SCFGR to activate the I2S functionalities and choose the
I2S standard through the I2SSTD[1:0] and PCMSYNC bits, the data length through the
DATLEN[1:0] bits and the number of bits per channel by configuring the CHLEN bit.
Select also the I2S master mode and direction (Transmitter or Receiver) through the
I2SCFG[1:0] bits in the SPI_I2SCFGR register.
4. If needed, select all the potential interruption sources and the DMA capabilities by
writing the SPI_CR2 register.
5. The I2SE bit in SPI_I2SCFGR register must be set.
WS and CK are configured in output mode. MCK is also an output, if the MCKOE bit in
SPI_I2SPR is set.
Transmission sequence
The transmission sequence begins when a half-word is written into the Tx buffer.
Assumedly, the first data written into the Tx buffer correspond to the channel Left data.
When data are transferred from the Tx buffer to the shift register, TXE is set and data
corresponding to the channel Right have to be written into the Tx buffer. The CHSIDE flag
indicates which channel is to be transmitted. It has a meaning when the TXE flag is set
because the CHSIDE flag is updated when TXE goes high.
A full frame has to be considered as a Left channel data transmission followed by a Right
channel data transmission. It is not possible to have a partial frame where only the left
channel is sent.
The data half-word is parallel loaded into the 16-bit shift register during the first bit
transmission, and then shifted out, serially, to the MOSI/SD pin, MSB first. The TXE flag is
set after each transfer from the Tx buffer to the shift register and an interrupt is generated if
the TXEIE bit in the SPI_CR2 register is set.
For more details about the write operations depending on the I2S standard mode selected,
refer to Section 28.4.3: Supported audio protocols).
To ensure a continuous audio data transmission, it is mandatory to write the SPI_DR with
the next data to transmit before the end of the current transmission.
To switch off the I2S, by clearing I2SE, it is mandatory to wait for TXE = 1 and BSY = 0.
Reception sequence
The operating mode is the same as for the transmission mode except for the point 3 (refer to
the procedure described in Section 28.4.5: I2S master mode), where the configuration
should set the master reception mode through the I2SCFG[1:0] bits.
Whatever the data or channel length, the audio data are received by 16-bit packets. This
means that each time the Rx buffer is full, the RXNE flag is set and an interrupt is generated
910/1751 RM0090 Rev 19

if the RXNEIE bit is set in SPI_CR2 register. Depending on the data and channel length
configuration, the audio value received for a right or left channel may result from one or two
receptions into the Rx buffer.
CHSIDE is updated after each reception. It is sensitive to the WS signal generated by the
I2S cell.
For more details about the read operations depending on the I2S standard mode selected,
refer to Section 28.4.3: Supported audio protocols.
If data are received while the previously received data have not been read yet, an overrun is
generated and the OVR flag is set. If the ERRIE bit is set in the SPI_CR2 register, an
interrupt is generated to indicate the error.
To switch off the I2S, specific actions are required to ensure that the I2S completes the
transfer cycle properly without initiating a new data transfer. The sequence depends on the
configuration of the data and channel lengths, and on the audio protocol mode selected. In
the case of:
• 16-bit data length extended on 32-bit channel length (DATLEN = 00 and CHLEN = 1)
using the LSB justified mode (I2SSTD = 10)
a) Wait for the second to last RXNE = 1 (n – 1)
b) Then wait 17 I2S clock cycles (using a software loop)
c) Disable the I2S (I2SE = 0)
• 16-bit data length extended on 32-bit channel length (DATLEN = 00 and CHLEN = 1) in
MSB justified, I2S or PCM modes (I2SSTD = 00, I2SSTD = 01 or I2SSTD = 11,
respectively)
a) Wait for the last RXNE
b) Then wait 1 I2S clock cycle (using a software loop)
• For all other combinations of DATLEN and CHLEN, whatever the audio mode selected
through the I2SSTD bits, carry out the following sequence to switch off the I2S:
a) Wait for the second to last RXNE = 1 (n – 1)
b) Then wait one I2S clock cycle (using a software loop)
Note: The BSY flag is kept low during transfers.
28.4.6 I2S slave mode

The I2S can be configured as follows:
• In slave mode for transmission or reception (half-duplex mode using I2Sx)
• In slave mode transmission and reception (full duplex mode using I2Sx and I2Sx_ext).
The operating mode is following mainly the same rules as described for the I2S master
configuration. In slave mode, there is no clock to be generated by the I2S interface. The
clock and WS signals are input from the external master connected to the I2S interface.
There is then no need, for the user, to configure the clock.
The configuration steps to follow are listed below:
RM0090 Rev 19 911/1751

925
1. Set the I2SMOD bit in the SPI_I2SCFGR register to reach the I2S functionalities and
choose the I2S standard through the I2SSTD[1:0] bits, the data length through the
DATLEN[1:0] bits and the number of bits per channel for the frame configuring the
CHLEN bit. Select also the mode (transmission or reception) for the slave through the
I2SCFG[1:0] bits in SPI_I2SCFGR register.
2. If needed, select all the potential interrupt sources and the DMA capabilities by writing
the SPI_CR2 register.
3. The I2SE bit in SPI_I2SCFGR register must be set.
Transmission sequence
The transmission sequence begins when the external master device sends the clock and
when the NSS_WS signal requests the transfer of data. The slave has to be enabled before
the external master starts the communication. The I2S data register has to be loaded before
the master initiates the communication.
For the I2S, MSB justified and LSB justified modes, the first data item to be written into the
data register corresponds to the data for the left channel. When the communication starts,
the data are transferred from the Tx buffer to the shift register. The TXE flag is then set in
order to request the right channel data to be written into the I2S data register.
The CHSIDE flag indicates which channel is to be transmitted. Compared to the master
transmission mode, in slave mode, CHSIDE is sensitive to the WS signal coming from the
external master. This means that the slave needs to be ready to transmit the first data
before the clock is generated by the master. WS assertion corresponds to left channel
transmitted first.
Note: The I2SE has to be written at least two PCLK cycles before the first clock of the master
comes on the CK line.
The data half-word is parallel-loaded into the 16-bit shift register (from the internal bus)
during the first bit transmission, and then shifted out serially to the MOSI/SD pin MSB first.
The TXE flag is set after each transfer from the Tx buffer to the shift register and an interrupt
is generated if the TXEIE bit in the SPI_CR2 register is set.
Note that the TXE flag should be checked to be at 1 before attempting to write the Tx buffer.
For more details about the write operations depending on the I2S standard mode selected,
refer to Section 28.4.3: Supported audio protocols.
To secure a continuous audio data transmission, it is mandatory to write the SPI_DR
register with the next data to transmit before the end of the current transmission. An
underrun flag is set and an interrupt may be generated if the data are not written into the
SPI_DR register before the first clock edge of the next data communication. This indicates
to the software that the transferred data are wrong. If the ERRIE bit is set into the SPI_CR2
register, an interrupt is generated when the UDR flag in the SPI_SR register goes high. In
this case, it is mandatory to switch off the I2S and to restart a data transfer starting from the
left channel.
To switch off the I2S, by clearing the I2SE bit, it is mandatory to wait for TXE = 1 and
BSY = 0.
Reception sequence
The operating mode is the same as for the transmission mode except for the point 1 (refer to
the procedure described in Section 28.4.6: I2S slave mode), where the configuration should
set the master reception mode using the I2SCFG[1:0] bits in the SPI_I2SCFGR register.
912/1751 RM0090 Rev 19

Whatever the data length or the channel length, the audio data are received by 16-bit
packets. This means that each time the RX buffer is full, the RXNE flag in the SPI_SR
register is set and an interrupt is generated if the RXNEIE bit is set in the SPI_CR2 register.
Depending on the data length and channel length configuration, the audio value received for
a right or left channel may result from one or two receptions into the RX buffer.
The CHSIDE flag is updated each time data are received to be read from SPI_DR. It is
sensitive to the external WS line managed by the external master component.
For more details about the read operations depending the I2S standard mode selected, refer
to Section 28.4.3: Supported audio protocols.
If data are received while the precedent received data have not yet been read, an overrun is
generated and the OVR flag is set. If the bit ERRIE is set in the SPI_CR2 register, an
interrupt is generated to indicate the error.
To switch off the I2S in reception mode, I2SE has to be cleared immediately after receiving
the last RXNE = 1.
Note: The external master components should have the capability of sending/receiving data in 16-
bit or 32-bit packets via an audio channel.
28.4.7 Status flags

Three status flags are provided for the application to fully monitor the state of the I2S bus.
Busy flag (BSY)

The BSY flag is set and cleared by hardware (writing to this flag has no effect). It indicates
the state of the communication layer of the I2S.
When BSY is set, it indicates that the I2S is busy communicating. There is one exception in
master receive mode (I2SCFG = 11) where the BSY flag is kept low during reception.
The BSY flag is useful to detect the end of a transfer if the software needs to disable the I2S.
This avoids corrupting the last transfer. For this, the procedure described below must be
strictly respected.
The BSY flag is set when a transfer starts, except when the I2S is in master receiver mode.
The BSY flag is cleared:
• when a transfer completes (except in master transmit mode, in which the
communication is supposed to be continuous)
• when the I2S is disabled
When communication is continuous:
• In master transmit mode, the BSY flag is kept high during all the transfers
• In slave mode, the BSY flag goes low for one I2S clock cycle between each transfer
Note: Do not use the BSY flag to handle each data transmission or reception. It is better to use the
TXE and RXNE flags instead.
RM0090 Rev 19 913/1751

925
Tx buffer empty flag (TXE)

When set, this flag indicates that the Tx buffer is empty and the next data to be transmitted
can then be loaded into it. The TXE flag is reset when the Tx buffer already contains data to
be transmitted. It is also reset when the I2S is disabled (I2SE bit is reset).
RX buffer not empty (RXNE)

When set, this flag indicates that there are valid received data in the RX Buffer. It is reset
when SPI_DR register is read.
Channel Side flag (CHSIDE)

In transmission mode, this flag is refreshed when TXE goes high. It indicates the channel
side to which the data to transfer on SD has to belong. In case of an underrun error event in
slave transmission mode, this flag is not reliable and I2S needs to be switched off and
switched on before resuming the communication.
In reception mode, this flag is refreshed when data are received into SPI_DR. It indicates
from which channel side data have been received. Note that in case of error (like OVR) this
flag becomes meaningless and the I2S should be reset by disabling and then enabling it
(with configuration if it needs changing).
This flag has no meaning in the PCM standard (for both Short and Long frame modes).
When the OVR or UDR flag in the SPI_SR is set and the ERRIE bit in SPI_CR2 is also set,
an interrupt is generated. This interrupt can be cleared by reading the SPI_SR status
register (once the interrupt source has been cleared).
28.4.8 Error flags

There are three error flags for the I2S cell.
Underrun flag (UDR)

In slave transmission mode this flag is set when the first clock for data transmission appears
while the software has not yet loaded any value into SPI_DR. It is available when the
I2SMOD bit in SPI_I2SCFGR is set. An interrupt may be generated if the ERRIE bit in
SPI_CR2 is set.
The UDR bit is cleared by a read operation on the SPI_SR register.
Overrun flag (OVR)

This flag is set when data are received and the previous data have not yet been read from
SPI_DR. As a result, the incoming data are lost. An interrupt may be generated if the ERRIE
bit is set in SPI_CR2.
In this case, the receive buffer contents are not updated with the newly received data from
the transmitter device. A read operation to the SPI_DR register returns the previous
correctly received data. All other subsequently transmitted half-words are lost.
Clearing the OVR bit is done by a read operation on the SPI_DR register followed by a read
access to the SPI_SR register.
Frame error flag (FRE)

This flag can be set by hardware only if the I2S is configured in Slave mode. It is set if the
external master is changing the WS line at a moment when the slave is not expected this
914/1751 RM0090 Rev 19

change. If the synchronization is lost, to recover from this state and resynchronize the
external master device with the I2S slave device, follow the steps below:
1. Disable the I2S
2. Re-enable it when the correct level is detected on the WS line (WS line is high in I2S
mode, or low for MSB- or LSB-justified or PCM modes).
Desynchronization between the master and slave device may be due to noisy environment
on the SCK communication clock or on the WS frame synchronization line. An error interrupt
can be generated if the ERRIE bit is set. The desynchronization flag (FRE) is cleared by
software when the status register is read.
28.4.9 I2S interrupts

Table 128 provides the list of I2S interrupts.
Table 128. I2S interrupt requests

Transmit buffer empty flag TXE TXEIE

Receive buffer not empty flag RXNE RXNEIE
Overrun error OVR
ERRIE
Underrun error UDR
Frame error flag FRE ERRIE
28.4.10 DMA features

DMA is working in exactly the same way as for the SPI mode. There is no difference on the
I2S. Only the CRC feature is not available in I2S mode since there is no data transfer
protection system.
RM0090 Rev 19 915/1751

925
28.5 SPI and I2S registers

28.5.1 SPI control register 1 (SPI_CR1) (not used in I2S mode)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
BIDI BIDI CRC CRC RX LSB
DFF SSM SSI SPE BR [2:0] MSTR CPOL CPHA
MODE OE EN NEXT ONLY FIRST
Bit 15 BIDIMODE: Bidirectional data mode enable

0: 2-line unidirectional data mode selected
1: 1-line bidirectional data mode selected
Note: This bit is not used in I2S mode
Bit 14 BIDIOE: Output enable in bidirectional mode
This bit combined with the BIDImode bit selects the direction of transfer in bidirectional mode
0: Output disabled (receive-only mode)
1: Output enabled (transmit-only mode)
Note: This bit is not used in I2S mode.
In master mode, the MOSI pin is used while the MISO pin is used in slave mode.
Bit 13 CRCEN: Hardware CRC calculation enable
0: CRC calculation disabled
1: CRC calculation enabled
Note: This bit should be written only when SPI is disabled (SPE = ‘0’) for correct operation.
It is not used in I2S mode.
Bit 12 CRCNEXT: CRC transfer next
0: Data phase (no CRC phase)
1: Next transfer is CRC (CRC phase)
Note: When the SPI is configured in full duplex or transmitter only modes, CRCNEXT must be
written as soon as the last data is written to the SPI_DR register.
When the SPI is configured in receiver only mode, CRCNEXT must be set after the
second last data reception.
This bit should be kept cleared when the transfers are managed by DMA.
Bit 11 DFF: Data frame format
0: 8-bit data frame format is selected for transmission/reception
1: 16-bit data frame format is selected for transmission/reception
Note: This bit should be written only when SPI is disabled (SPE = ‘0’) for correct operation.
916/1751 RM0090 Rev 19

Bit 10 RXONLY: Receive only

This bit combined with the BIDImode bit selects the direction of transfer in 2-line
unidirectional mode. This bit is also useful in a multislave system in which this particular
slave is not accessed, the output from the accessed slave is not corrupted.
0: Full duplex (Transmit and receive)
1: Output disabled (Receive-only mode)
Bit 9 SSM: Software slave management
When the SSM bit is set, the NSS pin input is replaced with the value from the SSI bit.
0: Software slave management disabled
1: Software slave management enabled
Note: This bit is not used in I2S mode and SPI TI mode
Bit 8 SSI: Internal slave select
This bit has an effect only when the SSM bit is set. The value of this bit is forced onto the
NSS pin and the IO value of the NSS pin is ignored.
Note: This bit is not used in I2S mode and SPI TI mode
Bit 7 LSBFIRST: Frame format
0: MSB transmitted first
1: LSB transmitted first
Note: This bit should not be changed when communication is ongoing.
It is not used in I2S mode and SPI TI mode
Bit 6 SPE: SPI enable
0: Peripheral disabled
1: Peripheral enabled
When disabling the SPI, follow the procedure described in Section 28.3.8.
Bits 5:3 BR[2:0]: Baud rate control
000: fPCLK/2
001: fPCLK/4
010: fPCLK/8
011: fPCLK/16
100: fPCLK/32
101: fPCLK/64
110: fPCLK/128
111: fPCLK/256
Note: These bits should not be changed when communication is ongoing.
They are not used in I2S mode.
Bit 2 MSTR: Master selection
0: Slave configuration
1: Master configuration
RM0090 Rev 19 917/1751

925
Bit1 CPOL: Clock polarity

0: CK to 0 when idle
1: CK to 1 when idle
It is not used in I2S mode and SPI TI mode.
Bit 0 CPHA: Clock phase
0: The first clock transition is the first data capture edge
1: The second clock transition is the first data capture edge
It is not used in I2S mode and SPI TI mode.
28.5.2 SPI control register 2 (SPI_CR2)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TXEIE RXNEIE ERRIE FRF SSOE TXDMAEN RXDMAEN

Reserved Res.

Bit 7 TXEIE: Tx buffer empty interrupt enable
0: TXE interrupt masked
1: TXE interrupt not masked. Used to generate an interrupt request when the TXE flag is set.
Bit 6 RXNEIE: RX buffer not empty interrupt enable
0: RXNE interrupt masked
1: RXNE interrupt not masked. Used to generate an interrupt request when the RXNE flag is
set.
This bit controls the generation of an interrupt when an error condition occurs )(CRCERR,
OVR, MODF in SPI mode, FRE in TI mode and UDR, OVR, and FRE in I2S mode).
0: Error interrupt is masked
1: Error interrupt is enabled
Bit 4 FRF: Frame format
0: SPI Motorola mode
1 SPI TI mode
Bit 3 Reserved. Forced to 0 by hardware.
Bit 2 SSOE: SS output enable
0: SS output is disabled in master mode and the cell can work in multimaster configuration
1: SS output is enabled in master mode and when the cell is enabled. The cell cannot work
in a multimaster environment.
Note: This bit is not used in I2S mode and SPI TI mode.
918/1751 RM0090 Rev 19

Bit 1 TXDMAEN: Tx buffer DMA enable

When this bit is set, the DMA request is made whenever the TXE flag is set.
0: Tx buffer DMA disabled
1: Tx buffer DMA enabled
Bit 0 RXDMAEN: Rx buffer DMA enable
When this bit is set, the DMA request is made whenever the RXNE flag is set.
0: Rx buffer DMA disabled
1: Rx buffer DMA enabled
28.5.3 SPI status register (SPI_SR)

Reset value: 0x0002
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CRC
FRE BSY OVR MODF UDR CHSIDE TXE RXNE
Reserved ERR
r r r r rc_w0 r r r r
Bits 15:9 Reserved. Forced to 0 by hardware.

Bit 8 FRE: Frame format error
0: No frame format error
1: A frame format error occurred
This flag is set by hardware and cleared by software when the SPIx_SR register is read.
Note: This flag is used when the SPI operates in TI slave mode or I2S slave mode (refer to
Section 28.3.10).
Bit 7 BSY: Busy flag
0: SPI (or I2S) not busy
1: SPI (or I2S) is busy in communication or Tx buffer is not empty
This flag is set and cleared by hardware.
Note: BSY flag must be used with caution: refer to Section 28.3.7 and Section 28.3.8.
Bit 6 OVR: Overrun flag
0: No overrun occurred
1: Overrun occurred
This flag is set by hardware and reset by a software sequence. Refer to Section 28.4.8:
Error flags for the software sequence.
Bit 5 MODF: Mode fault
0: No mode fault occurred
1: Mode fault occurred
Bit 4 CRCERR: CRC error flag
0: CRC value received matches the SPI_RXCRCR value
1: CRC value received does not match the SPI_RXCRCR value
This flag is set by hardware and cleared by software writing 0.
RM0090 Rev 19 919/1751

925
Bit 3 UDR: Underrun flag

0: No underrun occurred
1: Underrun occurred
Note: This bit is not used in SPI mode.
Bit 2 CHSIDE: Channel side
0: Channel Left has to be transmitted or has been received
1: Channel Right has to be transmitted or has been received
Note: This bit is not used for SPI mode and is meaningless in PCM mode.
Bit 1 TXE: Transmit buffer empty
0: Tx buffer not empty
1: Tx buffer empty
Bit 0 RXNE: Receive buffer not empty
0: Rx buffer empty
1: Rx buffer not empty
28.5.4 SPI data register (SPI_DR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DR[15:0]
Bits 15:0 DR[15:0]: Data register

Data received or to be transmitted.
The data register is split into 2 buffers - one for writing (Transmit Buffer) and another one for
reading (Receive buffer). A write to the data register will write into the Tx buffer and a read
from the data register will return the value held in the Rx buffer.
Note: These notes apply to SPI mode:
Depending on the data frame format selection bit (DFF in SPI_CR1 register), the data
sent or received is either 8-bit or 16-bit. This selection has to be made before enabling
the SPI to ensure correct operation.
For an 8-bit data frame, the buffers are 8-bit and only the LSB of the register
(SPI_DR[7:0]) is used for transmission/reception. When in reception mode, the MSB of
the register (SPI_DR[15:8]) is forced to 0.
For a 16-bit data frame, the buffers are 16-bit and the entire register, SPI_DR[15:0] is
used for transmission/reception.
920/1751 RM0090 Rev 19

28.5.5 SPI CRC polynomial register (SPI_CRCPR) (not used in I2S

mode)
Reset value: 0x0007
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CRCPOLY[15:0]
Bits 15:0 CRCPOLY[15:0]: CRC polynomial register

This register contains the polynomial for the CRC calculation.
The CRC polynomial (0007h) is the reset value of this register. Another polynomial can be
configured as required.
Note: These bits are not used for the I2S mode.
28.5.6 SPI RX CRC register (SPI_RXCRCR) (not used in I2S mode)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RXCRC[15:0]
Bits 15:0 RXCRC[15:0]: Rx CRC register

When CRC calculation is enabled, the RxCRC[15:0] bits contain the computed CRC value of
the subsequently received bytes. This register is reset when the CRCEN bit in SPI_CR1
register is written to 1. The CRC is calculated serially using the polynomial programmed in
the SPI_CRCPR register.
Only the 8 LSB bits are considered when the data frame format is set to be 8-bit data (DFF
bit of SPI_CR1 is cleared). CRC calculation is done based on any CRC8 standard.
The entire 16-bits of this register are considered when a 16-bit data frame format is selected
(DFF bit of the SPI_CR1 register is set). CRC calculation is done based on any CRC16
standard.
Note: A read to this register when the BSY Flag is set could return an incorrect value.
These bits are not used for I2S mode.
RM0090 Rev 19 921/1751

925
28.5.7 SPI TX CRC register (SPI_TXCRCR) (not used in I2S mode)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TXCRC[15:0]
Bits 15:0 TXCRC[15:0]: Tx CRC register

When CRC calculation is enabled, the TxCRC[7:0] bits contain the computed CRC value of
the subsequently transmitted bytes. This register is reset when the CRCEN bit of SPI_CR1
is written to 1. The CRC is calculated serially using the polynomial programmed in the
SPI_CRCPR register.
Only the 8 LSB bits are considered when the data frame format is set to be 8-bit data (DFF
bit of SPI_CR1 is cleared). CRC calculation is done based on any CRC8 standard.
The entire 16-bits of this register are considered when a 16-bit data frame format is selected
(DFF bit of the SPI_CR1 register is set). CRC calculation is done based on any CRC16
standard.
Note: A read to this register when the BSY flag is set could return an incorrect value.
These bits are not used for I2S mode.
28.5.8 SPI_I2S configuration register (SPI_I2SCFGR)

Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PCMSY
I2SMOD I2SE I2SCFG I2SSTD CKPOL DATLEN CHLEN
NC
Reserved Res.

Bit 11 I2SMOD: I2S mode selection
0: SPI mode is selected
1: I2S mode is selected
Note: This bit should be configured when the SPI or I2S is disabled
Bit 10 I2SE: I2S Enable
0: I2S peripheral is disabled
1: I2S peripheral is enabled
Note: This bit is not used in SPI mode.
Bits 9:8 I2SCFG: I2S configuration mode
00: Slave - transmit
01: Slave - receive
10: Master - transmit
11: Master - receive
Note: This bit should be configured when the I2S is disabled.
It is not used in SPI mode.
922/1751 RM0090 Rev 19

Bit 7 PCMSYNC: PCM frame synchronization

0: Short frame synchronization
1: Long frame synchronization
Note: This bit has a meaning only if I2SSTD = 11 (PCM standard is used)
It is not used in SPI mode.
Bit 6 Reserved: forced at 0 by hardware
Bits 5:4 I2SSTD: I2S standard selection
00: I2S Philips standard.
01: MSB justified standard (left justified)
10: LSB justified standard (right justified)
11: PCM standard
For more details on I2S standards, refer to Section 28.4.3: Supported audio protocols. Not used
in SPI mode.
Note: For correct operation, these bits should be configured when the I2S is disabled.
Bit 3 CKPOL: Steady state clock polarity
0: I2S clock steady state is low level
1: I2S clock steady state is high level
Note: For correct operation, this bit should be configured when the I2S is disabled.
This bit is not used in SPI mode
Bits 2:1 DATLEN: Data length to be transferred
00: 16-bit data length
11: Not allowed
Note: For correct operation, these bits should be configured when the I2S is disabled.
This bit is not used in SPI mode.
Bit 0 CHLEN: Channel length (number of bits per audio channel)
0: 16-bit wide
1: 32-bit wide
The bit write operation has a meaning only if DATLEN = 00 otherwise the channel length is fixed to
32-bit by hardware whatever the value filled in. Not used in SPI mode.
Note: For correct operation, this bit should be configured when the I2S is disabled.
RM0090 Rev 19 923/1751

925
28.5.9 SPI_I2S prescaler register (SPI_I2SPR)

Reset value: 0000 0010 (0x0002)
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MCKOE ODD I2SDIV
Reserved
rw rw rw

Bit 9 MCKOE: Master clock output enable
0: Master clock output is disabled
1: Master clock output is enabled
Note: This bit should be configured when the I2S is disabled. It is used only when the I2S is in master
mode.
This bit is not used in SPI mode.
Bit 8 ODD: Odd factor for the prescaler
0: real divider value is = I2SDIV *2
1: real divider value is = (I2SDIV * 2)+1
Refer to Section 28.4.4: Clock generator. Not used in SPI mode.
Note: This bit should be configured when the I2S is disabled. It is used only when the I2S is in master
mode.
Bits 7:0 I2SDIV: I2S Linear prescaler
I2SDIV [7:0] = 0 or I2SDIV [7:0] = 1 are forbidden values.
Refer to Section 28.4.4: Clock generator. Not used in SPI mode.
Note: These bits should be configured when the I2S is disabled. It is used only when the I2S is in
master mode.
924/1751 RM0090 Rev 19

28.5.10 SPI register map

The table provides shows the SPI register map and reset values.
Table 129. SPI register map and reset values

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
BIDIMODE
CRCNEXT
LSBFIRST
RXONLY
CRCEN
BIDIOE
MSTR
CPHA
CPOL
SSM
SPE
DFF
SPI_CR1 BR [2:0]
SSI
0x00 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RXDMAEN
TXDMAEN
RXNEIE
ERRIE
TXEIE
SSOE
Reserved
FRF
SPI_CR2
0x04 Reserved
CRCERR
CHSIDE
MODF
RXNE
OVR
UDR
FRE
BSY
TXE
SPI_SR
0x08 Reserved
Reset value 0 0 0 0 0 0 0 1 0
SPI_DR DR[15:0]
0x0C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
SPI_CRCPR CRCPOLY[15:0]
0x10 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1
SPI_RXCRCR RxCRC[15:0]
0x14 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
SPI_TXCRCR TxCRC[15:0]
0x18 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
PCMSYNC
DATLEN
I2SMOD
I2SCFG
I2SSTD
CKPOL
CHLEN
Reserved
I2SE
SPI_I2SCFGR
0x1C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0
MCKOE
ODD
SPI_I2SPR I2SDIV
0x20 Reserved
Reset value 0 0 0 0 0 0 0 0 1 0
RM0090 Rev 19 925/1751

925
Serial audio interface (SAI) RM0090
29 Serial audio interface (SAI)
This section applies to the STM32F42xxx and STM32F43xxx family.
29.1 Introduction
The SAI interface (serial audio interface) offers a wide set of audio protocols due to its
flexibility and wide range of configurations. Many stereo or mono audio applications may be
targeted. I2S standards, LSB or MSB-justified, PCM/DSP, TDM, and AC’97 protocols may
be addressed for example.
To bring this level of flexibility and configurability, the SAI contains two audio sub-blocks that
are fully independent of each other. Each audio sub-block is connected to up to 4 pins (SD,
SCK, FS, MCLK). Some of these pins can be shared if the two sub-blocks are declared as
synchronous to leave some free to be used as general purpose I/Os. The MCLK pin can be
output, or not, depending on the application, the decoder requirement and whether the
audio block is configured as the master.
The SAI can work in master or slave configuration. The audio sub-blocks can be either
receiver or transmitter and can work synchronously or not (with respect to the other one).
926/1751 RM0090 Rev 19

RM0090 Serial audio interface (SAI)
29.2 Main features

• Two independent audio sub-blocks which can be transmitters or receivers with their
respective FIFO.
• 8-word integrated FIFOs for each audio sub-block.
• Synchronous or asynchronous mode between the audio sub-blocks.
• Master or slave configuration independent for both audio sub-blocks.
• Clock generator for each audio block to target independent audio frequency sampling
when both audio sub-blocks are configured in master mode.
• Data size configurable: 8-, 10-, 16-, 20-, 24-, 32-bit.
• Peripheral with large configurability and flexibility allowing to target as example the
following audio protocol: I2S, LSB or MSB-justified, PCM/DSP, TDM, AC’97
• Up to 16 slots available with configurable size and with the possibility to select which
ones are active in the audio frame.
• Number of bits by frame may be configurable.
• Frame synchronization active level configurable (offset, bit length, level).
• First active bit position in the slot is configurable.
• LSB first or MSB first for data transfer.
• Mute mode.
• Stereo/Mono audio frame capability.
• Communication clock strobing edge configurable (SCK).
• Error flags with associated interrupts if enabled respectively.
– Overrun and underrun detection,
– Anticipated frame synchronization signal detection in slave mode,
– Late frame synchronization signal detection in slave mode,
– Codec not ready for the AC’97 mode in reception.
• Interruption sources when enabled:
– Errors,
– FIFO requests.
• DMA interface with 2 dedicated channels to handle access to the dedicated integrated
FIFO of each SAI audio sub-block.
RM0090 Rev 19 927/1751

964
29.3 Functional block diagram

The block diagram of the SAI is shown in Figure 283.
Figure 283. Functional block diagram
APB
Serial Audio Interface

(SAI) APB interface
SAI_XCR1 Audio block A

synchro int_sck
FIFO ctrl ctrl out int_FS
FIFO
FS_A
SAI_CK_A Clock generator FSM
Configuration SCK_A
Audio block A SD_A
I/O line Management

registers and
Status register MCLK_A
32-bit shift register
Audio block B
FIFO ctrl FS_B
SAI_CK_B Clock generator FIFO
SCK_B
Audio block B SD_B
FSM MCLK_B
Configuration
registers and
Status register 32-bit shift register
SAI_XCR1
APB interface
APB
MS30032V1
The SAI is mainly composed of two audio sub-blocks with their own clock generator. Each
audio block integrates a 32-bit shift register controlled by their own functional state machine.
Data are stored or read from the dedicated FIFO. FIFO may be accessed by the CPU, or by
DMA in order to leave the CPU free during the communication. Each audio block is
independent. They can be synchronous with each other.
An I/O line controller manages each dedicated pins for a given audio block in the SAI. If the
two blocks are synchronized, this controller reduces the number of I/Os used, freeing up an
FS pin, an SCK pin and eventually an MCLK pin, making them general purpose I/Os.
The functional state machine can be configured to address a wide range of audio protocols.
Some registers are present to set-up the desired protocols (audio frame waveform
generator).
The audio block can be a transmitter or receiver, in master or slave mode. The master mode
means the bit clock SCK and the frame synchronization signal are generated from the SAI,
whereas in slave mode, they come from another external or internal master. There is a
particular case for which the FS signal direction is not directly linked to the master or slave
928/1751 RM0090 Rev 19

mode definition. In AC’97 protocol, it will be an SAI output even if the SAI (link controller) is
set-up to consume the SCK clock (and so to be in Slave mode).
29.4 Main SAI modes

Each audio sub-block of the SAI can be configured to be master or slave via bit MODE[0] in
the SAI_xCR1 register of the selected audio block.
In master mode:
• The bit clock is generated by the SAI using the clock generator on pin SCK_A or
SCK_B (depending which audio block is declared as a master in the SAI).
• The dedicated pin SCK_x is considered as an output.
In slave mode:
• The slave must be enabled before the master is enabled.
• The slave audio block’s SCK clock I/O pin is considered input if it is configured in
asynchronous mode.
• If the audio block is declared synchronous with the second audio block in the SAI, its
SCK I/O pin is released to leave it free to be used as a general purpose I/O and is
connected internally to the SCK pin of the device with which it will be synchronized.
Each audio sub-block can be independently defined as a transmitter or receiver by bit
MODE[1] in the SAI_xCR1 register of the relevant audio block. The I/O pin SD will be
defined respectively as an output or an input.
It is possible to declare two master audio blocks in the same SAI with two different MCLK
and SCK clock frequencies (they have to be declared asynchronous).
Each of the audio blocks in the SAI are enabled by bit SAIxEN in the SAI_xCR1 register. As
soon as this bit is active, the transmitter or the receiver is sensitive to the activity on the
clock line, data line and synchronization line in slave mode.
In master TX mode, enabling the audio block immediately generates the bit clock for the
external slaves even if there is no data in the FIFO, However FS signal generation is
conditioned by the presence of data in the FIFO. After the FIFO receives the first data to
transmit, this data is output to external slaves. If there is no data to transmit in the FIFO, 0
values are then sent in the audio frame with an underrun flag generation.
In slave mode, the audio frame starts when the audio block is enabled and when a start of
frame is detected.
In Slave TX mode, no underrun event is possible on the first frame after the audio block is
enabled, because the mandatory operating sequence in this case is:
1. Write into the SAI_xDR (by software or by DMA).
2. Wait until the FIFO threshold (FLH) flag is different from 000b (FIFO empty).
3. Enable the audio block in slave transmitter mode.
RM0090 Rev 19 929/1751

964
29.5 SAI synchronization mode

Internal synchronization
An audio block can be declared synchronous with the second audio block. In this case, the
bit clock and the frame synchronization signals are shared to reduce the number of external
pins used for the communication. The audio block declared as synchronous with the other
one will see its own SCK_x, FS_x, and MCLK_x pins released to bring them back as GPIOs.
The one declared asynchronous is the one for which the I/O pins FS_x and SCK_x ad
MCLK_x (if the audio block is considered as master) are considered.
Typically, the audio block synchronous mode may be used to configure the SAI in full duplex
mode. One of the two audio blocks can be configured as master and the other as slave, or
both can be slaves; with one block declared as asynchronous (respective bit SYNCEN[1:0]
= 00 in SAI_xCR1) and the other one declared as synchronous with the other audio block
(respective bit SYNCEN[1:0] = 01 in the SAI_xCR1).
Note: APB frequency PCLK must be greater or equal to twice the bit rate clock frequency (due to
internal resynchronization stages).
29.6 Audio data size

The audio frame can target different data sizes by configuring bit DS[2:0] in the SAI_xCR1
register. The data sizes may be 8-, 10-, 16-, 20-, 24- or 32-bit. During the transfer, either the
MSB or the LSB of the data are sent first, depending on the configuration of bit LSBFIRST in
the SAI_xCR1 register.
29.7 Frame synchronization

The FS signal acts as the Frame synchronization signal in the audio frame (start of frame).
The shape of this signal is completely configurable in order to target the different audio
protocols with their own specificities concerning this Frame synchronization behavior. This
configurability is done using register SAI_xFRCR. Figure 284 gives a view of this flexibility.
Figure 284. Audio frame
Fs length = < 256 bits

Fs active length = < 128 bits
Fs offset
Fs
sck
sd Slot 0 Slot 1 Slot 2 Slot 3 … … Slot n
MS30037V1
930/1751 RM0090 Rev 19

In AC’97 mode (bit PRTCFG[1:0] = 10 in the SAI_xCR1 register), the frame synchronization
shape is forced to be configured to target these protocols. The SAI_xFRCR register value is
ignored.
Each audio block is independent and so each requires a specific configuration.
29.7.1 Frame length

• Master mode: The audio frame length can be configured up to 256 bit clock, setting bit
FRL[7:0] in the SAI_xFRCR register. If the frame length is greater than the number of
declared slots for the frame, the remaining bits to transmit will be extended to 0 or the
SD line will be released to HI-z depending the state of bit TRIS in the SAI_xCR2
register (refer to Section 29.12.4). In reception mode, the remaining bit is ignored.
• Slave mode: The audio frame length is mainly used in order to specify to the slave the
number of bit clocks per audio frame sent by the external master. It is used mainly to
detect from the master, any anticipated or late occurrence of the Frame
synchronization signal during an on-going audio frame. An error will be generated in
such case. For more details please refer to the Section 29.13.
The number of bits in the frame is equal to FRL[7:0] + 1.
The minimum number of bits to transfer in an audio frame is 8. This is the case when the
data size is 8-bit and only one slot is defined in NBSLOT[3:0] in the SAI_xSLOTR register
(NBSLOT[3:0] = 0000 for slot 0).
In master mode:
• If bit NODIV in the SAI_xCR1 register is cleared, the frame length should be aligned to
a number equal to a power of 2, from 8 to 256. This is to ensure that an audio frame
contains an integer number of MCLK pulses per bit clock, which ensures correct
operation of the external DAC/ADC inside the decoders. If the value set in FRL[7:0]
does not respect this rule, flag WCKCFG is set when the audio block is enabled and an
interrupt is generated if bit WCKCFGIE is set in the SAI_xIM register. The SAI is
automatically disabled.
• If bit NODIV in the SAI_xCR1 register is set, the FRL[7:0] bit can take any of the values
without constraint since the input clock of the audio block should be equal to the bit
clock. There is no MCLK_x clock which can be output. MCLK_x output pad is
automatically disabled.
In slave mode, there are no constraints for the FRL[7:0] bit configuration in the SAI_xFRCR
register.
29.7.2 Frame synchronization polarity

Bit FSPOL in the SAI_xFRCR register sets the active polarity of the FS pin from which a
frame is started. The start of frame is edge sensitive.
In slave mode, the audio block waits for a valid frame to start to transmit or to receive. Start
of frame is synchronized to this signal. It is effective only if the start of frame is not detected
during an on-going communication and assimilated to an anticipated start of frame (refer to
Section 29.13).
In master mode, the frame synchronization is sent continuously each time an audio frame is
complete until the SAIxEN bit in the SAI_xCR1 register is cleared. If no data is present in the
FIFO at the end of the previous audio frame, an underrun condition will be managed as
RM0090 Rev 19 931/1751

964
described in Section 29.13), but there will be no interruption in the audio communication
flow.
29.7.3 Frame synchronization active level length

Bit FSALL[6:0] in the SAI_xFRCR register configures the length of the active level of the
Frame synchronization signal. The length can be set from 1 to 128 bit clock SCK.
The active length may be half of the frame length in I2S, LSB or MSB-justified modes for
instance, or one-bit wide for PCM/DSP or TDM mode, or even 16-bit length in AC’97.
29.7.4 Frame synchronization offset

Depending on the audio protocol targeted in the application, the Frame synchronization
signal can be asserted when transmitting the last bit or the first bit of the audio frame (as for
instance, respectively, in I2S standard protocol and in MSB-justified protocol). Bit FSOFF in
the SAI_xFRCR register makes the choice.
29.7.5 FS signal role

The FS signal may have a different meaning depending on the FS function. Bit FSDEF in
the SAI_xFRCR register selects which meaning it will have. It may be either:
• 0: a start of frame, like for instance the PCM/DSP, TDM, AC’97, audio protocols,
• 1: a start of frame and a channel side identification within the audio frame like for the
I2S, the MSB or LSB-justified protocols.
When the FS signal is considered as a start of frame and channel side identification within
the frame, the number of declared slots must be considered to be half of the number for the
left channel and half of the number for the right channel. If the number of bit clock on half
audio frame is greater than the number of slots dedicated to a channel side, if TRIS = 0, 0 is
sent for transmission for the remaining bit clock in the SAI_xCR2 register, otherwise if TRIS
= 1, the SD line is released to HI-Z. In reception, the remaining bit clock are not considered
until the channel side changes.
932/1751 RM0090 Rev 19

Figure 285. FS role is start of frame + channel side identification (FSDEF = TRIS = 1)
Number of slots not aligned with the audio frame
Audio frame
Half of frame
FS
sck
slot Slot 0 ON Slot 1 OFF Slot 2 ON Slot 3 ON Slot 4 OFF Slot 5 ON
Number of slots aligned with the audio frame

Audio frame
Half of frame
FS
sck
slot Slot 0 Slot 1 Slot 2 Slot 3 Slot 4 Slot 5

MS30038V1
1. The frame length should be even.

If bit FSDEF in SAI_xFRCR is kept clear, so FS signal is equivalent to a start of frame, and
if the number of slots defined in bit NBSLOT[3:0] in SAI_xSLOTR multiplied by
the number of bits by slot configured in bit SLOTSZ[1:0] in SAI_xSLOTR is less than the
frame size (bit FRL[7:0] in the SAI_xFRCR register), then,
• if TRIS = 0 in the SAI_xCR2 register, the remaining bit after the last slot will be forced to
0 until the end of frame in case of transmission,
• if TRIS = 1, the line will be released to HI-Z during the transfer of these remaining bits.
In reception mode, these bits are discarded.
Figure 286. FS role is start of frame (FSDEF = 0)
Audio frame
sck
slot Slot 0 Slot 1 Slot 2 ... Slot n
Data = 0 after slot n if TRIS = 0

SD output released (HI-Z) after slot n if TRIS = 1
MS30039V1
29.8 Slot configuration

The slot is the basic element in the audio frame. The number of slots in the audio frame is
equal to the configured setting of bit NBSLOT[3:0] in the SAI_xSLOTR register +1.
The maximum number of slots per audio frame is fixed at 16.
For AC’97 protocol (when bit PRTCFG[1:0] = 10), the number of slots is automatically set to
target the protocol specification, and the value of NBSLOT[3:0] is ignored.
RM0090 Rev 19 933/1751

964
Each slot can be defined as a valid slot, or not, by setting bit SLOTEN[15:0] in the
SAI_xSLOTR register. In an audio frame, during the transfer of a non-valid slot, 0 value will
be forced on the data line or the SD data line will be released to HI-z (refer to
Section 29.12.4) if the audio block is transmitter, or the received value from the end of this
slot will be ignored. Consequently, there will be no FIFO access and so no request to read
or write the FIFO linked to this inactive slot status.
The slot size is also configurable as shown in the Figure 287. The size of the slots is
selected by setting bit SLOTSZ[1:0] in the SAI_xSLOTR register. The size is applied
identically for each slot in an audio frame.
Figure 287. Slot size configuration with FBOFF = 0 in SAI_xSLOTR
Audio block is transmitter Audio block is receiver
Slot size = data size Slot size = data size
slotx data size slotx data size
data size data size
slotx data size 00..00 slotx data size X..X
16-bit 16-bit
slotx data size 00..00 slotx data size XX..XX
32-bit 32-bit
X: don’t care
MS30033V1
It is possible to choose the position of the first data bit to transfer within the slots, this offset
is configured by bit FBOFF[5:0] in the SAI_xSLOTR register. 0 values will be injected in
transmitter mode from the beginning of the slot until this offset position is reached. In
reception, the bit in the offset phase is ignored. This feature targets the LSB justified
protocol (if the offset is equal to the slot size minus the data size).
Figure 288. First bit offset
Audio block is transmitter Audio block is receiver
Slot size = data size Slot size = data size
slotx data size slotx data size
data size data size
FBOFF FBOFF
00 data size 00 XX data size XX
slotx 16-bit slotx 16-bit
FBOFF = SLOT SZ -DS FBOFF = SLOT SZ -DS
slotx 00..00 data size slotx XX .. XX data size
32-bit 32-bit
X: don’t care
MS30034V1
934/1751 RM0090 Rev 19

It is mandatory to respect the following conditions in order to avoid bad SAI behavior:
FBOFF ≤ (SLOTSZ - DS),
DS ≤ SLOTSZ,
NBSLOT x SLOTSZ ≤ FRL (frame length),
The number of slots should be even when bit FSDEF in the SAI_xFRCR register is set.
In AC’97 (bit PRTCFG[1:0] = 10), the slot size is automatically set as defined in
Section 29.11.
29.9 SAI clock generator

Each audio block has its own clock generator to make these two blocks completely
independent. There is no difference in terms of functionality between these two clock
generators. They are exactly the same.
When the audio block is defined as Master, the clock generator generates the
communication clock (the bit clock) and the master clock for external decoders.
When the audio block is defined as slave, the clock generator is OFF.
Figure 289 illustrates the architecture of the audio block clock generator.
Figure 289. Audio block clock generator overview
NODIV
MCLK_x
MCKDIV[3:0]
0
FRL[7:0]
1 NODIV NODIV
SAI_CK_x 0
Master clock
divider Bit clock divider 0 SCK_x
0
1 1
MS30040V1
Note: If NoDiv is set to 1, the MCLK_x signal will be set at 0 level if this pin is configured as the
SAI pin in GPIO peripherals.
The clock source for the clock generator comes from the product clock controller. The
SAI_CK_x clock is equivalent to the master clock which may be divided for the external
decoders using bit MCKDIV[3:0]:
MCLK_x = SAI_CK_x / (MCKDIV[3:0] * 2), if MCKDIV[3:0] is not equal to 0000.
MCLK_x = SAI_CK_x, if MCKDIV[3:0] is equal to 0000.
MCLK_x signal is used only in TDM.
The division must be even in order to keep 50% on the Duty cycle on the MCLK output and
on the SCK_x clock. If bit MCKDIV[3:0] = 0000, division by one is applied to have MCLK_x
= SAI_CK_x.
In the SAI, the single ratio MCLK/FS = 256 is considered. Mostly, three frequency ranges
will be encountered as illustrated in the Table 130.
RM0090 Rev 19 935/1751

964
Table 130. Example of possible audio frequency sampling range

Input SAI_CK_x clock Most usual audio frequency
MCKDIV[3:0]
frequency sampling achievable
192 kHz MCKDIV[3:0] = 0000

96 kHz MCKDIV[3:0] = 0001
192 kHz x 256 48 kHz MCKDIV[3:0] = 0010
16 kHz MCKDIV[3:0] = 0100
8 kHz MCKDIV[3:0] = 1000
44.1 kHz MCKDIV[3:0] = 0000
44.1 kHz x 256 22.05 kHz MCKDIV[3:0] = 0001
11.025 kHz MCKDIV[3:0] = 0010
SAI_CK_x = MCLK(1) MCLK MCKDIV[3:0] = 0000
1. This may happen when the product clock controller selects an external clock source, instead of PLL clock.
The master clock may be generated externally on an I/O pad for external decoders if the
corresponding audio block is declared as master with bit NODIV = 0 in the SAI_xCR1
register. In slave, the value set in this last bit is ignored since the clock generator is OFF,
and the MCLK_x I/O pin is released for use as a general purpose I/O.
The bit clock is derived from the master clock. The bit clock divider sets the divider factor
between the bit clock SCK_x and the master clock MCLK_x following the formula:
SCK_x = MCLK x (FRL[7:0] +1) / 256
where:
256 is the fixed ratio between MCLK and the audio frequency sampling.
FRL[7:0] is the number of bit clock - 1 in the audio frame, configured in the SAI_xFRCR
register.
It is mandatory in master mode that (FRL[7:0] +1) should be equal to a number with a power
of 2 (refer to Section 29.7) in order to have an even integer number of MCLK_x pulses by bit
clock. The 50% duty cycle is guaranteed on the bit clock SCK_x.
The SAI_CK_x clock can be also equal to the bit clock frequency. In this case, bit NODIV in
the SAI_xCR1 register should be set and the value inside the MCKDIV divider and the bit
clock divider will be ignored. In this case, the number of bits per frame is fully configurable
without the need to be equal to a power of two.
The bit clock strobing edge on SCK can be configured by bit CKSTR in the SAI_xCR1
register.
29.10 Internal FIFOs

Each audio block in the SAI has its own FIFO. Depending if the block is defined to be a
transmitter or a receiver, the FIFO will be written or read, respectively. There is therefore
only one FIFO request linked to FREQ bit in the SAI_xSR register.
936/1751 RM0090 Rev 19

An interrupt is generated if FREQIE bit is enabled in the SAI_xIM register. This depends on:
• FIFO threshold setting (FLTH bits in SAI_CR2)
• Communication direction transmitter or receiver (see Section : Interrupt generation in
transmitter mode and Section : Interrupt generation in reception mode)
Interrupt generation in transmitter mode

The interrupt generation depends on the FIFO configuration in transmitter mode:
• When the FIFO threshold bits in SAI_XCR2 register are configured as FIFO empty
(FTH[2:0] set to 000b), an interrupt is generated (FREQ bit set by hardware to 1 in
SAI_XSR register) if no data are available in SAI_xDR register (FLTH[2:0] bits in SAI_xSR
is less than 001b). This Interrupt (FREQ bit in SAI_XSR register) is cleared by hardware
when the FIFO became not empty (FLTH[2:0] bits in SAI_xSR are different from 000b) i.e
one or more data are stored in the FIFO.
• When the FIFO threshold bits in SAI_XCR2 register are configured as FIFO quarter full
SAI_XSR register) if less than a quarter of the FIFO contains data (FLTH[2:0] bits in
SAI_xSR are less than 010b). This Interrupt (FREQ bit in SAI_XSR register) is cleared by
hardware when at least a quarter of the FIFO contains data (FLTH[2:0] bits in SAI_xSR
are higher or equal to 010b).
• When the FIFO threshold bits in SAI_XCR2 register are configured as FIFO half full
SAI_XSR register) if less than half of the FIFO contains data (FLTH[2:0] bits in SAI_xSR
are less than 011b). This Interrupt (FREQ bit in SAI_XSR register) is cleared by hardware
when at least half of the FIFO contains data (FLTH[2:0] bits in SAI_xSR are higher or
equal to 011b).
• When the FIFO threshold bits in SAI_XCR2 register are configured as FIFO three quarter
(FTH[2:0] set to 011b), an interrupt is generated (FREQ bit is set by hardware to 1 in
SAI_XSR register) if less than three quarters of the FIFO contain data (FLTH[2:0] bits in
SAI_xSR are less than 100b). This Interrupt (FREQ bit in SAI_XSR register) is cleared by
hardware when at least three quarters of the FIFO contain data (FLTH[2:0] bits in
SAI_xSR are higher or equal to 100b).
• When the FIFO threshold bits in SAI_XCR2 register are configured as FIFO full (FTH[2:0]
set to 100b), an interrupt is generated (FREQ bit is set by hardware to 1 in SAI_XSR
register) if the FIFO is not full (FLTH[2:0] bits in SAI_xSR is less than 101b). This Interrupt
(FREQ bit in SAI_XSR register) is cleared by hardware when the FIFO is full (FLTH[2:0]
bits in SAI_xSR is equal to 101b value).
Interrupt generation in reception mode

The interrupt generation depends on the FIFO configuration in reception mode:
• When the FIFO threshold bits in SAI_XCR2 register are configured as FIFO empty
SAI_XSR register) if at least one data is available in SAI_xDR register(FLTH[2:0] bits in
SAI_xSR is higher or equal to 001b). This Interrupt (FREQ bit in SAI_XSR register) is
cleared by hardware when the FIFO became empty (FLTH[2:0] bits in SAI_xSR is equal
to 000b) i.e no data is stored in FIFO.
• When the FIFO threshold bits in SAI_XCR2 register are configured as FIFO quarter fully
SAI_XSR register) if at less one quarter of the FIFO data locations are available
RM0090 Rev 19 937/1751

964
(FLTH[2:0] bits in SAI_xSR is higher or equal to 010b). This Interrupt (FREQ bit in
SAI_XSR register) is cleared by hardware when less than a quarter of the FIFO data
locations become available (FLTH[2:0] bits in SAI_xSR is less than 010b).
• When the FIFO threshold bits in SAI_XCR2 register are configured as FIFO half fully
(FTH[2:0] set to 010b value), an interrupt is generated (FREQ bit is set by hardware to 1
in SAI_XSR register) if at least half of the FIFO data locations are available (FLTH[2:0]
bits in SAI_xSR is higher or equal to 011b). This Interrupt (FREQ bit in SAI_XSR register)
is cleared by hardware when less than half of the FIFO data locations become available
(FLTH[2:0] bits in SAI_xSR is less than 011b).
• When the FIFO threshold bits in SAI_XCR2 register are configured as FIFO three quarter
full(FTH[2:0] set to 011b value), an interrupt is generated (FREQ bit is set by hardware to
1 in SAI_XSR register) if at least three quarters of the FIFO data locations are available
(FLTH[2:0] bits in SAI_xSR is higher or equal to 100b). This Interrupt (FREQ bit in
SAI_XSR register) is cleared by hardware when the FIFO has less than three quarters of
the FIFO data locations avalable(FLTH[2:0] bits in SAI_xSR is less than 100b).
• When the FIFO threshold bits in SAI_XCR2 register are configured as FIFO full(FTH[2:0]
set to 100b), an interrupt is generated (FREQ bit is set by hardware to 1 in SAI_XSR
register) if the FIFO is full (FLTH[2:0] bits in SAI_xSR is equal to 101b). This Interrupt
(FREQ bit in SAI_XSR register) is cleared by hardware when the FIFO is not full
(FLTH[2:0] bits in SAI_xSR is less than 101b).
Like interrupt generation, the SAI can use the DMA if DMAEN bit in the SAI_xCR1 register is
set. The FREQ bit assertion mechanism is the same as the interruption generation
mechanism described above for FREQIE.
Each FIFO is an 8-word FIFO. Each read or write operation from/to the FIFO targets one
word FIFO allocation whatever the access size. Each FIFO word contains one audio frame.
FIFO pointers are incremented by one word after each access to the SAI_xDR register.
Data should be right aligned when it is written in the SAI_xDR.
Data received will be right aligned in the SAI_xDR.
The FIFO pointers can be reinitialized when the SAI is disabled by setting bit FFLUSH in the
SAI_xCR2 register. If FFLUSH is set when the SAI is enabled the data present in the FIFO
will be lost automatically.
938/1751 RM0090 Rev 19

29.11 AC’97 link controller

The SAI is able to work as an AC’97 link controller. In this protocol:
• The slot number and the slot size are fixed.
• The frame synchronization signal is perfectly defined and has a fixed shape.
To select this protocol, set bit PRTCFG[1:0] in the SAI_xCR1 register to 10. When AC’97
mode is selected the data sizes that can be used are 16-bit or 20-bit only, else SAI behavior
is not guaranteed.
• Bits NBSLOT[3:0] and SLOTSZ[1:0] are consequently ignored.
• The number of slots is fixed at 13 slots. The first one is 16 bits wide and all the others
are 20 bits wide (data slots).
• Bit FBOFF[5:0] in the SAI_xSLOTR register is ignored
• The SAI_xFRCR register is ignored.
The FS signal from the block defined as asynchronous is configured automatically as an
output, since the AC’97 controller link drives the FS signal whatever the master or slave
configuration.
Figure 290 presents an AC’97 audio frame structure.
Figure 290. AC’97 audio frame
FS
1 2 3 4 5 6 7 8 9 10 11 12
CMD CMD PCM PCM LINE1 PCM PCM PCM PCM LINE2 HSET IO
SDI Tag ADDR DATA LFRONT RFRONT DAC CENTER LSURR RSURR LFE DAC DAC CTRL
STATUS STATUS PCM PCM LINE1 PCM RSR RSR RSR LINE2 IO
Tag HSET
SDO ADDR DATA LEFT RIGHT ADC MIC VD VD LVD ADC STATUS
MS192343V1
Note: In AC’97 protocol, bit 2 of the tag is reserved (always 0), so whatever the value written in the
SAI FIFO, bit 2 of the TAG is forced to 0 level.
For more details about TAG representation, please refer to the AC’97 protocol standard.
One SAI can be used to target an AC’97 point-to-point communication.
In receiver mode, the SAI acting as an AC’97 link controller will require no FIFO request and
so no data storage in the FIFO when the codec ready bit in the slot 0 is decoded low. If bit
CNRDYIE is enabled in the SAI_xIM register, flag CNRDY will be set in the SAI_xSR
register and an interrupt is generated. This flag is dedicated to the AC’97 protocol.
RM0090 Rev 19 939/1751

964
29.12 Specific features

The SAI has some specific functions which can be useful depending on the audio protocol
selected. These functions are accessible through specific bits in the SAI_xCR2 register.
29.12.1 Mute mode

Mute mode may be used when the audio block is a transmitter or receiver.
Transmitter
In transmitter mode, Mute mode can be selected at anytime. Mute mode is active for entire
audio frames. The bit MUTE in the SAI_xCR2 register requests Mute mode when it is set
during an on-going frame.
The mute mode bit is strobed only at the end of the frame. If set at this time, the mute mode
is active at the beginning of the new audio frame, for a complete frame, until the next end of
frame, it then strobes the bit to determine if the next frame will still be a mute frame.
If the number of slots set in bit NBSLOT[3:0] in the SAI_xSLOTR register is lower than or
equal to two, it is possible to specify if the value sent during the Mute mode is 0 or if it is the
last value of each slot. The selection is done via bit MUTEVAL in the SAI_xCR2 register.
If the number of slots set in bit NBSLOT[3:0] in the SAI_xSLOTR register is greater than
two, MUTEVAL bit in the SAI_xCR2 has no meaning as 0 values are sent on each bit on
each slot.
During Mute mode, the FIFO pointers are still incremented, meaning that data which was
present in the FIFO and for which the Mute mode is requested is discarded.
Receiver
In receiver mode, it is possible to detect a Mute mode sent from the external transmitter
when all the declared and valid slots of the audio frame receive 0 for a given consecutive
number of audio frames (bit MUTECNT[5:0] in the SAI_xCR2 register).
When the number of MUTE frames is detected, flag MUTEDET in the SAI_xSR register is
set and an interrupt can be generated if bit MUTEDETIE is set in the SAI_xCR2.
The mute frame counter is cleared when the audio block is disabled or when a valid slot
receives at least one data in an audio frame. The interrupt is generated just once, when the
counter reaches the specified value in bit MUTECNT[5:0]. Then the interrupt event is re-
armed when the counter is cleared.
29.12.2 MONO/STEREO function

In transmission mode, it is possible to address the Mono mode without any data pre-
processing in memory when the number of slot is equal to 2 (NBSLOT[3:0] = 0001 in the
SAI_xSLOTR). In such a case, the access to and from the FIFO will be reduced by two
since in transmission, the data for slot 0 is duplicated into data slot 1.
To select the Mono feature, set bit MONO in the SAI_xCR1 register.
In reception mode, bit MONO can be set and has a meaning only if the number of slots is
equal to 2 like for the transmission mode. When it is set, only the data of slot 0 will be stored
in the FIFO. The data belonging to slot 1 will be discarded since in this case, it is supposed
to be the same as the previous slot. If the data flux in reception is a real stereo audio flow
940/1751 RM0090 Rev 19

with a distinct and different left and right data, bit MONO has no meaning. The conversion
from the output stereo file to the equivalent mono file is done by software.
Note: To enable Mono mode, NBSLOT and SLOTEN must equal two and MONO bit set to 1.
29.12.3 Companding mode

Telecommunication applications may require to process the data to transmit or to receive
with a data companding algorithm.
Depending on the COMP[1:0] bit in the SAI_xCR2 register (used only when TDM mode is
selected), the software may choose to process or not the data before sending it on SD serial
output line (compression) or to expand the data after the reception on SD serial input line
(expansion) as illustrated in Figure 291,.The two companding modes supported are the µ-
Law and the A-Law log which are a part of the CCITT G.711 recommendation.
The companding standard employed in the United States and Japan is the µ-Law and allows
14 bits of dynamic range (COMP[1:0] = 10 in the SAI_xCR2 register).
The European companding standard is A-Law and allows 13 bits of dynamic range
(COMP[1:0] = 11 in the SAI_xCR2 register).
Companding standard (µ-Law or A-Law) can be computed based on 1’s complement or 2’s
complement representation depending on the CPL bit setting in the SAI_xCR2 register.
The µ-Law and A-Law formats encode data into 8-bit code elements with MSB alignment.
Companded data is always 8 bits wide. For this reason, bit DS[2:0] in the SAI_xCR1 register
will be forced to 010 when the SAI audio block is enabled (bit SAIxEN = 1 in the SAI_xCR1
register) and when the COMP[1:0] bit selects one of these two companding modes.
If no companding processing is required, COMP[1:0] bit in the SAI_xCR2 register should be
kept cleared.
Figure 291. Data companding hardware in an audio block in the SAI
Receiver mode (bit MODE[0] = 1 in SAI_xCR1)
COMP[1]
1 expand SD
FIFO 32-bit shift register
0
Transmitter mode (bit MODE[0] = 0 in SAI_xCR1)
SD
0
FIFO 32-bit shift register
compress 1
COMP[1]
MS19244V1
Note: Not applicable when AC’97 selected.
RM0090 Rev 19 941/1751

964
Expansion or compression mode is automatically selected by the SAI configuration.

• If the SAI audio block is configured to be a transmitter, and if the COMP[1] bit is set in
the SAI_xCR2 register, the compression mode will be applied.
• If the SAI audio block is declared as a receiver, the expansion algorithm will be applied.
29.12.4 Output data line management on an inactive slot

In transmitter mode, it is possible to choose the behavior of the SD line in output when an
inactive slot is sent on the data line (via bit TRIS in the SAI_xCR2 register when the SAI is
disabled).
• Either the SAI forces 0 on the SD output line when an inactive slot is transmitted, or
• The line is released in HI-z state at the end of the last bit of data transferred, to release
the line for other transmitters connected to this node.
It is important to note that the two transmitters do not attempt to drive the same SD output
pin simultaneously, which could result in a short circuit. In order to ensure a gap between
transmissions, if the data is lower than 32-bit, the data can be extended to 32-bit by setting
bit SLOTSZ[1:0] = 10 in the SAI_xSLOTR register. Then, the SD output pin will be tristated
at the end of the LSB of the active slot (during the padding to 0 phase to extend the data to
32-bit) if the following slot is declared inactive.
In addition, if the number of slots multiplied by the slot size is lower than the frame length,
the SD output line will be tristated when the padding to 0 is done to complete the audio
frame.
Figure 292 illustrates these behaviors.
942/1751 RM0090 Rev 19

Figure 292. Tristate strategy on SD output line on an inactive slot

Bit TRIS = 1 in the SAI_xCR1 and frame length = number of slots
Audio frame
sck
Slot size = data size

slot Slot 0 ON Slot 1 OFF Slot 2 OFF Slot 3 ON .. ON .. ON Slot n ON
SD (output) Data 0 Data 1 .. .. Data m
Slot size > data size
SD (output) Data 0 Data 1 .. .. Data m
SD (output) Data 0 .. .. Data m
Bit TRIS = 1 in the SAI_xCR1 and frame length > number of slots
Audio frame
sck

slot Slot 0 ON Slot 1 OFF Slot 2 OFF .. ON Slot n ON
SD (output) Data 0 .. Data m

slot Slot 0 ON Slot 1 OFF Slot 2 OFF .. ON Slot n ON
SD (output) Data 0 .. Data m
slot Slot 0 ON Slot 1 OFF Slot 2 OFF .. ON .. ON
SD (output) .. Data m
MS192345V1
When the selected audio protocol uses the FS signal as a start of frame and a channel side
identification (bit FSDEF = 1 in the SAI_xFRCR register), the tristate mode is managed
according to Figure 293 (where bit TRIS in the SAI_xCR1 register = 1, and FSDEF=1, and
half frame length > number of slots/2, and NBSLOT=6).
RM0090 Rev 19 943/1751

964
Figure 293. Tristate on output data line in a protocol like I2S
sck

SD (output) Data 0 Data 1 Data 2 Data 3
SD (output) Data 0 Data 1 Data 2 Data 3
SD (output) Data 0 Data 1 Data m
MS192346V1
If the TRIS bit in the SAI_xCR2 register is cleared, all the High impedance states on the SD
output line on Figure 292 and Figure 293 are replaced by a drive with a value of 0.
29.13 Error flags

The SAI embeds some error flags:
– FIFO overrun/underrun,
– Anticipated frame synchronization detection,
– Late frame synchronization detection,
– Codec not ready (AC’97 exclusively),
– Wrong clock configuration in master mode.
29.13.1 FIFO overrun/underrun (OVRUDR)

The FIFO Overrun/Underrun bit is called OVRUDR in the SAI_xSR register.
The overrun or underrun errors occupy the same bit since an audio block can be either
receiver or transmitter and each audio block in an SAI has its own SAI_xSR register.
Overrun
When the audio block is configured as receiver, an overrun condition may appear if data is
received in an audio frame when the FIFO is full and is not able to store the received data.
In this case, the received data is lost, the flag OVRUDR in the SAI_xSR register is set and
an interrupt is generated if bit OVRUDRIE is set in the SAI_xIM register. The slot number
from which the overrun occurs, is stored internally. No more data will be stored into the FIFO
until it becomes free to store new data. When the FIFO has at least one data free, the SAI
audio block receiver will store new data (from new audio frame) from the slot number which
944/1751 RM0090 Rev 19

was stored internally when the overrun condition was detected, and this, to avoid data slot
de-alignment in the destination memory (refer to Figure 294).
The OVRUDR flag is cleared when bit COVRUDR is set in the SAI_xCLRFR register.
Figure 294. Overrun detection error
Example: FIFO overrun on Slot 1

Audio frame Audio frame
sck
data Slot 0 ON Slot 1 ON Slot 1 ON Slot 0 ON Slot 1 ON ... ON Slot n ON
FIFO full
Received data discarded Data stored again in FIFO
OVRUDR
COVRUDR = 1
MS192348V2
Underrun
An underrun may occur when the audio block in the SAI is a transmitter and the FIFO is
empty when data needs to be transmitted (the audio block configuration (Master or Slave) is
not relevant). If an underrun is detected, the software must resynchronize data and slot.
Proceed as follows:
1. Disable the SAI peripheral by resetting the SAIEN bit of the SAI_xCR1 register. Check
that the SAI has been disabled by reading back the SAIEN bit (SAIEN should be equal
to 0).
2. Flush the Tx FIFO through the FFLUS bit of the SAI_xCR2 register.
3. Re-assigned to the correct data to be transferred on the first active slot of the new
frame.
4. Re-enabling the SAI peripheral (SAIEN bit set to 1).
The underrun event sets the OVRUDR flag in the SAI_xSR register and an interrupt is
generated if the OVRUDRIE bit is set in the SAI_xIM register. To clear this flag, set the
COVRUDR bit in the SAI_xCLRFR register.
RM0090 Rev 19 945/1751

964
Figure 295. FIFO underrun event
Example: FIFO underrun on Slot 1

Audio frame Audio frame
sck
data
SD (output) Slot 0 ON MUTE MUTE MUTE Slot 1 ON ... ON Slot 0 ON
FIFO empty
OVRUND
OVRUND=1
MS192347V2
29.13.2 Anticipated frame synchronisation detection (AFSDET)

This flag AFSDET is used only in Slave mode. In master mode, it is never asserted. It
informs about the detection of a frame synchronisation (FS) earlier than expected since the
frame length, the frame polarity, the frame offset are defined and known.
Early detection sets flag AFSDET in the SAI_xSR register.
This detection has no effect on the current audio frame which is not sensitive to the
anticipated FS. This means that “parasitic” events on signal FS are flagged without any
perturbation of the current audio frame.
If bit AFSDETIE is set in the SAI_xIM register, an interrupt is generated. To clear the flag
AFSDET, bit CAFSDET in the SAI_xCLRFR register has to be set.
To resynchronize with the master after Anticipated frame detection error, four steps should
be respected:
1. SAI block should be disabled by resetting SAIEN bit in SAI_xCR1 register, to be sure
that the SAI is disabled SAIEN bit is should be equal to 0 (reading back this bit).
2. FIFO should be flushed via FFLUS bit in SAI_xCR2 register.
3. Re-enabling the SAI peripheral (SAIEN bit set to 1) then the SAI.
4. SAI block will wait for the assertion on FS to restart the synchronization with master.
Note: This flag is not asserted in AC’97 since the SAI audio block acts as a link controller and
generates the FS signal even when declared as slave.
29.13.3 Late frame synchronization detection

Flag LFSDET in the SAI_xSR register can be set only when the SAI audio block is defined
as slave. The frame length, the frame polarity and the frame offset configuration are known
in register SAI_xFRCR.
If the external master does not send the FS signal at the expecting time (generating the
signal too late), the flag LFSDET in the SAI_xSR register will be set and an interrupt is
generated if bit LFSDETIE in the SAI_xIM register is set.
The flag is cleared when bit CLFSDET is set in the SAI_xCLRFR register.
946/1751 RM0090 Rev 19

The late frame synchronisation detection flag is set when the error is detected, SAI needs to
be resynchronized with the master (the four steps described above should be respected).
This detection and flag assertion can detect glitches on the SCK clock in a noisy
environment, detected by the state machine of the audio block. It could incorrectly shift the
SAI audio block state machine from one state in the current audio frame, thus corrupting the
frame.
There is no corruption if the external master is not managing the audio data frame transfer in
a continuous mode, which should not be the case for most application purposes. In this
case, flag LFSDET will be set.
Note: This flag is not asserted in AC’97 mode since the SAI audio block acts as a link controller
and generates the FS signal even when declared as slave.
29.13.4 Codec not ready (CNRDY AC’97)

The flag CNRDY in the SAI_xSR register is relevant only if the SAI audio block is configured
to work in AC’97 mode (bit PRTCFG[1:0] = 10 in the SAI_xCR1 register). If bit CNRDYIE is
set in the SAI_xIM register, an interrupt will be generated when the flag CNRDY is set.
It is asserted when the codec is not ready to communicate during the reception of the TAG 0
(slot0) of the AC’97 audio frame. In this case, there will be no data automatically stored into
the FIFO since the codec is not ready, until the TAG 0 indicates that the codec is ready. All
the active slots defined in the SAI_xSLOTR register will be captured when the codec is
ready.
To clear the flag, bit CCNRDY in the SAI_xCLRFR register has to be set.
29.13.5 Wrong clock configuration in master mode (with NODIV = 0)

When the audio block is master (MODE[1] = 0 in the SAI_xCR1 register) and if bit NODIV in
the SAI_xCR1 is clear, the flag WCKCFG will be set if bit FRL[7:0] in the SAI_xFRCR is not
set with a proper value when the SAIxEN bit in the SAI_xCR1 register is set, in order to
respect this following rule:
n
( FRL[7,0] ) + 1 = 2
where n is in the range from 3 to 8.

If bit WCKCFGIE is set, an interrupt is generated when flag WCKCFG is set in the SAI_xSR
register. To clear the flag, set bit CWCKCFG bit in the SAI_xCLRFR register.
When bit WCKCFG is set, the audio block is automatically disabled, clearing bit SAIxEN in
the SAI_xCR1 register via hardware.
The above formula is intended to guarantee that the number of MCLK pulses by bit clock is
an even integer in the audio frame with a 50% duty cycle bit clock generation to guarantee
the good quality of the audio sounds or acquisitions.
RM0090 Rev 19 947/1751

964
29.14 Interrupt sources

The SAI has 7 possible interrupt sources as illustrated by Table .
Table 131. Interrupt sources

Interru
Interrupt
pt Audio block mode Interrupt enable Interrupt clear
source
group
Depend on:
- FIFO threshold setting
(FLTH bits in SAI_CR2)
Master or Slave FREQIE in
FREQ FREQ - Communication direction
Receiver or transmitter SAI_xIM register
transmitter or receiver
for more details please refer
to Internal FIFOs section
Master or Slave OVRUDRIE in COVRUDR = 1 in
OVRUDR ERROR
Receiver or transmitter SAI_xIM register SAI_xCLRFR register
Slave
AFSDETIE in CAFSDET = 1 in
AFSDET ERROR (Not used in AC’97
SAI_xIM register SAI_xCLRFR register
mode)
Slave
LFSDETIE in CLFSDET = 1 in
LFSDET ERROR (Not used in AC’97
mode)
Slave CNRDYIE in CCNRDY = 1 in
CNRDY ERROR
(Only in AC’97 mode) SAI_xIM register SAI_xCLRFR register
MUTEDE Master or slave MUTEDETIE in CMUTEDET = 1 in

MUTE
T Receiver mode only SAI_xIM register SAI_xCLRFR register
Master with NODIV = 0

WCKCFGIE in CWCKCFG = 1 in
WCKCFG ERROR in the SAI_xCR1
register
Below are the SAI configuration steps to follow when an interrupt occurs:
1. Disable SAI interrupt.
2. Configure SAI.
3. Configure SAI interrupt source.
4. Enable SAI.
29.15 Disabling the SAI

The audio block in the SAI can be disabled at any moment by clearing bit SAIxEN in the
SAI_xCR1 register. All the frames that have already started will be automatically completed
before the total extinction of the SAI. Bit SAIxEN in the SAI_xCR1 register will stay high until
the SAI is completely switched-off at the end of the current audio frame transfer.
948/1751 RM0090 Rev 19

If there is an audio block in the SAI synchronous with the other one, the one which is the
master must be disabled first.
29.16 SAI DMA interface

In order to free the CPU and to optimize the bus bandwidth, each SAI audio block has an
independent DMA interface in order to read or to write into the SAI_xDR register (to hit the
internal FIFO). There is one DMA channel per audio block following basic DMA
request/acknowledge protocol.
To configure the audio block to transfer through the DMA interface, set bit DMAEN in the
SAI_xCR1 register. The DMA request is managed directly by the FIFO controller depend of
FIFO threshold level (for more details please refer to Internal FIFOs section). DMA direction
is linked to the SAI audio block configuration:
• If the audio block is a transmitter, the audio block’s FIFO controller outputs a DMA
request to load the FIFO with data written in the SAI_xDR register.
• If the audio block is a receiver, the DMA request will concern read operations from the
SAI_xDR register.
Below are the SAI configuration steps followed when DMA is used:
1. Configure SAI and FIFO Threshold level (in order to specify when the DMA request to
be launched)
2. Configure SAI DMA channel
3. Enable DMA
4. Enable SAI
Note: Before configuring the SAI block, the SAI DMA channel must be disabled.
RM0090 Rev 19 949/1751

964
29.17 SAI registers
29.17.1 SAI xConfiguration register 1 (SAI_xCR1) where x is A or B

Address offset: Block A: 0x004
Address offset: Block B: 0x024
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
MCKDIV[3:0] NODIV DMAEN SAIxEN

Reserved Res.
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OutDri LSBFIR
MONO SYNCEN[1:0] CKSTR DS[2:0] PRTCFG[1:0] MODE[1:0]
Reserved v ST Res.
Bits 31:24 Reserved, always read as 0.

Bit 23:20 MCKDIV[3:0]: Master clock divider. These bits are set and cleared by software.
0000: Divides by 1 the master clock input.
Otherwise, The Master clock frequency is calculated accordingly to the following formula:
MCLK_x = SAI_CK_x / (MCKDIV[3:0] * 2)
These bits have no meaning when the audio block is slave.
They have to be configured when the audio block is disabled.
Bit 19 NODIV: No divider. This bit is set and cleared by software.
0: Master Clock divider is enabled
1: No divider used in the clock generator (in this case Master Clock Divider bit has no effect)
Bit 18 Reserved, always read as 0.
Bit 17 DMAEN: DMA enable. This bit is set and cleared by software.
0: DMA is disabled
1: DMA is enabled
Note: In receiver mode, the bits MODE must be configured before setting bit DMAEN to avoid a DMA
request since the audio block is transmitter after reset (default setting)
Bit 16 SAIxEN: Audio block enable where x is A or B. This bit is set by software. It is cleared by hardware,
after disabling it by software (writing the bit low), the audio is completely disabled (waiting for the end
of the current frame).
0: Audio block is disabled
1: Audio block is enabled: this bit can be set only if it is at 0 during the write operation (means the
SAI is completely disabled before being re-enabled).
This bit allows to control the state of the audio block. If it is disabled somewhere in an audio frame,
the on-going transfer will be completed and the cell will be totally disabled at the end of this audio
frame transfer.
Note: When SAIx block is configured as master mode, clock must be present on the input of the SAI
before setting SAIxEN bit.
Bit 13 OUTDRIV: Output drive. This bit is set and cleared by software.
0: Audio block output driven when SAIEN is set
1: Audio block output driven immediately after the setting of this bit.
Note: This bit has to be set before enabling the audio block but after the audio block configuration.
950/1751 RM0090 Rev 19

Bit 12 MONO: Mono mode. This bit is set and cleared by software.
0: Stereo mode
1: Mono mode.
This bit has a meaning only when the number of slots is equal to 2.
When the Mono mode is selected, the data of the slot 0 data is duplicated on the slot 1 when the
audio block is a transmitter. In reception mode, the slot1 is discarded and only the data received
from the slot 0 will be stored.
Refer to Section 29.12.2 for more details.
Bits 11:10 SYNCEN[1:0]: Synchronization enable. This bit is set and cleared by software.
00: audio block is asynchronous.
01: audio block is synchronous with the other internal audio block. In this case audio block should be
configured in Slave mode
10: Reserved.
11: Not used
These bits have to be configured when the audio block is disabled.
Bit 9 CKSTR: Clock strobing edge. This bit is set and cleared by software.
0: data strobing edge is falling edge of SCK
1: data strobing edge is rising edge of SCK
This bit has to be configured when the audio block is disabled.
Bit 8 LSBFIRST: Least significant bit first. This bit is set and cleared by software.
0: data is transferred with the MSB of the data first
1: data is transferred with the LSB of the data first
This bit has to be configured when the audio block is disabled.
This bit has no meaning in AC’97 audio protocol since in AC’97 data is transferred with the MSB of
the data first.
Bits 7:5 DS[2:0]: Data size. These bits are set and cleared by software.
000: Not used
001: Not used
010: 8-bit
011: 10-bit
100: 16-bit
101: 20-bit
110: 24-bit
111: 32-bit
When the companding mode is selected (bit COMP[1:0]), these DS[1:0] are ignored since the data
size is fixed to 8-bit mode by the algorithm itself.
These bits must be configured when the audio block is disabled.
Note: When AC’97 mode is selected the data sizes that can be used are: 16-bit or 20-bit only, else
SAI behavior is not guaranteed.
RM0090 Rev 19 951/1751

964

Bits 3:2 PRTCFG[1:0]: Protocol configuration. These bits are set and cleared by software.
00: Free protocol
01: Not used
10: AC’97 protocol
11: Not used
Free protocol selection allows to use the powerful configuration of the audio block to address a
specific audio protocol (like I2S, LSB/MSB justified, TDM, PCM/DSP...) setting most of the
configuration register bits as well as frame configuration register.
Bits 1:0 MODE[1:0]: Audio block mode. These bits are set and cleared by software.
00: Master transmitter
01: Master receiver
10: Slave transmitter
11: Slave receiver
Note: In Master transmitter mode the audio block will start to generate the FS and clocks
952/1751 RM0090 Rev 19

29.17.2 SAI xConfiguration register 2 (SAI_xCR2) where x is A or B

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MUTE
COMP[1:0] CPL MUTECNT[5:0] Mute TRIS FFLUS FTH
VAL
Bits 31:16 Reserved, always read as 0

Bits 15:14 COMP[1:0]: Companding mode. These bits are set and cleared by software.
00: No companding algorithm
01: Reserved.
10: µ-Law algorithm
11: A-Law algorithm
The µ-Law and the A-Law log are a part of the CCITT G.711 recommendation, the type of
complement that will be used depends on ComPLement bit.
The data expansion or data compression are determined by the state of bit MODE[0].
The data compression is applied if the audio block is configured as a transmitter.
The data expansion is automatically applied when the audio block is configured as a receiver.
Note: Companding mode is applicable only when TDM is selected.
Bit 13 CPL: Complement bit. This bit is set and cleared by software.
It defines the type of complement to be used for companding mode
0: 1’s complement representation.
1: 2’s complement representation.
Note: This bit has effect only when the companding mode is µ-Law algorithm or A-Law algorithm.
Bits 12:7 MUTECNT[5:0]: Mute counter. These bits are set and cleared by software.
These bits are used only in reception mode.
The value set in these bits is compared to the number of consecutive mute frames detected in
reception. When the number of mute frames is equal to this value, the flag MUTEDET will be set and
an interrupt will be generated if bit MUTEDETIE is set.
RM0090 Rev 19 953/1751

964
Bit 6 MUTEVAL: Mute value. This bit is set and cleared by software.This bit has to be written before
enabling the audio block: SAIxEN.
0: Bit value 0 is sent during the MUTE mode.
1: Last values are sent during the MUTE mode.
This bit has a meaning only when the audio block is a transmitter and when the number of slots is
lower or equal to 2 and if the MUTE bit is set.
If more slots are declared, the bit value sent during the transmission in mute mode will be equal to 0,
whatever the value of this MUTEVAL bit.
if the number of slot is lower or equal to 2 and MUTEVAL = 1, the mute value transmitted for each
slot will be the ones sent during the previous frame.
Bit 5 MUTE: Mute. This bit is set and cleared by software.

0: No Mute mode.
1: Mute mode enabled.
This bit has a meaning only when the audio block is a transmitter. The MUTE value is linked to the
MUTEVAL value if the number of slots is lower or equal to 2, or equal to 0 if it is greater than 2.
Bit 4 TRIS: Tristate management on data line. This bit is set and cleared by software.
0: SD output line is still driven by the SAI when a slot is inactive.
1: SD output line is released (HI-Z) at the end of the last data bit of the last active slot if the next one
is inactive.
This bit has a meaning only if the audio block is configured to be a transmitter.
This bit should be configured when SAI is disabled.
Bit 3 FFLUSH: FIFO flush. This bit is set by software. It is always read low.
0: No FIFO flush.
1: FIFO flush.
Writing 1 to the bit triggers the FIFO Flush. All the internal FIFO pointers (read and write) are
cleared.
Data still present in the FIFO will be lost in such case (no more transmission or received data lost).
This bit should be configured when SAI is disabled.
Before flushing SAI, DMA stream/interruption must be disabled
Bits 2:0 FTH: FIFO threshold. This bit is set and cleared by software.
000: FIFO empty
001: ¼ FIFO
010: ½ FIFO
011: ¾ FIFO
100: FIFO full
101: Reserved
110: Reserved
111: Reserved
954/1751 RM0090 Rev 19

29.17.3 SAI xFrame configuration register (SAI_XFRCR) where x is A or B

Address offset: Block A: 0x00C
Address offset: Block B: 0x02C
Note: This register has no meaning in AC’97 audio protocol
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
FSOFF FSPOL FSDEF

Reserved
rw rw r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FSALL[6:0] FRL[7:0]
Res.

Bit 18 FSOFF: Frame synchronization offset. This bit is set and cleared by software.
0: FS is asserted on the first bit of the slot 0.
1: FS is asserted one bit before the first bit of the slot 0.
This bit has no meaning and is not used in AC’97 audio block configuration.
This bit must be configured when the audio block is disabled.
Bit 17 FSPOL: Frame synchronization polarity. This bit is set and cleared by software
0: FS is active low (falling edge)
1: FS is active high (rising edge)
This bit is used to configure the level of the start of frame on the FS signal.
Bit 16 FSDEF: Frame synchronization definition. This bit is set and cleared by software.
0: FS signal is a start frame signal
1: FS signal is a start of frame signal + channel side identification
When the bit is set, the number of slots defined in the SAI_ASLOTR register has to be even. It
means that there will be half of this number of slots dedicated for the left channel and the other slots
for the right channel (e.g: this bit has to be set for I2S or MSB/LSB-justified protocols...)
RM0090 Rev 19 955/1751

964

Bits 14:8 FSALL[6:0]: Frame synchronization active level length. These bits are set and cleared by software
The value set in these bits specifies the length in number of bit clock (SCK) + 1 (FSALL[6:0] + 1) of
the active level of the FS signal in the audio frame
These bits have no meaning and are not used in AC’97 audio block configuration.
These bits must be configured when the audio block is disabled.
Bits 7:0 FRL[7:0]: Frame length. These bits are set and cleared by software.
They define the length of the audio frame. More precisely, these bits define the number of SCK
clocks for each audio frame.
The number of bits in the frame is equal to FRL[7:0] + 1.
The minimum number of bits to transfer in an audio frame has to be equal to 8 or else the audio
block will have unexpected behavior. This is the case when the data size is 8-bit and only one slot 0
is defined in NBSLOT[4:0] in the SAI_ASLOTR register (NBSLOT[3:0] = 0000).
In master mode, if the master clock MCLK_x pin is declared as an output, the frame length should
be aligned to a number equal to a power of 2, from 8 to 256 in order to keep in an audio frame, an
integer number of MCLK pulses by bit clock for correct operation for external DAC/ADC inside the
decoders.
The Frame length should be even.
These bits have no meaning and are not used in AC’97 audio block configuration.
956/1751 RM0090 Rev 19

29.17.4 SAI xSlot register (SAI_xSLOTR) where x is A or B

Note: This register has no meaning in AC’97 audio protocol
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
SLOTEN[15:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
NBSLOT[3:0] SLOTSZ[1:0] FBOFF[4:0]

Reserved Res
Bits 31:16 SLOTEN[15:0]: Slot enable. These bits are set and cleared by software.
Each bit of the SLOTEN bits identify a slot position from 0 to 15 (maximum 16 slots)
0: Inactive slot.
1: Active slot.
These bits must be set when the audio block is disabled.
They are ignored in AC’97 mode.
Bits 11:8 NBSLOT[3:0]: Number of slots in an audio frame. These bits are set and cleared by software.
The value set in these bits register represents the number of slots + 1 in the audio frame (including
the number of inactive slots). The maximum number of slots is 16.
The number of slots should be even if bit FSDEF in the SAI_AFRCR register is set.
If the size is greater than the data size, the remaining bits will be forced to 0 if bit TRIS in the
SAI_xCR1 register is clear, otherwise they will be forced to 0 if the next slot is active or the SD line
will be forced to HI-Z if the next slot is inactive and bit TRIS = 1.
They are ignored in AC’97 omode.
Bits 7:6 SLOTSZ[1:0]: Slot size
This bits is set and cleared by software.
00: The slot size is equivalent to the data size (specified in DS[3:0] in the SAI_ACR1 register).
01: 16-bit
10: 32-bit
11: Reserved
The slot size must be greater or equal to the data size. If this condition is not respected, the behavior
of the SAI will be undetermined.
Bits 4:0 FBOFF[4:0]: First bit offset
The value set in these bits represents the position of the first data transfer bit in the slot. It represents
an offset value. During this offset phase 0 value are sent on the data line for transmission mode. For
reception mode, the received bit are discarded during the offset phase.
RM0090 Rev 19 957/1751

964
29.17.5 SAI xInterrupt mask register (SAI_xIM) where x is A or B

Address offset: blockA: 0x014
Address offset: block B: 0x034
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MUT
LFSDETI AFSDET CNRDY FREQI WCKC OVRU
EDET
Reserved E IE IE E FGIE DRIE
IE

Bit 6 LFSDETIE: Late frame synchronization detection interrupt enable. This bit is set and cleared by
software.
0: Interrupt is disabled
1: Interrupt is enabled
When this bit is set, an interrupt will be generated if the LFSDET bit is set in the SAI_ASR register.
This bit has no meaning in AC’97 mode. It has no meaning also if the audio block is master.
Bit 5 AFSDETIE: Anticipated frame synchronization detection interrupt enable. This bit is set and cleared by
software.
When this bit is set, an interrupt will be generated if the AFSDET bit in the SAI_ASR register is set.
This bit has no meaning in AC’97 mode. It has no meaning also if the audio block is master.
Bit 4 CNRDYIE: Codec not ready interrupt enable (ac’97). This bit is set and cleared by software.
When the interrupt is enabled, the audio block will detect in the slot 0 (tag0) of the AC’97 frame if the
codec connected on this line is ready or not. If not, the flag CNRDY in the SAI_ASR register will be
set and an interruption will be generated.
This bit has a meaning only if the AC97 mode is selected (bit PRTCFG[1:0]) and the audio block is a
receiver.
Bit 3 FREQIE: FIFO request interrupt enable. This bit is set and cleared by software.
When this bit is set, an interrupt will be generated if the FREQ bit in the SAI_ASR register is set.
In receiver mode, the bit MODE must be configured before setting bit FREQIE to avoid a parasitic
interruption since the audio block is a transmitter (default setting).
958/1751 RM0090 Rev 19

Bit 2 WCKCFGIE: Wrong clock configuration interrupt enable. This bit is set and cleared by software.
This bit is considered only if the audio block is configured as master (MODE[1] = 0 in the SAI_ACR1
register) and bit NODIV = 0 in the SAI_xCR1 register.
It generates an interrupt if the flag WCKCFG in the SAI_ASR register is set.
Note: This bit is used only in TDM mode and has no meaning for other modes.
Bit 1 MUTEDETIE: Mute detection interrupt enable. This bit is set and cleared by software.
When this bit is set, an interrupt will be generated if the MUTEDET bit in the SAI_ASR register is set.
This bit has a meaning only if the audio block is configured in receiver mode.
Bit 0 OVRUDRIE: Overrun/underrun interrupt enable. This bit is set and cleared by software.
When this bit is set, an interrupt will be generated if the OVRUDR bit in the SAI_ASR register is set.
RM0090 Rev 19 959/1751

964
29.17.6 SAI xStatus register (SAI_xSR) where x is A or B

Address offset: block A: 0x018
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
FLTH
Reserved
r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MUTED
LFSDET AFSDET CNRDY FREQ WCKCFG OVRUDR
Reserved ET
r r r r r r r

Bits 18:16 FLTH: FIFO level threshold. This bit is read only. The FIFO level threshold flag is managed only by
hardware and its setting depends on SAI block configuration (transmitter or receiver mode).
If SAI block is configured as transmitter:
000: FIFO_empty
001: FIFO <= ¼ but not empty
010: ¼ < FIFO <= ½
011: ½ < FIFO <= ¾
100: ¾ < FIFO but not full
101: FIFO full
If SAI block is configured as receiver:
000: FIFO_empty
001: FIFO < ¼ but not empty
010: ¼ <= FIFO < ½
011: ½ =< FIFO < ¾
100: ¾ =< FIFO but not full
101: FIFO full
Bit 6 LFSDET: Late frame synchronization detection. This bit is read only.
0: No error.
1: Frame synchronization signal is not present at the right time.
This flag can be set only if the audio block is configured in Slave mode.
It is not used in AC’97 mode.
It may generate an interrupt if bit LFSDETIE in the SAI_xIM register is set.
This flag is cleared when the software sets bit CLFSDET in the SAI_xCLRFR register
Bit 5 AFSDET: Anticipated frame synchronization detection. This bit is read only.
0: No error.
1: Frame synchronization signal is detected earlier than expected.
This flag can be set only if the audio block is configured in Slave mode.
It is not used in AC’97.
It may generate an interrupt if bit AFSDETIE in the SAI_xIM register is set.
This flag is cleared when the software sets bit CAFSDET in the SAI_xCLRFR register
960/1751 RM0090 Rev 19

Bit 4 CNRDY: Codec not ready. This bit is read only.

0: The external AC’97 codec is ready
1: The external AC’97 codec is not ready
This bit is used only when the AC’97 audio protocol is selected in the SAI_xCR1 register and is
configured in receiver mode.
It may generate an interrupt if bit CNRDYIE in the SAI_xIM register is set.
This flag is cleared when the software sets bit CCNRDY in the SAI_xCLRFR register
Bit 3 FREQ: FIFO request. This bit is read only.
0: No FIFO request.
1: FIFO request to read or to write the SAI_xDR.
The request depends on the audio block configuration.
If configured in transmission, the FIFO request concerns a write request operation in the SAI_xDR.
If configured in reception, the FIFO request concerns a read request operation from the SAI_xDR.
This flag can generate an interrupt if bit FREQIE in the SAI_xIM register is set.
Bit 2 WCKCFG: Wrong clock configuration flag. This bit is read only.
0: The clock configuration is correct
1: The clock configuration does not respect the rule concerning the frame length specification
defined in Section 29.7 (configuration of FRL[7:0] bit in the SAI_x FRCR register)
This bit is used only when the audio block is master (MODE[1] = 0 in the SAI_xCR1 register) and
when NODIV = 0 in the SAI_xCR1 register.
It may generate an interrupt if bit WCKCFGIE in the SAI_xIM register is set.
This flag is cleared when the software sets bit CWCKCFG in the SAI_xCLRFR register
Bit 1 MUTEDET: Mute detection. This bit is read only.
0: No MUTE detection on the SD input line
1: MUTE value detected on the SD input line (0 value) for a specified number of consecutive audio
frame
This flag is set if consecutive 0 values are received in each slot of an audio frame and for a
consecutive number of audio frames (set in the MUTECNT bit in the SAI_xCR2 register).
It may generate an interrupt if bit MUTEDETIE in the SAI_xIM register is set.
This flag is cleared when the software sets bit CMUTEDET in the SAI_xCLRFR register.
Bit 0 OVRUDR: Overrun / underrun. This bit is read only.
0: No overrun/underrun error.
1: Overrun/underrun error detection.
The overrun condition can occur only when the audio block is configured in reception.
The underrun condition can occur only when the audio block is configured in transmission.
It may generate an interrupt if bit OVRUDRIE in the SAI_xIM register is set.
This flag is cleared when the software set bit COVRUDR bit in the SAI_xCLRFR register.
RM0090 Rev 19 961/1751

964
29.17.7 SAI xClear flag register (SAI_xCLRFR) where X is A or B

Address offset: block A: 0x01C
Address offset: block B: 0x03C
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CAFSDE CMUTE COVRUD
CLFSDET CCNRDY CWCKCFG
Reserved T Reserved DET R
rw rw rw rw rw rw

Bit 6 CLFSDET: Clear late frame synchronization detection flag. This bit is write only.
Writing 1 in this bit clears the flag LFSDET in the SAI_xSR register.
Reading this bit always returns the value 0.
Bit 5 .CAFSDET: Clear anticipated frame synchronization detection flag. This bit is write only.
Writing 1 in this bit clears the flag AFSDET in the SAI_xSR register.
Bit 4 CCNRDY: Clear codec not ready flag. This bit is write only.
Writing 1 in this bit clears the flag CNRDY in the SAI_xSR register.
This bit is used only when the AC’97 audio protocol is selected in the SAI_xCR1 register.
Bit 2 CWCKCFG: Clear wrong clock configuration flag. This bit is write only.
Writing 1 in this bit clears the flag WCKCFG in the SAI_xSR register.
This bit is used only when the audio block is set as master (MODE[1] = 0 in the SAI_ACR1 register)
and bit NODIV = 0 in the SAI_xCR1 register.
Bit 1 CMUTEDET: Mute detection flag. This bit is write only.
Writing 1 in this bit clears the flag MUTEDET in the SAI_xSR register.
Bit 0 COVRUDR: Clear overrun / underrun. This bit is write only.
Writing 1 in this bit clears the flag OVRUDR in the SAI_xSR register.
962/1751 RM0090 Rev 19

29.17.8 SAI xData register (SAI_xDR) where x is A or B

Address offset: block A: 0x020
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DATA[31:16]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DATA[15:0]
Bits 31:0 DATA[31:0]: Data

A write into this register has the effect of loading the FIFO if the FIFO is not full.
A read from this register has to effect of draining-up the FIFO if the FIFO is not empty.
29.17.9 SAI register map

The following table summarizes the SAI registers.
Table 132. SAI register map and reset values
Register
Offset and reset
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
value
MCJDIV[3:0]
LSBFIRST
DMAEN
Reserved.
SAIxEN
CKSTR
Reserved
NODIV
MONO
0x0004
OutDri
SYNCE PRTCF MODE[

SAI_xCR1 DS[2:0]
Res.
Res.
or N[1:0] G[1:0] 1:0]

0x0024
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0
MUTE VAL
Reserved
FFLUS
MUTE
COMP[
TRIS
FTH
CPL
0x0008 SAI_xCR2 MUTECN[5:0]

1:0]
or 0x0028
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
FSDEF
Reserved
FSOFF
Reserved
FSPOL
0x000Cor SAI_xFRCR FSALL[6:0] FRL[7:0]

0x002C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1
SLOTS
Reserved
AFSDETIE Reserved
0x0010 or SAI_xSLOTR SLOTEN[15:0] NBSLOT[3:0]

Z[1:0}
FBOFF[4:0]
0x0030 Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
OVRUDRIE
MUTEDET
CNRDYIE
WCKCFG
LFSDET
FREQIE
Reserved
0x0014 SAI_xIM
or 0x0034
MUTEDET
WCKCFG
OVRUDR
FLVL[2:0]
AFSDET
LFSDET
CNRDY
Reserved
Reserved
0x0018
FREQ
SAI_xSR
or
0x0038
Reset value 0 0 0 0 0 0 0 1 0 0
RM0090 Rev 19 963/1751

964
Table 132. SAI register map and reset values (continued)
Register
Offset and reset
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
value
MUTEDET
CAFSDET
WCKCFG
OVRUDR
LFSDET
CNRDY
Reserved
0x001C SAI_xCLRFR
Res.
or 0x003C
0x0020 SAI_xDR DATA[31:0]
or 0x0040 Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
964/1751 RM0090 Rev 19

RM0090 Universal synchronous asynchronous receiver transmitter (USART)
30 Universal synchronous asynchronous receiver

transmitter (USART)
30.1 USART introduction

The universal synchronous asynchronous receiver transmitter (USART) offers a flexible
means of full-duplex data exchange with external equipment requiring an industry standard
NRZ asynchronous serial data format. The USART offers a very wide range of baud rates
using a fractional baud rate generator.
It supports synchronous one-way communication and half-duplex single wire
communication. It also supports the LIN (local interconnection network), Smartcard Protocol
and IrDA (infrared data association) SIR ENDEC specifications, and modem operations
(CTS/RTS). It allows multiprocessor communication.
High speed data communication is possible by using the DMA for multibuffer configuration.
30.2 USART main features

• Full duplex, asynchronous communications
• NRZ standard format (Mark/Space)
• Configurable oversampling method by 16 or by 8 to give flexibility between speed and
clock tolerance
• Fractional baud rate generator systems
– Common programmable transmit and receive baud rate (refer to the datasheets
for the value of the baud rate at the maximum APB frequency.
• Programmable data word length (8 or 9 bits)
• Configurable stop bits - support for 1 or 2 stop bits
• LIN Master Synchronous Break send capability and LIN slave break detection
capability
– 13-bit break generation and 10/11 bit break detection when USART is hardware
configured for LIN
• Transmitter clock output for synchronous transmission
• IrDA SIR encoder decoder
– Support for 3/16 bit duration for normal mode
• Smartcard emulation capability
– The Smartcard interface supports the asynchronous protocol Smartcards as
defined in the ISO 7816-3 standards
– 0.5, 1.5 stop bits for Smartcard operation
• Single-wire half-duplex communication
• Configurable multibuffer communication using DMA (direct memory access)
– Buffering of received/transmitted bytes in reserved SRAM using centralized DMA
• Separate enable bits for transmitter and receiver
RM0090 Rev 19 965/1751

1018
Universal synchronous asynchronous receiver transmitter (USART) RM0090
• Transfer detection flags:

– Receive buffer full
– Transmit buffer empty
– End of transmission flags
• Parity control:
– Transmits parity bit
– Checks parity of received data byte
• Four error detection flags:
– Overrun error
– Noise detection
– Frame error
– Parity error
• Ten interrupt sources with flags:
– CTS changes
– LIN break detection
– Transmit data register empty
– Transmission complete
– Receive data register full
– Idle line received
– Overrun error
– Framing error
– Noise error
– Parity error
• Multiprocessor communication - enter into mute mode if address match does not occur
• Wake up from mute mode (by idle line detection or address mark detection)
• Two receiver wakeup modes: Address bit (MSB, 9th bit), Idle line
30.3 USART functional description

The interface is externally connected to another device by three pins (see Figure 296). Any
USART bidirectional communication requires a minimum of two pins: Receive Data In (RX)
and Transmit Data Out (TX):
RX: Receive Data Input is the serial data input. Oversampling techniques are used for data
recovery by discriminating between valid incoming data and noise.
TX: Transmit Data Output. When the transmitter is disabled, the output pin returns to its I/O
port configuration. When the transmitter is enabled and nothing is to be transmitted, the TX
pin is at high level. In single-wire and smartcard modes, this I/O is used to transmit and
receive the data (at USART level, data are then received on SW_RX).
966/1751 RM0090 Rev 19

Through these pins, serial data is transmitted and received in normal USART mode as
frames comprising:
• An Idle Line prior to transmission or reception
• A start bit
• A data word (8 or 9 bits) least significant bit first
• 0.5,1, 1.5, 2 Stop bits indicating that the frame is complete
• This interface uses a fractional baud rate generator - with a 12-bit mantissa and 4-bit
fraction
• A status register (USART_SR)
• Data Register (USART_DR)
• A baud rate register (USART_BRR) - 12-bit mantissa and 4-bit fraction.
• A Guardtime Register (USART_GTPR) in case of Smartcard mode.
Refer to Section 30.6: USART registers on page 1007 for the definitions of each bit.
The following pin is required to interface in synchronous mode:
• CK: Transmitter clock output. This pin outputs the transmitter data clock for
synchronous transmission corresponding to SPI master mode (no clock pulses on start
bit and stop bit, and a software option to send a clock pulse on the last data bit). In
parallel data can be received synchronously on RX. This can be used to control
peripherals that have shift registers (e.g. LCD drivers). The clock phase and polarity
are software programmable. In smartcard mode, CK can provide the clock to the
smartcard.
The following pins are required in Hardware flow control mode:
• CTS: Clear To Send blocks the data transmission at the end of the current transfer
when high
• RTS: Request to send indicates that the USART is ready to receive a data (when low).
RM0090 Rev 19 967/1751

1018
Figure 296. USART block diagram

PWDATA PRDATA
Write Read (Data register) DR
(CPU or DMA) (CPU or DMA)
Transmit data register (TDR) Receive data register (RDR)

TX
RX IrDA
SIR
SW_RX Receive Shift Register
ENDEC Transmit Shift Register
block
IRDA_OUT
IRDA_IN GTPR
GT PSC SCLK control CK
CR3 CR2
DMAT DMAR SCEN NACK HD IRLP IREN LINE STOP[1:0] CKEN CPOL CPHA LBCL
CR2 CR1
USART Address UE M WAKE PCE PS PEIE
RTS Hardware
flow
CTS controller
Wakeup Receiver
Transmit Receiver clock
control unit
control
CR1 SR
TXEIE TCIE RXNE
IE
IDLE TE RE RWU SBK CTS LBD TXE TC RXNE IDLE ORE NF FE PE
IE
USART
interrupt
control
CR1 USART_BRR
OVER8
TE Transmitter rate
Transmitter control
clock
/ [8 x (2 - OVER8)] /USARTDIV
SAMPLING
DIVIDER DIV_Mantissa DIV_Fraction
15 4 0
fPCLKx(x=1,2)
Receiver rate
RE control
Conventional baud rate generator
USARTDIV = DIV_Mantissa + (DIV_Fraction / 8 × (2 – OVER8)) ai16099b
968/1751 RM0090 Rev 19

30.3.1 USART character description

Word length may be selected as being either 8 or 9 bits by programming the M bit in the
USART_CR1 register (see Figure 297).
The TX pin is in low state during the start bit. It is in high state during the stop bit.
An Idle character is interpreted as an entire frame of “1”s followed by the start bit of the
next frame which contains data (The number of “1” ‘s will include the number of stop bits).
A Break character is interpreted on receiving “0”s for a frame period. At the end of the
break frame the transmitter inserts either 1 or 2 stop bits (logic “1” bit) to acknowledge the
start bit.
Transmission and reception are driven by a common baud rate generator, the clock for each
is generated when the enable bit is set respectively for the transmitter and receiver.
The details of each block is given below.
Figure 297. Word length programming
9-bit word length (M bit is set), 1 Stop bit

Possible
Data frame Parity Next data frame
bit
Next
Start Stop Start
bit Bit0 Bit1 Bit2 Bit3 Bit4 Bit5 Bit6 Bit7 Bit8 bit bit
Clock **
Start
Idle frame bit
Break frame Stop Start

bit bit
** LBCL bit controls last data clock pulse
8-bit word length (M bit is reset), 1 Stop bit

Possible
Data frame Parity Next data frame
bit
Next
Start Stop Start
bit Bit0 Bit1 Bit2 Bit3 Bit4 Bit5 Bit6 Bit7
bit bit
Clock **
Start
Idle frame bit
Break frame Stop Start

bit bit

MS19822V2
RM0090 Rev 19 969/1751

1018
30.3.2 Transmitter
The transmitter can send data words of either 8 or 9 bits depending on the M bit status.
When the transmit enable bit (TE) is set, the data in the transmit shift register is output on
the TX pin and the corresponding clock pulses are output on the CK pin.
Character transmission
During an USART transmission, data shifts out least significant bit first on the TX pin. In this
mode, the USART_DR register consists of a buffer (TDR) between the internal bus and the
transmit shift register (see Figure 296).
Every character is preceded by a start bit which is a logic level low for one bit period. The
character is terminated by a configurable number of stop bits.
The following stop bits are supported by USART: 0.5, 1, 1.5 and 2 stop bits.
Note: The TE bit should not be reset during transmission of data. Resetting the TE bit during the
transmission will corrupt the data on the TX pin as the baud rate counters will get frozen.
The current data being transmitted will be lost.
An idle frame will be sent after the TE bit is enabled.
Configurable stop bits

The number of stop bits to be transmitted with every character can be programmed in
Control register 2, bits 13,12.
• 1 stop bit: This is the default value of number of stop bits.
• 2 Stop bits: This will be supported by normal USART, single-wire and modem modes.
• 0.5 stop bit: To be used when receiving data in Smartcard mode.
• 1.5 stop bits: To be used when transmitting and receiving data in Smartcard mode.
An idle frame transmission will include the stop bits.
A break transmission will be 10 low bits followed by the configured number of stop bits
(when m = 0) and 11 low bits followed by the configured number of stop bits (when m = 1). It
is not possible to transmit long breaks (break of length greater than 10/11 low bits).
970/1751 RM0090 Rev 19

Figure 298. Configurable stop bits

8-bit Word length (M bit is reset)
Possible
parity Next data frame
Data frame bit Next
Start start
Bit Bit0 Bit1 Bit2 Bit3 Bit4 Bit5 Bit6 Bit7 Stop
bit
CLOCK ****
**
a) 1 Stop Bit
Possible
Parity Next data frame
Data frame Bit Next
Start start
Bit Bit0 Bit1 Bit2 Bit3 Bit4 Bit5 Bit6 Bit7 bit
1 1/2 stop bits

b) 1 1/2 stop Bits
Possible
parity Next data frame
Data frame bit Next
Start Start
Bit Bit0 Bit1 Bit2 Bit3 Bit4 Bit5 Bit6 Bit7 2 Stop Bits Bit
c) 2 Stop Bits Possible

parity
Data frame Next data frame
bit Next
Start Start
Bit Bit0 Bit1 Bit2 Bit3 Bit4 Bit5 Bit6 Bit7 Bit
d) 1/2 Stop Bit 1/2 Stop Bit
MSv42088V1
Procedure:
1. Enable the USART by writing the UE bit in USART_CR1 register to 1.
2. Program the M bit in USART_CR1 to define the word length.
3. Program the number of stop bits in USART_CR2.
4. Select DMA enable (DMAT) in USART_CR3 if Multi buffer Communication is to take
place. Configure the DMA register as explained in multibuffer communication.
5. Select the desired baud rate using the USART_BRR register.
6. Set the TE bit in USART_CR1 to send an idle frame as first transmission.
7. Write the data to send in the USART_DR register (this clears the TXE bit). Repeat this
for each data to be transmitted in case of single buffer.
8. After writing the last data into the USART_DR register, wait until TC=1. This indicates
that the transmission of the last frame is complete. This is required for instance when
the USART is disabled or enters the Halt mode to avoid corrupting the last
transmission.
Single byte communication

Clearing the TXE bit is always performed by a write to the data register.
The TXE bit is set by hardware and it indicates:
• The data has been moved from TDR to the shift register and the data transmission has
started.
• The TDR register is empty.
• The next data can be written in the USART_DR register without overwriting the
previous data.
This flag generates an interrupt if the TXEIE bit is set.
RM0090 Rev 19 971/1751

1018
When a transmission is taking place, a write instruction to the USART_DR register stores
the data in the TDR register and which is copied in the shift register at the end of the current
transmission.
When no transmission is taking place, a write instruction to the USART_DR register places
the data directly in the shift register, the data transmission starts, and the TXE bit is
immediately set.
If a frame is transmitted (after the stop bit) and the TXE bit is set, the TC bit goes high. An
interrupt is generated if the TCIE bit is set in the USART_CR1 register.
After writing the last data into the USART_DR register, it is mandatory to wait for TC=1
before disabling the USART or causing the microcontroller to enter the low-power mode
(see Figure 299: TC/TXE behavior when transmitting).
The TC bit is cleared by the following software sequence:
1. A read from the USART_SR register
2. A write to the USART_DR register
Note: The TC bit can also be cleared by writing a ‘0 to it. This clearing sequence is recommended
only for Multibuffer communication.
Figure 299. TC/TXE behavior when transmitting
Idle preamble Frame 1 Frame 2 Frame 3

TX line

TXE flag cleared by software cleared by software Set by hardware
USART_DR F1 F2 F3
TC flag set
by hardware
Software
enables the Software waits until TXE=1 TC is not set TC is not set TC is not set
USART and writes F2 into DR because TXE=0 because TXE=0 because TXE=1
Software waits until TXE=1 Software waits until TXE=1 Software wait until TC=1
and writes F1 into DR and writes F1 into DR
MS48961V1
Break characters
Setting the SBK bit transmits a break character. The break frame length depends on the M
bit (see Figure 297).
If the SBK bit is set to ‘1 a break character is sent on the TX line after completing the current
character transmission. This bit is reset by hardware when the break character is completed
(during the stop bit of the break character). The USART inserts a logic 1 bit at the end of the
last break frame to guarantee the recognition of the start bit of the next frame.
Note: If the software resets the SBK bit before the commencement of break transmission, the
break character will not be transmitted. For two consecutive breaks, the SBK bit should be
set after the stop bit of the previous break.
972/1751 RM0090 Rev 19

Idle characters
Setting the TE bit drives the USART to send an idle frame before the first data frame.
30.3.3 Receiver
The USART can receive data words of either 8 or 9 bits depending on the M bit in the
USART_CR1 register.
Start bit detection

The start bit detection sequence is the same when oversampling by 16 or by 8.
In the USART, the start bit is detected when a specific sequence of samples is recognized.
This sequence is: 1 1 1 0 X 0 X 0 X 0 0 0 0.
Figure 300. Start bit detection when oversampling by 16 or 8
RX state
Idle Start bit
RX line
Ideal
sample 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
clock
Sampled values
Real
sample X X X X X X X X 9 10 11 12 13 14 15 16
clock
6/16
7/16 7/16
One-bit time
Conditions
to validate 1 1 1 0 X 0 X 0 X 0 0 0 0 X X X X X X
the start bit
Falling edge At least 2 bits At least 2 bits
detection out of 3 at 0 out of 3 at 0
ai15471b
Note: If the sequence is not complete, the start bit detection aborts and the receiver returns to the
idle state (no flag is set) where it waits for a falling edge.
The start bit is confirmed (RXNE flag set, interrupt generated if RXNEIE=1) if the 3 sampled
bits are at 0 (first sampling on the 3rd, 5th and 7th bits finds the 3 bits at 0 and second
sampling on the 8th, 9th and 10th bits also finds the 3 bits at 0).
The start bit is validated (RXNE flag set, interrupt generated if RXNEIE=1) but the NE noise
flag is set if, for both samplings, at least 2 out of the 3 sampled bits are at 0 (sampling on the
RM0090 Rev 19 973/1751

1018
3rd, 5th and 7th bits and sampling on the 8th, 9th and 10th bits). If this condition is not met,
the start detection aborts and the receiver returns to the idle state (no flag is set).
If, for one of the samplings (sampling on the 3rd, 5th and 7th bits or sampling on the 8th, 9th
and 10th bits), 2 out of the 3 bits are found at 0, the start bit is validated but the NE noise
flag bit is set.
Character reception
During an USART reception, data shifts in least significant bit first through the RX pin. In this
mode, the USART_DR register consists of a buffer (RDR) between the internal bus and the
received shift register.
Procedure:
1. Enable the USART by writing the UE bit in USART_CR1 register to 1.
2. Program the M bit in USART_CR1 to define the word length.
3. Program the number of stop bits in USART_CR2.
4. Select DMA enable (DMAR) in USART_CR3 if multibuffer communication is to take
place. Configure the DMA register as explained in multibuffer communication. STEP 3
5. Select the desired baud rate using the baud rate register USART_BRR
6. Set the RE bit USART_CR1. This enables the receiver which begins searching for a
start bit.
When a character is received
• The RXNE bit is set. It indicates that the content of the shift register is transferred to the
RDR. In other words, data has been received and can be read (as well as its
associated error flags).
• An interrupt is generated if the RXNEIE bit is set.
• The error flags can be set if a frame error, noise or an overrun error has been detected
during reception.
• In multibuffer, RXNE is set after every byte received and is cleared by the DMA read to
the Data Register.
• In single buffer mode, clearing the RXNE bit is performed by a software read to the
USART_DR register. The RXNE flag can also be cleared by writing a zero to it. The
RXNE bit must be cleared before the end of the reception of the next character to avoid
an overrun error.
Note: The RE bit should not be reset while receiving data. If the RE bit is disabled during
reception, the reception of the current byte will be aborted.
Break character
When a break character is received, the USART handles it as a framing error.
Idle character
When an idle frame is detected, there is the same procedure as a data received character
plus an interrupt if the IDLEIE bit is set.
974/1751 RM0090 Rev 19

Overrun error
An overrun error occurs when a character is received when RXNE has not been reset. Data
can not be transferred from the shift register to the RDR register until the RXNE bit is
cleared.
The RXNE flag is set after every byte received. An overrun error occurs if RXNE flag is set
when the next data is received or the previous DMA request has not been serviced. When
an overrun error occurs:
• The ORE bit is set.
• The RDR content will not be lost. The previous data is available when a read to
USART_DR is performed.
• The shift register will be overwritten. After that point, any data received during overrun
is lost.
• An interrupt is generated if either the RXNEIE bit is set or both the EIE and DMAR bits
are set.
• The ORE bit is reset by a read to the USART_SR register followed by a USART_DR
register read operation.
Note: The ORE bit, when set, indicates that at least 1 data has been lost. There are two
possibilities:
• if RXNE=1, then the last valid data is stored in the receive register RDR and can be
read,
• if RXNE=0, then it means that the last valid data has already been read and thus there
is nothing to be read in the RDR. This case can occur when the last valid data is read in
the RDR at the same time as the new (and lost) data is received. It may also occur
when the new data is received during the reading sequence (between the USART_SR
register read access and the USART_DR read access).
Selecting the proper oversampling method

The receiver implements different user-configurable oversampling techniques (except in
synchronous mode) for data recovery by discriminating between valid incoming data and
noise.
The oversampling method can be selected by programming the OVER8 bit in the
USART_CR1 register and can be either 16 or 8 times the baud rate clock (Figure 301 and
Figure 302).
Depending on the application:
• select oversampling by 8 (OVER8=1) to achieve higher speed (up to fPCLK/8). In this
case the maximum receiver tolerance to clock deviation is reduced (refer to
Section 30.3.5: USART receiver tolerance to clock deviation on page 988)
• select oversampling by 16 (OVER8=0) to increase the tolerance of the receiver to clock
deviations. In this case, the maximum speed is limited to maximum fPCLK/16
RM0090 Rev 19 975/1751

1018
Programming the ONEBIT bit in the USART_CR3 register selects the method used to
evaluate the logic level. There are two options:
• the majority vote of the three samples in the center of the received bit. In this case,
when the 3 samples used for the majority vote are not equal, the NF bit is set
• a single sample in the center of the received bit
Depending on the application:
– select the three samples’ majority vote method (ONEBIT=0) when operating in a
noisy environment and reject the data when a noise is detected (refer to
Figure 133) because this indicates that a glitch occurred during the sampling.
– select the single sample method (ONEBIT=1) when the line is noise-free to
increase the receiver’s tolerance to clock deviations (see Section 30.3.5: USART
receiver tolerance to clock deviation on page 988). In this case the NF bit will
never be set.
When noise is detected in a frame:
• The NF bit is set at the rising edge of the RXNE bit.
• The invalid data is transferred from the Shift register to the USART_DR register.
• No interrupt is generated in case of single byte communication. However this bit rises
at the same time as the RXNE bit which itself generates an interrupt. In case of
multibuffer communication an interrupt will be issued if the EIE bit is set in the
USART_CR3 register.
The NF bit is reset by a USART_SR register read operation followed by a USART_DR
Note: Oversampling by 8 is not available in the Smartcard, IrDA and LIN modes. In those modes,
the OVER8 bit is forced to ‘0 by hardware.
Figure 301. Data sampling when oversampling by 16
RX line
sampled values
Sample clock 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
6/16
7/16 7/16
One bit time
MSv31152V1
976/1751 RM0090 Rev 19

Figure 302. Data sampling when oversampling by 8
RX line
sampled values
Sample
clock (x8) 1 2 3 4 5 6 7 8
2/8
3/8 3/8
One bit time
MSv31153V1
Table 133. Noise detection from sampled data

Sampled value NE status Received bit value
000 0 0
001 1 0
010 1 0
011 1 1
100 1 0
101 1 1
110 1 1
111 0 1
Framing error
A framing error is detected when:
The stop bit is not recognized on reception at the expected time, following either a de-
synchronization or excessive noise.
When the framing error is detected:
• The FE bit is set by hardware
• The invalid data is transferred from the Shift register to the USART_DR register.
• No interrupt is generated in case of single byte communication. However this bit rises
at the same time as the RXNE bit which itself generates an interrupt. In case of
multibuffer communication an interrupt will be issued if the EIE bit is set in the
USART_CR3 register.
The FE bit is reset by a USART_SR register read operation followed by a USART_DR
RM0090 Rev 19 977/1751

1018
Configurable stop bits during reception

The number of stop bits to be received can be configured through the control bits of Control
Register 2 - it can be either 1 or 2 in normal mode and 0.5 or 1.5 in Smartcard mode.
1. 0.5 stop bit (reception in Smartcard mode): No sampling is done for 0.5 stop bit. As
a consequence, no framing error and no break frame can be detected when 0.5 stop bit
is selected.
2. 1 stop bit: Sampling for 1 stop Bit is done on the 8th, 9th and 10th samples.
3. 1.5 stop bits (Smartcard mode): When transmitting in smartcard mode, the device
must check that the data is correctly sent. Thus the receiver block must be enabled (RE
=1 in the USART_CR1 register) and the stop bit is checked to test if the smartcard has
detected a parity error. In the event of a parity error, the smartcard forces the data
signal low during the sampling - NACK signal-, which is flagged as a framing error.
Then, the FE flag is set with the RXNE at the end of the 1.5 stop bit. Sampling for 1.5
stop bits is done on the 16th, 17th and 18th samples (1 baud clock period after the
beginning of the stop bit). The 1.5 stop bit can be decomposed into 2 parts: one 0.5
baud clock period during which nothing happens, followed by 1 normal stop bit period
during which sampling occurs halfway through. Refer to Section 30.3.11: Smartcard on
page 997 for more details.
4. 2 stop bits: Sampling for 2 stop bits is done on the 8th, 9th and 10th samples of the
first stop bit. If a framing error is detected during the first stop bit the framing error flag
will be set. The second stop bit is not checked for framing error. The RXNE flag will be
set at the end of the first stop bit.
30.3.4 Fractional baud rate generation

The baud rate for the receiver and transmitter (Rx and Tx) are both set to the same value as
programmed in the Mantissa and Fraction values of USARTDIV.
Equation 1: Baud rate for standard USART (SPI mode included)

f CK
Tx/Rx baud = ----------------------------------------------------------------------------------
-
8 × ( 2 – OVER8 ) × USARTDIV
Equation 2: Baud rate in Smartcard, LIN and IrDA modes

f CK
Tx/Rx baud = ---------------------------------------------
16 × USARTDIV
USARTDIV is an unsigned fixed point number that is coded on the USART_BRR register.
• When OVER8=0, the fractional part is coded on 4 bits and programmed by the
DIV_fraction[3:0] bits in the USART_BRR register
• When OVER8=1, the fractional part is coded on 3 bits and programmed by the
DIV_fraction[2:0] bits in the USART_BRR register, and bit DIV_fraction[3] must be kept
cleared.
Note: The baud counters are updated to the new value in the baud registers after a write operation
to USART_BRR. Hence the baud rate register value should not be changed during
communication.
978/1751 RM0090 Rev 19

How to derive USARTDIV from USART_BRR register values when OVER8=0

Example 1:
If DIV_Mantissa = 0d27 and DIV_Fraction = 0d12 (USART_BRR = 0x1BC), then
Mantissa (USARTDIV) = 0d27
Fraction (USARTDIV) = 12/16 = 0d0.75
Therefore USARTDIV = 0d27.75
Example 2:
To program USARTDIV = 0d25.62
This leads to:
DIV_Fraction = 16*0d0.62 = 0d9.92
The nearest real number is 0d10 = 0xA
DIV_Mantissa = mantissa (0d25.620) = 0d25 = 0x19
Then, USART_BRR = 0x19A hence USARTDIV = 0d25.625
Example 3:
This leads to:
DIV_Fraction = 16*0d0.99 = 0d15.84
The nearest real number is 0d16 = 0x10 => overflow of DIV_frac[3:0] => carry must be
added up to the mantissa
DIV_Mantissa = mantissa (0d50.990 + carry) = 0d51 = 0x33
Then, USART_BRR = 0x330 hence USARTDIV = 0d51.000
How to derive USARTDIV from USART_BRR register values when OVER8=1

Example 1:
If DIV_Mantissa = 0x27 and DIV_Fraction[2:0]= 0d6 (USART_BRR = 0x1B6), then
Mantissa (USARTDIV) = 0d27
Fraction (USARTDIV) = 6/8 = 0d0.75
Therefore USARTDIV = 0d27.75
Example 2:
This leads to:
DIV_Fraction = 8*0d0.62 = 0d4.96
The nearest real number is 0d5 = 0x5
DIV_Mantissa = mantissa (0d25.620) = 0d25 = 0x19
RM0090 Rev 19 979/1751

1018
Then, USART_BRR = 0x195 => USARTDIV = 0d25.625

Example 3:
This leads to:
DIV_Fraction = 8*0d0.99 = 0d7.92
The nearest real number is 0d8 = 0x8 => overflow of the DIV_frac[2:0] => carry must be
added up to the mantissa
DIV_Mantissa = mantissa (0d50.990 + carry) = 0d51 = 0x33
Then, USART_BRR = 0x0330 => USARTDIV = 0d51.000
oversampling by 16(1)
Oversampling by 16 (OVER8=0)
Baud rate7 fPCLK = 8 MHz fPCLK = 12 MHz
Value % Error = Value

programmed (Calculated - programmed
S.No Desired Actual Actual % Error
in the baud Desired) B.rate / in the baud
rate register Desired B.rate rate register
1 1.2 KBps 1.2 KBps 416.6875 0 1.2 KBps 625 0

2 2.4 KBps 2.4 KBps 208.3125 0.01 2.4 KBps 312.5 0
3 9.6 KBps 9.604 KBps 52.0625 0.04 9.6 KBps 78.125 0
4 19.2 KBps 19.185 KBps 26.0625 0.08 19.2 KBps 39.0625 0
5 38.4 KBps 38.462 KBps 13 0.16 38.339 KBps 19.5625 0.16
6 57.6 KBps 57.554 KBps 8.6875 0.08 57.692 KBps 13 0.16
7 115.2 KBps 115.942 KBps 4.3125 0.64 115.385 KBps 6.5 0.16
8 230.4 KBps 228.571 KBps 2.1875 0.79 230.769 KBps 3.25 0.16
9 460.8 KBps 470.588 KBps 1.0625 2.12 461.538 KBps 1.625 0.16
10 921.6 KBps NA NA NA NA NA NA
11 2 MBps NA NA NA NA NA NA
1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.
980/1751 RM0090 Rev 19

Table 135. Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK =12 MHz,
Oversampling by 8 (OVER8 = 1)
Baud rate fPCLK = 8 MHz fPCLK = 12 MHz
% Error =
Value (Calculated - Value
programmed Desired) programmed
in the baud B.rate / in the baud
rate register Desired rate register
B.rate
1 1.2 KBps 1.2 KBps 833.375 0 1.2 KBps 1250 0

2 2.4 KBps 2.4 KBps 416.625 0.01 2.4 KBps 625 0
3 9.6 KBps 9.604 KBps 104.125 0.04 9.6 KBps 156.25 0
4 19.2 KBps 19.185 KBps 52.125 0.08 19.2 KBps 78.125 0
5 38.4 KBps 38.462 KBps 26 0.16 38.339 KBps 39.125 0.16
6 57.6 KBps 57.554 KBps 17.375 0.08 57.692 KBps 26 0.16
7 115.2 KBps 115.942 KBps 8.625 0.64 115.385 KBps 13 0.16
8 230.4 KBps 228.571 KBps 4.375 0.79 230.769 KBps 6.5 0.16
9 460.8 KBps 470.588 KBps 2.125 2.12 461.538 KBps 3.25 0.16
10 921.6 KBps 888.889 KBps 1.125 3.55 923.077 KBps 1.625 0.16

1 1.2 KBps 1.2 KBps 833.3125 0 1.2 1250 0

2 2.4 KBps 2.4 KBps 416.6875 0 2.4 625 0
3 9.6 KBps 9.598 KBps 104.1875 0.02 9.6 156.25 0
4 19.2 KBps 19.208 KBps 52.0625 0.04 19.2 78.125 0
5 38.4 KBps 38.369 KBps 26.0625 0.08 38.4 39.0625 0
6 57.6 KBps 57.554 KBps 17.375 0.08 57.554 26.0625 0.08
RM0090 Rev 19 981/1751

1018
oversampling by 16(1) (continued)

7 115.2 KBps 115.108 KBps 8.6875 0.08 115.385 13 0.16

8 230.4 KBps 231.884 KBps 4.3125 0.64 230.769 6.5 0.16
9 460.8 KBps 457.143 KBps 2.1875 0.79 461.538 3.25 0.16
10 921.6 KBps 941.176 KBps 1.0625 2.12 923.077 1.625 0.16

1 1.2 KBps 1.2 KBps 1666.625 0 1.2 KBps 2500 0

2 2.4 KBps 2.4 KBps 833.375 0 2.4 KBps 1250 0
3 9.6 KBps 9.598 KBps 208.375 0.02 9.6 KBps 312.5 0
4 19.2 KBps 19.208 KBps 104.125 0.04 19.2 KBps 156.25 0
5 38.4 KBps 38.369 KBps 52.125 0.08 38.4 KBps 78.125 0
6 57.6 KBps 57.554 KBps 34.75 0.08 57.554 KBps 52.125 0.08
7 115.2 KBps 115.108 KBps 17.375 0.08 115.385 KBps 26 0.16
8 230.4 KBps 231.884 KBps 8.625 0.64 230.769 KBps 13 0.16
9 460.8 KBps 457.143 KBps 4.375 0.79 461.538 KBps 6.5 0.16
10 921.6 KBps 941.176 KBps 2.125 2.12 923.077 KBps 3.25 0.16
11 2 MBps 2000 KBps 1 0 2000 KBps 1.5 0
12 3 MBps NA NA NA 3000 KBps 1 0
982/1751 RM0090 Rev 19

Value
% Error = Value
programme
(Calculated - programmed
S.No Desired Actual d in the Actual % Error
Desired)B.Rate in the baud
baud rate
/Desired B.Rate rate register
register
1. 2.4 KBps 2.400 KBps 208.3125 0.00% 2.400 KBps 416.6875 0.00%
2. 9.6 KBps 9.604 KBps 52.0625 0.04% 9.598 KBps 104.1875 0.02%
3. 19.2 KBps 19.185 KBps 26.0625 0.08% 19.208 KBps 52.0625 0.04%
4. 57.6 KBps 57.554 KBps 8.6875 0.08% 57.554 KBps 17.3750 0.08%
5. 115.2 KBps 115.942 KBps 4.3125 0.64% 115.108 KBps 8.6875 0.08%
228.571
6. 230.4 KBps 2.1875 0.79% 231.884 KBps 4.3125 0.64%
KBps
470.588
7. 460.8 KBps 1.0625 2.12% 457.143 KBps 2.1875 0.79%
KBps
8. 896 KBps NA NA NA 888.889 KBps 1.1250 0.79%
9. 921.6 KBps NA NA NA 941.176 KBps 1.0625 2.12%
10. 1.792 MBps NA NA NA NA NA NA

programmed (Calculated - programmed %
S.No Desired Actual Actual
in the baud Desired)B.Rate in the baud Error
rate register /Desired B.Rate rate register
1. 2.4 KBps 2.400 KBps 416.625 0.01% 2.400 KBps 833.375 0.00%
2. 9.6 KBps 9.604 KBps 104.125 0.04% 9.598 KBps 208.375 0.02%
3. 19.2 KBps 19.185 KBps 52.125 0.08% 19.208 KBps 104.125 0.04%
RM0090 Rev 19 983/1751

1018
oversampling by 8(1) (continued)

4. 57.6 KBps 57.557 KBps 17.375 0.08% 57.554 KBps 34.750 0.08%
5. 115.2 KBps 115.942 KBps 8.625 0.64% 115.108 KBps 17.375 0.08%
6. 230.4 KBps 228.571 KBps 4.375 0.79% 231.884 KBps 8.625 0.64%
7. 460.8 KBps 470.588 KBps 2.125 2.12% 457.143 KBps 4.375 0.79%
8. 896 KBps 888.889 KBps 1.125 0.79% 888.889 KBps 2.250 0.79%
9. 921.6 KBps 888.889 KBps 1.125 3.55% 941.176 KBps 2.125 2.12%
10. 1.792 MBps NA NA NA 1.7777 MBps 1.125 0.79%
11. 1.8432 MBps NA NA NA 1.7777 MBps 1.125 3.55%
oversampling by 16(1)(2)

1. 2.4 KBps 2.400 KBps 781.2500 0.00% 2.400 KBps 1562.5000 0.00%
2. 9.6 KBps 9.600 KBps 195.3125 0.00% 9.600 KBps 390.6250 0.00%
3. 19.2 KBps 19.194 KBps 97.6875 0.03% 19.200 KBps 195.3125 0.00%
4. 57.6 KBps 57.582KBps 32.5625 0.03% 57.582 KBps 65.1250 0.03%
5. 115.2 KBps 115.385 KBps 16.2500 0.16% 115.163 KBps 32.5625 0.03%
6. 230.4 KBps 230.769 KBps 8.1250 0.16% 230.769KBps 16.2500 0.16%
7. 460.8 KBps 461.538 KBps 4.0625 0.16% 461.538 KBps 8.1250 0.16%
8. 896 KBps 909.091 KBps 2.0625 1.46% 895.522 KBps 4.1875 0.05%
984/1751 RM0090 Rev 19

oversampling by 16(1)(2) (continued)

9. 921.6 KBps 909.091 KBps 2.0625 1.36% 923.077 KBps 4.0625 0.16%
10. 1.792 MBps 1.1764 MBps 1.0625 1.52% 1.8182 MBps 2.0625 1.36%
1.8432
11. 1.8750 MBps 1.0000 1.73% 1.8182 MBps 2.0625 1.52%
MBps
12. 3.584 MBps NA NA NA 3.2594 MBps 1.0625 1.52%
3.6864
13. NA NA NA 3.7500 MBps 1.0000 1.73%
MBps
7.3728
15. NA NA NA NA NA NA
MBps
2. Only USART1 and USART6 are clocked with PCLK2. Other USARTs are clocked with PCLK1. Refer to the device
datasheets for the maximum values for PCLK1 and PCLK2.
oversampling by 8(1) (2)
Baud rate fPCLK = 30 MHz fPCLK =60 MHz

1. 2.4 KBps 2.400 KBps 1562.5000 0.00% 2.400 KBps 3125.0000 0.00%
2. 9.6 KBps 9.600 KBps 390.6250 0.00% 9.600 KBps 781.2500 0.00%
3. 19.2 KBps 19.194 KBps 195.3750 0.03% 19.200 KBps 390.6250 0.00%
4. 57.6 KBps 57.582 KBps 65.1250 0.16% 57.582 KBps 130.2500 0.03%
5. 115.2 KBps 115.385 KBps 32.5000 0.16% 115.163 KBps 65.1250 0.03%
6. 230.4 KBps 230.769 KBps 16.2500 0.16% 230.769 KBps 32.5000 0.16%
7. 460.8 KBps 461.538 KBps 8.1250 0.16% 461.538 KBps 16.2500 0.16%
8. 896 KBps 909.091 KBps 4.1250 1.46% 895.522 KBps 8.3750 0.05%
9. 921.6 KBps 909.091 KBps 4.1250 1.36% 923.077 KBps 8.1250 0.16%
10. 1.792 MBps 1.7647 MBps 2.1250 1.52% 1.8182 MBps 4.1250 1.46%
RM0090 Rev 19 985/1751

1018
oversampling by 8(1) (2) (continued)
Baud rate fPCLK = 30 MHz fPCLK =60 MHz

11. 1.8432 MBps 1.8750 MBps 2.0000 1.73% 1.8182 MBps 4.1250 1.36%
12. 3.584 MBps 3.7500 MBps 1.0000 4.63% 3.5294 MBps 2.1250 1.52%
13. 3.6864 MBps 3.7500 MBps 1.0000 1.73% 3.7500 MBps 2.0000 1.73%
14. 7.168 MBps NA NA NA 7.5000 MBps 1.0000 4.63%
15. 7.3728 MBps NA NA NA 7.5000 MBps 1.0000 1.73%

1. 1.2 KBps 1.2 KBps 2187.5 0 1.2 KBps NA 0

2. 2.4 KBps 2.4 KBps 1093.75 0 2.4 KBps 2187.5 0
3. 9.6 KBps 9.6 KBps 273.4375 0 9.6 KBps 546.875 0
4. 19.2 KBps 19.195 KBps 136.75 0.02 19.2 KBps 273.4375 0
5. 38.4 KBps 38.391 KBps 68.375 0.02 38.391 KBps 136.75 0.02
6. 57.6 KBps 57.613 KBps 45.5625 0.02 57.613 KBps 91.125 0.02
7. 115.2 KBps 115.068 KBps 22.8125 0.11 115.226 KBps 45.5625 0.02
8. 230.4 KBps 230.769 KBps 11.375 0.16 230.137 KBps 22.8125 0.11
9. 460.8 KBps 461.538 KBps 5.6875 0.16 461.538 KBps 11.375 0.16
10. 921.6 KBps 913.043 KBps 2.875 0.93 923.076 KBps 5.6875 0.93
11. 1.792 MBps 1.826 MBps 1.4375 1.9 1.787 MBps 2.9375 0.27
12. 1.8432 MBps 1.826 MBps 1.4375 0.93 1.826 MBps 2.875 0.93
13. 3.584 MBps N.A N.A N.A 3.652 MBps 1.4375 1.9
14. 3.6864 MBps N.A N.A N.A 3.652 MBps 1.4375 0.93
986/1751 RM0090 Rev 19


15. 7.168 MBps N.A N.A N.A N.A N.A N.A

17. 9 MBps N.A N.A N.A N.A N.A N.A
1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can be
fixed with these data.
Value Value Value

programmed programmed programmed %
in the baud in the baud in the baud Error
rate register rate register rate register
1. 2.4 KBps 2.4 KBps 2187.5 0 2.4 KBps NA 0

2. 9.6 KBps 9.6 KBps 546.875 0 9.6 KBps 1093.75 0
3. 19.2 KBps 19.195 KBps 273.5 0.02 19.2 KBps 546.875 0
4. 38.4 KBps 38.391 KBps 136.75 0.02 38.391 KBps 273.5 0.02
5. 57.6 KBps 57.613 KBps 91.125 0.02 57.613 KBps 182.25 0.02
6. 115.2 KBps 115.068 KBps 45.625 0.11 115.226 KBps 91.125 0.02
7. 230.4 KBps 230.769 KBps 22.75 0.11 230.137 KBps 45.625 0.11
8. 460.8 KBps 461.538 KBps 11.375 0.16 461.538 KBps 22.75 0.16
9. 921.6 KBps 913.043 KBps 5.75 0.93 923.076 KBps 11.375 0.93
10. 1.792 MBps 1.826 MBps 2.875 1.9 1.787Mbps 5.875 0.27
11. 1.8432 MBps 1.826 MBps 2.875 0.93 1.826 MBps 5.75 0.93
12. 3.584 MBps 3.5 MBps 1.5 2.34 3.652 MBps 2.875 1.9
13. 3.6864 MBps 3.82 MBps 1.375 3.57 3.652 MBps 2.875 0.93
14. 7.168 MBps N.A N.A N.A 7 MBps 1.5 2.34
15. 7.3728 MBps N.A N.A N.A 7.636 MBps 1.375 3.57
RM0090 Rev 19 987/1751

1018
Value Value Value

programmed programmed programmed %
in the baud in the baud in the baud Error
rate register rate register rate register
16. 9 MBps N.A N.A N.A 9.333 MBps 1.125 3.7

17. 10.5 MBps N.A N.A N.A 10.5 MBps 1 0
30.3.5 USART receiver tolerance to clock deviation

The USART asynchronous receiver works correctly only if the total clock system deviation is
smaller than the USART receiver’s tolerance. The causes which contribute to the total
deviation are:
• DTRA: Deviation due to the transmitter error (which also includes the deviation of the
transmitter’s local oscillator)
• DQUANT: Error due to the baud rate quantization of the receiver
• DREC: Deviation of the receiver’s local oscillator
• DTCL: Deviation due to the transmission line (generally due to the transceivers which
can introduce an asymmetry between the low-to-high transition timing and the high-to-
low transition timing)
DTRA + DQUANT + DREC + DTCL < USART receiver’s tolerance
The USART receiver’s tolerance to properly receive data is equal to the maximum tolerated
deviation and depends on the following choices:
• 10- or 11-bit character length defined by the M bit in the USART_CR1 register
• oversampling by 8 or 16 defined by the OVER8 bit in the USART_CR1 register
• use of fractional baud rate or not
• use of 1 bit or 3 bits to sample the data, depending on the value of the ONEBIT bit in
the USART_CR3 register
Table 144. USART receiver’s tolerance when DIV fraction is 0

OVER8 bit = 0 OVER8 bit = 1
M bit
ONEBIT=0 ONEBIT=1 ONEBIT=0 ONEBIT=1
0 3.75% 4.375% 2.50% 3.75%
1 3.41% 3.97% 2.27% 3.41%
988/1751 RM0090 Rev 19

Table 145. USART receiver tolerance when DIV_Fraction is different from 0

OVER8 bit = 0 OVER8 bit = 1
M bit
ONEBIT=0 ONEBIT=1 ONEBIT=0 ONEBIT=1
0 3.33% 3.88% 2% 3%
1 3.03% 3.53% 1.82% 2.73%
Note: The figures specified in Table 144 and Table 145 may slightly differ in the special case when
the received frames contain some Idle frames of exactly 10-bit times when M=0 (11-bit times
when M=1).
30.3.6 Multiprocessor communication

There is a possibility of performing multiprocessor communication with the USART (several
USARTs connected in a network). For instance one of the USARTs can be the master, its TX
output is connected to the RX input of the other USART. The others are slaves, their
respective TX outputs are logically ANDed together and connected to the RX input of the
master.
In multiprocessor configurations it is often desirable that only the intended message
recipient should actively receive the full message contents, thus reducing redundant USART
service overhead for all non addressed receivers.
The non addressed devices may be placed in mute mode by means of the muting function.
In mute mode:
• None of the reception status bits can be set.
• All the receive interrupts are inhibited.
• The RWU bit in USART_CR1 register is set to 1. RWU can be controlled automatically
by hardware or written by the software under certain conditions.
The USART can enter or exit from mute mode using one of two methods, depending on the
WAKE bit in the USART_CR1 register:
• Idle Line detection if the WAKE bit is reset,
• Address Mark detection if the WAKE bit is set.
Idle line detection (WAKE=0)

The USART enters mute mode when the RWU bit is written to 1.
It wakes up when an Idle frame is detected. Then the RWU bit is cleared by hardware but
the IDLE bit is not set in the USART_SR register. RWU can also be written to 0 by software.
An example of mute mode behavior using Idle line detection is given in Figure 303.
RM0090 Rev 19 989/1751

1018
Figure 303. Mute mode using Idle line detection
RXNE RXNE
RX Data 1 Data 2 Data 3 Data 4 IDLE Data 5 Data 6
RWU Mute mode Normal mode
RWU written to 1 Idle frame detected
MSv40881V1
Address mark detection (WAKE=1)

In this mode, bytes are recognized as addresses if their MSB is a ‘1 else they are
considered as data. In an address byte, the address of the targeted receiver is put on the 4
LSB. This 4-bit word is compared by the receiver with its own address which is programmed
in the ADD bits in the USART_CR2 register.
The USART enters mute mode when an address character is received which does not
match its programmed address. In this case, the RWU bit is set by hardware. The RXNE
flag is not set for this address byte and no interrupt nor DMA request is issued as the
USART would have entered mute mode.
It exits from mute mode when an address character is received which matches the
programmed address. Then the RWU bit is cleared and subsequent bytes are received
normally. The RXNE bit is set for the address character since the RWU bit has been
cleared.
The RWU bit can be written to as 0 or 1 when the receiver buffer contains no data (RXNE=0
in the USART_SR register). Otherwise the write attempt is ignored.
An example of mute mode behavior using address mark detection is given in Figure 304.
Figure 304. Mute mode using address mark detection
In this example, the current address of the receiver is 1

(programmed in the USART_CR2 register)
RXNE RXNE RXNE
RX IDLE Addr=0 Data 1 Data 2 IDLE Addr=1 Data 3 Data 4 Addr=2 Data 5
RWU Mute mode Normal mode Mute mode
RWU written to 1 Matching address Non-matching address

(RXNE was cleared)
Non-matching address
MSv40882V1
990/1751 RM0090 Rev 19

30.3.7 Parity control

Parity control (generation of parity bit in transmission and parity checking in reception) can
be enabled by setting the PCE bit in the USART_CR1 register. Depending on the frame
length defined by the M bit, the possible USART frame formats are as listed in Table 146.
Table 146. Frame formats

M bit PCE bit USART frame(1)
0 0 | SB | 8 bit data | STB |

0 1 | SB | 7-bit data | PB | STB |
1 0 | SB | 9-bit data | STB |
1 1 | SB | 8-bit data PB | STB |
1. Legends: SB: start bit, STB: stop bit, PB: parity bit.
Even parity
The parity bit is calculated to obtain an even number of “1s” inside the frame made of the 7
or 8 LSB bits (depending on whether M is equal to 0 or 1) and the parity bit.
E.g.: data=00110101; 4 bits set => parity bit will be 0 if even parity is selected (PS bit in
USART_CR1 = 0).
Odd parity
The parity bit is calculated to obtain an odd number of “1s” inside the frame made of the 7 or
8 LSB bits (depending on whether M is equal to 0 or 1) and the parity bit.
E.g.: data=00110101; 4 bits set => parity bit will be 1 if odd parity is selected (PS bit in
USART_CR1 = 1).
Parity checking in reception

If the parity check fails, the PE flag is set in the USART_SR register and an interrupt is
generated if PEIE is set in the USART_CR1 register. The PE flag is cleared by a software
sequence (a read from the status register followed by a read or write access to the
USART_DR data register).
Note: In case of wakeup by an address mark: the MSB bit of the data is taken into account to
identify an address but not the parity bit. And the receiver does not check the parity of the
address data (PE is not set in case of a parity error).
Parity generation in transmission

If the PCE bit is set in USART_CR1, then the MSB bit of the data written in the data register
is transmitted but is changed by the parity bit (even number of “1s” if even parity is selected
(PS=0) or an odd number of “1s” if odd parity is selected (PS=1)).
Note: The software routine that manages the transmission can activate the software sequence
which clears the PE flag (a read from the status register followed by a read or write access
to the data register). When operating in half-duplex mode, depending on the software, this
can cause the PE flag to be unexpectedly cleared.
RM0090 Rev 19 991/1751

1018
30.3.8 LIN (local interconnection network) mode

The LIN mode is selected by setting the LINEN bit in the USART_CR2 register. In LIN
mode, the following bits must be kept cleared:
• STOP[1:0] and CLKEN in the USART_CR2 register
• SCEN, HDSEL and IREN in the USART_CR3 register.
LIN transmission
The same procedure explained in Section 30.3.2 has to be applied for LIN Master
transmission than for normal USART transmission with the following differences:
• Clear the M bit to configure 8-bit word length.
• Set the LINEN bit to enter LIN mode. In this case, setting the SBK bit sends 13 ‘0 bits
as a break character. Then a bit of value ‘1 is sent to allow the next start detection.
LIN reception
A break detection circuit is implemented on the USART interface. The detection is totally
independent from the normal USART receiver. A break can be detected whenever it occurs,
during Idle state or during a frame.
When the receiver is enabled (RE=1 in USART_CR1), the circuit looks at the RX input for a
start signal. The method for detecting start bits is the same when searching break
characters or data. After a start bit has been detected, the circuit samples the next bits
exactly like for the data (on the 8th, 9th and 10th samples). If 10 (when the LBDL = 0 in
USART_CR2) or 11 (when LBDL=1 in USART_CR2) consecutive bits are detected as ‘0,
and are followed by a delimiter character, the LBD flag is set in USART_SR. If the LBDIE
bit=1, an interrupt is generated. Before validating the break, the delimiter is checked for as it
signifies that the RX line has returned to a high level.
If a ‘1 is sampled before the 10 or 11 have occurred, the break detection circuit cancels the
current detection and searches for a start bit again.
If the LIN mode is disabled (LINEN=0), the receiver continues working as normal USART,
without taking into account the break detection.
If the LIN mode is enabled (LINEN=1), as soon as a framing error occurs (i.e. stop bit
detected at ‘0, which will be the case for any break frame), the receiver stops until the break
detection circuit receives either a ‘1, if the break word was not complete, or a delimiter
character if a break has been detected.
The behavior of the break detector state machine and the break flag is shown on the
Figure 305: Break detection in LIN mode (11-bit break length - LBDL bit is set) on page 993.
Examples of break frames are given on Figure 306: Break detection in LIN mode vs.
Framing error detection on page 994.
992/1751 RM0090 Rev 19

Figure 305. Break detection in LIN mode (11-bit break length - LBDL bit is set)
Case 1: break signal not long enough => break discarded, LBD is not set
Break frame
RX line
Capture strobe
Break state
Idle Bit0 Bit1 Bit2 Bit3 Bit4 Bit5 Bit6 Bit7 Bit8 Bit9 Bit10 Idle
machine
Read samples 0 0 0 0 0 0 0 0 0 0 1
Case 2: break signal just long enough => break detected, LBD is set
Break frame
RX line
Delimiter is immediate
Capture strobe
Break state
Idle Bit0 Bit1 Bit2 Bit3 Bit4 Bit5 Bit6 Bit7 Bit8 Bit9 B10 Idle
machine
Read samples 0 0 0 0 0 0 0 0 0 0 0
LBD
Case 3: break signal long enough => break detected, LBD is set
Break frame
RX line
Capture strobe
Break state
Idle Bit0 Bit1 Bit2 Bit3 Bit4 Bit5 Bit6 Bit7 Bit8 Bit9 Bit10 wait delimiter Idle
machine
Read samples 0 0 0 0 0 0 0 0 0 0 0
LBD
MSv40883V1
RM0090 Rev 19 993/1751

1018
Figure 306. Break detection in LIN mode vs. Framing error detection
Case 1: break occurring after an Idle
RX line data 1 IDLE BREAK data 2 (0x55) data 3 (header)

1 data time 1 data time
RXNE /FE
LBDF
Case 2: break occurring while data is being received
RX line data 1 data2 BREAK data 2 (0x55) data 3 (header)

1 data time 1 data time
RXNE /FE
LBDF
MSv31157V1
30.3.9 USART synchronous mode

The synchronous mode is selected by writing the CLKEN bit in the USART_CR2 register to
1. In synchronous mode, the following bits must be kept cleared:
• LINEN bit in the USART_CR2 register,
• SCEN, HDSEL and IREN bits in the USART_CR3 register.
The USART allows the user to control a bidirectional synchronous serial communications in
master mode. The CK pin is the output of the USART transmitter clock. No clock pulses are
sent to the CK pin during start bit and stop bit. Depending on the state of the LBCL bit in the
USART_CR2 register clock pulses will or will not be generated during the last valid data bit
(address mark). The CPOL bit in the USART_CR2 register allows the user to select the
clock polarity, and the CPHA bit in the USART_CR2 register allows the user to select the
phase of the external clock (see Figure 307, Figure 308 & Figure 309).
During the Idle state, preamble and send break, the external CK clock is not activated.
In synchronous mode the USART transmitter works exactly like in asynchronous mode. But
as CK is synchronized with TX (according to CPOL and CPHA), the data on TX is
synchronous.
In this mode the USART receiver works in a different manner compared to the
asynchronous mode. If RE=1, the data is sampled on CK (rising or falling edge, depending
on CPOL and CPHA), without any oversampling. A setup and a hold time must be
respected (which depends on the baud rate: 1/16 bit time).
Note: The CK pin works in conjunction with the TX pin. Thus, the clock is provided only if the
transmitter is enabled (TE=1) and a data is being transmitted (the data register USART_DR
994/1751 RM0090 Rev 19

has been written). This means that it is not possible to receive a synchronous data without
transmitting data.
The LBCL, CPOL and CPHA bits have to be selected when both the transmitter and the
receiver are disabled (TE=RE=0) to ensure that the clock pulses function correctly. These
bits should not be changed while the transmitter or the receiver is enabled.
It is advised that TE and RE are set in the same instruction in order to minimize the setup
and the hold time of the receiver.
The USART supports master mode only: it cannot receive or send data related to an input
clock (CK is always an output).
Figure 307. USART example of synchronous transmission
RX Data out
TX Data in
Synchronous device
USART
(e.g. slave SPI)
CK Clock
MSv31158V2
Figure 308. USART data clock timing diagram (M=0)
Idle or preceding M=0 (8 data bits) Idle or next

transmission Start Stop transmission
Clock (CPOL=0, CPHA=0 *
Clock (CPOL=0, CPHA=1

*
Data on TX
0 1 2 3 4 5 6 7
(from master)
Start LSB MSB Stop
Data on RX
0 1 2 3 4 5 6 7
(from slave)
LSB MSB
*
Capture strobe
*LBCL bit controls last data pulse
MSv31159V1
RM0090 Rev 19 995/1751

1018
Figure 309. USART data clock timing diagram (M=1)
Idle or
Idle or next
preceding Start M=1 (9 data bits) Stop
transmission
transmission
Clock (CPOL=0,
CPHA=0 *
Clock (CPOL=0,
CPHA=1 *
Clock (CPOL=1, *
CPHA=0
Clock (CPOL=1, *
CPHA=1
Data on TX
0 1 2 3 4 5 6 7 8
(from master)
Start LSB MSB Stop
Data on RX
0 1 2 3 4 5 6 7 8
(from slave)
LSB MSB
Capture *
strobe
*LBCL bit controls last data pulse
MSv31160V1
Figure 310. RX data setup/hold time
CK
(capture strobe on CK rising
edge in this example)
Data on RX (from slave) Valid DATA bit
tSETUP tHOLD
tSETUP=tHOLD 1/16 bit time

MSv31161V2
Note: The function of CK is different in Smartcard mode. Refer to the Smartcard mode chapter for
more details.
30.3.10 Single-wire half-duplex communication

The single-wire half-duplex mode is selected by setting the HDSEL bit in the USART_CR3
register. In this mode, the following bits must be kept cleared:
• LINEN and CLKEN bits in the USART_CR2 register,
• SCEN and IREN bits in the USART_CR3 register.
The USART can be configured to follow a single-wire half-duplex protocol where the TX and
RX lines are internally connected. The selection between half- and full-duplex
communication is made with a control bit ‘HALF DUPLEX SEL’ (HDSEL in USART_CR3).
996/1751 RM0090 Rev 19

As soon as HDSEL is written to 1:

• the TX and RX lines are internally connected
• the RX pin is no longer used
• the TX pin is always released when no data is transmitted. Thus, it acts as a standard
I/O in idle or in reception. It means that the I/O must be configured so that TX is
configured as floating input (or output high open-drain) when not driven by the USART.
Apart from this, the communications are similar to what is done in normal USART mode.
The conflicts on the line must be managed by the software (by the use of a centralized
arbiter, for instance). In particular, the transmission is never blocked by hardware and
continue to occur as soon as a data is written in the data register while the TE bit is set.
30.3.11 Smartcard
The Smartcard mode is selected by setting the SCEN bit in the USART_CR3 register. In
smartcard mode, the following bits must be kept cleared:
• LINEN bit in the USART_CR2 register,
• HDSEL and IREN bits in the USART_CR3 register.
Moreover, the CLKEN bit may be set in order to provide a clock to the smartcard.
The Smartcard interface is designed to support asynchronous protocol Smartcards as
defined in the ISO 7816-3 standard. The USART should be configured as:
• 8 bits plus parity: where M=1 and PCE=1 in the USART_CR1 register
• 1.5 stop bits when transmitting and receiving: where STOP=11 in the USART_CR2
register.
Note: It is also possible to choose 0.5 stop bit for receiving but it is recommended to use 1.5 stop
bits for both transmitting and receiving to avoid switching between the two configurations.
Figure 311 shows examples of what can be seen on the data line with and without parity
error.
Figure 311. ISO 7816-3 asynchronous protocol
Without Parity error

Guard time
S 0 1 2 3 4 5 6 7 p
Start bit
WithParity error
Guard time
S 0 1 2 3 4 5 6 7 p
Start bit
Line pulled low by receiver
during stop in case of parity error
MSv31162V1
When connected to a Smartcard, the TX output of the USART drives a bidirectional line that
is also driven by the Smartcard. The TX pin must be configured as open-drain.
Smartcard is a single wire half duplex communication protocol.
• Transmission of data from the transmit shift register is guaranteed to be delayed by a
minimum of 1/2 baud clock. In normal operation a full transmit shift register will start
RM0090 Rev 19 997/1751

1018
shifting on the next baud clock edge. In Smartcard mode this transmission is further
delayed by a guaranteed 1/2 baud clock.
• If a parity error is detected during reception of a frame programmed with a 0.5 or 1.5
stop bit period, the transmit line is pulled low for a baud clock period after the
completion of the receive frame. This is to indicate to the Smartcard that the data
transmitted to USART has not been correctly received. This NACK signal (pulling
transmit line low for 1 baud clock) will cause a framing error on the transmitter side
(configured with 1.5 stop bits). The application can handle re-sending of data according
to the protocol. A parity error is ‘NACK’ed by the receiver if the NACK control bit is set,
otherwise a NACK is not transmitted.
• The assertion of the TC flag can be delayed by programming the Guard Time register.
In normal operation, TC is asserted when the transmit shift register is empty and no
further transmit requests are outstanding. In Smartcard mode an empty transmit shift
register triggers the guard time counter to count up to the programmed value in the
Guard Time register. TC is forced low during this time. When the guard time counter
reaches the programmed value TC is asserted high.
• The de-assertion of TC flag is unaffected by Smartcard mode.
• If a framing error is detected on the transmitter end (due to a NACK from the receiver),
the NACK will not be detected as a start bit by the receive block of the transmitter.
According to the ISO protocol, the duration of the received NACK can be 1 or 2 baud
clock periods.
• On the receiver side, if a parity error is detected and a NACK is transmitted the receiver
will not detect the NACK as a start bit.
Note: A break character is not significant in Smartcard mode. A 0x00 data with a framing error will
be treated as data and not as a break.
No Idle frame is transmitted when toggling the TE bit. The Idle frame (as defined for the
other configurations) is not defined by the ISO protocol.
Figure 312 details how the NACK signal is sampled by the USART. In this example the
USART is transmitting a data and is configured with 1.5 stop bits. The receiver part of the
USART is enabled in order to check the integrity of the data and the NACK signal.
Figure 312. Parity error detection using the 1.5 stop bits
Bit 7 Parity bit 1.5 Stop bit
1 bit time 1.5 bit time
Sampling at Sampling at
8th, 9th, 10th 8th, 9th, 10th
0.5 bit time
Sampling at Sampling at
8th, 9th, 10th 8th, 9th, 10th
MSv31163V1
The USART can provide a clock to the smartcard through the CK output. In smartcard
mode, CK is not associated to the communication but is simply derived from the internal
peripheral input clock through a 5-bit prescaler. The division ratio is configured in the
998/1751 RM0090 Rev 19

prescaler register USART_GTPR. CK frequency can be programmed from fCK/2 to fCK/62,

where fCK is the peripheral input clock.
30.3.12 IrDA SIR ENDEC block

The IrDA mode is selected by setting the IREN bit in the USART_CR3 register. In IrDA
mode, the following bits must be kept cleared:
• LINEN, STOP and CLKEN bits in the USART_CR2 register,
• SCEN and HDSEL bits in the USART_CR3 register.
The IrDA SIR physical layer specifies use of a Return to Zero, Inverted (RZI) modulation
scheme that represents logic 0 as an infrared light pulse (see Figure 313).
The SIR Transmit encoder modulates the Non Return to Zero (NRZ) transmit bit stream
output from USART. The output pulse stream is transmitted to an external output driver and
infrared LED. USART supports only bit rates up to 115.2Kbps for the SIR ENDEC. In normal
mode the transmitted pulse width is specified as 3/16 of a bit period.
The SIR receive decoder demodulates the return-to-zero bit stream from the infrared
detector and outputs the received NRZ serial bit stream to USART. The decoder input is
normally HIGH (marking state) in the Idle state. The transmit encoder output has the
opposite polarity to the decoder input. A start bit is detected when the decoder input is low.
• IrDA is a half duplex communication protocol. If the Transmitter is busy (i.e. the USART
is sending data to the IrDA encoder), any data on the IrDA receive line will be ignored
by the IrDA decoder and if the Receiver is busy (USART is receiving decoded data
from the USART), data on the TX from the USART to IrDA will not be encoded by IrDA.
While receiving data, transmission should be avoided as the data to be transmitted
could be corrupted.
• A ‘0 is transmitted as a high pulse and a ‘1 is transmitted as a ‘0. The width of the pulse
is specified as 3/16th of the selected bit period in normal mode (see Figure 314).
• The SIR decoder converts the IrDA compliant receive signal into a bit stream for
USART.
• The SIR receive logic interprets a high state as a logic one and low pulses as logic
zeros.
• The transmit encoder output has the opposite polarity to the decoder input. The SIR
output is in low state when Idle.
• The IrDA specification requires the acceptance of pulses greater than 1.41 us. The
acceptable pulse width is programmable. Glitch detection logic on the receiver end
filters out pulses of width less than 2 PSC periods (PSC is the prescaler value
programmed in the IrDA low-power Baud Register, USART_GTPR). Pulses of width
less than 1 PSC period are always rejected, but those of width greater than one and
less than two periods may be accepted or rejected, those greater than 2 periods will be
accepted as a pulse. The IrDA encoder/decoder doesn’t work when PSC=0.
• The receiver can communicate with a low-power transmitter.
• In IrDA mode, the STOP bits in the USART_CR2 register must be configured to “1 stop
bit”.
RM0090 Rev 19 999/1751

1018
IrDA low-power mode

Transmitter:
In low-power mode the pulse width is not maintained at 3/16 of the bit period. Instead, the
width of the pulse is 3 times the low-power baud rate which can be a minimum of 1.42 MHz.
Generally this value is 1.8432 MHz (1.42 MHz < PSC< 2.12 MHz). A low-power mode
programmable divisor divides the system clock to achieve this value.
Receiver:
Receiving in low-power mode is similar to receiving in normal mode. For glitch detection the
USART should discard pulses of duration shorter than 1/PSC. A valid low is accepted only if
its duration is greater than 2 periods of the IrDA low-power Baud clock (PSC value in
USART_GTPR).
Note: A pulse of width less than two and greater than one PSC period(s) may or may not be
rejected.
The receiver set up time should be managed by software. The IrDA physical layer
specification specifies a minimum of 10 ms delay between transmission and reception (IrDA
is a half duplex protocol).
Figure 313. IrDA SIR ENDEC- block diagram
SIREN
TX
OR USART_TX
SIR
Transmit IrDA_OUT
Encoder
USART
SIR
RX
Receive IrDA_IN
DEcoder
USART_RX
MSv31164V2
Figure 314. IrDA data modulation (3/16) -Normal mode
Start Stop
bit bit
0 1 0 1 0 0 1 1 0 1
TX
IrDA_OUT
Bit period 3/16
IrDA_IN
RX 0 1 0 1 0 0 1 1 0 1
MSv31165V1
1000/1751 RM0090 Rev 19

30.3.13 Continuous communication using DMA

The USART is capable of continuous communication using the DMA. The DMA requests for
Rx buffer and Tx buffer are generated independently.
Transmission using DMA

DMA mode can be enabled for transmission by setting DMAT bit in the USART_CR3
register. Data is loaded from a SRAM area configured using the DMA peripheral (refer to the
DMA specification) to the USART_DR register whenever the TXE bit is set. To map a DMA
channel for USART transmission, use the following procedure (x denotes the channel
number):
1. Write the USART_DR register address in the DMA control register to configure it as the
destination of the transfer. The data will be moved to this address from memory after
each TXE event.
2. Write the memory address in the DMA control register to configure it as the source of
the transfer. The data will be loaded into the USART_DR register from this memory
area after each TXE event.
3. Configure the total number of bytes to be transferred to the DMA control register.
4. Configure the channel priority in the DMA register
5. Configure DMA interrupt generation after half/ full transfer as required by the
application.
6. Clear the TC bit in the SR register by writing 0 to it.
7. Activate the channel in the DMA register.
When the number of data transfers programmed in the DMA Controller is reached, the DMA
controller generates an interrupt on the DMA channel interrupt vector.
In transmission mode, once the DMA has written all the data to be transmitted (the TCIF flag
is set in the DMA_ISR register), the TC flag can be monitored to make sure that the USART
communication is complete. This is required to avoid corrupting the last transmission before
disabling the USART or entering the Stop mode. The software must wait until TC=1. The TC
flag remains cleared during all data transfers and it is set by hardware at the last frame’s
end of transmission.
RM0090 Rev 19 1001/1751

1018
Figure 315. Transmission using DMA

Idle preamble Frame 1 Frame 2 Frame 3
TX line

TXE flag cleared by DMA read cleared by DMA read set by hardware
DMA request ignored by the DMA

because DMA transfer is complete
USART_DR F1 F2 F3
set
TC flag by hardware
DMA writes
USART_DR
clear
flag DMA TCIF set by hardware by software
(Transfer complete)
software configures DMA writes F1 DMA writes F2 DMA writes F3 The DMA transfer
the DMA to send 3 into into into is complete software waits until TC=1
data and enables the USART_DR USART_DR USART_DR. (TCIF=1 in
USART DMA_ISR)
ai17192b
Reception using DMA

DMA mode can be enabled for reception by setting the DMAR bit in USART_CR3 register.
Data is loaded from the USART_DR register to a SRAM area configured using the DMA
peripheral (refer to the DMA specification) whenever a data byte is received. To map a DMA
channel for USART reception, use the following procedure:
1. Write the USART_DR register address in the DMA control register to configure it as the
source of the transfer. The data will be moved from this address to the memory after
each RXNE event.
2. Write the memory address in the DMA control register to configure it as the destination
of the transfer. The data will be loaded from USART_DR to this memory area after each
RXNE event.
3. Configure the total number of bytes to be transferred in the DMA control register.
4. Configure the channel priority in the DMA control register
5. Configure interrupt generation after half/ full transfer as required by the application.
6. Activate the channel in the DMA control register.
When the number of data transfers programmed in the DMA Controller is reached, the DMA
controller generates an interrupt on the DMA channel interrupt vector. The DMAR bit should
be cleared by software in the USART_CR3 register during the interrupt subroutine.
1002/1751 RM0090 Rev 19

Figure 316. Reception using DMA

Frame 1 Frame 2 Frame 3
TX line
set by hardware
RXNE flag cleared by DMA read
DMA request
USART_DR F1 F2 F3
DMA reads USART_DR
cleared
DMA TCIF flag set by hardware by software
(Transfer complete)
software configures the DMA reads F1 DMA reads F2 DMA reads F3 The DMA transfer
DMA to receive 3 data from from from is complete
blocks and enables USART_DR USART_DR USART_DR (TCIF=1 in
the USART DMA_ISR)
ai17193b
Error flagging and interrupt generation in multibuffer communication

In case of multibuffer communication if any error occurs during the transaction the error flag
will be asserted after the current byte. An interrupt will be generated if the interrupt enable
flag is set. For framing error, overrun error and noise flag which are asserted with RXNE in
case of single byte reception, there will be separate error flag interrupt enable bit (EIE bit in
the USART_CR3 register), which if set will issue an interrupt after the current byte with
either of these errors.
30.3.14 Hardware flow control

It is possible to control the serial data flow between 2 devices by using the CTS input and
the RTS output. The Figure 317 shows how to connect 2 devices in this mode:
Figure 317. Hardware flow control between 2 USARTs
USART 1 USART 2
TX RX
TX circuit RX circuit
CTS RTS
RX TX
RX circuit TX circuit
RTS CTS
MSv31169V2
RTS and CTS flow control can be enabled independently by writing respectively RTSE and
CTSE bits to 1 (in the USART_CR3 register).
RM0090 Rev 19 1003/1751

1018
RTS flow control

If the RTS flow control is enabled (RTSE=1), then RTS is asserted (tied low) as long as the
USART receiver is ready to receive a new data. When the receive register is full, RTS is
deasserted, indicating that the transmission is expected to stop at the end of the current
frame. Figure 318 shows an example of communication with RTS flow control enabled.
Figure 318. RTS flow control
Start Stop Start Stop

RX Data 1 Idle Data 2
bit bit bit bit
RTS
RXNE Data 1 read RXNE

Data 2 can now be transmitted
MSv31168V2
CTS flow control

If the CTS flow control is enabled (CTSE=1), then the transmitter checks the CTS input
before transmitting the next frame. If CTS is asserted (tied low), then the next data is
transmitted (assuming that a data is to be transmitted, in other words, if TXE=0), else the
transmission does not occur. When CTS is deasserted during a transmission, the current
transmission is completed before the transmitter stops.
When CTSE=1, the CTSIF status bit is automatically set by hardware as soon as the CTS
input toggles. It indicates when the receiver becomes ready or not ready for communication.
An interrupt is generated if the CTSIE bit in the USART_CR3 register is set. The figure
below shows an example of communication with CTS flow control enabled.
1004/1751 RM0090 Rev 19

Figure 319. CTS flow control

CTS CTS
CTS
Transmit data register
TDR Data 2 empty Data 3 empty
Stop Start Stop Start

TX Data 1 Data 2 Idle Data 3
bit bit bit bit
Writing data 3 in TDR Transmission of Data 3 is

delayed until CTS = 0
MSv31167V2
Note: Special behavior of break frames: when the CTS flow is enabled, the transmitter does not
check the CTS input state to send a break.
RM0090 Rev 19 1005/1751

1018
30.4 USART interrupts

Table 147. USART interrupt requests
Enable control
Interrupt event Event flag
bit
Transmit Data Register Empty TXE TXEIE

CTS flag CTS CTSIE
Transmission Complete TC TCIE
Received Data Ready to be Read RXNE
RXNEIE
Overrun Error Detected ORE
Idle Line Detected IDLE IDLEIE
Parity Error PE PEIE
Break Flag LBD LBDIE
Noise Flag, Overrun error and Framing Error in multibuffer
NF or ORE or FE EIE
communication
The USART interrupt events are connected to the same interrupt vector (see Figure 320).
• During transmission: Transmission Complete, Clear to Send or Transmit Data Register
empty interrupt.
• While receiving: Idle Line detection, Overrun error, Receive Data register not empty,
Parity error, LIN break detection, Noise Flag (only in multi buffer communication) and
Framing Error (only in multi buffer communication).
These events generate an interrupt if the corresponding Enable Control Bit is set.
Figure 320. USART interrupt mapping diagram

TC
TCIE
TXE
TXEIE
CTSIF
CTSIE
IDLE
IDLEIE USART
RXNEIE interrupt
ORE
RXNEIE
RXNE
PE
PEIE
LBD
LBDIE
FE
NE
EIE
ORE DMAR
MSv42089V1
1006/1751 RM0090 Rev 19

30.5 USART mode configuration

Table 148. USART mode configuration(1)
USART USART USART USART
USART modes UART4 UART5
1 2 3 6
Asynchronous mode X X X X X X
Hardware flow control X X X NA NA X
Multibuffer communication (DMA) X X X X X X
Multiprocessor communication X X X X X X
Synchronous X X X NA NA X
Smartcard X X X NA NA X
Half-duplex (single-wire mode) X X X X X X
IrDA X X X X X X
LIN X X X X X X
1. X = supported; NA = not applicable.
30.6 USART registers

30.6.1 Status register (USART_SR)

Reset value: 0x0000 00C0
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CTS LBD TXE TC RXNE IDLE ORE NF FE PE
Reserved
rc_w0 rc_w0 r rc_w0 rc_w0 r r r r r
RM0090 Rev 19 1007/1751

1018

Bit 9 CTS: CTS flag
This bit is set by hardware when the CTS input toggles, if the CTSE bit is set. It is cleared by
software (by writing it to 0). An interrupt is generated if CTSIE=1 in the USART_CR3
register.
0: No change occurred on the CTS status line
1: A change occurred on the CTS status line
Note: This bit is not available for UART4 & UART5.
Bit 8 LBD: LIN break detection flag
This bit is set by hardware when the LIN break is detected. It is cleared by software (by
writing it to 0). An interrupt is generated if LBDIE = 1 in the USART_CR2 register.
0: LIN Break not detected
1: LIN break detected
Note: An interrupt is generated when LBD=1 if LBDIE=1
Bit 7 TXE: Transmit data register empty
This bit is set by hardware when the content of the TDR register has been transferred into
the shift register. An interrupt is generated if the TXEIE bit =1 in the USART_CR1 register. It
is cleared by a write to the USART_DR register.
0: Data is not transferred to the shift register
1: Data is transferred to the shift register)
Note: This bit is used during single buffer transmission.
Bit 6 TC: Transmission complete
This bit is set by hardware if the transmission of a frame containing data is complete and if
TXE is set. An interrupt is generated if TCIE=1 in the USART_CR1 register. It is cleared by
a software sequence (a read from the USART_SR register followed by a write to the
USART_DR register). The TC bit can also be cleared by writing a '0' to it. This clearing
sequence is recommended only for multibuffer communication.
0: Transmission is not complete
1: Transmission is complete
Bit 5 RXNE: Read data register not empty
This bit is set by hardware when the content of the RDR shift register has been transferred
to the USART_DR register. An interrupt is generated if RXNEIE=1 in the USART_CR1
register. It is cleared by a read to the USART_DR register. The RXNE flag can also be
cleared by writing a zero to it. This clearing sequence is recommended only for multibuffer
communication.
0: Data is not received
1: Received data is ready to be read.
Bit 4 IDLE: IDLE line detected
This bit is set by hardware when an Idle Line is detected. An interrupt is generated if the
IDLEIE=1 in the USART_CR1 register. It is cleared by a software sequence (an read to the
USART_SR register followed by a read to the USART_DR register).
0: No Idle Line is detected
1: Idle Line is detected
Note: The IDLE bit will not be set again until the RXNE bit has been set itself (i.e. a new idle
line occurs).
1008/1751 RM0090 Rev 19

Bit 3 ORE: Overrun error

This bit is set by hardware when the word currently being received in the shift register is
ready to be transferred into the RDR register while RXNE=1. An interrupt is generated if
RXNEIE=1 in the USART_CR1 register. It is cleared by a software sequence (an read to the
USART_SR register followed by a read to the USART_DR register).
0: No Overrun error
1: Overrun error is detected
Note: When this bit is set, the RDR register content will not be lost but the shift register will be
overwritten. An interrupt is generated on ORE flag in case of Multi Buffer
communication if the EIE bit is set.
Bit 2 NF: Noise detected flag
This bit is set by hardware when noise is detected on a received frame. It is cleared by a
software sequence (an read to the USART_SR register followed by a read to the
USART_DR register).
0: No noise is detected
1: Noise is detected
Note: This bit does not generate interrupt as it appears at the same time as the RXNE bit
which itself generates an interrupting interrupt is generated on NF flag in case of Multi
Buffer communication if the EIE bit is set.
Note: When the line is noise-free, the NF flag can be disabled by programming the ONEBIT
bit to 1 to increase the USART tolerance to deviations (Refer to Section 30.3.5: USART
receiver tolerance to clock deviation on page 988).
Bit 1 FE: Framing error
This bit is set by hardware when a de-synchronization, excessive noise or a break character
is detected. It is cleared by a software sequence (an read to the USART_SR register
followed by a read to the USART_DR register).
0: No Framing error is detected
1: Framing error or break character is detected
Note: This bit does not generate interrupt as it appears at the same time as the RXNE bit
which itself generates an interrupt. If the word currently being transferred causes both
frame error and overrun error, it will be transferred and only the ORE bit will be set.
An interrupt is generated on FE flag in case of Multi Buffer communication if the EIE bit
is set.
Bit 0 PE: Parity error
This bit is set by hardware when a parity error occurs in receiver mode. It is cleared by a
software sequence (a read from the status register followed by a read or write access to the
USART_DR data register). The software must wait for the RXNE flag to be set before
clearing the PE bit.
An interrupt is generated if PEIE = 1 in the USART_CR1 register.
0: No parity error
1: Parity error
RM0090 Rev 19 1009/1751

1018
30.6.2 Data register (USART_DR)

Reset value: 0xXXXX XXXX
Bits 8:0 DR[8:0]: Data value
Contains the Received or Transmitted data character, depending on whether it is read from
or written to.
The Data register performs a double function (read and write) since it is composed of two
registers, one for transmission (TDR) and one for reception (RDR)
The TDR register provides the parallel interface between the internal bus and the output shift
register (see Figure 1).
The RDR register provides the parallel interface between the input shift register and the
internal bus.
When transmitting with the parity enabled (PCE bit set to 1 in the USART_CR1 register), the
value written in the MSB (bit 7 or bit 8 depending on the data length) has no effect because
it is replaced by the parity.
When receiving with the parity enabled, the value read in the MSB bit is the received parity
bit.
30.6.3 Baud rate register (USART_BRR)

Note: The baud counters stop counting if the TE or RE bits are disabled respectively.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DIV_Mantissa[11:0] DIV_Fraction[3:0]

Bits 15:4 DIV_Mantissa[11:0]: mantissa of USARTDIV
These 12 bits define the mantissa of the USART Divider (USARTDIV)
Bits 3:0 DIV_Fraction[3:0]: fraction of USARTDIV
These 4 bits define the fraction of the USART Divider (USARTDIV). When OVER8=1, the
DIV_Fraction3 bit is not considered and must be kept cleared.
30.6.4 Control register 1 (USART_CR1)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OVER8 Reserved UE M WAKE PCE PS PEIE TXEIE TCIE RXNEIE IDLEIE TE RE RWU SBK
rw Res. rw rw rw rw rw rw rw rw rw rw rw rw rw rw
1010/1751 RM0090 Rev 19


Bit 15 OVER8: Oversampling mode
0: oversampling by 16
1: oversampling by 8
Note: Oversampling by 8 is not available in the Smartcard, IrDA and LIN modes: when
SCEN=1,IREN=1 or LINEN=1 then OVER8 is forced to ‘0 by hardware.
Bit 13 UE: USART enable
When this bit is cleared, the USART prescalers and outputs are stopped and the end of the
current byte transfer in order to reduce power consumption. This bit is set and cleared by
software.
0: USART prescaler and outputs disabled
1: USART enabled
Bit 12 M: Word length
This bit determines the word length. It is set or cleared by software.
0: 1 Start bit, 8 Data bits, n Stop bit
1: 1 Start bit, 9 Data bits, n Stop bit
Note: The M bit must not be modified during a data transfer (both transmission and reception)
Bit 11 WAKE: Wakeup method
This bit determines the USART wakeup method, it is set or cleared by software.
0: Idle Line
1: Address Mark
Bit 10 PCE: Parity control enable
This bit selects the hardware parity control (generation and detection). When the parity
control is enabled, the computed parity is inserted at the MSB position (9th bit if M=1; 8th bit
if M=0) and parity is checked on the received data. This bit is set and cleared by software.
Once it is set, PCE is active after the current byte (in reception and in transmission).
0: Parity control disabled
1: Parity control enabled
Bit 9 PS: Parity selection
This bit selects the odd or even parity when the parity generation/detection is enabled (PCE
bit set). It is set and cleared by software. The parity will be selected after the current byte.
0: Even parity
1: Odd parity
Bit 8 PEIE: PE interrupt enable
0: Interrupt is inhibited
1: An USART interrupt is generated whenever PE=1 in the USART_SR register
Bit 7 TXEIE: TXE interrupt enable
1: An USART interrupt is generated whenever TXE=1 in the USART_SR register
Bit 6 TCIE: Transmission complete interrupt enable
1: An USART interrupt is generated whenever TC=1 in the USART_SR register
RM0090 Rev 19 1011/1751

1018
Bit 5 RXNEIE: RXNE interrupt enable

1: An USART interrupt is generated whenever ORE=1 or RXNE=1 in the USART_SR
register
Bit 4 IDLEIE: IDLE interrupt enable
1: An USART interrupt is generated whenever IDLE=1 in the USART_SR register
Bit 3 TE: Transmitter enable
This bit enables the transmitter. It is set and cleared by software.
0: Transmitter is disabled
1: Transmitter is enabled
Note: During transmission, a “0” pulse on the TE bit (“0” followed by “1”) sends a preamble
(idle line) after the current word, except in smartcard mode.
When TE is set, there is a 1 bit-time delay before the transmission starts.
Bit 2 RE: Receiver enable
This bit enables the receiver. It is set and cleared by software.
0: Receiver is disabled
1: Receiver is enabled and begins searching for a start bit
Bit 1 RWU: Receiver wakeup
This bit determines if the USART is in mute mode or not. It is set and cleared by software
and can be cleared by hardware when a wakeup sequence is recognized.
0: Receiver in active mode
1: Receiver in mute mode
Note: Before selecting Mute mode (by setting the RWU bit) the USART must first receive a
data byte, otherwise it cannot function in Mute mode with wakeup by Idle line detection.
In Address Mark Detection wakeup configuration (WAKE bit=1) the RWU bit cannot be
modified by software while the RXNE bit is set.
Bit 0 SBK: Send break
This bit set is used to send break characters. It can be set and cleared by software. It should
be set by software, and will be reset by hardware during the stop bit of break.
0: No break character is transmitted
1: Break character will be transmitted
1012/1751 RM0090 Rev 19


31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
LINEN STOP[1:0] CLKEN CPOL CPHA LBCL Res. LBDIE LBDL Res. ADD[3:0]
Res.

Bit 14 LINEN: LIN mode enable
0: LIN mode disabled
1: LIN mode enabled
The LIN mode enables the capability to send LIN Synch Breaks (13 low bits) using the SBK bit in
the USART_CR1 register, and to detect LIN Sync breaks.
Bits 13:12 STOP: STOP bits
These bits are used for programming the stop bits.
00: 1 Stop bit
01: 0.5 Stop bit
10: 2 Stop bits
11: 1.5 Stop bit
Note: The 0.5 Stop bit and 1.5 Stop bit are not available for UART4 & UART5.
Bit 11 CLKEN: Clock enable
This bit allows the user to enable the CK pin.
0: CK pin disabled
1: CK pin enabled
This bit is not available for UART4 & UART5.
Bit 10 CPOL: Clock polarity
This bit allows the user to select the polarity of the clock output on the CK pin in synchronous mode.
It works in conjunction with the CPHA bit to produce the desired clock/data relationship
0: Steady low value on CK pin outside transmission window.
1: Steady high value on CK pin outside transmission window.
This bit is not available for UART4 & UART5.
Bit 9 CPHA: Clock phase
This bit allows the user to select the phase of the clock output on the CK pin in synchronous mode.
It works in conjunction with the CPOL bit to produce the desired clock/data relationship (see figures
308 to 309)
0: The first clock transition is the first data capture edge
1: The second clock transition is the first data capture edge
RM0090 Rev 19 1013/1751

1018
Bit 8 LBCL: Last bit clock pulse

This bit allows the user to select whether the clock pulse associated with the last data bit
transmitted (MSB) has to be output on the CK pin in synchronous mode.
0: The clock pulse of the last data bit is not output to the CK pin
1: The clock pulse of the last data bit is output to the CK pin
Note: 1: The last bit is the 8th or 9th data bit transmitted depending on the 8 or 9 bit format selected
by the M bit in the USART_CR1 register.
2: This bit is not available for UART4 & UART5.
Bit 6 LBDIE: LIN break detection interrupt enable
Break interrupt mask (break detection using break delimiter).
1: An interrupt is generated whenever LBD=1 in the USART_SR register
Bit 5 LBDL: lin break detection length
This bit is for selection between 11 bit or 10 bit break detection.
0: 10-bit break detection
1: 11-bit break detection
Bits 3:0 ADD[3:0]: Address of the USART node
This bit-field gives the address of the USART node.
This is used in multiprocessor communication during mute mode, for wake up with address mark
detection.
Note: These 3 bits (CPOL, CPHA, LBCL) should not be written while the transmitter is enabled.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ONEBIT CTSIE CTSE RTSE DMAT DMAR SCEN NACK HDSEL IRLP IREN EIE
Reserved

Bit 11 ONEBIT: One sample bit method enable
This bit allows the user to select the sample method. When the one sample bit method is
selected the noise detection flag (NF) is disabled.
0: Three sample bit method
1: One sample bit method
Note: The ONEBIT feature applies only to data bits. It does not apply to START bit.
Bit 10 CTSIE: CTS interrupt enable
1: An interrupt is generated whenever CTS=1 in the USART_SR register
1014/1751 RM0090 Rev 19

Bit 9 CTSE: CTS enable

0: CTS hardware flow control disabled
1: CTS mode enabled, data is only transmitted when the CTS input is asserted (tied to 0). If
the CTS input is deasserted while a data is being transmitted, then the transmission is
completed before stopping. If a data is written into the data register while CTS is
deasserted, the transmission is postponed until CTS is asserted.
Bit 8 RTSE: RTS enable
0: RTS hardware flow control disabled
1: RTS interrupt enabled, data is only requested when there is space in the receive buffer.
The transmission of data is expected to cease after the current character has been
transmitted. The RTS output is asserted (tied to 0) when a data can be received.
Bit 7 DMAT: DMA enable transmitter
This bit is set/reset by software
1: DMA mode is enabled for transmission.
0: DMA mode is disabled for transmission.
Bit 6 DMAR: DMA enable receiver
This bit is set/reset by software
1: DMA mode is enabled for reception
0: DMA mode is disabled for reception
Bit 5 SCEN: Smartcard mode enable
This bit is used for enabling Smartcard mode.
0: Smartcard Mode disabled
1: Smartcard Mode enabled
Bit 4 NACK: Smartcard NACK enable
0: NACK transmission in case of parity error is disabled
1: NACK transmission during parity error is enabled
Bit 3 HDSEL: Half-duplex selection
Selection of Single-wire Half-duplex mode
0: Half duplex mode is not selected
1: Half duplex mode is selected
RM0090 Rev 19 1015/1751

1018
Bit 2 IRLP: IrDA low-power

This bit is used for selecting between normal and low-power IrDA modes
0: Normal mode
1: Low-power mode
Bit 1 IREN: IrDA mode enable
0: IrDA disabled
1: IrDA enabled
Bit 0 EIE: Error interrupt enable
Error Interrupt Enable Bit is required to enable interrupt generation in case of a framing
error, overrun error or noise flag (FE=1 or ORE=1 or NF=1 in the USART_SR register) in
case of Multi Buffer Communication (DMAR=1 in the USART_CR3 register).
1: An interrupt is generated whenever DMAR=1 in the USART_CR3 register and FE=1 or
ORE=1 or NF=1 in the USART_SR register.
1016/1751 RM0090 Rev 19

30.6.7 Guard time and prescaler register (USART_GTPR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
GT[7:0] PSC[7:0]

Bits 15:8 GT[7:0]: Guard time value
This bit-field gives the Guard time value in terms of number of baud clocks.
This is used in Smartcard mode. The Transmission Complete flag is set after this guard time
value.
– In IrDA Low-power mode:
PSC[7:0] = IrDA Low-Power Baud Rate
Used for programming the prescaler for dividing the system clock to achieve the low-power
frequency:
The source clock is divided by the value given in the register (8 significant bits):
00000000: Reserved - do not program this value
00000001: divides the source clock by 1
...
– In normal IrDA mode: PSC must be set to 00000001.
– In smartcard mode:
PSC[4:0]: Prescaler value
Used for programming the prescaler for dividing the system clock to provide the smartcard
clock.
The value given in the register (5 significant bits) is multiplied by 2 to give the division factor
of the source clock frequency:
00000: Reserved - do not program this value
...
Note: 1: Bits [7:5] have no effect if Smartcard mode is used.
2: This bit is not available for UART4 & UART5.
RM0090 Rev 19 1017/1751

1018
30.6.8 USART register map

The table below gives the USART register map and reset values.
Table 149. USART register map and reset values

Offset Register 31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
RXNE
IDLE
ORE
CTS
TXE
LBD
USART_SR
TC
NF
PE
FE
0x00 Reserved
Reset value 0 0 1 1 0 0 0 0 0 0
USART_DR DR[8:0]
0x04 Reserved
Reset value 0 0 0 0 0 0 0 0 0
DIV_Fraction
USART_BRR DIV_Mantissa[15:4]
0x08 Reserved [3:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RXNEIE
OVER8
IDLEIE
WAKE
TXEIE
Reserved
RWU
PEIE
TCIE
PCE
SBK
USART_CR1
UE
RE
PS
TE
M
0x0C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CLKEN
LINEN
LBDIE
CPHA
STOP
CPOL
Reserved
Reserved
LBCL
LBDL
USART_CR2 ADD[3:0]
0x10 Reserved [1:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0
ONEBIT
HDSEL
DMAR
CTSIE
NACK
SCEN
DMAT
CTSE
RTSE
IREN
IRLP
EIE
USART_CR3
0x14 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0
USART_GTPR GT[7:0] PSC[7:0]

0x18 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1018/1751 RM0090 Rev 19

RM0090 Secure digital input/output interface (SDIO)
31 Secure digital input/output interface (SDIO)
31.1 SDIO main features

The SD/SDIO MMC card host interface (SDIO) provides an interface between the APB2
peripheral bus and MultiMediaCards (MMCs), SD memory cards, SDIO cards and CE-ATA
devices.
The MultiMediaCard system specifications are available through the MultiMediaCard
Association website at http://www.jedec.org/, published by the MMCA technical committee.
SD memory card and SD I/O card system specifications are available through the SD card
Association website at http://www.sdcard.org.
CE-ATA system specifications are available through the CE-ATA workgroup website.
The SDIO features include the following:
• Full compliance with MultiMediaCard System Specification Version 4.2. Card support
for three different databus modes: 1-bit (default), 4-bit and 8-bit
• Full compatibility with previous versions of MultiMediaCards (forward compatibility)
• Full compliance with SD Memory Card Specifications Version 2.0
• Full compliance with SD I/O Card Specification Version 2.0: card support for two
different databus modes: 1-bit (default) and 4-bit
• Full support of the CE-ATA features (full compliance with CE-ATA digital protocol
Rev1.1)
• Data transfer up to 50 MHz for the 8 bit mode
• Data and command output enable signals to control external bidirectional drivers.
Note: The SDIO does not have an SPI-compatible communication mode.
The SD memory card protocol is a superset of the MultiMediaCard protocol as defined in the
MultiMediaCard system specification V2.11. Several commands required for SD memory
devices are not supported by either SD I/O-only cards or the I/O portion of combo cards.
Some of these commands have no use in SD I/O devices, such as erase commands, and
thus are not supported in the SDIO. In addition, several commands are different between
SD memory cards and SD I/O cards and thus are not supported in the SDIO. For details
refer to SD I/O card Specification Version 1.0. CE-ATA is supported over the MMC electrical
interface using a protocol that utilizes the existing MMC access primitives. The interface
electrical and signaling definition is as defined in the MMC reference.
The MultiMediaCard/SD bus connects cards to the controller.
The current version of the SDIO supports only one SD/SDIO/MMC4.2 card at any one time
and a stack of MMC4.1 or previous.
RM0090 Rev 19 1019/1751

1075
Secure digital input/output interface (SDIO) RM0090
31.2 SDIO bus topology

Communication over the bus is based on command and data transfers.
The basic transaction on the MultiMediaCard/SD/SD I/O bus is the command/response
transaction. These types of bus transaction transfer their information directly within the
command or response structure. In addition, some operations have a data token.
Data transfers to/from SD/SDIO memory cards are done in data blocks. Data transfers
to/from MMC are done data blocks or streams. Data transfers to/from the CE-ATA Devices
are done in data blocks.
Figure 321. SDIO “no response” and “no data” operations
From host to card(s) From host to card From card to host
SDMMC_CMD Command Command Response
SDMMC_D
Operation (no response) Operation (no data)

ai14734b
Figure 322. SDIO (multiple) block read operation
From host to card From card to host

data from card to host Stop command
stops data transfer
SDMMC_CMD Command Response Command Response
SDMMC_D Data block crc Data block crc Data block crc
Block read operation

Data stop operation
Multiple block read operation
ai14735b
1020/1751 RM0090 Rev 19

Figure 323. SDIO (multiple) block write operation
From host to card From card to host

data from host to card Stop command
stops data transfer
SDMMC_D Data block crc Busy Data block crc Busy
Data stop operation

Block write operation
Multiple block write operation

ai14737b
Note: The SDIO will not send any data as long as the Busy signal is asserted (SDIO_D0 pulled
low).
Figure 324. SDIO sequential read operation
From host to
card(s) From card to host
Stop command
Data from card to host stops data transfer
SDMMC_D Data stream
Data stop operation

Data transfer operation
ai14738b
Figure 325. SDIO sequential write operation
From host to
card(s) From card to host Stop command
stops data transfer
Data from host to card
SDMMC_D Data stream
Data stop operation

Data transfer operation
ai14739b
RM0090 Rev 19 1021/1751

1075
31.3 SDIO functional description

The SDIO consists of two parts:
• The SDIO adapter block provides all functions specific to the MMC/SD/SD I/O card
such as the clock generation unit, command and data transfer.
• The APB2 interface accesses the SDIO adapter registers, and generates interrupt and
DMA request signals.
Figure 326. SDIO block diagram
SDMMC
SDMMC_CK
Interrupts and
DMA request SDMMC_CMD
APB2 SDMMC SDMMC_D[7:0]

interface adapter
APB2 bus
PCLK2 SDMMCCLK
ai15898b
By default SDIO_D0 is used for data transfer. After initialization, the host can change the
databus width.
If a MultiMediaCard is connected to the bus, SDIO_D0, SDIO_D[3:0] or SDIO_D[7:0] can be
used for data transfer. MMC V3.31 or previous, supports only 1 bit of data so only SDIO_D0
can be used.
If an SD or SD I/O card is connected to the bus, data transfer can be configured by the host
to use SDIO_D0 or SDIO_D[3:0]. All data lines are operating in push-pull mode.
SDIO_CMD has two operational modes:
• Open-drain for initialization (only for MMCV3.31 or previous)
• Push-pull for command transfer (SD/SD I/O card MMC4.2 use push-pull drivers also for
initialization)
SDIO_CK is the clock to the card: one bit is transferred on both command and data lines
with each clock cycle.
The SDIO uses two clock signals:
• SDIO adapter clock SDIOCLK up to 50 MHz (48 MHz when in use with USB)
• APB2 bus clock (PCLK2)
PCLK2 and SDIO_CK clock frequencies must respect the following condition:
Frequenc ( PCLK2 ) ≥ 3 ⁄ 8 × Frequency ( SDIO_CK )
The signals shown in Table 150 are used on the MultiMediaCard/SD/SD I/O card bus.
1022/1751 RM0090 Rev 19

Table 150. SDIO I/O definitions

Pin Direction Description
MultiMediaCard/SD/SDIO card clock. This pin is the clock from

SDIO_CK Output
host to card.
MultiMediaCard/SD/SDIO card command. This pin is the
SDIO_CMD Bidirectional
bidirectional command/response signal.
MultiMediaCard/SD/SDIO card data. These pins are the
SDIO_D[7:0] Bidirectional
bidirectional databus.
31.3.1 SDIO adapter

Figure 327 shows a simplified block diagram of an SDIO adapter.
Figure 327. SDIO adapter
SDMMC adapter
Control unit SDMMC_CK
Card bus
Command
SDMMC_CMD
Adapter path
registers
To APB2
Data path SDMMC_D[7:0]
interface FIFO
PCLK2 SDMMCCLK
ai15899b
The SDIO adapter is a multimedia/secure digital memory card bus master that provides an
interface to a multimedia card stack or to a secure digital memory card. It consists of five
subunits:
• Adapter register block
• Control unit
• Command path
• Data path
• Data FIFO
Note: The adapter registers and FIFO use the APB2 bus clock domain (PCLK2). The control unit,
command path and data path use the SDIO adapter clock domain (SDIOCLK).
Adapter register block

The adapter register block contains all system registers. This block also generates the
signals that clear the static flags in the multimedia card. The clear signals are generated
when 1 is written into the corresponding bit location in the SDIO Clear register.
RM0090 Rev 19 1023/1751

1075
Control unit
The control unit contains the power management functions and the clock divider for the
memory card clock.
There are three power phases:
• power-off
• power-up
• power-on
Figure 328. Control unit
Control unit
Power management
Adapter Clock management SDMMC_CK

registers
To command and data path

ai14804b
The control unit is illustrated in Figure 328. It consists of a power management subunit and
a clock management subunit.
The power management subunit disables the card bus output signals during the power-off
and power-up phases.
The clock management subunit generates and controls the SDIO_CK signal. The SDIO_CK
output can use either the clock divide or the clock bypass mode. The clock output is
inactive:
• after reset
• during the power-off or power-up phases
• if the power saving mode is enabled and the card bus is in the Idle state (eight clock
periods after both the command and data path subunits enter the Idle phase)
1024/1751 RM0090 Rev 19

Command path
The command path unit sends commands to and receives responses from the cards.
Figure 329. SDIO adapter command path
To control unit Status Control Command

flag logic timer
Adapter registers
SDMMC_CMDin
CMD
Argument
CRC SDMMC_CMDout
Shift
CMD register
Response
To APB2 interface
registers
ai15900b
• Command path state machine (CPSM)

– When the command register is written to and the enable bit is set, command
transfer starts. When the command has been sent, the command path state
machine (CPSM) sets the status flags and enters the Idle state if a response is not
required. If a response is required, it waits for the response (see Figure 330 on
page 1026). When the response is received, the received CRC code and the
internally generated code are compared, and the appropriate status flags are set.
RM0090 Rev 19 1025/1751

1075
Figure 330. Command path state machine (CPSM)
CE-ATA Command
On reset Completion signal
received or Wait_CPL
CPSM disabled or
Command CRC failed
CPSM Enabled and Idle Response received or Response Received in CE-ATA

pending command
disabled or command mode and no interrupt and
CPSM CRC failed wait for CE-ATA Command
disabled Completion signal enabled
Pend
Enabled and
CPSM Disabled or
command start Receive
CPSM disabled or command timeout
no response
Last Data
Response
started
Send
Wait for response Wait
Response Received in CE-ATA mode and

no interrupt and wait for CE-ATA
Command Completion signal disabled ai14806b
When the Wait state is entered, the command timer starts running. If the timeout is reached
before the CPSM moves to the Receive state, the timeout flag is set and the Idle state is
entered.
Note: The command timeout has a fixed value of 64 SDIO_CK clock periods.
If the interrupt bit is set in the command register, the timer is disabled and the CPSM waits
for an interrupt request from one of the cards. If a pending bit is set in the command register,
the CPSM enters the Pend state, and waits for a CmdPend signal from the data path
subunit. When CmdPend is detected, the CPSM moves to the Send state. This enables the
data counter to trigger the stop command transmission.
Note: The CPSM remains in the Idle state for at least eight SDIO_CK periods to meet the NCC and
NRC timing constraints. NCC is the minimum delay between two host commands, and NRC is
the minimum delay between the host command and the card response.
1026/1751 RM0090 Rev 19

Figure 331. SDIO command transfer
at least 8 SDMMC_CK
cycles
SDMMC_CK Command Response Command
State Idle Send Wait Receive Idle Send
SDMMC_CMD Hi-Z Controller drives Hi-Z Card drives Hi-Z Controller drives
ai14807b
• Command format
– Command: a command is a token that starts an operation. Command are sent
from the host either to a single card (addressed command) or to all connected
cards (broadcast command are available for MMC V3.31 or previous). Commands
are transferred serially on the CMD line. All commands have a fixed length of 48
bits. The general format for a command token for MultiMediaCards, SD-Memory
cards and SDIO-Cards is shown in Table 151. CE-ATA commands are an
extension of MMC commands V4.2, and so have the same format.
The command path operates in a half-duplex mode, so that commands and
responses can either be sent or received. If the CPSM is not in the Send state, the
SDIO_CMD output is in the Hi-Z state, as shown in Figure 331 on page 1027.
Data on SDIO_CMD are synchronous with the rising edge of SDIO_CK. Table
shows the command format.
Table 151. Command format

Bit position Width Value Description
47 1 0 Start bit
46 1 1 Transmission bit
[45:40] 6 - Command index
[39:8] 32 - Argument
[7:1] 7 - CRC7
0 1 1 End bit
– Response: a response is a token that is sent from an addressed card (or

synchronously from all connected cards for MMC V3.31 or previous), to the host
as an answer to a previously received command. Responses are transferred
serially on the CMD line.
The SDIO supports two response types. Both use CRC error checking:
• 48 bit short response
• 136 bit long response
Note: If the response does not contain a CRC (CMD1 response), the device driver must ignore the
CRC failed status.
RM0090 Rev 19 1027/1751

1075
Table 152. Short response format

47 1 0 Start bit
[45:40] 6 - Command index
[39:8] 32 - Argument
[7:1] 7 - CRC7(or 1111111)
0 1 1 End bit
Table 153. Long response format

135 1 0 Start bit

[133:128] 6 111111 Reserved
[127:1] 127 - CID or CSD (including internal CRC7)
0 1 1 End bit
The command register contains the command index (six bits sent to a card) and the
command type. These determine whether the command requires a response, and whether
the response is 48 or 136 bits long (see Section 31.9.4 on page 1062). The command path
implements the status flags shown in Table 154:
Table 154. Command path status flags

Flag Description
CMDREND Set if response CRC is OK.
CCRCFAIL Set if response CRC fails.
CMDSENT Set when command (that does not require response) is sent
CTIMEOUT Response timeout.
CMDACT Command transfer in progress.
The CRC generator calculates the CRC checksum for all bits before the CRC code. This
includes the start bit, transmitter bit, command index, and command argument (or card
status). The CRC checksum is calculated for the first 120 bits of CID or CSD for the long
response format. Note that the start bit, transmitter bit and the six reserved bits are not used
in the CRC calculation.
The CRC checksum is a 7-bit value:
CRC[6:0] = Remainder [(M(x) * x7) / G(x)]
G(x) = x7 + x3 + 1
M(x) = (start bit) * x39 + ... + (last bit before CRC) * x0, or
M(x) = (start bit) * x119 + ... + (last bit before CRC) * x0
1028/1751 RM0090 Rev 19

Data path
The data path subunit transfers data to and from cards. Figure 332 shows a block diagram
of the data path.
Figure 332. Data path
Data path
To control unit Status Control Data

flag logic timer
Data FIFO SDMMC_Din[7:0]
Transmit
CRC SDMMC_Dout[7:0]
Shift
register
Receive
ai14808b
The card databus width can be programmed using the clock control register. If the 4-bit wide
bus mode is enabled, data is transferred at four bits per clock cycle over all four data signals
(SDIO_D[3:0]). If the 8-bit wide bus mode is enabled, data is transferred at eight bits per
clock cycle over all eight data signals (SDIO_D[7:0]). If the wide bus mode is not enabled,
only one bit per clock cycle is transferred over SDIO_D0.
Depending on the transfer direction (send or receive), the data path state machine (DPSM)
moves to the Wait_S or Wait_R state when it is enabled:
• Send: the DPSM moves to the Wait_S state. If there is data in the transmit FIFO, the
DPSM moves to the Send state, and the data path subunit starts sending data to a
card.
• Receive: the DPSM moves to the Wait_R state and waits for a start bit. When it
receives a start bit, the DPSM moves to the Receive state, and the data path subunit
starts receiving data from a card.
Data path state machine (DPSM)
The DPSM operates at SDIO_CK frequency. Data on the card bus signals is synchronous to
the rising edge of SDIO_CK. The DPSM has six states, as shown in Figure 333: Data path
state machine (DPSM).
RM0090 Rev 19 1029/1751

1075
Figure 333. Data path state machine (DPSM)
On reset DPSM disabled
DPSM enabled and Read Wait

Read Wait Started
and SD I/O mode enabled
Disabled or FIFO underrun or
end of data or CRC fail Idle
Disabled or CRC fail
or timeout Enable and not send
ReadWait Stop
Disabled or
end of data
Disabled or
Busy Rx FIFO empty or timeout or
start bit error
Not busy
Enable and send Data received and
Wait_R Read Wait Started and
SD I/O mode enabled
End of packet
Wait_S End of packet or

end of data or
FIFO overrun
Disabled or CRC fail
Data ready Start bit
Send
Receive
ai14809b
• Idle: the data path is inactive, and the SDIO_D[7:0] outputs are in Hi-Z. When the data
control register is written and the enable bit is set, the DPSM loads the data counter
with a new value and, depending on the data direction bit, moves to either the Wait_S
or the Wait_R state.
• Wait_R: if the data counter equals zero, the DPSM moves to the Idle state when the
receive FIFO is empty. If the data counter is not zero, the DPSM waits for a start bit on
SDIO_D. The DPSM moves to the Receive state if it receives a start bit before a
timeout, and loads the data block counter. If it reaches a timeout before it detects a
start bit, or a start bit error occurs, it moves to the Idle state and sets the timeout status
flag.
• Receive: serial data received from a card is packed in bytes and written to the data
FIFO. Depending on the transfer mode bit in the data control register, the data transfer
mode can be either block or stream:
– In block mode, when the data block counter reaches zero, the DPSM waits until it
receives the CRC code. If the received code matches the internally generated
1030/1751 RM0090 Rev 19

CRC code, the DPSM moves to the Wait_R state. If not, the CRC fail status flag is
set and the DPSM moves to the Idle state.
– In stream mode, the DPSM receives data while the data counter is not zero. When
the counter is zero, the remaining data in the shift register is written to the data
FIFO, and the DPSM moves to the Wait_R state.
If a FIFO overrun error occurs, the DPSM sets the FIFO error flag and moves to the
Idle state:
• Wait_S: the DPSM moves to the Idle state if the data counter is zero. If not, it waits until
the data FIFO empty flag is deasserted, and moves to the Send state.
Note: The DPSM remains in the Wait_S state for at least two clock periods to meet the NWR timing
requirements, where NWR is the number of clock cycles between the reception of the card
response and the start of the data transfer from the host.
• Send: the DPSM starts sending data to a card. Depending on the transfer mode bit in
the data control register, the data transfer mode can be either block or stream:
– In block mode, when the data block counter reaches zero, the DPSM sends an
internally generated CRC code and end bit, and moves to the Busy state.
– In stream mode, the DPSM sends data to a card while the enable bit is high and
the data counter is not zero. It then moves to the Idle state.
If a FIFO underrun error occurs, the DPSM sets the FIFO error flag and moves to the
Idle state.
• Busy: the DPSM waits for the CRC status flag:
– If it does not receive a positive CRC status, it moves to the Idle state and sets the
CRC fail status flag.
– If it receives a positive CRC status, it moves to the Wait_S state if SDIO_D0 is not
low (the card is not busy).
If a timeout occurs while the DPSM is in the Busy state, it sets the data timeout flag and
moves to the Idle state.
The data timer is enabled when the DPSM is in the Wait_R or Busy state, and
generates the data timeout error:
– When transmitting data, the timeout occurs if the DPSM stays in the Busy state for
longer than the programmed timeout period
– When receiving data, the timeout occurs if the end of the data is not true, and if the
DPSM stays in the Wait_R state for longer than the programmed timeout period.
• Data: data can be transferred from the card to the host or vice versa. Data is
transferred via the data lines. They are stored in a FIFO of 32 words, each word is 32
bits wide.
Table 155. Data token format

Description Start bit Data CRC16 End bit
Block Data 0 - yes 1

Stream Data 0 - no 1
RM0090 Rev 19 1031/1751

1075
Data FIFO
The data FIFO (first-in-first-out) subunit is a data buffer with a transmit and receive unit.
The FIFO contains a 32-bit wide, 32-word deep data buffer, and transmit and receive logic.
Because the data FIFO operates in the APB2 clock domain (PCLK2), all signals from the
subunits in the SDIO clock domain (SDIOCLK) are resynchronized.
Depending on the TXACT and RXACT flags, the FIFO can be disabled, transmit enabled, or
receive enabled. TXACT and RXACT are driven by the data path subunit and are mutually
exclusive:
– The transmit FIFO refers to the transmit logic and data buffer when TXACT is
asserted
– The receive FIFO refers to the receive logic and data buffer when RXACT is
asserted
• Transmit FIFO:
Data can be written to the transmit FIFO through the APB2 interface when the SDIO is
enabled for transmission.
The transmit FIFO is accessible via 32 sequential addresses. The transmit FIFO
contains a data output register that holds the data word pointed to by the read pointer.
When the data path subunit has loaded its shift register, it increments the read pointer
and drives new data out.
If the transmit FIFO is disabled, all status flags are deasserted. The data path subunit
asserts TXACT when it transmits data.
Table 156. Transmit FIFO status flags

Flag Description
TXFIFOF Set to high when all 32 transmit FIFO words contain valid data.
TXFIFOE Set to high when the transmit FIFO does not contain valid data.
Set to high when 8 or more transmit FIFO words are empty. This flag can be used
TXFIFOHE
as a DMA request.
Set to high when the transmit FIFO contains valid data. This flag is the inverse of
TXDAVL
the TXFIFOE flag.
Set to high when an underrun error occurs. This flag is cleared by writing to the
TXUNDERR
SDIO Clear register.
• Receive FIFO
When the data path subunit receives a word of data, it drives the data on the write
databus. The write pointer is incremented after the write operation completes. On the
read side, the contents of the FIFO word pointed to by the current value of the read
pointer is driven onto the read databus. If the receive FIFO is disabled, all status flags
are deasserted, and the read and write pointers are reset. The data path subunit
asserts RXACT when it receives data. Table 157 lists the receive FIFO status flags.
The receive FIFO is accessible via 32 sequential addresses.
1032/1751 RM0090 Rev 19

Table 157. Receive FIFO status flags

Flag Description
RXFIFOF Set to high when all 32 receive FIFO words contain valid data
RXFIFOE Set to high when the receive FIFO does not contain valid data.
Set to high when 8 or more receive FIFO words contain valid data. This flag can be
RXFIFOHF
used as a DMA request.
Set to high when the receive FIFO is not empty. This flag is the inverse of the
RXDAVL
RXFIFOE flag.
Set to high when an overrun error occurs. This flag is cleared by writing to the SDIO
RXOVERR
Clear register.
31.3.2 SDIO APB2 interface

The APB2 interface generates the interrupt and DMA requests, and accesses the SDIO
adapter registers and the data FIFO. It consists of a data path, register decoder, and
interrupt/DMA logic.
SDIO interrupts
The interrupt logic generates an interrupt request signal that is asserted when at least one
of the selected status flags is high. A mask register is provided to allow selection of the
conditions that will generate an interrupt. A status flag generates the interrupt request if a
corresponding mask flag is set.
SDIO/DMA interface - procedure for data transfers between the SDIO and
memory
In the example shown, the transfer is from the SDIO host controller to an MMC (512 bytes
using CMD24 (WRITE_BLOCK). The SDIO FIFO is filled by data stored in a memory using
the DMA controller.
1. Do the card identification process
2. Increase the SDIO_CK frequency
3. Select the card by sending CMD7
4. Configure the DMA2 as follows:
a) Enable DMA2 controller and clear any pending interrupts.
b) Program the DMA2_Stream3 or DMA2_Stream6 Channel4 source address
register with the memory location’s base address and DMA2_Stream3 or
RM0090 Rev 19 1033/1751

1075
DMA2_Stream6 Channel4 destination address register with the SDIO_FIFO

register address.
c) Program DMA2_Stream3 or DMA2_Stream6 Channel4 control register (memory
increment, not peripheral increment, peripheral and source width is word size).
d) Program DMA2_Stream3 or DMA2_Stream6 Channel4 to select the peripheral as
flow controller (set PFCTRL bit in DMA_S3CR or DMA_S6CR configuration
register).
e) Configure the incremental burst transfer to 4 beats (at least from peripheral side)
in DMA2_Stream3 or DMA2_Stream6 Channel4.
f) Enable DMA2_Stream3 or DMA2_Stream6 Channel4
5. Send CMD24 (WRITE_BLOCK) as follows:
a) Program the SDIO data length register (SDIO data timer register should be
already programmed before the card identification process).
b) Program the SDIO argument register with the address location of the card where
data is to be transferred.
c) Program the SDIO command register: CmdIndex with 24 (WRITE_BLOCK);
WaitResp with ‘1’ (SDIO card host waits for a response); CPSMEN with ‘1’ (SDIO
card host enabled to send a command). Other fields are at their reset value.
d) Wait for SDIO_STA[6] = CMDREND interrupt, then program the SDIO data control
register: DTEN with ‘1’ (SDIO card host enabled to send data); DTDIR with ‘0’
(from controller to card); DTMODE with ‘0’ (block data transfer); DMAEN with ‘1’
(DMA enabled); DBLOCKSIZE with 0x9 (512 bytes). Other fields are don’t care.
e) Wait for SDIO_STA[10] = DBCKEND.
6. Check that no channels are still enabled by polling the DMA Enabled Channel Status
register.
31.4 Card functional description
31.4.1 Card identification mode

While in card identification mode the host resets all cards, validates the operation voltage
range, identifies cards and sets a relative card address (RCA) for each card on the bus. All
data communications in the card identification mode use the command line (CMD) only.
31.4.2 Card reset

The GO_IDLE_STATE command (CMD0) is the software reset command and it puts the
MultiMediaCard and SD memory in the Idle state. The IO_RW_DIRECT command (CMD52)
resets the SD I/O card. After power-up or CMD0, all cards output bus drivers are in the high-
impedance state and the cards are initialized with a default relative card address
(RCA=0x0001) and with a default driver stage register setting (lowest speed, highest driving
current capability).
31.4.3 Operating voltage range validation

All cards can communicate with the SDIO card host using any operating voltage within the
specification range. The supported minimum and maximum VDD values are defined in the
operation conditions register (OCR) on the card.
1034/1751 RM0090 Rev 19

Cards that store the card identification number (CID) and card specific data (CSD) in the
payload memory are able to communicate this information only under data-transfer VDD
conditions. When the SDIO card host module and the card have incompatible VDD ranges,
the card is not able to complete the identification cycle and cannot send CSD data. For this
purpose, the special commands, SEND_OP_COND (CMD1), SD_APP_OP_COND (ACMD41
for SD Memory), and IO_SEND_OP_COND (CMD5 for SD I/O), are designed to provide a
mechanism to identify and reject cards that do not match the VDD range desired by the
SDIO card host. The SDIO card host sends the required VDD voltage window as the
operand of these commands. Cards that cannot perform data transfer in the specified range
disconnect from the bus and go to the inactive state.
By using these commands without including the voltage range as the operand, the SDIO
card host can query each card and determine the common voltage range before placing out-
of-range cards in the inactive state. This query is used when the SDIO card host is able to
select a common voltage range or when the user requires notification that cards are not
usable.
31.4.4 Card identification process

The card identification process differs for MultiMediaCards and SD cards. For
MultiMediaCard cards, the identification process starts at clock rate Fod. The SDIO_CMD
line output drivers are open-drain and allow parallel card operation during this process. The
registration process is accomplished as follows:
1. The bus is activated.
2. The SDIO card host broadcasts SEND_OP_COND (CMD1) to receive operation
conditions.
3. The response is the wired AND operation of the operation condition registers from all
cards.
4. Incompatible cards are placed in the inactive state.
5. The SDIO card host broadcasts ALL_SEND_CID (CMD2) to all active cards.
6. The active cards simultaneously send their CID numbers serially. Cards with outgoing
CID bits that do not match the bits on the command line stop transmitting and must wait
for the next identification cycle. One card successfully transmits a full CID to the SDIO
card host and enters the Identification state.
7. The SDIO card host issues SET_RELATIVE_ADDR (CMD3) to that card. This new
address is called the relative card address (RCA); it is shorter than the CID and
addresses the card. The assigned card changes to the Standby state, it does not react
to further identification cycles, and its output switches from open-drain to push-pull.
8. The SDIO card host repeats steps 5 through 7 until it receives a timeout condition.
For the SD card, the identification process starts at clock rate Fod, and the SDIO_CMD line
output drives are push-pull drivers instead of open-drain. The registration process is
accomplished as follows:
RM0090 Rev 19 1035/1751

1075

2. The SDIO card host broadcasts SD_APP_OP_COND (ACMD41).
3. The cards respond with the contents of their operation condition registers.
4. The incompatible cards are placed in the inactive state.
5. The SDIO card host broadcasts ALL_SEND_CID (CMD2) to all active cards.
6. The cards send back their unique card identification numbers (CIDs) and enter the
Identification state.
7. The SDIO card host issues SET_RELATIVE_ADDR (CMD3) to an active card with an
address. This new address is called the relative card address (RCA); it is shorter than
the CID and addresses the card. The assigned card changes to the Standby state. The
SDIO card host can reissue this command to change the RCA. The RCA of the card is
the last assigned value.
8. The SDIO card host repeats steps 5 through 7 with all active cards.
For the SD I/O card, the registration process is accomplished as follows:
2. The SDIO card host sends IO_SEND_OP_COND (CMD5).
3. The cards respond with the contents of their operation condition registers.
4. The incompatible cards are set to the inactive state.
5. The SDIO card host issues SET_RELATIVE_ADDR (CMD3) to an active card with an
address. This new address is called the relative card address (RCA); it is shorter than
the CID and addresses the card. The assigned card changes to the Standby state. The
SDIO card host can reissue this command to change the RCA. The RCA of the card is
the last assigned value.
31.4.5 Block write

During block write (CMD24 - 27) one or more blocks of data are transferred from the host to
the card with a CRC appended to the end of each block by the host. A card supporting block
write is always able to accept a block of data defined by WRITE_BL_LEN. If the CRC fails,
the card indicates the failure on the SDIO_D line and the transferred data are discarded and
not written, and all further transmitted blocks (in multiple block write mode) are ignored.
If the host uses partial blocks whose accumulated length is not block aligned and, block
misalignment is not allowed (CSD parameter WRITE_BLK_MISALIGN is not set), the card
will detect the block misalignment error before the beginning of the first misaligned block.
(ADDRESS_ERROR error bit is set in the status register). The write operation will also be
aborted if the host tries to write over a write-protected area. In this case, however, the card
will set the WP_VIOLATION bit.
Programming of the CID and CSD registers does not require a previous block length setting.
The transferred data is also CRC protected. If a part of the CSD or CID register is stored in
ROM, then this unchangeable part must match the corresponding part of the receive buffer.
If this match fails, then the card reports an error and does not change any register contents.
Some cards may require long and unpredictable times to write a block of data. After
receiving a block of data and completing the CRC check, the card begins writing and holds
the SDIO_D line low if its write buffer is full and unable to accept new data from a new
WRITE_BLOCK command. The host may poll the status of the card with a SEND_STATUS
command (CMD13) at any time, and the card will respond with its status. The
READY_FOR_DATA status bit indicates whether the card can accept new data or whether
the write process is still in progress. The host may deselect the card by issuing CMD7 (to
1036/1751 RM0090 Rev 19

select a different card), which will place the card in the Disconnect state and release the
SDIO_D line(s) without interrupting the write operation. When selecting the card again, it will
reactivate busy indication by pulling SDIO_D to low if programming is still in progress and
the write buffer is unavailable.
31.4.6 Block read

In Block read mode the basic unit of data transfer is a block whose maximum size is defined
in the CSD (READ_BL_LEN). If READ_BL_PARTIAL is set, smaller blocks whose start and
end addresses are entirely contained within one physical block (as defined by
READ_BL_LEN) may also be transmitted. A CRC is appended to the end of each block,
ensuring data transfer integrity. CMD17 (READ_SINGLE_BLOCK) initiates a block read and
after completing the transfer, the card returns to the Transfer state.
CMD18 (READ_MULTIPLE_BLOCK) starts a transfer of several consecutive blocks.
The host can abort reading at any time, within a multiple block operation, regardless of its
type. Transaction abort is done by sending the stop transmission command.
If the card detects an error (for example, out of range, address misalignment or internal
error) during a multiple block read operation (both types) it stops the data transmission and
remains in the data state. The host must than abort the operation by sending the stop
transmission command. The read error is reported in the response to the stop transmission
command.
If the host sends a stop transmission command after the card transmits the last block of a
multiple block operation with a predefined number of blocks, it is responded to as an illegal
command, since the card is no longer in the data state. If the host uses partial blocks whose
accumulated length is not block-aligned and block misalignment is not allowed, the card
detects a block misalignment error condition at the beginning of the first misaligned block
(ADDRESS_ERROR error bit is set in the status register).
31.4.7 Stream access, stream write and stream read

(MultiMediaCard only)
In stream mode, data is transferred in bytes and no CRC is appended at the end of each
block.
Stream write (MultiMediaCard only)

WRITE_DAT_UNTIL_STOP (CMD20) starts the data transfer from the SDIO card host to the
card, beginning at the specified address and continuing until the SDIO card host issues a
stop command. When partial blocks are allowed (CSD parameter WRITE_BL_PARTIAL is
set), the data stream can start and stop at any address within the card address space,
otherwise it can only start and stop at block boundaries. Because the amount of data to be
transferred is not determined in advance, a CRC cannot be used. When the end of the
memory range is reached while sending data and no stop command is sent by the SD card
host, any additional transferred data are discarded.
RM0090 Rev 19 1037/1751

1075
The maximum clock frequency for a stream write operation is given by the following
equation fields of the card-specific data register:
8 × 2 writebllen ) ( – NSAC ))
Maximumspeed = MIN (TRANSPEED,(------------------------------------------------------------------------
TAAC × R2WFACTOR
• Maximumspeed = maximum write frequency

• TRANSPEED = maximum data transfer rate
• writebllen = maximum write data block length
• NSAC = data read access time 2 in CLK cycles
• TAAC = data read access time 1
• R2WFACTOR = write speed factor
If the host attempts to use a higher frequency, the card may not be able to process the data
and stop programming, set the OVERRUN error bit in the status register, and while ignoring
all further data transfer, wait (in the receive data state) for a stop command. The write
operation is also aborted if the host tries to write over a write-protected area. In this case,
however, the card sets the WP_VIOLATION bit.
Stream read (MultiMediaCard only)

READ_DAT_UNTIL_STOP (CMD11) controls a stream-oriented data transfer.
This command instructs the card to send its data, starting at a specified address, until the
SDIO card host sends STOP_TRANSMISSION (CMD12). The stop command has an
execution delay due to the serial command transmission and the data transfer stops after
the end bit of the stop command. When the end of the memory range is reached while
sending data and no stop command is sent by the SDIO card host, any subsequent data
sent are considered undefined.
The maximum clock frequency for a stream read operation is given by the following
equation and uses fields of the card specific data register.
8 × 2 readbllen ) ( – NSAC ))
Maximumspeed = MIN (TRANSPEED,(-----------------------------------------------------------------------
TAAC × R2WFACTOR
• Maximumspeed = maximum read frequency

• TRANSPEED = maximum data transfer rate
• readbllen = maximum read data block length
• writebllen = maximum write data block length
• NSAC = data read access time 2 in CLK cycles
• TAAC = data read access time 1
• R2WFACTOR = write speed factor
If the host attempts to use a higher frequency, the card is not able to sustain data transfer. If
this happens, the card sets the UNDERRUN error bit in the status register, aborts the
transmission and waits in the data state for a stop command.
1038/1751 RM0090 Rev 19

31.4.8 Erase: group erase and sector erase

The erasable unit of the MultiMediaCard is the erase group. The erase group is measured in
write blocks, which are the basic writable units of the card. The size of the erase group is a
card-specific parameter and defined in the CSD.
The host can erase a contiguous range of Erase Groups. Starting the erase process is a
three-step sequence.
First the host defines the start address of the range using the ERASE_GROUP_START
(CMD35) command, next it defines the last address of the range using the
ERASE_GROUP_END (CMD36) command and, finally, it starts the erase process by issuing
the ERASE (CMD38) command. The address field in the erase commands is an Erase
Group address in byte units. The card ignores all LSBs below the Erase Group size,
effectively rounding the address down to the Erase Group boundary.
If an erase command is received out of sequence, the card sets the ERASE_SEQ_ERROR
bit in the status register and resets the whole sequence.
If an out-of-sequence (neither of the erase commands, except SEND_STATUS) command
received, the card sets the ERASE_RESET status bit in the status register, resets the erase
sequence and executes the last command.
If the erase range includes write protected blocks, they are left intact and only unprotected
blocks are erased. The WP_ERASE_SKIP status bit in the status register is set.
The card indicates that an erase is in progress by holding SDIO_D low. The actual erase
time may be quite long, and the host may issue CMD7 to deselect the card.
31.4.9 Wide bus selection or deselection

Wide bus (4-bit bus width) operation mode is selected or deselected using
SET_BUS_WIDTH (ACMD6). The default bus width after power-up or GO_IDLE_STATE
(CMD0) is 1 bit. SET_BUS_WIDTH (ACMD6) is only valid in a transfer state, which means
that the bus width can be changed only after a card is selected by
SELECT/DESELECT_CARD (CMD7).
31.4.10 Protection management

Three write protection methods for the cards are supported in the SDIO card host module:
1. internal card write protection (card responsibility)
2. mechanical write protection switch (SDIO card host module responsibility only)
3. password-protected card lock operation
Internal card write protection

Card data can be protected against write and erase. By setting the permanent or temporary
write-protect bits in the CSD, the entire card can be permanently write-protected by the
manufacturer or content provider. For cards that support write protection of groups of
sectors by setting the WP_GRP_ENABLE bit in the CSD, portions of the data can be
protected, and the write protection can be changed by the application. The write protection
is in units of WP_GRP_SIZE sectors as specified in the CSD. The SET_WRITE_PROT and
CLR_WRITE_PROT commands control the protection of the addressed group. The
SEND_WRITE_PROT command is similar to a single block read command. The card sends
a data block containing 32 write protection bits (representing 32 write protect groups starting
RM0090 Rev 19 1039/1751

1075
at the specified address) followed by 16 CRC bits. The address field in the write protect
commands is a group address in byte units.
The card ignores all LSBs below the group size.
Mechanical write protect switch

A mechanical sliding tab on the side of the card allows the user to set or clear the write
protection on a card. When the sliding tab is positioned with the window open, the card is
write-protected, and when the window is closed, the card contents can be changed. A
matched switch on the socket side indicates to the SDIO card host module that the card is
write-protected. The SDIO card host module is responsible for protecting the card. The
position of the write protect switch is unknown to the internal circuitry of the card.
Password protect
The password protection feature enables the SDIO card host module to lock and unlock a
card with a password. The password is stored in the 128-bit PWD register and its size is set
in the 8-bit PWD_LEN register. These registers are nonvolatile so that a power cycle does
not erase them. Locked cards respond to and execute certain commands. This means that
the SDIO card host module is allowed to reset, initialize, select, and query for status,
however it is not allowed to access data on the card. When the password is set (as indicated
by a nonzero value of PWD_LEN), the card is locked automatically after power-up. As with
the CSD and CID register write commands, the lock/unlock commands are available in the
transfer state only. In this state, the command does not include an address argument and
the card must be selected before using it. The card lock/unlock commands have the
structure and bus transaction types of a regular single-block write command. The
transferred data block includes all of the required information for the command (the
password setting mode, the PWD itself, and card lock/unlock). The command data block
size is defined by the SDIO card host module before it sends the card lock/unlock
command, and has the structure shown in Table 171.
The bit settings are as follows:
• ERASE: setting it forces an erase operation. All other bits must be zero, and only the
command byte is sent
• LOCK_UNLOCK: setting it locks the card. LOCK_UNLOCK can be set simultaneously
with SET_PWD, however not with CLR_PWD
• CLR_PWD: setting it clears the password data
• SET_PWD: setting it saves the password data to memory
• PWD_LEN: it defines the length of the password in bytes
• PWD: the password (new or currently used, depending on the command)
The following sections list the command sequences to set/reset a password, lock/unlock the
card, and force an erase.
Setting the password

1. Select a card (SELECT/DESELECT_CARD, CMD7), if none is already selected.
2. Define the block length (SET_BLOCKLEN, CMD16) to send, given by the 8-bit card
lock/unlock mode, the 8-bit PWD_LEN, and the number of bytes of the new password.
1040/1751 RM0090 Rev 19

When a password replacement is done, the block size must take into account that both
the old and the new passwords are sent with the command.
3. Send LOCK/UNLOCK (CMD42) with the appropriate data block size on the data line
including the 16-bit CRC. The data block indicates the mode (SET_PWD = 1), the
length (PWD_LEN), and the password (PWD) itself. When a password replacement is
done, the length value (PWD_LEN) includes the length of both passwords, the old and
the new one, and the PWD field includes the old password (currently used) followed by
the new password.
4. When the password is matched, the new password and its size are saved into the PWD
and PWD_LEN fields, respectively. When the old password sent does not correspond
(in size and/or content) to the expected password, the LOCK_UNLOCK_FAILED error
bit is set in the card status register, and the password is not changed.
The password length field (PWD_LEN) indicates whether a password is currently set. When
this field is nonzero, there is a password set and the card locks itself after power-up. It is
possible to lock the card immediately in the current power session by setting the
LOCK_UNLOCK bit (while setting the password) or sending an additional command for card
locking.
Resetting the password

lock/unlock mode, the 8-bit PWD_LEN, and the number of bytes in the currently used
password.
including the 16-bit CRC. The data block indicates the mode (CLR_PWD = 1), the
length (PWD_LEN) and the password (PWD) itself. The LOCK_UNLOCK bit is ignored.
4. When the password is matched, the PWD field is cleared and PWD_LEN is set to 0.
When the password sent does not correspond (in size and/or content) to the expected
password, the LOCK_UNLOCK_FAILED error bit is set in the card status register, and
the password is not changed.
Locking a card
lock/unlock mode (byte 0 in Table 171), the 8-bit PWD_LEN, and the number of bytes
of the current password.
including the 16-bit CRC. The data block indicates the mode (LOCK_UNLOCK = 1), the
length (PWD_LEN), and the password (PWD) itself.
4. When the password is matched, the card is locked and the CARD_IS_LOCKED status
bit is set in the card status register. When the password sent does not correspond (in
size and/or content) to the expected password, the LOCK_UNLOCK_FAILED error bit
is set in the card status register, and the lock fails.
It is possible to set the password and to lock the card in the same sequence. In this case,
the SDIO card host module performs all the required steps for setting the password (see
Setting the password on page 1040), however it is necessary to set the LOCK_UNLOCK bit
in Step 3 when the new password command is sent.
RM0090 Rev 19 1041/1751

1075
When the password is previously set (PWD_LEN is not 0), the card is locked automatically
after power-on reset. An attempt to lock a locked card or to lock a card that does not have a
password fails and the LOCK_UNLOCK_FAILED error bit is set in the card status register.
Unlocking the card

2. Define the block length (SET_BLOCKLEN, CMD16) to send, given by the 8-bit
cardlock/unlock mode (byte 0 in Table 171), the 8-bit PWD_LEN, and the number of
bytes of the current password.
including the 16-bit CRC. The data block indicates the mode (LOCK_UNLOCK = 0), the
length (PWD_LEN), and the password (PWD) itself.
4. When the password is matched, the card is unlocked and the CARD_IS_LOCKED
status bit is cleared in the card status register. When the password sent is not correct in
size and/or content and does not correspond to the expected password, the
LOCK_UNLOCK_FAILED error bit is set in the card status register, and the card
remains locked.
The unlocking function is only valid for the current power session. When the PWD field is not
clear, the card is locked automatically on the next power-up.
An attempt to unlock an unlocked card fails and the LOCK_UNLOCK_FAILED error bit is set
in the card status register.
Forcing erase
If the user has forgotten the password (PWD content), it is possible to access the card after
clearing all the data on the card. This forced erase operation erases all card data and all
password data.
2. Set the block length (SET_BLOCKLEN, CMD16) to 1 byte. Only the 8-bit card
lock/unlock byte (byte 0 in Table 171) is sent.
3. Send LOCK/UNLOCK (CMD42) with the appropriate data byte on the data line including
the 16-bit CRC. The data block indicates the mode (ERASE = 1). All other bits must be
zero.
4. When the ERASE bit is the only bit set in the data field, all card contents are erased,
including the PWD and PWD_LEN fields, and the card is no longer locked. When any
other bits are set, the LOCK_UNLOCK_FAILED error bit is set in the card status
register and the card retains all of its data, and remains locked.
An attempt to use a force erase on an unlocked card fails and the LOCK_UNLOCK_FAILED
error bit is set in the card status register.
31.4.11 Card status register

The response format R1 contains a 32-bit field named card status. This field is intended to
transmit the card status information (which may be stored in a local status register) to the
host. If not specified otherwise, the status entries are always related to the previously issued
command.
Table 158 defines the different entries of the status. The type and clear condition fields in
the table are abbreviated as follows:
1042/1751 RM0090 Rev 19

Type:
• E: error bit
• S: status bit
• R: detected and set for the actual command response
• X: detected and set during command execution. The SDIO card host must poll the card
by issuing the status command to read these bits.
Clear condition:
• A: according to the card current state
• B: always related to the previous command. Reception of a valid command clears it
(with a delay of one command)
• C: clear by read
Table 158. Card status

Clear
Bits Identifier Type Value Description
condition
The command address argument was out

of the allowed range for this card.
ADDRESS_ ’0’= no error A multiple block or stream read/write
31 ERX C
OUT_OF_RANGE ’1’= error operation is (although started in a valid
address) attempting to read or write
beyond the card capacity.
The commands address argument (in
accordance with the currently set block
length) positions the first data block
misaligned to the card physical blocks.
’0’= no error A multiple block read/write operation
30 ADDRESS_MISALIGN - C
’1’= error (although started with a valid
address/block-length combination) is
attempting to read or write a data block
which is not aligned with the physical
blocks of the card.
Either the argument of a
SET_BLOCKLEN command exceeds the
maximum value allowed for the card, or
’0’= no error the previously defined block length is
29 BLOCK_LEN_ERROR - illegal for the current command (e.g. the C
’1’= error
host issues a write command, the current
block length is smaller than the maximum
allowed value for the card and it is not
allowed to write partial blocks)
’0’= no error An error in the sequence of erase
28 ERASE_SEQ_ERROR - C
’1’= error commands occurred.
’0’= no error An invalid selection of erase groups for
27 ERASE_PARAM EX C
’1’= error erase occurred.
’0’= no error Attempt to program a write-protected

26 WP_VIOLATION EX
’1’= error block. C
RM0090 Rev 19 1043/1751

1075
Table 158. Card status (continued)

Clear
condition
‘0’ = card
When set, signals that the card is locked
25 CARD_IS_LOCKED SR unlocked A
by the host
‘1’ = card locked
Set when a sequence or password error
LOCK_UNLOCK_ ’0’= no error
24 EX has been detected in lock/unlock card C
FAILED ’1’= error
command
’0’= no error The CRC check of the previous command
23 COM_CRC_ERROR ER B
’1’= error failed.
’0’= no error
22 ILLEGAL_COMMAND ER Command not legal for the card state B
’1’= error
’0’= success Card internal ECC was applied but failed
21 CARD_ECC_FAILED EX C
’1’= failure to correct the data.
(Undefined by the standard) A card error
’0’= no error
20 CC_ERROR ER occurred, which is not related to the host C
’1’= error
command.
(Undefined by the standard) A generic
’0’= no error card error related to the (and detected
19 ERROR EX C
’1’= error during) execution of the last host
command (e.g. read or write failures).
18 Reserved
17 Reserved
Can be either of the following errors:
– The CID register has already been
written and cannot be overwritten
’0’= no error ‘1’= – The read-only section of the CSD does C
16 CID/CSD_OVERWRITE EX
error not match the card contents
– An attempt to reverse the copy (set as
original) or permanent WP
(unprotected) bits was made
’0’= not protected Set when only partial address space

15 WP_ERASE_SKIP EX C
’1’= protected was erased due to existing write
’0’= enabled The command has been executed without

14 CARD_ECC_DISABLED S X A
’1’= disabled using the internal ECC.
An erase sequence was cleared before
executing because an out of erase
’0’= cleared
13 ERASE_RESET - sequence command was received C
’1’= set
(commands other than CMD35, CMD36,
CMD38 or CMD13)
1044/1751 RM0090 Rev 19

Table 158. Card status (continued)

Clear
condition
0 = Idle
1 = Ready
2 = Ident
The state of the card when receiving the
3 = Stby
command. If the command execution
4 = Tran
causes a state change, it will be visible to
12:9 CURRENT_STATE SR 5 = Data B
the host in the response on the next
6 = Rcv
command. The four bits are interpreted as
7 = Prg
a binary number between 0 and 15.
8 = Dis
9 = Btst
10-15 = reserved
’0’= not ready ‘1’ Corresponds to buffer empty signalling on
8 READY_FOR_DATA SR -
= ready the bus
If set, the card did not switch to the
’0’= no error
7 SWITCH_ERROR EX expected mode as requested by the B
’1’= switch error
SWITCH command
6 Reserved
The card will expect ACMD, or an
‘0’ = Disabled
5 APP_CMD SR indication that the command has been C
‘1’ = Enabled
interpreted as ACMD
4 Reserved for SD I/O Card
’0’= no error Error in the sequence of the
3 AKE_SEQ_ERROR ER C
’1’= error authentication process
2 Reserved for application specific commands
1
Reserved for manufacturer test mode
0
31.4.12 SD status register

The SD status contains status bits that are related to the SD memory card proprietary
features and may be used for future application-specific usage. The size of the SD Status is
one data block of 512 bits. The contents of this register are transmitted to the SDIO card
host if ACMD13 is sent (CMD55 followed with CMD13). ACMD13 can be sent to a card in
transfer state only (card is selected).
Table 159 defines the different entries of the SD status register. The type and clear condition
fields in the table are abbreviated as follows:
Type:
• E: error bit
• S: status bit
• R: detected and set for the actual command response
• X: detected and set during command execution. The SDIO card Host must poll the card
by issuing the status command to read these bits
RM0090 Rev 19 1045/1751

1075
Clear condition:
• A: according to the card current state
• B: always related to the previous command. Reception of a valid command clears it
(with a delay of one command)
• C: clear by read
Table 159. SD status

Clear
condition
Shows the currently defined

’00’= 1 (default)
databus width that was
‘01’= reserved
511: 510 DAT_BUS_WIDTH S R defined by A
‘10’= 4 bit width
SET_BUS_WIDTH
‘11’= reserved
command
Card is in Secured Mode of
’0’= Not in the mode
509 SECURED_MODE S R operation (refer to the “SD A
’1’= In Secured Mode
Security Specification”).
508: 496 Reserved
In the future, the 8 LSBs will
’00xxh’= SD Memory Cards as be used to define different
defined in Physical Spec Ver1.01- variations of an SD memory
2.00 (’x’= don’t care). The card (each bit will define
495: 480 SD_CARD_TYPE SR following cards are currently different SD types). The 8 A
defined: MSBs will be used to define
’0000’= Regular SD RD/WR Card. SD Cards that do not comply
’0001’= SD ROM Card with current SD physical
layer specification.
SIZE_OF_PROTE Size of protected area (See
479: 448 SR (See below) A
CT ED_AREA below)
Speed Class of the card (See
447: 440 SPEED_CLASS SR (See below) A
below)
Performance of move indicated by
PERFORMANCE_
439: 432 SR 1 [MB/s] step. (See below) A
MOVE
(See below)
Size of AU
431:428 AU_SIZE SR (See below) A
(See below)
427:424 Reserved
Number of AUs to be erased at a
423:408 ERASE_SIZE SR (See below) A
time
Timeout value for erasing areas
407:402 ERASE_TIMEOUT S R specified by (See below) A
UNIT_OF_ERASE_AU
Fixed offset value added to erase
401:400 ERASE_OFFSET SR (See below) A
time.
399:312 Reserved
311:0 Reserved for Manufacturer
1046/1751 RM0090 Rev 19

SIZE_OF_PROTECTED_AREA
Setting this field differs between standard- and high-capacity cards. In the case of a
standard-capacity card, the capacity of protected area is calculated as follows:
Protected area = SIZE_OF_PROTECTED_AREA_* MULT * BLOCK_LEN.
SIZE_OF_PROTECTED_AREA is specified by the unit in MULT*BLOCK_LEN.
In the case of a high-capacity card, the capacity of protected area is specified in this field:
Protected area = SIZE_OF_PROTECTED_AREA
SIZE_OF_PROTECTED_AREA is specified by the unit in bytes.
SPEED_CLASS
This 8-bit field indicates the speed class and the value can be calculated by PW/2 (where
PW is the write performance).
Table 160. Speed class code field

SPEED_CLASS Value definition
00h Class 0
01h Class 2
02h Class 4
03h Class 6
04h – FFh Reserved
PERFORMANCE_MOVE
This 8-bit field indicates Pm (performance move) and the value can be set by 1 [MB/sec]
steps. If the card does not move used RUs (recording units), Pm should be considered as
infinity. Setting the field to FFh means infinity.
Table 161. Performance move field

PERFORMANCE_MOVE Value definition
00h Not defined

01h 1 [MB/sec]
02h 02h 2 [MB/sec]
--------- ---------
FEh 254 [MB/sec]
FFh Infinity
RM0090 Rev 19 1047/1751

1075
AU_SIZE
This 4-bit field indicates the AU size and the value can be selected in the power of 2 base
from 16 KB.
Table 162. AU_SIZE field

AU_SIZE Value definition
00h Not defined

01h 16 KB
02h 32 KB
03h 64 KB
04h 128 KB
05h 256 KB
06h 512 KB
07h 1 MB
08h 2 MB
09h 4 MB
Ah – Fh Reserved
The maximum AU size, which depends on the card capacity, is defined in Table 163. The
card can be set to any AU size between RU size and maximum AU size.
Table 163. Maximum AU size

Capacity 16 MB-64 MB 128 MB-256 MB 512 MB 1 GB-32 GB
Maximum AU Size 512 KB 1 MB 2 MB 4 MB
ERASE_SIZE
This 16-bit field indicates NERASE. When NERASE numbers of AUs are erased, the
timeout value is specified by ERASE_TIMEOUT (Refer to ERASE_TIMEOUT). The host
should determine the proper number of AUs to be erased in one operation so that the host
can show the progress of the erase operation. If this field is set to 0, the erase timeout
calculation is not supported.
Table 164. Erase size field

ERASE_SIZE Value definition
0000h Erase timeout calculation is not supported.

0001h 1 AU
0002h 2 AU
0003h 3 AU
--------- ---------
FFFFh 65535 AU
1048/1751 RM0090 Rev 19

ERASE_TIMEOUT
This 6-bit field indicates TERASE and the value indicates the erase timeout from offset
when multiple AUs are being erased as specified by ERASE_SIZE. The range of
ERASE_TIMEOUT can be defined as up to 63 seconds and the card manufacturer can
choose any combination of ERASE_SIZE and ERASE_TIMEOUT depending on the
implementation. Determining ERASE_TIMEOUT determines the ERASE_SIZE.
Table 165. Erase timeout field

ERASE_TIMEOUT Value definition
00 Erase timeout calculation is not supported.

01 1 [sec]
02 2 [sec]
03 3 [sec]
--------- ---------
63 63 [sec]
ERASE_OFFSET
This 2-bit field indicates TOFFSET and one of four values can be selected. This field is
meaningless if the ERASE_SIZE and ERASE_TIMEOUT fields are set to 0.
Table 166. Erase offset field

ERASE_OFFSET Value definition
0h 0 [sec]
1h 1 [sec]
2h 2 [sec]
3h 3 [sec]
31.4.13 SD I/O mode

SD I/O interrupts
To allow the SD I/O card to interrupt the MultiMediaCard/SD module, an interrupt function is
available on a pin on the SD interface. Pin 8, used as SDIO_D1 when operating in the 4-bit
SD mode, signals the cards interrupt to the MultiMediaCard/SD module. The use of the
interrupt is optional for each card or function within a card. The SD I/O interrupt is level-
sensitive, which means that the interrupt line must be held active (low) until it is either
recognized and acted upon by the MultiMediaCard/SD module or deasserted due to the end
of the interrupt period. After the MultiMediaCard/SD module has serviced the interrupt, the
interrupt status bit is cleared via an I/O write to the appropriate bit in the SD I/O card’s
internal registers. The interrupt output of all SD I/O cards is active low and the application
must provide external pull-up resistors on all data lines (SDIO_D[3:0]). The
MultiMediaCard/SD module samples the level of pin 8 (SDIO_D/IRQ) into the interrupt
detector only during the interrupt period. At all other times, the MultiMediaCard/SD module
ignores this value.
RM0090 Rev 19 1049/1751

1075
The interrupt period is applicable for both memory and I/O operations. The definition of the
interrupt period for operations with single blocks is different from the definition for multiple-
block data transfers.
SD I/O suspend and resume

Within a multifunction SD I/O or a card with both I/O and memory functions, there are
multiple devices (I/O and memory) that share access to the MMC/SD bus. To share access
to the MMC/SD module among multiple devices, SD I/O and combo cards optionally
implement the concept of suspend/resume. When a card supports suspend/resume, the
MMC/SD module can temporarily halt a data transfer operation to one function or memory
(suspend) to free the bus for a higher-priority transfer to a different function or memory. After
this higher-priority transfer is complete, the original transfer is resumed (restarted) where it
left off. Support of suspend/resume is optional on a per-card basis. To perform the
suspend/resume operation on the MMC/SD bus, the MMC/SD module performs the
following steps:
1. Determines the function currently using the SDIO_D [3:0] line(s)
2. Requests the lower-priority or slower transaction to suspend
3. Waits for the transaction suspension to complete
4. Begins the higher-priority transaction
5. Waits for the completion of the higher priority transaction
6. Restores the suspended transaction
SD I/O ReadWait
The optional ReadWait (RW) operation is defined only for the SD 1-bit and 4-bit modes. The
ReadWait operation allows the MMC/SD module to signal a card that it is reading multiple
registers (IO_RW_EXTENDED, CMD53) to temporarily stall the data transfer while allowing
the MMC/SD module to send commands to any function within the SD I/O device. To
determine when a card supports the ReadWait protocol, the MMC/SD module must test
capability bits in the internal card registers. The timing for ReadWait is based on the
interrupt period.
31.4.14 Commands and responses

Application-specific and general commands
The SD card host module system is designed to provide a standard interface for a variety of
applications types. In this environment, there is a need for specific customer/application
features. To implement these features, two types of generic commands are defined in the
standard: application-specific commands (ACMD) and general commands (GEN_CMD).
When the card receives the APP_CMD (CMD55) command, the card expects the next
command to be an application-specific command. ACMDs have the same structure as
regular MultiMediaCard commands and can have the same CMD number. The card
recognizes it as ACMD because it appears after APP_CMD (CMD55). When the command
immediately following the APP_CMD (CMD55) is not a defined application-specific
command, the standard command is used. For example, when the card has a definition for
SD_STATUS (ACMD13), and receives CMD13 immediately following APP_CMD (CMD55),
this is interpreted as SD_STATUS (ACMD13). However, when the card receives CMD7
immediately following APP_CMD (CMD55) and the card does not have a definition for
ACMD7, this is interpreted as the standard (SELECT/DESELECT_CARD) CMD7.
1050/1751 RM0090 Rev 19

To use one of the manufacturer-specific ACMDs the SD card Host must perform the
following steps:
1. Send APP_CMD (CMD55)
The card responds to the MultiMediaCard/SD module, indicating that the APP_CMD bit
is set and an ACMD is now expected.
2. Send the required ACMD
The card responds to the MultiMediaCard/SD module, indicating that the APP_CMD bit
is set and that the accepted command is interpreted as an ACMD. When a nonACMD
is sent, it is handled by the card as a normal MultiMediaCard command and the
APP_CMD bit in the card status register stays clear.
When an invalid command is sent (neither ACMD nor CMD) it is handled as a standard
MultiMediaCard illegal command error.
The bus transaction for a GEN_CMD is the same as the single-block read or write
commands (WRITE_BLOCK, CMD24 or READ_SINGLE_BLOCK,CMD17). In this case, the
argument denotes the direction of the data transfer rather than the address, and the data
block has vendor-specific format and meaning.
The card must be selected (in transfer state) before sending GEN_CMD (CMD56). The data
block size is defined by SET_BLOCKLEN (CMD16). The response to GEN_CMD (CMD56)
is in R1b format.
Command types
Both application-specific and general commands are divided into the four following types:
• broadcast command (BC): sent to all cards; no responses returned.
• broadcast command with response (BCR): sent to all cards; responses received
from all cards simultaneously.
• addressed (point-to-point) command (AC): sent to the card that is selected; does
not include a data transfer on the SDIO_D line(s).
• addressed (point-to-point) data transfer command (ADTC): sent to the card that is
selected; includes a data transfer on the SDIO_D line(s).
Command formats
See Table 151 on page 1027 for command formats.
Commands for the MultiMediaCard/SD module
Table 167. Block-oriented write commands

CMD Response
Type Argument Abbreviation Description
index format
Defines the number of blocks which

[31:16] set to 0
are going to be transferred in the
CMD23 ac [15:0] number R1 SET_BLOCK_COUNT
multiple-block read or write command
of blocks
that follows.
[31:0] data Writes a block of the size selected by
CMD24 adtc R1 WRITE_BLOCK
address the SET_BLOCKLEN command.
RM0090 Rev 19 1051/1751

1075
Table 167. Block-oriented write commands (continued)

CMD Response
index format
Continuously writes blocks of data

[31:0] data until a STOP_TRANSMISSION
CMD25 adtc R1 WRITE_MULTIPLE_BLOCK
address follows or the requested number of
blocks has been received.
Programming of the card identification
register. This command must be
issued only once per card. The card
CMD26 adtc [31:0] stuff bits R1 PROGRAM_CID contains hardware to prevent this
operation after the first programming.
Normally this command is reserved
for manufacturer.
Programming of the programmable
CMD27 adtc [31:0] stuff bits R1 PROGRAM_CSD
bits of the CSD.
Table 168. Block-oriented write protection commands

CMD Response
index format
If the card has write protection features,

this command sets the write protection bit
[31:0] data
CMD28 ac R1b SET_WRITE_PROT of the addressed group. The properties of
address
write protection are coded in the card-
specific data (WP_GRP_SIZE).
If the card provides write protection
[31:0] data
CMD29 ac R1b CLR_WRITE_PROT features, this command clears the write
address
protection bit of the addressed group.
If the card provides write protection
[31:0] write
features, this command asks the card to
CMD30 adtc protect data R1 SEND_WRITE_PROT
send the status of the write protection
address
bits.
CMD31 Reserved
Table 169. Erase commands

CMD Response
index format
CMD32
Reserved. These command indexes cannot be used in order to maintain backward compatibility with older
...
versions of the MultiMediaCard.
CMD34
Sets the address of the first erase
CMD35 ac [31:0] data address R1 ERASE_GROUP_START group within a range to be selected
for erase.
Sets the address of the last erase
CMD36 ac [31:0] data address R1 ERASE_GROUP_END group within a continuous range to be
selected for erase.
1052/1751 RM0090 Rev 19

Table 169. Erase commands (continued)

CMD Response
index format
Reserved. This command index cannot be used in order to maintain backward compatibility with older
CMD37
versions of the MultiMediaCards
Erases all previously selected write
CMD38 ac [31:0] stuff bits R1 ERASE
blocks.
Table 170. I/O mode commands

CMD Response
index format
Used to write and read 8-bit (register) data

[31:16] RCA fields. The command addresses a card and a
[15:15] register register and provides the data for writing if
write flag the write flag is set. The R4 response
CMD39 ac R4 FAST_IO
[14:8] register contains data read from the addressed
address register. This command accesses
[7:0] register data application-dependent registers that are not
defined in the MultiMediaCard standard.
CMD40 bcr [31:0] stuff bits R5 GO_IRQ_STATE Places the system in the interrupt mode.
CMD41 Reserved
Table 171. Lock card

CMD Response
index format
Sets/resets the password or locks/unlocks

CMD42 adtc [31:0] stuff bits R1b LOCK_UNLOCK the card. The size of the data block is set
by the SET_BLOCK_LEN command.
CMD43
... Reserved
CMD54
Table 172. Application-specific commands

CMD Response
index format
Indicates to the card that the next command

[31:16] RCA
CMD55 ac R1 APP_CMD bits is an application specific command rather
[15:0] stuff bits than a standard command
Used either to transfer a data block to the card
or to get a data block from the card for general
[31:1] stuff bits
CMD56 adtc purpose/application-specific commands. The
[0]: RD/WR size of the data block shall be set by the
SET_BLOCK_LEN command.
RM0090 Rev 19 1053/1751

1075
Table 172. Application-specific commands (continued)

CMD Response
index format
CMD57
... Reserved.
CMD59
CMD60
... Reserved for manufacturer.
CMD63
31.5 Response formats

All responses are sent via the MCCMD command line SDIO_CMD. The response
transmission always starts with the left bit of the bit string corresponding to the response
code word. The code length depends on the response type.
A response always starts with a start bit (always 0), followed by the bit indicating the
direction of transmission (card = 0). A value denoted by x in the tables below indicates a
variable entry. All responses, except for the R3 response type, are protected by a CRC.
Every command code word is terminated by the end bit (always 1).
There are five types of responses. Their formats are defined as follows:
31.5.1 R1 (normal response command)

Code length = 48 bits. The 45:40 bits indicate the index of the command to be responded to,
this value being interpreted as a binary-coded number (between 0 and 63). The status of the
card is coded in 32 bits.
Table 173. R1 response

Bit position Width (bits Value Description
47 1 0 Start bit
[45:40] 6 X Command index
[39:8] 32 X Card status
[7:1] 7 X CRC7
0 1 1 End bit
31.5.2 R1b
It is identical to R1 with an optional busy signal transmitted on the data line. The card may
become busy after receiving these commands based on its state prior to the command
reception.
31.5.3 R2 (CID, CSD register)

Code length = 136 bits. The contents of the CID register are sent as a response to the
CMD2 and CMD10 commands. The contents of the CSD register are sent as a response to
1054/1751 RM0090 Rev 19

CMD9. Only the bits [127...1] of the CID and CSD are transferred, the reserved bit [0] of
these registers is replaced by the end bit of the response. The card indicates that an erase
is in progress by holding MCDAT low. The actual erase time may be quite long, and the host
may issue CMD7 to deselect the card.

135 1 0 Start bit

[133:128] 6 ‘111111’ Command index
[127:1] 127 X Card status
0 1 1 End bit
31.5.4 R3 (OCR register)

Code length: 48 bits. The contents of the OCR register are sent as a response to CMD1.
The level coding is as follows: restricted voltage windows = low, card busy = low.

47 1 0 Start bit
[45:40] 6 ‘111111’ Reserved
[39:8] 32 X OCR register
[7:1] 7 ‘1111111’ Reserved
0 1 1 End bit
31.5.5 R4 (Fast I/O)

Code length: 48 bits. The argument field contains the RCA of the addressed card, the
register address to be read from or written to, and its content.

47 1 0 Start bit
[45:40] 6 ‘100111’ CMD39
[31:16] 16 X RCA
[39:8] Argument field [15:8] 8 X register address
[7:0] 8 X read register contents
RM0090 Rev 19 1055/1751

1075
Table 176. R4 response (continued)

[7:1] 7 X CRC7
0 1 1 End bit
31.5.6 R4b
For SD I/O only: an SDIO card receiving the CMD5 will respond with a unique SDIO
response R4. The format is:
Table 177. R4b response

47 1 0 Start bit
[45:40] 6 x Reserved
39 16 X Card is ready
[38:36] 3 X Number of I/O functions
[39:8] Argument field 35 1 X Present memory
[34:32] 3 X Stuff bits
[31:8] 24 X I/O ORC
[7:1] 7 X Reserved
0 1 1 End bit
Once an SD I/O card has received a CMD5, the I/O portion of that card is enabled to
respond normally to all further commands. This I/O enable of the function within the I/O card
will remain set until a reset, power cycle or CMD52 with write to I/O reset is received by the
card. Note that an SD memory-only card may respond to a CMD5. The proper response for
a memory-only card would be Present memory = 1 and Number of I/O functions = 0. A
memory-only card built to meet the SD Memory Card specification version 1.0 would detect
the CMD5 as an illegal command and not respond. The I/O aware host will send CMD5. If
the card responds with response R4, the host determines the card’s configuration based on
the data contained within the R4 response.
31.5.7 R5 (interrupt request)

Only for MultiMediaCard. Code length: 48 bits. If the response is generated by the host, the
RCA field in the argument will be 0x0.

47 1 0 Start bit
[45:40] 6 ‘101000’ CMD40
1056/1751 RM0090 Rev 19

Table 178. R5 response (continued)

RCA [31:16] of winning

[31:16] 16 X
card or of the host
[39:8] Argument field
Not defined. May be used
[15:0] 16 X
for IRQ data
[7:1] 7 X CRC7
0 1 1 End bit
31.5.8 R6
Only for SD I/O. The normal response to CMD3 by a memory device. It is shown in
Table 179.

Bit position Width (bits) Value Description
47 1 0 Start bit
[45:40] 6 ‘101000’ CMD40
[39:8] Argument [31:16] 16 X RCA [31:16] of winning card or of the host

field [15:0] 16 X Not defined. May be used for IRQ data
[7:1] 7 X CRC7
0 1 1 End bit
The card [23:8] status bits are changed when CMD3 is sent to an I/O-only card. In this case,
the 16 bits of response are the SD I/O-only values:
• Bit [15] COM_CRC_ERROR
• Bit [14] ILLEGAL_COMMAND
• Bit [13] ERROR
• Bits [12:0] Reserved
31.6 SDIO I/O card-specific operations

The following features are SD I/O-specific operations:
• SDIO read wait operation by SDIO_D2 signalling
• SDIO read wait operation by stopping the clock
• SDIO suspend/resume operation (write and read suspend)
• SDIO interrupts
The SDIO supports these operations only if the SDIO_DCTRL[11] bit is set, except for read
suspend that does not need specific hardware implementation.
RM0090 Rev 19 1057/1751

1075
31.6.1 SDIO I/O read wait operation by SDIO_D2 signaling

It is possible to start the readwait interval before the first block is received: when the data
path is enabled (SDIO_DCTRL[0] bit set), the SDIO-specific operation is enabled
(SDIO_DCTRL[11] bit set), read wait starts (SDI0_DCTRL[10] =0 and SDI_DCTRL[8] =1)
and data direction is from card to SDIO (SDIO_DCTRL[1] = 1), the DPSM directly moves
from Idle to Readwait. In Readwait the DPSM drives SDIO_D2 to 0 after 2 SDIO_CK clock
cycles. In this state, when you set the RWSTOP bit (SDIO_DCTRL[9]), the DPSM remains
in Wait for two more SDIO_CK clock cycles to drive SDIO_D2 to 1 for one clock cycle (in
accordance with SDIO specification). The DPSM then starts waiting again until it receives
data from the card. The DPSM will not start a readwait interval while receiving a block even
if read wait start is set: the readwait interval will start after the CRC is received. The
RWSTOP bit has to be cleared to start a new read wait operation. During the readwait
interval, the SDIO can detect SDIO interrupts on SDIO_D1.
31.6.2 SDIO read wait operation by stopping SDIO_CK

If the SDIO card does not support the previous read wait method, the SDIO can perform a
read wait by stopping SDIO_CK (SDIO_DCTRL is set just like in the method presented in
Section 31.6.1, but SDIO_DCTRL[10] =1): DSPM stops the clock two SDIO_CK cycles after
the end bit of the current received block and starts the clock again after the read wait start bit
is set.
As SDIO_CK is stopped, any command can be issued to the card. During a read/wait
interval, the SDIO can detect SDIO interrupts on SDIO_D1.
31.6.3 SDIO suspend/resume operation

While sending data to the card, the SDIO can suspend the write operation. the
SDIO_CMD[11] bit is set and indicates to the CPSM that the current command is a suspend
command. The CPSM analyzes the response and when the ACK is received from the card
(suspend accepted), it acknowledges the DPSM that goes Idle after receiving the CRC
token of the current block.
The hardware does not save the number of the remaining block to be sent to complete the
suspended operation (resume).
The write operation can be suspended by software, just by disabling the DPSM
(SDIO_DCTRL[0] =0) when the ACK of the suspend command is received from the card.
The DPSM enters then the Idle state.
To suspend a read: the DPSM waits in the Wait_r state as the function to be suspended
sends a complete packet just before stopping the data transaction. The application
continues reading RxFIFO until the FIF0 is empty, and the DPSM goes Idle automatically.
31.6.4 SDIO interrupts

SDIO interrupts are detected on the SDIO_D1 line once the SDIO_DCTRL[11] bit is set.
1058/1751 RM0090 Rev 19

31.7 CE-ATA specific operations

The following features are CE-ATA specific operations:
• sending the command completion signal disable to the CE-ATA device
• receiving the command completion signal from the CE-ATA device
• signaling the completion of the CE-ATA command to the CPU, using the status bit
and/or interrupt.
The SDIO supports these operations only for the CE-ATA CMD61 command, that is, if
SDIO_CMD[14] is set.
31.7.1 Command completion signal disable

Command completion signal disable is sent 8 bit cycles after the reception of a short
response if the ‘enable CMD completion’ bit, SDIO_CMD[12], is not set and the ‘not interrupt
Enable’ bit, SDIO_CMD[13], is set.
The CPSM enters the Pend state, loading the command shift register with the disable
sequence “00001” and, the command counter with 43. Eight cycles after, a trigger moves
the CPSM to the Send state. When the command counter reaches 48, the CPSM becomes
Idle as no response is awaited.
31.7.2 Command completion signal enable

If the ‘enable CMD completion’ bit SDIO_CMD[12] is set and the ‘not interrupt Enable’ bit
SDIO_CMD[13] is set, the CPSM waits for the command completion signal in the Waitcpl
state.
When ‘0’ is received on the CMD line, the CPSM enters the Idle state. No new command
can be sent for 7 bit cycles. Then, for the last 5 cycles (out of the 7) the CMD line is driven to
‘1’ in push-pull mode.
31.7.3 CE-ATA interrupt

The command completion is signaled to the CPU by the status bit SDIO_STA[23]. This static
bit can be cleared with the clear bit SDIO_ICR[23].
The SDIO_STA[23] status bit can generate an interrupt on each interrupt line, depending on
the mask bit SDIO_MASKx[23].
31.7.4 Aborting CMD61

If the command completion disable signal has not been sent and CMD61 needs to be
aborted, the command state machine must be disabled. It then becomes Idle, and the
CMD12 command can be sent. No command completion disable signal is sent during the
operation.
RM0090 Rev 19 1059/1751

1075
31.8 HW flow control

The HW flow control functionality is used to avoid FIFO underrun (TX mode) and overrun
(RX mode) errors.
The behavior is to stop SDIO_CK and freeze SDIO state machines. The data transfer is
stalled while the FIFO is unable to transmit or receive data. Only state machines clocked by
SDIOCLK are frozen, the APB2 interface is still alive. The FIFO can thus be filled or emptied
even if flow control is activated.
To enable HW flow control, the SDIO_CLKCR[14] register bit must be set to 1. After reset
Flow Control is disabled.
31.9 SDIO registers

The device communicates to the system via 32-bit-wide control registers accessible via
APB2.
31.9.1 SDIO power control register (SDIO_POWER)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PWRC
Reserved TRL
rw rw

Bits 1:0 PWRCTRL: Power supply control bits.
These bits are used to define the current functional state of the card clock:
00: Power-off: the clock to card is stopped.
01: Reserved
10: Reserved power-up
11: Power-on: the card is clocked.
Note: At least seven HCLK clock periods are needed between two write accesses to this register.
After a data write, data cannot be written to this register for three SDIOCLK clock periods
plus two PCLK2 clock periods.
1060/1751 RM0090 Rev 19

31.9.2 SDI clock control register (SDIO_CLKCR)

The SDIO_CLKCR register controls the SDIO_CK output clock.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
NEGEDGE
HWFC_EN
PWRSAV
BYPASS
CLKEN
WID
CLKDIV
Reserved BUS

Bit 14 HWFC_EN: HW Flow Control enable
0b: HW Flow Control is disabled
1b: HW Flow Control is enabled
When HW Flow Control is enabled, the meaning of the TXFIFOE and RXFIFOF interrupt
signals, see SDIO Status register definition in Section 31.9.11.
Bit 13 NEGEDGE:SDIO_CK dephasing selection bit
0b: SDIO_CK generated on the rising edge of the master clock SDIOCLK
1b: SDIO_CK generated on the falling edge of the master clock SDIOCLK
Bits 12:11 WIDBUS: Wide bus mode enable bit
00: Default bus mode: SDIO_D0 used
01: 4-wide bus mode: SDIO_D[3:0] used
10: 8-wide bus mode: SDIO_D[7:0] used
Bit 10 BYPASS: Clock divider bypass enable bit
0: Disable bypass: SDIOCLK is divided according to the CLKDIV value before driving the
SDIO_CK output signal.
1: Enable bypass: SDIOCLK directly drives the SDIO_CK output signal.
Bit 9 PWRSAV: Power saving configuration bit
For power saving, the SDIO_CK clock output can be disabled when the bus is idle by setting
PWRSAV:
0: SDIO_CK clock is always enabled
1: SDIO_CK is only enabled when the bus is active
Bit 8 CLKEN: Clock enable bit
0: SDIO_CK is disabled
1: SDIO_CK is enabled
Bits 7:0 CLKDIV: Clock divide factor
This field defines the divide factor between the input clock (SDIOCLK) and the output clock
(SDIO_CK): SDIO_CK frequency = SDIOCLK / [CLKDIV + 2].
Note: In order to have a duty cycle of 50% it is recommended to select even values of
CLKDIV.
RM0090 Rev 19 1061/1751

1075
Note: While the SD/SDIO card or MultiMediaCard is in identification mode, the SDIO_CK
frequency must be less than 400 kHz.
The clock frequency can be changed to the maximum card bus frequency when relative
card addresses are assigned to all cards.
After a data write, data cannot be written to this register for three SDIOCLK clock periods
plus two PCLK2 clock periods. SDIO_CK can also be stopped during the read wait interval
for SD I/O cards: in this case the SDIO_CLKCR register does not control SDIO_CK.
31.9.3 SDIO argument register (SDIO_ARG)

The SDIO_ARG register contains a 32-bit command argument, which is sent to a card as
part of a command message.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CMDARG
Bits 31:0 CMDARG: Command argument

Command argument sent to a card as part of a command message. If a command contains
an argument, it must be loaded into this register before writing a command to the command
register.
31.9.4 SDIO command register (SDIO_CMD)

The SDIO_CMD register contains the command index and command type bits. The
command index is sent to a card as part of a command message. The command type bits
control the command path state machine (CPSM).
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SDIOSuspend
ENCMDcompl
CE-ATACMD
CMDINDEX
WAITPEND
WAITRESP
CPSMEN
WAITINT
nIEN
Reserved

Bit 14 ATACMD: CE-ATA command
If ATACMD is set, the CPSM transfers CMD61.
Bit 13 nIEN: not Interrupt Enable
if this bit is 0, interrupts in the CE-ATA device are enabled.
Bit 12 ENCMDcompl: Enable CMD completion
If this bit is set, the command completion signal is enabled.
1062/1751 RM0090 Rev 19

Bit 11 SDIOSuspend: SD I/O suspend command

If this bit is set, the command to be sent is a suspend command (to be used only with SDIO
card).
Bit 10 CPSMEN: Command path state machine (CPSM) Enable bit
If this bit is set, the CPSM is enabled.
Bit 9 WAITPEND: CPSM Waits for ends of data transfer (CmdPend internal signal).
If this bit is set, the CPSM waits for the end of data transfer before it starts sending a
command.
Bit 8 WAITINT: CPSM waits for interrupt request
If this bit is set, the CPSM disables command timeout and waits for an interrupt request.
Bits 7:6 WAITRESP: Wait for response bits
They are used to configure whether the CPSM is to wait for a response, and if yes, which
kind of response.
00: No response, expect CMDSENT flag
01: Short response, expect CMDREND or CCRCFAIL flag
10: No response, expect CMDSENT flag
11: Long response, expect CMDREND or CCRCFAIL flag
Bits 5:0 CMDINDEX: Command index
The command index is sent to the card as part of a command message.
Note: After a data write, data cannot be written to this register for three SDIOCLK clock periods
MultiMediaCards can send two kinds of response: short responses, 48 bits long, or long
responses,136 bits long. SD card and SD I/O card can send only short responses, the
argument can vary according to the type of response: the software will distinguish the type
of response according to the sent command. CE-ATA devices send only short responses.
31.9.5 SDIO command response register (SDIO_RESPCMD)

The SDIO_RESPCMD register contains the command index field of the last command
response received. If the command response transmission does not contain the command
index field (long or OCR response), the RESPCMD field is unknown, although it must
contain 111111b (the value of the reserved field from the response).
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RESPCMD
Reserved
r r r r r r

Bits 5:0 RESPCMD: Response command index
Read-only bit field. Contains the command index of the last command response received.
RM0090 Rev 19 1063/1751

1075
31.9.6 SDIO response 1..4 register (SDIO_RESPx)

Address offset: (0x10 + (4 × x)); x = 1..4
The SDIO_RESP1/2/3/4 registers contain the status of a card, which is part of the received
response.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CARDSTATUSx
Bits 31:0 CARDSTATUSx: see Table 180.
The Card Status size is 32 or 127 bits, depending on the response type.
Table 180. Response type and SDIO_RESPx registers

Register Short response Long response
SDIO_RESP1 Card Status[31:0] Card Status [127:96]

SDIO_RESP2 Unused Card Status [95:64]
SDIO_RESP3 Unused Card Status [63:32]
SDIO_RESP4 Unused Card Status [31:1]0b
The most significant bit of the card status is received first. The SDIO_RESP3 register LSB is
always 0b.
31.9.7 SDIO data timer register (SDIO_DTIMER)

The SDIO_DTIMER register contains the data timeout period, in card bus clock periods.
A counter loads the value from the SDIO_DTIMER register, and starts decrementing when
the data path state machine (DPSM) enters the Wait_R or Busy state. If the timer reaches 0
while the DPSM is in either of these states, the timeout status flag is set.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DATATIME
Bits 31:0 DATATIME: Data timeout period

Data timeout period expressed in card bus clock periods.
Note: A data transfer must be written to the data timer register and the data length register before
being written to the data control register.
1064/1751 RM0090 Rev 19

31.9.8 SDIO data length register (SDIO_DLEN)

The SDIO_DLEN register contains the number of data bytes to be transferred. The value is
loaded into the data counter when data transfer starts.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DATALENGTH
Reserved
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 24:0 DATALENGTH: Data length value
Number of data bytes to be transferred.
Note: For a block data transfer, the value in the data length register must be a multiple of the block
size (see SDIO_DCTRL). A data transfer must be written to the data timer register and the
data length register before being written to the data control register.
For an SDIO multibyte transfer the value in the data length register must be between 1 and
512.
RM0090 Rev 19 1065/1751

1075
31.9.9 SDIO data control register (SDIO_DCTRL)

The SDIO_DCTRL register control the data path state machine (DPSM).
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RWSTART
RWSTOP
DTMODE
RWMOD
SDIOEN
DMAEN
DTDIR
DTEN
DBLOCKSIZE
Reserved

Bit 11 SDIOEN: SD I/O enable functions
If this bit is set, the DPSM performs an SD I/O-card-specific operation.
Bit 10 RWMOD: Read wait mode
0: Read Wait control stopping SDIO_D2
1: Read Wait control using SDIO_CK
Bit 9 RWSTOP: Read wait stop
0: Read wait in progress if RWSTART bit is set
1: Enable for read wait stop if RWSTART bit is set
Bit 8 RWSTART: Read wait start
If this bit is set, read wait operation starts.
Bits 7:4 DBLOCKSIZE: Data block size
Define the data block length when the block data transfer mode is selected:
0000: (0 decimal) lock length = 20 = 1 byte
0001: (1 decimal) lock length = 21 = 2 bytes
1111: (15 decimal) reserved
Bit 3 DMAEN: DMA enable bit
0: DMA disabled.
1: DMA enabled.
1066/1751 RM0090 Rev 19

Bit 2 DTMODE: Data transfer mode selection 1: Stream or SDIO multibyte data transfer.
0: Block data transfer
1: Stream or SDIO multibyte data transfer
Bit 1 DTDIR: Data transfer direction selection
0: From controller to card.
1: From card to controller.
Bit 0 DTEN: Data transfer enabled bit
Data transfer starts if 1b is written to the DTEN bit. Depending on the direction bit, DTDIR,
the DPSM moves to the Wait_S, Wait_R state or Readwait if RW Start is set immediately at
the beginning of the transfer. It is not necessary to clear the enable bit after the end of a data
transfer but the SDIO_DCTRL must be updated to enable a new data transfer
Note: After a data write, data cannot be written to this register for three SDIOCLK clock periods
The meaning of the DTMODE bit changes according to the value of the SDIOEN bit. When
SDIOEN=0 and DTMODE=1, the MultiMediaCard stream mode is enabled, and when
SDIOEN=1 and DTMODE=1, the peripheral enables an SDIO multibyte transfer.
31.9.10 SDIO data counter register (SDIO_DCOUNT)

The SDIO_DCOUNT register loads the value from the data length register (see
SDIO_DLEN) when the DPSM moves from the Idle state to the Wait_R or Wait_S state. As
data is transferred, the counter decrements the value until it reaches 0. The DPSM then
moves to the Idle state and the data status end flag, DATAEND, is set.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DATACOUNT
Reserved
r r r r r r r r r r r r r r r r r r r r r r r r r

Bits 24:0 DATACOUNT: Data count value
When this bit is read, the number of remaining data bytes to be transferred is returned. Write
has no effect.
Note: This register should be read only when the data transfer is complete.
RM0090 Rev 19 1067/1751

1075
31.9.11 SDIO status register (SDIO_STA)

The SDIO_STA register is a read-only register. It contains two types of flag:
• Static flags (bits [23:22,10:0]): these bits remain asserted until they are cleared by
writing to the SDIO Interrupt Clear register (see SDIO_ICR)
• Dynamic flags (bits [21:11]): these bits change state depending on the state of the
underlying logic (for example, FIFO full and empty flags are asserted and deasserted
as data while written to the FIFO)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TXUNDERR
CEATAEND
DTIMEOUT
CTIMEOUT
CMDREND
RXFIFOHF
STBITERR
RXOVERR
CMDSENT
TXFIFOHE
DCRCFAIL
CCRCFAIL
DBCKEND
DATAEND
RXFIFOE
RXFIFOF
CMDACT
TXFIFOE
TXFIFOF
RXDAVL
TXDAVL
RXACT
SDIOIT
TXACT
Reserved
r r r r r r r r r r r r r r r r r r r r r r r r

Bit 23 CEATAEND: CE-ATA command completion signal received for CMD61
Bit 22 SDIOIT: SDIO interrupt received
Bit 21 RXDAVL: Data available in receive FIFO
Bit 20 TXDAVL: Data available in transmit FIFO
Bit 19 RXFIFOE: Receive FIFO empty
Bit 18 TXFIFOE: Transmit FIFO empty
When HW Flow Control is enabled, TXFIFOE signals becomes activated when the FIFO
contains 2 words.
Bit 17 RXFIFOF: Receive FIFO full
When HW Flow Control is enabled, RXFIFOF signals becomes activated 2 words before the
FIFO is full.
Bit 16 TXFIFOF: Transmit FIFO full
Bit 15 RXFIFOHF: Receive FIFO half full: there are at least 8 words in the FIFO
Bit 14 TXFIFOHE: Transmit FIFO half empty: at least 8 words can be written into the FIFO
Bit 13 RXACT: Data receive in progress
Bit 12 TXACT: Data transmit in progress
Bit 11 CMDACT: Command transfer in progress
Bit 10 DBCKEND: Data block sent/received (CRC check passed)
Bit 9 STBITERR: Start bit not detected on all data signals in wide bus mode
Bit 8 DATAEND: Data end (data counter, SDIDCOUNT, is zero)
Bit 7 CMDSENT: Command sent (no response required)
Bit 6 CMDREND: Command response received (CRC check passed)
Bit 5 RXOVERR: Received FIFO overrun error
1068/1751 RM0090 Rev 19

Bit 4 TXUNDERR: Transmit FIFO underrun error

Bit 3 DTIMEOUT: Data timeout
Bit 2 CTIMEOUT: Command response timeout
The Command TimeOut period has a fixed value of 64 SDIO_CK clock periods.
Bit 1 DCRCFAIL: Data block sent/received (CRC check failed)
Bit 0 CCRCFAIL: Command response received (CRC check failed)
31.9.12 SDIO interrupt clear register (SDIO_ICR)

The SDIO_ICR register is a write-only register. Writing a bit with 1b clears the corresponding
bit in the SDIO_STA Status register.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TXUNDERRC
CEATAENDC
DTIMEOUTC
CTIMEOUTC
CMDRENDC
STBITERRC
DCRCFAILC
CCRCFAILC
CMDSENTC
RXOVERRC
DBCKENDC
DATAENDC
SDIOITC
Reserved Reserved

Bit 23 CEATAENDC: CEATAEND flag clear bit
Set by software to clear the CEATAEND flag.
0: CEATAEND not cleared
1: CEATAEND cleared
Bit 22 SDIOITC: SDIOIT flag clear bit
Set by software to clear the SDIOIT flag.
0: SDIOIT not cleared
1: SDIOIT cleared
Bit 10 DBCKENDC: DBCKEND flag clear bit
Set by software to clear the DBCKEND flag.
0: DBCKEND not cleared
1: DBCKEND cleared
Bit 9 STBITERRC: STBITERR flag clear bit
Set by software to clear the STBITERR flag.
0: STBITERR not cleared
1: STBITERR cleared
Bit 8 DATAENDC: DATAEND flag clear bit
Set by software to clear the DATAEND flag.
0: DATAEND not cleared
1: DATAEND cleared
RM0090 Rev 19 1069/1751

1075
Bit 7 CMDSENTC: CMDSENT flag clear bit

Set by software to clear the CMDSENT flag.
0: CMDSENT not cleared
1: CMDSENT cleared
Bit 6 CMDRENDC: CMDREND flag clear bit
Set by software to clear the CMDREND flag.
0: CMDREND not cleared
1: CMDREND cleared
Bit 5 RXOVERRC: RXOVERR flag clear bit
Set by software to clear the RXOVERR flag.
0: RXOVERR not cleared
1: RXOVERR cleared
Bit 4 TXUNDERRC: TXUNDERR flag clear bit
Set by software to clear TXUNDERR flag.
0: TXUNDERR not cleared
1: TXUNDERR cleared
Bit 3 DTIMEOUTC: DTIMEOUT flag clear bit
Set by software to clear the DTIMEOUT flag.
0: DTIMEOUT not cleared
1: DTIMEOUT cleared
Bit 2 CTIMEOUTC: CTIMEOUT flag clear bit
Set by software to clear the CTIMEOUT flag.
0: CTIMEOUT not cleared
1: CTIMEOUT cleared
Bit 1 DCRCFAILC: DCRCFAIL flag clear bit
Set by software to clear the DCRCFAIL flag.
0: DCRCFAIL not cleared
1: DCRCFAIL cleared
Bit 0 CCRCFAILC: CCRCFAIL flag clear bit
Set by software to clear the CCRCFAIL flag.
0: CCRCFAIL not cleared
1: CCRCFAIL cleared
1070/1751 RM0090 Rev 19

31.9.13 SDIO mask register (SDIO_MASK)

The interrupt mask register determines which status flags generate an interrupt request by
setting the corresponding bit to 1b.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TXUNDERRIE
CEATAENDIE
DTIMEOUTIE
CTIMEOUTIE
CMDRENDIE
RXFIFOHFIE
STBITERRIE
CMDSENTIE
RXOVERRIE
DCRCFAILIE
CCRCFAILIE
TXFIFOHEIE
DBCKENDIE
DATAENDIE
RXFIFOEIE
RXFIFOFIE
CMDACTIE
TXFIFOEIE
TXFIFOFIE
RXDAVLIE
TXDAVLIE
RXACTIE
SDIOITIE
TXACTIE
Reserved
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 23 CEATAENDIE: CE-ATA command completion signal received interrupt enable
Set and cleared by software to enable/disable the interrupt generated when receiving the
CE-ATA command completion signal.
0: CE-ATA command completion signal received interrupt disabled
1: CE-ATA command completion signal received interrupt enabled
Bit 22 SDIOITIE: SDIO mode interrupt received interrupt enable
Set and cleared by software to enable/disable the interrupt generated when receiving the
SDIO mode interrupt.
0: SDIO Mode Interrupt Received interrupt disabled
1: SDIO Mode Interrupt Received interrupt enabled
Bit 21 RXDAVLIE: Data available in Rx FIFO interrupt enable
Set and cleared by software to enable/disable the interrupt generated by the presence of
data available in Rx FIFO.
0: Data available in Rx FIFO interrupt disabled
1: Data available in Rx FIFO interrupt enabled
Bit 20 TXDAVLIE: Data available in Tx FIFO interrupt enable
Set and cleared by software to enable/disable the interrupt generated by the presence of
data available in Tx FIFO.
0: Data available in Tx FIFO interrupt disabled
1: Data available in Tx FIFO interrupt enabled
Bit 19 RXFIFOEIE: Rx FIFO empty interrupt enable
Set and cleared by software to enable/disable interrupt caused by Rx FIFO empty.
0: Rx FIFO empty interrupt disabled
1: Rx FIFO empty interrupt enabled
Bit 18 TXFIFOEIE: Tx FIFO empty interrupt enable
Set and cleared by software to enable/disable interrupt caused by Tx FIFO empty.
0: Tx FIFO empty interrupt disabled
1: Tx FIFO empty interrupt enabled
Bit 17 RXFIFOFIE: Rx FIFO full interrupt enable
Set and cleared by software to enable/disable interrupt caused by Rx FIFO full.
0: Rx FIFO full interrupt disabled
1: Rx FIFO full interrupt enabled
RM0090 Rev 19 1071/1751

1075
Bit 16 TXFIFOFIE: Tx FIFO full interrupt enable

Set and cleared by software to enable/disable interrupt caused by Tx FIFO full.
0: Tx FIFO full interrupt disabled
1: Tx FIFO full interrupt enabled
Bit 15 RXFIFOHFIE: Rx FIFO half full interrupt enable
Set and cleared by software to enable/disable interrupt caused by Rx FIFO half full.
0: Rx FIFO half full interrupt disabled
1: Rx FIFO half full interrupt enabled
Bit 14 TXFIFOHEIE: Tx FIFO half empty interrupt enable
Set and cleared by software to enable/disable interrupt caused by Tx FIFO half empty.
0: Tx FIFO half empty interrupt disabled
1: Tx FIFO half empty interrupt enabled
Bit 13 RXACTIE: Data receive acting interrupt enable
Set and cleared by software to enable/disable interrupt caused by data being received (data
receive acting).
0: Data receive acting interrupt disabled
1: Data receive acting interrupt enabled
Bit 12 TXACTIE: Data transmit acting interrupt enable
Set and cleared by software to enable/disable interrupt caused by data being transferred
(data transmit acting).
0: Data transmit acting interrupt disabled
1: Data transmit acting interrupt enabled
Bit 11 CMDACTIE: Command acting interrupt enable
Set and cleared by software to enable/disable interrupt caused by a command being
transferred (command acting).
0: Command acting interrupt disabled
1: Command acting interrupt enabled
Bit 10 DBCKENDIE: Data block end interrupt enable
Set and cleared by software to enable/disable interrupt caused by data block end.
0: Data block end interrupt disabled
1: Data block end interrupt enabled
Bit 9 STBITERRIE: Start bit error interrupt enable
Set and cleared by software to enable/disable interrupt caused by start bit error.
0: Start bit error interrupt disabled
1: Start bit error interrupt enabled
Bit 8 DATAENDIE: Data end interrupt enable
Set and cleared by software to enable/disable interrupt caused by data end.
0: Data end interrupt disabled
1: Data end interrupt enabled
Bit 7 CMDSENTIE: Command sent interrupt enable
Set and cleared by software to enable/disable interrupt caused by sending command.
0: Command sent interrupt disabled
1: Command sent interrupt enabled
1072/1751 RM0090 Rev 19

Bit 6 CMDRENDIE: Command response received interrupt enable

Set and cleared by software to enable/disable interrupt caused by receiving command
response.
0: Command response received interrupt disabled
1: command Response Received interrupt enabled
Bit 5 RXOVERRIE: Rx FIFO overrun error interrupt enable
Set and cleared by software to enable/disable interrupt caused by Rx FIFO overrun error.
0: Rx FIFO overrun error interrupt disabled
1: Rx FIFO overrun error interrupt enabled
Bit 4 TXUNDERRIE: Tx FIFO underrun error interrupt enable
Set and cleared by software to enable/disable interrupt caused by Tx FIFO underrun error.
0: Tx FIFO underrun error interrupt disabled
1: Tx FIFO underrun error interrupt enabled
Bit 3 DTIMEOUTIE: Data timeout interrupt enable
Set and cleared by software to enable/disable interrupt caused by data timeout.
0: Data timeout interrupt disabled
1: Data timeout interrupt enabled
Bit 2 CTIMEOUTIE: Command timeout interrupt enable
Set and cleared by software to enable/disable interrupt caused by command timeout.
0: Command timeout interrupt disabled
1: Command timeout interrupt enabled
Bit 1 DCRCFAILIE: Data CRC fail interrupt enable
Set and cleared by software to enable/disable interrupt caused by data CRC failure.
0: Data CRC fail interrupt disabled
1: Data CRC fail interrupt enabled
Bit 0 CCRCFAILIE: Command CRC fail interrupt enable
Set and cleared by software to enable/disable interrupt caused by command CRC failure.
0: Command CRC fail interrupt disabled
1: Command CRC fail interrupt enabled
31.9.14 SDIO FIFO counter register (SDIO_FIFOCNT)

The SDIO_FIFOCNT register contains the remaining number of words to be written to or
read from the FIFO. The FIFO counter loads the value from the data length register (see
SDIO_DLEN) when the data transfer enable bit, DTEN, is set in the data control register
(SDIO_DCTRL register) and the DPSM is at the Idle state. If the data length is not word-
aligned (multiple of 4), the remaining 1 to 3 bytes are regarded as a word.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FIFOCOUNT
Reserved
r r r r r r r r r r r r r r r r r r r r r r r r

Bits 23:0 FIFOCOUNT: Remaining number of words to be written to or read from the FIFO.
RM0090 Rev 19 1073/1751

1075
31.9.15 SDIO data FIFO register (SDIO_FIFO)

The receive and transmit FIFOs can be read or written as 32-bit wide registers. The FIFOs
contain 32 entries on 32 sequential addresses. This allows the CPU to use its load and store
multiple operands to read from/write to the FIFO.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FIF0Data
bits 31:0 FIFOData: Receive and transmit FIFO data

The FIFO data occupies 32 entries of 32-bit words, from address:
SDIO base + 0x080 to SDIO base + 0xFC.
31.9.16 SDIO register map

The following table summarizes the SDIO registers.
Table 181. SDIO register map

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
PWRCTRL
Reserved
0x00 SDIO_POWER
NEGEDGE
HWFC_EN
Reserved
PWRSAV
WIDBUS
BYPASS
CLKDIV
CLKEN
0x04 SDIO_CLKCR
0x08 SDIO_ARG CMDARG

SDIOSuspend
ENCMDcompl
CE-ATACMD
CMDINDEX
WAITPEND
WAITRESP
CPSMEN
Reserved
WAITINT
nIEN
0x0C SDIO_CMD
0x10 SDIO_RESPCMD Reserved RESPCMD

0x14 SDIO_RESP1 CARDSTATUS1
0x1C SDIO_RESP3 CARDSTATUS3
0x24 SDIO_DTIMER DATATIME
0x28 SDIO_DLEN Reserved DATALENGTH
DBLOCKSIZE
RWSTART
RWSTOP
DTMODE
Reserved
RWMOD
SDIOEN
DMAEN
DTDIR
DTEN
0x2C SDIO_DCTRL
0x30 SDIO_DCOUNT Reserved DATACOUNT
1074/1751 RM0090 Rev 19

0x80
0x48
0x38
0x34
0x3C
Offset
RM0090
SDIO_ICR
SDIO_STA
Register
SDIO_FIFO
SDIO_MASK
SDIO_FIFOCNT
31
30
29
Reserved Reserved Reserved 28
27
Reserved
26
25
24
CEATAENDIE CEATAENDC CEATAEND 23
SDIOITIE SDIOITC SDIOIT 22
RXDAVLIE RXDAVL 21
TXDAVLIE TXDAVL 20
RXFIFOEIE RXFIFOE 19
TXFIFOEIE TXFIFOE 18
RXFIFOFIE RXFIFOF
RM0090 Rev 19
17
TXFIFOFIE Reserved TXFIFOF 16
RXFIFOHFIE RXFIFOHF 15
FIF0Data
TXFIFOHEIE TXFIFOHE 14
RXACTIE RXACT 13
TXACTIE TXACT 12
Table 181. SDIO register map (continued)
CMDACTIE CMDACT 11
DBCKENDIE DBCKENDC DBCKEND
FIFOCOUNT
10
STBITERRIE STBITERRC STBITERR 9
DATAENDIE DATAENDC DATAEND 8
CMDSENTIE CMDSENTC CMDSENT 7
CMDRENDIE CMDRENDC CMDREND 6
RXOVERRIE RXOVERRC RXOVERR 5
TXUNDERRIE TXUNDERRC TXUNDERR 4
DTIMEOUTIE DTIMEOUTC DTIMEOUT 3
CTIMEOUTIE CTIMEOUTC CTIMEOUT 2
DCRCFAILIE DCRCFAILC DCRCFAIL 1
CCRCFAILIE CCRCFAILC CCRCFAIL
1075/1751
Secure digital input/output interface (SDIO)
1075
Controller area network (bxCAN) RM0090
32 Controller area network (bxCAN)
32.1 bxCAN introduction

The Basic Extended CAN peripheral, named bxCAN, interfaces the CAN network. It
supports the CAN protocols version 2.0A and B. It has been designed to manage a high
number of incoming messages efficiently with a minimum CPU load. It also meets the
priority requirements for transmit messages.
For safety-critical applications, the CAN controller provides all hardware functions for
supporting the CAN Time Triggered Communication option.
32.2 bxCAN main features

• Supports CAN protocol version 2.0 A, B Active
• Bit rates up to 1 Mbit/s
• Supports the Time Triggered Communication option
Transmission
• Three transmit mailboxes
• Configurable transmit priority
• Time Stamp on SOF transmission
Reception
• Two receive FIFOs with three stages
• Scalable filter banks:
– 28 filter banks shared between CAN1 and CAN2
• Identifier list feature
• Configurable FIFO overrun
• Time Stamp on SOF reception
Time-triggered communication option

• Disable automatic retransmission mode
• 16-bit free running timer
• Time Stamp sent in last two data bytes
Management
• Maskable interrupts
• Software-efficient mailbox mapping at a unique address space
1076/1751 RM0090 Rev 19

RM0090 Controller area network (bxCAN)
Dual CAN
• CAN1: Master bxCAN for managing the communication between a Slave bxCAN and
the 512-byte SRAM memory
• CAN2: Slave bxCAN, with no direct access to the SRAM memory.
• The two bxCAN cells share the 512-byte SRAM memory (see Figure 335)
32.3 bxCAN general description

In today’s CAN applications, the number of nodes in a network is increasing and often
several networks are linked together via gateways. Typically the number of messages in the
system (and thus to be handled by each node) has significantly increased. In addition to the
application messages, Network Management and Diagnostic messages have been
introduced.
• An enhanced filtering mechanism is required to handle each type of message.
Furthermore, application tasks require more CPU time, therefore real-time constraints
caused by message reception have to be reduced.
• A receive FIFO scheme allows the CPU to be dedicated to application tasks for a long
time period without losing messages.
The standard HLP (Higher Layer Protocol) based on standard CAN drivers requires an
efficient interface to the CAN controller.
Figure 334. CAN network topology

CAN node 1
CAN node 2
CAN node n
MCU
Application
CAN
Controller
CAN CAN
Rx Tx
CAN
Transceiver
CAN CAN
High Low
CAN Bus
32.3.1 CAN 2.0B active core

The bxCAN module handles the transmission and the reception of CAN messages fully
autonomously. Standard identifiers (11-bit) and extended identifiers (29-bit) are fully
supported by hardware.
RM0090 Rev 19 1077/1751

1121
32.3.2 Control, status and configuration registers

The application uses these registers to:
• Configure CAN parameters, e.g. baud rate
• Request transmissions
• Handle receptions
• Manage interrupts
• Get diagnostic information
32.3.3 Tx mailboxes
Three transmit mailboxes are provided to the software for setting up messages. The
transmission Scheduler decides which mailbox has to be transmitted first.
32.3.4 Acceptance filters

The bxCAN provides 28 scalable/configurable identifier filter banks for selecting the
incoming messages the software needs and discarding the others.
Receive FIFO
Two receive FIFOs are used by hardware to store the incoming messages. Three complete
messages can be stored in each FIFO. The FIFOs are managed completely by hardware.
1078/1751 RM0090 Rev 19

Figure 335. Dual CAN block diagram
CAN1 (Master) with 512 bytes SRAM

Master
Tx Mailboxes Master Master
Receive FIFO 0 Receive FIFO 1
2
1 2 2
Mailbox 0 1 1
Mailbox 0 Mailbox 0
Master Control
Master Status
Tx Status
Rx FIFO 0 Status Transmission
Scheduler
Rx FIFO 1 Status
Control/Status/Configuration
Acceptance Filters
Interrupt Enable
27
CAN 2.0B Active Core .. .. 26
Error Status Memory 2 3
1
Access Filter 0
Bit Timing Controller
Filter Master Master Filters Slave Filters

(0 to 27) (0 to 27)
Filter Mode
Transmission
Filter Scale Scheduler
Slave Slave
Slave Receive FIFO 0 Receive FIFO 1
Filter FIFOAssign Tx Mailboxes 2 2
Filter Activation 2 1 1
1 Mailbox 0 Mailbox 0
Mailbox 0
CAN2 (Slave)
Master Control
Control/Status/Configuration
Master Status
Tx Status
Rx FIFO 0 Status
CAN 2.0B Active Core
Rx FIFO 1 Status
Interrupt Enable
Error Status
Note: CAN 2 start filter bank number n is confi gurable by writing to
Bit Timing the CAN2SB[5:0] bits in the CAN_ FMRregister.
ai16094b
32.4 bxCAN operating modes

bxCAN has three main operating modes: initialization, normal and Sleep. After a
hardware reset, bxCAN is in Sleep mode to reduce power consumption and an internal pull-
up is active on CANTX. The software requests bxCAN to enter initialization or Sleep mode
by setting the INRQ or SLEEP bits in the CAN_MCR register. Once the mode has been
entered, bxCAN confirms it by setting the INAK or SLAK bits in the CAN_MSR register and
the internal pull-up is disabled. When neither INAK nor SLAK are set, bxCAN is in normal
RM0090 Rev 19 1079/1751

1121
mode. Before entering normal mode bxCAN always has to synchronize on the CAN bus.
To synchronize, bxCAN waits until the CAN bus is idle, this means 11 consecutive recessive
bits have been monitored on CANRX.
32.4.1 Initialization mode

The software initialization can be done while the hardware is in Initialization mode. To enter
this mode the software sets the INRQ bit in the CAN_MCR register and waits until the
hardware has confirmed the request by setting the INAK bit in the CAN_MSR register.
To leave Initialization mode, the software clears the INRQ bit. bxCAN has left Initialization
mode once the INAK bit has been cleared by hardware.
While in Initialization Mode, all message transfers to and from the CAN bus are stopped and
the status of the CAN bus output CANTX is recessive (high).
Entering Initialization Mode does not change any of the configuration registers.
To initialize the CAN Controller, software has to set up the Bit Timing (CAN_BTR) and CAN
options (CAN_MCR) registers.
To initialize the registers associated with the CAN filter banks (mode, scale, FIFO
assignment, activation and filter values), software has to set the FINIT bit (CAN_FMR). Filter
initialization also can be done outside the initialization mode.
Note: When FINIT=1, CAN reception is deactivated.
The filter values also can be modified by deactivating the associated filter activation bits (in
the CAN_FA1R register).
If a filter bank is not used, it is recommended to leave it non active (leave the corresponding
FACT bit cleared).
32.4.2 Normal mode

Once the initialization is complete, the software must request the hardware to enter Normal
mode to be able to synchronize on the CAN bus and start reception and transmission.
The request to enter Normal mode is issued by clearing the INRQ bit in the CAN_MCR
register. The bxCAN enters Normal mode and is ready to take part in bus activities when it
has synchronized with the data transfer on the CAN bus. This is done by waiting for the
occurrence of a sequence of 11 consecutive recessive bits (Bus Idle state). The switch to
Normal mode is confirmed by the hardware by clearing the INAK bit in the CAN_MSR
register.
The initialization of the filter values is independent from Initialization Mode but must be done
while the filter is not active (corresponding FACTx bit cleared). The filter scale and mode
configuration must be configured before entering Normal Mode.
32.4.3 Sleep mode (low-power)

To reduce power consumption, bxCAN has a low-power mode called Sleep mode. This
mode is entered on software request by setting the SLEEP bit in the CAN_MCR register. In
this mode, the bxCAN clock is stopped, however software can still access the bxCAN
mailboxes.
If software requests entry to initialization mode by setting the INRQ bit while bxCAN is in
Sleep mode, it must also clear the SLEEP bit.
1080/1751 RM0090 Rev 19

bxCAN can be woken up (exit Sleep mode) either by software clearing the SLEEP bit or on
detection of CAN bus activity.
On CAN bus activity detection, hardware automatically performs the wakeup sequence by
clearing the SLEEP bit if the AWUM bit in the CAN_MCR register is set. If the AWUM bit is
cleared, software has to clear the SLEEP bit when a wakeup interrupt occurs, in order to exit
from Sleep mode.
Note: If the wakeup interrupt is enabled (WKUIE bit set in CAN_IER register) a wakeup interrupt
will be generated on detection of CAN bus activity, even if the bxCAN automatically
performs the wakeup sequence.
After the SLEEP bit has been cleared, Sleep mode is exited once bxCAN has synchronized
with the CAN bus, refer to Figure 336. The Sleep mode is exited once the SLAK bit has
been cleared by hardware.
Figure 336. bxCAN operating modes

Reset
Sleep
SLAK= 1
INAK = 0
Q
NR SL
.I EE
C P
YN .I
NR
.S SL Q
EP CK EE .A
E .A P CK
SL .I
E EP NR
SL Q
.A
CK
Normal INRQ . ACK Initialization

SLAK= 0 SLAK= 0
INAK = 0 INAK = 1
INRQ . SYNC . SLEEP
ai15902
1. ACK = The wait state during which hardware confirms a request by setting the INAK or SLAK bits in the
CAN_MSR register
2. SYNC = The state during which bxCAN waits until the CAN bus is idle, meaning 11 consecutive recessive
bits have been monitored on CANRX
32.5 Test mode

Test mode can be selected by the SILM and LBKM bits in the CAN_BTR register. These bits
must be configured while bxCAN is in Initialization mode. Once test mode has been
selected, the INRQ bit in the CAN_MCR register must be reset to enter Normal mode.
32.5.1 Silent mode

The bxCAN can be put in Silent mode by setting the SILM bit in the CAN_BTR register.
In Silent mode, the bxCAN is able to receive valid data frames and valid remote frames, but
it sends only recessive bits on the CAN bus and it cannot start a transmission. If the bxCAN
has to send a dominant bit (ACK bit, overload flag, active error flag), the bit is rerouted
internally so that the CAN Core monitors this dominant bit, although the CAN bus may
RM0090 Rev 19 1081/1751

1121
remain in recessive state. Silent mode can be used to analyze the traffic on a CAN bus
without affecting it by the transmission of dominant bits (Acknowledge Bits, Error Frames).
Figure 337. bxCAN in silent mode
bxCAN
Tx Rx
=1
CANTX CANRX
32.5.2 Loop back mode

The bxCAN can be set in Loop Back Mode by setting the LBKM bit in the CAN_BTR
register. In Loop Back Mode, the bxCAN treats its own transmitted messages as received
messages and stores them (if they pass acceptance filtering) in a Receive mailbox.
Figure 338. bxCAN in loop back mode
bxCAN
Tx Rx
CANTX CANRX
This mode is provided for self-test functions. To be independent of external events, the CAN
Core ignores acknowledge errors (no dominant bit sampled in the acknowledge slot of a
data / remote frame) in Loop Back Mode. In this mode, the bxCAN performs an internal
feedback from its Tx output to its Rx input. The actual value of the CANRX input pin is
disregarded by the bxCAN. The transmitted messages can be monitored on the CANTX pin.
32.5.3 Loop back combined with silent mode

It is also possible to combine Loop Back mode and Silent mode by setting the LBKM and
SILM bits in the CAN_BTR register. This mode can be used for a “Hot Selftest”, meaning the
bxCAN can be tested like in Loop Back mode but without affecting a running CAN system
connected to the CANTX and CANRX pins. In this mode, the CANRX pin is disconnected
from the bxCAN and the CANTX pin is held recessive.
1082/1751 RM0090 Rev 19

Figure 339. bxCAN in combined mode
bxCAN
Tx Rx
=1
CANTX CANRX
32.6 Debug mode

When the microcontroller enters the debug mode (Cortex®-M4 with FPU core halted), the
bxCAN continues to work normally or stops, depending on:
• the DBG_CAN1_STOP bit for CAN1 or the DBG_CAN2_STOP bit for CAN2 in the
DBG module. For more details, refer to Section 38.16.2: Debug support for timers,
watchdog, bxCAN and I2C.
• the DBF bit in CAN_MCR. For more details, refer to Section 32.9.2.
32.7 bxCAN functional description
32.7.1 Transmission handling

In order to transmit a message, the application must select one empty transmit mailbox, set
up the identifier, the data length code (DLC) and the data before requesting the transmission
by setting the corresponding TXRQ bit in the CAN_TIxR register. Once the mailbox has left
empty state, the software no longer has write access to the mailbox registers. Immediately
after the TXRQ bit has been set, the mailbox enters pending state and waits to become the
highest priority mailbox, see Transmit Priority. As soon as the mailbox has the highest
priority it will be scheduled for transmission. The transmission of the message of the
scheduled mailbox will start (enter transmit state) when the CAN bus becomes idle. Once
the mailbox has been successfully transmitted, it will become empty again. The hardware
indicates a successful transmission by setting the RQCP and TXOK bits in the CAN_TSR
register.
If the transmission fails, the cause is indicated by the ALST bit in the CAN_TSR register in
case of an Arbitration Lost, and/or the TERR bit, in case of transmission error detection.
RM0090 Rev 19 1083/1751

1121
Transmit priority
By identifier When more than one transmit mailbox is pending, the

transmission order is given by the identifier of the message
stored in the mailbox. The message with the lowest identifier
value has the highest priority according to the arbitration of the
CAN protocol. If the identifier values are equal, the lower mailbox
number will be scheduled first.
By transmit request order The transmit mailboxes can be configured as a transmit FIFO by
setting the TXFP bit in the CAN_MCR register. In this mode the
priority order is given by the transmit request order. This mode is
very useful for segmented transmission.
Abort
A transmission request can be aborted by the user setting the ABRQ bit in the CAN_TSR
register. In pending or scheduled state, the mailbox is aborted immediately. An abort
request while the mailbox is in transmit state can have two results. If the mailbox is
transmitted successfully the mailbox becomes empty with the TXOK bit set in the
CAN_TSR register. If the transmission fails, the mailbox becomes scheduled, the
transmission is aborted and becomes empty with TXOK cleared. In all cases the mailbox
will become empty again at least at the end of the current transmission.
Nonautomatic retransmission mode

This mode has been implemented in order to fulfil the requirement of the Time Triggered
Communication option of the CAN standard. To configure the hardware in this mode the
NART bit in the CAN_MCR register must be set.
In this mode, each transmission is started only once. If the first attempt fails, due to an
arbitration loss or an error, the hardware will not automatically restart the message
transmission.
At the end of the first transmission attempt, the hardware considers the request as
completed and sets the RQCP bit in the CAN_TSR register. The result of the transmission is
indicated in the CAN_TSR register by the TXOK, ALST and TERR bits.
1084/1751 RM0090 Rev 19

Figure 340. Transmit mailbox states
EMPTY
RQCP=X
TXOK=X
TXRQ=1
TME = 1
PENDING
RQCP=0 Mailbox has
TXOK=0 highest priority
ABRQ=1
TME = 0
Mailbox does not

have highest priority
EMPTY SCHEDULED
RQCP=1 ABRQ=1 RQCP=0
TXOK=0 TXOK=0
TME = 1 TME = 0
CAN Bus = IDLE
Transmit failed * NART TRANSMIT Transmit failed * NART

RQCP=0
TXOK=0
TME = 0
EMPTY
Transmit succeeded
RQCP=1
TXOK=1
TME = 1
32.7.2 Time triggered communication mode

In this mode, the internal counter of the CAN hardware is activated and used to generate the
Time Stamp value stored in the CAN_RDTxR/CAN_TDTxR registers, respectively (for Rx
and Tx mailboxes). The internal counter is incremented each CAN bit time (refer to
Section 32.7.7). The internal counter is captured on the sample point of the Start Of Frame
bit in both reception and transmission.
32.7.3 Reception handling

For the reception of CAN messages, three mailboxes organized as a FIFO are provided. In
order to save CPU load, simplify the software and guarantee data consistency, the FIFO is
managed completely by hardware. The application accesses the messages stored in the
FIFO through the FIFO output mailbox.
Valid message
A received message is considered as valid when it has been received correctly according to
the CAN protocol (no error until the last but one bit of the EOF field) and It passed through
the identifier filtering successfully, see Section 32.7.4.
RM0090 Rev 19 1085/1751

1121
Figure 341. Receive FIFO states

EMPTY
FMP=0x00 Valid Message
FOVR=0 Received
PENDING_1
Release FMP=0x01
Mailbox FOVR=0
Release Valid Message

Mailbox Received
RFOM=1
PENDING_2
FMP=0x10
FOVR=0
Release Valid Message

Mailbox Received
RFOM=1
PENDING_3
FMP=0x11 Valid Message
FOVR=0 Received
OVERRUN
Release FMP=0x11
Mailbox FOVR=1
RFOM=1
Valid Message
Received
FIFO management
Starting from the empty state, the first valid message received is stored in the FIFO which
becomes pending_1. The hardware signals the event setting the FMP[1:0] bits in the
CAN_RFR register to the value 01b. The message is available in the FIFO output mailbox.
The software reads out the mailbox content and releases it by setting the RFOM bit in the
CAN_RFR register. The FIFO becomes empty again. If a new valid message has been
received in the meantime, the FIFO stays in pending_1 state and the new message is
available in the output mailbox.
If the application does not release the mailbox, the next valid message will be stored in the
FIFO which enters pending_2 state (FMP[1:0] = 10b). The storage process is repeated for
the next valid message putting the FIFO into pending_3 state (FMP[1:0] = 11b). At this
point, the software must release the output mailbox by setting the RFOM bit, so that a
mailbox is free to store the next valid message. Otherwise the next valid message received
will cause a loss of message.
Refer also to Section 32.7.5
Overrun
Once the FIFO is in pending_3 state (i.e. the three mailboxes are full) the next valid
message reception will lead to an overrun and a message will be lost. The hardware
1086/1751 RM0090 Rev 19

signals the overrun condition by setting the FOVR bit in the CAN_RFR register. Which
message is lost depends on the configuration of the FIFO:
• If the FIFO lock function is disabled (RFLM bit in the CAN_MCR register cleared) the
last message stored in the FIFO will be overwritten by the new incoming message. In
this case the latest messages will be always available to the application.
• If the FIFO lock function is enabled (RFLM bit in the CAN_MCR register set) the most
recent message will be discarded and the software will have the three oldest messages
in the FIFO available.
Reception related interrupts

Once a message has been stored in the FIFO, the FMP[1:0] bits are updated and an
interrupt request is generated if the FMPIE bit in the CAN_IER register is set.
When the FIFO becomes full (i.e. a third message is stored) the FULL bit in the CAN_RFR
register is set and an interrupt is generated if the FFIE bit in the CAN_IER register is set.
On overrun condition, the FOVR bit is set and an interrupt is generated if the FOVIE bit in
the CAN_IER register is set.
32.7.4 Identifier filtering

In the CAN protocol the identifier of a message is not associated with the address of a node
but related to the content of the message. Consequently a transmitter broadcasts its
message to all receivers. On message reception a receiver node decides - depending on
the identifier value - whether the software needs the message or not. If the message is
needed, it is copied into the SRAM. If not, the message must be discarded without
intervention by the software.
To fulfill this requirement, the bxCAN Controller provides 28 configurable and scalable filter
banks (27-0) to the application. This hardware filtering saves CPU resources which would
be otherwise needed to perform filtering by software. Each filter bank x consists of two 32-bit
registers, CAN_FxR0 and CAN_FxR1.
Scalable width
To optimize and adapt the filters to the application needs, each filter bank can be scaled
independently. Depending on the filter scale a filter bank provides:
• One 32-bit filter for the STDID[10:0], EXTID[17:0], IDE and RTR bits.
• Two 16-bit filters for the STDID[10:0], RTR, IDE and EXTID[17:15] bits.
Refer to Figure 342.
Furthermore, the filters can be configured in mask mode or in identifier list mode.
Mask mode
In mask mode the identifier registers are associated with mask registers specifying which
bits of the identifier are handled as “must match” or as “don’t care”.
Identifier list mode

In identifier list mode, the mask registers are used as identifier registers. Thus instead of
defining an identifier and a mask, two identifiers are specified, doubling the number of single
identifiers. All bits of the incoming identifier must match the bits specified in the filter
registers.
RM0090 Rev 19 1087/1751

1121
Filter bank scale and mode configuration

The filter banks are configured by means of the corresponding CAN_FMR register. To
configure a filter bank it must be deactivated by clearing the FACT bit in the CAN_FAR
register. The filter scale is configured by means of the corresponding FSCx bit in the
CAN_FS1R register, refer to Figure 342. The identifier list or identifier mask mode for the
corresponding Mask/Identifier registers is configured by means of the FBMx bits in the
CAN_FMR register.
To filter a group of identifiers, configure the Mask/Identifier registers in mask mode.
To select single identifiers, configure the Mask/Identifier registers in identifier list mode.
Filters not used by the application should be left deactivated.
Each filter within a filter bank is numbered (called the Filter Number) from 0 to a maximum
dependent on the mode and the scale of each of the filter banks.
Concerning the filter configuration, refer to Figure 342.
Figure 342. Filter bank scale configuration - register organization

Filter
One 32-Bit Filter - Identifier Mask Num.
FBMx = 0
ID CAN_FxR1[31:24] CAN_FxR1[23:16] CAN_FxR1[15:8] CAN_FxR1[7:0]

n
Mask CAN_FxR2[31:24] CAN_FxR2[23:16] CAN_FxR2[15:8] CAN_FxR2[7:0]
FSCx = 1
Mapping STID[10:3] STID[2:0] EXID[17:13] EXID[12:5] EXID[4:0] IDE RTR 0
Two 32-Bit Filters - Identifier List

FBMx = 1
ID CAN_FxR1[31:24] CAN_FxR1[23:16] CAN_FxR1[15:8] CAN_FxR1[7:0] n

ID CAN_FxR2[31:24] CAN_FxR2[23:16] CAN_FxR2[15:8] CAN_FxR2[7:0] n+1
Mapping STID[10:3] STID[2:0] EXID[17:13] EXID[12:5] EXID[4:0] IDE RTR 0
Two 16-Bit Filters - Identifier Mask

ID CAN_FxR1[15:8] CAN_FxR1[7:0]
n
FBMx = 0
Mask CAN_FxR1[31:24] CAN_FxR1[23:16]
ID CAN_FxR2[15:8] CAN_FxR2[7:0]
n+1
Mask CAN_FxR2[31:24] CAN_FxR2[23:16]
FSCx = 0
Mapping STID[10:3] STID[2:0] RTR IDE EXID[17:15]
Four 16-Bit Filters - Identifier List

ID CAN_FxR1[15:8] CAN_FxR1[7:0] n
FBMx = 1
ID CAN_FxR1[31:24] CAN_FxR1[23:16] n+1

Mapping STID[10:3] STID[2:0] RTR IDE EXID[17:15]
Filter Bank Mode2
Filter Bank Scale
Config. Bits1
x = filter bank number

ID=Identifier
1
These bits are located in the CAN_FS1R register
2
These bits are located in the CAN_FM1R register
Filter match index

Once a message has been received in the FIFO it is available to the application. Typically,
application data is copied into SRAM locations. To copy the data to the right location the
1088/1751 RM0090 Rev 19

application has to identify the data by means of the identifier. To avoid this, and to ease the
access to the SRAM locations, the CAN controller provides a Filter Match Index.
This index is stored in the mailbox together with the message according to the filter priority
rules. Thus each received message has its associated filter match index.
The Filter Match index can be used in two ways:
• Compare the Filter Match index with a list of expected values.
• Use the Filter Match Index as an index on an array to access the data destination
location.
For nonmasked filters, the software no longer has to compare the identifier.
If the filter is masked the software reduces the comparison to the masked bits only.
The index value of the filter number does not take into account the activation state of the
filter banks. In addition, two independent numbering schemes are used, one for each FIFO.
Refer to Figure 343 for an example.
Figure 343. Example of filter numbering

Filter FIFO0 Filter Filter FIFO1 Filter
Bank Num. Bank Num.
0 0
0 ID List (32-bit) 2 ID Mask (16-bit)
1 1
2
1 ID Mask (32-bit) 2 4 ID List (32-bit) 3
3
4 Deactivated 4
3 ID List (16-bit) 7
5 ID Mask (16-bit) 5
6
Deactivated 7 6
5 8 ID Mask (16-bit)
ID List (32-bit) 8 7
8
9 Deactivated 9
6 ID Mask (16-bit) 10
10 ID List (16-bit) 10
11
11 12
9 ID List (32-bit) 11 ID List (32-bit)
12 13
13 ID Mask (32-bit) 13 12 ID Mask (32-bit) 14
ID = Identifier
RM0090 Rev 19 1089/1751

1121
Filter priority rules

Depending on the filter combination it may occur that an identifier passes successfully
through several filters. In this case the filter match value stored in the receive mailbox is
chosen according to the following priority rules:
• A 32-bit filter takes priority over a 16-bit filter.
• For filters of equal scale, priority is given to the Identifier List mode over the Identifier
Mask mode
• For filters of equal scale and mode, priority is given by the filter number (the lower the
number, the higher the priority).
Figure 344. Filtering mechanism - Example

Three filter banks in 32-bit Unidentified List mode,
the remaining in 32-bit Identifier Mask mode
Message Received
Identifier Ctrl Data
Filter bank
Num Receive FIFO
Identifier 0
0
Identifier 1 Message
Identifier List
Identifier 4 Identifier #4 Match Stored

2
Identifier 5
Identifier & Mask
1
Identifier
Mask 2 Filter number stored in the
FMI
Filter Match Index field
within the CAN_RDTxR
Identifier register
4 3
Mask
No Match
Found
Message Discarded
The example above shows the filtering principle of the bxCAN. On reception of a message,
the identifier is compared first with the filters configured in identifier list mode. If there is a
match, the message is stored in the associated FIFO and the index of the matching filter is
stored in the Filter Match Index. As shown in the example, the identifier matches with
Identifier #2 thus the message content and FMI 2 is stored in the FIFO.
If there is no match, the incoming identifier is then compared with the filters configured in
mask mode.
If the identifier does not match any of the identifiers configured in the filters, the message is
discarded by hardware without disturbing the software.
1090/1751 RM0090 Rev 19

32.7.5 Message storage

The interface between the software and the hardware for the CAN messages is
implemented by means of mailboxes. A mailbox contains all information related to a
message; identifier, data, control, status and time stamp information.
Transmit mailbox
The software sets up the message to be transmitted in an empty transmit mailbox. The
status of the transmission is indicated by hardware in the CAN_TSR register.
Table 182. Transmit mailbox mapping

Offset to transmit mailbox base address (bytes) Register name
0 CAN_TIxR
4 CAN_TDTxR
8 CAN_TDLxR
12 CAN_TDHxR
Receive mailbox
When a message has been received, it is available to the software in the FIFO output
mailbox. Once the software has handled the message (e.g. read it) the software must
release the FIFO output mailbox by means of the RFOM bit in the CAN_RFR register to
make the next incoming message available. The filter match index is stored in the MFMI
field of the CAN_RDTxR register. The 16-bit time stamp value is stored in the TIME[15:0]
field of CAN_RDTxR.
Table 183. Receive mailbox mapping

Offset to receive mailbox base address (bytes) Register name
0 CAN_RIxR
4 CAN_RDTxR
8 CAN_RDLxR
12 CAN_RDHxR
RM0090 Rev 19 1091/1751

1121
Figure 345. CAN error state diagram

When TEC or REC > 127
ERROR ACTIVE
ERROR PASSIVE
When TEC and REC < 128,
When 128 * 11 recessive bits occur: When TEC > 255
BUS OFF
ai15903
1092/1751 RM0090 Rev 19


The error management as described in the CAN protocol is handled entirely by hardware
using a Transmit Error Counter (TEC value, in CAN_ESR register) and a Receive Error
Counter (REC value, in the CAN_ESR register), which get incremented or decremented
according to the error condition. For detailed information about TEC and REC management,
refer to the CAN standard.
Both of them may be read by software to determine the stability of the network.
Furthermore, the CAN hardware provides detailed information on the current error status in
CAN_ESR register. By means of the CAN_IER register (ERRIE bit, etc.), the software can
configure the interrupt generation on error detection in a very flexible way.
Bus-Off recovery
The Bus-Off state is reached when TEC is greater than 255, this state is indicated by BOFF
bit in CAN_ESR register. In Bus-Off state, the bxCAN is no longer able to transmit and
receive messages.
Depending on the ABOM bit in the CAN_MCR register bxCAN will recover from Bus-Off
(become error active again) either automatically or on software request. But in both cases
the bxCAN has to wait at least for the recovery sequence specified in the CAN standard
(128 occurrences of 11 consecutive recessive bits monitored on CANRX).
If ABOM is set, the bxCAN will start the recovering sequence automatically after it has
entered Bus-Off state.
If ABOM is cleared, the software must initiate the recovering sequence by requesting
bxCAN to enter and to leave initialization mode.
Note: In initialization mode, bxCAN does not monitor the CANRX signal, therefore it cannot
complete the recovery sequence. To recover, bxCAN must be in normal mode.
32.7.7 Bit timing

The bit timing logic monitors the serial bus-line and performs sampling and adjustment of
the sample point by synchronizing on the start-bit edge and resynchronizing on the following
edges.
Its operation may be explained simply by splitting nominal bit time into three segments as
follows:
• Synchronization segment (SYNC_SEG): a bit change is expected to occur within this
time segment. It has a fixed length of one time quantum (1 x tq).
• Bit segment 1 (BS1): defines the location of the sample point. It includes the
PROP_SEG and PHASE_SEG1 of the CAN standard. Its duration is programmable
between 1 and 16 time quanta but may be automatically lengthened to compensate for
positive phase drifts due to differences in the frequency of the various nodes of the
network.
• Bit segment 2 (BS2): defines the location of the transmit point. It represents the
PHASE_SEG2 of the CAN standard. Its duration is programmable between 1 and 8
time quanta but may also be automatically shortened to compensate for negative
phase drifts.
The resynchronization Jump Width (SJW) defines an upper bound to the amount of
lengthening or shortening of the bit segments. It is programmable between 1 and 4 time
quanta.
RM0090 Rev 19 1093/1751

1121
A valid edge is defined as the first transition in a bit time from dominant to recessive bus
level provided the controller itself does not send a recessive bit.
If a valid edge is detected in BS1 instead of SYNC_SEG, BS1 is extended by up to SJW so
that the sample point is delayed.
Conversely, if a valid edge is detected in BS2 instead of SYNC_SEG, BS2 is shortened by
up to SJW so that the transmit point is moved earlier.
As a safeguard against programming errors, the configuration of the Bit Timing register
(CAN_BTR) is only possible while the device is in Standby mode.
Note: For a detailed description of the CAN bit timing and resynchronization mechanism, refer to
the ISO 11898 standard.
Figure 346. Bit timing
NOMINAL BIT TIME
SYNC_SEG BIT SEGMENT 1 (BS1) BIT SEGMENT 2 (BS2)
1 x tq tBS1 tBS2
SAMPLE POINT TRANSMIT POINT

1
BaudRate = ----------------------------------------------
NominalBitTime
NominalBitTime = 1 × t q + t BS1 + t BS2
with:
tBS1 = tq x (TS1[3:0] + 1),
tBS2 = tq x (TS2[2:0] + 1),
tq = (BRP[9:0] + 1) x tPCLK
where tq refers to the Time quantum
tPCLK = time period of the APB clock,
BRP[9:0], TS1[3:0] and TS2[2:0] are defined in the CAN_BTR register.
1094/1751 RM0090 Rev 19

Figure 347. CAN frames

Inter-Frame Space
Inter-Frame Space Data Frame (Standard identifier) or Overload Frame
44 + 8 * N
Arbitration Field Ctrl Field Data Field CRC Field Ack Field
2
32 6 8*N 16 7
ID DLC CRC EOF
IDE
r0
RTR
SOF
ACK
Inter-Frame Space
Inter-Frame Space Data Frame (Extended Identifier) or Overload Frame
64 + 8 * N
Arbitration Field Arbitration Field Ctrl Field Data Field CRC Field Ack Field
2
32 32 6 8*N 16 7
ID DLC CRC EOF
r1
RTR
IDE
r0
SRR
SOF
ACK
Inter-Frame Space
Inter-Frame Space Remote Frame or Overload Frame
44
Arbitration Field Ctrl Field CRC Field Ack Field
2
32 6 16 7
ID DLC CRC EOF

IDE
r0
RTR
SOF
ACK
Data Frame or Inter-Frame Space
Remote Frame Error Frame or Overload Frame Notes:
Error Flag Echo Error Delimiter 0 <= N <= 8
Flag
6 6 8 SOF = Start Of Frame
ID = Identifier
RTR = Remote Transmission Request
Data Frame or IDE = Identifier Extension Bit
Any Frame Inter-Frame Space Remote Frame r0 = Reserved Bit
Suspend D LC = Data Length Code
Intermission Transmission Bus Idle
3 C RC = Cyclic Redundancy Code
8
Error flag: 6 dominant bits if node is error
End Of Frame or active else 6 recessive bits.

Error Delimiter or Inter-Frame Space Suspend transmission: applies to error
Overload Delimiter Overload Frame or Error Frame passive nodes only.
Overload Overload Overload EOF = End of Frame
Flag Echo Delimiter
6 6 8 ACK = Acknowledge bit
C trl = Control
ai15154
32.8 bxCAN interrupts

Four interrupt vectors are dedicated to bxCAN. Each interrupt source can be independently
enabled or disabled by means of the CAN Interrupt Enable register (CAN_IER).
RM0090 Rev 19 1095/1751

1121
Figure 348. Event flags and interrupt generation

CAN_IER
TRANSMIT
INTERRUPT
TMEIE
CAN_TSR
RQCP0
RQCP1 + &
RQCP2
FMP0
FMPIE0
& FIFO 0
INTERRUPT
CAN_RF0R FULL0
FFIE0
& +
FOVR0
FOVIE0
&
FMP1
FMPIE1
& FIFO 1
INTERRUPT
CAN_RF1R FULL1
FFIE1
& +
FOVR1
FOVIE1
&
ERRIE
EWGIE
EWGF &
EPVIE
CAN_ESR EPVF & &
BOFIE
+
BOFF & ERRI
CAN_MSR STATUS CHANGE
ERROR
LECIE
1≤LEC≤6 & INTERRUPT
WKUI
WKUIE
&
CAN_MSR
SLAKI
SLKIE
&
• The transmit interrupt can be generated by the following events:

– Transmit mailbox 0 becomes empty, RQCP0 bit in the CAN_TSR register set.
• The FIFO 0 interrupt can be generated by the following events:
– Reception of a new message, FMP0 bits in the CAN_RF0R register are not ‘00’.
– FIFO0 full condition, FULL0 bit in the CAN_RF0R register set.
– FIFO0 overrun condition, FOVR0 bit in the CAN_RF0R register set.
• The FIFO 1 interrupt can be generated by the following events:
– Reception of a new message, FMP1 bits in the CAN_RF1R register are not ‘00’.
– FIFO1 full condition, FULL1 bit in the CAN_RF1R register set.
– FIFO1 overrun condition, FOVR1 bit in the CAN_RF1R register set.
• The error and status change interrupt can be generated by the following events:
– Error condition, for more details on error conditions refer to the CAN Error Status
register (CAN_ESR).
1096/1751 RM0090 Rev 19

– Wakeup condition, SOF monitored on the CAN Rx signal.

– Entry into Sleep mode.
32.9 CAN registers

32.9.1 Register access protection

Erroneous access to certain configuration registers can cause the hardware to temporarily
disturb the whole CAN network. Therefore the CAN_BTR register can be modified by
software only while the CAN hardware is in initialization mode.
Although the transmission of incorrect data will not cause problems at the CAN network
level, it can severely disturb the application. A transmit mailbox can be only modified by
software while it is in empty state, refer to Figure 340.
The filter values can be modified either deactivating the associated filter banks or by setting
the FINIT bit. Moreover, the modification of the filter configuration (scale, mode and FIFO
assignment) in CAN_FMxR, CAN_FSxR and CAN_FFAR registers can only be done when
the filter initialization mode is set (FINIT=1) in the CAN_FMR register.
32.9.2 CAN control and status registers

CAN master control register (CAN_MCR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DBF
Reserved
rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RESET TTCM ABOM AWUM NART RFLM TXFP SLEEP INRQ
Reserved
rs rw rw rw rw rw rw rw rw

Bit 16 DBF: Debug freeze
0: CAN working during debug
1: CAN reception/transmission frozen during debug. Reception FIFOs can still be
accessed/controlled normally.
Bit 15 RESET: bxCAN software master reset
0: Normal operation.
1: Force a master reset of the bxCAN -> Sleep mode activated after reset (FMP bits and
CAN_MCR register are initialized to the reset values). This bit is automatically reset to 0.
RM0090 Rev 19 1097/1751

1121
Bit 7 TTCM: Time triggered communication mode

0: Time Triggered Communication mode disabled.
1: Time Triggered Communication mode enabled
Note: For more information on Time Triggered Communication mode refer to Section 32.7.2.
Bit 6 ABOM: Automatic bus-off management
This bit controls the behavior of the CAN hardware on leaving the Bus-Off state.
0: The Bus-Off state is left on software request, once 128 occurrences of 11 recessive bits
have been monitored and the software has first set and cleared the INRQ bit of the
CAN_MCR register.
1: The Bus-Off state is left automatically by hardware once 128 occurrences of 11 recessive
bits have been monitored.
For detailed information on the Bus-Off state refer to Section 32.7.6.
Bit 5 AWUM: Automatic wakeup mode
This bit controls the behavior of the CAN hardware on message reception during Sleep
mode.
0: The Sleep mode is left on software request by clearing the SLEEP bit of the CAN_MCR
register.
1: The Sleep mode is left automatically by hardware on CAN message detection.
The SLEEP bit of the CAN_MCR register and the SLAK bit of the CAN_MSR register are
cleared by hardware.
Bit 4 NART: No automatic retransmission
0: The CAN hardware will automatically retransmit the message until it has been
successfully transmitted according to the CAN standard.
1: A message will be transmitted only once, independently of the transmission result
(successful, error or arbitration lost).
Bit 3 RFLM: Receive FIFO locked mode
0: Receive FIFO not locked on overrun. Once a receive FIFO is full the next incoming
message will overwrite the previous one.
1: Receive FIFO locked against overrun. Once a receive FIFO is full the next incoming
message will be discarded.
Bit 2 TXFP: Transmit FIFO priority
This bit controls the transmission order when several mailboxes are pending at the same
time.
0: Priority driven by the identifier of the message
1: Priority driven by the request order (chronologically)
Bit 1 SLEEP: Sleep mode request
This bit is set by software to request the CAN hardware to enter the Sleep mode. Sleep
mode will be entered as soon as the current CAN activity (transmission or reception of a
CAN frame) has been completed.
This bit is cleared by software to exit Sleep mode.
This bit is cleared by hardware when the AWUM bit is set and a SOF bit is detected on the
CAN Rx signal.
This bit is set after reset - CAN starts in Sleep mode.
1098/1751 RM0090 Rev 19

Bit 0 INRQ: Initialization request

The software clears this bit to switch the hardware into normal mode. Once 11 consecutive
recessive bits have been monitored on the Rx signal the CAN hardware is synchronized and
ready for transmission and reception. Hardware signals this event by clearing the INAK bit in
the CAN_MSR register.
Software sets this bit to request the CAN hardware to enter initialization mode. Once
software has set the INRQ bit, the CAN hardware waits until the current CAN activity
(transmission or reception) is completed before entering the initialization mode. Hardware
signals this event by setting the INAK bit in the CAN_MSR register.
CAN master status register (CAN_MSR)

Reset value: 0x0000 0C02
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RX SAMP RXM TXM SLAKI WKUI ERRI SLAK INAK
Reserved. Reserved
r r r r rc_w1 rc_w1 rc_w1 r r

Bit 11 RX: CAN Rx signal
Monitors the actual value of the CAN_RX Pin.
Bit 10 SAMP: Last sample point
The value of RX on the last sample point (current received bit value).
Bit 9 RXM: Receive mode
The CAN hardware is currently receiver.
Bit 8 TXM: Transmit mode
The CAN hardware is currently transmitter.
Bit 4 SLAKI: Sleep acknowledge interrupt
When SLKIE=1, this bit is set by hardware to signal that the bxCAN has entered Sleep
Mode. When set, this bit generates a status change interrupt if the SLKIE bit in the
CAN_IER register is set.
This bit is cleared by software or by hardware, when SLAK is cleared.
Note: When SLKIE=0, no polling on SLAKI is possible. In this case the SLAK bit can be
polled.
Bit 3 WKUI: Wakeup interrupt
This bit is set by hardware to signal that a SOF bit has been detected while the CAN
hardware was in Sleep mode. Setting this bit generates a status change interrupt if the
WKUIE bit in the CAN_IER register is set.
This bit is cleared by software.
RM0090 Rev 19 1099/1751

1121
Bit 2 ERRI: Error interrupt

This bit is set by hardware when a bit of the CAN_ESR has been set on error detection and
the corresponding interrupt in the CAN_IER is enabled. Setting this bit generates a status
change interrupt if the ERRIE bit in the CAN_IER register is set.
Bit 1 SLAK: Sleep acknowledge
This bit is set by hardware and indicates to the software that the CAN hardware is now in
Sleep mode. This bit acknowledges the Sleep mode request from the software (set SLEEP
bit in CAN_MCR register).
This bit is cleared by hardware when the CAN hardware has left Sleep mode (to be
synchronized on the CAN bus). To be synchronized the hardware has to monitor a
sequence of 11 consecutive recessive bits on the CAN RX signal.
Note: The process of leaving Sleep mode is triggered when the SLEEP bit in the CAN_MCR
register is cleared. Refer to the AWUM bit of the CAN_MCR register description for
detailed information for clearing SLEEP bit
Bit 0 INAK: Initialization acknowledge
This bit is set by hardware and indicates to the software that the CAN hardware is now in
initialization mode. This bit acknowledges the initialization request from the software (set
INRQ bit in CAN_MCR register).
This bit is cleared by hardware when the CAN hardware has left the initialization mode (to
be synchronized on the CAN bus). To be synchronized the hardware has to monitor a
sequence of 11 consecutive recessive bits on the CAN RX signal.
CAN transmit status register (CAN_TSR)

Reset value: 0x1C00 0000
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
LOW2 LOW1 LOW0 TME2 TME1 TME0 CODE[1:0] ABRQ2 TERR2 ALST2 TXOK2 RQCP2
Reserved
r r r r r r r r rs rc_w1 rc_w1 rc_w1 rc_w1
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ABRQ1 Reserved TERR1 ALST1 TXOK1 RQCP1 ABRQ0 TERR0 ALST0 TXOK0 RQCP0
Reserved
rs Res. rc_w1 rc_w1 rc_w1 rc_w1 rs rc_w1 rc_w1 rc_w1 rc_w1
Bit 31 LOW2: Lowest priority flag for mailbox 2

This bit is set by hardware when more than one mailbox are pending for transmission and
mailbox 2 has the lowest priority.
Note: The LOW[2:0] bits are set to zero when only one mailbox is pending.
Bit 28 TME2: Transmit mailbox 2 empty
This bit is set by hardware when no transmit request is pending for mailbox 2.
1100/1751 RM0090 Rev 19


Bits 25:24 CODE[1:0]: Mailbox code
In case at least one transmit mailbox is free, the code value is equal to the number of the
next transmit mailbox free.
In case all transmit mailboxes are pending, the code value is equal to the number of the
transmit mailbox with the lowest priority.
Bit 23 ABRQ2: Abort request for mailbox 2
Set by software to abort the transmission request for the corresponding mailbox.
Cleared by hardware when the mailbox becomes empty.
Setting this bit has no effect when the mailbox is not pending for transmission.
Bit 19 TERR2: Transmission error of mailbox 2
This bit is set when the previous TX failed due to an error.
Bit 18 ALST2: Arbitration lost for mailbox 2
This bit is set when the previous TX failed due to an arbitration lost.
Bit 17 TXOK2: Transmission OK of mailbox 2
The hardware updates this bit after each transmission attempt.
0: The previous transmission failed
1: The previous transmission was successful
This bit is set by hardware when the transmission request on mailbox 2 has been completed
successfully. Refer to Figure 340.
Bit 16 RQCP2: Request completed mailbox2
Set by hardware when the last request (transmit or abort) has been performed.
Cleared by software writing a “1” or by hardware on transmission request (TXRQ2 set in
CAN_TMID2R register).
Clearing this bit clears all the status bits (TXOK2, ALST2 and TERR2) for Mailbox 2.
Bit 15 ABRQ1: Abort request for mailbox 1
Bit 11 TERR1: Transmission error of mailbox1
Bit 10 ALST1: Arbitration lost for mailbox1
Bit 9 TXOK1: Transmission OK of mailbox1
successfully. Refer to Figure 340
CAN_TI1R register).
RM0090 Rev 19 1101/1751

1121
Bit 7 ABRQ0: Abort request for mailbox0

Bit 3 TERR0: Transmission error of mailbox0
Bit 2 ALST0: Arbitration lost for mailbox0
Bit 1 TXOK0: Transmission OK of mailbox0
successfully. Refer to Figure 340
CAN_TI0R register).
CAN receive FIFO 0 register (CAN_RF0R)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RFOM0 FOVR0 FULL0 FMP0[1:0]
Reserved Res.
rs rc_w1 rc_w1 r r

Bit 5 RFOM0: Release FIFO 0 output mailbox
Set by software to release the output mailbox of the FIFO. The output mailbox can only be
released when at least one message is pending in the FIFO. Setting this bit when the FIFO
is empty has no effect. If at least two messages are pending in the FIFO, the software has to
release the output mailbox to access the next message.
Cleared by hardware when the output mailbox has been released.
Bit 4 FOVR0: FIFO 0 overrun
This bit is set by hardware when a new message has been received and passed the filter
while the FIFO was full.
Bit 3 FULL0: FIFO 0 full
Set by hardware when three messages are stored in the FIFO.
1102/1751 RM0090 Rev 19

Bits 1:0 FMP0[1:0]: FIFO 0 message pending

These bits indicate how many messages are pending in the receive FIFO.
FMP is increased each time the hardware stores a new message in to the FIFO. FMP is
decreased each time the software releases the output mailbox by setting the RFOM0 bit.
CAN receive FIFO 1 register (CAN_RF1R)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RFOM1 FOVR1 FULL1 FMP1[1:0]
Reserved Res.
rs rc_w1 rc_w1 r r

Bit 5 RFOM1: Release FIFO 1 output mailbox
Set by software to release the output mailbox of the FIFO. The output mailbox can only be
released when at least one message is pending in the FIFO. Setting this bit when the FIFO
is empty has no effect. If at least two messages are pending in the FIFO, the software has to
release the output mailbox to access the next message.
Cleared by hardware when the output mailbox has been released.
Bit 4 FOVR1: FIFO 1 overrun
This bit is set by hardware when a new message has been received and passed the filter
while the FIFO was full.
Bit 3 FULL1: FIFO 1 full
Set by hardware when three messages are stored in the FIFO.
Bits 1:0 FMP1[1:0]: FIFO 1 message pending
These bits indicate how many messages are pending in the receive FIFO1.
FMP1 is increased each time the hardware stores a new message in to the FIFO1. FMP is
decreased each time the software releases the output mailbox by setting the RFOM1 bit.
CAN interrupt enable register (CAN_IER)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
SLKIE WKUIE
Reserved
rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
LEC BOF EPV EWG FOV FF FMP FOV FF FMP TME
ERRIE
Reserved IE IE IE IE Res. IE1 IE1 IE1 IE0 IE0 IE0 IE
RM0090 Rev 19 1103/1751

1121

Bit 17 SLKIE: Sleep interrupt enable
0: No interrupt when SLAKI bit is set.
1: Interrupt generated when SLAKI bit is set.
Bit 16 WKUIE: Wakeup interrupt enable
0: No interrupt when WKUI is set.
1: Interrupt generated when WKUI bit is set.
0: No interrupt will be generated when an error condition is pending in the CAN_ESR.
1: An interrupt will be generation when an error condition is pending in the CAN_ESR.
Bit 11 LECIE: Last error code interrupt enable
0: ERRI bit will not be set when the error code in LEC[2:0] is set by hardware on error
detection.
1: ERRI bit will be set when the error code in LEC[2:0] is set by hardware on error detection.
Bit 10 BOFIE: Bus-off interrupt enable
0: ERRI bit will not be set when BOFF is set.
1: ERRI bit will be set when BOFF is set.
Bit 9 EPVIE: Error passive interrupt enable
0: ERRI bit will not be set when EPVF is set.
1: ERRI bit will be set when EPVF is set.
Bit 8 EWGIE: Error warning interrupt enable
0: ERRI bit will not be set when EWGF is set.
1: ERRI bit will be set when EWGF is set.
Bit 6 FOVIE1: FIFO overrun interrupt enable
0: No interrupt when FOVR is set.
1: Interrupt generation when FOVR is set.
Bit 5 FFIE1: FIFO full interrupt enable
0: No interrupt when FULL bit is set.
1: Interrupt generated when FULL bit is set.
Bit 4 FMPIE1: FIFO message pending interrupt enable
0: No interrupt generated when state of FMP[1:0] bits are not 00b.
1: Interrupt generated when state of FMP[1:0] bits are not 00b.
Bit 3 FOVIE0: FIFO overrun interrupt enable
0: No interrupt when FOVR bit is set.
1: Interrupt generated when FOVR bit is set.
1104/1751 RM0090 Rev 19

Bit 2 FFIE0: FIFO full interrupt enable

0: No interrupt when FULL bit is set.
1: Interrupt generated when FULL bit is set.
Bit 1 FMPIE0: FIFO message pending interrupt enable
0: No interrupt generated when state of FMP[1:0] bits are not 00b.
1: Interrupt generated when state of FMP[1:0] bits are not 00b.
Bit 0 TMEIE: Transmit mailbox empty interrupt enable
0: No interrupt when RQCPx bit is set.
1: Interrupt generated when RQCPx bit is set.
Note: Refer to Section 32.8: bxCAN interrupts.
CAN error status register (CAN_ESR)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
REC[7:0] TEC[7:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
LEC[2:0] BOFF EPVF EWGF
Reserved Res.
rw rw rw r r r
Bits 31:24 REC[7:0]: Receive error counter

The implementing part of the fault confinement mechanism of the CAN protocol. In case of
an error during reception, this counter is incremented by 1 or by 8 depending on the error
condition as defined by the CAN standard. After every successful reception the counter is
decremented by 1 or reset to 120 if its value was higher than 128. When the counter value
exceeds 127, the CAN controller enters the error passive state.
Bits 23:16 TEC[7:0]: Least significant byte of the 9-bit transmit error counter
The implementing part of the fault confinement mechanism of the CAN protocol.
Bits 6:4 LEC[2:0]: Last error code
This field is set by hardware and holds a code which indicates the error condition of the last
error detected on the CAN bus. If a message has been transferred (reception or
transmission) without error, this field will be cleared to ‘0’.
The LEC[2:0] bits can be set to value 0b111 by software. They are updated by hardware to
indicate the current communication status.
000: No Error
001: Stuff Error
010: Form Error
011: Acknowledgment Error
100: Bit recessive Error
101: Bit dominant Error
110: CRC Error
111: Set by software
RM0090 Rev 19 1105/1751

1121
Bit 2 BOFF: Bus-off flag

This bit is set by hardware when it enters the bus-off state. The bus-off state is entered on
TEC overflow, greater than 255, refer to Section 32.7.6.
Bit 1 EPVF: Error passive flag
This bit is set by hardware when the Error Passive limit has been reached (Receive Error
Counter or Transmit Error Counter>127).
Bit 0 EWGF: Error warning flag
This bit is set by hardware when the warning limit has been reached
(Receive Error Counter or Transmit Error Counter≥96).
CAN bit timing register (CAN_BTR)

This register can only be accessed by the software when the CAN hardware is in
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
SILM LBKM SJW[1:0] Res. TS2[2:0] TS1[3:0]
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
BRP[9:0]
Reserved
Bit 31 SILM: Silent mode (debug)

0: Normal operation
1: Silent Mode
Bit 30 LBKM: Loop back mode (debug)
0: Loop Back Mode disabled
1: Loop Back Mode enabled
Bits 25:24 SJW[1:0]: Resynchronization jump width
These bits define the maximum number of time quanta the CAN hardware is allowed to
lengthen or shorten a bit to perform the resynchronization.
tRJW = tq x (SJW[1:0] + 1)
Bits 22:20 TS2[2:0]: Time segment 2
These bits define the number of time quanta in Time Segment 2.
tBS2 = tq x (TS2[2:0] + 1)
1106/1751 RM0090 Rev 19

Bits 19:16 TS1[3:0]: Time segment 1

These bits define the number of time quanta in Time Segment 1
tBS1 = tq x (TS1[3:0] + 1)
For more information on bit timing refer to Section 32.7.7.
Bits 9:0 BRP[9:0]: Baud rate prescaler
These bits define the length of a time quanta.
tq = (BRP[9:0]+1) x tPCLK
32.9.3 CAN mailbox registers

This chapter describes the registers of the transmit and receive mailboxes. Refer to
Section 32.7.5: Message storage for detailed register mapping.
Transmit and receive mailboxes have the same registers except:
• The FMI field in the CAN_RDTxR register.
• A receive mailbox is always write protected.
• A transmit mailbox is write-enabled only while empty, corresponding TME bit in the
CAN_TSR register set.
There are three TX Mailboxes and two RX Mailboxes , as shown in Figure 349. Each RX
Mailbox allows access to a 3-level depth FIFO, the access being offered only to the oldest
received message in the FIFO. Each mailbox consist of four registers.
Figure 349. RX and TX mailboxes
CAN_RI0R CAN_RI1R CAN_TI0R CAN_TI1R CAN_TI2R

CAN_RDT0R CAN_RDT1R CAN_TDT0R CAN_TDT1R CAN_TDT2R
CAN_RL0R CAN_RL1R CAN_TDL0R CAN_TDL1R CAN_TDL2R
CAN_RH0R CAN_RH1R CAN_TDH0R CAN_TDH1R CAN_TDH2R
FIFO0 FIFO1 Three Tx Mailboxes
RM0090 Rev 19 1107/1751

1121
CAN TX mailbox identifier register (CAN_TIxR) (x=0..2)

Address offsets: 0x180, 0x190, 0x1A0
Reset value: 0xXXXX XXXX (except bit 0, TXRQ = 0)
All TX registers are write protected when the mailbox is pending transmission (TMEx reset).
This register also implements the TX request control (bit 0) - reset value 0.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
STID[10:0]/EXID[28:18] EXID[17:13]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
EXID[12:0] IDE RTR TXRQ
Bits 31:21 STID[10:0]/EXID[28:18]: Standard identifier or extended identifier

The standard identifier or the MSBs of the extended identifier (depending on the IDE bit
value).
Bits 20:3 EXID[17:0]: Extended identifier
The LSBs of the extended identifier.
Bit 2 IDE: Identifier extension
This bit defines the identifier type of message in the mailbox.
0: Standard identifier.
1: Extended identifier.
Bit 1 RTR: Remote transmission request
0: Data frame
1: Remote frame
Bit 0 TXRQ: Transmit mailbox request
Set by software to request the transmission for the corresponding mailbox.
1108/1751 RM0090 Rev 19

CAN mailbox data length control and time stamp register (CAN_TDTxR)
(x=0..2)
All bits of this register are write protected when the mailbox is not in empty state.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
TIME[15:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TGT DLC[3:0]
Reserved Reserved
rw rw rw rw rw
Bits 31:16 TIME[15:0]: Message time stamp

This field contains the 16-bit timer value captured at the SOF transmission.
Bit 8 TGT: Transmit global time
This bit is active only when the hardware is in the Time Trigger Communication mode,
TTCM bit of the CAN_MCR register is set.
0: Time stamp TIME[15:0] is not sent.
1: Time stamp TIME[15:0] value is sent in the last two data bytes of the 8-byte message:
TIME[7:0] in data byte 7 and TIME[15:8] in data byte 6, replacing the data written in
CAN_TDHxR[31:16] register (DATA6[7:0] and DATA7[7:0]). DLC must be programmed as 8
in order these two bytes to be sent over the CAN bus.
Bits 3:0 DLC[3:0]: Data length code
This field defines the number of data bytes a data frame contains or a remote frame request.
A message can contain from 0 to 8 data bytes, depending on the value in the DLC field.
RM0090 Rev 19 1109/1751

1121
CAN mailbox data low register (CAN_TDLxR) (x=0..2)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DATA3[7:0] DATA2[7:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DATA1[7:0] DATA0[7:0]
Bits 31:24 DATA3[7:0]: Data byte 3

Data byte 3 of the message.
A message can contain from 0 to 8 data bytes and starts with byte 0.
CAN mailbox data high register (CAN_TDHxR) (x=0..2)

Address offsets: 0x18C, 0x19C, 0x1AC
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DATA7[7:0] DATA6[7:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DATA5[7:0] DATA4[7:0]

Note: If TGT of this message and TTCM are active, DATA7 and DATA6 will be replaced by
the TIME stamp value.
1110/1751 RM0090 Rev 19

CAN receive FIFO mailbox identifier register (CAN_RIxR) (x=0..1)

Address offsets: 0x1B0, 0x1C0
All RX registers are write protected.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
STID[10:0]/EXID[28:18] EXID[17:13]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
EXID[12:0] IDE RTR
Res.
r r r r r r r r r r r r r r r
Bits 31:21 STID[10:0]/EXID[28:18]: Standard identifier or extended identifier

The standard identifier or the MSBs of the extended identifier (depending on the IDE bit
value).
Bits 20:3 EXID[17:0]: Extended identifier
The LSBs of the extended identifier.
Bit 2 IDE: Identifier extension
This bit defines the identifier type of message in the mailbox.
0: Standard identifier.
1: Extended identifier.
Bit 1 RTR: Remote transmission request
0: Data frame
1: Remote frame
RM0090 Rev 19 1111/1751

1121
CAN receive FIFO mailbox data length control and time stamp register
(CAN_RDTxR) (x=0..1)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
TIME[15:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FMI[7:0] DLC[3:0]
Reserved
Bits 31:16 TIME[15:0]: Message time stamp

This field contains the 16-bit timer value captured at the SOF detection.
Bits 15:8 FMI[7:0]: Filter match index
This register contains the index of the filter the message stored in the mailbox passed
through. For more details on identifier filtering refer to Section 32.7.4
Bits 3:0 DLC[3:0]: Data length code
This field defines the number of data bytes a data frame contains (0 to 8). It is 0 in the case
of a remote frame request.
1112/1751 RM0090 Rev 19

CAN receive FIFO mailbox data low register (CAN_RDLxR) (x=0..1)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DATA3[7:0] DATA2[7:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DATA1[7:0] DATA0[7:0]
Bits 31:24 DATA3[7:0]: Data Byte 3

A message can contain from 0 to 8 data bytes and starts with byte 0.
CAN receive FIFO mailbox data high register (CAN_RDHxR) (x=0..1)

Address offsets: 0x1BC, 0x1CC
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DATA7[7:0] DATA6[7:0]
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DATA5[7:0] DATA4[7:0]

RM0090 Rev 19 1113/1751

1121
32.9.4 CAN filter registers

CAN filter master register (CAN_FMR)
Reset value: 0x2A1C 0E01
All bits of this register are set and cleared by software.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CAN2SB[5:0] FINIT
Reserved Reserved

Bits 13:8 CAN2SB[5:0]: CAN2 start bank
These bits are set and cleared by software. They define the start bank for the CAN2
interface (Slave) in the range 0 to 27.
Note: When CAN2SB[5:0] = 28d, all the filters to CAN1 can be used.
When CAN2SB[5:0] is set to 0, no filters are assigned to CAN1.
Bit 0 FINIT: Filter init mode
Initialization mode for filter banks
0: Active filters mode.
1: Initialization mode for the filters.
1114/1751 RM0090 Rev 19

CAN filter mode register (CAN_FM1R)

This register can be written only when the filter initialization mode is set (FINIT=1) in the
CAN_FMR register.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
FBM27 FBM26 FBM25 FBM24 FBM23 FBM22 FBM21 FBM20 FBM19 FBM18 FBM17 FBM16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FBM15 FBM14 FBM13 FBM12 FBM11 FBM10 FBM9 FBM8 FBM7 FBM6 FBM5 FBM4 FBM3 FBM2 FBM1 FBM0
Note: Refer to Figure 342.

Bits 27:0 FBMx: Filter mode
Mode of the registers of Filter x.
0: Two 32-bit registers of filter bank x are in Identifier Mask mode.
1: Two 32-bit registers of filter bank x are in Identifier List mode.
CAN filter scale register (CAN_FS1R)

CAN_FMR register.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
FSC27 FSC26 FSC25 FSC24 FSC23 FSC22 FSC21 FSC20 FSC19 FSC18 FSC17 FSC16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FSC15 FSC14 FSC13 FSC12 FSC11 FSC10 FSC9 FSC8 FSC7 FSC6 FSC5 FSC4 FSC3 FSC2 FSC1 FSC0

Bits 27:0 FSCx: Filter scale configuration
These bits define the scale configuration of Filters 13-0.
0: Dual 16-bit scale configuration
1: Single 32-bit scale configuration
Note: Refer to Figure 342.
RM0090 Rev 19 1115/1751

1121
CAN filter FIFO assignment register (CAN_FFA1R)

CAN_FMR register.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
FFA27 FFA26 FFA25 FFA24 FFA23 FFA22 FFA21 FFA20 FFA19 FFA18 FFA17 FFA16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FFA15 FFA14 FFA13 FFA12 FFA11 FFA10 FFA9 FFA8 FFA7 FFA6 FFA5 FFA4 FFA3 FFA2 FFA1 FFA0

Bits 27:0 FFAx: Filter FIFO assignment for filter x
The message passing through this filter will be stored in the specified FIFO.
0: Filter assigned to FIFO 0
1: Filter assigned to FIFO 1
CAN filter activation register (CAN_FA1R)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
FACT27 FACT26 FACT25 FACT24 FACT23 FACT22 FACT21 FACT20 FACT19 FACT18 FACT17 FACT16
Reserved
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FACT15 FACT14 FACT13 FACT12 FACT11 FACT10 FACT9 FACT8 FACT7 FACT6 FACT5 FACT4 FACT3 FACT2 FACT1 FACT0

Bits 27:0 FACTx: Filter active
The software sets this bit to activate Filter x. To modify the Filter x registers (CAN_FxR[0:7]),
the FACTx bit must be cleared or the FINIT bit of the CAN_FMR register must be set.
0: Filter x is not active
1: Filter x is active
1116/1751 RM0090 Rev 19

Filter bank i register x (CAN_FiRx) (i=0..27, x=1, 2)

Address offsets: 0x240..0x31C
There are 28 filter banks, i=0 .. 27. Each filter bank i is composed of two 32-bit registers,
CAN_FiR[2:1].
This register can only be modified when the FACTx bit of the CAN_FAxR register is cleared
or when the FINIT bit of the CAN_FMR register is set.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
FB31 FB30 FB29 FB28 FB27 FB26 FB25 FB24 FB23 FB22 FB21 FB20 FB19 FB18 FB17 FB16
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FB15 FB14 FB13 FB12 FB11 FB10 FB9 FB8 FB7 FB6 FB5 FB4 FB3 FB2 FB1 FB0
In all configurations:
Bits 31:0 FB[31:0]: Filter bits

Identifier
Each bit of the register specifies the level of the corresponding bit of the expected identifier.
0: Dominant bit is expected
1: Recessive bit is expected
Mask
Each bit of the register specifies whether the bit of the associated identifier register must
match with the corresponding bit of the expected identifier or not.
0: Don’t care, the bit is not used for the comparison
1: Must match, the bit of the incoming identifier must have the same level has specified in
the corresponding identifier register of the filter.
Note: Depending on the scale and mode configuration of the filter the function of each register can
differ. For the filter mapping, functions description and mask registers association, refer to
Section 32.7.4.
A Mask/Identifier register in mask mode has the same bit mapping as in identifier list
mode.
For the register mapping/addresses of the filter banks refer to Table 184.
RM0090 Rev 19 1117/1751

1121
32.9.5 bxCAN register map

Refer to Section 2.3: Memory map for the register boundary addresses. The registers from
offset 0x200 to 31C are present only in CAN1.
Table 184. bxCAN register map and reset values

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
RESET
SLEEP
AWUM
ABOM
TTCM
RFLM
NART
TXFP
INRQ
DBF
CAN_MCR
Reset value 1 0 0 0 0 0 0 0 1 0
SAMP
SLAKI
WKUI
SLAK
ERRI
INAK
RXM
TXM
CAN_MSR
RX
0x004 Reserved Res.
Reset value 1 1 0 0 0 0 0 1 0
CODE[1:0]
LOW[2:0]
TME[2:0]
RQCP2
RQCP1
RQCP0
ABRQ2
ABRQ1
ABRQ0
TERR2
TXOK2
TERR1
TXOK1
TERR0
TXOK0
ALST2
ALST1
ALST0
CAN_TSR
0x008 Res. Res. Res..
Reset value 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
FMP0[1:0]
RFOM0
FOVR0
FULL0
Reserved
CAN_RF0R
0x00C Reserved
FMP1[1:0]
RFOM1
FOVR1
FULL1
Reserved
CAN_RF1R
0x010 Reserved
FMPIE1
FMPIE0
FOVIE1
FOVIE0
EWGIE
WKUIE
ERRIE
TMEIE
BOFIE
EPVIE
LECIE
SLKIE
FFIE1
FFIE0
CAN_IER Reserved
0x014 Reserved Res.
Reset value 0 0 0 0 0 0 0 0 0
LEC[2:0] 0 0 0 0 0
EWGF
BOFF
EPVF
CAN_ESR REC[7:0] TEC[7:0] Reserved
0x018 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
SJW[1:0]
LBKM
Reserved
SILM
CAN_BTR TS2[2:0] TS1[3:0] BRP[9:0]

Reset value 0 0 0 0 0 1 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0
0x020-
Reserved
0x17F
TXRQ
RTR
IDE
CAN_TI0R STID[10:0]/EXID[28:18] EXID[17:0]

0x180
Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x 0
TGT
CAN_TDT0R TIME[15:0] DLC[3:0]

Reset value x x x x x x x x x x x x x x x x x x x x x
CAN_TDL0R DATA3[7:0] DATA2[7:0] DATA1[7:0] DATA0[7:0]

0x188
Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x
1118/1751 RM0090 Rev 19

Table 184. bxCAN register map and reset values (continued)

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
CAN_TDH0R DATA7[7:0] DATA6[7:0] DATA5[7:0] DATA4[7:0]
0x18C
TXRQ
RTR
IDE
0x190
TGT

0x198

0x19C
TXRQ
RTR
IDE
0x1A0
TGT
0x1A4 Reserved Reserved

0x1A8

0x1AC
Reserved
RTR
IDE
CAN_RI0R STID[10:0]/EXID[28:18] EXID[17:0]
0x1B0
Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x
CAN_RDT0R TIME[15:0] FMI[7:0] DLC[3:0]

0x1B4 Reserved
Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x
CAN_RDL0R DATA3[7:0] DATA2[7:0] DATA1[7:0] DATA0[7:0]

0x1B8
CAN_RDH0R DATA7[7:0] DATA6[7:0] DATA5[7:0] DATA4[7:0]

0x1BC
Reserved
RTR
IDE
CAN_RI1R STID[10:0]/EXID[28:18] EXID[17:0]

0x1C0
Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x
RM0090 Rev 19 1119/1751

1121

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
CAN_RDT1R TIME[15:0] FMI[7:0] DLC[3:0]
0x1C4 Reserved
Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x
CAN_RDL1R DATA3[7:0] DATA2[7:0] DATA1[7:0] DATA0[7:0]

0x1C8
CAN_RDH1R DATA7[7:0] DATA6[7:0] DATA5[7:0] DATA4[7:0]

0x1CC
0x1D0-
Reserved
0x1FF
FINIT
CAN_FMR CAN2SB[5:0]
CAN_FM1R FBM[27:0]
0x204 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x208 Reserved
CAN_FS1R FSC[27:0]
0x20C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x210 Reserved
CAN_FFA1R FFA[27:0]
0x214 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x218 Reserved
CAN_FA1R FACT[27:0]
0x21C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x220 Reserved
0x224-
Reserved
0x23F
CAN_F0R1 FB[31:0]
0x240
CAN_F0R2 FB[31:0]
0x244
CAN_F1R1 FB[31:0]
0x248
1120/1751 RM0090 Rev 19


Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
CAN_F1R2 FB[31:0]
0x24C
. . .
. . .
. . .
. . .
CAN_F27R1 FB[31:0]
0x318
CAN_F27R2 FB[31:0]
0x31C
RM0090 Rev 19 1121/1751

1121
Ethernet (ETH): media access control (MAC) with DMA controller RM0090
33 Ethernet (ETH): media access control (MAC) with

DMA controller
This section applies only to STM32F42xx/F43xx and STM32F407x/F417x devices.
33.1 Ethernet introduction

Portions Copyright (c) 2004, 2005 Synopsys, Inc. All rights reserved. Used with permission.
The Ethernet peripheral enables the STM32F4xx to transmit and receive data over Ethernet
in compliance with the IEEE 802.3-2002 standard.
The Ethernet provides a configurable, flexible peripheral to meet the needs of various
applications and customers. It supports two industry standard interfaces to the external
physical layer (PHY): the default media independent interface (MII) defined in the IEEE
802.3 specifications and the reduced media independent interface (RMII). It can be used in
number of applications such as switches, network interface cards, etc.
The Ethernet is compliant with the following standards:
• IEEE 802.3-2002 for Ethernet MAC
• IEEE 1588-2008 standard for precision networked clock synchronization
• AMBA 2.0 for AHB Master/Slave ports
• RMII specification from RMII consortium
33.2 Ethernet main features

The Ethernet (ETH) peripheral includes the following features, listed by category:
1122/1751 RM0090 Rev 19

RM0090 Ethernet (ETH): media access control (MAC) with DMA controller
33.2.1 MAC core features

• Supports 10/100 Mbit/s data transfer rates with external PHY interfaces
• IEEE 802.3-compliant MII interface to communicate with an external Fast Ethernet
PHY
• Supports both full-duplex and half-duplex operations
– Supports CSMA/CD Protocol for half-duplex operation
– Supports IEEE 802.3x flow control for full-duplex operation
– Optional forwarding of received pause control frames to the user application in full-
duplex operation
– Back-pressure support for half-duplex operation
– Automatic transmission of zero-quanta pause frame on deassertion of flow control
input in full-duplex operation
• Preamble and start-of-frame data (SFD) insertion in Transmit, and deletion in Receive
paths
• Automatic CRC and pad generation controllable on a per-frame basis
• Options for automatic pad/CRC stripping on receive frames
• Programmable frame length to support Standard frames with sizes up to 16 KB
• Programmable interframe gap (40-96 bit times in steps of 8)
• Supports a variety of flexible address filtering modes:
– Up to four 48-bit perfect (DA) address filters with masks for each byte
– Up to three 48-bit SA address comparison check with masks for each byte
– 64-bit Hash filter (optional) for multicast and unicast (DA) addresses
– Option to pass all multicast addressed frames
– Promiscuous mode support to pass all frames without any filtering for network
monitoring
– Passes all incoming packets (as per filter) with a status report
• Separate 32-bit status returned for transmission and reception packets
• Supports IEEE 802.1Q VLAN tag detection for reception frames
• Separate transmission, reception, and control interfaces to the Application
• Supports mandatory network statistics with RMON/MIB counters (RFC2819/RFC2665)
• MDIO interface for PHY device configuration and management
• Detection of LAN wakeup frames and AMD Magic Packet™ frames
• Receive feature for checksum off-load for received IPv4 and TCP packets
encapsulated by the Ethernet frame
• Enhanced receive feature for checking IPv4 header checksum and TCP, UDP, or ICMP
checksum encapsulated in IPv4 or IPv6 datagrams
• Support Ethernet frame time stamping as described in IEEE 1588-2008. Sixty-four-bit
time stamps are given in each frame’s transmit or receive status
• Two sets of FIFOs: a 2-KB Transmit FIFO with programmable threshold capability, and
a 2-KB Receive FIFO with a configurable threshold (default of 64 bytes)
• Receive Status vectors inserted into the Receive FIFO after the EOF transfer enables
multiple-frame storage in the Receive FIFO without requiring another FIFO to store
those frames’ Receive Status
• Option to filter all error frames on reception and not forward them to the application in
RM0090 Rev 19 1123/1751

1239
Store-and-Forward mode
• Option to forward under-sized good frames
• Supports statistics by generating pulses for frames dropped or corrupted (due to
overflow) in the Receive FIFO
• Supports Store and Forward mechanism for transmission to the MAC core
• Automatic generation of PAUSE frame control or back pressure signal to the MAC core
based on Receive FIFO-fill (threshold configurable) level
• Handles automatic retransmission of Collision frames for transmission
• Discards frames on late collision, excessive collisions, excessive deferral and underrun
conditions
• Software control to flush Tx FIFO
• Calculates and inserts IPv4 header checksum and TCP, UDP, or ICMP checksum in
frames transmitted in Store-and-Forward mode
• Supports internal loopback on the MII for debugging
33.2.2 DMA features

• Supports all AHB burst types in the AHB Slave Interface
• Software can select the type of AHB burst (fixed or indefinite burst) in the AHB Master
interface.
• Option to select address-aligned bursts from AHB master port
• Optimization for packet-oriented DMA transfers with frame delimiters
• Byte-aligned addressing for data buffer support
• Dual-buffer (ring) or linked-list (chained) descriptor chaining
• Descriptor architecture, allowing large blocks of data transfer with minimum CPU
intervention;
• each descriptor can transfer up to 8 KB of data
• Comprehensive status reporting for normal operation and transfers with errors
• Individual programmable burst size for Transmit and Receive DMA Engines for optimal
host bus utilization
• Programmable interrupt options for different operational conditions
• Per-frame Transmit/Receive complete interrupt control
• Round-robin or fixed-priority arbitration between Receive and Transmit engines
• Start/Stop modes
• Current Tx/Rx buffer pointer as status registers
• Current Tx/Rx descriptor pointer as status registers
33.2.3 PTP features

• Received and transmitted frames time stamping
• Coarse and fine correction methods
• Trigger interrupt when system time becomes greater than target time
• Pulse per second output (product alternate function output)
1124/1751 RM0090 Rev 19

33.3 Ethernet pins

Table 185 shows the MAC signals and the corresponding MII/RMII signal mapping. All MAC
signals are mapped onto AF11, some signals are mapped onto different I/O pins, and
should be configured in Alternate function mode (for more details, refer to Section 8.3.2: I/O
pin multiplexer and mapping).
Table 185. Alternate function mapping

AF11
Port
ETH
PA0-WKUP ETH_MII_CRS
PA1 ETH_MII _RX_CLK / ETH_RMII _REF_CLK
PA2 ETH _MDIO
PA3 ETH _MII_COL
PA7 ETH_MII _RX_DV / ETH_RMII _CRS_DV
PB0 ETH _MII_RXD2
PB1 ETH _MII_RXD3
PB5 ETH _PPS_OUT
PB8 ETH _MII_TXD3
PB10 ETH_ MII_RX_ER
PB11 ETH _MII_TX_EN / ETH _RMII_TX_EN
PB12 ETH _MII_TXD0 / ETH _RMII_TXD0
PB13 ETH _MII_TXD1 / ETH _RMII_TXD1
PC1 ETH _MDC
PC2 ETH _MII_TXD2
PC3 ETH _MII_TX_CLK
PC4 ETH_MII_RXD0 / ETH_RMII_RXD0
PC5 ETH _MII_RXD1/ ETH _RMII_RXD1
PE2 ETH_MII_TXD3
PG8 ETH_PPS_OUT
PG11 ETH _MII_TX_EN / ETH _RMII_TX_EN
PG13 ETH _MII_TXD0 / ETH _RMII_TXD0
PG14 ETH _MII_TXD1 / ETH _RMII_TXD1
PH2 ETH _MII_CRS
PH3 ETH _MII_COL
PH6 ETH _MII_RXD2
PH7 ETH _MII_RXD3
PI10 ETH _MII_RX_ER
RM0090 Rev 19 1125/1751

1239
33.4 Ethernet functional description: SMI, MII and RMII

The Ethernet peripheral consists of a MAC 802.3 (media access control) with a dedicated
DMA controller. It supports both default media-independent interface (MII) and reduced
media-independent interface (RMII) through one selection bit (refer to SYSCFG_PMC
register).
The DMA controller interfaces with the Core and memories through the AHB Master and
Slave interfaces. The AHB Master Interface controls data transfers while the AHB Slave
interface accesses Control and Status registers (CSR) space.
The Transmit FIFO (Tx FIFO) buffers data read from system memory by the DMA before
transmission by the MAC Core. Similarly, the Receive FIFO (Rx FIFO) stores the Ethernet
frames received from the line until they are transferred to system memory by the DMA.
The Ethernet peripheral also includes an SMI to communicate with external PHY. A set of
configuration registers permit the user to select the desired mode and features for the MAC
and the DMA controller.
Note: The AHB clock frequency must be at least 25 MHz when the Ethernet is used.
Figure 350. ETH block diagram

AHB Slave interface
Media access RMII

DMA Operation
control & control Interface
mode MAC 802.3
status register
registers
Select
MAC
control
registers MII External PHY
2 Kbyte
Checksum PTP MDC
Bus matrix
RX FIFO
offload IEEE1588
Ethernet
DMA
2 Kbyte MDIO
TX FIFO PMT MMC
ai15620c
1. For AHB connections refer to Figure 1: System architecture for STM32F405xx/07xx and
STM32F415xx/17xx devices and Figure 2: System architecture for STM32F42xxx and STM32F43xxx
devices.
33.4.1 Station management interface: SMI

The station management interface (SMI) allows the application to access any PHY registers
through a 2-wire clock and data lines. The interface supports accessing up to 32 PHYs.
The application can select one of the 32 PHYs and one of the 32 registers within any PHY
and send control data or receive status information. Only one register in one PHY can be
addressed at any given time.
Both the MDC clock line and the MDIO data line are implemented as alternate function I/O
in the microcontroller:
• MDC: a periodic clock that provides the timing reference for the data transfer at the
maximum frequency of 2.5 MHz. The minimum high and low times for MDC must be
1126/1751 RM0090 Rev 19

160 ns each, and the minimum period for MDC must be 400 ns. In idle state the SMI
management interface drives the MDC clock signal low.
• MDIO: data input/output bitstream to transfer status information to/from the PHY device
synchronously with the MDC clock signal
Figure 351. SMI interface signals
802.3 MAC
MDC
External
STM32 MCU
MDIO PHY
ai15621b
SMI frame format

The frame structure related to a read or write operation is shown in Table 13, the order of bit
transmission must be from left to right.
Table 186. Management frame format

Management frame fields
Preamble
Start Operation PADDR RADDR TA Data (16 bits) Idle
(32 bits)
Read 1... 1 01 10 ppppp rrrrr Z0 ddddddddddddddd Z

Write 1... 1 01 01 ppppp rrrrr 10 ddddddddddddddd Z
The management frame consists of eight fields:

• Preamble: each transaction (read or write) can be initiated with the preamble field that
corresponds to 32 contiguous logic one bits on the MDIO line with 32 corresponding
cycles on MDC. This field is used to establish synchronization with the PHY device.
• Start: the start of frame is defined by a <01> pattern to verify transitions on the line
from the default logic one state to zero and back to one.
• Operation: defines the type of transaction (read or write) in progress.
• PADDR: the PHY address is 5 bits, allowing 32 unique PHY addresses. The MSB bit of
the address is the first transmitted and received.
• RADDR: the register address is 5 bits, allowing 32 individual registers to be addressed
within the selected PHY device. The MSB bit of the address is the first transmitted and
received.
• TA: the turn-around field defines a 2-bit pattern between the RADDR and DATA fields
to avoid contention during a read transaction. For a read transaction the MAC controller
drives high-impedance on the MDIO line for the 2 bits of TA. The PHY device must
drive a high-impedance state on the first bit of TA, a zero bit on the second one.
RM0090 Rev 19 1127/1751

1239
For a write transaction, the MAC controller drives a <10> pattern during the TA field.
The PHY device must drive a high-impedance state for the 2 bits of TA.
• Data: the data field is 16-bit. The first bit transmitted and received must be bit 15 of the
ETH_MIID register.
• Idle: the MDIO line is driven in high-impedance state. All three-state drivers must be
disabled and the PHY’s pull-up resistor keeps the line at logic one.
SMI write operation

When the application sets the MII Write and Busy bits (in Ethernet MAC MII address register
(ETH_MACMIIAR)), the SMI initiates a write operation into the PHY registers by transferring
the PHY address, the register address in PHY, and the write data (in Ethernet MAC MII data
register (ETH_MACMIIDR). The application should not change the MII Address register
contents or the MII Data register while the transaction is ongoing. Write operations to the MII
Address register or the MII Data register during this period are ignored (the Busy bit is high),
and the transaction is completed without any error. After the Write operation has completed,
the SMI indicates this by resetting the Busy bit.
Figure 352 shows the frame format for the write operation.
Figure 352. MDIO timing and frame structure - Write cycle
MDC
MDIO 32 1's 0 1 0 1 A4 A3 A2 A1 A0 R4 R3 R2 R1 R0 D15 D14 D1 D0
Start
Preamble of OP PHY address Register address Turn data
code around
frame
Data to PHY
ai15626
SMI read operation

When the user sets the MII Busy bit in the Ethernet MAC MII address register
(ETH_MACMIIAR) with the MII Write bit at 0, the SMI initiates a read operation in the PHY
registers by transferring the PHY address and the register address in PHY. The application
should not change the MII Address register contents or the MII Data register while the
transaction is ongoing. Write operations to the MII Address register or MII Data register
during this period are ignored (the Busy bit is high) and the transaction is completed without
any error. After the read operation has completed, the SMI resets the Busy bit and then
updates the MII Data register with the data read from the PHY.
Figure 353 shows the frame format for the read operation.
1128/1751 RM0090 Rev 19

Figure 353. MDIO timing and frame structure - Read cycle
MDC
MDIO 32 1's 0 1 1 0 A4 A3 A2 A1 A0 R4 R3 R2 R1 R0 D15 D14 D1 D0
Start
Preamble of OP PHY address Register address Turn data
code around
frame
Data to PHY Data from PHY ai15627
SMI clock selection

The MAC initiates the Management Write/Read operation. The SMI clock is a divided clock
whose source is the application clock (AHB clock). The divide factor depends on the clock
range setting in the MII Address register.
Table 187 shows how to set the clock ranges.
Table 187. Clock range

Selection HCLK clock MDC clock
000 60-100 MHz AHB clock / 42

001 100-150 MHz AHB clock / 62
010 20-35 MHz AHB clock / 16
011 35-60 MHz AHB clock / 26
100 150-180 MHz AHB clock / 102
101, 110, 111 Reserved -
33.4.2 Media-independent interface: MII

The media-independent interface (MII) defines the interconnection between the MAC
sublayer and the PHY for data transfer at 10 Mbit/s and 100 Mbit/s.
Figure 354. Media independent interface signals
TX_CLK
TXD[3:0]
TX_EN
RX_CLK
802.3 MAC
RXD[3:0]
STM32 MCU RX_ER External
RX_DV PHY
CRS
COL
MDC
MDIO
ai15622c
RM0090 Rev 19 1129/1751

1239
• MII_TX_CLK: continuous clock that provides the timing reference for the TX data
transfer. The nominal frequency is: 2.5 MHz at 10 Mbit/s speed; 25 MHz at 100 Mbit/s
speed.
• MII_RX_CLK: continuous clock that provides the timing reference for the RX data
transfer. The nominal frequency is: 2.5 MHz at 10 Mbit/s speed; 25 MHz at 100 Mbit/s
speed.
• MII_TX_EN: transmission enable indicates that the MAC is presenting nibbles on the
MII for transmission. It must be asserted synchronously (MII_TX_CLK) with the first
nibble of the preamble and must remain asserted while all nibbles to be transmitted are
presented to the MII.
• MII_TXD[3:0]: transmit data is a bundle of 4 data signals driven synchronously by the
MAC sublayer and qualified (valid data) on the assertion of the MII_TX_EN signal.
MII_TXD[0] is the least significant bit, MII_TXD[3] is the most significant bit. While
MII_TX_EN is deasserted the transmit data must have no effect upon the PHY.
• MII_CRS: carrier sense is asserted by the PHY when either the transmit or receive
medium is non idle. It shall be deasserted by the PHY when both the transmit and
receive media are idle. The PHY must ensure that the MII_CS signal remains asserted
throughout the duration of a collision condition. This signal is not required to transition
synchronously with respect to the TX and RX clocks. In full duplex mode the state of
this signal is don’t care for the MAC sublayer.
• MII_COL: collision detection must be asserted by the PHY upon detection of a collision
on the medium and must remain asserted while the collision condition persists. This
signal is not required to transition synchronously with respect to the TX and RX clocks.
In full duplex mode the state of this signal is don’t care for the MAC sublayer.
• MII_RXD[3:0]: reception data is a bundle of 4 data signals driven synchronously by the
PHY and qualified (valid data) on the assertion of the MII_RX_DV signal. MII_RXD[0] is
the least significant bit, MII_RXD[3] is the most significant bit. While MII_RX_EN is
deasserted and MII_RX_ER is asserted, a specific MII_RXD[3:0] value is used to
transfer specific information from the PHY (see Table 189).
• MII_RX_DV: receive data valid indicates that the PHY is presenting recovered and
decoded nibbles on the MII for reception. It must be asserted synchronously
(MII_RX_CLK) with the first recovered nibble of the frame and must remain asserted
through the final recovered nibble. It must be deasserted prior to the first clock cycle
that follows the final nibble. In order to receive the frame correctly, the MII_RX_DV
signal must encompass the frame, starting no later than the SFD field.
• MII_RX_ER: receive error must be asserted for one or more clock periods
(MII_RX_CLK) to indicate to the MAC sublayer that an error was detected somewhere
in the frame. This error condition must be qualified by MII_RX_DV assertion as
described in Table 189.
Table 188. TX interface signal encoding

MII_TX_EN MII_TXD[3:0] Description
0 0000 through 1111 Normal inter-frame

1 0000 through 1111 Normal data transmission
1130/1751 RM0090 Rev 19

Table 189. RX interface signal encoding

MII_RX_DV MII_RX_ERR MII_RXD[3:0] Description
0 0 0000 through 1111 Normal inter-frame

0 1 0000 Normal inter-frame
0 1 0001 through 1101 Reserved
0 1 1110 False carrier indication
0 1 1111 Reserved
1 0 0000 through 1111 Normal data reception
1 1 0000 through 1111 Data reception with errors
MII clock sources

To generate both TX_CLK and RX_CLK clock signals, the external PHY must be clocked
with an external 25 MHz as shown in Figure 355. Instead of using an external 25 MHz
quartz to provide this clock, the STM32F4xxmicrocontroller can output this signal on its
MCO pin. In this case, the PLL multiplier has to be configured so as to get the desired
frequency on the MCO pin, from the 25 MHz external quartz.
Figure 355. MII clock sources

802.3 MAC
STM32 External
TX _CLK
MCU PHY
25 MHz
RX _CLK
HSE
MCO
25 MHz For 10/100 Mbit/s
ai15623b
33.4.3 Reduced media-independent interface: RMII

The reduced media-independent interface (RMII) specification reduces the pin count
between the microcontroller Ethernet peripheral and the external Ethernet in 10/100 Mbit/s.
According to the IEEE 802.3u standard, an MII contains 16 pins for data and control. The
RMII specification is dedicated to reduce the pin count to 7 pins (a 62.5% decrease in pin
count).
RM0090 Rev 19 1131/1751

1239
The RMII is instantiated between the MAC and the PHY. This helps translation of the MAC’s
MII into the RMII. The RMII block has the following characteristics:
• It supports 10-Mbit/s and 100-Mbit/s operating rates
• The clock reference must be doubled to 50 MHz
• The same clock reference must be sourced externally to both MAC and external
Ethernet PHY
• It provides independent 2-bit wide (dibit) transmit and receive data paths
Figure 356. Reduced media-independent interface signals
TXD[1:0]
TX_EN
802.3 MAC
RXD[1:0]
STM32 External
MCU CRS_
DV PHY
MDC
MDIO
REF
_ CLK
Clock source
ai15624b
RMII clock sources

Either clock the PHY from an external 50 MHz clock or use a PHY with an embedded PLL to
generate the 50 MHz frequency.
Figure 357. RMII clock sources

802.3 MAC
STM32
MCU External
PHY
25 MHz
PLL
REF_CLK
50 MHz
For 10/100 Mbit/s
MS19930V1
33.4.4 MII/RMII selection

The mode, MII or RMII, is selected using the configuration bit 23, MII_RMII_SEL, in the
SYSCFG_PMC register. The application has to set the MII/RMII mode while the Ethernet
controller is under reset or before enabling the clocks.
1132/1751 RM0090 Rev 19

MII/RMII internal clock scheme

The clock scheme required to support both the MII and RMII, as well as 10 and 100 Mbit/s
operations is described in Figure 358.
Figure 358. Clock scheme
MAC
MII_TX_CLK as AF GPIO and AF 25 MHz or 2.5 MHz

0 25 MHz MACTXCLK TX
(25 MHz or 2.5 MHz) controller
1 or 2.5 MHz
0 MII
50 MHz Sync. divider
/2 for 100 Mb/s 1 RMII(1)
/20 for 10 Mb/s
MII_RX_CLK as AF 0 25 MHz MACRXCLK AHB
(25 MHz or 2.5 MHz) GPIO and AF 25 MHz or 2.5 MHz RX
1 or 2.5 MHz
controller
RMII_REF_CK as AF
(50 MHz)
RMII
HCLK HCLK
must be greater
than 25 MHz
ai15650
1. The MII/RMII selection is controlled through bit 23, MII_RMII_SEL, in the SYSCFG_PMC register.
To save a pin, the two input clock signals, RMII_REF_CK and MII_RX_CLK, are multiplexed
on the same GPIO pin.
33.5 Ethernet functional description: MAC 802.3

The IEEE 802.3 International Standard for local area networks (LANs) employs the
CSMA/CD (carrier sense multiple access with collision detection) as the access method.
The Ethernet peripheral consists of a MAC 802.3 (media access control) controller with
media independent interface (MII) and a dedicated DMA controller.
The MAC block implements the LAN CSMA/CD sublayer for the following families of
systems: 10 Mbit/s and 100 Mbit/s of data rates for baseband and broadband systems. Half-
and full-duplex operation modes are supported. The collision detection access method is
applied only to the half-duplex operation mode. The MAC control frame sublayer is
supported.
The MAC sublayer performs the following functions associated with a data link control
procedure:
• Data encapsulation (transmit and receive)
– Framing (frame boundary delimitation, frame synchronization)
– Addressing (handling of source and destination addresses)
– Error detection
• Media access management
– Medium allocation (collision avoidance)
– Contention resolution (collision handling)
RM0090 Rev 19 1133/1751

1239
Basically there are two operating modes of the MAC sublayer:

• Half-duplex mode: the stations contend for the use of the physical medium, using the
CSMA/CD algorithms.
• Full duplex mode: simultaneous transmission and reception without contention
resolution (CSMA/CD algorithm are unnecessary) when all the following conditions are
met:
– physical medium capability to support simultaneous transmission and reception
– exactly 2 stations connected to the LAN
– both stations configured for full-duplex operation
33.5.1 MAC 802.3 frame format

The MAC block implements the MAC sublayer and the optional MAC control sublayer
(10/100 Mbit/s) as specified by the IEEE 802.3-2002 standard.
Two frame formats are specified for data communication systems using the CSMA/CD
MAC:
• Basic MAC frame format
• Tagged MAC frame format (extension of the basic MAC frame format)
Figure 360 and Figure 361 describe the frame structure (untagged and tagged) that
includes the following fields:
• Preamble: 7-byte field used for synchronization purposes (PLS circuitry)
Hexadecimal value: 55-55-55-55-55-55-55
Bit pattern: 01010101 01010101 01010101 01010101 01010101 01010101 01010101
(right-to-left bit transmission)
• Start frame delimiter (SFD): 1-byte field used to indicate the start of a frame.
Hexadecimal value: D5
Bit pattern: 11010101 (right-to-left bit transmission)
• Destination and Source Address fields: 6-byte fields to indicate the destination and
source station addresses as follows (see Figure 359):
– Each address is 48 bits in length
– The first LSB bit (I/G) in the destination address field is used to indicate an
individual (I/G = 0) or a group address (I/G = 1). A group address could identify
none, one or more, or all the stations connected to the LAN. In the source address
the first bit is reserved and reset to 0.
– The second bit (U/L) distinguishes between locally (U/L = 1) or globally (U/L = 0)
administered addresses. For broadcast addresses this bit is also 1.
– Each byte of each address field must be transmitted least significant bit first.
The address designation is based on the following types:
• Individual address: this is the physical address associated with a particular station on
the network.
• Group address. A multidestination address associated with one or more stations on a
given network. There are two kinds of multicast address:
– Multicast-group address: an address associated with a group of logically related
stations.
– Broadcast address: a distinguished, predefined multicast address (all 1’s in the
destination address field) that always denotes all the stations on a given LAN.
1134/1751 RM0090 Rev 19

Figure 359. Address field format
I/G = 0 Individual address

I/G = 1 Group address
MSB 46-bit address U/L I/G LSB
U/L = 0 Globally administered address
U/L = 1 Locally administered address
Bit transmission order (right to left)
ai15628
• QTag Prefix: 4-byte field inserted between the Source address field and the MAC Client
Length/Type field. This field is an extension of the basic frame (untagged) to obtain the
tagged MAC frame. The untagged MAC frames do not include this field. The
extensions for tagging are as follows:
– 2-byte constant Length/Type field value consistent with the Type interpretation
(greater than 0x0600) equal to the value of the 802.1Q Tag Protocol Type (0x8100
hexadecimal). This constant field is used to distinguish tagged and untagged MAC
frames.
– 2-byte field containing the Tag control information field subdivided as follows: a 3-
bit user priority, a canonical format indicator (CFI) bit and a 12-bit VLAN Identifier.
The length of the tagged MAC frame is extended by 4 bytes by the QTag Prefix.
• MAC client length/type: 2-byte field with different meaning (mutually exclusive),
depending on its value:
– If the value is less than or equal to maxValidFrame (0d1500) then this field
indicates the number of MAC client data bytes contained in the subsequent data
field of the 802.3 frame (length interpretation).
– If the value is greater than or equal to MinTypeValue (0d1536 decimal, 0x0600)
then this field indicates the nature of the MAC client protocol (Type interpretation)
related to the Ethernet frame.
Regardless of the interpretation of the length/type field, if the length of the data field is
less than the minimum required for proper operation of the protocol, a PAD field is
added after the data field but prior to the FCS (frame check sequence) field. The
length/type field is transmitted and received with the higher-order byte first.
For length/type field values in the range between maxValidLength and minTypeValue
(boundaries excluded), the behavior of the MAC sublayer is not specified: they may or
may not be passed by the MAC sublayer.
• Data and PAD fields: n-byte data field. Full data transparency is provided, it means that
any arbitrary sequence of byte values may appear in the data field. The size of the
PAD, if any, is determined by the size of the data field. Max and min length of the data
and PAD field are:
– Maximum length = 1500 bytes
– Minimum length for untagged MAC frames = 46 bytes
– Minimum length for tagged MAC frames = 42 bytes
When the data field length is less than the minimum required, the PAD field is added to
match the minimum length (42 bytes for tagged frames, 46 bytes for untagged frames).
• Frame check sequence: 4-byte field that contains the cyclic redundancy check (CRC)
value. The CRC computation is based on the following fields: source address,
destination address, QTag prefix, length/type, LLC data and PAD (that is, all fields
except the preamble, SFD). The generating polynomial is the following:
32 26 23 22 16 12 11 10 8 7 5 4 2
G( x) = x +x +x +x +x +x +x +x +x +x +x +x +x +x+1
RM0090 Rev 19 1135/1751

1239
The CRC value of a frame is computed as follows:

• The first 2 bits of the frame are complemented
• The n-bits of the frame are the coefficients of a polynomial M(x) of degree (n – 1). The
first bit of the destination address corresponds to the xn – 1 term and the last bit of the
data field corresponds to the x0 term
• M(x) is multiplied by x32 and divided by G(x), producing a remainder R(x) of degree
≤ 31
• The coefficients of R(x) are considered as a 32-bit sequence
• The bit sequence is complemented and the result is the CRC
• The 32-bits of the CRC value are placed in the frame check sequence. The x32 term is
the first transmitted, the x0 term is the last one
Figure 360. MAC frame format
7 bytes Preamble
1 byte SFD
6 bytes Destination address

Bytes within
6 bytes Source address frame transmitted
top to bottom
2 bytes MAC client length/type
MAC client data

46-1500 bytes
PAD
4 bytes Frame check sequence
MSB LSB
Bit transmission order (right to left)

ai15629
1136/1751 RM0090 Rev 19

Figure 361. Tagged MAC frame format
7 bytes Preamble bytes within

frame transmitted
1 byte SFD top to bottom
6 bytes Destination address
MSB LSB
6 bytes Source address
1 0 0 0 0 0 0 0 1
QTag Prefix Length/type = 802.1QTagType
4 bytes 0 0 0 0 0 0 0 0 0
Tag control information
2 bytes MAC client length/type
MAC client data
42-1500 bytes
Pad User priority CFI
4 bytes Frame check sequence VLAN identifier (VID, 12 bits)
MSB LSB
Bit transmission order (r ight to left)
ai15630
Each byte of the MAC frame, except the FCS field, is transmitted low-order bit first.
An invalid MAC frame is defined by one of the following conditions:
• The frame length is inconsistent with the expected value as specified by the length/type
field. If the length/type field contains a type value, then the frame length is assumed to
be consistent with this field (no invalid frame)
• The frame length is not an integer number of bytes (extra bits)
• The CRC value computed on the incoming frame does not match the included FCS
33.5.2 MAC frame transmission

The DMA controls all transactions for the transmit path. Ethernet frames read from the
system memory are pushed into the FIFO by the DMA. The frames are then popped out and
transferred to the MAC core. When the end-of-frame is transferred, the status of the
transmission is taken from the MAC core and transferred back to the DMA. The Transmit
FIFO has a depth of 2 Kbyte. FIFO-fill level is indicated to the DMA so that it can initiate a
data fetch in required bursts from the system memory, using the AHB interface. The data
from the AHB Master interface is pushed into the FIFO.
When the SOF is detected, the MAC accepts the data and begins transmitting to the MII.
The time required to transmit the frame data to the MII after the application initiates
transmission is variable, depending on delay factors like IFG delay, time to transmit
preamble/SFD, and any back-off delays for Half-duplex mode. After the EOF is transferred
to the MAC core, the core completes normal transmission and then gives the status of
transmission back to the DMA. If a normal collision (in Half-duplex mode) occurs during
transmission, the MAC core makes the transmit status valid, then accepts and drops all
further data until the next SOF is received. The same frame should be retransmitted from
SOF on observing a Retry request (in the Status) from the MAC. The MAC issues an
underflow status if the data are not provided continuously during the transmission. During
the normal transfer of a frame, if the MAC receives an SOF without getting an EOF for the
previous frame, then the SOF is ignored and the new frame is considered as the
continuation of the previous frame.
RM0090 Rev 19 1137/1751

1239
There are two modes of operation for popping data towards the MAC core:
• In Threshold mode, as soon as the number of bytes in the FIFO crosses the configured
threshold level (or when the end-of-frame is written before the threshold is crossed),
the data is ready to be popped out and forwarded to the MAC core. The threshold level
is configured using the TTC bits of ETH_DMABMR.
• In Store-and-forward mode, only after a complete frame is stored in the FIFO, the frame
is popped towards the MAC core. If the Tx FIFO size is smaller than the Ethernet frame
to be transmitted, then the frame is popped towards the MAC core when the Tx FIFO
becomes almost full.
The application can flush the Transmit FIFO of all contents by setting the FTF
(ETH_DMAOMR register [20]) bit. This bit is self-clearing and initializes the FIFO pointers to
the default state. If the FTF bit is set during a frame transfer to the MAC core, then transfer
is stopped as the FIFO is considered to be empty. Hence an underflow event occurs at the
MAC transmitter and the corresponding Status word is forwarded to the DMA.
Automatic CRC and pad generation

When the number of bytes received from the application falls below 60 (DA+SA+LT+Data),
zeros are appended to the transmitting frame to make the data length exactly 46 bytes to
meet the minimum data field requirement of IEEE 802.3. The MAC can be programmed not
to append any padding. The cyclic redundancy check (CRC) for the frame check sequence
(FCS) field is calculated and appended to the data being transmitted. When the MAC is
programmed to not append the CRC value to the end of Ethernet frames, the computed
CRC is not transmitted. An exception to this rule is that when the MAC is programmed to
append pads for frames (DA+SA+LT+Data) less than 60 bytes, CRC will be appended at the
end of the padded frames.
The CRC generator calculates the 32-bit CRC for the FCS field of the Ethernet frame. The
encoding is defined by the following polynomial.
32 26 23 22 16 12 11 10 8 7 5 4 2
Transmit protocol
The MAC controls the operation of Ethernet frame transmission. It performs the following
functions to meet the IEEE 802.3/802.3z specifications. It:
• generates the preamble and SFD
• generates the jam pattern in Half-duplex mode
• controls the Jabber timeout
• controls the flow for Half-duplex mode (back pressure)
• generates the transmit frame status
• contains time stamp snapshot logic in accordance with IEEE 1588
When a new frame transmission is requested, the MAC sends out the preamble and SFD,
followed by the data. The preamble is defined as 7 bytes of 0b10101010 pattern, and the
SFD is defined as 1 byte of 0b10101011 pattern. The collision window is defined as 1 slot
time (512 bit times for 10/100 Mbit/s Ethernet). The jam pattern generation is applicable only
to Half-duplex mode, not to Full-duplex mode.
In MII mode, if a collision occurs at any time from the beginning of the frame to the end of
the CRC field, the MAC sends a 32-bit jam pattern of 0x5555 5555 on the MII to inform all
1138/1751 RM0090 Rev 19

other stations that a collision has occurred. If the collision is seen during the preamble
transmission phase, the MAC completes the transmission of the preamble and SFD and
then sends the jam pattern.
A jabber timer is maintained to cut off the transmission of Ethernet frames if more than 2048
(default) bytes have to be transferred. The MAC uses the deferral mechanism for flow
control (back pressure) in Half-duplex mode. When the application requests to stop
receiving frames, the MAC sends a JAM pattern of 32 bytes whenever it senses the
reception of a frame, provided that transmit flow control is enabled. This results in a collision
and the remote station backs off. The application requests flow control by setting the BPA bit
(bit 0) in the ETH_MACFCR register. If the application requests a frame to be transmitted,
then it is scheduled and transmitted even when back pressure is activated. Note that if back
pressure is kept activated for a long time (and more than 16 consecutive collision events
occur) then the remote stations abort their transmissions due to excessive collisions. If IEEE
1588 time stamping is enabled for the transmit frame, this block takes a snapshot of the
system time when the SFD is put onto the transmit MII bus.
Transmit scheduler
The MAC is responsible for scheduling the frame transmission on the MII. It maintains the
interframe gap between two transmitted frames and follows the truncated binary exponential
backoff algorithm for Half-duplex mode. The MAC enables transmission after satisfying the
IFG and backoff delays. It maintains an idle period of the configured interframe gap (IFG bits
in the ETH_MACCR register) between any two transmitted frames. If frames to be
transmitted arrive sooner than the configured IFG time, the MII waits for the enable signal
from the MAC before starting the transmission on it. The MAC starts its IFG counter as soon
as the carrier signal of the MII goes inactive. At the end of the programmed IFG value, the
MAC enables transmission in Full-duplex mode. In Half-duplex mode and when IFG is
configured for 96 bit times, the MAC follows the rule of deference specified in Section
4.2.3.2.1 of the IEEE 802.3 specification. The MAC resets its IFG counter if a carrier is
detected during the first two-thirds (64-bit times for all IFG values) of the IFG interval. If the
carrier is detected during the final one third of the IFG interval, the MAC continues the IFG
count and enables the transmitter after the IFG interval. The MAC implements the truncated
binary exponential backoff algorithm when it operates in Half-duplex mode.
Transmit flow control

When the Transmit Flow Control Enable bit (TFE bit in ETH_MACFCR) is set, the MAC
generates Pause frames and transmits them as necessary, in Full-duplex mode. The Pause
frame is appended with the calculated CRC, and is sent. Pause frame generation can be
initiated in two ways.
A pause frame is sent either when the application sets the FCB bit in the ETH_MACFCR
register or when the receive FIFO is full (packet buffer).
• If the application has requested flow control by setting the FCB bit in ETH_MACFCR,
the MAC generates and transmits a single Pause frame. The value of the pause time in
the generated frame contains the programmed pause time value in ETH_MACFCR. To
extend the pause or end the pause prior to the time specified in the previously
transmitted Pause frame, the application must request another Pause frame
transmission after programming the Pause Time value (PT in ETH_MACFCR register)
with the appropriate value.
• If the application has requested flow control when the receive FIFO is full, the MAC
generates and transmits a Pause frame. The value of the pause time in the generated
frame is the programmed pause time value in ETH_MACFCR. If the receive FIFO
RM0090 Rev 19 1139/1751

1239
remains full at a configurable number of slot-times (PLT bits in ETH_MACFCR) before

this Pause time runs out, a second Pause frame is transmitted. The process is
repeated as long as the receive FIFO remains full. If this condition is no more satisfied
prior to the sampling time, the MAC transmits a Pause frame with zero pause time to
indicate to the remote end that the receive buffer is ready to receive new data frames.
Single-packet transmit operation
The general sequence of events for a transmit operation is as follows:
1. If the system has data to be transferred, the DMA controller fetches them from the
memory through the AHB Master interface and starts forwarding them to the FIFO. It
continues to receive the data until the end of frame is transferred.
2. When the threshold level is crossed or a full packet of data is received into the FIFO,
the frame data are popped and driven to the MAC core. The DMA continues to transfer
data from the FIFO until a complete packet has been transferred to the MAC. Upon
completion of the frame, the DMA controller is notified by the status coming from the
MAC.
Transmit operation—Two packets in the buffer
1. Because the DMA must update the descriptor status before releasing it to the Host,
there can be at the most two frames inside a transmit FIFO. The second frame is
fetched by the DMA and put into the FIFO only if the OSF (operate on second frame)
bit is set. If this bit is not set, the next frame is fetched from the memory only after the
MAC has completely processed the frame and the DMA has released the descriptors.
2. If the OSF bit is set, the DMA starts fetching the second frame immediately after
completing the transfer of the first frame to the FIFO. It does not wait for the status to
be updated. In the meantime, the second frame is received into the FIFO while the first
frame is being transmitted. As soon as the first frame has been transferred and the
status is received from the MAC, it is pushed to the DMA. If the DMA has already
completed sending the second packet to the FIFO, the second transmission must wait
for the status of the first packet before proceeding to the next frame.
Retransmission during collision

While a frame is being transferred to the MAC, a collision event may occur on the MAC line
interface in Half-duplex mode. The MAC would then indicate a retry attempt by giving the
status even before the end of frame is received. Then the retransmission is enabled and the
frame is popped out again from the FIFO. After more than 96 bytes have been popped
towards the MAC core, the FIFO controller frees up that space and makes it available to the
DMA to push in more data. This means that the retransmission is not possible after this
threshold is crossed or when the MAC core indicates a late collision event.
Transmit FIFO flush operation

The MAC provides a control to the software to flush the Transmit FIFO through the use of Bit
20 in the Operation mode register. The Flush operation is immediate and the Tx FIFO and
the corresponding pointers are cleared to the initial state even if the Tx FIFO is in the middle
of transferring a frame to the MAC Core. This results in an underflow event in the MAC
transmitter, and the frame transmission is aborted. The status of such a frame is marked
with both underflow and frame flush events (TDES0 bits 13 and 1). No data are coming to
the FIFO from the application (DMA) during the Flush operation. Transfer transmit status
words are transferred to the application for the number of frames that is flushed (including
partial frames). Frames that are completely flushed have the Frame flush status bit (TDES0
13) set. The Flush operation is completed when the application (DMA) has accepted all of
1140/1751 RM0090 Rev 19

the Status words for the frames that were flushed. The Transmit FIFO Flush control register
bit is then cleared. At this point, new frames from the application (DMA) are accepted. All
data presented for transmission after a Flush operation are discarded unless they start with
an SOF marker.
Transmit status word

At the end of the Ethernet frame transfer to the MAC core and after the core has completed
the transmission of the frame, the transmit status is given to the application. The detailed
description of the Transmit Status is the same as for bits [23:0] in TDES0. If IEEE 1588 time
stamping is enabled, a specific frames’ 64-bit time stamp is returned, along with the transmit
status.
Transmit checksum offload

Communication protocols such as TCP and UDP implement checksum fields, which helps
determine the integrity of data transmitted over a network. Because the most widespread
use of Ethernet is to encapsulate TCP and UDP over IP datagrams, the Ethernet controller
has a transmit checksum offload feature that supports checksum calculation and insertion in
the transmit path, and error detection in the receive path. This section explains the operation
of the checksum offload feature for transmitted frames.
Note: The checksum for TCP, UDP or ICMP is calculated over a complete frame, then inserted
into its corresponding header field. Due to this requirement, this function is enabled only
when the Transmit FIFO is configured for Store-and-forward mode (that is, when the TSF bit
is set in the ETH_ETH_DMAOMR register). If the core is configured for Threshold (cut-
through) mode, the Transmit checksum offload is bypassed.
You must make sure the Transmit FIFO is deep enough to store a complete frame before
that frame is transferred to the MAC Core transmitter. If the FIFO depth is less than the input
Ethernet frame size, the payload (TCP/UDP/ICMP) checksum insertion function is bypassed
and only the frame’s IPv4 Header checksum is modified, even in Store-and-forward mode.
The transmit checksum offload supports two types of checksum calculation and insertion.
This checksum can be controlled for each frame by setting the CIC bits (Bits 28:27 in
TDES1, described in TDES1: Transmit descriptor Word1).
See IETF specifications RFC 791, RFC 793, RFC 768, RFC 792, RFC 2460 and RFC 4443
for IPv4, TCP, UDP, ICMP, IPv6 and ICMPv6 packet header specifications, respectively.
• IP header checksum
In IPv4 datagrams, the integrity of the header fields is indicated by the 16-bit header
checksum field (the eleventh and twelfth bytes of the IPv4 datagram). The checksum
offload detects an IPv4 datagram when the Ethernet frame’s Type field has the value
0x0800 and the IP datagram’s Version field has the value 0x4. The input frame’s
checksum field is ignored during calculation and replaced by the calculated value. IPv6
headers do not have a checksum field; thus, the checksum offload does not modify
IPv6 header fields. The result of this IP header checksum calculation is indicated by the
IP Header Error status bit in the Transmit status (Bit 16). This status bit is set whenever
the values of the Ethernet Type field and the IP header’s Version field are not
consistent, or when the Ethernet frame does not have enough data, as indicated by the
RM0090 Rev 19 1141/1751

1239
IP header Length field. In other words, this bit is set when an IP header error is
asserted under the following circumstances:
a) For IPv4 datagrams:
– The received Ethernet type is 0x0800, but the IP header’s Version field does not
equal 0x4
– The IPv4 Header Length field indicates a value less than 0x5 (20 bytes)
– The total frame length is less than the value given in the IPv4 Header Length field
b) For IPv6 datagrams:
– The Ethernet type is 0x86DD but the IP header Version field does not equal 0x6
– The frame ends before the IPv6 header (40 bytes) or extension header (as given
in the corresponding Header Length field in an extension header) has been
completely received. Even when the checksum offload detects such an IP header
error, it inserts an IPv4 header checksum if the Ethernet Type field indicates an
IPv4 payload.
• TCP/UDP/ICMP checksum
The TCP/UDP/ICMP checksum processes the IPv4 or IPv6 header (including
extension headers) and determines whether the encapsulated payload is TCP, UDP or
ICMP.
Note that:
a) For non-TCP, -UDP, or -ICMP/ICMPv6 payloads, this checksum is bypassed and
nothing further is modified in the frame.
b) Fragmented IP frames (IPv4 or IPv6), IP frames with security features (such as an
authentication header or encapsulated security payload), and IPv6 frames with
routing headers are bypassed and not processed by the checksum.
The checksum is calculated for the TCP, UDP, or ICMP payload and inserted into its
corresponding field in the header. It can work in the following two modes:
– In the first mode, the TCP, UDP, or ICMPv6 pseudo-header is not included in the
checksum calculation and is assumed to be present in the input frame’s checksum
field. The checksum field is included in the checksum calculation, and then
replaced by the final calculated checksum.
– In the second mode, the checksum field is ignored, the TCP, UDP, or ICMPv6
pseudo-header data are included into the checksum calculation, and the
checksum field is overwritten with the final calculated value.
Note that: for ICMP-over-IPv4 packets, the checksum field in the ICMP packet must
always be 0x0000 in both modes, because pseudo-headers are not defined for such
packets. If it does not equal 0x0000, an incorrect checksum may be inserted into the
packet.
The result of this operation is indicated by the payload checksum error status bit in the
Transmit Status vector (bit 12). The payload checksum error status bit is set when
either of the following is detected:
– the frame has been forwarded to the MAC transmitter in Store-and-forward mode
without the end of frame being written to the FIFO
– the packet ends before the number of bytes indicated by the payload length field in
the IP header is received.
When the packet is longer than the indicated payload length, the bytes are ignored as
stuff bytes, and no error is reported. When the first type of error is detected, the TCP,
1142/1751 RM0090 Rev 19

UDP or ICMP header is not modified. For the second error type, still, the calculated
checksum is inserted into the corresponding header field.
MII/RMII transmit bit order

Each nibble from the MII is transmitted on the RMII a dibit at a time with the order of dibit
transmission shown in Figure 362. Lower order bits (D1 and D0) are transmitted first
followed by higher order bits (D2 and D3).
Figure 362. Transmission bit order
RMII_TXD[1:0]
LSB MSB
D0 D1 Bibit stream
LSB D0
D1
MII_TXD[3:0]
D2
MSB D3
Nibble stream ai15632
MII/RMII transmit timing diagrams
Figure 363. Transmission with no collision
MII_TX_CLK
MII_TX_EN
MII_TXD[3:0] PR EA MB LE
MII_CS
MII_COL
Low
ai15631
RM0090 Rev 19 1143/1751

1239
Figure 364. Transmission with collision
MII_TX_CLK
MII_TX_EN
MII_TXD[3:0] PR EAM BLE SFD DA DA JAM JAM JAM JAM
MII_CS
MII_COL
ai15651
Figure 365 shows a frame transmission in MII and RMII.
Figure 365. Frame transmission in MMI and RMII modes
MII_RX_CLK
MII_TX_EN
MII_TXD[3:0]
RMII_REF_CLK
RMII_TX_EN
RMII_TXD[1:0]
ai15652
33.5.3 MAC frame reception

The MAC received frames are pushes into the Rx FIFO. The status (fill level) of this FIFO is
indicated to the DMA once it crosses the configured receive threshold (RTC in the
ETH_DMAOMR register) so that the DMA can initiate pre-configured burst transfers
towards the AHB interface.
In the default Cut-through mode, when 64 bytes (configured with the RTC bits in the
ETH_DMAOMR register) or a full packet of data are received into the FIFO, the data are
popped out and the DMA is notified of its availability. Once the DMA has initiated the
transfer to the AHB interface, the data transfer continues from the FIFO until a complete
1144/1751 RM0090 Rev 19

packet has been transferred. Upon completion of the EOF frame transfer, the status word is
popped out and sent to the DMA controller.
In Rx FIFO Store-and-forward mode (configured by the RSF bit in the ETH_DMAOMR
register), a frame is read only after being written completely into the Receive FIFO. In this
mode, all error frames are dropped (if the core is configured to do so) such that only valid
frames are read and forwarded to the application. In Cut-through mode, some error frames
are not dropped, because the error status is received at the end of the frame, by which time
the start of that frame has already been read of the FIFO.
A receive operation is initiated when the MAC detects an SFD on the MII. The core strips the
preamble and SFD before proceeding to process the frame. The header fields are checked
for the filtering and the FCS field used to verify the CRC for the frame. The frame is dropped
in the core if it fails the address filter.
Receive protocol
The received frame preamble and SFD are stripped. Once the SFD has been detected, the
MAC starts sending the Ethernet frame data to the receive FIFO, beginning with the first
byte following the SFD (destination address). If IEEE 1588 time stamping is enabled, a
snapshot of the system time is taken when any frame's SFD is detected on the MII. Unless
the MAC filters out and drops the frame, this time stamp is passed on to the application.
If the received frame length/type field is less than 0x600 and if the MAC is programmed for
the auto CRC/pad stripping option, the MAC sends the data of the frame to RxFIFO up to
the count specified in the length/type field, then starts dropping bytes (including the FCS
field). If the Length/Type field is greater than or equal to 0x600, the MAC sends all received
Ethernet frame data to Rx FIFO, regardless of the value on the programmed auto-CRC strip
option. The MAC watchdog timer is enabled by default, that is, frames above 2048 bytes
(DA + SA + LT + Data + pad + FCS) are cut off. This feature can be disabled by
programming the watchdog disable (WD) bit in the MAC configuration register. However,
even if the watchdog timer is disabled, frames greater than 16 KB in size are cut off and a
watchdog timeout status is given.
Receive CRC: automatic CRC and pad stripping

The MAC checks for any CRC error in the receiving frame. It calculates the 32-bit CRC for
the received frame that includes the Destination address field through the FCS field. The
encoding is defined by the following polynomial.
32 26 23 22 16 12 11 10 8 7 5 4 2
Regardless of the auto-pad/CRC strip, the MAC receives the entire frame to compute the
CRC check for the received frame.
Receive checksum offload

Both IPv4 and IPv6 frames in the received Ethernet frames are detected and processed for
data integrity. You can enable the receive checksum offload by setting the IPCO bit in the
ETH_MACCR register. The MAC receiver identifies IPv4 or IPv6 frames by checking for
value 0x0800 or 0x86DD, respectively, in the received Ethernet frame Type field. This
identification applies to VLAN-tagged frames as well. The receive checksum offload
calculates IPv4 header checksums and checks that they match the received IPv4 header
checksums. The IP Header Error bit is set for any mismatch between the indicated payload
RM0090 Rev 19 1145/1751

1239
type (Ethernet Type field) and the IP header version, or when the received frame does not
have enough bytes, as indicated by the IPv4 header’s Length field (or when fewer than 20
bytes are available in an IPv4 or IPv6 header). The receive checksum offload also identifies
a TCP, UDP or ICMP payload in the received IP datagrams (IPv4 or IPv6) and calculates the
checksum of such payloads properly, as defined in the TCP, UDP or ICMP specifications. It
includes the TCP/UDP/ICMPv6 pseudo-header bytes for checksum calculation and checks
whether the received checksum field matches the calculated value. The result of this
operation is given as a Payload Checksum Error bit in the receive status word. This status
bit is also set if the length of the TCP, UDP or ICMP payload does not match the expected
payload length given in the IP header. As mentioned in TCP/UDP/ICMP checksum, the
receive checksum offload bypasses the payload of fragmented IP datagrams, IP datagrams
with security features, IPv6 routing headers, and payloads other than TCP, UDP or ICMP.
This information (whether the checksum is bypassed or not) is given in the receive status,
as described in the RDES0: Receive descriptor Word0 section. In this configuration, the
core does not append any payload checksum bytes to the received Ethernet frames.
As mentioned in RDES0: Receive descriptor Word0, the meaning of certain register bits
changes as shown in Table 190.
Table 190. Frame statuses

Bit 18: Bit 27: Header Bit 28: Payload
Frame status
Ethernet frame checksum error checksum error
The frame is an IEEE 802.3 frame (Length

0 0 0
field value is less than 0x0600).
IPv4/IPv6 Type frame in which no checksum
1 0 0
error is detected.
IPv4/IPv6 Type frame in which a payload
1 0 1 checksum error (as described for PCE) is
detected
IPv4/IPv6 Type frame in which IP header
1 1 0 checksum error (as described for IPCO HCE)
is detected.
IPv4/IPv6 Type frame in which both PCE and
1 1 1
IPCO HCE are detected.
IPv4/IPv6 Type frame in which there is no IP
0 0 1 HCE and the payload check is bypassed due
to unsupported payload.
Type frame which is neither IPv4 or IPv6
0 1 1 (checksum offload bypasses the checksum
check completely)
0 1 0 Reserved
Receive frame controller

If the RA bit is reset in the MAC CSR frame filter register, the MAC performs frame filtering
based on the destination/source address (the application still needs to perform another level
of filtering if it decides not to receive any bad frames like runt, CRC error frames, etc.). On
detecting a filter-fail, the frame is dropped and not transferred to the application. When the
filtering parameters are changed dynamically, and in case of (DA-SA) filter-fail, the rest of
the frame is dropped and the Rx Status Word is immediately updated (with zero frame
1146/1751 RM0090 Rev 19

length, CRC error and Runt Error bits set), indicating the filter fail. In Ethernet power down
mode, all received frames are dropped, and are not forwarded to the application.
Receive flow control

The MAC detects the receiving Pause frame and pauses the frame transmission for the
delay specified within the received Pause frame (only in Full-duplex mode). The Pause
frame detection function can be enabled or disabled with the RFCE bit in ETH_MACFCR.
Once receive flow control has been enabled, the received frame destination address begins
to be monitored for any match with the multicast address of the control frame
(0x0180 C200 0001). If a match is detected (the destination address of the received frame
matches the reserved control frame destination address), the MAC then decides whether or
not to transfer the received control frame to the application, based on the level of the PCF
bit in ETH_MACFFR.
The MAC also decodes the type, opcode, and Pause Timer fields of the receiving control
frame. If the byte count of the status indicates 64 bytes, and if there is no CRC error, the
MAC transmitter pauses the transmission of any data frame for the duration of the decoded
Pause time value, multiplied by the slot time (64 byte times for both 10/100 Mbit/s modes).
Meanwhile, if another Pause frame is detected with a zero Pause time value, the MAC
resets the Pause time and manages this new pause request.
If the received control frame matches neither the type field (0x8808), the opcode (0x00001),
nor the byte length (64 bytes), or if there is a CRC error, the MAC does not generate a
Pause.
In the case of a pause frame with a multicast destination address, the MAC filters the frame
based on the address match.
For a pause frame with a unicast destination address, the MAC filtering depends on whether
the DA matched the contents of the MAC address 0 register and whether the UPDF bit in
ETH_MACFCR is set (detecting a pause frame even with a unicast destination address).
The PCF register bits (bits [7:6] in ETH_MACFFR) control filtering for control frames in
addition to address filtering.
Receive operation multiframe handling

Since the status is available immediately following the data, the FIFO is capable of storing
any number of frames into it, as long as it is not full.
Error handling
If the Rx FIFO is full before it receives the EOF data from the MAC, an overflow is declared
and the whole frame is dropped, and the overflow counter in the (ETH_DMAMFBOCR
register) is incremented. The status indicates a partial frame due to overflow. The Rx FIFO
can filter error and undersized frames, if enabled (using the FEF and FUGF bits in
ETH_DMAOMR).
If the Receive FIFO is configured to operate in Store-and-forward mode, all error frames can
be filtered and dropped.
In Cut-through mode, if a frame's status and length are available when that frame's SOF is
read from the Rx FIFO, then the complete erroneous frame can be dropped. The DMA can
flush the error frame being read from the FIFO, by enabling the receive frame flash bit. The
data transfer to the application (DMA) is then stopped and the rest of the frame is internally
read and dropped. The next frame transfer can then be started, if available.
RM0090 Rev 19 1147/1751

1239
Receive status word

At the end of the Ethernet frame reception, the MAC outputs the receive status to the
application (DMA). The detailed description of the receive status is the same as for
bits[31:0] in RDES0, given in RDES0: Receive descriptor Word0.
Frame length interface

In case of switch applications, data transmission and reception between the application and
MAC happen as complete frame transfers. The application layer should be aware of the
length of the frames received from the ingress port in order to transfer the frame to the
egress port. The MAC core provides the frame length of each received frame inside the
status at the end of each frame reception.
Note: A frame length value of 0 is given for partial frames written into the Rx FIFO due to overflow.
MII/RMII receive bit order

Each nibble is transmitted to the MII from the dibit received from the RMII in the nibble
transmission order shown in Figure 366. The lower-order bits (D0 and D1) are received first,
followed by the higher-order bits (D2 and D3).
Figure 366. Receive bit order

RMII_RXD[1:0]
LSB MSB
D0 D1 Di-bit stream
LSB D0
D1
MII_RXD[3:0]
D2
MSB D3
Nibble stream ai15633
1148/1751 RM0090 Rev 19

Figure 367. Reception with no error
MII_RX_CLK
MII_RX_DV
MII_RXD[3:0] PREAMBLE SFD FCS
MII_RX_ERR
ai15634
Figure 368. Reception with errors
MII_RX_CLK
MII_RX_DV
MII_RXD[3:0] PREAMBLE SFD DA DA XX XX XX
MII_RX_ERR
ai15635
Figure 369. Reception with false carrier indication
MII_RX_CLK
MII_RX_DV
MII_RXD[3:0] XX XX XX XX 0E XX XX XX XX
MII_RX_ERR
ai15636
33.5.4 MAC interrupts

Interrupts can be generated from the MAC core as a result of various events.
The ETH_MACSR register describes the events that can cause an interrupt from the MAC
core. You can prevent each event from asserting the interrupt by setting the corresponding
mask bits in the Interrupt Mask register.
RM0090 Rev 19 1149/1751

1239
The interrupt register bits only indicate the block from which the event is reported. You have
to read the corresponding status registers and other registers to clear the interrupt. For
example, bit 3 of the Interrupt register, set high, indicates that the Magic packet or Wake-on-
LAN frame is received in Power-down mode. You must read the ETH_MACPMTCSR
register to clear this interrupt event.
Figure 370. MAC core interrupt masking scheme
TSTS
TSTI
AND
TSTIM
Interrupt
OR
PMTS
PMTI
AND
PMTIM
ai15637
33.5.5 MAC filtering

Address filtering
Address filtering checks the destination and source addresses on all received frames and
the address filtering status is reported accordingly. Address checking is based on different
parameters (Frame filter register) chosen by the application. The filtered frame can also be
identified: multicast or broadcast frame.
Address filtering uses the station's physical (MAC) address and the Multicast Hash table for
address checking purposes.
Unicast destination address filter

The MAC supports up to 4 MAC addresses for unicast perfect filtering. If perfect filtering is
selected (HU bit in the Frame filter register is reset), the MAC compares all 48 bits of the
received unicast address with the programmed MAC address for any match. Default
MacAddr0 is always enabled, other addresses MacAddr1–MacAddr3 are selected with an
individual enable bit. Each byte of these other addresses (MacAddr1–MacAddr3) can be
masked during comparison with the corresponding received DA byte by setting the
corresponding Mask Byte Control bit in the register. This helps group address filtering for the
DA. In Hash filtering mode (when HU bit is set), the MAC performs imperfect filtering for
unicast addresses using a 64-bit Hash table. For hash filtering, the MAC uses the 6 upper
CRC (see note 1 below) bits of the received destination address to index the content of the
Hash table. A value of 000000 selects bit 0 in the selected register, and a value of 111111
selects bit 63 in the Hash Table register. If the corresponding bit (indicated by the 6-bit CRC)
is set to 1, the unicast frame is said to have passed the Hash filter; otherwise, the frame has
failed the Hash filter.
Note: This CRC is a 32-bit value coded by the following polynomial (for more details refer to
Section 33.5.3):
32 26 23 22 16 12 11 10 8 7 5 4 2
G(x ) = x +x +x +x +x +x +x +x +x +x +x +x +x +x+1
1150/1751 RM0090 Rev 19

Multicast destination address filter

The MAC can be programmed to pass all multicast frames by setting the PAM bit in the
Frame filter register. If the PAM bit is reset, the MAC performs the filtering for multicast
addresses based on the HM bit in the Frame filter register. In Perfect filtering mode, the
multicast address is compared with the programmed MAC destination address registers
(1–3). Group address filtering is also supported. In Hash filtering mode, the MAC performs
imperfect filtering using a 64-bit Hash table. For hash filtering, the MAC uses the 6 upper
CRC (see note 1 below) bits of the received multicast address to index the content of the
Hash table. A value of 000000 selects bit 0 in the selected register and a value of 111111
selects bit 63 in the Hash Table register. If the corresponding bit is set to 1, then the
multicast frame is said to have passed the Hash filter; otherwise, the frame has failed the
Hash filter.
Note: This CRC is a 32-bit value coded by the following polynomial (for more details refer to
Section 33.5.3):
32 26 23 22 16 12 11 10 8 7 5 4 2
G(x ) = x +x +x +x +x +x +x +x +x +x +x +x +x +x+1
Hash or perfect address filter

The DA filter can be configured to pass a frame when its DA matches either the Hash filter
or the Perfect filter by setting the HPF bit in the Frame filter register and setting the
corresponding HU or HM bits. This configuration applies to both unicast and multicast
frames. If the HPF bit is reset, only one of the filters (Hash or Perfect) is applied to the
received frame.
Broadcast address filter

The MAC does not filter any broadcast frames in the default mode. However, if the MAC is
programmed to reject all broadcast frames by setting the BFD bit in the Frame filter register,
any broadcast frames are dropped.
Unicast source address filter

The MAC can also perform perfect filtering based on the source address field of the
received frames. By default, the MAC compares the SA field with the values programmed in
the SA registers. The MAC address registers [1:3] can be configured to contain SA instead
of DA for comparison, by setting bit 30 in the corresponding register. Group filtering with SA
is also supported. The frames that fail the SA filter are dropped by the MAC if the SAF bit in
the Frame filter register is set. Otherwise, the result of the SA filter is given as a status bit in
the Receive Status word (see RDES0: Receive descriptor Word0).
When the SAF bit is set, the result of the SA and DA filters is AND’ed to decide whether the
frame needs to be forwarded. This means that either of the filter fail result will drop the
frame. Both filters have to pass the frame for the frame to be forwarded to the application.
Inverse filtering operation

For both destination and source address filtering, there is an option to invert the filter-match
result at the final output. These are controlled by the DAIF and SAIF bits in the Frame filter
register, respectively. The DAIF bit is applicable for both Unicast and Multicast DA frames.
The result of the unicast/multicast destination address filter is inverted in this mode.
Similarly, when the SAIF bit is set, the result of the unicast SA filter is inverted. Table 191
RM0090 Rev 19 1151/1751

1239
and Table 192 summarize destination and source address filtering based on the type of
frame received.
Table 191. Destination address filtering

Frame
PM HPF HU DAIF HM PAM DB DA filter operation
type
1 X X X X X X Pass
Broadcast 0 X X X X X 0 Pass
0 X X X X X 1 Fail
1 X X X X X X Pass all frames
0 X 0 0 X X X Pass on perfect/group filter match
0 X 0 1 X X X Fail on perfect/Group filter match
0 0 1 0 X X X Pass on hash filter match
Unicast
0 0 1 1 X X X Fail on hash filter match
Pass on hash or perfect/Group filter
0 1 1 0 X X X
match
0 1 1 1 X X X Fail on hash or perfect/Group filter match
1 X X X X X X Pass all frames
X X X X X 1 X Pass all frames
Pass on Perfect/Group filter match and
0 X X 0 0 0 X
drop PAUSE control frames if PCF = 0x
Pass on hash filter match and drop
0 0 X 0 1 0 X
PAUSE control frames if PCF = 0x
Pass on hash or perfect/Group filter
Multicast 0 1 X 0 1 0 X match and drop PAUSE control frames if
PCF = 0x
Fail on perfect/Group filter match and
0 X X 1 0 0 X
drop PAUSE control frames if PCF = 0x
Fail on hash filter match and drop PAUSE
0 0 X 1 1 0 X
control frames if PCF = 0x
Fail on hash or perfect/Group filter match
0 1 X 1 1 0 X and drop PAUSE control frames if PCF =
0x
1152/1751 RM0090 Rev 19

Table 192. Source address filtering

Frame type PM SAIF SAF SA filter operation
1 X X Pass all frames

Pass status on perfect/Group filter match but do not drop
0 0 0
frames that fail
Unicast
0 1 0 Fail status on perfect/group filter match but do not drop frame
0 0 1 Pass on perfect/group filter match and drop frames that fail
0 1 1 Fail on perfect/group filter match and drop frames that fail
33.5.6 MAC loopback mode

The MAC supports loopback of transmitted frames onto its receiver. By default, the MAC
loopback function is disabled, but this feature can be enabled by programming the
Loopback bit in the MAC ETH_MACCR register.
33.5.7 MAC management counters: MMC

The MAC management counters (MMC) maintain a set of registers for gathering statistics
on the received and transmitted frames. These include a control register for controlling the
behavior of the registers, two 32-bit registers containing generated interrupts (receive and
transmit), and two 32-bit registers containing masks for the Interrupt register (receive and
transmit). These registers are accessible from the application. Each register is 32 bits wide.
Section 33.8 describes the various counters and lists the addresses of each of the statistics
counters. This address is used for read/write accesses to the desired transmit/receive
counter.
The Receive MMC counters are updated for frames that pass address filtering. Dropped
frames statistics are not updated unless the dropped frames are runt frames of less than 6
bytes (DA bytes are not received fully).
Good transmitted and received frames

Transmitted frames are considered “good” if transmitted successfully. In other words, a
transmitted frame is good if the frame transmission is not aborted due to any of the following
errors:
+ Jabber Timeout
+ No Carrier/Loss of Carrier
+ Late Collision
+ Frame Underflow
+ Excessive Deferral
+ Excessive Collision
RM0090 Rev 19 1153/1751

1239
Received frames are considered “good” if none of the following errors exists:
+ CRC error
+ Runt Frame (shorter than 64 bytes)
+ Alignment error (in 10/ 100 Mbit/s only)
+ Length error (non-Type frames only)
+ Out of Range (non-Type frames only, longer than maximum size)
+ MII_RXER Input error
The maximum frame size depends on the frame type, as follows:
+ Untagged frame maxsize = 1518
+ VLAN Frame maxsize = 1522
33.5.8 Power management: PMT

This section describes the power management (PMT) mechanisms supported by the MAC.
PMT supports the reception of network (remote) wakeup frames and Magic Packet frames.
PMT generates interrupts for wakeup frames and Magic Packets received by the MAC. The
PMT block is enabled with remote wakeup frame enable and Magic Packet enable. These
enable bits (WFE and MPE) are in the ETH_MACPMTCSR register and are programmed by
the application. When the power down mode is enabled in the PMT, then all received frames
are dropped by the MAC and they are not forwarded to the application. The MAC comes out
of the power down mode only when either a Magic Packet or a Remote wakeup frame is
received and the corresponding detection is enabled.
Remote wakeup frame filter register

There are eight wakeup frame filter registers. To write on each of them, load the wakeup
frame filter register value by value. The wanted values of the wakeup frame filter are loaded
by sequentially loading eight times the wakeup frame filter register. The read operation is
identical to the write operation. To read the eight values, you have to read eight times the
wakeup frame filter register to reach the last register. Each read/write points the wakeup
frame filter register to the next filter register.
Figure 371. Wakeup frame filter register
Wakeup frame filter reg0 Filter 0 Byte Mask
Filter 3 Filter 2 Filter 1 Filter 0

Wakeup frame filter reg4 RSVD
Command
RSVD
Command
RSVD
Command
RSVD
Command
Wakeup frame filter reg5 Filter 3 Offset Filter 2 Offset Filter 1 Offset Filter 0 Offset
Wakeup frame filter reg6 Filter 1 CRC - 16 Filter 0 CRC - 16
ai15647
1154/1751 RM0090 Rev 19

• Filter i Byte Mask

This register defines which bytes of the frame are examined by filter i (0, 1, 2, and 3) in
order to determine whether or not the frame is a wakeup frame. The MSB (thirty-first
bit) must be zero. Bit j [30:0] is the Byte Mask. If bit j (byte number) of the Byte Mask is
set, then Filter i Offset + j of the incoming frame is processed by the CRC block;
otherwise Filter i Offset + j is ignored.
• Filter i Command
This 4-bit command controls the filter i operation. Bit 3 specifies the address type,
defining the pattern’s destination address type. When the bit is set, the pattern applies
to only multicast frames. When the bit is reset, the pattern applies only to unicast
frames. Bit 2 and bit 1 are reserved. Bit 0 is the enable bit for filter i; if bit 0 is not set,
filter i is disabled.
• Filter i Offset
This register defines the offset (within the frame) from which the frames are examined
by filter i. This 8-bit pattern offset is the offset for the filter i first byte to be examined.
The minimum allowed is 12, which refers to the 13th byte of the frame (offset value 0
refers to the first byte of the frame).
• Filter i CRC-16
This register contains the CRC_16 value calculated from the pattern, as well as the
byte mask programmed to the wakeup filter register block.
Remote wakeup frame detection

When the MAC is in sleep mode and the remote wakeup bit is enabled in the
ETH_MACPMTCSR register, normal operation is resumed after receiving a remote wakeup
frame. The application writes all eight wakeup filter registers, by performing a sequential
write to the wakeup frame filter register address. The application enables remote wakeup by
writing a 1 to bit 2 in the ETH_MACPMTCSR register. PMT supports four programmable
filters that provide different receive frame patterns. If the incoming frame passes the
address filtering of Filter Command, and if Filter CRC-16 matches the incoming examined
pattern, then the wakeup frame is received. Filter_offset (minimum value 12, which refers to
the 13th byte of the frame) determines the offset from which the frame is to be examined.
Filter Byte Mask determines which bytes of the frame must be examined. The thirty-first bit
of Byte Mask must be set to zero. The wakeup frame is checked only for length error, FCS
error, dribble bit error, MII error, collision, and to ensure that it is not a runt frame. Even if the
wakeup frame is more than 512 bytes long, if the frame has a valid CRC value, it is
considered valid. Wakeup frame detection is updated in the ETH_MACPMTCSR register for
every remote wakeup frame received. If enabled, a PMT interrupt is generated to indicate
the reception of a remote wakeup frame.
Magic packet detection

The Magic Packet frame is based on a method that uses Advanced Micro Device’s Magic
Packet technology to power up the sleeping device on the network. The MAC receives a
specific packet of information, called a Magic Packet, addressed to the node on the network.
Only Magic Packets that are addressed to the device or a broadcast address are checked to
determine whether they meet the wakeup requirements. Magic Packets that pass address
filtering (unicast or broadcast) are checked to determine whether they meet the remote
Wake-on-LAN data format of 6 bytes of all ones followed by a MAC address appearing 16
times. The application enables Magic Packet wakeup by writing a 1 to bit 1 in the
ETH_MACPMTCSR register. The PMT block constantly monitors each frame addressed to
RM0090 Rev 19 1155/1751

1239
the node for a specific Magic Packet pattern. Each received frame is checked for a
0xFFFF FFFF FFFF pattern following the destination and source address field. The PMT
block then checks the frame for 16 repetitions of the MAC address without any breaks or
interruptions. In case of a break in the 16 repetitions of the address, the 0xFFFF FFFF FFFF
pattern is scanned for again in the incoming frame. The 16 repetitions can be anywhere in
the frame, but must be preceded by the synchronization stream (0xFFFF FFFF FFFF). The
device also accepts a multicast frame, as long as the 16 duplications of the MAC address
are detected. If the MAC address of a node is 0x0011 2233 4455, then the MAC scans for
the data sequence:
Destination address source address ……………….. FFFF FFFF FFFF
0011 2233 4455 0011 2233 4455 0011 2233 4455 0011 2233 4455
0011 2233 4455 0011 2233 4455 0011 2233 4455 0011 2233 4455
0011 2233 4455 0011 2233 4455 0011 2233 4455 0011 2233 4455
0011 2233 4455 0011 2233 4455 0011 2233 4455 0011 2233 4455
…CRC
Magic Packet detection is updated in the ETH_MACPMTCSR register for received Magic
Packet. If enabled, a PMT interrupt is generated to indicate the reception of a Magic Packet.
System consideration during power-down

The Ethernet PMT block is able to detect frames while the system is in the Stop mode,
provided that the EXTI line 19 is enabled.
The MAC receiver state machine should remain enabled during the power-down mode. This
means that the RE bit has to remain set in the ETH_MACCR register because it is involved
in magic packet/ wake-on-LAN frame detection. The transmit state machine should however
be turned off during the power-down mode by clearing the TE bit in the ETH_MACCR
register. Moreover, the Ethernet DMA should be disabled during the power-down mode,
because it is not necessary to copy the magic packet/wake-on-LAN frame into the SRAM.
To disable the Ethernet DMA, clear the ST bit and the SR bit (for the transmit DMA and the
receive DMA, respectively) in the ETH_DMAOMR register.
The recommended power-down and wakeup sequences are as follows:
1156/1751 RM0090 Rev 19

1. Disable the transmit DMA and wait for any previous frame transmissions to complete.
These transmissions can be detected when the transmit interrupt ETH_DMASR
register[0] is received.
2. Disable the MAC transmitter and MAC receiver by clearing the RE and TE bits in the
ETH_MACCR configuration register.
3. Wait for the receive DMA to have emptied all the frames in the Rx FIFO.
4. Disable the receive DMA.
5. Configure and enable the EXTI line 19 to generate either an event or an interrupt.
6. If you configure the EXTI line 19 to generate an interrupt, you also have to correctly
configure the ETH_WKUP_IRQ Handler function, which should clear the pending bit of
the EXTI line 19.
7. Enable Magic packet/Wake-on-LAN frame detection by setting the MFE/ WFE bit in the
ETH_MACPMTCSR register.
8. Enable the MAC power-down mode, by setting the PD bit in the ETH_MACPMTCSR
register.
9. Enable the MAC Receiver by setting the RE bit in the ETH_MACCR register.
10. Enter the system’s Stop mode (for more details refer to Section 5.3.4: Stop mode):
11. On receiving a valid wakeup frame, the Ethernet peripheral exits the power-down
mode.
12. Read the ETH_MACPMTCSR to clear the power management event flag, enable the
MAC transmitter state machine, and the receive and transmit DMA.
13. Configure the system clock: enable the HSE and set the clocks.
33.5.9 Precision time protocol (IEEE1588 PTP)

The IEEE 1588 standard defines a protocol that allows precise clock synchronization in
measurement and control systems implemented with technologies such as network
communication, local computing and distributed objects. The protocol applies to systems
that communicate by local area networks supporting multicast messaging, including (but not
limited to) Ethernet. This protocol is used to synchronize heterogeneous systems that
include clocks of varying inherent precision, resolution and stability. The protocol supports
system-wide synchronization accuracy in the submicrosecond range with minimum network
and local clock computing resources. The message-based protocol, known as the precision
time protocol (PTP), is transported over UDP/IP. The system or network is classified into
Master and Slave nodes for distributing the timing/clock information. The protocol’s
technique for synchronizing a slave node to a master node by exchanging PTP messages is
described in Figure 372.
RM0090 Rev 19 1157/1751

1239
Figure 372. Networked time synchronization
Master clock time Slave clock time
t1 Sync message
Data at
slave clock
t2m t2 t2
Follow_up message
containing value of t1
t1, t2
t3m Delay_Req message t3 t1, t2, t3
t4
Delay_Resp message
containing value of t4
t1, t2, t3, t4

time
ai15669
1. The master broadcasts PTP Sync messages to all its nodes. The Sync message
contains the master’s reference time information. The time at which this message
leaves the master’s system is t1. For Ethernet ports, this time has to be captured at the
MII.
2. A slave receives the Sync message and also captures the exact time, t2, using its
timing reference.
3. The master then sends the slave a Follow_up message, which contains the t1
information for later use.
4. The slave sends the master a Delay_Req message, noting the exact time, t3, at which
this frame leaves the MII.
5. The master receives this message and captures the exact time, t4, at which it enters its
system.
6. The master sends the t4 information to the slave in the Delay_Resp message.
7. The slave uses the four values of t1, t2, t3, and t4 to synchronize its local timing
reference to the master’s timing reference.
Most of the protocol implementation occurs in the software, above the UDP layer. As
described above, however, hardware support is required to capture the exact time when
specific PTP packets enter or leave the Ethernet port at the MII. This timing information has
to be captured and returned to the software for a proper, high-accuracy implementation of
PTP.
Reference timing source

To get a snapshot of the time, the core requires a reference time in 64-bit format (split into
two 32-bit channels, with the upper 32 bits providing time in seconds, and the lower 32 bits
indicating time in nanoseconds) as defined in the IEEE 1588 specification.
The PTP reference clock input is used to internally generate the reference time (also called
the System Time) and to capture time stamps. The frequency of this reference clock must
1158/1751 RM0090 Rev 19

be greater than or equal to the resolution of time stamp counter. The synchronization
accuracy target between the master node and the slaves is around 100 ns.
The generation, update and modification of the System Time are described in System Time
correction methods.
The accuracy depends on the PTP reference clock input period, the characteristics of the
oscillator (drift) and the frequency of the synchronization procedure.
Due to the synchronization from the Tx and Rx clock input domain to the PTP reference
clock domain, the uncertainty on the time stamp latched value is 1 reference clock period. If
we add the uncertainty due to resolution, we will add half the period for time stamping.
Transmission of frames with the PTP feature

When a frame’s SFD is output on the MII, a time stamp is captured. Frames for which time
stamp capture is required are controllable on a per-frame basis. In other words, each
transmitted frame can be marked to indicate whether a time stamp must be captured or not
for that frame. The transmitted frames are not processed to identify PTP frames. Frame
control is exercised through the control bits in the transmit descriptor. Captured time stamps
are returned to the application in the same way as the status is provided for frames. The
time stamp is sent back along with the Transmit status of the frame, inside the
corresponding transmit descriptor, thus connecting the time stamp automatically to the
specific PTP frame. The 64-bit time stamp information is written back to the TDES2 and
TDES3 fields, with TDES2 holding the time stamp’s 32 least significant bits.
Reception of frames with the PTP feature

When the IEEE 1588 time stamping feature is enabled, the Ethernet MAC captures the time
stamp of all frames received on the MII. The MAC provides the time stamp as soon as the
frame reception is complete. Captured time stamps are returned to the application in the
same way as the frame status is provided. The time stamp is sent back along with the
Receive status of the frame, inside the corresponding receive descriptor. The 64-bit time
stamp information is written back to the RDES2 and RDES3 fields, with RDES2 holding the
time stamp’s 32 least significant bits.
System Time correction methods

The 64-bit PTP time is updated using the PTP input reference clock, HCLK. This PTP time
is used as a source to take snapshots (time stamps) of the Ethernet frames being
transmitted or received at the MII. The System Time counter can be initialized or corrected
using either the Coarse or the Fine correction method.
In the Coarse correction method, the initial value or the offset value is written to the Time
stamp update register (refer to Section 33.8.3). For initialization, the System Time counter is
written with the value in the Time stamp update registers, whereas for system time
correction, the offset value (Time stamp update register) is added to or subtracted from the
system time.
In the Fine correction method, the slave clock (reference clock) frequency drift with respect
to the master clock (as defined in IEEE 1588) is corrected over a period of time, unlike in the
Coarse correction method where it is corrected in a single clock cycle. The longer correction
time helps maintain linear time and does not introduce drastic changes (or a large jitter) in
the reference time between PTP Sync message intervals. In this method, an accumulator
sums up the contents of the Addend register as shown in Figure 373. The arithmetic carry
that the accumulator generates is used as a pulse to increment the system time counter.
RM0090 Rev 19 1159/1751

1239
The accumulator and the addend are 32-bit registers. Here, the accumulator acts as a high-
precision frequency multiplier or divider. Figure 373 shows this algorithm.
Figure 373. System time update using the Fine correction method
Addend register
Addend update
+
Accumulator register
Constant value
Increment Subsecond
register
+
Subsecond register
Increment Second register
Second register
ai15670
The system time update logic requires a 50 MHz clock frequency to achieve 20 ns accuracy.
The frequency division is the ratio of the reference clock frequency to the required clock
frequency. Hence, if the reference clock (HCLK) is, let us say, 66 MHz, the ratio is calculated
as 66 MHz/50 MHz = 1.32. Hence, the default addend value to be set in the register is
232/1.32, which is equal to 0xC1F0 7C1F.
If the reference clock drifts lower, to 65 MHz for example, the ratio is 65/50 or 1.3 and the
value to set in the addend register is 232/1.30 equal to 0xC4EC 4EC4. If the clock drifts
higher, to 67 MHz for example, the addend register must be set to 0xBF0 B7672. When the
clock drift is zero, the default addend value of 0xC1F0 7C1F (232/1.32) should be
programmed.
In Figure 373, the constant value used to increment the subsecond register is 0d43. This
makes an accuracy of 20 ns in the system time (in other words, it is incremented by 20 ns
steps).
The software has to calculate the drift in frequency based on the Sync messages, and to
update the Addend register accordingly. Initially, the slave clock is set with
FreqCompensationValue0 in the Addend register. This value is as follows:
FreqCompensationValue0 = 232 / FreqDivisionRatio
If MasterToSlaveDelay is initially assumed to be the same for consecutive Sync messages,
the algorithm described below must be applied. After a few Sync cycles, frequency lock
occurs. The slave clock can then determine a precise MasterToSlaveDelay value and re-
synchronize with the master using the new value.
1160/1751 RM0090 Rev 19

The algorithm is as follows:

• At time MasterSyncTime (n) the master sends the slave clock a Sync message. The
slave receives this message when its local clock is SlaveClockTime (n) and computes
MasterClockTime (n) as:
MasterClockTime (n) = MasterSyncTime (n) + MasterToSlaveDelay (n)
• The master clock count for current Sync cycle, MasterClockCount (n) is given by:
MasterClockCount (n) = MasterClockTime (n) – MasterClockTime (n – 1) (assuming
that MasterToSlaveDelay is the same for Sync cycles n and n – 1)
• The slave clock count for current Sync cycle, SlaveClockCount (n) is given by:
SlaveClockCount (n) = SlaveClockTime (n) – SlaveClockTime (n – 1)
• The difference between master and slave clock counts for current Sync cycle,
ClockDiffCount (n) is given by:
ClockDiffCount (n) = MasterClockCount (n) – SlaveClockCount (n)
• The frequency-scaling factor for slave clock, FreqScaleFactor (n) is given by:
FreqScaleFactor (n) = (MasterClockCount (n) + ClockDiffCount (n)) /
SlaveClockCount (n)
• The frequency compensation value for Addend register, FreqCompensationValue (n) is
given by:
FreqCompensationValue (n) = FreqScaleFactor (n) × FreqCompensationValue (n – 1)
In theory, this algorithm achieves lock in one Sync cycle; however, it may take several
cycles, due to changing network propagation delays and operating conditions.
This algorithm is self-correcting: if for any reason the slave clock is initially set to a value
from the master that is incorrect, the algorithm corrects it at the cost of more Sync cycles.
Programming steps for system time generation initialization

The time stamping feature can be enabled by setting bit 0 in the Time stamp control register
(ETH__PTPTSCR). However, it is essential to initialize the time stamp counter after this bit
is set to start time stamp operation. The proper sequence is the following:
1. Mask the Time stamp trigger interrupt by setting bit 9 in the MACIMR register.
2. Program Time stamp register bit 0 to enable time stamping.
3. Program the Subsecond increment register based on the PTP clock frequency.
4. If you are using the Fine correction method, program the Time stamp addend register
and set Time stamp control register bit 5 (addend register update).
5. Poll the Time stamp control register until bit 5 is cleared.
6. To select the Fine correction method (if required), program Time stamp control register
bit 1.
7. Program the Time stamp high update and Time stamp low update registers with the
appropriate time value.
8. Set Time stamp control register bit 2 (Time stamp init).
9. The Time stamp counter starts operation as soon as it is initialized with the value
written in the Time stamp update register.
10. Enable the MAC receiver and transmitter for proper time stamping.
Note: If time stamp operation is disabled by clearing bit 0 in the ETH_PTPTSCR register, the
above steps must be repeated to restart the time stamp operation.
RM0090 Rev 19 1161/1751

1239
Programming steps for system time update in the Coarse correction method
To synchronize or update the system time in one process (coarse correction method),
perform the following steps:
1. Write the offset (positive or negative) in the Time stamp update high and low registers.
2. Set bit 3 (TSSTU) in the Time stamp control register.
3. The value in the Time stamp update registers is added to or subtracted from the system
time when the TSSTU bit is cleared.
Programming steps for system time update in the Fine correction method
To synchronize or update the system time to reduce system-time jitter (fine correction
method), perform the following steps:
1. With the help of the algorithm explained in System Time correction methods, calculate
the rate by which you want to speed up or slow down the system time increments.
2. Update the time stamp.
3. Wait the time you want the new value of the Addend register to be active. You can do
this by activating the Time stamp trigger interrupt after the system time reaches the
target value.
4. Program the required target time in the Target time high and low registers. Unmask the
Time stamp interrupt by clearing bit 9 in the ETH_MACIMR register.
5. Set Time stamp control register bit 4 (TSARU).
6. When this trigger causes an interrupt, read the ETH_MACSR register.
7. Reprogram the Time stamp addend register with the old value and set ETH_TPTSCR
bit 5 again.
PTP trigger internal connection with TIM2

The MAC provides a trigger interrupt when the system time becomes greater than the target
time. Using an interrupt introduces a known latency plus an uncertainty in the command
execution time.
In order to avoid this uncertainty, a PTP trigger output signal is set high when the system
time is greater than the target time. It is internally connected to the TIM2 input trigger. With
this signal, the input capture feature, the output compare feature and the waveforms of the
timer can be used, triggered by the synchronized PTP system time. No uncertainty is
introduced since the clock of the timer (PCLK1: TIM2 APB1 clock) and PTP reference clock
(HCLK) are synchronous.
This PTP trigger signal is connected to the TIM2 ITR1 input selectable by software. The
connection is enabled through bits 11 and 10 in the TIM2 option register (TIM2_OR).
Figure 374 shows the connection.
Figure 374. PTP trigger output to TIM2 ITR1 connection
PTP trigger ITR1

Ethernet MAC TIM2
ai15671
1162/1751 RM0090 Rev 19

PTP pulse-per-second output signal

This PTP pulse output is used to check the synchronization between all nodes in the
network. To be able to test the difference between the local slave clock and the master
reference clock, both clocks were given a pulse-per-second (PPS) output signal that may be
connected to an oscilloscope if necessary. The deviation between the two signals can
therefore be measured. The pulse width of the PPS output is 125 ms.
The PPS output is enabled through a GPIO alternate function. (GPIO_AFR register).
The default frequency of the PPS output is 1 Hz. PPSFREQ[3:0] (in ETH_PTPPPSCR) can
be used to set the frequency of the PPS output to 2PPSFREQ Hz.
When set to 1 Hz, the PPS pulse width is 125 ms with binary rollover (TSSSR=0, bit 9 in
ETH_PTPTSCR) and 100 ms with digital rollover (TSSSR=1). When set to 2 Hz and higher,
the duty cycle of the PPS output is 50% with binary rollover.
With digital rollover (TSSSR=1), it is recommended not to use the PPS output with a
frequency other than 1 Hz as it would have irregular waveforms (though its average
frequency would always be correct during any one-second window).
Figure 375. PPS output
PPS output
Ethernet MAC
ai15672
33.6 Ethernet functional description: DMA controller operation

The DMA has independent transmit and receive engines, and a CSR space. The transmit
engine transfers data from system memory into the Tx FIFO while the receive engine
transfers data from the Rx FIFO into system memory. The controller utilizes descriptors to
efficiently move data from source to destination with minimum CPU intervention. The DMA
is designed for packet-oriented data transfers such as frames in Ethernet. The controller can
be programmed to interrupt the CPU in cases such as frame transmit and receive transfer
completion, and other normal/error conditions. The DMA and the STM32F4xx communicate
through two data structures:
• Control and status registers (CSR)
• Descriptor lists and data buffers.
Control and status registers are described in detail in Section 33.8. Descriptors are
described in detail in Normal Tx DMA descriptors.
The DMA transfers the received data frames to the receive buffer in the STM32F4xx
memory, and transmits data frames from the transmit buffer in the STM32F4xx memory.
Descriptors that reside in the STM32F4xx memory act as pointers to these buffers. There
are two descriptor lists: one for reception, and one for transmission. The base address of
each list is written into DMA registers 3 and 4, respectively. A descriptor list is forward-linked
(either implicitly or explicitly). The last descriptor may point back to the first entry to create a
ring structure. Explicit chaining of descriptors is accomplished by configuring the second
address chained in both the receive and transmit descriptors (RDES1[14] and TDES0[20]).
The descriptor lists reside in the Host’s physical memory space. Each descriptor can point to
a maximum of two buffers. This enables the use of two physically addressed buffers,
RM0090 Rev 19 1163/1751

1239
instead of two contiguous buffers in memory. A data buffer resides in the Host’s physical
memory space, and consists of an entire frame or part of a frame, but cannot exceed a
single frame. Buffers contain only data. The buffer status is maintained in the descriptor.
Data chaining refers to frames that span multiple data buffers. However, a single descriptor
cannot span multiple frames. The DMA skips to the next frame buffer when the end of frame
is detected. Data chaining can be enabled or disabled. The descriptor ring and chain
structure is shown in Figure 376.
Figure 376. Descriptor ring and chain structure

Ring structure Chain structure
Buffer 1 Buffer 1
Descriptor 0 Descriptor 0
Buffer 2
Buffer 1
Descriptor 1
Buffer 2 Buffer 1
Descriptor 1
Buffer 1
Descriptor 2
Buffer 2
Descriptor 2 Buffer 1
Buffer 1
Descriptor n
Buffer 2
Next descriptor
ai15638
33.6.1 Initialization of a transfer using DMA

Initialization for the MAC is as follows:
1. Write to ETH_DMABMR to set STM32F4xx bus access parameters.
2. Write to the ETH_DMAIER register to mask unnecessary interrupt causes.
3. The software driver creates the transmit and receive descriptor lists. Then it writes to
both the ETH_DMARDLAR and ETH_DMATDLAR registers, providing the DMA with
the start address of each list.
4. Write to MAC registers 1, 2, and 3 to choose the desired filtering options.
5. Write to the MAC ETH_MACCR register to configure and enable the transmit and
receive operating modes. The PS and DM bits are set based on the auto-negotiation
result (read from the PHY).
6. Write to the ETH_DMAOMR register to set bits 13 and 1 and start transmission and
reception.
7. The transmit and receive engines enter the running state and attempt to acquire
descriptors from the respective descriptor lists. The receive and transmit engines then
begin processing receive and transmit operations. The transmit and receive processes
are independent of each other and can be started or stopped separately.
33.6.2 Host bus burst access

The DMA attempts to execute fixed-length burst transfers on the AHB master interface if
configured to do so (FB bit in ETH_DMABMR). The maximum burst length is indicated and
limited by the PBL field (ETH_DMABMR [13:8]). The receive and transmit descriptors are
1164/1751 RM0090 Rev 19

always accessed in the maximum possible burst size (limited by PBL) for the 16 bytes to be
read.
The Transmit DMA initiates a data transfer only when there is sufficient space in the
Transmit FIFO to accommodate the configured burst or the number of bytes until the end of
frame (when it is less than the configured burst length). The DMA indicates the start address
and the number of transfers required to the AHB Master Interface. When the AHB Interface
is configured for fixed-length burst, then it transfers data using the best combination of
INCR4, INCR8, INCR16 and SINGLE transactions. Otherwise (no fixed-length burst), it
transfers data using INCR (undefined length) and SINGLE transactions.
The Receive DMA initiates a data transfer only when sufficient data for the configured burst
is available in Receive FIFO or when the end of frame (when it is less than the configured
burst length) is detected in the Receive FIFO. The DMA indicates the start address and the
number of transfers required to the AHB master interface. When the AHB interface is
configured for fixed-length burst, then it transfers data using the best combination of INCR4,
INCR8, INCR16 and SINGLE transactions. If the end of frame is reached before the fixed-
burst ends on the AHB interface, then dummy transfers are performed in order to complete
the fixed-length burst. Otherwise (FB bit in ETH_DMABMR is reset), it transfers data using
INCR (undefined length) and SINGLE transactions.
When the AHB interface is configured for address-aligned beats, both DMA engines ensure
that the first burst transfer the AHB initiates is less than or equal to the size of the configured
PBL. Thus, all subsequent beats start at an address that is aligned to the configured PBL.
The DMA can only align the address for beats up to size 16 (for PBL > 16), because the
AHB interface does not support more than INCR16.
33.6.3 Host data buffer alignment

The transmit and receive data buffers do not have any restrictions on start address
alignment. In our system with 32-bit memory, the start address for the buffers can be aligned
to any of the four bytes. However, the DMA always initiates transfers with address aligned to
the bus width with dummy data for the byte lanes not required. This typically happens during
the transfer of the beginning or end of an Ethernet frame.
• Example of buffer read:
If the Transmit buffer address is 0x0000 0FF2, and 15 bytes need to be transferred,
then the DMA will read five full words from address 0x0000 0FF0, but when transferring
data to the Transmit FIFO, the extra bytes (the first two bytes) will be dropped or
ignored. Similarly, the last 3 bytes of the last transfer will also be ignored. The DMA
always ensures it transfers a full 32-bit data items to the Transmit FIFO, unless it is the
end of frame.
• Example of buffer write:
If the Receive buffer address is 0x0000 0FF2, and 16 bytes of a received frame need to
be transferred, then the DMA will write five full 32-bit data items from address
0x0000 0FF0. But the first 2 bytes of the first transfer and the last 2 bytes of the third
transfer will have dummy data.
33.6.4 Buffer size calculations

The DMA does not update the size fields in the transmit and receive descriptors. The DMA
updates only the status fields (xDES0) of the descriptors. The driver has to calculate the
sizes. The transmit DMA transfers the exact number of bytes (indicated by buffer size field in
TDES1) towards the MAC core. If a descriptor is marked as first (FS bit in TDES0 is set),
RM0090 Rev 19 1165/1751

1239
then the DMA marks the first transfer from the buffer as the start of frame. If a descriptor is
marked as last (LS bit in TDES0), then the DMA marks the last transfer from that data buffer
as the end of frame. The receive DMA transfers data to a buffer until the buffer is full or the
end of frame is received. If a descriptor is not marked as last (LS bit in RDES0), then the
buffer(s) that correspond to the descriptor are full and the amount of valid data in a buffer is
accurately indicated by the buffer size field minus the data buffer pointer offset when the
descriptor’s FS bit is set. The offset is zero when the data buffer pointer is aligned to the
databus width. If a descriptor is marked as last, then the buffer may not be full (as indicated
by the buffer size in RDES1). To compute the amount of valid data in this final buffer, the
driver must read the frame length (FL bits in RDES0[29:16]) and subtract the sum of the
buffer sizes of the preceding buffers in this frame. The receive DMA always transfers the
start of next frame with a new descriptor.
Note: Even when the start address of a receive buffer is not aligned to the system databus width
the system should allocate a receive buffer of a size aligned to the system bus width. For
example, if the system allocates a 1024 byte (1 KB) receive buffer starting from address
0x1000, the software can program the buffer start address in the receive descriptor to have
a 0x1002 offset. The receive DMA writes the frame to this buffer with dummy data in the first
two locations (0x1000 and 0x1001). The actual frame is written from location 0x1002. Thus,
the actual useful space in this buffer is 1022 bytes, even though the buffer size is
programmed as 1024 bytes, due to the start address offset.
33.6.5 DMA arbiter

The arbiter inside the DMA takes care of the arbitration between transmit and receive
channel accesses to the AHB master interface. Two types of arbitrations are possible:
round-robin, and fixed-priority. When round-robin arbitration is selected (DA bit in
ETH_DMABMR is reset), the arbiter allocates the databus in the ratio set by the PM bits in
ETH_DMABMR, when both transmit and receive DMAs request access simultaneously.
When the DA bit is set, the receive DMA always gets priority over the transmit DMA for data
access.
33.6.6 Error response to DMA

For any data transfer initiated by a DMA channel, if the slave replies with an error response,
that DMA stops all operations and updates the error bits and the fatal bus error bit in the
Status register (ETH_DMASR register). That DMA controller can resume operation only
after soft- or hard-resetting the peripheral and re-initializing the DMA.
33.6.7 Tx DMA configuration

TxDMA operation: default (non-OSF) mode
The transmit DMA engine in default mode proceeds as follows:
1. The user sets up the transmit descriptor (TDES0-TDES3) and sets the OWN bit
(TDES0[31]) after setting up the corresponding data buffer(s) with Ethernet frame data.
2. Once the ST bit (ETH_DMAOMR register[13]) is set, the DMA enters the Run state.
1166/1751 RM0090 Rev 19

3. While in the Run state, the DMA polls the transmit descriptor list for frames requiring
transmission. After polling starts, it continues in either sequential descriptor ring order
or chained order. If the DMA detects a descriptor flagged as owned by the CPU, or if an
error condition occurs, transmission is suspended and both the Transmit buffer
Unavailable (ETH_DMASR register[2]) and Normal Interrupt Summary (ETH_DMASR
register[16]) bits are set. The transmit engine proceeds to Step 9.
4. If the acquired descriptor is flagged as owned by DMA (TDES0[31] is set), the DMA
decodes the transmit data buffer address from the acquired descriptor.
5. The DMA fetches the transmit data from the STM32F4xx memory and transfers the
data.
6. If an Ethernet frame is stored over data buffers in multiple descriptors, the DMA closes
the intermediate descriptor and fetches the next descriptor. Steps 3, 4, and 5 are
repeated until the end of Ethernet frame data is transferred.
7. When frame transmission is complete, if IEEE 1588 time stamping was enabled for the
frame (as indicated in the transmit status) the time stamp value is written to the transmit
descriptor (TDES2 and TDES3) that contains the end-of-frame buffer. The status
information is then written to this transmit descriptor (TDES0). Because the OWN bit is
cleared during this step, the CPU now owns this descriptor. If time stamping was not
enabled for this frame, the DMA does not alter the contents of TDES2 and TDES3.
8. Transmit Interrupt (ETH_DMASR register [0]) is set after completing the transmission
of a frame that has Interrupt on Completion (TDES1[31]) set in its last descriptor. The
DMA engine then returns to Step 3.
9. In the Suspend state, the DMA tries to re-acquire the descriptor (and thereby returns to
Step 3) when it receives a transmit poll demand, and the Underflow Interrupt Status bit
is cleared.
Figure 377 shows the TxDMA transmission flow in default mode.
RM0090 Rev 19 1167/1751

1239
Figure 377. TxDMA operation in Default mode
Start TxDMA Start Stop TxDMA
(Re-)fetch next
descriptor
Poll demand (AHB) Yes

error?
No
TxDMA suspended No
Own
bit set?
Yes
Transfer data from

buffer(s)
(AHB) Yes
error?
No
No
Frame xfer
complete?
Yes
Close intermediate
descriptor Wait for Tx status
Time stamp Yes Write time stamp to

present? TDES2 and TDES3
No
Write status word No

(AHB) Yes
to TDES0 error?
No
(AHB) Yes
error?
ai15639
TxDMA operation: OSF mode

While in the Run state, the transmit process can simultaneously acquire two frames without
closing the Status descriptor of the first (if the OSF bit is set in ETH_DMAOMR register[2]).
As the transmit process finishes transferring the first frame, it immediately polls the transmit
descriptor list for the second frame. If the second frame is valid, the transmit process
transfers this frame before writing the first frame’s status information. In OSF mode, the
Run-state transmit DMA operates according to the following sequence:
1168/1751 RM0090 Rev 19

1. The DMA operates as described in steps 1–6 of the TxDMA (default mode).
2. Without closing the previous frame’s last descriptor, the DMA fetches the next
descriptor.
3. If the DMA owns the acquired descriptor, the DMA decodes the transmit buffer address
in this descriptor. If the DMA does not own the descriptor, the DMA goes into Suspend
mode and skips to Step 7.
4. The DMA fetches the Transmit frame from the STM32F4xx memory and transfers the
frame until the end of frame data are transferred, closing the intermediate descriptors if
this frame is split across multiple descriptors.
5. The DMA waits for the transmission status and time stamp of the previous frame. When
the status is available, the DMA writes the time stamp to TDES2 and TDES3, if such
time stamp was captured (as indicated by a status bit). The DMA then writes the status,
with a cleared OWN bit, to the corresponding TDES0, thus closing the descriptor. If
time stamping was not enabled for the previous frame, the DMA does not alter the
contents of TDES2 and TDES3.
6. If enabled, the Transmit interrupt is set, the DMA fetches the next descriptor, then
proceeds to Step 3 (when Status is normal). If the previous transmission status shows
an underflow error, the DMA goes into Suspend mode (Step 7).
7. In Suspend mode, if a pending status and time stamp are received by the DMA, it
writes the time stamp (if enabled for the current frame) to TDES2 and TDES3, then
writes the status to the corresponding TDES0. It then sets relevant interrupts and
returns to Suspend mode.
8. The DMA can exit Suspend mode and enter the Run state (go to Step 1 or Step 2
depending on pending status) only after receiving a Transmit Poll demand
(ETH_DMATPDR register).
Figure 378 shows the basic flowchart in OSF mode.
RM0090 Rev 19 1169/1751

1239
Figure 378. TxDMA operation in OSF mode
Start TxDMA Start Stop TxDMA
(Re-)fetch next
descriptor
(AHB) Yes
Poll error?
demand
No
TxDMA suspended No Own

bit set?
Yes
Previous frame
status available Transfer data from
buffer(s)
Time stamp (AHB) Yes

present? error?
No No
Yes
No Frame xfer Second

Write time stamp to Yes
complete? frame?
TDES2 & TDES3 No
for previous frame Yes
Close intermediate Wait for previous

descriptor frame’s Tx status
(AHB) Yes
Write time stamp to
error? Time stamp Yes TDES2 & TDES3
present? for previous frame
No
No
Write status word to Write status word to No

(AHB) Yes
prev. frame’s TDES0 prev. frame’s TDES0 error?
(AHB) No (AHB)
No
error? error?
Yes
Yes
ai15640
Transmit frame processing

The transmit DMA expects that the data buffers contain complete Ethernet frames,
excluding preamble, pad bytes, and FCS fields. The DA, SA, and Type/Len fields contain
valid data. If the transmit descriptor indicates that the MAC core must disable CRC or pad
insertion, the buffer must have complete Ethernet frames (excluding preamble), including
the CRC bytes. Frames can be data-chained and span over several buffers. Frames have to
be delimited by the first descriptor (TDES0[28]) and the last descriptor (TDES0[29]). As the
transmission starts, TDES0[28] has to be set in the first descriptor. When this occurs, the
frame data are transferred from the memory buffer to the Transmit FIFO. Concurrently, if the
last descriptor (TDES0[29]) of the current frame is cleared, the transmit process attempts to
acquire the next descriptor. The transmit process expects TDES0[28] to be cleared in this
descriptor. If TDES0[29] is cleared, it indicates an intermediary buffer. If TDES0[29] is set, it
1170/1751 RM0090 Rev 19

indicates the last buffer of the frame. After the last buffer of the frame has been transmitted,
the DMA writes back the final status information to the transmit descriptor 0 (TDES0) word
of the descriptor that has the last segment set in transmit descriptor 0 (TDES0[29]). At this
time, if Interrupt on Completion (TDES0[30]) is set, Transmit Interrupt (in ETH_DMASR
register [0]) is set, the next descriptor is fetched, and the process repeats. Actual frame
transmission begins after the Transmit FIFO has reached either a programmable transmit
threshold (ETH_DMAOMR register[16:14]), or a full frame is contained in the FIFO. There is
also an option for the Store and forward mode (ETH_DMAOMR register[21]). Descriptors
are released (OWN bit TDES0[31] is cleared) when the DMA finishes transferring the frame.
Transmit polling suspended

Transmit polling can be suspended by either of the following conditions:
• The DMA detects a descriptor owned by the CPU (TDES0[31]=0) and the Transmit
buffer unavailable flag is set (ETH_DMASR register[2]). To resume, the driver must
give descriptor ownership to the DMA and then issue a Poll Demand command.
• A frame transmission is aborted when a transmit error due to underflow is detected.
The appropriate Transmit Descriptor 0 (TDES0) bit is set. If the second condition
occurs, both the Abnormal Interrupt Summary (in ETH_DMASR register [15]) and
Transmit Underflow bits (in ETH_DMASR register[5]) are set, and the information is
written to Transmit Descriptor 0, causing the suspension. If the DMA goes into
Suspend state due to the first condition, then both the Normal Interrupt Summary
(ETH_DMASR register [16]) and Transmit Buffer Unavailable (ETH_DMASR
register[2]) bits are set. In both cases, the position in the transmit list is retained. The
retained position is that of the descriptor following the last descriptor closed by the
DMA. The driver must explicitly issue a Transmit Poll Demand command after rectifying
the suspension cause.
Normal Tx DMA descriptors

The normal transmit descriptor structure consists of four 32-bit words as shown in
Figure 379. The bit descriptions of TDES0, TDES1, TDES2 and TDES3 are given below.
Note that enhanced descriptors must be used if time stamping is activated (ETH_PTPTSCR
bit 0, TSE=1) or if IPv4 checksum offload is activated (ETH_MACCR bit 10, IPCO=1).
Figure 379. Normal transmit descriptor

31 0
O T T
Ctrl T Res. Ctrl Reserved T
TDES 0 W Status [16:0]
[30:26] S 24 [23:20] [19:18] S
N
E S
TDES 1 Reserved Buffer 2 byte count Reserved Buffer 1 byte count

[31:29] [28:16] [15:13] [12:0]
TDES 2 Buffer 1 address [31:0] / Time stamp low [31:0]
TDES 3 Buffer 2 address [31:0] or Next descriptor address [31:0] / Time stamp high [31:0]
ai15642b
RM0090 Rev 19 1171/1751

1239
• TDES0: Transmit descriptor Word0

The application software has to program the control bits [30:26]+[23:20] plus the OWN
bit [31] during descriptor initialization. When the DMA updates the descriptor (or writes
it back), it resets all the control bits plus the OWN bit, and reports only the status bits.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
OWN IC LS FS DC DP TTSE CIC TER TCH TTSS IHE
Res Res.
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ES JT FF IPE LCA NC LCO EC VF CC ED UF DB
Bit 31 OWN: Own bit

When set, this bit indicates that the descriptor is owned by the DMA. When this bit is reset, it
indicates that the descriptor is owned by the CPU. The DMA clears this bit either when it
completes the frame transmission or when the buffers allocated in the descriptor are read
completely. The ownership bit of the frame’s first descriptor must be set after all subsequent
descriptors belonging to the same frame have been set.
Bit 30 IC: Interrupt on completion
When set, this bit sets the Transmit Interrupt (ETH_DMASR[0]) after the present frame has
been transmitted.
Bit 29 LS: Last segment
When set, this bit indicates that the buffer contains the last segment of the frame.
Bit 28 FS: First segment
When set, this bit indicates that the buffer contains the first segment of a frame.
Bit 27 DC: Disable CRC
When this bit is set, the MAC does not append a cyclic redundancy check (CRC) to the end
of the transmitted frame. This is valid only when the first segment (TDES0[28]) is set.
Bit 26 DP: Disable pad
When set, the MAC does not automatically add padding to a frame shorter than 64 bytes.
When this bit is reset, the DMA automatically adds padding and CRC to a frame shorter than
64 bytes, and the CRC field is added despite the state of the DC (TDES0[27]) bit. This is
valid only when the first segment (TDES0[28]) is set.
Bit 25 TTSE: Transmit time stamp enable
When TTSE is set and when TSE is set (ETH_PTPTSCR bit 0), IEEE1588 hardware time
stamping is activated for the transmit frame described by the descriptor. This field is only valid
when the First segment control bit (TDES0[28]) is set.
Bits 23:22 CIC: Checksum insertion control
These bits control the checksum calculation and insertion. Bit encoding is as shown below:
00: Checksum Insertion disabled
01: Only IP header checksum calculation and insertion are enabled
10: IP header checksum and payload checksum calculation and insertion are enabled, but
pseudo-header checksum is not calculated in hardware
11: IP Header checksum and payload checksum calculation and insertion are enabled, and
pseudo-header checksum is calculated in hardware.
1172/1751 RM0090 Rev 19

Bit 21 TER: Transmit end of ring

When set, this bit indicates that the descriptor list reached its final descriptor. The DMA
returns to the base address of the list, creating a descriptor ring.
Bit 20 TCH: Second address chained
When set, this bit indicates that the second address in the descriptor is the next descriptor
address rather than the second buffer address. When TDES0[20] is set, TBS2
(TDES1[28:16]) is a “don’t care” value. TDES0[21] takes precedence over TDES0[20].
Bit 17 TTSS: Transmit time stamp status
This field is used as a status bit to indicate that a time stamp was captured for the described
transmit frame. When this bit is set, TDES2 and TDES3 have a time stamp value captured for the
transmit frame. This field is only valid when the descriptor’s Last segment control bit (TDES0[29])
is set.
Note that when enhanced descriptors are enabled (EDFE=1 in ETH_DMABMR), TTSS=1
indicates that TDES6 and TDES7 have the time stamp value.
Bit 16 IHE: IP header error
When set, this bit indicates that the MAC transmitter detected an error in the IP datagram
header. The transmitter checks the header length in the IPv4 packet against the number of
header bytes received from the application and indicates an error status if there is a
mismatch. For IPv6 frames, a header error is reported if the main header length is not 40
bytes. Furthermore, the Ethernet length/type field value for an IPv4 or IPv6 frame must
match the IP header version received with the packet. For IPv4 frames, an error status is
also indicated if the Header Length field has a value less than 0x5.
Bit 15 ES: Error summary
Indicates the logical OR of the following bits:
TDES0[14]: Jabber timeout
TDES0[13]: Frame flush
TDES0[11]: Loss of carrier
TDES0[10]: No carrier
TDES0[9]: Late collision
TDES0[8]: Excessive collision
TDES0[2]:Excessive deferral
TDES0[1]: Underflow error
TDES0[16]: IP header error
TDES0[12]: IP payload error
Bit 14 JT: Jabber timeout
When set, this bit indicates the MAC transmitter has experienced a jabber timeout. This bit is
only set when the MAC configuration register’s JD bit is not set.
Bit 13 FF: Frame flushed
When set, this bit indicates that the DMA/MTL flushed the frame due to a software Flush
command given by the CPU.
Bit 12 IPE: IP payload error
When set, this bit indicates that MAC transmitter detected an error in the TCP, UDP, or ICMP
IP datagram payload. The transmitter checks the payload length received in the IPv4 or IPv6
header against the actual number of TCP, UDP or ICMP packet bytes received from the
application and issues an error status in case of a mismatch.
RM0090 Rev 19 1173/1751

1239
Bit 11 LCA: Loss of carrier

When set, this bit indicates that a loss of carrier occurred during frame transmission (that is,
the MII_CRS signal was inactive for one or more transmit clock periods during frame
transmission). This is valid only for the frames transmitted without collision when the MAC
operates in Half-duplex mode.
Bit 10 NC: No carrier
When set, this bit indicates that the Carrier Sense signal form the PHY was not asserted
during transmission.
Bit 9 LCO: Late collision
When set, this bit indicates that frame transmission was aborted due to a collision occurring
after the collision window (64 byte times, including preamble, in MII mode). This bit is not
valid if the Underflow Error bit is set.
Bit 8 EC: Excessive collision
When set, this bit indicates that the transmission was aborted after 16 successive collisions
while attempting to transmit the current frame. If the RD (Disable retry) bit in the MAC
Configuration register is set, this bit is set after the first collision, and the transmission of the
frame is aborted.
Bit 7 VF: VLAN frame
When set, this bit indicates that the transmitted frame was a VLAN-type frame.
Bits 6:3 CC: Collision count
This 4-bit counter value indicates the number of collisions occurring before the frame was
transmitted. The count is not valid when the Excessive collisions bit (TDES0[8]) is set.
Bit 2 ED: Excessive deferral
When set, this bit indicates that the transmission has ended because of excessive deferral
of over 24 288 bit times if the Deferral check (DC) bit in the MAC Control register is set high.
Bit 1 UF: Underflow error
When set, this bit indicates that the MAC aborted the frame because data arrived late from
the RAM memory. Underflow error indicates that the DMA encountered an empty transmit
buffer while transmitting the frame. The transmission process enters the Suspended state
and sets both Transmit underflow (ETH_DMASR[5]) and Transmit interrupt
(ETH_DMASR[0]).
Bit 0 DB: Deferred bit
When set, this bit indicates that the MAC defers before transmission because of the
presence of the carrier. This bit is valid only in Half-duplex mode.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TBS2 TBS1
Reserved Reserved
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
31:29 Reserved, must be kept at reset value.
1174/1751 RM0090 Rev 19

28:16 TBS2: Transmit buffer 2 size

These bits indicate the second data buffer size in bytes. This field is not valid if TDES0[20] is
set.
12:0 TBS1: Transmit buffer 1 size
These bits indicate the first data buffer byte size, in bytes. If this field is 0, the DMA ignores
this buffer and uses Buffer 2 or the next descriptor, depending on the value of TCH
(TDES0[20]).

TDES2 contains the address pointer to the first buffer of the descriptor or it contains
time stamp data.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TBAP1/TBAP/TTSL
rw
Bits 31:0 TBAP1: Transmit buffer 1 address pointer / Transmit frame time stamp low
These bits have two different functions: they indicate to the DMA the location of data in
memory, and after all data are transferred, the DMA can then use these bits to pass back time
stamp data.
TBAP: When the software makes this descriptor available to the DMA (at the moment that the
OWN bit is set to 1 in TDES0), these bits indicate the physical address of Buffer 1. There is no
limitation on the buffer address alignment. See Host data buffer alignment for further details on
buffer address alignment.
TTSL: Before it clears the OWN bit in TDES0, the DMA updates this field with the 32 least
significant bits of the time stamp captured for the corresponding transmit frame (overwriting
the value for TBAP1). This field has the time stamp only if time stamping is activated for this
frame (see TTSE, TDES0 bit 25) and if the Last segment control bit (LS) in the descriptor is
set.

TDES3 contains the address pointer either to the second buffer of the descriptor or the
next descriptor, or it contains time stamp data.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TBAP2/TBAP2/TTSH
rw
RM0090 Rev 19 1175/1751

1239
Bits 31:0 TBAP2: Transmit buffer 2 address pointer (Next descriptor address) / Transmit frame time
stamp high
These bits have two different functions: they indicate to the DMA the location of data in
memory, and after all data are transferred, the DMA can then use these bits to pass back
time stamp data.
TBAP2: When the software makes this descriptor available to the DMA (at the moment when
the OWN bit is set to 1 in TDES0), these bits indicate the physical address of Buffer 2 when a
descriptor ring structure is used. If the Second address chained (TDES1 [20]) bit is set, this
address contains the pointer to the physical memory where the next descriptor is present. The
buffer address pointer must be aligned to the bus width only when TDES1 [20] is set. (LSBs are
ignored internally.)
TTSH: Before it clears the OWN bit in TDES0, the DMA updates this field with the 32 most
significant bits of the time stamp captured for the corresponding transmit frame (overwriting
the value for TBAP2). This field has the time stamp only if time stamping is activated for this
frame (see TDES0 bit 25, TTSE) and if the Last segment control bit (LS) in the descriptor is
set.
Enhanced Tx DMA descriptors

Enhanced descriptors (enabled with EDFE=1, ETHDMABMR bit 7), must be used if time
stamping is activated (TSE=1, ETH_PTPTSCR bit 0) or if IPv4 checksum offload is
activated (IPCO=1, ETH_MACCR bit 10).
Enhanced descriptors comprise eight 32-bit words, twice the size of normal descriptors.
TDES0, TDES1, TDES2 and TDES3 have the same definitions as for normal transmit
descriptors (refer to Normal Tx DMA descriptors). TDES6 and TDES7 hold the time stamp.
TDES4, TDES5, TDES6 and TDES7 are defined below.
When the Enhanced descriptor mode is selected, the software needs to allocate 32-bytes (8
words) of memory for every descriptor. When time stamping or IPv4 checksum offload are
not being used, the enhanced descriptor format may be disabled and the software can use
normal descriptors with the default size of 16 bytes.
1176/1751 RM0090 Rev 19

Figure 380. Enhanced transmit descriptor

31 0
O T T
Ctrl T Res. Ctrl Reserved T
TDES 0 W Status [16:0]
[30:26] S 24 [23:20] [19:18] S
N
E S
TDES 1 Reserved Buffer 2 byte count Reserved Buffer 1 byte count

[31:29] [28:16] [15:13] [12:0]
TDES 2 Buffer 1 address [31:0]
TDES 3 Buffer 2 address [31:0] or Next descriptor address [31:0]
TDES 4 Reserved
TDES 5 Reserved
TDES 6 Time stamp low [31:0]
TDES 7 Time stamp high [31:0]
ai17105b

Reserved
Reserved
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TTSL
rw
Bits 31:0 TTSL: Transmit frame time stamp low

This field is updated by DMA with the 32 least significant bits of the time stamp captured
for the corresponding transmit frame. This field has the time stamp only if the Last segment
control bit (LS) in the descriptor is set.
RM0090 Rev 19 1177/1751

1239
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TTSH
rw
Bits 31:0 TTSH: Transmit frame time stamp high

This field is updated by DMA with the 32 most significant bits of the time stamp captured for
the corresponding transmit frame. This field has the time stamp only if the Last segment control
bit (LS) in the descriptor is set.
33.6.8 Rx DMA configuration

The Receive DMA engine’s reception sequence is illustrated in Figure 381 and described
below:
1. The CPU sets up Receive descriptors (RDES0-RDES3) and sets the OWN bit
(RDES0[31]).
2. Once the SR (ETH_DMAOMR register[1]) bit is set, the DMA enters the Run state.
While in the Run state, the DMA polls the receive descriptor list, attempting to acquire
free descriptors. If the fetched descriptor is not free (is owned by the CPU), the DMA
enters the Suspend state and jumps to Step 9.
3. The DMA decodes the receive data buffer address from the acquired descriptors.
4. Incoming frames are processed and placed in the acquired descriptor’s data buffers.
5. When the buffer is full or the frame transfer is complete, the Receive engine fetches the
next descriptor.
6. If the current frame transfer is complete, the DMA proceeds to step 7. If the DMA does
not own the next fetched descriptor and the frame transfer is not complete (EOF is not
yet transferred), the DMA sets the Descriptor error bit in RDES0 (unless flushing is
disabled). The DMA closes the current descriptor (clears the OWN bit) and marks it as
intermediate by clearing the Last segment (LS) bit in the RDES1 value (marks it as last
descriptor if flushing is not disabled), then proceeds to step 8. If the DMA owns the next
descriptor but the current frame transfer is not complete, the DMA closes the current
descriptor as intermediate and returns to step 4.
7. If IEEE 1588 time stamping is enabled, the DMA writes the time stamp (if available) to
the current descriptor’s RDES2 and RDES3. It then takes the received frame’s status
and writes the status word to the current descriptor’s RDES0, with the OWN bit cleared
and the Last segment bit set.
8. The Receive engine checks the latest descriptor’s OWN bit. If the CPU owns the
descriptor (OWN bit is at 0) the Receive buffer unavailable bit (in ETH_DMASR
register[7]) is set and the DMA Receive engine enters the Suspended state (step 9). If
the DMA owns the descriptor, the engine returns to step 4 and awaits the next frame.
9. Before the Receive engine enters the Suspend state, partial frames are flushed from
the Receive FIFO (you can control flushing using bit 24 in the ETH_DMAOMR
register).
10. The Receive DMA exits the Suspend state when a Receive Poll demand is given or the
start of next frame is available from the Receive FIFO. The engine proceeds to step 2
and re-fetches the next descriptor.
1178/1751 RM0090 Rev 19

The DMA does not acknowledge accepting the status until it has completed the time stamp
write-back and is ready to perform status write-back to the descriptor. If software has
enabled time stamping through CSR, when a valid time stamp value is not available for the
frame (for example, because the receive FIFO was full before the time stamp could be
written to it), the DMA writes all ones to RDES2 and RDES3. Otherwise (that is, if time
stamping is not enabled), RDES2 and RDES3 remain unchanged.
Figure 381. Receive DMA operation
Start RxDMA Start Stop RxDMA
Poll demand / (Re-)Fetch next

new frame available descriptor
(AHB)
RxDMA suspended Yes
error?
No
Yes
Frame transfer
No Own bit set?
complete?
No Yes
Frame data
Yes Flush disabled ? No
available ?
No Yes
Flush the
Write data to buffer(s) Wait for frame data
remaining frame
(AHB)
Yes
error?
No
Fetch next descriptor
(AHB)
Yes
error?
No
Flush Own bit set Frame transfer

No No
disabled ? for next desc? complete?
No
Yes Yes Yes
Close RDES0 as Time stamp Write time stamp to

Set descriptor error Yes
intermediate descriptor present? RDES2 & RDES3
No
Close RDES0 as last (AHB)

No Yes
descriptor error?
(AHB)
No
error?
Yes
ai15643
RM0090 Rev 19 1179/1751

1239
Receive descriptor acquisition

The receive engine always attempts to acquire an extra descriptor in anticipation of an
incoming frame. Descriptor acquisition is attempted if any of the following conditions is/are
satisfied:
• The receive Start/Stop bit (ETH_DMAOMR register[1]) has been set immediately after
the DMA has been placed in the Run state.
• The data buffer of the current descriptor is full before the end of the frame currently
being transferred
• The controller has completed frame reception, but the current receive descriptor has
not yet been closed.
• The receive process has been suspended because of a CPU-owned buffer
(RDES0[31] = 0) and a new frame is received.
• A Receive poll demand has been issued.
Receive frame processing

The MAC transfers the received frames to the STM32F4xx memory only when the frame
passes the address filter and the frame size is greater than or equal to the configurable
threshold bytes set for the Receive FIFO, or when the complete frame is written to the FIFO
in Store-and-forward mode. If the frame fails the address filtering, it is dropped in the MAC
block itself (unless Receive All ETH_MACFFR [31] bit is set). Frames that are shorter than
64 bytes, because of collision or premature termination, can be purged from the Receive
FIFO. After 64 (configurable threshold) bytes have been received, the DMA block begins
transferring the frame data to the receive buffer pointed to by the current descriptor. The
DMA sets the first descriptor (RDES0[9]) after the DMA AHB Interface becomes ready to
receive a data transfer (if DMA is not fetching transmit data from the memory), to delimit the
frame. The descriptors are released when the OWN (RDES0[31]) bit is reset to 0, either as
the data buffer fills up or as the last segment of the frame is transferred to the receive buffer.
If the frame is contained in a single descriptor, both the last descriptor (RDES0[8]) and first
descriptor (RDES0[9]) bits are set. The DMA fetches the next descriptor, sets the last
descriptor (RDES0[8]) bit, and releases the RDES0 status bits in the previous frame
descriptor. Then the DMA sets the receive interrupt bit (ETH_DMASR register [6]). The
same process repeats unless the DMA encounters a descriptor flagged as being owned by
the CPU. If this occurs, the receive process sets the receive buffer unavailable bit
(ETH_DMASR register[7]) and then enters the Suspend state. The position in the receive
list is retained.
Receive process suspended

If a new receive frame arrives while the receive process is in Suspend state, the DMA re-
fetches the current descriptor in the STM32F4xx memory. If the descriptor is now owned by
the DMA, the receive process re-enters the Run state and starts frame reception. If the
descriptor is still owned by the host, by default, the DMA discards the current frame at the
top of the Rx FIFO and increments the missed frame counter. If more than one frame is
stored in the Rx FIFO, the process repeats. The discarding or flushing of the frame at the
top of the Rx FIFO can be avoided by setting the DMA Operation mode register bit 24
(DFRF). In such conditions, the receive process sets the receive buffer unavailable status
bit and returns to the Suspend state.
1180/1751 RM0090 Rev 19

Normal Rx DMA descriptors

The normal receive descriptor structure consists of four 32-bit words (16 bytes). These are
shown in Figure 382. The bit descriptions of RDES0, RDES1, RDES2 and RDES3 are given
below.
Note that enhanced descriptors must be used if time stamping is activated (TSE=1,
ETH_PTPTSCR bit 0) or if IPv4 checksum offload is activated (IPCO=1, ETH_MACCR bit
10).
Figure 382. Normal Rx DMA descriptor structure

31 0
O
RDES 0 W Status [30:0]
N
RDES 1 CT Reserved Buffer 2 byte count CTRL Res. Buffer 1 byte count
RL [30:29] [28:16] [15:14] [12:0]
RDES 2 Buffer 1 address [31:0]
RDES 3 Buffer 2 address [31:0] or Next descriptor address [31:0]
ai15644
• RDES0: Receive descriptor Word0

RDES0 contains the received frame status, the frame length and the descriptor
ownership information.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
IPHCE/TSV
PCE/ESA
VLAN
OWN
RWT
AFM
LCO
SAF
OE
DE
RE
DE
CE
ES
FS
FT
LE
LS
FL
rw
Bit 31 OWN: Own bit

When set, this bit indicates that the descriptor is owned by the DMA of the MAC Subsystem.
When this bit is reset, it indicates that the descriptor is owned by the Host. The DMA clears this bit
either when it completes the frame reception or when the buffers that are associated with this
descriptor are full.
Bit 30 AFM: Destination address filter fail
When set, this bit indicates a frame that failed the DA filter in the MAC Core.
Bits 29:16 FL: Frame length
These bits indicate the byte length of the received frame that was transferred to host memory
(including CRC). This field is valid only when last descriptor (RDES0[8]) is set and descriptor error
(RDES0[14]) is reset.
This field is valid when last descriptor (RDES0[8]) is set. When the last descriptor and error
summary bits are not set, this field indicates the accumulated number of bytes that have been
transferred for the current frame.
RM0090 Rev 19 1181/1751

1239
Bit 15 ES: Error summary

Indicates the logical OR of the following bits:
RDES0[1]: CRC error
RDES0[3]: Receive error
RDES0[4]: Watchdog timeout
RDES0[6]: Late collision
RDES0[7]: Giant frame (This is not applicable when RDES0[7] indicates an IPV4 header
checksum error.)
RDES0[11]: Overflow error
RDES0[14]: Descriptor error.
This field is valid only when the last descriptor (RDES0[8]) is set.
Bit 14 DE: Descriptor error
When set, this bit indicates a frame truncation caused by a frame that does not fit within the current
descriptor buffers, and that the DMA does not own the next descriptor. The frame is truncated.
This field is valid only when the last descriptor (RDES0[8]) is set.
Bit 13 SAF: Source address filter fail
When set, this bit indicates that the SA field of frame failed the SA filter in the MAC Core.
Bit 12 LE: Length error
When set, this bit indicates that the actual length of the received frame does not match the value in
the Length/ Type field. This bit is valid only when the Frame type (RDES0[5]) bit is reset.
Bit 11 OE: Overflow error
When set, this bit indicates that the received frame was damaged due to buffer overflow.
Bit 10 VLAN: VLAN tag
When set, this bit indicates that the frame pointed to by this descriptor is a VLAN frame tagged
by the MAC core.
Bit 9 FS: First descriptor
When set, this bit indicates that this descriptor contains the first buffer of the frame. If the size of the
first buffer is 0, the second buffer contains the beginning of the frame. If the size of the second buffer
is also 0, the next descriptor contains the beginning of the frame.
Bit 8 LS: Last descriptor
When set, this bit indicates that the buffers pointed to by this descriptor are the last buffers of
the frame.
Bit 7 IPHCE/TSV: IPv header checksum error / time stamp valid
If IPHCE is set, it indicates an error in the IPv4 or IPv6 header. This error can be due to inconsistent
Ethernet Type field and IP header Version field values, a header checksum mismatch in IPv4, or an
Ethernet frame lacking the expected number of IP header bytes. This bit can take on special
meaning as specified in Table 193.
If enhanced descriptor format is enabled (EDFE=1, bit 7 of ETH_DMABMR), this bit takes on
the TSV function (otherwise it is IPHCE). When TSV is set, it indicates that a snapshot of the
timestamp is written in descriptor words 6 (RDES6) and 7 (RDES7). TSV is valid only when
the Last descriptor bit (RDES0[8]) is set.
Bit 6 LCO: Late collision
When set, this bit indicates that a late collision has occurred while receiving the frame in Half-
duplex mode.
1182/1751 RM0090 Rev 19

Bit 5 FT: Frame type

When set, this bit indicates that the Receive frame is an Ethernet-type frame (the LT field is greater
than or equal to 0x0600). When this bit is reset, it indicates that the received frame is an
IEEE802.3 frame. This bit is not valid for Runt frames less than 14 bytes. When the normal
descriptor format is used (ETH_DMABMR EDFE=0), FT can take on special meaning as
specified in Table 193.
Bit 4 RWT: Receive watchdog timeout
When set, this bit indicates that the Receive watchdog timer has expired while receiving the
current frame and the current frame is truncated after the watchdog timeout.
Bit 3 RE: Receive error
When set, this bit indicates that the RX_ERR signal is asserted while RX_DV is asserted
during frame reception.
Bit 2 DE: Dribble bit error
When set, this bit indicates that the received frame has a non-integer multiple of bytes (odd
nibbles). This bit is valid only in MII mode.
Bit 1 CE: CRC error
When set, this bit indicates that a cyclic redundancy check (CRC) error occurred on the
received frame. This field is valid only when the last descriptor (RDES0[8]) is set.
Bit 0 PCE/ESA: Payload checksum error / extended status available
When set, it indicates that the TCP, UDP or ICMP checksum the core calculated does not
match the received encapsulated TCP, UDP or ICMP segment’s Checksum field. This bit is
also set when the received number of payload bytes does not match the value indicated in
the Length field of the encapsulated IPv4 or IPv6 datagram in the received Ethernet frame.
This bit can take on special meaning as specified in Table 193.
If the enhanced descriptor format is enabled (EDFE=1, bit 7 in ETH_DMABMR), this bit takes
on the ESA function (otherwise it is PCE). When ESA is set, it indicates that the extended
status is available in descriptor word 4 (RDES4). ESA is valid only when the last descriptor bit
(RDES0[8]) is set.
Bits 5, 7, and 0 reflect the conditions discussed in Table 193.
RM0090 Rev 19 1183/1751

1239
Table 193. Receive descriptor 0 - encoding for bits 7, 5 and 0 (normal descriptor
format only, EDFE=0)
Bit 5: Bit 7: IPC Bit 0: payload
frame checksum checksum Frame status
type error error
IEEE 802.3 Type frame (Length field value is less than

0 0 0
0x0600.)
1 0 0 IPv4/IPv6 Type frame, no checksum error detected
IPv4/IPv6 Type frame with a payload checksum error (as described
1 0 1
for PCE) detected
IPv4/IPv6 Type frame with an IP header checksum error (as
1 1 0
described for IPC CE) detected
IPv4/IPv6 Type frame with both IP header and payload checksum
1 1 1
errors detected
IPv4/IPv6 Type frame with no IP header checksum error and the
0 0 1
payload check bypassed, due to an unsupported payload
A Type frame that is neither IPv4 or IPv6 (the checksum offload
0 1 1
engine bypasses checksum completely.)
0 1 0 Reserved
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserved
RCH
RER
DIC
RBS2 RBS2 RBS
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
Bit 31 DIC: Disable interrupt on completion

When set, this bit prevents setting the Status register’s RS bit (CSR5[6]) for the received frame
ending in the buffer indicated by this descriptor. This, in turn, disables the assertion of the interrupt to
Host due to RS for that frame.
Bits 28:16 RBS2: Receive buffer 2 size
These bits indicate the second data buffer size, in bytes. The buffer size must be a multiple of 4,
8, or 16, depending on the bus widths (32, 64 or 128, respectively), even if the value of RDES3
(buffer2 address pointer) is not aligned to bus width. If the buffer size is not an appropriate multiple
of 4, 8 or 16, the resulting behavior is undefined. This field is not valid if RDES1 [14] is set.
Bit 15 RER: Receive end of ring
When set, this bit indicates that the descriptor list reached its final descriptor. The DMA returns to the
base address of the list, creating a descriptor ring.
1184/1751 RM0090 Rev 19

Bit 14 RCH: Second address chained

When set, this bit indicates that the second address in the descriptor is the next descriptor address
rather than the second buffer address. When this bit is set, RBS2 (RDES1[28:16]) is a “don’t
care” value. RDES1[15] takes precedence over RDES1[14].
Bits 12:0 RBS1: Receive buffer 1 size
Indicates the first data buffer size in bytes. The buffer size must be a multiple of 4, 8 or 16,
depending upon the bus widths (32, 64 or 128), even if the value of RDES2 (buffer1 address
pointer) is not aligned. When the buffer size is not a multiple of 4, 8 or 16, the resulting behavior is
undefined. If this field is 0, the DMA ignores this buffer and uses Buffer 2 or next descriptor
depending on the value of RCH (bit 14).

RDES2 contains the address pointer to the first data buffer in the descriptor, or it
contains time stamp data.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RBP1 / RTSL
Bits 31:0 RBAP1 / RTSL: Receive buffer 1 address pointer / Receive frame time stamp low
These bits take on two different functions: the application uses them to indicate to the DMA
where to store the data in memory, and then after transferring all the data the DMA may use
these bits to pass back time stamp data.
RBAP1: When the software makes this descriptor available to the DMA (at the moment that
the OWN bit is set to 1 in RDES0), these bits indicate the physical address of Buffer 1. There are
no limitations on the buffer address alignment except for the following condition: the DMA uses the
configured value for its address generation when the RDES2 value is used to store the start of
frame. Note that the DMA performs a write operation with the RDES2[3/2/1:0] bits as 0 during the
transfer of the start of frame but the frame data is shifted as per the actual buffer address pointer. The
DMA ignores RDES2[3/2/1:0] (corresponding to bus width of 128/64/32) if the address pointer is to
a buffer where the middle or last part of the frame is stored.
RTSL: Before it clears the OWN bit in RDES0, the DMA updates this field with the 32 least
significant bits of the time stamp captured for the corresponding receive frame (overwriting
the value for RBAP1). This field has the time stamp only if time stamping is activated for this
frame and if the Last segment control bit (LS) in the descriptor is set.
RM0090 Rev 19 1185/1751

1239

RDES3 contains the address pointer either to the second data buffer in the descriptor
or to the next descriptor, or it contains time stamp data.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RBP2 / RTSH
Bits 31:0 RBAP2 / RTSH: Receive buffer 2 address pointer (next descriptor address) / Receive frame
time stamp high
These bits take on two different functions: the application uses them to indicate to the DMA
the location of where to store the data in memory, and then after transferring all the data the
DMA may use these bits to pass back time stamp data.
RBAP1: When the software makes this descriptor available to the DMA (at the moment that
the OWN bit is set to 1 in RDES0), these bits indicate the physical address of buffer 2 when a
descriptor ring structure is used. If the second address chained (RDES1 [24]) bit is set, this address
contains the pointer to the physical memory where the next descriptor is present. If RDES1 [24] is
set, the buffer (next descriptor) address pointer must be bus width-aligned (RDES3[3, 2, or
1:0] = 0, corresponding to a bus width of 128, 64 or 32. LSBs are ignored internally.)
However, when RDES1 [24] is reset, there are no limitations on the RDES3 value, except for the
following condition: the DMA uses the configured value for its buffer address generation when the
RDES3 value is used to store the start of frame. The DMA ignores RDES3[3, 2, or 1:0]
(corresponding to a bus width of 128, 64 or 32) if the address pointer is to a buffer where the
middle or last part of the frame is stored.
RTSH: Before it clears the OWN bit in RDES0, the DMA updates this field with the 32 most
significant bits of the time stamp captured for the corresponding receive frame (overwriting
the value for RBAP2). This field has the time stamp only if time stamping is activated and if
the Last segment control bit (LS) in the descriptor is set.
Enhanced Rx DMA descriptors format with IEEE1588 time stamp

Enhanced descriptors (enabled with EDFE=1, ETHDMABMR bit 7), must be used if time
stamping is activated (TSE=1, ETH_PTPTSCR bit 0) or if IPv4 checksum offload is
activated (IPCO=1, ETH_MACCR bit 10).
Enhanced descriptors comprise eight 32-bit words, twice the size of normal descriptors.
RDES0, RDES1, RDES2 and RDES3 have the same definitions as for normal receive
descriptors (refer to Normal Rx DMA descriptors). RDES4 contains extended status while
RDES6 and RDES7 hold the time stamp. RDES4, RDES5, RDES6 and RDES7 are defined
below.
When the Enhanced descriptor mode is selected, the software needs to allocate 32 bytes (8
words) of memory for every descriptor. When time stamping or IPv4 checksum offload are
not being used, the enhanced descriptor format may be disabled and the software can use
normal descriptors with the default size of 16 bytes.
1186/1751 RM0090 Rev 19

Figure 383. Enhanced receive descriptor field format with IEEE1588 time stamp
enabled
31 0
O
RDES 0 W Status [30:0]
N
RDES 1 CT Reserved Buffer 2 byte count CTRL Res. Buffer 1 byte count
RL [30:29] [28:16] [15:14] [12:0]
RDES 2 Buffer 1 address [31:0]
RDES 3 Buffer 2 address [31:0] or Next descriptor address [31:0]
RDES 4 Extended Status [31:0]
RDES 5 Reserved
RDES 6 Time stamp low [31:0]
RDES 7 Time stamp high [31:0]
ai17104

The extended status, shown below, is valid only when there is status related to IPv4
checksum or time stamp available as indicated by bit 0 in RDES0.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7
IPV6PR 6 5 4 3 2 1 0
IPV4PR
IPCB
IPHE
IPPE
PFT
PV PMT IPPT
Reserved

Bit 13 PV: PTP version
When set, indicates that the received PTP message uses the IEEE 1588 version 2 format.
When cleared, it uses version 1 format. This is valid only if the message type is non-zero.
Bit 12 PFT: PTP frame type
When set, this bit indicates that the PTP message is sent directly over Ethernet. When this bit
is cleared and the message type is non-zero, it indicates that the PTP message is sent over
UDP-IPv4 or UDP-IPv6. The information on IPv4 or IPv6 can be obtained from bits 6 and 7.
RM0090 Rev 19 1187/1751

1239
Bits 11:8 PMT: PTP message type

These bits are encoded to give the type of the message received.
– 0000: No PTP message received
– 0001: SYNC (all clock types)
– 0010: Follow_Up (all clock types)
– 0011: Delay_Req (all clock types)
– 0100: Delay_Resp (all clock types)
– 0101: Pdelay_Req (in peer-to-peer transparent clock) or Announce (in ordinary or boundary
clock)
– 0110: Pdelay_Resp (in peer-to-peer transparent clock) or Management (in ordinary or
boundary clock)
– 0111: Pdelay_Resp_Follow_Up (in peer-to-peer transparent clock) or Signaling (for ordinary
or boundary clock)
– 1xxx - Reserved
Bit 7 IPV6PR: IPv6 packet received
When set, this bit indicates that the received packet is an IPv6 packet.
Bit 6 IPV4PR: IPv4 packet received
When set, this bit indicates that the received packet is an IPv4 packet.
Bit 5 IPCB: IP checksum bypassed
When set, this bit indicates that the checksum offload engine is bypassed.
Bit 4 IPPE: IP payload error
When set, this bit indicates that the 16-bit IP payload checksum (that is, the TCP, UDP, or
ICMP checksum) that the core calculated does not match the corresponding checksum field
in the received segment. It is also set when the TCP, UDP, or ICMP segment length does not
match the payload length value in the IP Header field.
Bit 3 IPHE: IP header error
When set, this bit indicates either that the 16-bit IPv4 header checksum calculated by the
core does not match the received checksum bytes, or that the IP datagram version is not
consistent with the Ethernet Type value.
Bits 2:0 IPPT: IP payload type
if IPv4 checksum offload is activated (IPCO=1, ETH_MACCR bit 10), these bits
indicate the type of payload encapsulated in the IP datagram. These bits are ‘00’ if there is an IP
header error or fragmented IP.
– 000: Unknown or did not process IP payload
– 001: UDP
– 010: TCP
– 011: ICMP
– 1xx: Reserved

Reserved.
The table below describes the fields that have different meaning for RDES6 when the
receive descriptor is closed and time stamping is enabled.
1188/1751 RM0090 Rev 19

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RTSL
Bits 31:0 RTSL: Receive frame time stamp low

The DMA updates this field with the 32 least significant bits of the time stamp captured for the
corresponding receive frame. The DMA updates this field only for the last descriptor of the receive
frame indicated by last descriptor status bit (RDES0[8]). When this field and the RTSH field in
RDES7 show all ones, the time stamp must be treated as corrupt.

The table below describes the fields that have a different meaning for RDES7 when the
receive descriptor is closed and time stamping is enabled.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RTSH
Bits 31:0 RTSH: Receive frame time stamp high

The DMA updates this field with the 32 most significant bits of the time stamp captured for the
corresponding receive frame. The DMA updates this field only for the last descriptor of the receive
frame indicated by last descriptor status bit (RDES0[8]).
When this field and RDES7’s RTSL field show all ones, the time stamp must be treated as
corrupt.
33.6.9 DMA interrupts

Interrupts can be generated as a result of various events. The ETH_DMASR register
contains all the bits that might cause an interrupt. The ETH_DMAIER register contains an
enable bit for each of the events that can cause an interrupt.
There are two groups of interrupts, Normal and Abnormal, as described in the ETH_DMASR
register. Interrupts are cleared by writing a 1 to the corresponding bit position. When all the
enabled interrupts within a group are cleared, the corresponding summary bit is cleared. If
the MAC core is the cause for assertion of the interrupt, then any of the TSTS or PMTS bits
in the ETH_DMASR register is set high.
Interrupts are not queued and if the interrupt event occurs before the driver has responded
to it, no additional interrupts are generated. For example, the Receive Interrupt bit
(ETH_DMASR register [6]) indicates that one or more frames were transferred to the
STM32F4xx buffer. The driver must scan all descriptors, from the last recorded position to
the first one owned by the DMA.
An interrupt is generated only once for simultaneous, multiple events. The driver must scan
the ETH_DMASR register for the cause of the interrupt. The interrupt is not generated again
unless a new interrupting event occurs, after the driver has cleared the appropriate bit in the
ETH_DMASR register. For example, the controller generates a Receive interrupt
(ETH_DMASR register[6]) and the driver begins reading the ETH_DMASR register. Next,
receive buffer unavailable (ETH_DMASR register[7]) occurs. The driver clears the Receive
interrupt. Even then, a new interrupt is generated, due to the active or pending Receive
buffer unavailable interrupt.
RM0090 Rev 19 1189/1751

1239
Figure 384. Interrupt scheme

TS
MMCI
AND
TIE PMTI
TBUS
AND TSTI
TBUIE RS
OR NIS
AND
RIE ERS AND
NISE
AND
ERIE
OR Interrupt
FBES
AND
TPSS FBEIE
AND TJTS AIS
TPSSIE
AND AND
ROS TJTIE
AISE
AND
TUS ROIE
AND
TUIE RBU OR
AND RPSS
RWTS RBUIE
AND
AND RPSSIE
RWTIE
ETS
AND
ETIE
AI15646
33.7 Ethernet interrupts

The Ethernet controller has two interrupt vectors: one dedicated to normal Ethernet
operations and the other, used only for the Ethernet wakeup event (with wakeup frame or
Magic Packet detection) when it is mapped on EXTI lIne19.
The first Ethernet vector is reserved for interrupts generated by the MAC and the DMA as
listed in the MAC interrupts and DMA interrupts sections.
The second vector is reserved for interrupts generated by the PMT on wakeup events. The
mapping of a wakeup event on EXTI line19 causes the STM32F4xx to exit the low-power
mode, and generates an interrupt.
When an Ethernet wakeup event mapped on EXTI Line19 occurs and the MAC PMT
interrupt is enabled and the EXTI Line19 interrupt, with detection on rising edge, is also
enabled, both interrupts are generated.
A watchdog timer (see ETH_DMARSWTR register) is given for flexible control of the RS bit
(ETH_DMASR register). When this watchdog timer is programmed with a non-zero value, it
gets activated as soon as the RxDMA completes a transfer of a received frame to system
memory without asserting the Receive Status because it is not enabled in the corresponding
Receive descriptor (RDES1[31]). When this timer runs out as per the programmed value,
the RS bit is set and the interrupt is asserted if the corresponding RIE is enabled in the
ETH_DMAIER register. This timer is disabled before it runs out, when a frame is transferred
to memory and the RS is set because it is enabled for that descriptor.
1190/1751 RM0090 Rev 19

Note: Reading the PMT control and status register automatically clears the Wakeup Frame
Received and Magic Packet Received PMT interrupt flags. However, since the registers for
these flags are in the CLK_RX domain, there may be a significant delay before this update
is visible by the firmware. The delay is especially long when the RX clock is slow (in 10 Mbit
mode) and when the AHB bus is high-frequency.
Since interrupt requests from the PMT to the CPU are based on the same registers in the
CLK_RX domain, the CPU may spuriously call the interrupt routine a second time even after
reading PMT_CSR. Thus, it may be necessary that the firmware polls the Wakeup Frame
Received and Magic Packet Received bits and exits the interrupt service routine only when
they are found to be at ‘0’.
33.8 Ethernet register descriptions

The peripheral registers can be accessed by bytes (8-bit), half-words (16-bit) or words (32-
bits).
33.8.1 MAC register description

Ethernet MAC configuration register (ETH_MACCR)
The MAC configuration register is the operation mode register of the MAC. It establishes
receive and transmit operating modes.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7
APCS 6 5 4 3 2 1 0
CSTF
IPCO
Reserved
Reserved
Reserved
Reserved
Reserved
ROD
CSD
FES
WD
DM
RD
DC
RE
LM
TE
BL
JD
IFG
Reserved

CSTF: CRC stripping for Type frames
Bit 25 When set, the last 4 bytes (FCS) of all frames of Ether type (type field greater than
0x0600) will be stripped and dropped before forwarding the frame to the application.
Bit 24Reserved, must be kept at reset value.
Bit 23 WD: Watchdog disable
When this bit is set, the MAC disables the watchdog timer on the receiver, and can receive
frames of up to 16 384 bytes.
When this bit is reset, the MAC allows no more than 2 048 bytes of the frame being received
and cuts off any bytes received after that.
Bit 22 JD: Jabber disable
When this bit is set, the MAC disables the jabber timer on the transmitter, and can transfer
frames of up to 16 384 bytes.
When this bit is reset, the MAC cuts off the transmitter if the application sends out more than
2 048 bytes of data during transmission.
RM0090 Rev 19 1191/1751

1239
Bits 19:17 IFG: Interframe gap

These bits control the minimum interframe gap between frames during transmission.
000: 96 bit times
001: 88 bit times
010: 80 bit times
….
111: 40 bit times
Note: In Half-duplex mode, the minimum IFG can be configured for 64 bit times (IFG = 100)
only. Lower values are not considered.
Bit 16 CSD: Carrier sense disable
When set high, this bit makes the MAC transmitter ignore the MII CRS signal during frame
transmission in Half-duplex mode. No error is generated due to Loss of Carrier or No Carrier
during such transmission.
When this bit is low, the MAC transmitter generates such errors due to Carrier Sense and
even aborts the transmissions.
Bit 14 FES: Fast Ethernet speed
Indicates the speed in Fast Ethernet (MII) mode:
0: 10 Mbit/s
1: 100 Mbit/s
Bit 13 ROD: Receive own disable
When this bit is set, the MAC disables the reception of frames in Half-duplex mode.
When this bit is reset, the MAC receives all packets that are given by the PHY while
transmitting.
This bit is not applicable if the MAC is operating in Full-duplex mode.
Bit 12 LM: Loopback mode
When this bit is set, the MAC operates in loopback mode at the MII. The MII receive clock
input (RX_CLK) is required for the loopback to work properly, as the transmit clock is not
looped-back internally.
Bit 11 DM: Duplex mode
When this bit is set, the MAC operates in a Full-duplex mode where it can transmit and
receive simultaneously.
Bit 10 IPCO: IPv4 checksum offload
When set, this bit enables IPv4 checksum checking for received frame payloads'
TCP/UDP/ICMP headers. When this bit is reset, the checksum offload function in the
receiver is disabled and the corresponding PCE and IP HCE status bits (see Table 190) are
always cleared.
Bit 9 RD: Retry disable
When this bit is set, the MAC attempts only 1 transmission. When a collision occurs on the
MII, the MAC ignores the current frame transmission and reports a Frame Abort with
excessive collision error in the transmit frame status.
When this bit is reset, the MAC attempts retries based on the settings of BL.
Note: This bit is applicable only in the Half-duplex mode.
1192/1751 RM0090 Rev 19

Bit 7 APCS: Automatic pad/CRC stripping

When this bit is set, the MAC strips the Pad/FCS field on incoming frames only if the
length’s field value is less than or equal to 1 500 bytes. All received frames with length field
greater than or equal to 1 501 bytes are passed on to the application without stripping the
Pad/FCS field.
When this bit is reset, the MAC passes all incoming frames unmodified.
Bits 6:5 BL: Back-off limit
The Back-off limit determines the random integer number (r) of slot time delays (4 096 bit
times for 1000 Mbit/s and 512 bit times for 10/100 Mbit/s) the MAC waits before
rescheduling a transmission attempt during retries after a collision.
Note: This bit is applicable only to Half-duplex mode.
00: k = min (n, 10)
01: k = min (n, 8)
10: k = min (n, 4)
11: k = min (n, 1),
where n = retransmission attempt. The random integer r takes the value in the range 0 ≤ r <
2k
Bit 4 DC: Deferral check
When this bit is set, the deferral check function is enabled in the MAC. The MAC issues a
Frame Abort status, along with the excessive deferral error bit set in the transmit frame
status when the transmit state machine is deferred for more than 24 288 bit times in 10/100-
Mbit/s mode. Deferral begins when the transmitter is ready to transmit, but is prevented
because of an active CRS (carrier sense) signal on the MII. Defer time is not cumulative. If
the transmitter defers for 10 000 bit times, then transmits, collides, backs off, and then has
to defer again after completion of back-off, the deferral timer resets to 0 and restarts.
When this bit is reset, the deferral check function is disabled and the MAC defers until the
CRS signal goes inactive. This bit is applicable only in Half-duplex mode.
Bit 3 TE: Transmitter enable
When this bit is set, the transmit state machine of the MAC is enabled for transmission on
the MII. When this bit is reset, the MAC transmit state machine is disabled after the
completion of the transmission of the current frame, and does not transmit any further
frames.
Bit 2 RE: Receiver enable
When this bit is set, the receiver state machine of the MAC is enabled for receiving frames
from the MII. When this bit is reset, the MAC receive state machine is disabled after the
completion of the reception of the current frame, and will not receive any further frames from
the MII.
RM0090 Rev 19 1193/1751

1239
Ethernet MAC frame filter register (ETH_MACFFR)

The MAC frame filter register contains the filter controls for receiving frames. Some of the
controls from this register go to the address check block of the MAC, which performs the
first level of address filtering. The second level of filtering is performed on the incoming
frame, based on other controls such as pass bad frames and pass control frames.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DAIF
SAIF
PAM
HPF
PCF
BFD
SAF
HM
PM
HU
RA
Reserved
Bit 31 RA: Receive all

When this bit is set, the MAC receiver passes all received frames on to the application,
irrespective of whether they have passed the address filter. The result of the SA/DA filtering
is updated (pass or fail) in the corresponding bits in the receive status word. When this bit is
reset, the MAC receiver passes on to the application only those frames that have passed the
SA/DA address filter.
Bit 10 HPF: Hash or perfect filter
When this bit is set and if the HM or HU bit is set, the address filter passes frames that
match either the perfect filtering or the hash filtering.
When this bit is cleared and if the HU or HM bit is set, only frames that match the Hash filter
are passed.
Bit 9 SAF: Source address filter
The MAC core compares the SA field of the received frames with the values programmed in
the enabled SA registers. If the comparison matches, then the SAMatch bit in the RxStatus
word is set high. When this bit is set high and the SA filter fails, the MAC drops the frame.
When this bit is reset, the MAC core forwards the received frame to the application. It also
forwards the updated SA Match bit in RxStatus depending on the SA address comparison.
Bit 8 SAIF: Source address inverse filtering
When this bit is set, the address check block operates in inverse filtering mode for the SA
address comparison. The frames whose SA matches the SA registers are marked as failing
the SA address filter.
When this bit is reset, frames whose SA does not match the SA registers are marked as
failing the SA address filter.
1194/1751 RM0090 Rev 19

Bits 7:6 PCF: Pass control frames

These bits control the forwarding of all control frames (including unicast and multicast
PAUSE frames). Note that the processing of PAUSE control frames depends only on RFCE
in Flow Control Register[2].
00: MAC prevents all control frames from reaching the application
01: MAC forwards all control frames to application except Pause control frames
10: MAC forwards all control frames to application even if they fail the address filter
11: MAC forwards control frames that pass the address filter.
These bits control the forwarding of all control frames (including unicast and multicast
PAUSE frames). Note that the processing of PAUSE control frames depends only on RFCE
in Flow Control register[2].
00 or 01: MAC prevents all control frames from reaching the application
10: MAC forwards all control frames to application even if they fail the address filter
11: MAC forwards control frames that pass the address filter.
Bit 5 BFD: Broadcast frames disable
When this bit is set, the address filters filter all incoming broadcast frames.
When this bit is reset, the address filters pass all received broadcast frames.
Bit 4 PAM: Pass all multicast
When set, this bit indicates that all received frames with a multicast destination address (first
bit in the destination address field is '1') are passed.
When reset, filtering of multicast frame depends on the HM bit.
Bit 3 DAIF: Destination address inverse filtering
When this bit is set, the address check block operates in inverse filtering mode for the DA
address comparison for both unicast and multicast frames.
When reset, normal filtering of frames is performed.
Bit 2 HM: Hash multicast
When set, MAC performs destination address filtering of received multicast frames
according to the hash table.
When reset, the MAC performs a perfect destination address filtering for multicast frames,
that is, it compares the DA field with the values programmed in DA registers.
Bit 1 HU: Hash unicast
When set, MAC performs destination address filtering of unicast frames according to the
hash table.
When reset, the MAC performs a perfect destination address filtering for unicast frames, that
is, it compares the DA field with the values programmed in DA registers.
Bit 0 PM: Promiscuous mode
When this bit is set, the address filters pass all incoming frames regardless of their
destination or source address. The SA/DA filter fails status bits in the receive status word
are always cleared when PM is set.
Ethernet MAC hash table high register (ETH_MACHTHR)

The 64-bit Hash table is used for group address filtering. For hash filtering, the contents of
the destination address in the incoming frame are passed through the CRC logic, and the
upper 6 bits in the CRC register are used to index the contents of the Hash table. This CRC
is a 32-bit value coded by the following polynomial (for more details refer to Section 33.5.3):
32 26 23 22 16 12 11 10 8 7 5 4 2
RM0090 Rev 19 1195/1751

1239
The most significant bit determines the register to be used (hash table high/hash table low),
and the other 5 bits determine which bit within the register. A hash value of 0b0 0000 selects
bit 0 in the selected register, and a value of 0b1 1111 selects bit 31 in the selected register.
For example, if the DA of the incoming frame is received as 0x1F52 419C B6AF (0x1F is the
first byte received on the MII interface), then the internally calculated 6-bit Hash value is
0x2C and the HTH register bit[12] is checked for filtering. If the DA of the incoming frame is
received as 0xA00A 9800 0045, then the calculated 6-bit Hash value is 0x07 and the HTL
register bit[7] is checked for filtering.
If the corresponding bit value in the register is 1, the frame is accepted. Otherwise, it is
rejected. If the PAM (pass all multicast) bit is set in the ETH_MACFFR register, then all
multicast frames are accepted regardless of the multicast hash values.
The Hash table high register contains the higher 32 bits of the multicast Hash table.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
HTH
Bits 31:0 HTH: Hash table high

This field contains the upper 32 bits of Hash table.
Ethernet MAC hash table low register (ETH_MACHTLR)

The Hash table low register contains the lower 32 bits of the multi-cast Hash table.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
HTL
Bits 31:0 HTL: Hash table low

This field contains the lower 32 bits of the Hash table.
Ethernet MAC MII address register (ETH_MACMIIAR)

The MII address register controls the management cycles to the external PHY through the
management interface.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserved
PA MR CR MW MB
Reserved rc_
w1
1196/1751 RM0090 Rev 19


Bits 15:11 PA: PHY address
This field tells which of the 32 possible PHY devices are being accessed.
Bits 10:6 MR: MII register
These bits select the desired MII register in the selected PHY device.
Bits 4:2 CR: Clock range
The CR clock range selection determines the HCLK frequency and is used to decide the
frequency of the MDC clock:
Selection HCLK MDC Clock
000 60-100 MHz HCLK/42
001 100-150 MHzHCLK/62
010 20-35 MHz HCLK/16
011 35-60 MHz HCLK/26
100 150-168 MHz HCLK/102
101, 110, 111 Reserved -
Bit 1 MW: MII write
When set, this bit tells the PHY that this will be a Write operation using the MII Data register. If
this bit is not set, this will be a Read operation, placing the data in the MII Data register.
Bit 0 MB: MII busy
This bit should read a logic 0 before writing to ETH_MACMIIAR and ETH_MACMIIDR. This
bit must also be reset to 0 during a Write to ETH_MACMIIAR. During a PHY register access,
this bit is set to 0b1 by the application to indicate that a read or write access is in progress.
ETH_MACMIIDR (MII Data) should be kept valid until this bit is cleared by the MAC during a
PHY Write operation. The ETH_MACMIIDR is invalid until this bit is cleared by the MAC
during a PHY Read operation. The ETH_MACMIIAR (MII Address) should not be written to
until this bit is cleared.
Ethernet MAC MII data register (ETH_MACMIIDR)

The MAC MII Data register stores write data to be written to the PHY register located at the
address specified in ETH_MACMIIAR. ETH_MACMIIDR also stores read data from the PHY
register located at the address specified by ETH_MACMIIAR.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MD
Reserved

Bits 15:0 MD: MII data
This contains the 16-bit data value read from the PHY after a Management Read operation,
or the 16-bit data value to be written to the PHY before a Management Write operation.
RM0090 Rev 19 1197/1751

1239
Ethernet MAC flow control register (ETH_MACFCR)

The Flow control register controls the generation and reception of the control (Pause
Command) frames by the MAC. A write to a register with the Busy bit set to '1' causes the
MAC to generate a pause control frame. The fields of the control frame are selected as
specified in the 802.3x specification, and the Pause Time value from this register is used in
the Pause Time field of the control frame. The Busy bit remains set until the control frame is
transferred onto the cable. The Host must make sure that the Busy bit is cleared before
writing to the register.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ZQPD
UPFD
RFCE
TFCE
FCB/
Reserved
PT PLT
BPA
Reserved
rc_w1
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
/rw
Bits 31:16 PT: Pause time

This field holds the value to be used in the Pause Time field in the transmit control frame. If
the Pause Time bits is configured to be double-synchronized to the MII clock domain, then
consecutive write operations to this register should be performed only after at least 4 clock
cycles in the destination clock domain.
Bit 7 ZQPD: Zero-quanta pause disable
When set, this bit disables the automatic generation of Zero-quanta pause control frames on
the deassertion of the flow-control signal from the FIFO layer.
When this bit is reset, normal operation with automatic Zero-quanta pause control frame
generation is enabled.
Bits 5:4 PLT: Pause low threshold
This field configures the threshold of the Pause timer at which the Pause frame is
automatically retransmitted. The threshold values should always be less than the Pause
Time configured in bits[31:16]. For example, if PT = 100H (256 slot-times), and PLT = 01,
then a second PAUSE frame is automatically transmitted if initiated at 228 (256 – 28) slot-
times after the first PAUSE frame is transmitted.
Selection Threshold
00 Pause time minus 4 slot times
Slot time is defined as time taken to transmit 512 bits (64 bytes) on the MII interface.
Bit 3 UPFD: Unicast pause frame detect
When this bit is set, the MAC detects the Pause frames with the station’s unicast address
specified in the ETH_MACA0HR and ETH_MACA0LR registers, in addition to detecting
Pause frames with the unique multicast address.
When this bit is reset, the MAC detects only a Pause frame with the unique multicast
address specified in the 802.3x standard.
1198/1751 RM0090 Rev 19

Bit 2 RFCE: Receive flow control enable

When this bit is set, the MAC decodes the received Pause frame and disables its transmitter
for a specified (Pause Time) time.
When this bit is reset, the decode function of the Pause frame is disabled.
Bit 1 TFCE: Transmit flow control enable
In Full-duplex mode, when this bit is set, the MAC enables the flow control operation to
transmit Pause frames. When this bit is reset, the flow control operation in the MAC is
disabled, and the MAC does not transmit any Pause frames.
In Half-duplex mode, when this bit is set, the MAC enables the back-pressure operation.
When this bit is reset, the back pressure feature is disabled.
Bit 0 FCB/BPA: Flow control busy/back pressure activate
This bit initiates a Pause Control frame in Full-duplex mode and activates the back pressure
function in Half-duplex mode if TFCE bit is set.
In Full-duplex mode, this bit should be read as 0 before writing to the Flow control register.
To initiate a Pause control frame, the Application must set this bit to 1. During a transfer of
the Control frame, this bit continues to be set to signify that a frame transmission is in
progress. After completion of the Pause control frame transmission, the MAC resets this bit
to 0. The Flow control register should not be written to until this bit is cleared.
In Half-duplex mode, when this bit is set (and TFCE is set), back pressure is asserted by the
MAC core. During back pressure, when the MAC receives a new frame, the transmitter
starts sending a JAM pattern resulting in a collision. When the MAC is configured to Full-
duplex mode, the BPA is automatically disabled.
Ethernet MAC VLAN tag register (ETH_MACVLANTR)

The VLAN tag register contains the IEEE 802.1Q VLAN Tag to identify the VLAN frames.
The MAC compares the 13th and 14th bytes of the receiving frame (Length/Type) with
0x8100, and the following 2 bytes are compared with the VLAN tag; if a match occurs, the
received VLAN bit in the receive frame status is set. The legal length of the frame is
increased from 1518 bytes to 1522 bytes.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
VLANTC
VLANTI
Reserved
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
RM0090 Rev 19 1199/1751

1239

Bit 16 VLANTC: 12-bit VLAN tag comparison
When this bit is set, a 12-bit VLAN identifier, rather than the complete 16-bit VLAN tag, is
used for comparison and filtering. Bits[11:0] of the VLAN tag are compared with the
corresponding field in the received VLAN-tagged frame.
When this bit is reset, all 16 bits of the received VLAN frame’s fifteenth and sixteenth bytes
are used for comparison.
Bits 15:0 VLANTI: VLAN tag identifier (for receive frames)
This contains the 802.1Q VLAN tag to identify VLAN frames, and is compared to the fifteenth
and sixteenth bytes of the frames being received for VLAN frames. Bits[15:13] are the user
priority, Bit[12] is the canonical format indicator (CFI) and bits[11:0] are the VLAN tag’s VLAN
identifier (VID) field. When the VLANTC bit is set, only the VID (bits[11:0]) is used for
comparison.
If VLANTI (VLANTI[11:0] if VLANTC is set) is all zeros, the MAC does not check the fifteenth
and sixteenth bytes for VLAN tag comparison, and declares all frames with a Type field value
of 0x8100 as VLAN frames.
Ethernet MAC remote wakeup frame filter register (ETH_MACRWUFFR)

This is the address through which the remote wakeup frame filter registers are written/read
by the application. The Wakeup frame filter register is actually a pointer to eight (not
transparent) such wakeup frame filter registers. Eight sequential write operations to this
address with the offset (0x0028) will write all wakeup frame filter registers. Eight sequential
read operations from this address with the offset (0x0028) will read all wakeup frame filter
registers. This register contains the higher 16 bits of the 7th MAC address. Refer to Remote
wakeup frame filter register section for additional information.
Figure 385. Ethernet MAC remote wakeup frame filter register (ETH_MACRWUFFR)
Filter 3 Filter 2 Filter 1 Filter 0

Wakeup frame filter reg4 RSVD RSVD RSVD RSVD
Command Command Command Command
Wakeup frame filter reg5 Filter 3 Offset Filter 2 Offset Filter 1 Offset Filter 0 Offset
ai15648
1200/1751 RM0090 Rev 19

Ethernet MAC PMT control and status register (ETH_MACPMTCSR)

The ETH_MACPMTCSR programs the request wakeup events and monitors the wakeup
events.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
WFFRPR
Reserved
Reserved
WFR
WFE
MPR
MPE
GU
PD
Reserved
rs Res. rw rc_r rc_r rw rw rs
Bit 31 WFFRPR: Wakeup frame filter register pointer reset

When set, it resets the Remote wakeup frame filter register pointer to 0b000. It is
automatically cleared after 1 clock cycle.
Bit 9 GU: Global unicast
When set, it enables any unicast packet filtered by the MAC (DAF) address recognition to be
a wakeup frame.
Bit 6 WFR: Wakeup frame received
When set, this bit indicates the power management event was generated due to reception of
a wakeup frame. This bit is cleared by a read into this register.
Bit 5 MPR: Magic packet received
When set, this bit indicates the power management event was generated by the reception of
a Magic Packet. This bit is cleared by a read into this register.
Bit 2 WFE: Wakeup frame enable
When set, this bit enables the generation of a power management event due to wakeup
frame reception.
Bit 1 MPE: Magic Packet enable
When set, this bit enables the generation of a power management event due to Magic
Packet reception.
Bit 0 PD: Power down
When this bit is set, all received frames will be dropped. This bit is cleared automatically
when a magic packet or wakeup frame is received, and Power-down mode is disabled.
Frames received after this bit is cleared are forwarded to the application. This bit must only
be set when either the Magic Packet Enable or Wakeup Frame Enable bit is set high.
RM0090 Rev 19 1201/1751

1239
Ethernet MAC debug register (ETH_MACDBGR)

This debug register gives the status of all the main modules of the transmit and receive data
paths and the FIFOs. An all-zero status indicates that the MAC core is in Idle state (and
FIFOs are empty) and no activity is going on in the data paths.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MSFRWCS
MMRPEA
RFWRA
MMTEA
MTFCS
RFRCS
TFWA
TFNE
TFRS
RFFL
Reserved
Reserved
Reserved
Reserved
MTP
TFF
Reserved
ro ro ro ro ro ro ro ro ro ro ro ro ro ro ro ro ro

Bit 25 TFF: Tx FIFO full
When high, it indicates that the Tx FIFO is full and hence no more frames will be accepted
for transmission.
Bit 24 TFNE: Tx FIFO not empty
When high, it indicates that the TxFIFO is not empty and has some data left for
transmission.
Bit 22 TFWA: Tx FIFO write active
When high, it indicates that the TxFIFO write controller is active and transferring data to the
TxFIFO.
Bits 21:20 TFRS: Tx FIFO read status
This indicates the state of the TxFIFO read controller:
00: Idle state
01: Read state (transferring data to the MAC transmitter)
10: Waiting for TxStatus from MAC transmitter
11: Writing the received TxStatus or flushing the TxFIFO
Bit 19 MTP: MAC transmitter in pause
When high, it indicates that the MAC transmitter is in Pause condition (in full-duplex mode
only) and hence will not schedule any frame for transmission
Bits 18:17 MTFCS: MAC transmit frame controller status
This indicates the state of the MAC transmit frame controller:
00: Idle
01: Waiting for Status of previous frame or IFG/backoff period to be over
10: Generating and transmitting a Pause control frame (in full duplex mode)
11: Transferring input frame for transmission
Bit 16 MMTEA: MAC MII transmit engine active
When high, it indicates that the MAC MII transmit engine is actively transmitting data and
that it is not in the Idle state.
1202/1751 RM0090 Rev 19

Bits 9:8 RFFL: Rx FIFO fill level

This gives the status of the Rx FIFO fill-level:
00: RxFIFO empty
01: RxFIFO fill-level below flow-control de-activate threshold
10: RxFIFO fill-level above flow-control activate threshold
11: RxFIFO full
Bits 6:5 RFRCS: Rx FIFO read controller status
It gives the state of the Rx FIFO read controller:
00: IDLE state
01: Reading frame data
10: Reading frame status (or time-stamp)
11: Flushing the frame data and status
Bit 4 RFWRA: Rx FIFO write controller active
When high, it indicates that the Rx FIFO write controller is active and transferring a received
frame to the FIFO.
Bits 2:1 MSFRWCS: MAC small FIFO read / write controllers status
When high, these bits indicate the respective active state of the small FIFO read and write
controllers of the MAC receive frame controller module.
Bit 0 MMRPEA: MAC MII receive protocol engine active
When high, it indicates that the MAC MII receive protocol engine is actively receiving data
and is not in the Idle state.
RM0090 Rev 19 1203/1751

1239
Ethernet MAC interrupt status register (ETH_MACSR)

The ETH_MACSR register contents identify the events in the MAC that can generate an
interrupt.
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TSTS MMCTS MMCRS MMCS PMTS
rc_r r r r r

Bit 9 TSTS: Time stamp trigger status
This bit is set high when the system time value equals or exceeds the value specified in the
Target time high and low registers. This bit is cleared by reading the ETH_PTPTSSR
register.
Bit 6 MMCTS: MMC transmit status
This bit is set high whenever an interrupt is generated in the ETH_MMCTIR register. This bit
is cleared when all the bits in this interrupt register (ETH_MMCTIR) are cleared.
Bit 5 MMCRS: MMC receive status
This bit is set high whenever an interrupt is generated in the ETH_MMCRIR register. This bit
is cleared when all the bits in this interrupt register (ETH_MMCRIR) are cleared.
Bit 4 MMCS: MMC status
This bit is set high whenever any of bits 6:5 is set high. It is cleared only when both bits are
low.
Bit 3 PMTS: PMT status
This bit is set whenever a Magic packet or Wake-on-LAN frame is received in Power-down
mode (see bits 5 and 6 in the ETH_MACPMTCSR register Ethernet MAC PMT control and
status register (ETH_MACPMTCSR)). This bit is cleared when both bits[6:5], of this last
register, are cleared due to a read operation to the ETH_MACPMTCSR register.
1204/1751 RM0090 Rev 19

Ethernet MAC interrupt mask register (ETH_MACIMR)

The ETH_MACIMR register bits make it possible to mask the interrupt signal due to the
corresponding event in the ETH_MACSR register.
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TSTIM PMTIM
rw rw

Bit 9 TSTIM: Time stamp trigger interrupt mask
When set, this bit disables the time stamp interrupt generation.
Bit 3 PMTIM: PMT interrupt mask
When set, this bit disables the assertion of the interrupt signal due to the setting of the PMT
Status bit in ETH_MACSR.
Ethernet MAC address 0 high register (ETH_MACA0HR)

Reset value: 0x8000 FFFF
The MAC address 0 high register holds the upper 16 bits of the 6-byte first MAC address of
the station. Note that the first DA byte that is received on the MII interface corresponds to
the LS Byte (bits [7:0]) of the MAC address low register. For example, if 0x1122 3344 5566
is received (0x11 is the first byte) on the MII as the destination address, then the MAC
address 0 register [47:0] is compared with 0x6655 4433 2211.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MO
MACA0H
Reserved
1 rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
Bit 31 MO: Always 1.

Bits 15:0 MACA0H: MAC address0 high [47:32]
This field contains the upper 16 bits (47:32) of the 6-byte MAC address0. This is used by the
MAC for filtering for received frames and for inserting the MAC address in the transmit flow
control (Pause) frames.
RM0090 Rev 19 1205/1751

1239
Ethernet MAC address 0 low register (ETH_MACA0LR)

The MAC address 0 low register holds the lower 32 bits of the 6-byte first MAC address of
the station.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MACA0L
Bits 31:0 MACA0L: MAC address0 low [31:0]

This field contains the lower 32 bits of the 6-byte MAC address0. This is used by the MAC
for filtering for received frames and for inserting the MAC address in the transmit flow control
(Pause) frames.

The MAC address 1 high register holds the upper 16 bits of the 6-byte second MAC address
of the station.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
AE SA MBC MACA1H
Reserved
Bit 31 AE: Address enable

When this bit is set, the address filters use the MAC address1 for perfect filtering. When this
bit is cleared, the address filters ignore the address for filtering.
Bit 30 SA: Source address
When this bit is set, the MAC address1[47:0] is used for comparison with the SA fields of the
received frame.
When this bit is cleared, the MAC address1[47:0] is used for comparison with the DA fields
of the received frame.
Bits 29:24 MBC: Mask byte control
These bits are mask control bits for comparison of each of the MAC address1 bytes. When
they are set high, the MAC core does not compare the corresponding byte of received
DA/SA with the contents of the MAC address1 registers. Each bit controls the masking of the
bytes as follows:
– Bit 29: ETH_MACA1HR [15:8]
– Bit 27: ETH_MACA1LR [31:24]
…
1206/1751 RM0090 Rev 19


This field contains the upper 16 bits (47:32) of the 6-byte second MAC address.
Ethernet MAC address1 low register (ETH_MACA1LR)

The MAC address 1 low register holds the lower 32 bits of the 6-byte second MAC address
of the station.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MACA1L

This field contains the lower 32 bits of the 6-byte MAC address1. The content of this field is
undefined until loaded by the application after the initialization process.

of the station.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
AE SA MBC MACA2H
Reserved
RM0090 Rev 19 1207/1751

1239

When this bit is set, the address filters use the MAC address2 for perfect filtering. When
reset, the address filters ignore the address for filtering.
When this bit is set, the MAC address 2 [47:0] is used for comparison with the SA fields of
the received frame.
When this bit is reset, the MAC address 2 [47:0] is used for comparison with the DA fields of
the received frame.
set high, the MAC core does not compare the corresponding byte of received DA/SA with the
contents of the MAC address 2 registers. Each bit controls the masking of the bytes as
follows:
…
This field contains the upper 16 bits (47:32) of the 6-byte MAC address2.

of the station.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MACA2L

This field contains the lower 32 bits of the 6-byte second MAC address2. The content of this
field is undefined until loaded by the application after the initialization process.
1208/1751 RM0090 Rev 19


of the station.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
AE SA MBC MACA3H
Reserved

When this bit is set, the address filters use the MAC address3 for perfect filtering. When this
bit is cleared, the address filters ignore the address for filtering.
When this bit is set, the MAC address 3 [47:0] is used for comparison with the SA fields of the
received frame.
When this bit is cleared, the MAC address 3[47:0] is used for comparison with the DA fields
of the received frame.
these bits are set high, the MAC core does not compare the corresponding byte of received
DA/SA with the contents of the MAC address 3 registers. Each bit controls the masking of the
bytes as follows:
…
This field contains the upper 16 bits (47:32) of the 6-byte MAC address3.

of the station.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MACA3L

This field contains the lower 32 bits of the 6-byte second MAC address3. The content of this
field is undefined until loaded by the application after the initialization process.
RM0090 Rev 19 1209/1751

1239
33.8.2 MMC register description

Ethernet MMC control register (ETH_MMCCR)
The Ethernet MMC Control register establishes the operating mode of the management
counters.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MCFHP
ROR
MCP
MCF
CSR
CR
Reserved
rw rw rw rw rw rw

MCFHP: MMC counter Full-Half preset
When MCFHP is low and bit4 is set, all MMC counters get preset to almost-half value. All
Bit 5 frame-counters get preset to 0x7FFF_FFF0 (half - 16)
When MCFHP is high and bit4 is set, all MMC counters get preset to almost-full value. All
frame-counters get preset to 0xFFFF_FFF0 (full - 16)
MCP: MMC counter preset
When set, all counters will be initialized or preset to almost full or almost half as per
Bit 4 Bit5 above. This bit will be cleared automatically after 1 clock cycle. This bit along
with bit5 is useful for debugging and testing the assertion of interrupts due to MMC
counter becoming half-full or full.
Bit 3 MCF: MMC counter freeze
When set, this bit freezes all the MMC counters to their current value. (None of the MMC
counters are updated due to any transmitted or received frame until this bit is cleared to 0. If
any MMC counter is read with the Reset on Read bit set, then that counter is also cleared in
this mode.)
Bit 2 ROR: Reset on read
When this bit is set, the MMC counters is reset to zero after read (self-clearing after reset).
The counters are cleared when the least significant byte lane (bits [7:0]) is read.
Bit 1 CSR: Counter stop rollover
When this bit is set, the counter does not roll over to zero after it reaches the maximum
value.
Bit 0 CR: Counter reset
When it is set, all counters are reset. This bit is cleared automatically after 1 clock cycle.
Ethernet MMC receive interrupt register (ETH_MMCRIR)

The Ethernet MMC receive interrupt register maintains the interrupts generated when
receive statistic counters reach half their maximum values. (MSB of the counter is set.) It is
a 32-bit wide register. An interrupt bit is cleared when the respective MMC counter that
1210/1751 RM0090 Rev 19

caused the interrupt is read. The least significant byte lane (bits [7:0]) of the respective
counter must be read in order to clear the interrupt bit.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RGUFS
RFCES
RFAES
rc_r rc_r rc_r

Bit 17 RGUFS: Received Good Unicast Frames Status
This bit is set when the received, good unicast frames, counter reaches half the maximum
value.
Bit 6 RFAES: Received frames alignment error status
This bit is set when the received frames, with alignment error, counter reaches half the
maximum value.
Bit 5 RFCES: Received frames CRC error status
This bit is set when the received frames, with CRC error, counter reaches half the maximum
value.
Ethernet MMC transmit interrupt register (ETH_MMCTIR)

The Ethernet MMC transmit Interrupt register maintains the interrupts generated when
transmit statistic counters reach half their maximum values. (MSB of the counter is set.) It is
a 32-bit wide register. An interrupt bit is cleared when the respective MMC counter that
caused the interrupt is read. The least significant byte lane (bits [7:0]) of the respective
counter must be read in order to clear the interrupt bit.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TGFMSCS
TGFSCS
TGFS
rc_r rc_r rc_r

Bit 21 TGFS: Transmitted good frames status
This bit is set when the transmitted, good frames, counter reaches half the maximum value.
RM0090 Rev 19 1211/1751

1239
Bit 15 TGFMSCS: Transmitted good frames more single collision status

This bit is set when the transmitted, good frames after more than a single collision, counter
reaches half the maximum value.
Bit 14 TGFSCS: Transmitted good frames single collision status
This bit is set when the transmitted, good frames after a single collision, counter reaches half
the maximum value.
Ethernet MMC receive interrupt mask register (ETH_MMCRIMR)

The Ethernet MMC receive interrupt mask register maintains the masks for interrupts
generated when the receive statistic counters reach half their maximum value. (MSB of the
counter is set.) It is a 32-bit wide register.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RGUFM
RFCEM
RFAEM
rw rw rw

Bit 17 RGUFM: Received good unicast frames mask
Setting this bit masks the interrupt when the received, good unicast frames, counter
Bit 6 RFAEM: Received frames alignment error mask
Setting this bit masks the interrupt when the received frames, with alignment error, counter
Bit 5 RFCEM: Received frame CRC error mask
Setting this bit masks the interrupt when the received frames, with CRC error, counter
1212/1751 RM0090 Rev 19

Ethernet MMC transmit interrupt mask register (ETH_MMCTIMR)

The Ethernet MMC transmit interrupt mask register maintains the masks for interrupts
generated when the transmit statistic counters reach half their maximum value. (MSB of the
counter is set). It is a 32-bit wide register.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TGFMSCM
TGFSCM
TGFM
rw rw rw

Bit 21 TGFM: Transmitted good frames mask
Setting this bit masks the interrupt when the transmitted, good frames, counter reaches half
the maximum value.
Bit 15 TGFMSCM: Transmitted good frames more single collision mask
Setting this bit masks the interrupt when the transmitted good frames after more than a
single collision counter reaches half the maximum value.
Bit 14 TGFSCM: Transmitted good frames single collision mask
Setting this bit masks the interrupt when the transmitted good frames after a single collision
counter reaches half the maximum value.
Ethernet MMC transmitted good frames after a single collision counter

register (ETH_MMCTGFSCCR)
This register contains the number of successfully transmitted frames after a single collision
in Half-duplex mode.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TGFSCC
Bits 31:0 TGFSCC: Transmitted good frames single collision counter

Transmitted good frames after a single collision counter.
RM0090 Rev 19 1213/1751

1239
Ethernet MMC transmitted good frames after more than a single collision
counter register (ETH_MMCTGFMSCCR)
This register contains the number of successfully transmitted frames after more than a
single collision in Half-duplex mode.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TGFMSCC
Bits 31:0 TGFMSCC: Transmitted good frames more single collision counter
Transmitted good frames after more than a single collision counter
Ethernet MMC transmitted good frames counter register (ETH_MMCTGFCR)

This register contains the number of good frames transmitted.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TGFC
Bits 31:0 TGFC: Transmitted good frames counter
Ethernet MMC received frames with CRC error counter register

(ETH_MMCRFCECR)
This register contains the number of frames received with CRC error.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RFCEC
Bits 31:0 RFCEC: Received frames CRC error counter

Received frames with CRC error counter
1214/1751 RM0090 Rev 19

Ethernet MMC received frames with alignment error counter register

(ETH_MMCRFAECR)
This register contains the number of frames received with alignment (dribble) error.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RFAEC
Bits 31:0 RFAEC: Received frames alignment error counter

Received frames with alignment error counter
MMC received good unicast frames counter register (ETH_MMCRGUFCR)

Address offset: 0x01C4
This register contains the number of good unicast frames received.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RGUFC
Bits 31:0 RGUFC: Received good unicast frames counter
33.8.3 IEEE 1588 time stamp registers

This section describes the registers required to support precision network clock
synchronization functions under the IEEE 1588 standard.
Ethernet PTP time stamp control register (ETH_PTPTSCR)

Reset value: 0x0000 00002000
This register controls the time stamp generation and update logic.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TSSPTPOEFE
TSPTPPSV2E
TSSIPV4FE
TSSIPV6FE
TSPFFMAE
TSSMRME
TSSARFE
TSSEME
TTSARU
TSCNT
TSSSR
TSFCU
TSSTU
TSITE
TSSTI
Reserved
TSE
Reserved
RM0090 Rev 19 1215/1751

1239

Bit 18 TSPFFMAE: Time stamp PTP frame filtering MAC address enable
When set, this bit uses the MAC address (except for MAC address 0) to filter the PTP frames
when PTP is sent directly over Ethernet.
Bits 17:16 TSCNT: Time stamp clock node type
The following are the available types of clock node:
00: Ordinary clock
01: Boundary clock
10: End-to-end transparent clock
11: Peer-to-peer transparent clock
Bit 15 TSSMRME: Time stamp snapshot for message relevant to master enable
When this bit is set, the snapshot is taken for messages relevant to the master node only.
When this bit is cleared the snapshot is taken for messages relevant to the slave node only.
This is valid only for the ordinary clock and boundary clock nodes.
Bit 14 TSSEME: Time stamp snapshot for event message enable
When this bit is set, the time stamp snapshot is taken for event messages only (SYNC,
Delay_Req, Pdelay_Req or Pdelay_Resp). When this bit is cleared the snapshot is taken for
all other messages except for Announce, Management and Signaling.
Bit 13 TSSIPV4FE: Time stamp snapshot for IPv4 frames enable
When this bit is set, the time stamp snapshot is taken for IPv4 frames.
Bit 12 TSSIPV6FE: Time stamp snapshot for IPv6 frames enable
When this bit is set, the time stamp snapshot is taken for IPv6 frames.
Bit 11 TSSPTPOEFE: Time stamp snapshot for PTP over ethernet frames enable
When this bit is set, the time stamp snapshot is taken for frames which have PTP messages
in Ethernet frames (PTP over Ethernet) also. By default snapshots are taken for UDP-
IPEthernet PTP packets.
Bit 10 TSPTPPSV2E: Time stamp PTP packet snooping for version2 format enable
When this bit is set, the PTP packets are snooped using the version 2 format. When the bit is
cleared, the PTP packets are snooped using the version 1 format.
Note: IEEE 1588 Version 1 and Version 2 formats as indicated in IEEE standard 1588-2008
(Revision of IEEE STD. 1588-2002).
Bit 9 TSSSR: Time stamp subsecond rollover: digital or binary rollover control
When this bit is set, the Time stamp low register rolls over when the subsecond counter
reaches the value 0x3B9A C9FF (999 999 999 in decimal), and increments the Time Stamp
(high) seconds.
When this bit is cleared, the rollover value of the subsecond register reaches 0x7FFF FFFF.
The subsecond increment has to be programmed correctly depending on the PTP’s
reference clock frequency and this bit value.
Bit 8 TSSARFE: Time stamp snapshot for all received frames enable
When this bit is set, the time stamp snapshot is enabled for all frames received by the core.
Bit 5 TSARU: Time stamp addend register update
When this bit is set, the Time stamp addend register’s contents are updated to the PTP block
for fine correction. This bit is cleared when the update is complete. This register bit must be
read as zero before you can set it.
1216/1751 RM0090 Rev 19

Bit 4 TSITE: Time stamp interrupt trigger enable

When this bit is set, a time stamp interrupt is generated when the system time becomes
greater than the value written in the Target time register. When the Time stamp trigger
interrupt is generated, this bit is cleared.
Bit 3 TSSTU: Time stamp system time update
When this bit is set, the system time is updated (added to or subtracted from) with the value
specified in the Time stamp high update and Time stamp low update registers. Both the
TSSTU and TSSTI bits must be read as zero before you can set this bit. Once the update is
completed in hardware, this bit is cleared.
Bit 2 TSSTI: Time stamp system time initialize
When this bit is set, the system time is initialized (overwritten) with the value specified in the
Time stamp high update and Time stamp low update registers. This bit must be read as zero
before you can set it. When initialization is complete, this bit is cleared.
Bit 1 TSFCU: Time stamp fine or coarse update
When set, this bit indicates that the system time stamp is to be updated using the Fine
Update method. When cleared, it indicates the system time stamp is to be updated using the
Coarse method.
Bit 0 TSE: Time stamp enable
When this bit is set, time stamping is enabled for transmit and receive frames. When this bit
is cleared, the time stamp function is suspended and time stamps are not added for transmit
and receive frames. Because the maintained system time is suspended, you must always
initialize the time stamp feature (system time) after setting this bit high.
The table below indicates the messages for which a snapshot is taken depending on the
clock, enable master and enable snapshot for event message register settings.
Table 194. Time stamp snapshot dependency on registers bits

TSCNT TSSMRME TSSEME
Messages for which snapshots are taken
(bits 17:16) (bit 15)(1) (bit 14)
00 or 01 X(2) 0 SYNC, Follow_Up, Delay_Req, Delay_Resp

00 or 01 1 1 Delay_Req
00 or 01 0 1 SYNC
10 N/A 0 SYNC, Follow_Up, Delay_Req, Delay_Resp
10 N/A 1 SYNC, Follow_Up
SYNC, Follow_Up, Delay_Req, Delay_Resp,
11 N/A 0
Pdelay_Req, Pdelay_Resp
11 N/A 1 SYNC, Pdelay_Req, Pdelay_Resp
1. N/A = not applicable.
2. X = don’t care.
RM0090 Rev 19 1217/1751

1239
Ethernet PTP subsecond increment register (ETH_PTPSSIR)

This register contains the 8-bit value by which the subsecond register is incremented. In
Coarse update mode (TSFCU bit in ETH_PTPTSCR), the value in this register is added to
the system time every clock cycle of HCLK. In Fine update mode, the value in this register is
added to the system time whenever the accumulator gets an overflow.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
STSSI
Reserved

Bits 7:0 STSSI: System time subsecond increment
The value programmed in this register is added to the contents of the subsecond value of the
system time in every update.
For example, to achieve 20 ns accuracy, the value is: 20 / 0.467 = ~ 43 (or 0x2A).
Ethernet PTP time stamp high register (ETH_PTPTSHR)

This register contains the most significant (higher) 32 time bits. This read-only register
contains the seconds system time value. The Time stamp high register, along with Time
stamp low register, indicates the current value of the system time maintained by the MAC.
Though it is updated on a continuous basis.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
STS
Bits 31:0 STS: System time second

The value in this field indicates the current value in seconds of the System Time maintained
by the core.
1218/1751 RM0090 Rev 19

Ethernet PTP time stamp low register (ETH_PTPTSLR)

This register contains the least significant (lower) 32 time bits. This read-only register
contains the subsecond system time value.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
STPNS
STSS
Bit 31 STPNS: System time positive or negative sign

This bit indicates a positive or negative time value. When set, the bit indicates that time
representation is negative. When cleared, it indicates that time representation is positive.
Because the system time should always be positive, this bit is normally zero.
Bits 30:0 STSS: System time subseconds
The value in this field has the subsecond time representation, with 0.46 ns accuracy.
Ethernet PTP time stamp high update register (ETH_PTPTSHUR)

This register contains the most significant (higher) 32 bits of the time to be written to, added
to, or subtracted from the System Time value. The Time stamp high update register, along
with the Time stamp update low register, initializes or updates the system time maintained
by the MAC. You have to write both of these registers before setting the TSSTI or TSSTU
bits in the Time stamp control register.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TSUS
Bits 31:0 TSUS: Time stamp update second

The value in this field indicates the time, in seconds, to be initialized or added to the system
time.
RM0090 Rev 19 1219/1751

1239
Ethernet PTP time stamp low update register (ETH_PTPTSLUR)

This register contains the least significant (lower) 32 bits of the time to be written to, added
to, or subtracted from the System Time value.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TSUPNS
TSUSS
Bit 31 TSUPNS: Time stamp update positive or negative sign

This bit indicates positive or negative time value. When set, the bit indicates that time
representation is negative. When cleared, it indicates that time representation is positive.
When TSSTI is set (system time initialization) this bit should be zero. If this bit is set when
TSSTU is set, the value in the Time stamp update registers is subtracted from the system
time. Otherwise it is added to the system time.
Bits 30:0 TSUSS: Time stamp update subseconds
The value in this field indicates the subsecond time to be initialized or added to the system
time. This value has an accuracy of 0.46 ns (in other words, a value of 0x0000_0001 is
0.46 ns).
Ethernet PTP time stamp addend register (ETH_PTPTSAR)

This register is used by the software to readjust the clock frequency linearly to match the
master clock frequency. This register value is used only when the system time is configured
for Fine update mode (TSFCU bit in ETH_PTPTSCR). This register content is added to a
32-bit accumulator in every clock cycle and the system time is updated whenever the
accumulator overflows.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TSA
Bits 31:0 TSA: Time stamp addend

This register indicates the 32-bit time value to be added to the Accumulator register to
achieve time synchronization.
1220/1751 RM0090 Rev 19

Ethernet PTP target time high register (ETH_PTPTTHR)

This register contains the higher 32 bits of time to be compared with the system time for
interrupt event generation. The Target time high register, along with Target time low register,
is used to schedule an interrupt event (TSARU bit in ETH_PTPTSCR) when the system time
exceeds the value programmed in these registers.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TTSH
Bits 31:0 TTSH: Target time stamp high

This register stores the time in seconds. When the time stamp value matches or exceeds
both Target time stamp registers, the MAC, if enabled, generates an interrupt.
Ethernet PTP target time low register (ETH_PTPTTLR)

This register contains the lower 32 bits of time to be compared with the system time for
interrupt event generation.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TTSL
Bits 31:0 TTSL: Target time stamp low

This register stores the time in (signed) nanoseconds. When the value of the time stamp
matches or exceeds both Target time stamp registers, the MAC, if enabled, generates an
interrupt.
Ethernet PTP time stamp status register (ETH_PTPTSSR)

This register contains the time stamp status register.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TSTTR
TSSO
Reserved
ro ro
RM0090 Rev 19 1221/1751

1239

Bit 1 TSTTR: Time stamp target time reached
When set, this bit indicates that the value of the system time is greater than or equal to the
value specified in the Target time high and low registers. This bit is cleared when the
ETH_PTPTSSR register is read.
Bit 0 TSSO: Time stamp second overflow
When set, this bit indicates that the second value of the time stamp has overflowed beyond
0xFFFF FFFF.
Ethernet PTP PPS control register (ETH_PTPPPSCR)

This register controls the frequency of the PPS output.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PPSFREQ
Reserved
ro ro

Bits 3:0 PPSFREQ: PPS frequency selection
The PPS output frequency is set to 2PPSFREQ Hz.
0000: 1 Hz with a pulse width of 125 ms for binary rollover and, of 100 ms for digital rollover
0001: 2 Hz with 50% duty cycle for binary rollover (digital rollover not recommended)
...
Note: If digital rollover is used (TSSSR=1, bit 9 in ETH_PTPTSCR), it is recommended not to
use the PPS output with a frequency other than 1 Hz. Otherwise, with digital rollover,
the PPS output has irregular waveforms at higher frequencies (though its average
frequency will always be correct during any one-second window).
1222/1751 RM0090 Rev 19

33.8.4 DMA register description

This section defines the bits for each DMA register. Non-32 bit accesses are allowed as long
as the address is word-aligned.
Ethernet DMA bus mode register (ETH_DMABMR)

The bus mode register establishes the bus operating modes for the DMA.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FPM
USP
AAB
MB
SR
DA
FB
RDP PM PBL DSL
EDFE
Reserved
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rs

Bit 26 MB: Mixed burst
When this bit is set high and the FB bit is low, the AHB master interface starts all bursts of a
length greater than 16 with INCR (undefined burst). When this bit is cleared, it reverts to
fixed burst transfers (INCRx and SINGLE) for burst lengths of 16 and below.
Bit 25 AAB: Address-aligned beats
When this bit is set high and the FB bit equals 1, the AHB interface generates all bursts
aligned to the start address LS bits. If the FB bit equals 0, the first burst (accessing the data
buffer’s start address) is not aligned, but subsequent bursts are aligned to the address.
Bit 24 FPM: 4xPBL mode
When set high, this bit multiplies the PBL value programmed (bits [22:17] and bits [13:8])
four times. Thus the DMA transfers data in a maximum of 4, 8, 16, 32, 64 and 128 beats
depending on the PBL value.
Bit 23 USP: Use separate PBL
When set high, it configures the RxDMA to use the value configured in bits [22:17] as PBL
while the PBL value in bits [13:8] is applicable to TxDMA operations only. When this bit is
cleared, the PBL value in bits [13:8] is applicable for both DMA engines.
Bits 22:17 RDP: Rx DMA PBL
These bits indicate the maximum number of beats to be transferred in one RxDMA
transaction. This is the maximum value that is used in a single block read/write operation.
The RxDMA always attempts to burst as specified in RDP each time it starts a burst transfer
on the host bus. RDP can be programmed with permissible values of 1, 2, 4, 8, 16, and 32.
Any other value results in undefined behavior.
These bits are valid and applicable only when USP is set high.
Bit 16 FB: Fixed burst
This bit controls whether the AHB Master interface performs fixed burst transfers or not.
When set, the AHB uses only SINGLE, INCR4, INCR8 or INCR16 during start of normal
burst transfers. When reset, the AHB uses SINGLE and INCR burst transfer operations.
RM0090 Rev 19 1223/1751

1239
Bits 15:14 PM: Rx Tx priority ratio

RxDMA requests are given priority over TxDMA requests in the following ratio:
00: 1:1
01: 2:1
10: 3:1
11: 4:1
This is valid only when the DA bit is cleared.
Bits 13:8 PBL: Programmable burst length
These bits indicate the maximum number of beats to be transferred in one DMA transaction.
This is the maximum value that is used in a single block read/write operation. The DMA
always attempts to burst as specified in PBL each time it starts a burst transfer on the host
bus. PBL can be programmed with permissible values of 1, 2, 4, 8, 16, and 32. Any other
value results in undefined behavior. When USP is set, this PBL value is applicable for
TxDMA transactions only.
The PBL values have the following limitations:
– The maximum number of beats (PBL) possible is limited by the size of the Tx FIFO and Rx
FIFO.
– The FIFO has a constraint that the maximum beat supported is half the depth of the FIFO.
– If the PBL is common for both transmit and receive DMA, the minimum Rx FIFO and Tx
FIFO depths must be considered.
– Do not program out-of-range PBL values, because the system may not behave properly.
Bit 7 EDFE: Enhanced descriptor format enable
When this bit is set, the enhanced descriptor format is enabled and the descriptor size is
increased to 32 bytes (8 words). This is required when time stamping is activated (TSE=1,
ETH_PTPTSCR bit 0) or if IPv4 checksum offload is activated (IPCO=1, ETH_MACCR bit
10).
Bits 6:2 DSL: Descriptor skip length
This bit specifies the number of words to skip between two unchained descriptors. The
address skipping starts from the end of current descriptor to the start of next descriptor.
When DSL value equals zero, the descriptor table is taken as contiguous by the DMA, in
Ring mode.
Bit 1 DA: DMA Arbitration
0: Round-robin with Rx:Tx priority given in bits [15:14]
1: Rx has priority over Tx
Bit 0 SR: Software reset
When this bit is set, the MAC DMA controller resets all MAC Subsystem internal registers
and logic. It is cleared automatically after the reset operation has completed in all of the core
clock domains. Read a 0 value in this bit before re-programming any register of the core.
Ethernet DMA transmit poll demand register (ETH_DMATPDR)

This register is used by the application to instruct the DMA to poll the transmit descriptor list.
The transmit poll demand register enables the Transmit DMA to check whether or not the
current descriptor is owned by DMA. The Transmit Poll Demand command is given to wake
up the TxDMA if it is in Suspend mode. The TxDMA can go into Suspend mode due to an
underflow error in a transmitted frame or due to the unavailability of descriptors owned by
1224/1751 RM0090 Rev 19

transmit DMA. You can issue this command anytime and the TxDMA resets it once it starts
re-fetching the current descriptor from host memory.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TPD
rw_wt
Bits 31:0 TPD: Transmit poll demand

When these bits are written with any value, the DMA reads the current descriptor pointed to
by the ETH_DMACHTDR register. If that descriptor is not available (owned by Host),
transmission returns to the Suspend state and ETH_DMASR register bit 2 is asserted. If the
descriptor is available, transmission resumes.
EHERNET DMA receive poll demand register (ETH_DMARPDR)

This register is used by the application to instruct the DMA to poll the receive descriptor list.
The Receive poll demand register enables the receive DMA to check for new descriptors.
This command is given to wake up the RxDMA from Suspend state. The RxDMA can go into
Suspend state only due to the unavailability of descriptors owned by it.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RPD
rw_wt
Bits 31:0 RPD: Receive poll demand

When these bits are written with any value, the DMA reads the current descriptor pointed to
by the ETH_DMACHRDR register. If that descriptor is not available (owned by Host),
reception returns to the Suspended state and ETH_DMASR register bit 7 is not asserted. If
the descriptor is available, the Receive DMA returns to active state.
Ethernet DMA receive descriptor list address register (ETH_DMARDLAR)

The Receive descriptor list address register points to the start of the receive descriptor list.
The descriptor lists reside in the STM32F4xx's physical memory space and must be word-
aligned. The DMA internally converts it to bus-width aligned address by making the
corresponding LS bits low. Writing to the ETH_DMARDLAR register is permitted only when
reception is stopped. When stopped, the ETH_DMARDLAR register must be written to
before the receive Start command is given.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SRL
RM0090 Rev 19 1225/1751

1239
Bits 31:0 SRL: Start of receive list

This field contains the base address of the first descriptor in the receive descriptor list. The
LSB bits [1/2/3:0] for 32/64/128-bit bus width) are internally ignored and taken as all-zero by
the DMA. Hence these LSB bits are read only.
Ethernet DMA transmit descriptor list address register (ETH_DMATDLAR)

The Transmit descriptor list address register points to the start of the transmit descriptor list.
The descriptor lists reside in the STM32F4xx's physical memory space and must be word-
aligned. The DMA internally converts it to bus-width-aligned address by taking the
corresponding LSB to low. Writing to the ETH_DMATDLAR register is permitted only when
transmission has stopped. Once transmission has stopped, the ETH_DMATDLAR register
can be written before the transmission Start command is given.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
STL
Bits 31:0 STL: Start of transmit list

This field contains the base address of the first descriptor in the transmit descriptor list. The
LSB bits [1/2/3:0] for 32/64/128-bit bus width) are internally ignored and taken as all-zero
by the DMA. Hence these LSB bits are read-only.
Ethernet DMA status register (ETH_DMASR)

The Status register contains all the status bits that the DMA reports to the application. The
ETH_DMASR register is usually read by the software driver during an interrupt service
routine or polling. Most of the fields in this register cause the host to be interrupted. The
ETH_DMASR register bits are not cleared when read. Writing 1 to (unreserved) bits in
ETH_DMASR register[16:0] clears them and writing 0 has no effect. Each field (bits [16:0])
can be masked by masking the appropriate bit in the ETH_DMAIER register.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MMCS
RWTS
PMTS
RBUS
RPSS
TBUS
FBES
TPSS
TSTS
TJTS
ROS
RPS
ERS
EBS
TUS
TPS
ETS
NIS
AIS
RS
Reserved
Reserved
Reserved
TS
rc- rc- rc- rc- rc- rc- rc- rc- rc- rc- rc- rc- rc- rc- rc-
w1 w1 w1 w1 w1 w1 w1 w1 w1 w1 w1 w1 w1 w1 w1
1226/1751 RM0090 Rev 19


Bit 29 TSTS: Time stamp trigger status
This bit indicates an interrupt event in the MAC core's Time stamp generator block. The
software must read the MAC core’s status register, clearing its source (bit 9), to reset this bit
to 0. When this bit is high an interrupt is generated if enabled.
Bit 28 PMTS: PMT status
This bit indicates an event in the MAC core’s PMT. The software must read the
corresponding registers in the MAC core to get the exact cause of interrupt and clear its
source to reset this bit to 0. The interrupt is generated when this bit is high if enabled.
Bit 27 MMCS: MMC status
This bit reflects an event in the MMC of the MAC core. The software must read the
corresponding registers in the MAC core to get the exact cause of interrupt and clear the
source of interrupt to make this bit as 0. The interrupt is generated when this bit is high if
enabled.
Bits 25:23 EBS: Error bits status
These bits indicate the type of error that caused a bus error (error response on the AHB
interface). Valid only with the fatal bus error bit (ETH_DMASR register [13]) set. This field
does not generate an interrupt.
Bit 231 Error during data transfer by TxDMA
0 Error during data transfer by RxDMA
Bit 24 1 Error during read transfer
0 Error during write transfer
Bit 25 1 Error during descriptor access
0 Error during data buffer access
Bits 22:20 TPS: Transmit process state
These bits indicate the Transmit DMA FSM state. This field does not generate an interrupt.
000: Stopped; Reset or Stop Transmit Command issued
001: Running; Fetching transmit transfer descriptor
010: Running; Waiting for status
011: Running; Reading Data from host memory buffer and queuing it to transmit buffer (Tx
FIFO)
100, 101: Reserved for future use
110: Suspended; Transmit descriptor unavailable or transmit buffer underflow
111: Running; Closing transmit descriptor
Bits 19:17 RPS: Receive process state
These bits indicate the Receive DMA FSM state. This field does not generate an interrupt.
000: Stopped: Reset or Stop Receive Command issued
001: Running: Fetching receive transfer descriptor
010: Reserved for future use
011: Running: Waiting for receive packet
100: Suspended: Receive descriptor unavailable
101: Running: Closing receive descriptor
110: Reserved for future use
111: Running: Transferring the receive packet data from receive buffer to host memory
RM0090 Rev 19 1227/1751

1239
Bit 16 NIS: Normal interrupt summary

The normal interrupt summary bit value is the logical OR of the following when the
corresponding interrupt bits are enabled in the ETH_DMAIER register:
– ETH_DMASR [0]: Transmit interrupt
– ETH_DMASR [2]: Transmit buffer unavailable
– ETH_DMASR [6]: Receive interrupt
– ETH_DMASR [14]: Early receive interrupt
Only unmasked bits affect the normal interrupt summary bit.
This is a sticky bit and it must be cleared (by writing a 1 to this bit) each time a corresponding
bit that causes NIS to be set is cleared.
Bit 15 AIS: Abnormal interrupt summary
The abnormal interrupt summary bit value is the logical OR of the following when the
corresponding interrupt bits are enabled in the ETH_DMAIER register:
– ETH_DMASR [1]:Transmit process stopped
– ETH_DMASR [3]:Transmit jabber timeout
– ETH_DMASR [4]: Receive FIFO overflow
– ETH_DMASR [5]: Transmit underflow
– ETH_DMASR [7]: Receive buffer unavailable
– ETH_DMASR [8]: Receive process stopped
– ETH_DMASR [9]: Receive watchdog timeout
– ETH_DMASR [10]: Early transmit interrupt
– ETH_DMASR [13]: Fatal bus error
Only unmasked bits affect the abnormal interrupt summary bit.
This is a sticky bit and it must be cleared each time a corresponding bit that causes AIS to be
set is cleared.
Bit 14 ERS: Early receive status
This bit indicates that the DMA had filled the first data buffer of the packet. Receive Interrupt
ETH_DMASR [6] automatically clears this bit.
Bit 13 FBES: Fatal bus error status
This bit indicates that a bus error occurred, as detailed in [25:23]. When this bit is set, the
corresponding DMA engine disables all its bus accesses.
Bit 10 ETS: Early transmit status
This bit indicates that the frame to be transmitted was fully transferred to the Transmit FIFO.
Bit 9 RWTS: Receive watchdog timeout status
This bit is asserted when a frame with a length greater than 2 048 bytes is received.
Bit 8 RPSS: Receive process stopped status
This bit is asserted when the receive process enters the Stopped state.
Bit 7 RBUS: Receive buffer unavailable status
This bit indicates that the next descriptor in the receive list is owned by the host and cannot
be acquired by the DMA. Receive process is suspended. To resume processing receive
descriptors, the host should change the ownership of the descriptor and issue a Receive Poll
Demand command. If no Receive Poll Demand is issued, receive process resumes when the
next recognized incoming frame is received. ETH_DMASR [7] is set only when the previous
receive descriptor was owned by the DMA.
1228/1751 RM0090 Rev 19

Bit 6 RS: Receive status

This bit indicates the completion of the frame reception. Specific frame status information
has been posted in the descriptor. Reception remains in the Running state.
Bit 5 TUS: Transmit underflow status
This bit indicates that the transmit buffer had an underflow during frame transmission.
Transmission is suspended and an underflow error TDES0[1] is set.
Bit 4 ROS: Receive overflow status
This bit indicates that the receive buffer had an overflow during frame reception. If the partial
frame is transferred to the application, the overflow status is set in RDES0[11].
Bit 3 TJTS: Transmit jabber timeout status
This bit indicates that the transmit jabber timer expired, meaning that the transmitter had
been excessively active. The transmission process is aborted and placed in the Stopped
state. This causes the transmit jabber timeout TDES0[14] flag to be asserted.
Bit 2 TBUS: Transmit buffer unavailable status
This bit indicates that the next descriptor in the transmit list is owned by the host and cannot
be acquired by the DMA. Transmission is suspended. Bits [22:20] explain the transmit
process state transitions. To resume processing transmit descriptors, the host should change
the ownership of the bit of the descriptor and then issue a Transmit Poll Demand command.
Bit 1 TPSS: Transmit process stopped status
This bit is set when the transmission is stopped.
Bit 0 TS: Transmit status
This bit indicates that frame transmission is finished and TDES1[31] is set in the first
descriptor.
Ethernet DMA operation mode register (ETH_DMAOMR)

The operation mode register establishes the Transmit and Receive operating modes and
commands. The ETH_DMAOMR register should be the last CSR to be written as part of
DMA initialization.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DTCEFD
FUGF
DFRF
Reserved
Reserved
Reserved
OSF
RTC
RSF
TTC
TSF
FEF
FTF
SR
ST
rw rw rw rw rs rw rw rw rw rw rw rw rw rw rw

Bit 26 DTCEFD: Dropping of TCP/IP checksum error frames disable
When this bit is set, the core does not drop frames that only have errors detected by the
receive checksum offload engine. Such frames do not have any errors (including FCS error)
in the Ethernet frame received by the MAC but have errors in the encapsulated payload only.
When this bit is cleared, all error frames are dropped if the FEF bit is reset.
RM0090 Rev 19 1229/1751

1239
Bit 25 RSF: Receive store and forward

When this bit is set, a frame is read from the Rx FIFO after the complete frame has been
written to it, ignoring RTC bits. When this bit is cleared, the Rx FIFO operates in Cut-through
mode, subject to the threshold specified by the RTC bits.
Bit 24 DFRF: Disable flushing of received frames
When this bit is set, the RxDMA does not flush any frames due to the unavailability of
receive descriptors/buffers as it does normally when this bit is cleared (see Receive process
suspended).
Bit 21 TSF: Transmit store and forward
When this bit is set, transmission starts when a full frame resides in the Transmit FIFO.
When this bit is set, the TTC values specified by the ETH_DMAOMR register bits [16:14] are
ignored.
When this bit is cleared, the TTC values specified by the ETH_DMAOMR register bits
[16:14] are taken into account.
This bit should be changed only when transmission is stopped.
Bit 20 FTF: Flush transmit FIFO
When this bit is set, the transmit FIFO controller logic is reset to its default values and thus
all data in the Tx FIFO are lost/flushed. This bit is cleared internally when the flushing
operation is complete. The Operation mode register should not be written to until this bit is
cleared.
Bits 16:14 TTC: Transmit threshold control
These three bits control the threshold level of the Transmit FIFO. Transmission starts when
the frame size within the Transmit FIFO is larger than the threshold. In addition, full frames
with a length less than the threshold are also transmitted. These bits are used only when the
TSF bit (Bit 21) is cleared.
000: 64
001: 128
010: 192
011: 256
100: 40
101: 32
110: 24
111: 16
Bit 13 ST: Start/stop transmission
When this bit is set, transmission is placed in the Running state, and the DMA checks the
transmit list at the current position for a frame to be transmitted. Descriptor acquisition is
attempted either from the current position in the list, which is the transmit list base address
set by the ETH_DMATDLAR register, or from the position retained when transmission was
stopped previously. If the current descriptor is not owned by the DMA, transmission enters
the Suspended state and the transmit buffer unavailable bit (ETH_DMASR [2]) is set. The
Start Transmission command is effective only when transmission is stopped. If the command
is issued before setting the DMA ETH_DMATDLAR register, the DMA behavior is
unpredictable.
When this bit is cleared, the transmission process is placed in the Stopped state after
completing the transmission of the current frame. The next descriptor position in the transmit
list is saved, and becomes the current position when transmission is restarted. The Stop
Transmission command is effective only when the transmission of the current frame is
complete or when the transmission is in the Suspended state.
1230/1751 RM0090 Rev 19


Bit 7 FEF: Forward error frames
When this bit is set, all frames except runt error frames are forwarded to the DMA.
When this bit is cleared, the Rx FIFO drops frames with error status (CRC error, collision
error, giant frame, watchdog timeout, overflow). However, if the frame’s start byte (write)
pointer is already transferred to the read controller side (in Threshold mode), then the
frames are not dropped. The Rx FIFO drops the error frames if that frame's start byte is not
transferred (output) on the ARI bus.
Bit 6 FUGF: Forward undersized good frames
When this bit is set, the Rx FIFO forwards undersized frames (frames with no error and
length less than 64 bytes) including pad-bytes and CRC).
When this bit is cleared, the Rx FIFO drops all frames of less than 64 bytes, unless such a
frame has already been transferred due to lower value of receive threshold (e.g., RTC = 01).
Bits 4:3 RTC: Receive threshold control
These two bits control the threshold level of the Receive FIFO. Transfer (request) to DMA
starts when the frame size within the Receive FIFO is larger than the threshold. In addition,
full frames with a length less than the threshold are transferred automatically.
Note: Note that value of 11 is not applicable if the configured Receive FIFO size is 128 bytes.
Note: These bits are valid only when the RSF bit is zero, and are ignored when the RSF bit is
set to 1.
00: 64
01: 32
10: 96
11: 128
Bit 2 OSF: Operate on second frame
When this bit is set, this bit instructs the DMA to process a second frame of Transmit data
even before status for first frame is obtained.
Bit 1 SR: Start/stop receive
When this bit is set, the receive process is placed in the Running state. The DMA attempts to
acquire the descriptor from the receive list and processes incoming frames. Descriptor
acquisition is attempted from the current position in the list, which is the address set by the
DMA ETH_DMARDLAR register or the position retained when the receive process was
previously stopped. If no descriptor is owned by the DMA, reception is suspended and the
receive buffer unavailable bit (ETH_DMASR [7]) is set. The Start Receive command is
effective only when reception has stopped. If the command was issued before setting the
DMA ETH_DMARDLAR register, the DMA behavior is unpredictable.
When this bit is cleared, RxDMA operation is stopped after the transfer of the current frame.
The next descriptor position in the receive list is saved and becomes the current position
when the receive process is restarted. The Stop Receive command is effective only when
the Receive process is in either the Running (waiting for receive packet) or the Suspended
state.
RM0090 Rev 19 1231/1751

1239
Ethernet DMA interrupt enable register (ETH_DMAIER)

The Interrupt enable register enables the interrupts reported by ETH_DMASR. Setting a bit
to 1 enables a corresponding interrupt. After a hardware or software reset, all interrupts are
disabled.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RWTIE
RBUIE
RPSIE
TBUIE
FBEIE
TPSIE
TJTIE
ROIE
Reserved
NISE
ERIE
AISE
TUIE
ETIE
RIE
TIE
Reserved

Bit 16 NISE: Normal interrupt summary enable
When this bit is set, a normal interrupt is enabled. When this bit is cleared, a normal
interrupt is disabled. This bit enables the following bits:
– ETH_DMASR [0]: Transmit Interrupt
– ETH_DMASR [2]: Transmit buffer unavailable
– ETH_DMASR [6]: Receive interrupt
– ETH_DMASR [14]: Early receive interrupt
Bit 15 AISE: Abnormal interrupt summary enable
When this bit is set, an abnormal interrupt is enabled. When this bit is cleared, an abnormal
interrupt is disabled. This bit enables the following bits:
– ETH_DMASR [1]: Transmit process stopped
– ETH_DMASR [3]: Transmit jabber timeout
– ETH_DMASR [4]: Receive overflow
– ETH_DMASR [5]: Transmit underflow
– ETH_DMASR [7]: Receive buffer unavailable
– ETH_DMASR [8]: Receive process stopped
– ETH_DMASR [9]: Receive watchdog timeout
– ETH_DMASR [10]: Early transmit interrupt
– ETH_DMASR [13]: Fatal bus error
Bit 14 ERIE: Early receive interrupt enable
When this bit is set with the normal interrupt summary enable bit (ETH_DMAIER
register[16]), the early receive interrupt is enabled.
When this bit is cleared, the early receive interrupt is disabled.
Bit 13 FBEIE: Fatal bus error interrupt enable
When this bit is set with the abnormal interrupt summary enable bit (ETH_DMAIER
register[15]), the fatal bus error interrupt is enabled.
When this bit is cleared, the fatal bus error enable interrupt is disabled.
Bit 10 ETIE: Early transmit interrupt enable
When this bit is set with the abnormal interrupt summary enable bit (ETH_DMAIER register
[15]), the early transmit interrupt is enabled.
When this bit is cleared, the early transmit interrupt is disabled.
1232/1751 RM0090 Rev 19

Bit 9 RWTIE: receive watchdog timeout interrupt enable

register[15]), the receive watchdog timeout interrupt is enabled.
When this bit is cleared, the receive watchdog timeout interrupt is disabled.
Bit 8 RPSIE: Receive process stopped interrupt enable
register[15]), the receive stopped interrupt is enabled. When this bit is cleared, the receive
stopped interrupt is disabled.
Bit 7 RBUIE: Receive buffer unavailable interrupt enable
register[15]), the receive buffer unavailable interrupt is enabled.
When this bit is cleared, the receive buffer unavailable interrupt is disabled.
Bit 6 RIE: Receive interrupt enable
register[16]), the receive interrupt is enabled.
When this bit is cleared, the receive interrupt is disabled.
Bit 5 TUIE: Underflow interrupt enable
register[15]), the transmit underflow interrupt is enabled.
When this bit is cleared, the underflow interrupt is disabled.
Bit 4 ROIE: Overflow interrupt enable
register[15]), the receive overflow interrupt is enabled.
When this bit is cleared, the overflow interrupt is disabled.
Bit 3 TJTIE: Transmit jabber timeout interrupt enable
register[15]), the transmit jabber timeout interrupt is enabled.
When this bit is cleared, the transmit jabber timeout interrupt is disabled.
Bit 2 TBUIE: Transmit buffer unavailable interrupt enable
register[16]), the transmit buffer unavailable interrupt is enabled.
When this bit is cleared, the transmit buffer unavailable interrupt is disabled.
Bit 1 TPSIE: Transmit process stopped interrupt enable
register[15]), the transmission stopped interrupt is enabled.
When this bit is cleared, the transmission stopped interrupt is disabled.
Bit 0 TIE: Transmit interrupt enable
register[16]), the transmit interrupt is enabled.
When this bit is cleared, the transmit interrupt is disabled.
The Ethernet interrupt is generated only when the TSTS or PMTS bits of the DMA Status
register is asserted with their corresponding interrupt are unmasked, or when the NIS/AIS
Status bit is asserted and the corresponding Interrupt Enable bits (NISE/AISE) are enabled.
RM0090 Rev 19 1233/1751

1239
Ethernet DMA missed frame and buffer overflow counter register

(ETH_DMAMFBOCR)
The DMA maintains two counters to track the number of missed frames during reception.
This register reports the current value of the counter. The counter is used for diagnostic
purposes. Bits [15:0] indicate missed frames due to the STM32F4xx buffer being
unavailable (no receive descriptor was available). Bits [27:17] indicate missed frames due to
Rx FIFO overflow conditions and runt frames (good frames of less than 64 bytes).
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OMFC
OFOC
MFA
MFC
Reserved
rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_
r r r r r r r r r r r r r r r r r r r r r r r r r r r r r

Bit 28 OFOC: Overflow bit for FIFO overflow counter
Bits 27:17 MFA: Missed frames by the application
Indicates the number of frames missed by the application
Bit 16 OMFC: Overflow bit for missed frame counter
Bits 15:0 MFC: Missed frames by the controller
Indicates the number of frames missed by the Controller due to the host receive buffer being
unavailable. This counter is incremented each time the DMA discards an incoming frame.
Ethernet DMA receive status watchdog timer register (ETH_DMARSWTR)

This register, when written with a non-zero value, enables the watchdog timer for the receive
status (RS, ETH_DMASR[6]).
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RSWTC
Reserved

Bits 7:0 RSWTC: Receive status (RS) watchdog timer count
Indicates the number of HCLK clock cycles multiplied by 256 for which the watchdog timer
is set. The watchdog timer gets triggered with the programmed value after the RxDMA
completes the transfer of a frame for which the RS status bit is not set due to the setting of
RDES1[31] in the corresponding descriptor. When the watchdog timer runs out, the RS bit
is set and the timer is stopped. The watchdog timer is reset when the RS bit is set high due
to automatic setting of RS as per RDES1[31] of any received frame.
1234/1751 RM0090 Rev 19

Ethernet DMA current host transmit descriptor register (ETH_DMACHTDR)

The Current host transmit descriptor register points to the start address of the current
transmit descriptor read by the DMA.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
HTDAP
Bits 31:0 HTDAP: Host transmit descriptor address pointer

Cleared . Pointer updated by DMA during operation.
Ethernet DMA current host receive descriptor register (ETH_DMACHRDR)

The Current host receive descriptor register points to the start address of the current receive
descriptor read by the DMA.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
HRDAP
Bits 31:0 HRDAP: Host receive descriptor address pointer

Cleared On Reset. Pointer updated by DMA during operation.
Ethernet DMA current host transmit buffer address register

(ETH_DMACHTBAR)
The Current host transmit buffer address register points to the current transmit buffer
address being read by the DMA.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
HTBAP
Bits 31:0 HTBAP: Host transmit buffer address pointer

RM0090 Rev 19 1235/1751

1239
Ethernet DMA current host receive buffer address register

(ETH_DMACHRBAR)
The current host receive buffer address register points to the current receive buffer address
being read by the DMA.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
HRBAP
Bits 31:0 HRBAP: Host receive buffer address pointer

33.8.5 Ethernet register maps

Table 195 gives the ETH register map and reset values.
Table 195. Ethernet register map and reset values

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
eserved
APCS
CSTF
Reserved
Reserved
Reserved
Reserved
IPCO
ROD
CSD
FES
WD
ETH_MACCR IFG
DM
RD
DC
LM
RE
TE
JD
BL
0x00 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DAIF
SAIF
PAM
HPF
PCF
BFD
SAF
ETH_MACFFR
HM
PM
HU
RA
0x04 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0
ETH_MACHTHR HTH[31:0]
0x08
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ETH_MACHTLR HTL[31:0]
0x0C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
M M
ETH_MACMIIAR PA MR
CR
0x10 Reserved W B
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ETH_MACMIIDR MD
0x14 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
FCB/BPA
ZQPD
UPFD
RFCE
TFCE
Reserved
ETH_MACFCR PT PLT
0x18 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
VLANTC
ETH_
VLANTI
0x1C MACVLANTR Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ETH_
Frame filter reg0\Frame filter reg1\Frame filter reg2\Frame filter reg3\Frame filter reg4\...\Frame filter reg7
0x28 MACRWUFFR
Reset value 0
1236/1751 RM0090 Rev 19

Table 195. Ethernet register map and reset values (continued)

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
WFFRPR
ETH_
Reserved
Reserved
WFR
MPR
WFE
MPE
0x2C
GU
PD
MACPMTCSR Reserved
MSFRWCS
MMRPEA
TFNEGU
RFWRA
MMTEA
MTFCS
RFRCS
TFWA
ETH_
TFRS
Reserved
Reserved
Reserved
Reserved
RFFL
MTP
TFF
0x34 MACDBGR Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
MMCRS
MMCTS
MMCS
PMTS
Reserved
TSTS
ETH_MACSR Reserve
0x38 Reserved
d
PMTIM
TSTIM
ETH_MACIMR Reserve
d
Reset value 0 0
ETH_MACA0HR Reserved MACA0H

MO
0x40
Reset value 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
ETH_MACA0LR MACA0L
0x44
Reset value 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
ETH_MACA1HR MBC[6:0] MACA1H

AE
SA
0x48 Reserved
Reset value 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
ETH_MACA1LR MACA1L
0x4C
Reset value 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
ETH_MACA2HR MBC MACA2H

AE
SA
0x50 Reserved
Reset value 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
ETH_MACA2LR MACA2L
0x54
Reset value 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
ETH_MACA3HR MBC MACA3H

AE
SA
0x58 Reserved
Reset value 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
ETH_MACA3LR MACA3L
0x5C
Reset value 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
MCFHP
ROR
MCP
MCF
CSR
ETH_MMCCR
CR
0x100 Reserved
RGUFS
RFCES
RFAES
ETH_MMCRIR
Reset value 0 0 0
TGFMSCS
TGFSCS
TGFS
ETH_MMCTIR
Reset value 0 0 0
RGUFM
RFCEM
RFAEM
ETH_MMCRIMR
0x10C Reserved Reserved Reserved
Reset value 0 0 0
RM0090 Rev 19 1237/1751

1239

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
TGFMSCM
TGFSCM
TGFM
ETH_MMCTIMR
Reset value 0 0 0
ETH_MMCTGFS
TGFSCC
0x14C CCR
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ETH_MMCTGF
TGFMSCC
0x150 MSCCR
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ETH_MMCTGF
TGFC
0x168 CR
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ETH_MMCRFC
RFCEC
0x194 ECR
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ETH_MMCRFAE
RFAEC
0x198 CR
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ETH_MMCRGU
RGUFC
0x1C4 FCR
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
TSSPTPOEFE
TSPTPPSV2E
TSSIPV4FE
TSSIPV6FE
TSPFFMAE
TSSMRME
TSSARFE
TSSEME
TTSARU
TSCNT
TSSSR
TSFCU
TSSTU
TSITE
TSSTI
Reserved
TSE
ETH_PTPTSCR
0x700 Reserved
Reset value 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0
ETH_PTPSSIR STSSI
0x704 Reserved
Reset value 0 0 0 0 0 0 0 0
ETH_PTPTSHR STS[31:0]
0x708
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
STPNS
ETH_PTPTSLR STSS
0x70C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ETH_PTPTSHU
TSUS
0x710 R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
TSUPNS
ETH_PTPTSLU
TSUSS
0x714 R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ETH_PTPTSAR TSA
0x718
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ETH_PTPTTHR TTSH
0x71C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ETH_PTPTTLR TTSL
0x720
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1238/1751 RM0090 Rev 19


Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
TSTTR
TSSO
ETH_PTPTSSR
0x728 Reserved
Reset value 0 0
ETH_PTPPPSC PPS
0x72C R Reserved FREQ
Reset value 0 0 0
EDFE
FPM
USP
AAB
ETH_DMABMR MB RDP PM PBL DSL
SR
DA
FB
0x1000 Reserved
Reset value 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1
ETH_DMATPDR TPD
0x1004
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ETH_DMARPDR RPD
0x1008
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ETH_DMARDLA
SRL
0x100C R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ETH_DMATDLA
STL
0x1010 R
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
MMCS
RWTS
PMTS
RBUS
RPSS
TBUS
FBES
TPSS
Reserved
TSTS
Reserved
Reserved
TJTS
ROS
RPS
ERS
EBS
TUS
TPS
ETS
NIS
ETH_DMASR
AIS
RS
TS
0x1014
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DTCEFD
FUGF
DFRF
Reserved
Reserved
Reserved
OSF
RTC
RSF
TTC
TSF
FEF
FTF
ETH_DMAOMR
SR
ST
Reserve
d
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RWTIE
RBUIE
RPSIE
TBUIE
FBEIE
TPSIE
TJTIE
Reserved
ROIE
NISE
ERIE
AISE
TUIE
ETIE
RIE
ETH_DMAIER
TIE
0x101C Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
OMFC
OFOC
ETH_DMAMFB
MFA
Reserve MFC
0x1020 OCR
d
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ETH_
RSWTC
0x1024 DMARSWTR Reserved
Reset value 0 0 0 0 0 0 0 0
ETH_
HTDAP
0x1048 DMACHTDR
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ETH_
HRDAP
0x104C DMACHRDR
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ETH_
HTBAP
0x1050 DMACHTBAR
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ETH_
HRBAP
0x1054 DMACHRBAR
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RM0090 Rev 19 1239/1751

1239
USB on-the-go full-speed (OTG_FS) RM0090
34 USB on-the-go full-speed (OTG_FS)
34.1 OTG_FS introduction

Portions Copyright (c) 2004, 2005 Synopsys, Inc. All rights reserved. Used with permission.
This section presents the architecture and the programming model of the OTG_FS
controller.
The following acronyms are used throughout this section:
FS Full-speed
LS Low-speed
MAC Media access controller
OTG On-the-go
PFC Packet FIFO controller
PHY Physical layer
USB Universal serial bus
UTMI USB 2.0 transceiver macrocell interface (UTMI)
References are made to the following documents:

• USB On-The-Go Supplement, Revision 1.3
• Universal Serial Bus Revision 2.0 Specification
The OTG_FS is a dual-role device (DRD) controller that supports both device and host
functions and is fully compliant with the On-The-Go Supplement to the USB 2.0
Specification. It can also be configured as a host-only or device-only controller, fully
compliant with the USB 2.0 Specification. In host mode, the OTG_FS supports full-speed
(FS, 12 Mbits/s) and low-speed (LS, 1.5 Mbits/s) transfers whereas in device mode, it only
supports full-speed (FS, 12 Mbits/s) transfers. The OTG_FS supports both HNP and SRP.
The only external device required is a charge pump for VBUS in host mode.
1240/1751 RM0090 Rev 19

RM0090 USB on-the-go full-speed (OTG_FS)
34.2 OTG_FS main features

The main features can be divided into three categories: general, host-mode and device-
mode features.
34.2.1 General features

The OTG_FS interface general features are the following:
• It is USB-IF certified to the Universal Serial Bus Specification Rev 2.0
• It includes full support (PHY) for the optional On-The-Go (OTG) protocol detailed in the
On-The-Go Supplement Rev 1.3 specification
– Integrated support for A-B Device Identification (ID line)
– Integrated support for host Negotiation Protocol (HNP) and Session Request
Protocol (SRP)
– It allows host to turn VBUS off to conserve battery power in OTG applications
– It supports OTG monitoring of VBUS levels with internal comparators
– It supports dynamic host-peripheral switch of role
• It is software-configurable to operate as:
– SRP capable USB FS Peripheral (B-device)
– SRP capable USB FS/LS host (A-device)
– USB On-The-Go Full-Speed Dual Role device
• It supports FS SOF and LS Keep-alives with
– SOF pulse PAD connectivity (OTG_FS_SOF)
– SOF pulse internal connection to timer2 (TIM2)
– Configurable framing period
– Configurable end of frame interrupt
• It includes power saving features such as system stop during USB Suspend, switch-off
of clock domains internal to the digital core, PHY and DFIFO power management
• It features a dedicated RAM of 1.25 Kbytes with advanced FIFO control:
– Configurable partitioning of RAM space into different FIFOs for flexible and
efficient use of RAM
– Each FIFO can hold multiple packets
– Dynamic memory allocation
– Configurable FIFO sizes that are not powers of 2 to allow the use of contiguous
memory locations
• It guarantees max USB bandwidth for up to one frame (1ms) without system
intervention
RM0090 Rev 19 1241/1751

1380
34.2.2 Host-mode features

The OTG_FS interface main features and requirements in host-mode are the following:
• External charge pump for VBUS voltage generation.
• Up to 8 host channels (pipes): each channel is dynamically reconfigurable to allocate
any type of USB transfer.
• Built-in hardware scheduler holding:
– Up to 8 interrupt plus isochronous transfer requests in the periodic hardware
queue
– Up to 8 control plus bulk transfer requests in the non-periodic hardware queue
• Management of a shared RX FIFO, a periodic TX FIFO and a nonperiodic TX FIFO for
efficient usage of the USB data RAM.
34.2.3 Peripheral-mode features

The OTG_FS interface main features in peripheral-mode are the following:
• 1 bidirectional control endpoint0
• 3 IN endpoints (EPs) configurable to support Bulk, Interrupt or Isochronous transfers
• 3 OUT endpoints configurable to support Bulk, Interrupt or Isochronous transfers
• Management of a shared Rx FIFO and a Tx-OUT FIFO for efficient usage of the USB
data RAM
• Management of up to 4 dedicated Tx-IN FIFOs (one for each active IN EP) to put less
load on the application
• Support for the soft disconnect feature.
1242/1751 RM0090 Rev 19

34.3 OTG_FS functional description

Figure 386. OTG full-speed block diagram
Cortex® core
OTG_FS_DP
Power
USB2.0 OTG OTG_FS_DM
and
OTG FS UTMIFS FS
clock OTG_FS_ID
core PHY
controller USB suspend
USB clock at 48 MHz System clock domain
USB clock OTG_FS_VBUS
domain
RAM bus
Universal serial bus
OTG_FS_SOF
1.25 Kbyte
USB data
FIFOs
MS19928V4
34.3.1 OTG pins
Table 196. OTG_FS input/output pins

Signal name Signal type Description
OTG_FS_DP Digital input/output USB OTG D+ line

OTG_FS_DM Digital input/output USB OTG D- line
OTG_FS_ID Digital input USB OTG ID
OTG_FS_VBUS Analog input USB OTG VBUS
OTG_FS_SOF Digital output USB OTG Start Of Frame (visibility)
34.3.2 OTG full-speed core

The USB OTG FS receives the 48 MHz ±0.25% clock from the reset and clock controller
(RCC), via an external quartz. The USB clock is used for driving the 48 MHz domain at full-
speed (12 Mbit/s) and must be enabled prior to configuring the OTG FS core.
The CPU reads and writes from/to the OTG FS core registers through the AHB peripheral
bus. It is informed of USB events through the single USB OTG interrupt line described in
Section 34.15: OTG_FS interrupts.
RM0090 Rev 19 1243/1751

1380
The CPU submits data over the USB by writing 32-bit words to dedicated OTG_FS locations
(push registers). The data are then automatically stored into Tx-data FIFOs configured
within the USB data RAM. There is one Tx-FIFO push register for each in-endpoint
(peripheral mode) or out-channel (host mode).
The CPU receives the data from the USB by reading 32-bit words from dedicated OTG_FS
addresses (pop registers). The data are then automatically retrieved from a shared Rx-FIFO
configured within the 1.25 KB USB data RAM. There is one Rx-FIFO pop register for each
out-endpoint or in-channel.
The USB protocol layer is driven by the serial interface engine (SIE) and serialized over the
USB by the full-/low-speed transceiver module within the on-chip physical layer (PHY).
34.3.3 Full-speed OTG PHY

The embedded full-speed OTG PHY is controlled by the OTG FS core and conveys USB
control & data signals through the full-speed subset of the UTMI+ Bus (UTMIFS). It provides
the physical support to USB connectivity.
The full-speed OTG PHY includes the following components:
• FS/LS transceiver module used by both host and device. It directly drives transmission
and reception on the single-ended USB lines.
• integrated ID pull-up resistor used to sample the ID line for A/B device identification.
• DP/DM integrated pull-up and pull-down resistors controlled by the OTG_FS core
depending on the current role of the device. As a peripheral, it enables the DP pull-up
resistor to signal full-speed peripheral connections as soon as VBUS is sensed to be at
a valid level (B-session valid). In host mode, pull-down resistors are enabled on both
DP/DM. Pull-up and pull-down resistors are dynamically switched when the device’s
role is changed via the host negotiation protocol (HNP).
• Pull-up/pull-down resistor ECN circuit. The DP pull-up consists of 2 resistors controlled
separately from the OTG_FS as per the resistor Engineering Change Notice applied to
USB Rev2.0. The dynamic trimming of the DP pull-up strength allows for better noise
rejection and Tx/Rx signal quality.
• VBUS sensing comparators with hysteresis used to detect VBUS Valid, A-B Session
Valid and session-end voltage thresholds. They are used to drive the session request
protocol (SRP), detect valid startup and end-of-session conditions, and constantly
monitor the VBUS supply during USB operations.
• VBUS pulsing method circuit used to charge/discharge VBUS through resistors during
the SRP (weak drive).
Caution: To guarantee a correct operation for the USB OTG FS peripheral, the AHB frequency should
be higher than 14.2 MHz.
1244/1751 RM0090 Rev 19

34.4 OTG dual role device (DRD)

Figure 387. OTG A-B device connection
VDD
5 V to VDD
voltage regulator
VDD
STM32 MCU EN
STMPS2141STR
GPIO
Current-limited 5 V Pwr
Overcurrent power distribution
GPIO+IRQ switch
(2)
USBmicro-AB connector
VBUS
PA9
DM
PA11
OSC_IN DP
PA12
ID
PA10
OSC_OUT
VSS
MS19904V4
1. External voltage regulator only needed when building a VBUS powered device
2. STMPS2141STR needed only if the application has to support a VBUS powered device. A basic power
switch can be used if 5 V are available on the application board.
34.4.1 ID line detection

The host or peripheral (the default) role is assumed depending on the ID input pin
(OTG_FS_ID). The ID line status is determined on plugging in the USB, depending on which
side of the USB cable is connected to the micro-AB receptacle.
• If the B-side of the USB cable is connected with a floating ID wire, the integrated pull-
up resistor detects a high ID level and the default Peripheral role is confirmed. In this
configuration the OTG_FS complies with the standard FSM described by section 6.8.2:
On-The-Go B-device of the On-The-Go Specification Rev1.3 supplement to the
USB2.0.
• If the A-side of the USB cable is connected with a grounded ID, the OTG_FS issues an
ID line status change interrupt (CIDSCHG bit in OTG_FS_GINTSTS) for host software
initialization, and automatically switches to the host role. In this configuration the
OTG_FS complies with the standard FSM described by section 6.8.1: On-The-Go A-
device of the On-The-Go Specification Rev1.3 supplement to the USB2.0.
34.4.2 HNP dual role device

The HNP capable bit in the Global USB configuration register (HNPCAP bit in OTG_FS_
GUSBCFG) enables the OTG_FS core to dynamically change its role from A-host to A-
peripheral and vice-versa, or from B-Peripheral to B-host and vice-versa according to the
host negotiation protocol (HNP). The current device status can be read by the combined
values of the Connector ID Status bit in the Global OTG control and status register (CIDSTS
bit in OTG_FS_GOTGCTL) and the current mode of operation bit in the global interrupt and
status register (CMOD bit in OTG_FS_GINTSTS).
The HNP program model is described in detail in Section 34.17: OTG_FS programming
model.
RM0090 Rev 19 1245/1751

1380
34.4.3 SRP dual role device

The SRP capable bit in the global USB configuration register (SRPCAP bit in
OTG_FS_GUSBCFG) enables the OTG_FS core to switch off the generation of VBUS for
the A-device to save power. Note that the A-device is always in charge of driving VBUS
regardless of the host or peripheral role of the OTG_FS.
the SRP A/B-device program model is described in detail in Section 34.17: OTG_FS
programming model.
34.5 USB peripheral

This section gives the functional description of the OTG_FS in the USB peripheral mode.
The OTG_FS works as an USB peripheral in the following circumstances:
• OTG B-Peripheral
– OTG B-device default state if B-side of USB cable is plugged in
• OTG A-Peripheral
– OTG A-device state after the HNP switches the OTG_FS to its peripheral role
• B-device
– If the ID line is present, functional and connected to the B-side of the USB cable,
and the HNP-capable bit in the Global USB Configuration register (HNPCAP bit in
OTG_FS_GUSBCFG) is cleared (see On-The-Go Rev1.3 par. 6.8.3).
• Peripheral only (see Figure 388: USB peripheral-only connection)
– The force device mode bit in the Global USB configuration register (FDMOD in
OTG_FS_GUSBCFG) is set to 1, forcing the OTG_FS core to work as a USB
peripheral-only (see On-The-Go Rev1.3 par. 6.8.3). In this case, the ID line is
ignored even if present on the USB connector.
Note: To build a bus-powered device implementation in case of the B-device or peripheral-only
configuration, an external regulator has to be added that generates the VDD chip-supply
from VBUS.
The VBUS pin can be freed by disabling the VBUS sensing option. This is done by setting the
NOVBUSSENS bit in the OTG_FS_GCCFG register. In this case the VBUS is considered
internally to be always at VBUS valid level (5 V).
1246/1751 RM0090 Rev 19

Figure 388. USB peripheral-only connection

VDD
5V to VDD
Volatge regulator
STM32 MCU
VBUS
USB Std-B connector

PA9
DM
PA11
OSC_IN DP
PA12
VSS
OSC_OUT
MS19905V4
1. Use a regulator to build a bus-powered device.
34.5.1 SRP-capable peripheral

The SRP capable bit in the Global USB configuration register (SRPCAP bit in
OTG_FS_GUSBCFG) enables the OTG_FS to support the session request protocol (SRP).
In this way, it allows the remote A-device to save power by switching off VBUS while the USB
session is suspended.
The SRP peripheral mode program model is described in detail in the B-device session
request protocol section.
34.5.2 Peripheral states

Powered state
The VBUS input detects the B-Session valid voltage by which the USB peripheral is allowed
to enter the powered state (see USB2.0 par9.1). The OTG_FS then automatically connects
the DP pull-up resistor to signal full-speed device connection to the host and generates the
session request interrupt (SRQINT bit in OTG_FS_GINTSTS) to notify the powered state.
The VBUS input also ensures that valid VBUS levels are supplied by the host during USB
operations. If a drop in VBUS below B-session valid happens to be detected (for instance
because of a power disturbance or if the host port has been switched off), the OTG_FS
automatically disconnects and the session end detected (SEDET bit in
OTG_FS_GOTGINT) interrupt is generated to notify that the OTG_FS has exited the
powered state.
In the powered state, the OTG_FS expects to receive some reset signaling from the host.
No other USB operation is possible. When a reset signaling is received the reset detected
interrupt (USBRST in OTG_FS_GINTSTS) is generated. When the reset signaling is
complete, the enumeration done interrupt (ENUMDNE bit in OTG_FS_GINTSTS) is
generated and the OTG_FS enters the Default state.
RM0090 Rev 19 1247/1751

1380
Soft disconnect
The powered state can be exited by software with the soft disconnect feature. The DP pull-
up resistor is removed by setting the soft disconnect bit in the device control register (SDIS
bit in OTG_FS_DCTL), causing a device disconnect detection interrupt on the host side
even though the USB cable was not really removed from the host port.
Default state
In the Default state the OTG_FS expects to receive a SET_ADDRESS command from the
host. No other USB operation is possible. When a valid SET_ADDRESS command is
decoded on the USB, the application writes the corresponding number into the device
address field in the device configuration register (DAD bit in OTG_FS_DCFG). The
OTG_FS then enters the address state and is ready to answer host transactions at the
configured USB address.
Suspended state
The OTG_FS peripheral constantly monitors the USB activity. After counting 3 ms of USB
idleness, the early suspend interrupt (ESUSP bit in OTG_FS_GINTSTS) is issued, and
confirmed 3 ms later, if appropriate, by the suspend interrupt (USBSUSP bit in
OTG_FS_GINTSTS). The device suspend bit is then automatically set in the device status
register (SUSPSTS bit in OTG_FS_DSTS) and the OTG_FS enters the suspended state.
The suspended state may optionally be exited by the device itself. In this case the
application sets the remote wakeup signaling bit in the device control register (RWUSIG bit
in OTG_FS_DCTL) and clears it after 1 to 15 ms.
When a resume signaling is detected from the host, the resume interrupt (WKUPINT bit in
OTG_FS_GINTSTS) is generated and the device suspend bit is automatically cleared.
34.5.3 Peripheral endpoints

The OTG_FS core instantiates the following USB endpoints:
• Control endpoint 0:
– Bidirectional and handles control messages only
– Separate set of registers to handle in and out transactions
– Proper control (OTG_FS_DIEPCTL0/OTG_FS_DOEPCTL0), transfer
configuration (OTG_FS_DIEPTSIZ0/OTG_FS_DIEPTSIZ0), and status-interrupt
(OTG_FS_DIEPINTx/)OTG_FS_DOEPINT0) registers. The available set of bits
inside the control and transfer size registers slightly differs from that of other
endpoints
• 3 IN endpoints
– Each of them can be configured to support the isochronous, bulk or interrupt
transfer type
– Each of them has proper control (OTG_FS_DIEPCTLx), transfer configuration
(OTG_FS_DIEPTSIZx), and status-interrupt (OTG_FS_DIEPINTx) registers
– The Device IN endpoints common interrupt mask register (OTG_FS_DIEPMSK) is
available to enable/disable a single kind of endpoint interrupt source on all of the
IN endpoints (EP0 included)
– Support for incomplete isochronous IN transfer interrupt (IISOIXFR bit in
OTG_FS_GINTSTS), asserted when there is at least one isochronous IN endpoint
1248/1751 RM0090 Rev 19

on which the transfer is not completed in the current frame. This interrupt is
asserted along with the end of periodic frame interrupt
(OTG_FS_GINTSTS/EOPF).
• 3 OUT endpoints
– Each of them can be configured to support the isochronous, bulk or interrupt
transfer type
– Each of them has a proper control (OTG_FS_DOEPCTLx), transfer configuration
(OTG_FS_DOEPTSIZx) and status-interrupt (OTG_FS_DOEPINTx) register
– Device Out endpoints common interrupt mask register (OTG_FS_DOEPMSK) is
available to enable/disable a single kind of endpoint interrupt source on all of the
OUT endpoints (EP0 included)
– Support for incomplete isochronous OUT transfer interrupt (INCOMPISOOUT bit
in OTG_FS_GINTSTS), asserted when there is at least one isochronous OUT
endpoint on which the transfer is not completed in the current frame. This interrupt
is asserted along with the end of periodic frame interrupt
(OTG_FS_GINTSTS/EOPF).
Endpoint control
• The following endpoint controls are available to the application through the device
endpoint-x IN/OUT control register (DIEPCTLx/DOEPCTLx):
– Endpoint enable/disable
– Endpoint activate in current configuration
– Program USB transfer type (isochronous, bulk, interrupt)
– Program supported packet size
– Program Tx-FIFO number associated with the IN endpoint
– Program the expected or transmitted data0/data1 PID (bulk/interrupt only)
– Program the even/odd frame during which the transaction is received or
transmitted (isochronous only)
– Optionally program the NAK bit to always negative-acknowledge the host
regardless of the FIFO status
– Optionally program the STALL bit to always stall host tokens to that endpoint
– Optionally program the SNOOP mode for OUT endpoint not to check the CRC
field of received data
Endpoint transfer
The device endpoint-x transfer size registers (DIEPTSIZx/DOEPTSIZx) allow the application
to program the transfer size parameters and read the transfer status. Programming must be
done before setting the endpoint enable bit in the endpoint control register. Once the
endpoint is enabled, these fields are read-only as the OTG FS core updates them with the
current transfer status.
The following transfer parameters can be programmed:
• Transfer size in bytes
• Number of packets that constitute the overall transfer size
RM0090 Rev 19 1249/1751

1380
Endpoint status/interrupt
The device endpoint-x interrupt registers (DIEPINTx/DOPEPINTx) indicate the status of an
endpoint with respect to USB- and AHB-related events. The application must read these
registers when the OUT endpoint interrupt bit or the IN endpoint interrupt bit in the core
interrupt register (OEPINT bit in OTG_FS_GINTSTS or IEPINT bit in OTG_FS_GINTSTS,
respectively) is set. Before the application can read these registers, it must first read the
device all endpoints interrupt (OTG_FS_DAINT) register to get the exact endpoint number
for the device endpoint-x interrupt register. The application must clear the appropriate bit in
this register to clear the corresponding bits in the DAINT and GINTSTS registers
The peripheral core provides the following status checks and interrupt generation:
• Transfer completed interrupt, indicating that data transfer was completed on both the
application (AHB) and USB sides
• Setup stage has been done (control-out only)
• Associated transmit FIFO is half or completely empty (in endpoints)
• NAK acknowledge has been transmitted to the host (isochronous-in only)
• IN token received when Tx-FIFO was empty (bulk-in/interrupt-in only)
• Out token received when endpoint was not yet enabled
• Babble error condition has been detected
• Endpoint disable by application is effective
• Endpoint NAK by application is effective (isochronous-in only)
• More than 3 back-to-back setup packets were received (control-out only)
• Timeout condition detected (control-in only)
• Isochronous out packet has been dropped, without generating an interrupt
34.6 USB host

This section gives the functional description of the OTG_FS in the USB host mode. The
OTG_FS works as a USB host in the following circumstances:
• OTG A-host
– OTG A-device default state when the A-side of the USB cable is plugged in
• OTG B-host
– OTG B-device after HNP switching to the host role
• A-device
– If the ID line is present, functional and connected to the A-side of the USB cable,
and the HNP-capable bit is cleared in the Global USB Configuration register
(HNPCAP bit in OTG_FS_GUSBCFG). Integrated pull-down resistors are
automatically set on the DP/DM lines.
• Host only (see Figure 389: USB host-only connection).
– The force host mode bit in the global USB configuration register (FHMOD bit in
OTG_FS_GUSBCFG) forces the OTG_FS core to work as a USB host-only. In this
case, the ID line is ignored even if present on the USB connector. Integrated pull-
down resistors are automatically set on the DP/DM lines.
Note: On-chip 5 V VBUS generation is not supported. For this reason, a charge pump or, if 5 V are
available on the application board, a basic power switch must be added externally to drive
1250/1751 RM0090 Rev 19

the 5 V VBUS line. The external charge pump can be driven by any GPIO output. This is
required for the OTG A-host, A-device and host-only configurations.
The VBUS input ensures that valid VBUS levels are supplied by the charge pump during USB
operations while the charge pump overcurrent output can be input to any GPIO pin
configured to generate port interrupts. The overcurrent ISR must promptly disable the VBUS
generation.
The VBUS pin can be freed by disabling the VBUS sensing option. This is done by setting the
NOVBUSSENS bit in the OTG_FS_GCCFG register. In this case the VBUS is considered
internally to be always at VBUS valid level (5 V).
Figure 389. USB host-only connection
VDD
5V
EN STMPS2141STR
GPIO
Current-limited 5 V Pwr
GPIO + IRQ Overcurrent power distribution

(2)
switch
USB Std-A connector

VBUS
DM
OSC_IN
DP
VSS
OSC_OUT
MSv36915V2
1. STMPS2141STR needed only if the application has to support a VBUS powered device. A basic power
switch can be used if 5 V are available on the application board.
2. VDD range is between 2 V and 3.6 V.
34.6.1 SRP-capable host

SRP support is available through the SRP capable bit in the global USB configuration
register (SRPCAP bit in OTG_FS_GUSBCFG). With the SRP feature enabled, the host can
save power by switching off the VBUS power while the USB session is suspended.
The SRP host mode program model is described in detail in the A-device session request
protocol section.
34.6.2 USB host states

Host port power
On-chip 5 V VBUS generation is not supported. For this reason, a charge pump or, if 5 V are
available on the application board, a basic power switch, must be added externally to drive
the 5 V VBUS line. The external charge pump can be driven by any GPIO output. When the
application decides to power on VBUS using the chosen GPIO, it must also set the port
power bit in the host port control and status register (PPWR bit in OTG_FS_HPRT).
RM0090 Rev 19 1251/1751

1380
VBUS valid
When HNP or SRP is enabled the VBUS sensing pin (PA9) pin should be connected to
VBUS. The VBUS input ensures that valid VBUS levels are supplied by the charge pump
during USB operations. Any unforeseen VBUS voltage drop below the VBUS valid threshold
(4.25 V) leads to an OTG interrupt triggered by the session end detected bit (SEDET bit in
OTG_FS_GOTGINT). The application is then required to remove the VBUS power and clear
the port power bit.
When HNP and SRP are both disabled, the VBUS sensing pin (PA9) should not be
connected to VBUS. This pin can be can be used as GPIO.
The charge pump overcurrent flag can also be used to prevent electrical damage. Connect
the overcurrent flag output from the charge pump to any GPIO input and configure it to
generate a port interrupt on the active level. The overcurrent ISR must promptly disable the
VBUS generation and clear the port power bit.
Host detection of a peripheral connection

If SRP or HNP are enabled, even if USB peripherals or B-devices can be attached at any
time, the OTG_FS will not detect any bus connection until VBUS is no longer sensed at a
valid level (5 V). When VBUS is at a valid level and a remote B-device is attached, the
OTG_FS core issues a host port interrupt triggered by the device connected bit in the host
port control and status register (PCDET bit in OTG_FS_HPRT).
When HNP and SRP are both disabled, USB peripherals or B-device are detected as soon
as they are connected. The OTG_FS core issues a host port interrupt triggered by the
device connected bit in the host port control and status (PCDET bit in OTG_FS_HPRT).
Host detection of peripheral a disconnection

The peripheral disconnection event triggers the disconnect detected interrupt (DISCINT bit
in OTG_FS_GINTSTS).
Host enumeration
After detecting a peripheral connection the host must start the enumeration process by
sending USB reset and configuration commands to the new peripheral.
Before starting to drive a USB reset, the application waits for the OTG interrupt triggered by
the debounce done bit (DBCDNE bit in OTG_FS_GOTGINT), which indicates that the bus is
stable again after the electrical debounce caused by the attachment of a pull-up resistor on
DP (FS) or DM (LS).
The application drives a USB reset signaling (single-ended zero) over the USB by keeping
the port reset bit set in the host port control and status register (PRST bit in
OTG_FS_HPRT) for a minimum of 10 ms and a maximum of 20 ms. The application takes
care of the timing count and then of clearing the port reset bit.
Once the USB reset sequence has completed, the host port interrupt is triggered by the port
enable/disable change bit (PENCHNG bit in OTG_FS_HPRT). This informs the application
that the speed of the enumerated peripheral can be read from the port speed field in the
host port control and status register (PSPD bit in OTG_FS_HPRT) and that the host is
starting to drive SOFs (FS) or Keep alives (LS). The host is now ready to complete the
peripheral enumeration by sending peripheral configuration commands.
1252/1751 RM0090 Rev 19

Host suspend
The application decides to suspend the USB activity by setting the port suspend bit in the
host port control and status register (PSUSP bit in OTG_FS_HPRT). The OTG_FS core
stops sending SOFs and enters the suspended state.
The suspended state can be optionally exited on the remote device’s initiative (remote
wakeup). In this case the remote wakeup interrupt (WKUPINT bit in OTG_FS_GINTSTS) is
generated upon detection of a remote wakeup signaling, the port resume bit in the host port
control and status register (PRES bit in OTG_FS_HPRT) self-sets, and resume signaling is
automatically driven over the USB. The application must time the resume window and then
clear the port resume bit to exit the suspended state and restart the SOF.
If the suspended state is exited on the host initiative, the application must set the port
resume bit to start resume signaling on the host port, time the resume window and finally
clear the port resume bit.
34.6.3 Host channels

The OTG_FS core instantiates 8 host channels. Each host channel supports an USB host
transfer (USB pipe). The host is not able to support more than 8 transfer requests at the
same time. If more than 8 transfer requests are pending from the application, the host
controller driver (HCD) must re-allocate channels when they become available from
previous duty, that is, after receiving the transfer completed and channel halted interrupts.
Each host channel can be configured to support in/out and any type of periodic/nonperiodic
transaction. Each host channel makes us of proper control (HCCHARx), transfer
configuration (HCTSIZx) and status/interrupt (HCINTx) registers with associated mask
(HCINTMSKx) registers.
Host channel control

• The following host channel controls are available to the application through the host
channel-x characteristics register (HCCHARx):
– Channel enable/disable
– Program the FS/LS speed of target USB peripheral
– Program the address of target USB peripheral
– Program the endpoint number of target USB peripheral
– Program the transfer IN/OUT direction
– Program the USB transfer type (control, bulk, interrupt, isochronous)
– Program the maximum packet size (MPS)
– Program the periodic transfer to be executed during odd/even frames
Host channel transfer

The host channel transfer size registers (HCTSIZx) allow the application to program the
transfer size parameters, and read the transfer status. Programming must be done before
setting the channel enable bit in the host channel characteristics register. Once the endpoint
RM0090 Rev 19 1253/1751

1380
is enabled the packet count field is read-only as the OTG FS core updates it according to
the current transfer status.
• The following transfer parameters can be programmed:
– transfer size in bytes
– number of packets making up the overall transfer size
– initial data PID
Host channel status/interrupt

The host channel-x interrupt register (HCINTx) indicates the status of an endpoint with
respect to USB- and AHB-related events. The application must read these register when the
host channels interrupt bit in the core interrupt register (HCINT bit in OTG_FS_GINTSTS) is
set. Before the application can read these registers, it must first read the host all channels
interrupt (HCAINT) register to get the exact channel number for the host channel-x interrupt
register. The application must clear the appropriate bit in this register to clear the
corresponding bits in the HAINT and GINTSTS registers. The mask bits for each interrupt
source of each channel are also available in the OTG_FS_HCINTMSK-x register.
• The host core provides the following status checks and interrupt generation:
– Transfer completed interrupt, indicating that the data transfer is complete on both
the application (AHB) and USB sides
– Channel has stopped due to transfer completed, USB transaction error or disable
command from the application
– Associated transmit FIFO is half or completely empty (IN endpoints)
– ACK response received
– NAK response received
– STALL response received
– USB transaction error due to CRC failure, timeout, bit stuff error, false EOP
– Babble error
– fraMe overrun
– dAta toggle error
34.6.4 Host scheduler

The host core features a built-in hardware scheduler which is able to autonomously re-order
and manage the USB transaction requests posted by the application. At the beginning of
each frame the host executes the periodic (isochronous and interrupt) transactions first,
followed by the nonperiodic (control and bulk) transactions to achieve the higher level of
priority granted to the isochronous and interrupt transfer types by the USB specification.
The host processes the USB transactions through request queues (one for periodic and one
for nonperiodic). Each request queue can hold up to 8 entries. Each entry represents a
pending transaction request from the application, and holds the IN or OUT channel number
along with other information to perform a transaction on the USB. The order in which the
requests are written to the queue determines the sequence of the transactions on the USB
interface.
At the beginning of each frame, the host processes the periodic request queue first, followed
by the nonperiodic request queue. The host issues an incomplete periodic transfer interrupt
(IPXFR bit in OTG_FS_GINTSTS) if an isochronous or interrupt transaction scheduled for
the current frame is still pending at the end of the current frame. The OTG HS core is fully
1254/1751 RM0090 Rev 19

responsible for the management of the periodic and nonperiodic request queues.The
periodic transmit FIFO and queue status register (HPTXSTS) and nonperiodic transmit
FIFO and queue status register (HNPTXSTS) are read-only registers which can be used by
the application to read the status of each request queue. They contain:
• The number of free entries currently available in the periodic (nonperiodic) request
queue (8 max)
• Free space currently available in the periodic (nonperiodic) Tx-FIFO (out-transactions)
• IN/OUT token, host channel number and other status information.
As request queues can hold a maximum of 8 entries each, the application can push to
schedule host transactions in advance with respect to the moment they physically reach the
SB for a maximum of 8 pending periodic transactions plus 8 pending nonperiodic
transactions.
To post a transaction request to the host scheduler (queue) the application must check that
there is at least 1 entry available in the periodic (nonperiodic) request queue by reading the
PTXQSAV bits in the OTG_FS_HNPTXSTS register or NPTQXSAV bits in the
OTG_FS_HNPTXSTS register.
34.7 SOF trigger

Figure 390. SOF connectivity
STM32 MCU
PA8
SOF pulse output, to
external audio control
PA9 VBUS
USB mic ro-AB connector

ITR1 SOF
PA11 D-
pulse
PA12 D+
TIM2 SOFgen PA10 ID
VSS
MS19907V3
The OTG FS core provides means to monitor, track and configure SOF framing in the host
and peripheral, as well as an SOF pulse output connectivity feature.
Such utilities are especially useful for adaptive audio clock generation techniques, where
the audio peripheral needs to synchronize to the isochronous stream provided by the PC, or
the host needs to trim its framing rate according to the requirements of the audio peripheral.
34.7.1 Host SOFs

In host mode the number of PHY clocks occurring between the generation of two
consecutive SOF (FS) or Keep-alive (LS) tokens is programmable in the host frame interval
register (HFIR), thus providing application control over the SOF framing period. An interrupt
RM0090 Rev 19 1255/1751

1380
is generated at any start of frame (SOF bit in OTH_FS_GINTSTS). The current frame
number and the time remaining until the next SOF are tracked in the host frame number
register (HFNUM).
An SOF pulse signal, generated at any SOF starting token and with a width of 20 HCLK
cycles, can be made available externally on the OTG_FS_SOF pin using the SOFOUTEN
bit in the global control and configuration register. The SOF pulse is also internally
connected to the input trigger of timer 2 (TIM2), so that the input capture feature, the output
compare feature and the timer can be triggered by the SOF pulse. The TIM2 connection is
enabled through the ITR1_RMP bits of TIM2_OR register.
34.7.2 Peripheral SOFs

In device mode, the start of frame interrupt is generated each time an SOF token is received
on the USB (SOF bit in OTH_FS_GINTSTS). The corresponding frame number can be read
from the device status register (FNSOF bit in OTG_FS_DSTS). An SOF pulse signal with a
width of 20 HCLK cycles is also generated and can be made available externally on the
OTG_FS_SOF pin by using the SOF output enable bit in the global control and configuration
register (SOFOUTEN bit in OTG_FS_GCCFG). The SOF pulse signal is also internally
connected to the TIM2 input trigger, so that the input capture feature, the output compare
feature and the timer can be triggered by the SOF pulse. The TIM2 connection is enabled
through the ITR1_RMP bits of the TIM2 option register (TIM2_OR).
The end of periodic frame interrupt (GINTSTS/EOPF) is used to notify the application when
80%, 85%, 90% or 95% of the time frame interval elapsed depending on the periodic frame
interval field in the device configuration register (PFIVL bit in OTG_FS_DCFG). This feature
can be used to determine if all of the isochronous traffic for that frame is complete.
34.8 OTG low-power modes

Table 197 below defines the STM32 low power modes and their compatibility with the OTG.
Table 197. Compatibility of STM32 low power modes with the OTG
Mode Description USB compatibility
Required when USB not in

Run MCU fully active
suspend state.
USB suspend exit causes the device to exit Sleep mode. Available while USB is in
Sleep
Peripheral registers content is kept. suspend state.
USB suspend exit causes the device to exit Stop mode. Available while USB is in
Stop
Peripheral registers content is kept(1). suspend state.
Powered-down. The peripheral must be reinitialized after Not compatible with USB
Standby
exiting Standby mode. applications.
1. Within Stop mode there are different possible settings. Some restrictions may also exist, please refer to
Section 5: Power controller (PWR) to understand which (if any) restrictions apply when using OTG.
1256/1751 RM0090 Rev 19

The power consumption of the OTG PHY is controlled by three bits in the general core
configuration register:
• PHY power down (GCCFG/PWRDWN)
It switches on/off the full-speed transceiver module of the PHY. It must be preliminarily
set to allow any USB operation.
• A-VBUS sensing enable (GCCFG/VBUSASEN)
It switches on/off the VBUS comparators associated with A-device operations. It must
be set when in A-device (USB host) mode and during HNP.
• B-VBUS sensing enable (GCCFG/VBUSASEN)
It switches on/off the VBUS comparators associated with B-device operations. It must
be set when in B-device (USB peripheral) mode and during HNP.
Power reduction techniques are available while in the USB suspended state, when the USB
session is not yet valid or the device is disconnected.
• Stop PHY clock (STPPCLK bit in OTG_FS_PCGCCTL)
When setting the stop PHY clock bit in the clock gating control register, most of the
48 MHz clock domain internal to the OTG full-speed core is switched off by clock
gating. The dynamic power consumption due to the USB clock switching activity is cut
even if the 48 MHz clock input is kept running by the application
Most of the transceiver is also disabled, and only the part in charge of detecting the
asynchronous resume or remote wakeup event is kept alive.
• Gate HCLK (GATEHCLK bit in OTG_FS_PCGCCTL)
When setting the Gate HCLK bit in the clock gating control register, most of the system
clock domain internal to the OTG_FS core is switched off by clock gating. Only the
register read and write interface is kept alive. The dynamic power consumption due to
the USB clock switching activity is cut even if the system clock is kept running by the
application for other purposes.
• USB system stop
When the OTG_FS is in the USB suspended state, the application may decide to
drastically reduce the overall power consumption by a complete shut down of all the
clock sources in the system. USB System Stop is activated by first setting the Stop
PHY clock bit and then configuring the system deep sleep mode in the power control
system module (PWR).
The OTG_FS core automatically reactivates both system and USB clocks by
asynchronous detection of remote wakeup (as an host) or resume (as a device)
signaling on the USB.
To save dynamic power, the USB data FIFO is clocked only when accessed by the OTG_FS
core.
34.9 Dynamic update of the OTG_FS_HFIR register

The USB core embeds a dynamic trimming capability of SOF framing period in host mode
allowing to synchronize an external device with the SOF frames.
When the OTG_FS_HFIR register is changed within a current SOF frame, the SOF period
correction is applied in the next frame as described in Figure 391.
RM0090 Rev 19 1257/1751

1380
Figure 391. Updating OTG_FS_HFIR dynamically
Old OTG_FS_HIFR value OTG_FS_HIFR value New OTG_FS_HIFR value

= 400 periods = 450 periods+HIFR write latency = 450 periods
SOF
reload
Latency
OTG_FS_HFIR
write
OTG_FS_HFIR 400 450
value
Frame
450
449
400
399
400
399
450
449
450
449
… … … …
1
0
1
0
1
0
1
0
timer
ai184
34.10 USB data FIFOs

The USB system features 1.25 Kbyte of dedicated RAM with a sophisticated FIFO control
mechanism. The packet FIFO controller module in the OTG_FS core organizes RAM space
into Tx-FIFOs into which the application pushes the data to be temporarily stored before the
USB transmission, and into a single Rx FIFO where the data received from the USB are
temporarily stored before retrieval (popped) by the application. The number of instructed
FIFOs and how these are organized inside the RAM depends on the device’s role. In
peripheral mode an additional Tx-FIFO is instructed for each active IN endpoint. Any FIFO
size is software configured to better meet the application requirements.
1258/1751 RM0090 Rev 19

34.11 Peripheral FIFO architecture

Figure 392. Device-mode FIFO address mapping and AHB FIFO access mapping
Single data
FIFO
IN endpoint Tx FIFO #n DIEPTXF2[31:16]

Dedicated Tx Tx FIFO #n
DFIFO push access FIFO #n control packet
from AHB DIEPTXFx[15:0]
(optional)
MAC pop
.. .. ..
. . .
DIEPTXF2[15:0]
IN endpoint Tx FIFO #1
Dedicated Tx
DFIFO push access Tx FIFO #1 packet
FIFO #1 control
from AHB DIEPTXF1[31:16]
(optional)
DIEPTXF1[15:0]
MAC pop
IN endpoint Tx FIFO #0 Dedicated Tx

Tx FIFO #0 packet GNPTXFSIZ[31:16]
DFIFO push access FIFO #0 control
from AHB (optional)
MAC pop
GNPTXFSIZ[15:0]
Any OUT endpoint DFIFO pop GRXFSIZ[31:16]

access from AHB Rx FIFO control Rx packets
MAC push (Rx start

A1 = 0 address
fixed to 0)
ai15611
34.11.1 Peripheral Rx FIFO

The OTG peripheral uses a single receive FIFO that receives the data directed to all OUT
endpoints. Received packets are stacked back-to-back until free space is available in the
Rx-FIFO. The status of the received packet (which contains the OUT endpoint destination
number, the byte count, the data PID and the validity of the received data) is also stored by
the core on top of the data payload. When no more space is available, host transactions are
NACKed and an interrupt is received on the addressed endpoint. The size of the receive
FIFO is configured in the receive FIFO Size register (GRXFSIZ).
The single receive FIFO architecture makes it more efficient for the USB peripheral to fill in
the receive RAM buffer:
• All OUT endpoints share the same RAM buffer (shared FIFO)
• The OTG FS core can fill in the receive FIFO up to the limit for any host sequence of
OUT tokens
The application keeps receiving the Rx-FIFO non-empty interrupt (RXFLVL bit in
OTG_FS_GINTSTS) as long as there is at least one packet available for download. It reads
the packet information from the receive status read and pop register (GRXSTSP) and finally
pops data off the receive FIFO by reading from the endpoint-related pop address.
RM0090 Rev 19 1259/1751

1380
34.11.2 Peripheral Tx FIFOs

The core has a dedicated FIFO for each IN endpoint. The application configures FIFO sizes
by writing the non periodic transmit FIFO size register (OTG_FS_TX0FSIZ) for IN endpoint0
and the device IN endpoint transmit FIFOx registers (DIEPTXFx) for IN endpoint-x.
34.12 Host FIFO architecture

Figure 393. Host-mode FIFO address mapping and AHB FIFO access mapping
Single data
FIFO
Any periodic channel Periodic Tx Periodic Tx packets HPTXFSIZ[31:16]

DFIFO push access FIFO control
from AHB (optional)
MAC pop HPTXFSIZ[15:0]
Any non-periodic Periodic Tx packets NPTXFSIZ[31:16]

channel DFIFO push Non-periodic
access from AHB Tx FIFO control
NPTXFSIZ[15:0]
MAC pop
Rx packets RXFSIZ[31:16]
Any channel DFIFO pop
access from AHB Rx FIFO control
Rx start address
fixed to 0
MAC push A1 = 0
ai15610
34.12.1 Host Rx FIFO

The host uses one receiver FIFO for all periodic and nonperiodic transactions. The FIFO is
used as a receive buffer to hold the received data (payload of the received packet) from the
USB until it is transferred to the system memory. Packets received from any remote IN
endpoint are stacked back-to-back until free space is available. The status of each received
packet with the host channel destination, byte count, data PID and validity of the received
data are also stored into the FIFO. The size of the receive FIFO is configured in the receive
FIFO size register (GRXFSIZ).
The single receive FIFO architecture makes it highly efficient for the USB host to fill in the
receive data buffer:
• All IN configured host channels share the same RAM buffer (shared FIFO)
• The OTG FS core can fill in the receive FIFO up to the limit for any sequence of IN
tokens driven by the host software
The application receives the Rx FIFO not-empty interrupt as long as there is at least one
packet available for download. It reads the packet information from the receive status read
and pop register and finally pops the data off the receive FIFO.
1260/1751 RM0090 Rev 19

34.12.2 Host Tx FIFOs

The host uses one transmit FIFO for all non-periodic (control and bulk) OUT transactions
and one transmit FIFO for all periodic (isochronous and interrupt) OUT transactions. FIFOs
are used as transmit buffers to hold the data (payload of the transmit packet) to be
transmitted over the USB. The size of the periodic (nonperiodic) Tx FIFO is configured in the
host periodic (nonperiodic) transmit FIFO size (HPTXFSIZ/HNPTXFSIZ) register.
The two Tx FIFO implementation derives from the higher priority granted to the periodic type
of traffic over the USB frame. At the beginning of each frame, the built-in host scheduler
processes the periodic request queue first, followed by the nonperiodic request queue.
The two transmit FIFO architecture provides the USB host with separate optimization for
periodic and nonperiodic transmit data buffer management:
• All host channels configured to support periodic (nonperiodic) transactions in the OUT
direction share the same RAM buffer (shared FIFOs)
• The OTG FS core can fill in the periodic (nonperiodic) transmit FIFO up to the limit for
any sequence of OUT tokens driven by the host software
The OTG_FS core issues the periodic Tx FIFO empty interrupt (PTXFE bit in
OTG_FS_GINTSTS) as long as the periodic Tx-FIFO is half or completely empty,
depending on the value of the periodic Tx-FIFO empty level bit in the AHB configuration
register (PTXFELVL bit in OTG_FS_GAHBCFG). The application can push the transmission
data in advance as long as free space is available in both the periodic Tx FIFO and the
periodic request queue. The host periodic transmit FIFO and queue status register
(HPTXSTS) can be read to know how much space is available in both.
OTG_FS core issues the non periodic Tx FIFO empty interrupt (NPTXFE bit in
OTG_FS_GINTSTS) as long as the nonperiodic Tx FIFO is half or completely empty
depending on the non periodic Tx FIFO empty level bit in the AHB configuration register
(TXFELVL bit in OTG_FS_GAHBCFG). The application can push the transmission data as
long as free space is available in both the nonperiodic Tx FIFO and nonperiodic request
queue. The host nonperiodic transmit FIFO and queue status register (HNPTXSTS) can be
read to know how much space is available in both.
34.13 FIFO RAM allocation
34.13.1 Device mode

Receive FIFO RAM allocation: the application should allocate RAM for SETUP Packets:
10 locations must be reserved in the receive FIFO to receive SETUP packets on control
endpoint. The core does not use these locations, which are reserved for SETUP packets, to
write any other data. One location is to be allocated for Global OUT NAK. Status information
is written to the FIFO along with each received packet. Therefore, a minimum space of
(Largest Packet Size / 4) + 1 must be allocated to receive packets. If multiple isochronous
endpoints are enabled, then at least two (Largest Packet Size / 4) + 1 spaces must be
allocated to receive back-to-back packets. Typically, two (Largest Packet Size / 4) + 1
spaces are recommended so that when the previous packet is being transferred to the CPU,
the USB can receive the subsequent packet.
Along with the last packet for each endpoint, transfer complete status information is also
pushed to the FIFO. Typically, one location for each OUT endpoint is recommended.
RM0090 Rev 19 1261/1751

1380
Transmit FIFO RAM allocation: the minimum RAM space required for each IN Endpoint
Transmit FIFO is the maximum packet size for that particular IN endpoint.
Note: More space allocated in the transmit IN Endpoint FIFO results in better performance on the
USB.
34.13.2 Host mode

Receive FIFO RAM allocation
Status information is written to the FIFO along with each received packet. Therefore, a
minimum space of (Largest Packet Size / 4) + 1 must be allocated to receive packets. If
multiple isochronous channels are enabled, then at least two (Largest Packet Size / 4) + 1
spaces must be allocated to receive back-to-back packets. Typically, two (Largest Packet
Size / 4) + 1 spaces are recommended so that when the previous packet is being
transferred to the CPU, the USB can receive the subsequent packet.
Along with the last packet in the host channel, transfer complete status information is also
pushed to the FIFO. So one location must be allocated for this.
Transmit FIFO RAM allocation

The minimum amount of RAM required for the host Non-periodic Transmit FIFO is the
largest maximum packet size among all supported non-periodic OUT channels.
Typically, two Largest Packet Sizes worth of space is recommended, so that when the
current packet is under transfer to the USB, the CPU can get the next packet.
The minimum amount of RAM required for host periodic Transmit FIFO is the largest
maximum packet size out of all the supported periodic OUT channels. If there is at least one
Isochronous OUT endpoint, then the space must be at least two times the maximum packet
size of that channel.
Note: More space allocated in the Transmit Non-periodic FIFO results in better performance on
the USB.
34.14 USB system performance

Best USB and system performance is achieved owing to the large RAM buffers, the highly
configurable FIFO sizes, the quick 32-bit FIFO access through AHB push/pop registers and,
especially, the advanced FIFO control mechanism. Indeed, this mechanism allows the
OTG_FS to fill in the available RAM space at best regardless of the current USB sequence.
With these features:
• The application gains good margins to calibrate its intervention in order to optimize the
CPU bandwidth usage:
– It can accumulate large amounts of transmission data in advance compared to
when they are effectively sent over the USB
– It benefits of a large time margin to download data from the single receive FIFO
• The USB Core is able to maintain its full operating rate, that is to provide maximum full-
speed bandwidth with a great margin of autonomy versus application intervention:
– It has a large reserve of transmission data at its disposal to autonomously manage
the sending of data over the USB
1262/1751 RM0090 Rev 19

– It has a lot of empty space available in the receive buffer to autonomously fill it in
with the data coming from the USB
As the OTG_FS core is able to fill in the 1.25 Kbyte RAM buffer very efficiently, and as
1.25 Kbyte of transmit/receive data is more than enough to cover a full speed frame, the
USB system is able to withstand the maximum full-speed data rate for up to one USB frame
(1 ms) without any CPU intervention.
34.15 OTG_FS interrupts

When the OTG_FS controller is operating in one mode, either device or host, the application
must not access registers from the other mode. If an illegal access occurs, a mode
mismatch interrupt is generated and reflected in the Core interrupt register (MMIS bit in the
OTG_FS_GINTSTS register). When the core switches from one mode to the other, the
registers in the new mode of operation must be reprogrammed as they would be after a
power-on reset.
Figure 394 shows the interrupt hierarchy.
RM0090 Rev 19 1263/1751

1380
Figure 394. Interrupt hierarchy

Wakeup interrupt
(1) OTG_FS_WKUP
Global interrupt
OTG_FS
AND
OR
Global interrupt mask (bit 0)

T
IEP P
NT
IN
OTG_AHBCFG
T
PIN
T
RT
IN
IN
GI
AND AHB configuration register
HC
OE
HP
OT
OTG_GINTSTS
Core register interrupt
31:26 25 24 23:20 19 18 17:3 2 1:0
OTG_GINTMSK
Core interrupt mask register
OTG_GOTGINT
OTG interrupt register
(15 + #EP):16 (#EP-1):0

OUT endpoints IN endpoints
OTG_DAINTMSK
Device all endpoints interrupt mask
register
OTG_DAINT
Device all endpoints interrupt register
OTG_DIEPMSK/
OTG_DOEPMSK
Device IN/OUT endpoints common
interrupt mask register
x=0
OTG_DIEPINTx/
...
OTG_DOEPINTx x = #HC-1
Device IN/OUT endpoint interrupt
registers
OTG_HPRT
Host port control and status register
OTG_HAINTMSK
Host all channels interrupt mask register
OTG_HAINT
Host all channels interrupt register
x=0
OTG_HCTINTMSKx
...
Host channels interrupt mask registers x = #HC-1

OTG_HCTINTx
Host channels interrupt registers
MSv36921V3
1. OTG_FS_WKUP become active (high state) when resume condition occurs during L1 SLEEP or L2 SUSPEND states.
1264/1751 RM0090 Rev 19

34.16 OTG_FS control and status registers

By reading from and writing to the control and status registers (CSRs) through the AHB
slave interface, the application controls the OTG_FS controller. These registers are 32 bits
wide, and the addresses are 32-bit block aligned. The OTG_FS registers must be accessed
by words (32 bits).
CSRs are classified as follows:
• Core global registers
• Host-mode registers
• Host global registers
• Host port CSRs
• Host channel-specific registers
• Device-mode registers
• Device global registers
• Device endpoint-specific registers
• Power and clock-gating registers
• Data FIFO (DFIFO) access registers
Only the Core global, Power and clock-gating, Data FIFO access, and host port control and
status registers can be accessed in both host and device modes. When the OTG_FS
controller is operating in one mode, either device or host, the application must not access
registers from the other mode. If an illegal access occurs, a mode mismatch interrupt is
generated and reflected in the Core interrupt register (MMIS bit in the OTG_FS_GINTSTS
register). When the core switches from one mode to the other, the registers in the new mode
of operation must be reprogrammed as they would be after a power-on reset.
RM0090 Rev 19 1265/1751

1380
34.16.1 CSR memory map

The host and device mode registers occupy different addresses. All registers are
implemented in the AHB clock domain.
Figure 395. CSR memory map

0000h
Core global CSRs (1 Kbyte)
0400h
Host mode CSRs (1 Kbyte)
0800h
Device mode CSRs (1.5 Kbyte)
0E00h
Power and clock gating CSRs (0.5 Kbyte)

1000h
Device EP 0/Host channel 0 FIFO (4 Kbyte)
2000h
Device EP1/Host channel 1 FIFO (4 Kbyte)

3000h DFIFO
push/pop
to this region
Device EP (x – 1)(1)/Host channel (x – 1)(1) FIFO (4 Kbyte)
Device EP x(1)/Host channel x(1) FIFO (4 Kbyte)
Reserved
2 0000h
DFIFO
debug read/
Direct access to data FIFO RAM
write to this
for debugging (128 Kbyte)
region
3 FFFFh
ai15615b
1. x = 3 in device mode and x = 7 in host mode.
Global CSR map

These registers are available in both host and device modes.
Table 198. Core global control and status registers (CSRs)

Address
Acronym Register name
offset
OTG_FS_GOTGCTL 0x000 OTG_FS control and status register (OTG_FS_GOTGCTL) on page 1271
OTG_FS_GOTGINT 0x004 OTG_FS interrupt register (OTG_FS_GOTGINT) on page 1272
OTG_FS_GAHBCFG 0x008 OTG_FS AHB configuration register (OTG_FS_GAHBCFG) on page 1274
OTG_FS_GUSBCFG 0x00C OTG_FS USB configuration register (OTG_FS_GUSBCFG) on page 1275
OTG_FS_GRSTCTL 0x010 OTG_FS reset register (OTG_FS_GRSTCTL) on page 1277
1266/1751 RM0090 Rev 19

Table 198. Core global control and status registers (CSRs) (continued)
Address
offset
OTG_FS_GINTSTS 0x014 OTG_FS core interrupt register (OTG_FS_GINTSTS) on page 1279
OTG_FS_GINTMSK 0x018 OTG_FS interrupt mask register (OTG_FS_GINTMSK) on page 1283
OTG_FS_GRXSTSR 0x01C OTG_FS Receive status debug read/OTG status read and pop registers
OTG_FS_GRXSTSP 0x020 (OTG_FS_GRXSTSR/OTG_FS_GRXSTSP) on page 1286
OTG_FS_GRXFSIZ 0x024 OTG_FS Receive FIFO size register (OTG_FS_GRXFSIZ) on page 1287
OTG_FS Host non-periodic transmit FIFO size register

OTG_FS_HNPTXFSIZ/
0x028 (OTG_FS_HNPTXFSIZ)/Endpoint 0 Transmit FIFO size
OTG_FS_DIEPTXF0(1)
(OTG_FS_DIEPTXF0)
OTG_FS non-periodic transmit FIFO/queue status register

OTG_FS_HNPTXSTS 0x02C
(OTG_FS_HNPTXSTS) on page 1288
OTG_FS general core configuration register (OTG_FS_GCCFG) on

OTG_FS_GCCFG 0x038
page 1289
OTG_FS_CID 0x03C OTG_FS core ID register (OTG_FS_CID) on page 1290
OTG_FS Host periodic transmit FIFO size register (OTG_FS_HPTXFSIZ) on

OTG_FS_HPTXFSIZ 0x100
page 1291
0x104
OTG_FS device IN endpoint transmit FIFO size register
OTG_FS_DIEPTXFx 0x108
(OTG_FS_DIEPTXFx) (x = 1..3, where x is the FIFO_number) on page 1291
0x10C
1. The general rule is to use OTG_FS_HNPTXFSIZ for host mode and OTG_FS_DIEPTXF0 for device mode.
Host-mode CSR map

These registers must be programmed every time the core changes to host mode.
Table 199. Host-mode control and status registers (CSRs)

Offset
address
OTG_FS_HCFG 0x400 OTG_FS Host configuration register (OTG_FS_HCFG) on page 1292
OTG_FS_HFIR 0x404 OTG_FS Host frame interval register (OTG_FS_HFIR) on page 1292
OTG_FS Host frame number/frame time remaining register

OTG_FS_HFNUM 0x408
(OTG_FS_HFNUM) on page 1293
OTG_FS_Host periodic transmit FIFO/queue status register

OTG_FS_HPTXSTS 0x410
(OTG_FS_HPTXSTS) on page 1293
OTG_FS Host all channels interrupt register (OTG_FS_HAINT) on

OTG_FS_HAINT 0x414
page 1294
OTG_FS Host all channels interrupt mask register (OTG_FS_HAINTMSK)
OTG_FS_HAINTMSK 0x418
on page 1295
RM0090 Rev 19 1267/1751

1380
Table 199. Host-mode control and status registers (CSRs) (continued)

Offset
address
OTG_FS Host port control and status register (OTG_FS_HPRT) on

OTG_FS_HPRT 0x440
page 1295
0x500
0x520 OTG_FS Host channel-x characteristics register (OTG_FS_HCCHARx)
OTG_FS_HCCHARx
... (x = 0..7, where x = Channel_number) on page 1298
0x5E0
OTG_FS Host channel-x interrupt register (OTG_FS_HCINTx) (x = 0..7,
OTG_FS_HCINTx 0x508
where x = Channel_number) on page 1299
OTG_FS Host channel-x interrupt mask register (OTG_FS_HCINTMSKx)
OTG_FS_HCINTMSKx 0x50C
(x = 0..7, where x = Channel_number) on page 1300
OTG_FS Host channel-x transfer size register (OTG_FS_HCTSIZx)
OTG_FS_HCTSIZx 0x510
(x = 0..7, where x = Channel_number) on page 1301
Device-mode CSR map

These registers must be programmed every time the core changes to device mode.
Table 200. Device-mode control and status registers

Offset
address
OTG_FS_DCFG 0x800 OTG_FS device configuration register (OTG_FS_DCFG) on page 1302
OTG_FS_DCTL 0x804 OTG_FS device control register (OTG_FS_DCTL) on page 1303
OTG_FS_DSTS 0x808 OTG_FS device status register (OTG_FS_DSTS) on page 1304
OTG_FS device IN endpoint common interrupt mask register

OTG_FS_DIEPMSK 0x810
(OTG_FS_DIEPMSK) on page 1305
OTG_FS device OUT endpoint common interrupt mask register

OTG_FS_DOEPMSK 0x814
(OTG_FS_DOEPMSK) on page 1306
OTG_FS device all endpoints interrupt register (OTG_FS_DAINT) on

OTG_FS_DAINT 0x818
page 1307
OTG_FS all endpoints interrupt mask register (OTG_FS_DAINTMSK) on

OTG_FS_DAINTMSK 0x81C
page 1308
OTG_FS device VBUS discharge time register (OTG_FS_DVBUSDIS) on

OTG_FS_DVBUSDIS 0x828
page 1308
OTG_FS device VBUS pulsing time register (OTG_FS_DVBUSPULSE)
OTG_FS_DVBUSPULSE 0x82C
on page 1308
OTG_FS device IN endpoint FIFO empty interrupt mask register:

OTG_FS_DIEPEMPMSK 0x834
(OTG_FS_DIEPEMPMSK) on page 1309
OTG_FS device control IN endpoint 0 control register

OTG_FS_DIEPCTL0 0x900
(OTG_FS_DIEPCTL0) on page 1309
1268/1751 RM0090 Rev 19

Table 200. Device-mode control and status registers (continued)

Offset
address
0x920
OTG device endpoint x control register (OTG_FS_DIEPCTLx) (x = 1..3,
OTG_FS_DIEPCTLx 0x940
where x = Endpoint_number) on page 1311
0x960
OTG_FS device endpoint-x interrupt register (OTG_FS_DIEPINTx)

OTG_FS_DIEPINTx 0x908
(x = 0..3, where x = Endpoint_number) on page 1318
OTG_FS device IN endpoint 0 transfer size register

OTG_FS_DIEPTSIZ0 0x910
(OTG_FS_DIEPTSIZ0) on page 1320
OTG_FS device IN endpoint transmit FIFO status register

OTG_FS_DTXFSTSx 0x918 (OTG_FS_DTXFSTSx) (x = 0..3, where x = Endpoint_number) on
page 1324
0x930 OTG_FS device OUT endpoint-x transfer size register
OTG_FS_DIEPTSIZx 0x950 (OTG_FS_DOEPTSIZx) (x = 1..3, where x = Endpoint_number) on
0x970 page 1324
OTG_FS device control OUT endpoint 0 control register

OTG_FS_DOEPCTL0 0xB00
(OTG_FS_DOEPCTL0) on page 1314
0xB20
OTG device endpoint x control register (OTG_FS_DIEPCTLx) (x = 1..3,
OTG_FS_DOEPCTLx 0xB40
where x = Endpoint_number) on page 1311
0xB60
OTG_FS device endpoint-x interrupt register (OTG_FS_DOEPINTx)

OTG_FS_DOEPINTx 0xB08
(x = 0..3, where x = Endpoint_number) on page 1319
OTG_FS device OUT endpoint 0 transfer size register

OTG_FS_DOEPTSIZ0 0xB10
(OTG_FS_DOEPTSIZ0) on page 1322
0xB30 OTG_FS device OUT endpoint-x transfer size register

OTG_FS_DOEPTSIZx 0xB50 (OTG_FS_DOEPTSIZx) (x = 1..3, where x = Endpoint_number) on
0xB70 page 1324
Data FIFO (DFIFO) access register map

These registers, available in both host and device modes, are used to read or write the FIFO
space for a specific endpoint or a channel, in a given direction. If a host channel is of type
IN, the FIFO can only be read on the channel. Similarly, if a host channel is of type OUT, the
FIFO can only be written on the channel.
Table 201. Data FIFO (DFIFO) access register map

FIFO access register section Address range Access
Device IN Endpoint 0/Host OUT Channel 0: DFIFO Write Access w

0x1000–0x1FFC
Device OUT Endpoint 0/Host IN Channel 0: DFIFO Read Access r
Device IN Endpoint 1/Host OUT Channel 1: DFIFO Write Access w

0x2000–0x2FFC
Device OUT Endpoint 1/Host IN Channel 1: DFIFO Read Access r
RM0090 Rev 19 1269/1751

1380
Table 201. Data FIFO (DFIFO) access register map (continued)

FIFO access register section Address range Access
... ... ...
Device IN Endpoint x(1)/Host OUT Channel x(1): DFIFO Write Access w

0xX000–0xXFFC
Device OUT Endpoint x(1)/Host IN Channel x(1): DFIFO Read Access r
1. Where x is 3 in device mode and 7 in host mode.
Power and clock gating CSR map

There is a single register for power and clock gating. It is available in both host and device
modes.
Table 202. Power and clock gating control and status registers
Register name Acronym Offset address: 0xE00–0xFFF
Power and clock gating control register OTG_FS_PCGCCTL 0xE00-0xE04
Reserved - 0xE05–0xFFF
1270/1751 RM0090 Rev 19

34.16.2 OTG_FS global registers

These registers are available in both host and device modes, and do not need to be
reprogrammed when switching between these modes.
Bit values in the register descriptions are expressed in binary unless otherwise specified.
OTG_FS control and status register (OTG_FS_GOTGCTL)

The OTG_FS_GOTGCTL register controls the behavior and reflects the status of the OTG
function of the core.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
HSHNPEN
HNGSCS
SRQSCS
DHNPEN
CIDSTS
HNPRQ
BSVLD
ASVLD
DBCT
SRQ
r r r r rw rw rw r rw r

Bit 19 BSVLD: B-session valid
Indicates the device mode transceiver status.
0: B-session is not valid.
1: B-session is valid.
In OTG mode, you can use this bit to determine if the device is connected or disconnected.
Note: Only accessible in device mode.
Bit 18 ASVLD: A-session valid
Indicates the host mode transceiver status.
0: A-session is not valid
1: A-session is valid
Note: Only accessible in host mode.
Bit 17 DBCT: Long/short debounce time
Indicates the debounce time of a detected connection.
0: Long debounce time, used for physical connections (100 ms + 2.5 µs)
1: Short debounce time, used for soft connections (2.5 µs)
Bit 16 CIDSTS: Connector ID status
Indicates the connector ID status on a connect event.
0: The OTG_FS controller is in A-device mode
1: The OTG_FS controller is in B-device mode
Note: Accessible in both device and host modes.
RM0090 Rev 19 1271/1751

1380
Bit 11 DHNPEN: Device HNP enabled

The application sets this bit when it successfully receives a SetFeature.SetHNPEnable
command from the connected USB host.
0: HNP is not enabled in the application
1: HNP is enabled in the application
Bit 10 HSHNPEN: host set HNP enable
The application sets this bit when it has successfully enabled HNP (using the
SetFeature.SetHNPEnable command) on the connected device.
0: Host Set HNP is not enabled
1: Host Set HNP is enabled
Bit 9 HNPRQ: HNP request
The application sets this bit to initiate an HNP request to the connected USB host. The
application can clear this bit by writing a 0 when the host negotiation success status change
bit in the OTG_FS_GOTGINT register (HNSSCHG bit in OTG_FS_GOTGINT) is set. The
core clears this bit when the HNSSCHG bit is cleared.
0: No HNP request
1: HNP request
Bit 8 HNGSCS: Host negotiation success
The core sets this bit when host negotiation is successful. The core clears this bit when the
HNP Request (HNPRQ) bit in this register is set.
0: Host negotiation failure
1: Host negotiation success
Bit 1 SRQ: Session request
The application sets this bit to initiate a session request on the USB. The application can
clear this bit by writing a 0 when the host negotiation success status change bit in the
OTG_FS_GOTGINT register (HNSSCHG bit in OTG_FS_GOTGINT) is set. The core clears
this bit when the HNSSCHG bit is cleared.
If you use the USB 1.1 full-speed serial transceiver interface to initiate the session request,
the application must wait until VBUS discharges to 0.2 V, after the B-Session Valid bit in this
register (BSVLD bit in OTG_FS_GOTGCTL) is cleared.
0: No session request
1: Session request
Bit 0 SRQSCS: Session request success
The core sets this bit when a session request initiation is successful.
0: Session request failure
1: Session request success
OTG_FS interrupt register (OTG_FS_GOTGINT)

1272/1751 RM0090 Rev 19

The application reads this register whenever there is an OTG interrupt and clears the bits in
this register to clear the OTG interrupt.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
HNSSCHG
ADTOCHG
SRSSCHG
HNGDET
DBCDNE
SEDET
Reserved Reserved Reserved Res.
rc_ rc_ rc_ rc_ rc_ rc_

w1 w1 w1 w1 w1 w1

Bit 19 DBCDNE: Debounce done
The core sets this bit when the debounce is completed after the device connect. The
application can start driving USB reset after seeing this interrupt. This bit is only valid when
the HNP Capable or SRP Capable bit is set in the OTG_FS_GUSBCFG register (HNPCAP
bit or SRPCAP bit in OTG_FS_GUSBCFG, respectively).
Bit 18 ADTOCHG: A-device timeout change
The core sets this bit to indicate that the A-device has timed out while waiting for the B-device
to connect.
Bit 17 HNGDET: Host negotiation detected
The core sets this bit when it detects a host negotiation request on the USB.
Bit 9 HNSSCHG: Host negotiation success status change
The core sets this bit on the success or failure of a USB host negotiation request. The
application must read the host negotiation success bit of the OTG_FS_GOTGCTL register
(HNGSCS in OTG_FS_GOTGCTL) to check for success or failure.
Bit 8 SRSSCHG: Session request success status change
The core sets this bit on the success or failure of a session request. The application must
read the session request success bit in the OTG_FS_GOTGCTL register (SRQSCS bit in
OTG_FS_GOTGCTL) to check for success or failure.
Bit 2 SEDET: Session end detected
The core sets this bit to indicate that the level of the voltage on VBUS is no longer valid for a
B-Peripheral session when VBUS < 0.8 V.
RM0090 Rev 19 1273/1751

1380
OTG_FS AHB configuration register (OTG_FS_GAHBCFG)

This register can be used to configure the core after power-on or a change in mode. This
register mainly contains AHB system-related configuration parameters. Do not change this
register after the initial programming. The application must program this register before
starting any transactions on either the AHB or the USB.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PTXFELVL
GINTMSK
TXFELVL
Reserved Reserved
rw rw rw

Bit 8 PTXFELVL: Periodic TxFIFO empty level
Indicates when the periodic TxFIFO empty interrupt bit in the OTG_FS_GINTSTS register
(PTXFE bit in OTG_FS_GINTSTS) is triggered.
0: PTXFE (in OTG_FS_GINTSTS) interrupt indicates that the Periodic TxFIFO is half empty
1: PTXFE (in OTG_FS_GINTSTS) interrupt indicates that the Periodic TxFIFO is completely
empty
Bit 7 TXFELVL: TxFIFO empty level
In device mode, this bit indicates when IN endpoint Transmit FIFO empty interrupt (TXFE in
OTG_FS_DIEPINTx.) is triggered.
0: the TXFE (in OTG_FS_DIEPINTx) interrupt indicates that the IN Endpoint TxFIFO is half
empty
1: the TXFE (in OTG_FS_DIEPINTx) interrupt indicates that the IN Endpoint TxFIFO is
completely empty
In host mode, this bit indicates when the nonperiodic Tx FIFO empty interrupt (NPTXFE bit in
OTG_FS_GINTSTS) is triggered:
0: the NPTXFE (in OTG_FS_GINTSTS) interrupt indicates that the nonperiodic Tx FIFO is
half empty
1: the NPTXFE (in OTG_FS_GINTSTS) interrupt indicates that the nonperiodic Tx FIFO is
completely empty
Bit 0 GINTMSK: Global interrupt mask
The application uses this bit to mask or unmask the interrupt line assertion to itself.
Irrespective of this bit’s setting, the interrupt status registers are updated by the core.
0: Mask the interrupt assertion to the application.
1: Unmask the interrupt assertion to the application.
1274/1751 RM0090 Rev 19

OTG_FS USB configuration register (OTG_FS_GUSBCFG)

This register can be used to configure the core after power-on or a changing to host mode
or device mode. It contains USB and USB-PHY related configuration parameters. The
application must program this register before starting any transactions on either the AHB or
the USB. Do not make changes to this register after the initial programming.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SRPCAP
CTXPKT
PHYSEL
FDMOD
FHMOD
HNPCA
TRDT TOCAL
P
Reserved Res. Reserved
rw rw rw rw rw rw r rw
Bit 31 CTXPKT: Corrupt Tx packet

This bit is for debug purposes only. Never set this bit to 1.
Bit 30 FDMOD: Force device mode
Writing a 1 to this bit forces the core to device mode irrespective of the OTG_FS_ID input
pin.
0: Normal mode
1: Force device mode
After setting the force bit, the application must wait at least 25 ms before the change takes
effect.
Bit 29 FHMOD: Force host mode
Writing a 1 to this bit forces the core to host mode irrespective of the OTG_FS_ID input pin.
0: Normal mode
1: Force host mode
After setting the force bit, the application must wait at least 25 ms before the change takes
effect.
Bits 13:10 TRDT: USB turnaround time
These bits allow setting the turnaround time in PHY clocks. They must be configured
according to Table 203: TRDT values, depending on the application AHB frequency. Higher
TRDT values allow stretching the USB response time to IN tokens in order to compensate for
longer AHB read access latency to the Data FIFO.
Bit 9 HNPCAP: HNP-capable
The application uses this bit to control the OTG_FS controller’s HNP capabilities.
0: HNP capability is not enabled.
1: HNP capability is enabled.
RM0090 Rev 19 1275/1751

1380
Bit 8 SRPCAP: SRP-capable

The application uses this bit to control the OTG_FS controller’s SRP capabilities. If the core
operates as a non-SRP-capable
B-device, it cannot request the connected A-device (host) to activate VBUS and start a
session.
0: SRP capability is not enabled.
1: SRP capability is enabled.
Bit 6 PHYSEL: Full speed serial transceiver select
This bit is always 1 with read-only access.
Bits 2:0 TOCAL: FS timeout calibration
The number of PHY clocks that the application programs in this field is added to the full-
speed interpacket timeout duration in the core to account for any additional delays
introduced by the PHY. This can be required, because the delay introduced by the PHY in
generating the line state condition can vary from one PHY to another.
The USB standard timeout value for full-speed operation is 16 to 18 (inclusive) bit times. The
application must program this field based on the speed of enumeration. The number of bit
times added per PHY clock is 0.25 bit times.
Table 203. TRDT values

AHB frequency range (MHz)
TRDT minimum value
Min. Max
14.2 15 0xF
15 16 0xE
16 17.2 0xD
17.2 18.5 0xC
18.5 20 0xB
20 21.8 0xA
21.8 24 0x9
24 27.5 0x8
27.5 32 0x7
32 - 0x6
1276/1751 RM0090 Rev 19

OTG_FS reset register (OTG_FS_GRSTCTL)

The application uses this register to reset various hardware features inside the core.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RXFFLSH
TXFFLSH
AHBIDL
HSRST
CSRST
FCRST
Reserved
TXFNUM
Reserved
r rw rs rs rs rs rs
Bit 31 AHBIDL: AHB master idle

Indicates that the AHB master state machine is in the Idle condition.
Bits 10:6 TXFNUM: TxFIFO number
This is the FIFO number that must be flushed using the TxFIFO Flush bit. This field must not
be changed until the core clears the TxFIFO Flush bit.
00000:
– Non-periodic TxFIFO flush in host mode
– Tx FIFO 0 flush in device mode
00001:
– Periodic TxFIFO flush in host mode
– TXFIFO 1 flush in device mode
00010: TXFIFO 2 flush in device mode
...
00101: TXFIFO 15 flush in device mode
10000: Flush all the transmit FIFOs in device or host mode.
Bit 5 TXFFLSH: TxFIFO flush
This bit selectively flushes a single or all transmit FIFOs, but cannot do so if the core is in the
midst of a transaction.
The application must write this bit only after checking that the core is neither writing to the
TxFIFO nor reading from the TxFIFO. Verify using these registers:
Read—NAK Effective Interrupt ensures the core is not reading from the FIFO
Write—AHBIDL bit in OTG_FS_GRSTCTL ensures the core is not writing anything to the
FIFO.
Bit 4 RXFFLSH: RxFIFO flush
The application can flush the entire RxFIFO using this bit, but must first ensure that the core
is not in the middle of a transaction.
The application must only write to this bit after checking that the core is neither reading from
the RxFIFO nor writing to the RxFIFO.
The application must wait until the bit is cleared before performing any other operations. This
bit requires 8 clocks (slowest of PHY or AHB clock) to clear.
RM0090 Rev 19 1277/1751

1380
Bit 2 FCRST: Host frame counter reset

The application writes this bit to reset the frame number counter inside the core. When the
frame counter is reset, the subsequent SOF sent out by the core has a frame number of 0.
Bit 1 HSRST: HCLK soft reset
The application uses this bit to flush the control logic in the AHB Clock domain. Only AHB
Clock Domain pipelines are reset.
FIFOs are not flushed with this bit.
All state machines in the AHB clock domain are reset to the Idle state after terminating the
transactions on the AHB, following the protocol.
CSR control bits used by the AHB clock domain state machines are cleared.
To clear this interrupt, status mask bits that control the interrupt status and are generated by
the AHB clock domain state machine are cleared.
Because interrupt status bits are not cleared, the application can get the status of any core
events that occurred after it set this bit.
This is a self-clearing bit that the core clears after all necessary logic is reset in the core. This
can take several clocks, depending on the core’s current state.
Bit 0 CSRST: Core soft reset
Resets the HCLK and PCLK domains as follows:
Clears the interrupts and all the CSR register bits except for the following bits:
– RSTPDMODL bit in OTG_FS_PCGCCTL
– GAYEHCLK bit in OTG_FS_PCGCCTL
– PWRCLMP bit in OTG_FS_PCGCCTL
– STPPCLK bit in OTG_FS_PCGCCTL
– FSLSPCS bit in OTG_FS_HCFG
– DSPD bit in OTG_FS_DCFG
All module state machines (except for the AHB slave unit) are reset to the Idle state, and all
the transmit FIFOs and the receive FIFO are flushed.
Any transactions on the AHB Master are terminated as soon as possible, after completing the
last data phase of an AHB transfer. Any transactions on the USB are terminated immediately.
The application can write to this bit any time it wants to reset the core. This is a self-clearing
bit and the core clears this bit after all the necessary logic is reset in the core, which can take
several clocks, depending on the current state of the core. Once this bit has been cleared,
the software must wait at least 3 PHY clocks before accessing the PHY domain
(synchronization delay). The software must also check that bit 31 in this register is set to 1
(AHB Master is Idle) before starting any operation.
Typically, the software reset is used during software development and also when you
dynamically change the PHY selection bits in the above listed USB configuration registers.
When you change the PHY, the corresponding clock for the PHY is selected and used in the
PHY domain. Once a new clock is selected, the PHY domain has to be reset for proper
operation.
1278/1751 RM0090 Rev 19

OTG_FS core interrupt register (OTG_FS_GINTSTS)

This register interrupts the application for system-level events in the current mode (device
mode or host mode).
Some of the bits in this register are valid only in host mode, while others are valid in device
mode only. This register also indicates the current mode. To clear the interrupt status bits of
the rc_w1 type, the application must write 1 into the bit.
The FIFO status interrupts are read-only; once software reads from or writes to the FIFO
while servicing these interrupts, FIFO interrupt conditions are cleared automatically.
The application must clear the OTG_FS_GINTSTS register at initialization before
unmasking the interrupt bit to avoid any interrupts generated prior to initialization.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
IPXFR/INCOMPISOOUT
GONAKEFF
ENUMDNE
GINAKEFF
USBSUSP
ISOODRP
CIDSCHG
HPRTINT
IISOIXFR
Reserved
USBRST
DISCINT
NPTXFE
WKUINT
SRQINT
OTGINT
OEPINT
RXFLVL
ESUSP
PTXFE
IEPINT
HCINT
CMOD
EOPF
MMIS
SOF
Reserved
Reserved
Reserved
rc_w1
rc_w1
rc_w1 r r r Res. rc_w1 r r rc_w1 r r r r r r
Bit 31 WKUPINT: Resume/remote wakeup detected interrupt

In device mode, this interrupt is asserted when a resume is detected on the USB. In host
mode, this interrupt is asserted when a remote wakeup is detected on the USB.
Bit 30 SRQINT: Session request/new session detected interrupt
In host mode, this interrupt is asserted when a session request is detected from the device.
In device mode, this interrupt is asserted when VBUS is in the valid range for a B-peripheral
device. Accessible in both device and host modes.
Bit 29 DISCINT: Disconnect detected interrupt
Asserted when a device disconnect is detected.
Bit 28 CIDSCHG: Connector ID status change
The core sets this bit when there is a change in connector ID status.
Bit 26 PTXFE: Periodic TxFIFO empty
Asserted when the periodic transmit FIFO is either half or completely empty and there is
space for at least one entry to be written in the periodic request queue. The half or
completely empty status is determined by the periodic TxFIFO empty level bit in the
OTG_FS_GAHBCFG register (PTXFELVL bit in OTG_FS_GAHBCFG).
RM0090 Rev 19 1279/1751

1380
Bit 25 HCINT: Host channels interrupt

The core sets this bit to indicate that an interrupt is pending on one of the channels of the
core (in host mode). The application must read the OTG_FS_HAINT register to determine
the exact number of the channel on which the interrupt occurred, and then read the
corresponding OTG_FS_HCINTx register to determine the exact cause of the interrupt. The
application must clear the appropriate status bit in the OTG_FS_HCINTx register to clear
this bit.
Bit 24 HPRTINT: Host port interrupt
The core sets this bit to indicate a change in port status of one of the OTG_FS controller
ports in host mode. The application must read the OTG_FS_HPRT register to determine the
exact event that caused this interrupt. The application must clear the appropriate status bit in
the OTG_FS_HPRT register to clear this bit.
Bit 21 IPXFR: Incomplete periodic transfer
In host mode, the core sets this interrupt bit when there are incomplete periodic transactions
still pending, which are scheduled for the current frame.
INCOMPISOOUT: Incomplete isochronous OUT transfer
In device mode, the core sets this interrupt to indicate that there is at least one isochronous
OUT endpoint on which the transfer is not completed in the current frame. This interrupt is
asserted along with the End of periodic frame interrupt (EOPF) bit in this register.
Bit 20 IISOIXFR: Incomplete isochronous IN transfer
The core sets this interrupt to indicate that there is at least one isochronous IN endpoint on
which the transfer is not completed in the current frame. This interrupt is asserted along with
the End of periodic frame interrupt (EOPF) bit in this register.
Bit 19 OEPINT: OUT endpoint interrupt
The core sets this bit to indicate that an interrupt is pending on one of the OUT endpoints of
the core (in device mode). The application must read the OTG_FS_DAINT register to
determine the exact number of the OUT endpoint on which the interrupt occurred, and then
read the corresponding OTG_FS_DOEPINTx register to determine the exact cause of the
interrupt. The application must clear the appropriate status bit in the corresponding
OTG_FS_DOEPINTx register to clear this bit.
Bit 18 IEPINT: IN endpoint interrupt
The core sets this bit to indicate that an interrupt is pending on one of the IN endpoints of the
core (in device mode). The application must read the OTG_FS_DAINT register to determine
the exact number of the IN endpoint on which the interrupt occurred, and then read the
corresponding OTG_FS_DIEPINTx register to determine the exact cause of the interrupt.
The application must clear the appropriate status bit in the corresponding
OTG_FS_DIEPINTx register to clear this bit.
Bit 15 EOPF: End of periodic frame interrupt
Indicates that the period specified in the periodic frame interval field of the OTG_FS_DCFG
register (PFIVL bit in OTG_FS_DCFG) has been reached in the current frame.
1280/1751 RM0090 Rev 19

Bit 14 ISOODRP: Isochronous OUT packet dropped interrupt

The core sets this bit when it fails to write an isochronous OUT packet into the RxFIFO
because the RxFIFO does not have enough space to accommodate a maximum size packet
for the isochronous OUT endpoint.
Bit 13 ENUMDNE: Enumeration done
The core sets this bit to indicate that speed enumeration is complete. The application must
read the OTG_FS_DSTS register to obtain the enumerated speed.
Bit 12 USBRST: USB reset
The core sets this bit to indicate that a reset is detected on the USB.
Bit 11 USBSUSP: USB suspend
The core sets this bit to indicate that a suspend was detected on the USB. The core enters
the Suspended state when there is no activity on the data lines for a period of 3 ms.
Bit 10 ESUSP: Early suspend
The core sets this bit to indicate that an Idle state has been detected on the USB for 3 ms.
Bit 7 GONAKEFF: Global OUT NAK effective
Indicates that the Set global OUT NAK bit in the OTG_FS_DCTL register (SGONAK bit in
OTG_FS_DCTL), set by the application, has taken effect in the core. This bit can be cleared
by writing the Clear global OUT NAK bit in the OTG_FS_DCTL register (CGONAK bit in
OTG_FS_DCTL).
Bit 6 GINAKEFF: Global IN non-periodic NAK effective
Indicates that the Set global non-periodic IN NAK bit in the OTG_FS_DCTL register
(SGINAK bit in OTG_FS_DCTL), set by the application, has taken effect in the core. That is,
the core has sampled the Global IN NAK bit set by the application. This bit can be cleared by
clearing the Clear global non-periodic IN NAK bit in the OTG_FS_DCTL register (CGINAK
bit in OTG_FS_DCTL).
This interrupt does not necessarily mean that a NAK handshake is sent out on the USB. The
STALL bit takes precedence over the NAK bit.
Bit 5 NPTXFE: Non-periodic TxFIFO empty
This interrupt is asserted when the non-periodic TxFIFO is either half or completely empty,
and there is space for at least one entry to be written to the non-periodic transmit request
queue. The half or completely empty status is determined by the non-periodic TxFIFO empty
level bit in the OTG_FS_GAHBCFG register (TXFELVL bit in OTG_FS_GAHBCFG).
Note: Accessible in host mode only.
Bit 4 RXFLVL: RxFIFO non-empty
Indicates that there is at least one packet pending to be read from the RxFIFO.
Note: Accessible in both host and device modes.
RM0090 Rev 19 1281/1751

1380
Bit 3 SOF: Start of frame

In host mode, the core sets this bit to indicate that an SOF (FS), or Keep-Alive (LS) is
transmitted on the USB. The application must write a 1 to this bit to clear the interrupt.
In device mode, in the core sets this bit to indicate that an SOF token has been received on
the USB. The application can read the Device Status register to get the current frame
number. This interrupt is seen only when the core is operating in FS.
Bit 2 OTGINT: OTG interrupt
The core sets this bit to indicate an OTG protocol event. The application must read the OTG
Interrupt Status (OTG_FS_GOTGINT) register to determine the exact event that caused this
interrupt. The application must clear the appropriate status bit in the OTG_FS_GOTGINT
register to clear this bit.
Bit 1 MMIS: Mode mismatch interrupt
The core sets this bit when the application is trying to access:
– A host mode register, when the core is operating in device mode
– A device mode register, when the core is operating in host mode
The register access is completed on the AHB with an OKAY response, but is ignored by the
core internally and does not affect the operation of the core.
Bit 0 CMOD: Current mode of operation
Indicates the current mode.
0: Device mode
1: Host mode
1282/1751 RM0090 Rev 19

OTG_FS interrupt mask register (OTG_FS_GINTMSK)

This register works with the Core interrupt register to interrupt the application. When an
interrupt bit is masked, the interrupt associated with that bit is not generated. However, the
Core Interrupt (OTG_FS_GINTSTS) register bit corresponding to that interrupt is still set.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
IPXFRM/IISOOXFRM
GONAKEFFM
ENUMDNEM
GINAKEFFM
USBSUSPM
ISOODRPM
CIDSCHGM
IISOIXFRM
NPTXFEM
RXFLVLM
ESUSPM
USBRST
PTXFEM
DISCINT
OTGINT
OEPINT
EOPFM
MMISM
SRQIM
IEPINT
PRTIM
SOFM
WUIM
HCIM
Reserved
Reserved
Reserved
Reserved
Reserved
Bit 31 WUIM: Resume/remote wakeup detected interrupt mask

0: Masked interrupt
1: Unmasked interrupt
Bit 30 SRQIM: Session request/new session detected interrupt mask
0: Masked interrupt
Bit 29 DISCINT: Disconnect detected interrupt mask
0: Masked interrupt
Bit 28 CIDSCHGM: Connector ID status change mask
0: Masked interrupt
Bit 26 PTXFEM: Periodic TxFIFO empty mask
0: Masked interrupt
Bit 25 HCIM: Host channels interrupt mask
0: Masked interrupt
Bit 24 PRTIM: Host port interrupt mask
0: Masked interrupt
RM0090 Rev 19 1283/1751

1380

Bit 21 IPXFRM: Incomplete periodic transfer mask
0: Masked interrupt
IISOOXFRM: Incomplete isochronous OUT transfer mask
0: Masked interrupt
Bit 20 IISOIXFRM: Incomplete isochronous IN transfer mask
0: Masked interrupt
Bit 19 OEPINT: OUT endpoints interrupt mask
0: Masked interrupt
Bit 18 IEPINT: IN endpoints interrupt mask
0: Masked interrupt
Bit 15 EOPFM: End of periodic frame interrupt mask
0: Masked interrupt
Bit 14 ISOODRPM: Isochronous OUT packet dropped interrupt mask
0: Masked interrupt
Bit 13 ENUMDNEM: Enumeration done mask
0: Masked interrupt
Bit 12 USBRST: USB reset mask
0: Masked interrupt
Bit 11 USBSUSPM: USB suspend mask
0: Masked interrupt
1284/1751 RM0090 Rev 19

Bit 10 ESUSPM: Early suspend mask

0: Masked interrupt
Bit 7 GONAKEFFM: Global OUT NAK effective mask
0: Masked interrupt
Bit 6 GINAKEFFM: Global non-periodic IN NAK effective mask
0: Masked interrupt
Bit 5 NPTXFEM: Non-periodic TxFIFO empty mask
0: Masked interrupt
Note: Only accessible in Host mode.
Bit 4 RXFLVLM: Receive FIFO non-empty mask
0: Masked interrupt
Bit 3 SOFM: Start of frame mask
0: Masked interrupt
Bit 2 OTGINT: OTG interrupt mask
0: Masked interrupt
Bit 1 MMISM: Mode mismatch interrupt mask
0: Masked interrupt
RM0090 Rev 19 1285/1751

1380
OTG_FS Receive status debug read/OTG status read and pop registers
(OTG_FS_GRXSTSR/OTG_FS_GRXSTSP)
Address offset for Read: 0x01C
Address offset for Pop: 0x020
A read to the Receive status debug read register returns the contents of the top of the
Receive FIFO. A read to the Receive status read and pop register additionally pops the top
data entry out of the RxFIFO.
The receive status contents must be interpreted differently in host and device modes. The
core ignores the receive status pop/read when the receive FIFO is empty and returns a
value of 0x0000 0000. The application must only pop the Receive Status FIFO when the
Receive FIFO non-empty bit of the Core interrupt register (RXFLVL bit in
OTG_FS_GINTSTS) is asserted.
Host mode
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PKTSTS DPID BCNT CHNUM
Reserved
r r r r

Bits 20:17 PKTSTS: Packet status
Indicates the status of the received packet
0010: IN data packet received
0011: IN transfer completed (triggers an interrupt)
0101: Data toggle error (triggers an interrupt)
0111: Channel halted (triggers an interrupt)
Others: Reserved
Bits 16:15 DPID: Data PID
Indicates the Data PID of the received packet
00: DATA0
10: DATA1
01: DATA2
11: MDATA
Bits 14:4 BCNT: Byte count
Indicates the byte count of the received IN data packet.
Bits 3:0 CHNUM: Channel number
Indicates the channel number to which the current received packet belongs.
1286/1751 RM0090 Rev 19

Device mode
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FRMNUM PKTSTS DPID BCNT EPNUM

Reserved
r r r r r

Bits 24:21 FRMNUM: Frame number
This is the least significant 4 bits of the frame number in which the packet is received on the
USB. This field is supported only when isochronous OUT endpoints are supported.
Bits 20:17 PKTSTS: Packet status
Indicates the status of the received packet
0001: Global OUT NAK (triggers an interrupt)
0010: OUT data packet received
0011: OUT transfer completed (triggers an interrupt)
0100: SETUP transaction completed (triggers an interrupt)
0110: SETUP data packet received
Others: Reserved
Indicates the Data PID of the received OUT data packet
00: DATA0
10: DATA1
01: DATA2
11: MDATA
Bits 14:4 BCNT: Byte count
Indicates the byte count of the received data packet.
Bits 3:0 EPNUM: Endpoint number
Indicates the endpoint number to which the current received packet belongs.
OTG_FS Receive FIFO size register (OTG_FS_GRXFSIZ)

The application can program the RAM size that must be allocated to the RxFIFO.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RXFD
Reserved
rw

Bits 15:0 RXFD: RxFIFO depth
This value is in terms of 32-bit words.
Minimum value is 16
Maximum value is 256
The power-on reset value of this register is specified as the largest Rx data FIFO depth.
RM0090 Rev 19 1287/1751

1380
OTG_FS Host non-periodic transmit FIFO size register

(OTG_FS_HNPTXFSIZ)/Endpoint 0 Transmit FIFO size (OTG_FS_DIEPTXF0)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
NPTXFD/TX0FD NPTXFSA/TX0FSA
rw rw
Host mode
Bits 31:16 NPTXFD: Non-periodic TxFIFO depth

Minimum value is 16
Bits 15:0 NPTXFSA: Non-periodic transmit RAM start address
This field contains the memory start address for non-periodic transmit FIFO RAM.
Device mode
Bits 31:16 TX0FD: Endpoint 0 TxFIFO depth

Minimum value is 16
Bits 15:0 TX0FSA: Endpoint 0 transmit RAM start address
This field contains the memory start address for the endpoint 0 transmit FIFO RAM.
OTG_FS non-periodic transmit FIFO/queue status register

(OTG_FS_HNPTXSTS)
Note: In Device mode, this register is not valid.
This read-only register contains the free space information for the non-periodic TxFIFO and
the non-periodic transmit request queue.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserved
NPTXQTOP NPTQXSAV NPTXFSAV
r r r
1288/1751 RM0090 Rev 19


Bits 30:24 NPTXQTOP: Top of the non-periodic transmit request queue
Entry in the non-periodic Tx request queue that is currently being processed by the MAC.
Bits 30:27: Channel/endpoint number
Bits 26:25:
– 00: IN/OUT token
– 01: Zero-length transmit packet (device IN/host OUT)
– 11: Channel halt command
Bit 24: Terminate (last entry for selected channel/endpoint)
Bits 23:16 NPTQXSAV: Non-periodic transmit request queue space available
Indicates the amount of free space available in the non-periodic transmit request queue.
This queue holds both IN and OUT requests in host mode. Device mode has only IN
requests.
00: Non-periodic transmit request queue is full
01: 1 location available
10: 2 locations available
bxn: n locations available (0 ≤ n ≤ 8)
Others: Reserved
Bits 15:0 NPTXFSAV: Non-periodic TxFIFO space available
Indicates the amount of free space available in the non-periodic TxFIFO.
Values are in terms of 32-bit words.
00: Non-periodic TxFIFO is full
01: 1 word available
10: 2 words available
0xn: n words available (where 0 ≤ n ≤ 256)
Others: Reserved
OTG_FS general core configuration register (OTG_FS_GCCFG)

Reset value: 0x0000 XXXX
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
NOVBUSSENS
SOFOUTEN
VBUSBSEN
VBUSASEN
.PWRDWN
Reserved
Reserved Reserved
rw rw rw rw rw
RM0090 Rev 19 1289/1751

1380

Bit 21 NOVBUSSENS: VBUS sensing disable option
When this bit is set, VBUS is considered internally to be always at VBUS valid level (5 V). This
option removes the need for a dedicated VBUS pad, and leave this pad free to be used for
other purposes such as a shared functionality. VBUS connection can be remapped on
another general purpose input pad and monitored by software.
This option is only suitable for host-only or device-only applications.
0: VBUS sensing available by hardware
1: VBUS sensing not available by hardware.
Bit 20 SOFOUTEN: SOF output enable
0: SOF pulse not available on PAD (OTG_FS_SOF)
1: SOF pulse available on PAD (OTG_FS_SOF)
Bit 19 VBUSBSEN: Enable the VBUS sensing “B” device
0: VBUS sensing “B” disabled
1: VBUS sensing “B” enabled
Bit 18 VBUSASEN: Enable the VBUS sensing “A” device
0: VBUS sensing “A” disabled
1: VBUS sensing “A” enabled
Bit 16 PWRDWN: Power down
Used to activate the transceiver in transmission/reception
0: Power down active
1: Power down deactivated (“Transceiver active”)
OTG_FS core ID register (OTG_FS_CID)

Reset value:0x0000 1200
This is a register containing the Product ID as reset value.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PRODUCT_ID
Bits 31:0 PRODUCT_ID: Product ID field

Application-programmable ID field.
1290/1751 RM0090 Rev 19

OTG_FS Host periodic transmit FIFO size register (OTG_FS_HPTXFSIZ)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PTXFSIZ PTXSA
Bits 31:16 PTXFD: Host periodic TxFIFO depth

Minimum value is 16
Bits 15:0 PTXSA: Host periodic TxFIFO start address
The power-on reset value of this register is the sum of the largest Rx data FIFO depth and
largest non-periodic Tx data FIFO depth.
OTG_FS device IN endpoint transmit FIFO size register (OTG_FS_DIEPTXFx)

(x = 1..3, where x is the FIFO_number)
Address offset: 0x104 + 0x04 * (x -1)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
INEPTXFD INEPTXSA
Bits 31:16 INEPTXFD: IN endpoint TxFIFO depth

Minimum value is 16
The power-on reset value of this register is specified as the largest IN endpoint FIFO
number depth.
Bits 15:0 INEPTXSA: IN endpoint FIFOx transmit RAM start address
This field contains the memory start address for IN endpoint transmit FIFOx. The address
must be aligned with a 32-bit memory location.
RM0090 Rev 19 1291/1751

1380
34.16.3 Host-mode registers

Bit values in the register descriptions are expressed in binary unless otherwise specified.
Host-mode registers affect the operation of the core in the host mode. Host mode registers
must not be accessed in device mode, as the results are undefined. Host mode registers
can be categorized as follows:
OTG_FS Host configuration register (OTG_FS_HCFG)

This register configures the core after power-on. Do not make changes to this register after
initializing the host.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FSLSPCS
FSLSS
Reserved
r rw rw

Bit 2 FSLSS: FS- and LS-only support
The application uses this bit to control the core’s enumeration speed. Using this bit, the
application can make the core enumerate as an FS host, even if the connected device
supports HS traffic. Do not make changes to this field after initial programming.
1: FS/LS-only, even if the connected device can support HS (read-only)
Bits 1:0 FSLSPCS: FS/LS PHY clock select
When the core is in FS host mode
01: PHY clock is running at 48 MHz
Others: Reserved
When the core is in LS host mode
00: Reserved
01: Select 48 MHz PHY clock frequency
10: Select 6 MHz PHY clock frequency
11: Reserved
Note: The FSLSPCS must be set on a connection event according to the speed of the
connected device (after changing this bit, a software reset must be performed).
OTG_FS Host frame interval register (OTG_FS_HFIR)

Reset value: 0x0000 EA60
This register stores the frame interval information for the current speed to which the
OTG_FS controller has enumerated.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FRIVL
Reserved
1292/1751 RM0090 Rev 19


Bits 15:0 FRIVL: Frame interval
The value that the application programs to this field specifies the interval between two
consecutive SOFs (FS) or Keep-Alive tokens (LS). This field contains the number of PHY
clocks that constitute the required frame interval. The application can write a value to this
register only after the Port enable bit of the host port control and status register (PENA bit in
OTG_FS_HPRT) has been set. If no value is programmed, the core calculates the value
based on the PHY clock specified in the FS/LS PHY Clock Select field of the host
configuration register (FSLSPCS in OTG_FS_HCFG). Do not change the value of this field
after the initial configuration.
– Frame interval = 1 ms × (FRIVL - 1)
OTG_FS Host frame number/frame time remaining register (OTG_FS_HFNUM)

Reset value: 0x0000 3FFF
This register indicates the current frame number. It also indicates the time remaining (in
terms of the number of PHY clocks) in the current frame.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FTREM FRNUM
Bits 31:16 FTREM: Frame time remaining

Indicates the amount of time remaining in the current frame, in terms of PHY clocks. This
field decrements on each PHY clock. When it reaches zero, this field is reloaded with the
value in the Frame interval register and a new SOF is transmitted on the USB.
Bits 15:0 FRNUM: Frame number
This field increments when a new SOF is transmitted on the USB, and is cleared to 0 when
it reaches 0x3FFF.
OTG_FS_Host periodic transmit FIFO/queue status register

(OTG_FS_HPTXSTS)
This read-only register contains the free space information for the periodic TxFIFO and the
periodic transmit request queue.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PTXQTOP PTXQSAV PTXFSAVL
RM0090 Rev 19 1293/1751

1380
Bits 31:24 PTXQTOP: Top of the periodic transmit request queue

This indicates the entry in the periodic Tx request queue that is currently being processed by
the MAC.
This register is used for debugging.
Bit 31: Odd/Even frame
– 0: send in even frame
– 1: send in odd frame
Bits 30:27: Channel/endpoint number
Bits 26:25: Type
– 00: IN/OUT
– 01: Zero-length packet
– 11: Disable channel command
Bit 24: Terminate (last entry for the selected channel/endpoint)
Bits 23:16 PTXQSAV: Periodic transmit request queue space available
Indicates the number of free locations available to be written in the periodic transmit request
queue. This queue holds both IN and OUT requests.
00: Periodic transmit request queue is full
01: 1 location available
10: 2 locations available
bxn: n locations available (0 ≤ n ≤ 8)
Others: Reserved
Bits 15:0 PTXFSAVL: Periodic transmit data FIFO space available
Indicates the number of free locations available to be written to in the periodic TxFIFO.
Values are in terms of 32-bit words
0000: Periodic TxFIFO is full
0001: 1 word available
0010: 2 words available
bxn: n words available (where 0 ≤ n ≤ PTXFD)
Others: Reserved
OTG_FS Host all channels interrupt register (OTG_FS_HAINT)

When a significant event occurs on a channel, the host all channels interrupt register
interrupts the application using the host channels interrupt bit of the Core interrupt register
(HCINT bit in OTG_FS_GINTSTS). This is shown in Figure 394. There is one interrupt bit
per channel, up to a maximum of 16 bits. Bits in this register are set and cleared when the
application sets and clears bits in the corresponding host channel-x interrupt register.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
HAINT
Reserved

Bits 15:0 HAINT: Channel interrupts
One bit per channel: Bit 0 for Channel 0, bit 15 for Channel 15
1294/1751 RM0090 Rev 19

OTG_FS Host all channels interrupt mask register (OTG_FS_HAINTMSK)

The host all channel interrupt mask register works with the host all channel interrupt register
to interrupt the application when an event occurs on a channel. There is one interrupt mask
bit per channel, up to a maximum of 16 bits.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
HAINTM
Reserved

Bits 15:0 HAINTM: Channel interrupt mask
0: Masked interrupt
One bit per channel: Bit 0 for channel 0, bit 15 for channel 15
OTG_FS Host port control and status register (OTG_FS_HPRT)

This register is available only in host mode. Currently, the OTG host supports only one port.
A single register holds USB port-related information such as USB reset, enable, suspend,
resume, connect status, and test mode for each port. It is shown in Figure 394. The rc_w1
bits in this register can trigger an interrupt to the application through the host port interrupt
bit of the core interrupt register (HPRTINT bit in OTG_FS_GINTSTS). On a Port Interrupt,
the application must read this register and clear the bit that caused the interrupt. For the
rc_w1 bits, the application must write a 1 to the bit to clear the interrupt.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
POCCHNG
PENCHNG
PCDET
PSUSP
PCSTS
PLSTS
PPWR
POCA
PRES
PENA
PRST
Reserved
PSPD PTCTL
Reserved
rc_ rc_ rc_ rc_

r r rw rw rw rw rw r r rw rs rw r r
w1 w1 w0 w1

Bits 18:17 PSPD: Port speed
Indicates the speed of the device attached to this port.
01: Full speed
10: Low speed
11: Reserved
RM0090 Rev 19 1295/1751

1380
Bits 16:13 PTCTL: Port test control

The application writes a nonzero value to this field to put the port into a Test mode, and the
corresponding pattern is signaled on the port.
0000: Test mode disabled
0001: Test_J mode
0010: Test_K mode
0011: Test_SE0_NAK mode
0100: Test_Packet mode
0101: Test_Force_Enable
Others: Reserved
Bit 12 PPWR: Port power
The application uses this field to control power to this port, and the core clears this bit on an
overcurrent condition.
0: Power off
1: Power on
Bits 11:10 PLSTS: Port line status
Indicates the current logic level USB data lines
Bit 10: Logic level of OTG_FS_DP
Bit 11: Logic level of OTG_FS_DM
Bit 8 PRST: Port reset
When the application sets this bit, a reset sequence is started on this port. The application
must time the reset period and clear this bit after the reset sequence is complete.
0: Port not in reset
1: Port in reset
The application must leave this bit set for a minimum duration of at least 10 ms to start a
reset on the port. The application can leave it set for another 10 ms in addition to the
required minimum duration, before clearing the bit, even though there is no maximum limit
set by the USB standard.
Bit 7 PSUSP: Port suspend
The application sets this bit to put this port in Suspend mode. The core only stops sending
SOFs when this is set. To stop the PHY clock, the application must set the Port clock stop
bit, which asserts the suspend input pin of the PHY.
The read value of this bit reflects the current suspend status of the port. This bit is cleared
by the core after a remote wakeup signal is detected or the application sets the Port reset bit
or Port resume bit in this register or the Resume/remote wakeup detected interrupt bit or
Disconnect detected interrupt bit in the Core interrupt register (WKUINT or DISCINT in
OTG_FS_GINTSTS, respectively).
0: Port not in Suspend mode
1: Port in Suspend mode
Bit 6 PRES: Port resume
The application sets this bit to drive resume signaling on the port. The core continues to
drive the resume signal until the application clears this bit.
If the core detects a USB remote wakeup sequence, as indicated by the Port
resume/remote wakeup detected interrupt bit of the Core interrupt register (WKUINT bit in
OTG_FS_GINTSTS), the core starts driving resume signaling without application
intervention and clears this bit when it detects a disconnect condition. The read value of this
bit indicates whether the core is currently driving resume signaling.
0: No resume driven
1: Resume driven
1296/1751 RM0090 Rev 19

Bit 5 POCCHNG: Port overcurrent change

The core sets this bit when the status of the Port overcurrent active bit (bit 4) in this register
changes.
Bit 4 POCA: Port overcurrent active
Indicates the overcurrent condition of the port.
0: No overcurrent condition
1: Overcurrent condition
Bit 3 PENCHNG: Port enable/disable change
The core sets this bit when the status of the Port enable bit 2 in this register changes.
Bit 2 PENA: Port enable
A port is enabled only by the core after a reset sequence, and is disabled by an overcurrent
condition, a disconnect condition, or by the application clearing this bit. The application
cannot set this bit by a register write. It can only clear it to disable the port. This bit does not
trigger any interrupt to the application.
0: Port disabled
1: Port enabled
Bit 1 PCDET: Port connect detected
The core sets this bit when a device connection is detected to trigger an interrupt to the
application using the host port interrupt bit in the Core interrupt register (HPRTINT bit in
OTG_FS_GINTSTS). The application must write a 1 to this bit to clear the interrupt.
Bit 0 PCSTS: Port connect status
0: No device is attached to the port
1: A device is attached to the port
RM0090 Rev 19 1297/1751

1380
OTG_FS Host channel-x characteristics register (OTG_FS_HCCHARx)

(x = 0..7, where x = Channel_number)
Address offset: 0x500 + 0x20 * x
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ODDFRM
CHENA
EPTYP
LSDEV
CHDIS
EPDIR
Reserved
DAD MCNT EPNUM MPSIZ
rs rs rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
Bit 31 CHENA: Channel enable

This field is set by the application and cleared by the OTG host.
0: Channel disabled
1: Channel enabled
Bit 30 CHDIS: Channel disable
The application sets this bit to stop transmitting/receiving data on a channel, even before
the transfer for that channel is complete. The application must wait for the Channel disabled
interrupt before treating the channel as disabled.
Bit 29 ODDFRM: Odd frame
This field is set (reset) by the application to indicate that the OTG host must perform a
transfer in an odd frame. This field is applicable for only periodic (isochronous and interrupt)
transactions.
0: Even frame
1: Odd frame
Bits 28:22 DAD: Device address
This field selects the specific device serving as the data source or sink.
Bits 21:20 MCNT: Multicount
This field indicates to the host the number of transactions that must be executed per frame
for this periodic endpoint. For non-periodic transfers, this field is not used
00: Reserved. This field yields undefined results
01: 1 transaction
10: 2 transactions per frame to be issued for this endpoint
11: 3 transactions per frame to be issued for this endpoint
Note: This field must be set to at least 01.
Bits 19:18 EPTYP: Endpoint type
Indicates the transfer type selected.
00: Control
01: Isochronous
10: Bulk
11: Interrupt
Bit 17 LSDEV: Low-speed device
This field is set by the application to indicate that this channel is communicating to a low-
speed device.
1298/1751 RM0090 Rev 19

Bit 15 EPDIR: Endpoint direction

Indicates whether the transaction is IN or OUT.
0: OUT
1: IN
Bits 14:11 EPNUM: Endpoint number
Indicates the endpoint number on the device serving as the data source or sink.
Bits 10:0 MPSIZ: Maximum packet size
Indicates the maximum packet size of the associated endpoint.
OTG_FS Host channel-x interrupt register (OTG_FS_HCINTx) (x = 0..7, where

x = Channel_number)
This register indicates the status of a channel with respect to USB- and AHB-related events.
It is shown in Figure 394. The application must read this register when the host channels
interrupt bit in the Core interrupt register (HCINT bit in OTG_FS_GINTSTS) is set. Before
the application can read this register, it must first read the host all channels interrupt
(OTG_FS_HAINT) register to get the exact channel number for the host channel-x interrupt
corresponding bits in the OTG_FS_HAINT and OTG_FS_GINTSTS registers.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FRMOR
DTERR
BBERR
TXERR
STALL
XFRC
CHH
ACK
NAK
Reserved
Reserved
Reserved
rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_
w1 w1 w1 w1 w1 w1 w1 w1 w1

Bit 10 DTERR: Data toggle error
Bit 9 FRMOR: Frame overrun
Bit 8 BBERR: Babble error
Bit 7 TXERR: Transaction error
Indicates one of the following errors occurred on the USB.
CRC check failure
Timeout
Bit stuff error
False EOP
Bit 5 ACK: ACK response received/transmitted interrupt
Bit 4 NAK: NAK response received interrupt
Bit 3 STALL: STALL response received interrupt
RM0090 Rev 19 1299/1751

1380

Bit 1 CHH: Channel halted
Indicates the transfer completed abnormally either because of any USB transaction error or
in response to disable request by the application.
Bit 0 XFRC: Transfer completed
Transfer completed normally without any errors.
OTG_FS Host channel-x interrupt mask register (OTG_FS_HCINTMSKx)

(x = 0..7, where x = Channel_number)
Address offset: 0x50C + 0x20 * x
This register reflects the mask for each channel status described in the previous section.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FRMORM
DTERRM
BBERRM
TXERRM
STALLM
XFRCM
CHHM
ACKM
NAKM
Reserved
Reserved
Reserved

Bit 10 DTERRM: Data toggle error mask
0: Masked interrupt
Bit 9 FRMORM: Frame overrun mask
0: Masked interrupt
Bit 8 BBERRM: Babble error mask
0: Masked interrupt
Bit 7 TXERRM: Transaction error mask
0: Masked interrupt
Bit 5 ACKM: ACK response received/transmitted interrupt mask
0: Masked interrupt
Bit 4 NAKM: NAK response received interrupt mask
0: Masked interrupt
Bit 3 STALLM: STALL response received interrupt mask
0: Masked interrupt
1300/1751 RM0090 Rev 19


Bit 1 CHHM: Channel halted mask
0: Masked interrupt
Bit 0 XFRCM: Transfer completed mask
0: Masked interrupt
OTG_FS Host channel-x transfer size register (OTG_FS_HCTSIZx) (x = 0..7,

where x = Channel_number)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserved
DPID PKTCNT XFRSIZ

The application programs this field with the type of PID to use for the initial transaction. The
host maintains this field for the rest of the transfer.
00: DATA0
01: DATA2
10: DATA1
11: MDATA (non-control)/SETUP (control)
Bits 28:19 PKTCNT: Packet count
This field is programmed by the application with the expected number of packets to be
transmitted (OUT) or received (IN).
The host decrements this count on every successful transmission or reception of an OUT/IN
packet. Once this count reaches zero, the application is interrupted to indicate normal
completion.
Bits 18:0 XFRSIZ: Transfer size
For an OUT, this field is the number of data bytes the host sends during the transfer.
For an IN, this field is the buffer size that the application has reserved for the transfer. The
application is expected to program this field as an integer multiple of the maximum packet
size for IN transactions (periodic and non-periodic).
RM0090 Rev 19 1301/1751

1380
34.16.4 Device-mode registers

OTG_FS device configuration register (OTG_FS_DCFG)
This register configures the core in device mode after power-on or after certain control
commands or enumeration. Do not make changes to this register after initial programming.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
NZLSOHSK
PFIVL
DSPD
Reserved
DAD
Reserved

Bits 12:11 PFIVL: Periodic frame interval
Indicates the time within a frame at which the application must be notified using the end of
periodic frame interrupt. This can be used to determine if all the isochronous traffic for that
frame is complete.
00: 80% of the frame interval
Bits 10:4 DAD: Device address
The application must program this field after every SetAddress control command.
Bit 2 NZLSOHSK: Non-zero-length status OUT handshake
The application can use this field to select the handshake the core sends on receiving a
nonzero-length data packet during the OUT transaction of a control transfer’s Status stage.
1: Send a STALL handshake on a nonzero-length status OUT transaction and do not send
the received OUT packet to the application.
0: Send the received OUT packet to the application (zero-length or nonzero-length) and
send a handshake based on the NAK and STALL bits for the endpoint in the Device
endpoint control register.
Bits 1:0 DSPD: Device speed
Indicates the speed at which the application requires the core to enumerate, or the
maximum speed the application can support. However, the actual bus speed is determined
only after the chirp sequence is completed, and is based on the speed of the USB host to
which the core is connected.
00: Reserved
01: Reserved
10: Reserved
11: Full speed (USB 1.1 transceiver clock is 48 MHz)
1302/1751 RM0090 Rev 19

OTG_FS device control register (OTG_FS_DCTL)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
POPRGDNE
CGONAK
SGONAK
GONSTS
RWUSIG
CGINAK
SGINAK
GINSTS
TCTL
SDIS
Reserved
rw w w w w rw rw rw r r rw rw

Bit 11 POPRGDNE: Power-on programming done
The application uses this bit to indicate that register programming is completed after a
wakeup from power down mode.
Bit 10 CGONAK: Clear global OUT NAK
Writing 1 to this field clears the Global OUT NAK.
Bit 9 SGONAK: Set global OUT NAK
Writing 1 to this field sets the Global OUT NAK.
The application uses this bit to send a NAK handshake on all OUT endpoints.
The application must set the this bit only after making sure that the Global OUT NAK
effective bit in the Core interrupt register (GONAKEFF bit in OTG_FS_GINTSTS) is cleared.
Bit 8 CGINAK: Clear global IN NAK
Writing 1 to this field clears the Global IN NAK.
Bit 7 SGINAK: Set global IN NAK
Writing 1 to this field sets the Global non-periodic IN NAK. The application uses this bit to
send a NAK handshake on all non-periodic IN endpoints.
The application must set this bit only after making sure that the Global IN NAK effective bit
in the Core interrupt register (GINAKEFF bit in OTG_FS_GINTSTS) is cleared.
Bits 6:4 TCTL: Test control
000: Test mode disabled
001: Test_J mode
010: Test_K mode
011: Test_SE0_NAK mode
100: Test_Packet mode
101: Test_Force_Enable
Others: Reserved
Bit 3 GONSTS: Global OUT NAK status
0: A handshake is sent based on the FIFO Status and the NAK and STALL bit settings.
1: No data is written to the RxFIFO, irrespective of space availability. Sends a NAK
handshake on all packets, except on SETUP transactions. All isochronous OUT packets are
dropped.
RM0090 Rev 19 1303/1751

1380
Bit 2 GINSTS: Global IN NAK status

0: A handshake is sent out based on the data availability in the transmit FIFO.
1: A NAK handshake is sent out on all non-periodic IN endpoints, irrespective of the data
availability in the transmit FIFO.
Bit 1 SDIS: Soft disconnect
The application uses this bit to signal the USB OTG core to perform a soft disconnect. As
long as this bit is set, the host does not see that the device is connected, and the device
does not receive signals on the USB. The core stays in the disconnected state until the
application clears this bit.
0: Normal operation. When this bit is cleared after a soft disconnect, the core generates a
device connect event to the USB host. When the device is reconnected, the USB host
restarts device enumeration.
1: The core generates a device disconnect event to the USB host.
Bit 0 RWUSIG: Remote wakeup signaling
When the application sets this bit, the core initiates remote signaling to wake up the USB
host. The application must set this bit to instruct the core to exit the Suspend state. As
specified in the USB 2.0 specification, the application must clear this bit 1 ms to 15 ms after
setting it.
Table 204 contains the minimum duration (according to device state) for which the Soft
disconnect (SDIS) bit must be set for the USB host to detect a device disconnect. To
accommodate clock jitter, it is recommended that the application add some extra delay to
the specified minimum duration.
Table 204. Minimum duration for soft disconnect

Operating speed Device state Minimum duration
Full speed Suspended 1 ms + 2.5 µs

Full speed Idle 2.5 µs
Full speed Not Idle or Suspended (Performing transactions) 2.5 µs
OTG_FS device status register (OTG_FS_DSTS)

This register indicates the status of the core with respect to USB-related events. It must be
read on interrupts from the device all interrupts (OTG_FS_DAINT) register.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ENUMSPD
SUSPSTS
EERR
FNSOF
Reserved Reserved
r r r r r r r r r r r r r r r r r r

Bits 21:8 FNSOF: Frame number of the received SOF
1304/1751 RM0090 Rev 19

Bit 3 EERR: Erratic error

The core sets this bit to report any erratic errors.
Due to erratic errors, the OTG_FS controller goes into Suspended state and an interrupt is
generated to the application with Early suspend bit of the OTG_FS_GINTSTS register
(ESUSP bit in OTG_FS_GINTSTS). If the early suspend is asserted due to an erratic error,
the application can only perform a soft disconnect recover.
Bits 2:1 ENUMSPD: Enumerated speed
Indicates the speed at which the OTG_FS controller has come up after speed detection
through a chirp sequence.
01: Reserved
10: Reserved
11: Full speed (PHY clock is running at 48 MHz)
Others: reserved
Bit 0 SUSPSTS: Suspend status
In device mode, this bit is set as long as a Suspend condition is detected on the USB. The
core enters the Suspended state when there is no activity on the USB data lines for a period
of 3 ms. The core comes out of the suspend:
– When there is an activity on the USB data lines
– When the application writes to the Remote wakeup signaling bit in the OTG_FS_DCTL
register (RWUSIG bit in OTG_FS_DCTL).
OTG_FS device IN endpoint common interrupt mask register

(OTG_FS_DIEPMSK)
This register works with each of the OTG_FS_DIEPINTx registers for all endpoints to
generate an interrupt per IN endpoint. The IN endpoint interrupt for a specific status in the
OTG_FS_DIEPINTx register can be masked by writing to the corresponding bit in this
register. Status bits are masked by default.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ITTXFEMSK
INEPNMM
INEPNEM
XFRCM
NAKM
EPDM
Reserved
TOM
Reserved Reserved

Bit 13 NAKM: NAK interrupt mask
0: Masked interrupt
Bit 6 INEPNEM: IN endpoint NAK effective mask
0: Masked interrupt
RM0090 Rev 19 1305/1751

1380
Bit 5 INEPNMM: IN token received with EP mismatch mask

0: Masked interrupt
Bit 4 ITTXFEMSK: IN token received when TxFIFO empty mask
0: Masked interrupt
Bit 3 TOM: Timeout condition mask (Non-isochronous endpoints)
0: Masked interrupt
Bit 1 EPDM: Endpoint disabled interrupt mask
0: Masked interrupt
Bit 0 XFRCM: Transfer completed interrupt mask
0: Masked interrupt
OTG_FS device OUT endpoint common interrupt mask register

(OTG_FS_DOEPMSK)
This register works with each of the OTG_FS_DOEPINTx registers for all endpoints to
generate an interrupt per OUT endpoint. The OUT endpoint interrupt for a specific status in
the OTG_FS_DOEPINTx register can be masked by writing into the corresponding bit in this
register. Status bits are masked by default.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OUTPKTERRM
STSPHSRXM
OTEPDM
BERRM
XFRCM
STUPM
EPDM
Reserved
Reserved
NAK
Reserved Reserved

Bit 13 NAKMSK: NAK interrupt mask
0: Masked interrupt
Bit 12 BERRM: Babble error interrupt mask
0: Masked interrupt
Bit 8 OUTPKTERRM: Out packet error mask
0: Masked interrupt
1306/1751 RM0090 Rev 19


Bit 5 STSPHSRXM: Status phase received for control write mask
0: Masked interrupt
Bit 4 OTEPDM: OUT token received when endpoint disabled mask
Applies to control OUT endpoints only.
0: Masked interrupt
Bit 3 STUPM: SETUP phase done mask
Applies to control endpoints only.
0: Masked interrupt
Bit 1 EPDM: Endpoint disabled interrupt mask
0: Masked interrupt
Bit 0 XFRCM: Transfer completed interrupt mask
0: Masked interrupt
OTG_FS device all endpoints interrupt register (OTG_FS_DAINT)

When a significant event occurs on an endpoint, a OTG_FS_DAINT register interrupts the
application using the Device OUT endpoints interrupt bit or Device IN endpoints interrupt bit
of the OTG_FS_GINTSTS register (OEPINT or IEPINT in OTG_FS_GINTSTS,
respectively). There is one interrupt bit per endpoint, up to a maximum of 16 bits for OUT
endpoints and 16 bits for IN endpoints. For a bidirectional endpoint, the corresponding IN
and OUT interrupt bits are used. Bits in this register are set and cleared when the
application sets and clears bits in the corresponding Device Endpoint-x interrupt register
(OTG_FS_DIEPINTx/OTG_FS_DOEPINTx).
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OEPINT IEPINT
Bits 31:16 OEPINT: OUT endpoint interrupt bits

One bit per OUT endpoint:
Bit 16 for OUT endpoint 0, bit 19 for OUT endpoint 3.
Bits 15:0 IEPINT: IN endpoint interrupt bits
One bit per IN endpoint:
Bit 0 for IN endpoint 0, bit 3 for endpoint 3.
RM0090 Rev 19 1307/1751

1380
OTG_FS all endpoints interrupt mask register (OTG_FS_DAINTMSK)

The OTG_FS_DAINTMSK register works with the Device endpoint interrupt register to
interrupt the application when an event occurs on a device endpoint. However, the
OTG_FS_DAINT register bit corresponding to that interrupt is still set.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OEPM IEPM
Bits 31:16 OEPM: OUT EP interrupt mask bits

One per OUT endpoint:
Bit 16 for OUT EP 0, bit 19 for OUT EP 3
0: Masked interrupt
Bits 15:0 IEPM: IN EP interrupt mask bits
One bit per IN endpoint:
Bit 0 for IN EP 0, bit 3 for IN EP 3
0: Masked interrupt
OTG_FS device VBUS discharge time register (OTG_FS_DVBUSDIS)

Reset value: 0x0000 17D7
This register specifies the VBUS discharge time after VBUS pulsing during SRP.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
VBUSDT
Reserved

Bits 15:0 VBUSDT: Device VBUS discharge time
Specifies the VBUS discharge time after VBUS pulsing during SRP. This value equals:
VBUS discharge time in PHY clocks / 1 024
Depending on your VBUS load, this value may need adjusting.
OTG_FS device VBUS pulsing time register (OTG_FS_DVBUSPULSE)

Reset value: 0x0000 05B8
This register specifies the VBUS pulsing time during SRP.
1308/1751 RM0090 Rev 19

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DVBUSP
Reserved

Bits 11:0 DVBUSP: Device VBUS pulsing time
Specifies the VBUS pulsing time during SRP. This value equals:
VBUS pulsing time in PHY clocks / 1 024
OTG_FS device IN endpoint FIFO empty interrupt mask register:

(OTG_FS_DIEPEMPMSK)
This register is used to control the IN endpoint FIFO empty interrupt generation
(TXFE_OTG_FS_DIEPINTx).
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
INEPTXFEM
Reserved

Bits 15:0 INEPTXFEM: IN EP Tx FIFO empty interrupt mask bits
These bits act as mask bits for OTG_FS_DIEPINTx.
TXFE interrupt one bit per IN endpoint:
Bit 0 for IN endpoint 0, bit 3 for IN endpoint 3
0: Masked interrupt
OTG_FS device control IN endpoint 0 control register (OTG_FS_DIEPCTL0)

This section describes the OTG_FS_DIEPCTL0 register. Nonzero control endpoints use
registers for endpoints 1–3.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
USBAEP
NAKSTS
EPENA
EPDIS
STALL
CNAK
SNAK
Reserved
Reserved
Reserved
TXFNUM EPTYP MPSIZ

Reserved
r r w w rw rw rw rw rs r r r r rw rw
RM0090 Rev 19 1309/1751

1380
Bit 31 EPENA: Endpoint enable

The application sets this bit to start transmitting data on the endpoint 0.
The core clears this bit before setting any of the following interrupts on this endpoint:
– Endpoint disabled
– Transfer completed
Bit 30 EPDIS: Endpoint disable
The application sets this bit to stop transmitting data on an endpoint, even before the
transfer for that endpoint is complete. The application must wait for the Endpoint disabled
interrupt before treating the endpoint as disabled. The core clears this bit before setting the
Endpoint disabled interrupt. The application must set this bit only if Endpoint enable is
already set for this endpoint.
Bit 27 SNAK: Set NAK
A write to this bit sets the NAK bit for the endpoint.
Using this bit, the application can control the transmission of NAK handshakes on an
endpoint. The core can also set this bit for an endpoint after a SETUP packet is received on
that endpoint.
Bit 26 CNAK: Clear NAK
A write to this bit clears the NAK bit for the endpoint.
This value is set to the FIFO number that is assigned to IN endpoint 0.
Bit 21 STALL: STALL handshake
The application can only set this bit, and the core clears it when a SETUP token is received
for this endpoint. If a NAK bit, a Global IN NAK or Global OUT NAK is set along with this bit,
the STALL bit takes priority.
Hardcoded to ‘00’ for control.
Bit 17 NAKSTS: NAK status
Indicates the following:
0: The core is transmitting non-NAK handshakes based on the FIFO status
1: The core is transmitting NAK handshakes on this endpoint.
When this bit is set, either by the application or core, the core stops transmitting data, even
if there are data available in the TxFIFO. Irrespective of this bit’s setting, the core always
responds to SETUP data packets with an ACK handshake.
1310/1751 RM0090 Rev 19

Bit 15 USBAEP: USB active endpoint

This bit is always set to 1, indicating that control endpoint 0 is always active in all
configurations and interfaces.
The application must program this field with the maximum packet size for the current logical
endpoint.
00: 64 bytes
01: 32 bytes
10: 16 bytes
11: 8 bytes
OTG device endpoint x control register (OTG_FS_DIEPCTLx) (x = 1..3, where

x = Endpoint_number)
The application uses this register to control the behavior of each logical endpoint other than
endpoint 0.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SD0PID/SEVNFRM
EONUM/DPID
SODDFRM
USBAEP
NAKSTS
EPENA
EPTYP
EPDIS
STALL
CNAK
SNAK
Reserved
TXFNUM MPSIZ
Reserved
rs rs w w w w rw rw rw rw rw rw rw r r rw rw rw rw rw rw rw rw rw rw rw rw

The application sets this bit to start transmitting data on an endpoint.
– SETUP phase done
The application sets this bit to stop transmitting/receiving data on an endpoint, even before
the transfer for that endpoint is complete. The application must wait for the Endpoint
disabled interrupt before treating the endpoint as disabled. The core clears this bit before
setting the Endpoint disabled interrupt. The application must set this bit only if Endpoint
enable is already set for this endpoint.
Bit 29 SODDFRM: Set odd frame
Applies to isochronous IN and OUT endpoints only.
Writing to this field sets the Even/Odd frame (EONUM) field to odd frame.
RM0090 Rev 19 1311/1751

1380
Bit 28 SD0PID: Set DATA0 PID

Applies to interrupt/bulk IN endpoints only.
Writing to this field sets the endpoint data PID (DPID) field in this register to DATA0.
SEVNFRM: Set even frame
Applies to isochronous IN endpoints only.
Writing to this field sets the Even/Odd frame (EONUM) field to even frame.
endpoint. The core can also set this bit for OUT endpoints on a Transfer completed interrupt,
or after a SETUP is received on the endpoint.
These bits specify the FIFO number associated with this endpoint. Each active IN endpoint
must be programmed to a separate FIFO number.
This field is valid only for IN endpoints.
Applies to non-control, non-isochronous IN endpoints only (access type is rw).
The application sets this bit to stall all tokens from the USB host to this endpoint. If a NAK
bit, Global IN NAK, or Global OUT NAK is set along with this bit, the STALL bit takes priority.
Only the application can clear this bit, never the core.
This is the transfer type supported by this logical endpoint.
00: Control
01: Isochronous
10: Bulk
11: Interrupt
It indicates the following:
0: The core is transmitting non-NAK handshakes based on the FIFO status.
When either the application or the core sets this bit:
For non-isochronous IN endpoints: The core stops transmitting any data on an IN endpoint,
even if there are data available in the TxFIFO.
For isochronous IN endpoints: The core sends out a zero-length data packet, even if there
are data available in the TxFIFO.
Irrespective of this bit’s setting, the core always responds to SETUP data packets with an
ACK handshake.
1312/1751 RM0090 Rev 19

Bit 16 EONUM: Even/odd frame

Applies to isochronous IN endpoints only.
Indicates the frame number in which the core transmits/receives isochronous data for this
endpoint. The application must program the even/odd frame number in which it intends to
transmit/receive isochronous data for this endpoint using the SEVNFRM and SODDFRM
fields in this register.
0: Even frame
1: Odd frame
DPID: Endpoint data PID
Applies to interrupt/bulk IN endpoints only.
Contains the PID of the packet to be received or transmitted on this endpoint. The
application must program the PID of the first packet to be received or transmitted on this
endpoint, after the endpoint is activated. The application uses the SD0PID register field to
program either DATA0 or DATA1 PID.
0: DATA0
1: DATA1
Indicates whether this endpoint is active in the current configuration and interface. The core
clears this bit for all endpoints (other than EP 0) after detecting a USB reset. After receiving
the SetConfiguration and SetInterface commands, the application must program endpoint
registers accordingly and set this bit.
endpoint. This value is in bytes.
RM0090 Rev 19 1313/1751

1380
OTG_FS device control OUT endpoint 0 control register

(OTG_FS_DOEPCTL0)
Address offset: 0xB00
This section describes the OTG_FS_DOEPCTL0 register. Nonzero control endpoints use
registers for endpoints 1–3.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
USBAEP
NAKSTS
EPENA
EPDIS
STALL
SNPM
CNAK
SNAK
Reserved
Reserved
EPTYP MPSIZ
Reserved Reserved
w r w w rs rw r r r r r r

The application sets this bit to start transmitting data on endpoint 0.
The application cannot disable control OUT endpoint 0.
endpoint. The core can also set this bit on a Transfer completed interrupt, or after a SETUP
is received on the endpoint.
The application can only set this bit, and the core clears it, when a SETUP token is received
for this endpoint. If a NAK bit or Global OUT NAK is set along with this bit, the STALL bit
takes priority. Irrespective of this bit’s setting, the core always responds to SETUP data
packets with an ACK handshake.
Bit 20 SNPM: Snoop mode
This bit configures the endpoint to Snoop mode. In Snoop mode, the core does not check
the correctness of OUT packets before transferring them to application memory.
Hardcoded to 2’b00 for control.
1314/1751 RM0090 Rev 19


When either the application or the core sets this bit, the core stops receiving data, even if
there is space in the RxFIFO to accommodate the incoming packet. Irrespective of this bit’s
setting, the core always responds to SETUP data packets with an ACK handshake.
This bit is always set to 1, indicating that a control endpoint 0 is always active in all
configurations and interfaces.
The maximum packet size for control OUT endpoint 0 is the same as what is programmed in
control IN endpoint 0.
00: 64 bytes
01: 32 bytes
10: 16 bytes
11: 8 bytes
OTG_FS device endpoint-x control register (OTG_FS_DOEPCTLx) (x = 1..3,

where x = Endpoint_number)
Address offset for OUT endpoints: 0xB00 + 0x20 * x
The application uses this register to control the behavior of each logical endpoint other than
endpoint 0.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SODDFRM/SD1PID
SD0PID/SEVNFRM
EONUM/DPID
USBAEP
NAKSTS
EPENA
EPTYP
EPDIS
STALL
SNPM
CNAK
SNAK
MPSIZ
Reserved Reserved
rs rs w w w w rw rw rw rw r r rw rw rw rw rw rw rw rw rw rw rw rw

Applies to IN and OUT endpoints.
The application sets this bit to start transmitting data on an endpoint.
RM0090 Rev 19 1315/1751

1380

The application sets this bit to stop transmitting/receiving data on an endpoint, even before
the transfer for that endpoint is complete. The application must wait for the Endpoint
disabled interrupt before treating the endpoint as disabled. The core clears this bit before
setting the Endpoint disabled interrupt. The application must set this bit only if Endpoint
enable is already set for this endpoint.
Applies to interrupt/bulk IN and OUT endpoints only. Writing to this field sets the endpoint
data PID (DPID) field in this register to DATA1.
SODDFRM: Set odd frame
Applies to isochronous IN and OUT endpoints only. Writing to this field sets the Even/Odd
frame (EONUM) field to odd frame.
Applies to interrupt/bulk OUT endpoints only.
Writing to this field sets the endpoint data PID (DPID) field in this register to DATA0.
SEVNFRM: Set even frame
Applies to isochronous OUT endpoints only.
Writing to this field sets the Even/Odd frame (EONUM) field to even frame.
endpoint. The core can also set this bit for OUT endpoints on a Transfer Completed
interrupt, or after a SETUP is received on the endpoint.
Applies to non-control, non-isochronous OUT endpoints only (access type is rw).
The application sets this bit to stall all tokens from the USB host to this endpoint. If a NAK
bit, Global IN NAK, or Global OUT NAK is set along with this bit, the STALL bit takes
priority. Only the application can clear this bit, never the core.
Bit 20 SNPM: Snoop mode
This bit configures the endpoint to Snoop mode. In Snoop mode, the core does not check
the correctness of OUT packets before transferring them to application memory.
This is the transfer type supported by this logical endpoint.
00: Control
01: Isochronous
10: Bulk
11: Interrupt
1316/1751 RM0090 Rev 19


When either the application or the core sets this bit:
The core stops receiving any data on an OUT endpoint, even if there is space in the
RxFIFO to accommodate the incoming packet.
Irrespective of this bit’s setting, the core always responds to SETUP data packets with an
ACK handshake.
Bit 16 EONUM: Even/odd frame
Applies to isochronous IN and OUT endpoints only.
Indicates the frame number in which the core transmits/receives isochronous data for this
endpoint. The application must program the even/odd frame number in which it intends to
transmit/receive isochronous data for this endpoint using the SEVNFRM and SODDFRM
fields in this register.
0: Even frame
1: Odd frame
DPID: Endpoint data PID
Applies to interrupt/bulk OUT endpoints only.
Contains the PID of the packet to be received or transmitted on this endpoint. The
application must program the PID of the first packet to be received or transmitted on this
endpoint, after the endpoint is activated. The application uses the SD0PID register field to
program either DATA0 or DATA1 PID.
0: DATA0
1: DATA1
Indicates whether this endpoint is active in the current configuration and interface. The core
clears this bit for all endpoints (other than EP 0) after detecting a USB reset. After receiving
the SetConfiguration and SetInterface commands, the application must program endpoint
registers accordingly and set this bit.
endpoint. This value is in bytes.
RM0090 Rev 19 1317/1751

1380
OTG_FS device endpoint-x interrupt register (OTG_FS_DIEPINTx) (x = 0..3,

This register indicates the status of an endpoint with respect to USB- and AHB-related
events. It is shown in Figure 394. The application must read this register when the IN
endpoints interrupt bit of the Core interrupt register (IEPINT in OTG_FS_GINTSTS) is set.
Before the application can read this register, it must first read the device all endpoints
interrupt (OTG_FS_DAINT) register to get the exact endpoint number for the Device
endpoint-x interrupt register. The application must clear the appropriate bit in this register to
clear the corresponding bits in the OTG_FS_DAINT and OTG_FS_GINTSTS registers.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PKTDRPSTS
INEPNM
EPDISD
INEPNE
ITTXFE
XFRC
TXFE
TOC
NAK
Reserved
Reserved
Reserved Reserved
rc_ rc_ rc_ rc_ rc_ rc_ rc_ rc_

r
w1 w1 w1 w1 w1 w1 w1 w1

Bit 13 NAK: NAK input
The core generates this interrupt when a NAK is transmitted or received by the device. In
case of isochronous IN endpoints the interrupt gets generated when a zero length packet is
transmitted due to unavailability of data in the Tx FIFO.
Bit 11 PKTDRPSTS: Packet dropped status
This bit indicates to the application that an ISOC OUT packet has been dropped. This bit
does not have an associated mask bit and does not generate an interrupt.
Bit 7 TXFE: Transmit FIFO empty
This interrupt is asserted when the TxFIFO for this endpoint is either half or completely
empty. The half or completely empty status is determined by the TxFIFO Empty Level bit in
the OTG_FS_GAHBCFG register (TXFELVL bit in OTG_FS_GAHBCFG).
Bit 6 INEPNE: IN endpoint NAK effective
This bit can be cleared when the application clears the IN endpoint NAK by writing to the
CNAK bit in OTG_FS_DIEPCTLx.
This interrupt indicates that the core has sampled the NAK bit set (either by the application
or by the core). The interrupt indicates that the IN endpoint NAK bit set by the application
has taken effect in the core.
This interrupt does not guarantee that a NAK handshake is sent on the USB. A STALL bit
takes priority over a NAK bit.
Bit 5 INEPNM: IN token received with EP mismatch.
Indicates that the data in the top of the non-periodic TxFIFO belongs to an endpoint other
than the one for which the IN token was received. This interrupt is asserted on the endpoint
for which the IN token was received.
1318/1751 RM0090 Rev 19

Bit 4 ITTXFE: IN token received when TxFIFO is empty

Applies to non-periodic IN endpoints only.
Indicates that an IN token was received when the associated TxFIFO (periodic/non-periodic)
was empty. This interrupt is asserted on the endpoint for which the IN token was received.
Bit 3 TOC: Timeout condition
Applies only to Control IN endpoints.
Indicates that the core has detected a timeout condition on the USB for the last IN token on
this endpoint.
Bit 1 EPDISD: Endpoint disabled interrupt
This bit indicates that the endpoint is disabled per the application’s request.
Bit 0 XFRC: Transfer completed interrupt
This field indicates that the programmed transfer is complete on the AHB as well as on the
USB, for this endpoint.
OTG_FS device endpoint-x interrupt register (OTG_FS_DOEPINTx) (x = 0..3,

Address offset: 0xB08 + 0x20 * x
This register indicates the status of an endpoint with respect to USB- and AHB-related
events. It is shown in Figure 394. The application must read this register when the OUT
Endpoints Interrupt bit of the OTG_FS_GINTSTS register (OEPINT bit in
OTG_FS_GINTSTS) is set. Before the application can read this register, it must first read
the OTG_FS_DAINT register to get the exact endpoint number for the OTG_FS_DOEPINTx
corresponding bits in the OTG_FS_DAINT and OTG_FS_GINTSTS registers.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
rc_w1 OUTPKTERR
STSPHSRX
OTEPDIS
EPDISD
BERR
XFRC
STUP
NAK
Reserved
Reserved
Reserved Reserved
rc_w1
rc_w1
rc_w1
rc_w1
rc_w1
rc_w1
rc_w1

Bit 13 NAK: NAK input
The core generates this interrupt when a NAK is transmitted or received by the device. In
case of isochronous IN endpoints the interrupt gets generated when a zero length packet is
transmitted due to unavailability of data in the Tx FIFO.
Bit 12 BERR: Babble error interrupt
The core generates this interrupt when babble is received for the endpoint.
RM0090 Rev 19 1319/1751

1380
Bit 8 OUTPKTERR: OUT packet error

This interrupt is asserted when the core detects an overflow or a CRC error for an OUT
packet. This interrupt is valid only when thresholding is enabled.
Bit 5 STSPHSRX: Status phase received for control write
This interrupt is generated only after the core has transferred all the data that the host has
sent during the data phase of a control write transfer, to the system memory buffer. The
interrupt indicates to the application that the host has switched from data phase to the status
phase of a control write transfer. The application can use this interrupt to ACK or STALL the
status phase, after it has decoded the data phase.
Bit 4 OTEPDIS: OUT token received when endpoint disabled
Applies only to control OUT endpoints.
Indicates that an OUT token was received when the endpoint was not yet enabled. This
interrupt is asserted on the endpoint for which the OUT token was received.
Bit 3 STUP: SETUP phase done
Applies to control OUT endpoint only.
Indicates that the SETUP phase for the control endpoint is complete and no more back-to-
back SETUP packets were received for the current control transfer. On this interrupt, the
application can decode the received SETUP data packet.
Bit 1 EPDISD: Endpoint disabled interrupt
This bit indicates that the endpoint is disabled per the application’s request.
Bit 0 XFRC: Transfer completed interrupt
This field indicates that the programmed transfer is complete on the AHB as well as on the
USB, for this endpoint.
OTG_FS device IN endpoint 0 transfer size register (OTG_FS_DIEPTSIZ0)

The application must modify this register before enabling endpoint 0. Once endpoint 0 is
enabled using the endpoint enable bit in the device control endpoint 0 control registers
(EPENA in OTG_FS_DIEPCTL0), the core modifies this register. The application can only
read this register once the core has cleared the Endpoint enable bit.
Nonzero endpoints use the registers for endpoints 1–3.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PKTCNT XFRSIZ
Reserved Reserved
1320/1751 RM0090 Rev 19


Bits 20:19 PKTCNT: Packet count
Indicates the total number of USB packets that constitute the Transfer Size amount of data
for endpoint 0.
This field is decremented every time a packet (maximum size or short packet) is read from
the TxFIFO.
Indicates the transfer size in bytes for endpoint 0. The core interrupts the application only
after it has exhausted the transfer size amount of data. The transfer size can be set to the
maximum packet size of the endpoint, to be interrupted at the end of each packet.
The core decrements this field every time a packet from the external memory is written to
the TxFIFO.
RM0090 Rev 19 1321/1751

1380
OTG_FS device OUT endpoint 0 transfer size register (OTG_FS_DOEPTSIZ0)

Address offset: 0xB10
The application must modify this register before enabling endpoint 0. Once endpoint 0 is
enabled using the Endpoint enable bit in the OTG_FS_DOEPCTL0 registers (EPENA bit in
OTG_FS_DOEPCTL0), the core modifies this register. The application can only read this
register once the core has cleared the Endpoint enable bit.
Nonzero endpoints use the registers for endpoints 1–3.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PKTCNT
Reserved
STUPCNT XFRSIZ
Reserved Reserved

Bits 30:29 STUPCNT: SETUP packet count
This field specifies the number of back-to-back SETUP data packets the endpoint can
receive.
01: 1 packet
10: 2 packets
11: 3 packets
Bit 19 PKTCNT: Packet count
This field is decremented to zero after a packet is written into the RxFIFO.
Indicates the transfer size in bytes for endpoint 0. The core interrupts the application only
after it has exhausted the transfer size amount of data. The transfer size can be set to the
maximum packet size of the endpoint, to be interrupted at the end of each packet.
The core decrements this field every time a packet is read from the RxFIFO and written to
the external memory.
1322/1751 RM0090 Rev 19

OTG_FS device endpoint-x transfer size register (OTG_FS_DIEPTSIZx)

(x = 1..3, where x = Endpoint_number)
The application must modify this register before enabling the endpoint. Once the endpoint is
enabled using the Endpoint enable bit in the OTG_FS_DIEPCTLx registers (EPENA bit in
OTG_FS_DIEPCTLx), the core modifies this register. The application can only read this
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PKTCNT XFRSIZ
Reserved
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 28:19 PKTCNT: Packet count
for this endpoint.
This field is decremented every time a packet (maximum size or short packet) is read from
the TxFIFO.
This field contains the transfer size in bytes for the current endpoint. The core only interrupts
the application after it has exhausted the transfer size amount of data. The transfer size can
be set to the maximum packet size of the endpoint, to be interrupted at the end of each
packet.
The core decrements this field every time a packet from the external memory is written to the
TxFIFO.
RM0090 Rev 19 1323/1751

1380
OTG_FS device IN endpoint transmit FIFO status register

(OTG_FS_DTXFSTSx) (x = 0..3, where x = Endpoint_number)
Address offset for IN endpoints: 0x918 + 0x20 * x
This read-only register contains the free space information for the Device IN endpoint
TxFIFO.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
INEPTFSAV
Reserved

15:0 INEPTFSAV: IN endpoint TxFIFO space available
Indicates the amount of free space available in the Endpoint TxFIFO.
Values are in terms of 32-bit words:
0x0: Endpoint TxFIFO is full
0x1: 1 word available
0x2: 2 words available
0xn: n words available
Others: Reserved
OTG_FS device OUT endpoint-x transfer size register (OTG_FS_DOEPTSIZx)

(x = 1..3, where x = Endpoint_number)
Address offset: 0xB10 + 0x20 * x
The application must modify this register before enabling the endpoint. Once the endpoint is
enabled using Endpoint Enable bit of the OTG_FS_DOEPCTLx registers (EPENA bit in
OTG_FS_DOEPCTLx), the core modifies this register. The application can only read this
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Reserved
RXDPID/S
PKTCNT XFRSIZ
TUPCNT
r/rw r/rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 30:29 RXDPID: Received data PID
Applies to isochronous OUT endpoints only.
This is the data PID received in the last packet for this endpoint.
00: DATA0
01: DATA2
10: DATA1
11: MDATA
1324/1751 RM0090 Rev 19

STUPCNT: SETUP packet count

Applies to control OUT Endpoints only.
This field specifies the number of back-to-back SETUP data packets the endpoint can
receive.
01: 1 packet
10: 2 packets
11: 3 packets
Bit 28:19 PKTCNT: Packet count
for this endpoint.
This field is decremented every time a packet (maximum size or short packet) is written to
the RxFIFO.
This field contains the transfer size in bytes for the current endpoint. The core only interrupts
the application after it has exhausted the transfer size amount of data. The transfer size can
be set to the maximum packet size of the endpoint, to be interrupted at the end of each
packet.
The core decrements this field every time a packet is read from the RxFIFO and written to
the external memory.
34.16.5 OTG_FS power and clock gating control register

(OTG_FS_PCGCCTL)
Address offset: 0xE00
This register is available in host and device modes.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
GATEHCLK
PHYSUSP
STPPCLK
Reserved Reserved
rw rw rw

Bit 4 PHYSUSP: PHY Suspended
Indicates that the PHY has been suspended. This bit is updated once the PHY is suspended
after the application has set the STPPCLK bit (bit 0).
Bit 1 GATEHCLK: Gate HCLK
The application sets this bit to gate HCLK to modules other than the AHB Slave and Master
and wakeup logic when the USB is suspended or the session is not valid. The application
clears this bit when the USB is resumed or a new session starts.
Bit 0 STPPCLK: Stop PHY clock
The application sets this bit to stop the PHY clock when the USB is suspended, the session
is not valid, or the device is disconnected. The application clears this bit when the USB is
resumed or a new session starts.
RM0090 Rev 19 1325/1751

1380
34.16.6 OTG_FS register map

The table below gives the USB OTG register map and reset values.
Table 205. OTG_FS register map and reset values
Offset Register 31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
HSHNPEN
HNGSCS
DHNPEN
SRQSCS
CIDSTS
HNPRQ
BSVLD
ASVLD
DBCT
OTG_FS_GOTG
SRQ
CTL
Reset value 0 0 0 1 0 0 0 0 0 0
HNSSCHG
ADTOCHG
SRSSCHG
DBCDNE
HNGDET
SEDET
OTG_FS_GOTG
INT
0x004 Reserved Reserved Reserved Res.
PTXFELVL
GINTMSK
TXFELVL
OTG_FS_GAHB
0x008 CFG Reserved Reserved
Reset value 0 0 0
SRPCAP
CTXPKT
PHYSEL
FDMOD
FHMOD
HNPCA
OTG_FS_GUSB
Reserved
Reserved
TRDT TOCAL
CFG
0x00C Reserved
Reset value 0 0 0 0 0 1 0 1 0 0 0 0 0
RXFFLSH
TXFFLSH
AHBIDL
HSRST
CSRST
FCRST
OTG_FS_GRST
Reserved
TXFNUM
CTL
0x010 Reserved
Reset value 1 0 0 0 0 0 0 0 0 0 0
IPXFR/INCOMPISOOUT
GONAKEFF
ENUMDNE
GINAKEFF
USBSUSP
CIDSCHG
ISOODRP
HPRTINT
IISOIXFR
USBRST
DISCINT
WKUINT
NPTXFE
SRQINT
OTGINT
OEPINT
RXFLVL
ESUSP
IEPINT
PTXFE
HCINT
CMOD
EOPF
MMIS
OTG_FS_GINTS
Reserved
Reserved
Reserved
Reserved
SOF
TS
0x014
Reset value 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0
IPXFRM/IISOOXFRM
GONAKEFFM
ENUMDNEM
GINAKEFFM
USBSUSPM
ISOODRPM
CIDSCHGM
IISOIXFRM
NPTXFEM
RXFLVLM
ESUSPM
USBRST
PTXFEM
DISCINT
OTGINT
OEPINT
EOPFM
MMISM
SRQIM
IEPINT
PRTIM
SOFM
WUIM
HCIM
OTG_FS_GINT
Reserved
Reserved
Reserved
Reserved
Reserved
MSK
0x018
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
OTG_FS_GRXS
TSR (host PKTSTS DPID BCNT CHNUM
mode) Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x01C
OTG_FS_GRXS
TSR (Device FRMNUM PKTSTS DPID BCNT EPNUM
mode) Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1326/1751 RM0090 Rev 19

Table 205. OTG_FS register map and reset values (continued)
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
OTG_FS_GRXS
TSR (host PKTSTS DPID BCNT CHNUM
mode) Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x020
OTG_FS_GRXS
TSPR (Device FRMNUM PKTSTS DPID BCNT EPNUM
mode) Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
OTG_FS_GRXF
RXFD
SIZ
0x024 Reserved
Reset value 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0
OTG_FS_HNPT
XFSIZ/
NPTXFD/TX0FD NPTXFSA/TX0FSA
OTG_FS_DIEPT
0x028
XF0
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0
OTG_FS_HNPT
NPTXQTOP NPTQXSAV NPTXFSAV
XSTS
Res.
0x02C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0
NOVBUSSENS
SOFOUTEN
VBUSBSEN
VBUSASEN
.PWRDWN
OTG_FS_
Reserved
GCCFG
OTG_FS_CID PRODUCT_ID
0x03C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0
OTG_FS_HPTX
PTXFSIZ PTXSA
0x100 FSIZ
Reset value 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0
OTG_FS_DIEPT
INEPTXFD INEPTXSA
XF1
0x104
Reset value 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0
OTG_FS_DIEPT
INEPTXFD INEPTXSA
XF2
0x108
Reset value 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0
OTG_FS_DIEPT
INEPTXFD INEPTXSA
XF3
0x10C
Reset value 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0
FSLSPCS
FSLSS
OTG_FS_HCFG
0x400 Reserved
Reset value 0 0 0
OTG_FS_HFIR FRIVL
0x404 Reserved
Reset value 1 1 1 0 1 0 1 0 0 1 1 0 0 0 0 0
OTG_FS_HFNU
FTREM FRNUM
M
0x408
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1
RM0090 Rev 19 1327/1751

1380
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
OTG_FS_HPTX
PTXQTOP PTXQSAV PTXFSAVL
STS
0x410
Reset value 0 0 0 0 0 0 0 0 Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
OTG_FS_HAINT HAINT
0x414 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
OTG_FS_HAINT
HAINTM
MSK
0x418 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
POCCHNG
PENCHNG
PSUSP
PCDET
PCSTS
PLSTS
PPWR
POCA
PRES
PENA
PRST
Reserved
OTG_FS_HPRT PSPD PTCTL
0x440 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ODDFRM
CHENA
EPTYP
LSDEV
CHDIS
EPDIR
OTG_FS_HCCH
Reserved
AR0
0x500
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ODDFRM
CHENA
EPTYP
LSDEV
CHDIS
OTG_FS_HCCH EPDIR
Reserved

AR1
0x520
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ODDFRM
CHENA
EPTYP
LSDEV
CHDIS
EPDIR
OTG_FS_HCCH
Reserved

AR2
0x540
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ODDFRM
CHENA
EPTYP
LSDEV
CHDIS
EPDIR
OTG_FS_HCCH
Reserved

AR3
0x560
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ODDFRM
CHENA
EPTYP
LSDEV
CHDIS
EPDIR
OTG_FS_HCCH
Reserved

AR4
0x580
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ODDFRM
CHENA
EPTYP
LSDEV
CHDIS
EPDIR
OTG_FS_HCCH
Reserved

0x5A0 AR5
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ODDFRM
CHENA
EPTYP
LSDEV
CHDIS
EPDIR
OTG_FS_HCCH
Reserved

AR6
0x5C0
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ODDFRM
CHENA
EPTYP
LSDEV
CHDIS
EPDIR
OTG_FS_HCCH
Reserved

0x5E0 AR7
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
FRMOR
DTERR
BBERR
TXERR
STALL
XFRC
OTG_FS_HCINT
CHH
Reserved
Reserved
ACK
NAK
0
0x508 Reserved
Reset value 0 0 0 0 0 0 0 0 0
1328/1751 RM0090 Rev 19

0x588
0x568
0x548
0x528
0x5E8
0x5A8
0x58C
0x56C
0x54C
0x52C
0x50C
0x5C8
Offset
RM0090
7
6
5
4
3
2
1
MSK4
MSK3
MSK2
MSK1
MSK0
Register
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
OTG_FS_HCINT
OTG_FS_HCINT
OTG_FS_HCINT
OTG_FS_HCINT
OTG_FS_HCINT
OTG_FS_HCINT
OTG_FS_HCINT
OTG_FS_HCINT
OTG_FS_HCINT
OTG_FS_HCINT
OTG_FS_HCINT
OTG_FS_HCINT
31
30
29
28
27
26
25
24
23
22
21
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
20
19
18
RM0090 Rev 19
17
16
15
14
13
12
11
0
0
0
0
0
0
0
0
0
0
0
0
DTERRM DTERRM DTERRM DTERRM DTERRM DTERR DTERR DTERR DTERR DTERR DTERR DTERR 10
0
0
0
0
0
0
0
0
0
0
0
0
FRMORM FRMORM FRMORM FRMORM FRMORM FRMOR FRMOR FRMOR FRMOR FRMOR FRMOR FRMOR 9
0
0
0
0
0
0
0
0
0
0
0
0
BBERRM BBERRM BBERRM BBERRM BBERRM BBERR BBERR BBERR BBERR BBERR BBERR BBERR 8
0
0
0
0
0
0
0
0
0
0
0
0
TXERRM TXERRM TXERRM TXERRM TXERRM TXERR TXERR TXERR TXERR TXERR TXERR TXERR 7
Reserved Reserved Reserved Reserved Reserved Reserved Reserved Reserved Reserved Reserved Reserved Reserved 6
0
0
0
0
0
0
0
0
0
0
0
0
ACKM ACKM ACKM ACKM ACKM ACK ACK ACK ACK ACK ACK ACK 5
0
0
0
0
0
0
0
0
0
0
0
0
NAKM NAKM NAKM NAKM NAKM NAK NAK NAK NAK NAK NAK NAK 4
0
0
0
0
0
0
0
0
0
0
0
0
STALLM STALLM STALLM STALLM STALLM STALL STALL STALL STALL STALL STALL STALL 3
Reserved Reserved Reserved Reserved Reserved Reserved Reserved Reserved Reserved Reserved Reserved Reserved 2
0
0
0
0
0
0
0
0
0
0
0
0
CHHM CHHM CHHM CHHM CHHM CHH CHH CHH CHH CHH CHH CHH 1
0
0
0
0
0
0
0
0
0
0
0
0
XFRCM XFRCM XFRCM XFRCM XFRCM XFRC XFRC XFRC XFRC XFRC XFRC XFRC 0
USB on-the-go full-speed (OTG_FS)
1329/1751
1380
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
FRMORM
DTERRM
BBERRM
TXERRM
STALLM
XFRCM
CHHM
ACKM
NAKM
OTG_FS_HCINT
Reserved
Reserved
MSK5
0x5AC Reserved
Reset value 0 0 0 0 0 0 0 0 0
FRMORM
DTERRM
BBERRM
TXERRM
STALLM
XFRCM
CHHM
ACKM
NAKM
OTG_FS_HCINT
Reserved
Reserved
MSK6
0x5CC Reserved
Reset value 0 0 0 0 0 0 0 0 0
FRMORM
DTERRM
BBERRM
TXERRM
STALLM
XFRCM
CHHM
ACKM
NAKM
OTG_FS_HCINT
Reserved
Reserved
MSK7
0x5EC Reserved
Reset value 0 0 0 0 0 0 0 0 0
OTG_FS_HCTSI
Reserved
DPID PKTCNT XFRSIZ

Z0
0x510
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
OTG_FS_HCTSI
Reserved
DPID PKTCNT XFRSIZ

Z1
0x530
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
OTG_FS_HCTSI
Reserved
DPID PKTCNT XFRSIZ

Z2
0x550
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
OTG_FS_HCTSI
Reserved
DPID PKTCNT XFRSIZ

0x570 Z3
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
OTG_FS_HCTSI
Reserved
DPID PKTCNT XFRSIZ

Z4
0x590
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
OTG_FS_HCTSI
Reserved
DPID PKTCNT XFRSIZ

Z5
0x5B0
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
OTG_FS_HCTSI
Reserved
DPID PKTCNT XFRSIZ

Z6
0x5D0
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
OTG_FS_HCTSI
Reserved
DPID PKTCNT XFRSIZ

Z7
0x5F0
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
NZLSOHSK
PFIVL
DSPD
Reserved
DAD
OTG_FS_DCFG
0x800 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0
POPRGDNE
CGONAK
SGONAK
GONSTS
RWUSIG
CGINAK
SGINAK
GINSTS
TCTL
SDIS
OTG_FS_DCTL
0x804 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0
1330/1751 RM0090 Rev 19

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
ENUMSPD
SUSPSTS
EERR
OTG_FS_DSTS FNSOF
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ITTXFEMSK
INEPNMM
INEPNEM
XFRCM
NAKM
EPDM
OTG_FS_DIEPM
TOM
Reserved
SK
OUTPKTERRM
STSPHSRXM
OTEPDM
NAKMSK
BERRM
XFRCM
STUPM
EPDM
OTG_FS_DOEP
Reserved
Reserved
Reserved
MSK
0x814 Reserved
Reset value 0 0 0 0 0 0 0 0
OTG_FS_DAINT OEPINT IEPINT

0x818
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
OTG_FS_DAINT
OEPM IEPM
0x81C MSK
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
OTG_FS_DVBU
VBUSDT
0x828 SDIS
Reserved
Reset value 0 0 0 1 0 1 1 1 1 1 0 1 0 1 1 1
OTG_FS_DVBU
DVBUSP
SPULSE
0x82C Reserved
Reset value 0 1 0 1 1 0 1 1 1 0 0 0
OTG_FS_DIEPE
INEPTXFEM
MPMSK
0x834 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
USBAEP
NAKSTS
EPENA
EPDIS
STALL
CNAK
SNAK
OTG_FS_DIEPC EPTY MPSI

Reserved
Reserved
Reserved
TXFNUM
TL0 P Z
0x900 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0
TG_FS_DTXFST
INEPTFSAV
0x918 S0 Reserved
Reset value 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0
SODDFRM/SD1PID
SD0PID/SEVNFRM
EONUM/DPID
USBAEP
NAKSTS
EPENA
EPTYP
EPDIS
STALL
CNAK
SNAK
OTG_FS_DIEPC
Reserved
TXFNUM MPSIZ
TL1
0x920 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
TG_FS_DTXFST
INEPTFSAV
S1
0x938 Reserved
Reset value 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0
RM0090 Rev 19 1331/1751

1380
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
SD0PID/SEVNFRM
EONUM/DPID
SODDFRM
USBAEP
NAKSTS
EPENA
EPTYP
EPDIS
STALL
CNAK
SNAK
OTG_FS_DIEPC
Reserved
TXFNUM MPSIZ
TL2
0x940 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
TG_FS_DTXFST
INEPTFSAV
S2
0x958 Reserved
Reset value 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0
SD0PID/SEVNFRM
EONUM/DPID
SODDFRM
USBAEP
NAKSTS
EPENA
EPTYP
EPDIS
STALL
CNAK
SNAK
OTG_FS_DIEPC
Reserved
TXFNUM MPSIZ
TL3
0x960 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
TG_FS_DTXFST
INEPTFSAV
S3
0x978 Reserved
Reset value 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0
USBAEP
NAKSTS
EPENA
EPDIS
STALL
SNPM
CNAK
SNAK
OTG_FS_DOEP EPTY MPSI

Reserved
Reserved
0xB00 CTL0 Reserved P Reserved

Z
Reset value 0 0 0 0 0 0 0 0 0 1 0 0
SD0PID/SEVNFRM
EONUM/DPID
SODDFRM
USBAEP
NAKSTS
EPENA
EPTYP
EPDIS
STALL
SNPM
CNAK
SNAK
OTG_FS_DOEP
MPSIZ
CTL1
0xB20 Reserved Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
SD0PID/SEVNFRM
EONUM/DPID
SODDFRM
USBAEP
NAKSTS
EPENA
EPTYP
EPDIS
STALL
SNPM
CNAK
SNAK
OTG_FS_DOEP
MPSIZ
CTL2
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
SD0PID/SEVNFRM
EONUM/DPID
SODDFRM
USBAEP
NAKSTS
EPENA
EPTYP
EPDIS
STALL
SNPM
CNAK
SNAK
OTG_FS_DOEP
MPSIZ
CTL3
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
PKTDRPSTS
INEPNM
EPDISD
INEPNE
ITTXFE
XFRC
TXFE
OTG_FS_DIEPI
Reserved
Reserved
Reserved
TOC
NAK
0x908 NT0 Reserved
Reset value 0 0 1 0 0 0 0 0 0
1332/1751 RM0090 Rev 19

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
PKTDRPSTS
INEPNM
EPDISD
INEPNE
ITTXFE
XFRC
TXFE
OTG_FS_DIEPI
Reserved
Reserved
Reserved
TOC
NAK
NT1
0x928 Reserved
Reset value 0 0 1 0 0 0 0 0 0
PKTDRPSTS
INEPNM
INEPNE
EPDISD
ITTXFE
XFRC
TXFE
OTG_FS_DIEPI
Reserved
Reserved
Reserved
TOC
NAK
NT2
0x948 Reserved
Reset value 0 0 1 0 0 0 0 0 0
PKTDRPSTS
INEPNM
EPDISD
INEPNE
ITTXFE
XFRC
TXFE
OTG_FS_DIEPI
Reserved
Reserved
Reserved
TOC
NAK
0x968 NT3 Reserved
Reset value 0 0 1 0 0 0 0 0 0
OUTPKTERR
STSPHSRX
OTEPDIS
EPDISD
BERR
XFRC
STUP
OTG_FS_DOEPI
Reserved
Reserved
Reserved
NAK
NT0
0xB08 Reserved
Reset value 0 0 0 0 0 0 0 0
OUTPKTERR
STSPHSRX
OTEPDIS
EPDISD
BERR
XFRC
STUP
OTG_FS_DOEPI
Reserved
Reserved
Reserved
NAK
NT1
0xB28 Reserved
Reset value 0 0 0 0 0 0 0 0
OUTPKTERR
STSPHSRX
OTEPDIS
EPDISD
BERR
XFRC
STUP
OTG_FS_DOEPI
Reserved
Reserved
Reserved
NAK
NT2
0xB48 Reserved
Reset value 0 0 0 0 0 0 0 0
OUTPKTERR
STSPHSRX
OTEPDIS
EPDISD
BERR
XFRC
STUP
OTG_FS_DOEPI
Reserved
Reserved
Reserved
NAK
NT3
0xB68 Reserved
Reset value 0 0 0 0 0 0 0 0
OTG_FS_DIEPT PKTC
XFRSIZ
SIZ0 NT
Reset value 0 0 0 0 0 0 0 0 0
OTG_FS_DIEPT
PKTCNT XFRSIZ
SIZ1
0x930 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
OTG_FS_DIEPT
PKTCNT XFRSIZ
SIZ2
0x950 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
OTG_FS_DIEPT
PKTCNT XFRSIZ
SIZ3
0x970 Reserved
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
RM0090 Rev 19 1333/1751

1380
Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
10
11
9
8
7
6
5
4
3
2
1
0
PKTCNT
OTG_FS_DOEP STUP
Reserved
XFRSIZ
TSIZ0 CNT
Reset value 0 0 0 0 0 0 0 0 0 0
STUPCNT
RXDPID/
OTG_FS_DOEP
Reserved
PKTCNT XFRSIZ
0xB30 TSIZ1
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
STUPCNT
RXDPID/
OTG_FS_DOEP
Reserved
PKTCNT XFRSIZ
TSIZ2
0xB50
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
STUPCNT
RXDPID/
OTG_FS_DOEP
Reserved
PKTCNT XFRSIZ
TSIZ3
0xB70
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
GATEHCLK
PHYSUSP
STPPCLK
OTG_FS_PCGC
Reserved
CTL
0xE00 Reserved
Reset value
1334/1751 RM0090 Rev 19

34.17 OTG_FS programming model
34.17.1 Core initialization

The application must perform the core initialization sequence. If the cable is connected
during power-up, the current mode of operation bit in the OTG_FS_GINTSTS (CMOD bit in
OTG_FS_GINTSTS) reflects the mode. The OTG_FS controller enters host mode when an
“A” plug is connected or device mode when a “B” plug is connected.
This section explains the initialization of the OTG_FS controller after power-on. The
application must follow the initialization sequence irrespective of host or device mode
operation. All core global registers are initialized according to the core’s configuration:
1. Program the following fields in the OTG_FS_GAHBCFG register:
– Global interrupt mask bit GINTMSK = 1
– RxFIFO non-empty (RXFLVL bit in OTG_FS_GINTSTS)
– Periodic TxFIFO empty level
2. Program the following fields in the OTG_FS_GUSBCFG register:
– HNP capable bit
– SRP capable bit
– FS timeout calibration field
– USB turnaround time field
3. The software must unmask the following bits in the OTG_FS_GINTMSK register:
OTG interrupt mask
Mode mismatch interrupt mask
4. The software can read the CMOD bit in OTG_FS_GINTSTS to determine whether the
OTG_FS controller is operating in host or device mode.
RM0090 Rev 19 1335/1751

1380
34.17.2 Host initialization

To initialize the core as host, the application must perform the following steps:
1. Program the HPRTINT in the OTG_FS_GINTMSK register to unmask
2. Program the OTG_FS_HCFG register to select full-speed host
3. Program the PPWR bit in OTG_FS_HPRT to 1. This drives VBUS on the USB.
4. Wait for the PCDET interrupt in OTG_FS_HPRT0. This indicates that a device is
connecting to the port.
5. Program the PRST bit in OTG_FS_HPRT to 1. This starts the reset process.
6. Wait at least 10 ms for the reset process to complete.
7. Program the PRST bit in OTG_FS_HPRT to 0.
8. Wait for the PENCHNG interrupt in OTG_FS_HPRT.
9. Read the PSPD bit in OTG_FS_HPRT to get the enumerated speed.
10. Program the HFIR register with a value corresponding to the selected PHY clock 1
11. Program the FSLSPCS field in the OTG_FS_HCFG register following the speed of the
device detected in step 9. If FSLSPCS has been changed a port reset must be
performed.
12. Program the OTG_FS_GRXFSIZ register to select the size of the receive FIFO.
13. Program the OTG_FS_HNPTXFSIZ register to select the size and the start address of
the Non-periodic transmit FIFO for non-periodic transactions.
14. Program the OTG_FS_HPTXFSIZ register to select the size and start address of the
periodic transmit FIFO for periodic transactions.
To communicate with devices, the system software must initialize and enable at least one
channel.
34.17.3 Device initialization

The application must perform the following steps to initialize the core as a device on power-
up or after a mode change from host to device.
1. Program the following fields in the OTG_FS_DCFG register:
– Device speed
– Non-zero-length status OUT handshake
2. Program the OTG_FS_GINTMSK register to unmask the following interrupts:
– USB reset
– Enumeration done
– Early suspend
– USB suspend
– SOF
3. Program the VBUSBSEN bit in the OTG_FS_GCCFG register to enable VBUS sensing
in “B” device mode and supply the 5 volts across the pull-up resistor on the DP line.
4. Wait for the USBRST interrupt in OTG_FS_GINTSTS. It indicates that a reset has been
detected on the USB that lasts for about 10 ms on receiving this interrupt.
Wait for the ENUMDNE interrupt in OTG_FS_GINTSTS. This interrupt indicates the end of
reset on the USB. On receiving this interrupt, the application must read the OTG_FS_DSTS
1336/1751 RM0090 Rev 19

register to determine the enumeration speed and perform the steps listed in Endpoint
initialization on enumeration completion on page 1353.
At this point, the device is ready to accept SOF packets and perform control transfers on
control endpoint 0.
34.17.4 Host programming model

Channel initialization
The application must initialize one or more channels before it can communicate with
connected devices. To initialize and enable a channel, the application must perform the
following steps:
1. Program the OTG_FS_GINTMSK register to unmask the following:
2. Channel interrupt
– Non-periodic transmit FIFO empty for OUT transactions (applicable when
operating in pipelined transaction-level with the packet count field programmed
with more than one).
– Non-periodic transmit FIFO half-empty for OUT transactions (applicable when
operating in pipelined transaction-level with the packet count field programmed
with more than one).
3. Program the OTG_FS_HAINTMSK register to unmask the selected channels’
interrupts.
4. Program the OTG_FS_HCINTMSK register to unmask the transaction-related
interrupts of interest given in the host channel interrupt register.
5. Program the selected channel’s OTG_FS_HCTSIZx register with the total transfer size,
in bytes, and the expected number of packets, including short packets. The application
must program the PID field with the initial data PID (to be used on the first OUT
transaction or to be expected from the first IN transaction).
6. Program the OTG_FS_HCCHARx register of the selected channel with the device’s
endpoint characteristics, such as type, speed, direction, and so forth. (The channel can
be enabled by setting the channel enable bit to 1 only when the application is ready to
transmit or receive any packet).
Halting a channel
The application can disable any channel by programming the OTG_FS_HCCHARx register
with the CHDIS and CHENA bits set to 1. This enables the OTG_FS host to flush the posted
requests (if any) and generates a channel halted interrupt. The application must wait for the
CHH interrupt in OTG_FS_HCINTx before reallocating the channel for other transactions.
The OTG_FS host does not interrupt the transaction that has already been started on the
USB.
Before disabling a channel, the application must ensure that there is at least one free space
available in the non-periodic request queue (when disabling a non-periodic channel) or the
periodic request queue (when disabling a periodic channel). The application can simply
flush the posted requests when the Request queue is full (before disabling the channel), by
programming the OTG_FS_HCCHARx register with the CHDIS bit set to 1, and the CHENA
bit cleared to 0.
The application is expected to disable a channel on any of the following conditions:
RM0090 Rev 19 1337/1751

1380
1. When an STALL, TXERR, BBERR or DTERR interrupt in OTG_FS_HCINTx is received

for an IN or OUT channel. The application must be able to receive other interrupts
(DTERR, Nak, Data, TXERR) for the same channel before receiving the halt.
2. When a DISCINT (Disconnect Device) interrupt in OTG_FS_GINTSTS is received.
(The application is expected to disable all enabled channels).
3. When the application aborts a transfer before normal completion.
Operational model
The application must initialize a channel before communicating to the connected device.
This section explains the sequence of operation to be performed for different types of USB
transactions.
• Writing the transmit FIFO
The OTG_FS host automatically writes an entry (OUT request) to the periodic/non-
periodic request queue, along with the last word write of a packet. The application must
ensure that at least one free space is available in the periodic/non-periodic request
queue before starting to write to the transmit FIFO. The application must always write
to the transmit FIFO in words. If the packet size is non-word aligned, the application
must use padding. The OTG_FS host determines the actual packet size based on the
programmed maximum packet size and transfer size.
Figure 396. Transmit FIFO write task
Start
Read GNPTXSTS/HPTXFSIZ
registers for available FIFO
and queue spaces
Wait for NPTXFE/PTXFE 1 MPS or

interrupt in LPS FIFO space
No Yes
OTG_FS_GINTSTS available?
Yes
Write 1 packet
data to
transmit FIFO
More packets
to send?
No
MPS: Maximum packet size

LPS: Last packet size
Done
ai15673b
• Reading the receive FIFO
1338/1751 RM0090 Rev 19

The application must ignore all packet statuses other than IN data packet (bx0010).
Figure 397. Receive FIFO read task
Start
No
RXFLVL
interrupt ?
Yes
Unmask RXFLVL Mask RXFLVL Unmask RXFLVL

interrupt interrupt interrupt
Read the received

Read
packet from the
Receive FIFO OTG_FS_GRXSTSP
PKTSTS
No
0b0010?
No
Yes
Yes
BCNT > 0?
ai15674
• Bulk and control OUT/SETUP transactions

A typical bulk or control OUT/SETUP pipelined transaction-level operation is shown in
Figure 398. See channel 1 (ch_1). Two bulk OUT packets are transmitted. A control
SETUP transaction operates in the same way but has only one packet. The
assumptions are:
– The application is attempting to send two maximum-packet-size packets (transfer
size = 1, 024 bytes).
– The non-periodic transmit FIFO can hold two packets (128 bytes for FS).
– The non-periodic request queue depth = 4.
• Normal bulk and control OUT/SETUP operations
The sequence of operations in (channel 1) is as follows:
a) Initialize channel 1
b) Write the first packet for channel 1
c) Along with the last word write, the core writes an entry to the non-periodic request
queue
d) As soon as the non-periodic queue becomes non-empty, the core attempts to
send an OUT token in the current frame
e) Write the second (last) packet for channel 1
f) The core generates the XFRC interrupt as soon as the last transaction is
completed successfully
RM0090 Rev 19 1339/1751

1380
g) In response to the XFRC interrupt, de-allocate the channel for other transfers
h) Handling non-ACK responses
Figure 398. Normal bulk/control OUT/SETUP and bulk/control IN transactions
Application AHB Host USB Device

1 init_reg(ch _1)
init _reg(ch_2) Non-Periodic Request
1 Queue
write_tx_fifo 4 Assume that this queue
(ch_1) 1 3 can hold 4 entries.
MPS
set _ch_en 2
(ch _2)
ch_1
2 write_tx_fifo
(ch_1) 1 ch_2
MPS OU T
set _ch_en 5 ch_1
(ch _2)
ch_2 D AT A0
MPS
3
AC K
set _ch_en
(ch _2) IN
4
D AT A0
5
RXFLVL interrupt
ACK
1 ch_1
read_rx_sts
read_rx_fifo
MPS
O UT
ch_2
set _ch_en
(ch _2) ch_2 D AT A1
MPS
ch_2
7 ACK
XFRC interrupt
6
De-allocate IN
(ch_1)
D AT A1
RXFLVL interrupt
1 6 ACK
read_rx_stsre
MPS
ad_rx_fifo
RXFLVL interrupt 8 ch_2

read_rx_sts
7
XFRC interrupt
Disable
(ch _2) 9
RXFLVL interrupt
read_rx_sts 11 10
CHH interrupt
r
De-allocate 12
(ch _2) 13
ai15675
The channel-specific interrupt service routine for bulk and control OUT/SETUP
transactions is shown in the following code samples.
• Interrupt service routine for bulk/control OUT/SETUP and bulk/control IN
transactions
a) Bulk/Control OUT/SETUP
Unmask (NAK/TXERR/STALL/XFRC)
if (XFRC)
{
1340/1751 RM0090 Rev 19

Reset Error Count

Mask ACK
De-allocate Channel
}
else if (STALL)
{
Transfer Done = 1
Unmask CHH
Disable Channel
}
else if (NAK or TXERR )
{
Rewind Buffer Pointers
Unmask CHH
Disable Channel
if (TXERR)
{
Increment Error Count
Unmask ACK
}
else
{
Reset Error Count
}
}
else if (CHH)
{
Mask CHH
if (Transfer Done or (Error_count == 3))
{
De-allocate Channel
}
else
{
Re-initialize Channel
}
}
else if (ACK)
{
Reset Error Count
Mask ACK
}
The application is expected to write the data packets into the transmit FIFO as and
when the space is available in the transmit FIFO and the Request queue. The
application can make use of the NPTXFE interrupt in OTG_FS_GINTSTS to find the
transmit FIFO space.
b) Bulk/Control IN
Unmask (TXERR/XFRC/BBERR/STALL/DTERR)
if (XFRC)
{
Reset Error Count
RM0090 Rev 19 1341/1751

1380
Unmask CHH
Disable Channel
Reset Error Count
Mask ACK
}
else if (TXERR or BBERR or STALL)
{
Unmask CHH
Disable Channel
if (TXERR)
{
Unmask ACK
}
}
else if (CHH)
{
Mask CHH
{
De-allocate Channel
}
else
{
}
}
else if (ACK)
{
Reset Error Count
Mask ACK
}
else if (DTERR)
{
Reset Error Count
}
The application is expected to write the requests as and when the Request queue space is
available and until the XFRC interrupt is received.
• Bulk and control IN transactions
A typical bulk or control IN pipelined transaction-level operation is shown in Figure 399.
See channel 2 (ch_2). The assumptions are:
– The application is attempting to receive two maximum-packet-size packets
(transfer size = 1 024 bytes).
– The receive FIFO can contain at least one maximum-packet-size packet and two
status words per packet (72 bytes for FS).
– The non-periodic request queue depth = 4.
1342/1751 RM0090 Rev 19

Figure 399. Bulk/control IN transactions

1 init_reg(ch _1)
init _reg(ch_2) Non-Periodic Request
1 Queue
write_tx_fifo 4 Assume that this queue
(ch_1) 1 3 can hold 4 entries.
MPS
set _ch_en 2
(ch _2)
ch_1
2 write_tx_fifo
(ch_1) 1 ch_2
MPS OU T
set _ch_en 5 ch_1
(ch _2)
ch_2 D AT A0
MPS
3
AC K
set _ch_en
(ch _2) IN
4
D AT A0
5
RXFLVL interrupt
ACK
1 ch_1
read_rx_sts
read_rx_fifo
MPS
O UT
ch_2
set _ch_en
(ch _2) ch_2 D AT A1
MPS
ch_2
7 ACK
XFRC interrupt
6
De-allocate IN
(ch_1)
D AT A1
RXFLVL interrupt
1 6 ACK
read_rx_stsre
MPS
ad_rx_fifo
RXFLVL interrupt 8 ch_2

read_rx_sts
7
XFRC interrupt
Disable
(ch _2) 9
RXFLVL interrupt
read_rx_sts 11 10
CHH interrupt
r
De-allocate 12
(ch _2) 13
ai15675
The sequence of operations is as follows:

a) Initialize channel 2.
b) Set the CHENA bit in HCCHAR2 to write an IN request to the non-periodic request
queue.
c) The core attempts to send an IN token after completing the current OUT
transaction.
d) The core generates an RXFLVL interrupt as soon as the received packet is written
to the receive FIFO.
e) In response to the RXFLVL interrupt, mask the RXFLVL interrupt and read the
received packet status to determine the number of bytes received, then read the
receive FIFO accordingly. Following this, unmask the RXFLVL interrupt.
RM0090 Rev 19 1343/1751

1380
f) The core generates the RXFLVL interrupt for the transfer completion status entry
in the receive FIFO.
g) The application must read and ignore the receive packet status when the receive
packet status is not an IN data packet (PKTSTS in GRXSTSR ≠ 0b0010).
h) The core generates the XFRC interrupt as soon as the receive packet status is
read.
i) In response to the XFRC interrupt, disable the channel and stop writing the
OTG_FS_HCCHAR2 register for further requests. The core writes a channel
disable request to the non-periodic request queue as soon as the
OTG_FS_HCCHAR2 register is written.
j) The core generates the RXFLVL interrupt as soon as the halt status is written to
the receive FIFO.
k) Read and ignore the receive packet status.
l) The core generates a CHH interrupt as soon as the halt status is popped from the
receive FIFO.
m) In response to the CHH interrupt, de-allocate the channel for other transfers.
n) Handling non-ACK responses
• Control transactions
Setup, Data, and Status stages of a control transfer must be performed as three
separate transfers. Setup-, Data- or Status-stage OUT transactions are performed
similarly to the bulk OUT transactions explained previously. Data- or Status-stage IN
transactions are performed similarly to the bulk IN transactions explained previously.
For all three stages, the application is expected to set the EPTYP field in
OTG_FS_HCCHAR1 to Control. During the Setup stage, the application is expected to
set the PID field in OTG_FS_HCTSIZ1 to SETUP.
• Interrupt OUT transactions
A typical interrupt OUT operation is shown in Figure 400. The assumptions are:
– The application is attempting to send one packet in every frame (up to 1 maximum
packet size), starting with the odd frame (transfer size = 1 024 bytes)
– The periodic transmit FIFO can hold one packet (1 KB)
– Periodic request queue depth = 4
a) Initialize and enable channel 1. The application must set the ODDFRM bit in
OTG_FS_HCCHAR1.
b) Write the first packet for channel 1.
c) Along with the last word write of each packet, the OTG_FS host writes an entry to
the periodic request queue.
d) The OTG_FS host attempts to send an OUT token in the next (odd) frame.
e) The OTG_FS host generates an XFRC interrupt as soon as the last packet is
transmitted successfully.
f) In response to the XFRC interrupt, reinitialize the channel for the next transfer.
1344/1751 RM0090 Rev 19

Figure 400. Normal interrupt OUT/IN transactions

1 init _reg(ch_1)
init_reg(ch _2) Periodic Request Queue
1 write_tx_fifo 3
Assume that this queue
1 can hold 4 entries.
(ch_1)
MPS
set_ch_en 2
(ch_2) ch_1 4
2 ch_2
3
OU T Odd
(micro)
DATA0
frame
M PS
5
6 XFRC interrupt ACK
init _reg(ch_1) 4 IN
write_tx_fifo
(ch_1) 1 5
MPS DATA0
RXFLVL interrupt
ACK
read_rx_sts
read_rx_fifo
1
6 MPS
RXFLVL interrupt
read_rx_sts ch_1
7 8
XFRC interrupt
ch_2
init_reg(ch _2)
9
set_ch_en
(ch_2)
Even
OU T (micro)
XFRC interrupt frame
DATA1
init _reg(ch_1)
MPS
write_tx_fifo 1
ACK
(ch_1) MPS
IN
DATA1
ai15676
• Interrupt service routine for interrupt OUT/IN transactions

a) Interrupt OUT
Unmask (NAK/TXERR/STALL/XFRC/FRMOR)
if (XFRC)
{
Reset Error Count
Mask ACK
De-allocate Channel
}
else
if (STALL or FRMOR)
{
Mask ACK
Unmask CHH
RM0090 Rev 19 1345/1751

1380
Disable Channel
if (STALL)
{
Transfer Done = 1
}
}
else
if (NAK or TXERR)
{
Rewind Buffer Pointers
Reset Error Count
Mask ACK
Unmask CHH
Disable Channel
}
else
if (CHH)
{
Mask CHH
{
De-allocate Channel
}
else
{
Re-initialize Channel (in next b_interval - 1 Frame)
}
}
else
if (ACK)
{
Reset Error Count
Mask ACK
}
The application uses the NPTXFE interrupt in OTG_FS_GINTSTS to find the transmit
FIFO space.
b) Interrupt IN
Unmask (NAK/TXERR/XFRC/BBERR/STALL/FRMOR/DTERR)
if (XFRC)
{
Reset Error Count
Mask ACK
if (OTG_FS_HCTSIZx.PKTCNT == 0)
{
De-allocate Channel
}
else
{
Transfer Done = 1
Unmask CHH
Disable Channel
1346/1751 RM0090 Rev 19

}
}
else
if (STALL or FRMOR or NAK or DTERR or BBERR)
{
Mask ACK
Unmask CHH
Disable Channel
if (STALL or BBERR)
{
Reset Error Count
Transfer Done = 1
}
else
if (!FRMOR)
{
Reset Error Count
}
}
else
if (TXERR)
{
Unmask ACK
Unmask CHH
Disable Channel
}
else
if (CHH)
{
Mask CHH
{
De-allocate Channel
}
else
Re-initialize Channel (in next b_interval - 1 /Frame)
}
}
else
if (ACK)
{
Reset Error Count
Mask ACK
RM0090 Rev 19 1347/1751

1380
• Interrupt IN transactions
The assumptions are:
– The application is attempting to receive one packet (up to 1 maximum packet size)
in every frame, starting with odd (transfer size = 1 024 bytes).
– The receive FIFO can hold at least one maximum-packet-size packet and two
status words per packet (1 031 bytes).
– Periodic request queue depth = 4.
• Normal interrupt IN operation
a) Initialize channel 2. The application must set the ODDFRM bit in
OTG_FS_HCCHAR2.
b) Set the CHENA bit in OTG_FS_HCCHAR2 to write an IN request to the periodic
request queue.
c) The OTG_FS host writes an IN request to the periodic request queue for each
OTG_FS_HCCHAR2 register write with the CHENA bit set.
d) The OTG_FS host attempts to send an IN token in the next (odd) frame.
e) As soon as the IN packet is received and written to the receive FIFO, the OTG_FS
host generates an RXFLVL interrupt.
f) In response to the RXFLVL interrupt, read the received packet status to determine
the number of bytes received, then read the receive FIFO accordingly. The
application must mask the RXFLVL interrupt before reading the receive FIFO, and
unmask after reading the entire packet.
g) The core generates the RXFLVL interrupt for the transfer completion status entry
in the receive FIFO. The application must read and ignore the receive packet
status when the receive packet status is not an IN data packet (PKTSTS in
GRXSTSR ≠ 0b0010).
h) The core generates an XFRC interrupt as soon as the receive packet status is
read.
i) In response to the XFRC interrupt, read the PKTCNT field in OTG_FS_HCTSIZ2.
If the PKTCNT bit in OTG_FS_HCTSIZ2 is not equal to 0, disable the channel
before re-initializing the channel for the next transfer, if any). If PKTCNT bit in
1348/1751 RM0090 Rev 19

OTG_FS_HCTSIZ2 = 0, reinitialize the channel for the next transfer. This time, the
application must reset the ODDFRM bit in OTG_FS_HCCHAR2.
• Isochronous OUT transactions
A typical isochronous OUT operation is shown in Figure 401. The assumptions are:
– The application is attempting to send one packet every frame (up to 1 maximum
packet size), starting with an odd frame. (transfer size = 1 024 bytes).
– The periodic transmit FIFO can hold one packet (1 KB).
a) Initialize and enable channel 1. The application must set the ODDFRM bit in
OTG_FS_HCCHAR1.
b) Write the first packet for channel 1.
c) Along with the last word write of each packet, the OTG_FS host writes an entry to
the periodic request queue.
d) The OTG_FS host attempts to send the OUT token in the next frame (odd).
e) The OTG_FS host generates the XFRC interrupt as soon as the last packet is
transmitted successfully.
f) In response to the XFRC interrupt, reinitialize the channel for the next transfer.
g) Handling non-ACK responses
RM0090 Rev 19 1349/1751

1380
Figure 401. Normal isochronous OUT/IN transactions

1 init _reg(ch_1)
init_reg(ch _2) Periodic Request Queue
1 write_tx_fifo 3
Assume that this queue
1 can hold 4 entries.
(ch_1)
MPS
set_ch_en 2
(ch_2) ch_1 4
2 ch_2
3
OU T Odd
(micro)
DATA0
frame
M PS
5
6 XFRC interrupt ACK
init _reg(ch_1) 4 IN
write_tx_fifo
(ch_1) 1 5
MPS DATA0
RXFLVL interrupt
ACK
read_rx_sts
read_rx_fifo
1
6 MPS
RXFLVL interrupt
read_rx_sts ch_1
7 8
XFRC interrupt
ch_2
init_reg(ch _2)
9
set_ch_en
(ch_2)
Even
OU T (micro)
XFRC interrupt frame
DATA1
init _reg(ch_1)
MPS
write_tx_fifo 1
ACK
(ch_1) MPS
IN
DATA1
ai15676
• Interrupt service routine for isochronous OUT/IN transactions

Code sample: Isochronous OUT
Unmask (FRMOR/XFRC)
if (XFRC)
{
De-allocate Channel
}
else
if (FRMOR)
{
Unmask CHH
Disable Channel
}
1350/1751 RM0090 Rev 19

else
if (CHH)
{
Mask CHH
De-allocate Channel
}
Code sample: Isochronous IN
Unmask (TXERR/XFRC/FRMOR/BBERR)
if (XFRC or FRMOR)
{
if (XFRC and (OTG_FS_HCTSIZx.PKTCNT == 0))
{
Reset Error Count
De-allocate Channel
}
else
{
Unmask CHH
Disable Channel
}
}
else
if (TXERR or BBERR)
{
Unmask CHH
Disable Channel
}
else
if (CHH)
{
Mask CHH
{
De-allocate Channel
}
else
{
}
}
RM0090 Rev 19 1351/1751

1380
• Isochronous IN transactions
The assumptions are:
– The application is attempting to receive one packet (up to 1 maximum packet size)
in every frame starting with the next odd frame (transfer size = 1 024 bytes).
– The receive FIFO can hold at least one maximum-packet-size packet and two
status word per packet (1 031 bytes).
a) Initialize channel 2. The application must set the ODDFRM bit in
OTG_FS_HCCHAR2.
b) Set the CHENA bit in OTG_FS_HCCHAR2 to write an IN request to the periodic
request queue.
c) The OTG_FS host writes an IN request to the periodic request queue for each
OTG_FS_HCCHAR2 register write with the CHENA bit set.
d) The OTG_FS host attempts to send an IN token in the next odd frame.
e) As soon as the IN packet is received and written to the receive FIFO, the OTG_FS
host generates an RXFLVL interrupt.
f) In response to the RXFLVL interrupt, read the received packet status to determine
the number of bytes received, then read the receive FIFO accordingly. The
application must mask the RXFLVL interrupt before reading the receive FIFO, and
unmask it after reading the entire packet.
g) The core generates an RXFLVL interrupt for the transfer completion status entry in
the receive FIFO. This time, the application must read and ignore the receive
packet status when the receive packet status is not an IN data packet (PKTSTS bit
in OTG_FS_GRXSTSR ≠ 0b0010).
h) The core generates an XFRC interrupt as soon as the receive packet status is
read.
i) In response to the XFRC interrupt, read the PKTCNT field in OTG_FS_HCTSIZ2.
If PKTCNT ≠ 0 in OTG_FS_HCTSIZ2, disable the channel before re-initializing the
channel for the next transfer, if any. If PKTCNT = 0 in OTG_FS_HCTSIZ2,
reinitialize the channel for the next transfer. This time, the application must reset
the ODDFRM bit in OTG_FS_HCCHAR2.
• Selecting the queue depth
Choose the periodic and non-periodic request queue depths carefully to match the
number of periodic/non-periodic endpoints accessed.
The non-periodic request queue depth affects the performance of non-periodic
transfers. The deeper the queue (along with sufficient FIFO size), the more often the
core is able to pipeline non-periodic transfers. If the queue size is small, the core is
able to put in new requests only when the queue space is freed up.
The core’s periodic request queue depth is critical to perform periodic transfers as
scheduled. Select the periodic queue depth, based on the number of periodic transfers
scheduled in a microframe. If the periodic request queue depth is smaller than the
periodic transfers scheduled in a microframe, a frame overrun condition occurs.
• Handling babble conditions
OTG_FS controller handles two cases of babble: packet babble and port babble.
Packet babble occurs if the device sends more data than the maximum packet size for
1352/1751 RM0090 Rev 19

the channel. Port babble occurs if the core continues to receive data from the device at
EOF2 (the end of frame 2, which is very close to SOF).
When OTG_FS controller detects a packet babble, it stops writing data into the Rx
buffer and waits for the end of packet (EOP). When it detects an EOP, it flushes already
written data in the Rx buffer and generates a Babble interrupt to the application.
When OTG_FS controller detects a port babble, it flushes the RxFIFO and disables the
port. The core then generates a Port disabled interrupt (HPRTINT in
OTG_FS_GINTSTS, PENCHNG in OTG_FS_HPRT). On receiving this interrupt, the
application must determine that this is not due to an overcurrent condition (another
cause of the Port Disabled interrupt) by checking POCA in OTG_FS_HPRT, then
perform a soft reset. The core does not send any more tokens after it has detected a
port babble condition.
34.17.5 Device programming model

Endpoint initialization on USB reset
1. Set the NAK bit for all OUT endpoints
– SNAK = 1 in OTG_FS_DOEPCTLx (for all OUT endpoints)
2. Unmask the following interrupt bits
– INEP0 = 1 in OTG_FS_DAINTMSK (control 0 IN endpoint)
– OUTEP0 = 1 in OTG_FS_DAINTMSK (control 0 OUT endpoint)
– STUP = 1 in DOEPMSK
– XFRC = 1 in DOEPMSK
– XFRC = 1 in DIEPMSK
– TOC = 1 in DIEPMSK
3. Set up the Data FIFO RAM for each of the FIFOs
– Program the OTG_FS_GRXFSIZ register, to be able to receive control OUT data
and setup data. If thresholding is not enabled, at a minimum, this must be equal to
1 max packet size of control endpoint 0 + 2 words (for the status of the control
OUT data packet) + 10 words (for setup packets).
– Program the OTG_FS_TX0FSIZ register (depending on the FIFO number chosen)
to be able to transmit control IN data. At a minimum, this must be equal to 1 max
packet size of control endpoint 0.
4. Program the following fields in the endpoint-specific registers for control OUT endpoint
0 to receive a SETUP packet
– STUPCNT = 3 in OTG_FS_DOEPTSIZ0 (to receive up to 3 back-to-back SETUP
packets)
At this point, all initialization required to receive SETUP packets is done.
Endpoint initialization on enumeration completion

1. On the Enumeration Done interrupt (ENUMDNE in OTG_FS_GINTSTS), read the
OTG_FS_DSTS register to determine the enumeration speed.
2. Program the MPSIZ field in OTG_FS_DIEPCTL0 to set the maximum packet size. This
step configures control endpoint 0. The maximum packet size for a control endpoint
depends on the enumeration speed.
RM0090 Rev 19 1353/1751

1380
At this point, the device is ready to receive SOF packets and is configured to perform control
transfers on control endpoint 0.
Endpoint initialization on SetAddress command

This section describes what the application must do when it receives a SetAddress
command in a SETUP packet.
1. Program the OTG_FS_DCFG register with the device address received in the
SetAddress command
2. Program the core to send out a status IN packet
Endpoint initialization on SetConfiguration/SetInterface command

This section describes what the application must do when it receives a SetConfiguration or
SetInterface command in a SETUP packet.
1. When a SetConfiguration command is received, the application must program the
endpoint registers to configure them with the characteristics of the valid endpoints in
the new configuration.
2. When a SetInterface command is received, the application must program the endpoint
registers of the endpoints affected by this command.
3. Some endpoints that were active in the prior configuration or alternate setting are not
valid in the new configuration or alternate setting. These invalid endpoints must be
deactivated.
4. Unmask the interrupt for each active endpoint and mask the interrupts for all inactive
endpoints in the OTG_FS_DAINTMSK register.
5. Set up the Data FIFO RAM for each FIFO.
6. After all required endpoints are configured; the application must program the core to
send a status IN packet.
At this point, the device core is configured to receive and transmit any type of data packet.
Endpoint activation
This section describes the steps required to activate a device endpoint or to configure an
existing device endpoint to a new type.
1. Program the characteristics of the required endpoint into the following fields of the
OTG_FS_DIEPCTLx register (for IN or bidirectional endpoints) or the
OTG_FS_DOEPCTLx register (for OUT or bidirectional endpoints).
– Maximum packet size
– USB active endpoint = 1
– Endpoint start data toggle (for interrupt and bulk endpoints)
– Endpoint type
– TxFIFO number
2. Once the endpoint is activated, the core starts decoding the tokens addressed to that
endpoint and sends out a valid handshake for each valid token received for the
endpoint.
1354/1751 RM0090 Rev 19

Endpoint deactivation
This section describes the steps required to deactivate an existing endpoint.
1. In the endpoint to be deactivated, clear the USB active endpoint bit in the
OTG_FS_DIEPCTLx register (for IN or bidirectional endpoints) or the
OTG_FS_DOEPCTLx register (for OUT or bidirectional endpoints).
2. Once the endpoint is deactivated, the core ignores tokens addressed to that endpoint,
which results in a timeout on the USB.
Note: The application must meet the following conditions to set up the device core to handle
traffic:
NPTXFEM and RXFLVLM in the OTG_FS_GINTMSK register must be cleared.
34.17.6 Operational model

SETUP and OUT data transfers
This section describes the internal data flow and application-level operations during data
OUT transfers and SETUP transactions.
• Packet read
This section describes how to read packets (OUT data and SETUP packets) from the
receive FIFO.
1. On catching an RXFLVL interrupt (OTG_FS_GINTSTS register), the application must
read the Receive status pop register (OTG_FS_GRXSTSP).
2. The application can mask the RXFLVL interrupt (in OTG_FS_GINTSTS) by writing to
RXFLVL = 0 (in OTG_FS_GINTMSK), until it has read the packet from the receive
FIFO.
3. If the received packet’s byte count is not 0, the byte count amount of data is popped
from the receive Data FIFO and stored in memory. If the received packet byte count is
0, no data is popped from the receive data FIFO.
4. The receive FIFO’s packet status readout indicates one of the following:
a) Global OUT NAK pattern:
PKTSTS = Global OUT NAK, BCNT = 0x000, EPNUM = Don’t Care (0x0),
DPID = Don’t Care (0b00).
These data indicate that the global OUT NAK bit has taken effect.
b) SETUP packet pattern:
PKTSTS = SETUP, BCNT = 0x008, EPNUM = Control EP Num, DPID = D0.
These data indicate that a SETUP packet for the specified endpoint is now
available for reading from the receive FIFO.
c) Setup stage done pattern:
PKTSTS = Setup Stage Done, BCNT = 0x0, EPNUM = Control EP Num,
DPID = Don’t Care (0b00).
These data indicate that the Setup stage for the specified endpoint has completed
and the Data stage has started. After this entry is popped from the receive FIFO,
the core asserts a Setup interrupt on the specified control OUT endpoint.
d) Data OUT packet pattern:
PKTSTS = DataOUT, BCNT = size of the received data OUT packet (0 ≤ BCNT
≤ 1 024), EPNUM = EPNUM on which the packet was received, DPID = Actual
Data PID.
RM0090 Rev 19 1355/1751

1380
e) Data transfer completed pattern:

PKTSTS = Data OUT Transfer Done, BCNT = 0x0, EPNUM = OUT EP Num
on which the data transfer is complete, DPID = Don’t Care (0b00).
These data indicate that an OUT data transfer for the specified OUT endpoint has
completed. After this entry is popped from the receive FIFO, the core asserts a
Transfer Completed interrupt on the specified OUT endpoint.
5. After the data payload is popped from the receive FIFO, the RXFLVL interrupt
(OTG_FS_GINTSTS) must be unmasked.
6. Steps 1–5 are repeated every time the application detects assertion of the interrupt line
due to RXFLVL in OTG_FS_GINTSTS. Reading an empty receive FIFO can result in
undefined core behavior.
Figure 402 provides a flowchart of the above procedure.
Figure 402. Receive FIFO packet read
wait until RXFLVL in OTG_FS_GINTSTSG
rd_data = rd_reg (OTG_FS_GRXSTSP);
Y rd_data.BCNT = 0 rcv_out_pkt ()
word_cnt =
packet mem[0: word_cnt – 1] = BCNT[11:2]
C +
store in rd_rxfifo(rd_data.EPNUM, (BCNT[1] | BCNT[1])
memory word_cnt)
ai15677b
• SETUP transactions
This section describes how the core handles SETUP packets and the application’s
sequence for handling SETUP transactions.
• Application requirements
1. To receive a SETUP packet, the STUPCNT field (OTG_FS_DOEPTSIZx) in a control
OUT endpoint must be programmed to a non-zero value. When the application
programs the STUPCNT field to a non-zero value, the core receives SETUP packets
and writes them to the receive FIFO, irrespective of the NAK status and EPENA bit
setting in OTG_FS_DOEPCTLx. The STUPCNT field is decremented every time the
control endpoint receives a SETUP packet. If the STUPCNT field is not programmed to
a proper value before receiving a SETUP packet, the core still receives the SETUP
packet and decrements the STUPCNT field, but the application may not be able to
1356/1751 RM0090 Rev 19

determine the correct number of SETUP packets received in the Setup stage of a
control transfer.
– STUPCNT = 3 in OTG_FS_DOEPTSIZx
2. The application must always allocate some extra space in the Receive data FIFO, to be
able to receive up to three SETUP packets on a control endpoint.
– The space to be reserved is 10 words. Three words are required for the first
SETUP packet, 1 word is required for the Setup stage done word and 6 words are
required to store two extra SETUP packets among all control endpoints.
– 3 words per SETUP packet are required to store 8 bytes of SETUP data and 4
bytes of SETUP status (Setup packet pattern). The core reserves this space in the
receive data FIFO to write SETUP data only, and never uses this space for data
packets.
3. The application must read the 2 words of the SETUP packet from the receive FIFO.
4. The application must read and discard the Setup stage done word from the receive
FIFO.
• Internal data flow
1. When a SETUP packet is received, the core writes the received data to the receive
FIFO, without checking for available space in the receive FIFO and irrespective of the
endpoint’s NAK and STALL bit settings.
– The core internally sets the IN NAK and OUT NAK bits for the control IN/OUT
endpoints on which the SETUP packet was received.
2. For every SETUP packet received on the USB, 3 words of data are written to the
receive FIFO, and the STUPCNT field is decremented by 1.
– The first word contains control information used internally by the core
– The second word contains the first 4 bytes of the SETUP command
– The third word contains the last 4 bytes of the SETUP command
3. When the Setup stage changes to a Data IN/OUT stage, the core writes an entry
(Setup stage done word) to the receive FIFO, indicating the completion of the Setup
stage.
4. On the AHB side, SETUP packets are emptied by the application.
5. When the application pops the Setup stage done word from the receive FIFO, the core
interrupts the application with an STUP interrupt (OTG_FS_DOEPINTx), indicating it
can process the received SETUP packet.
– The core clears the endpoint enable bit for control OUT endpoints.
• Application programming sequence
1. Program the OTG_FS_DOEPTSIZx register.
– STUPCNT = 3
2. Wait for the RXFLVL interrupt (OTG_FS_GINTSTS) and empty the data packets from
the receive FIFO.
3. Assertion of the STUP interrupt (OTG_FS_DOEPINTx) marks a successful completion
of the SETUP Data Transfer.
– On this interrupt, the application must read the OTG_FS_DOEPTSIZx register to
determine the number of SETUP packets received and process the last received
SETUP packet.
RM0090 Rev 19 1357/1751

1380
Figure 403. Processing a SETUP packet
Wait for STUP in OTG_FS_DOEPINTx
rem_supcnt =
rd_reg(DOEPTSIZx)
setup_cmd[31:0] = mem[4 – 2 * rem_supcnt]

setup_cmd[63:32] = mem[5 – 2 * rem_supcnt]
Find setup cmd type
Read ctrl-rd/wr/2 stage Write
2-stage
setup_np_in_pkt setup_np_in_pkt rcv_out_pkt

Data IN phase Status IN phase Data OUT phase
ai15678
• Handling more than three back-to-back SETUP packets

Per the USB 2.0 specification, normally, during a SETUP packet error, a host does not send
more than three back-to-back SETUP packets to the same endpoint. However, the USB 2.0
specification does not limit the number of back-to-back SETUP packets a host can send to
the same endpoint. When this condition occurs, the OTG_FS controller generates an
interrupt (B2BSTUP in OTG_FS_DOEPINTx).
• Setting the global OUT NAK
Internal data flow
1. When the application sets the Global OUT NAK (SGONAK bit in OTG_FS_DCTL), the
core stops writing data, except SETUP packets, to the receive FIFO. Irrespective of the
space availability in the receive FIFO, non-isochronous OUT tokens receive a NAK
handshake response, and the core ignores isochronous OUT data packets
2. The core writes the Global OUT NAK pattern to the receive FIFO. The application must
reserve enough receive FIFO space to write this data pattern.
3. When the application pops the Global OUT NAK pattern word from the receive FIFO,
the core sets the GONAKEFF interrupt (OTG_FS_GINTSTS).
4. Once the application detects this interrupt, it can assume that the core is in Global OUT
NAK mode. The application can clear this interrupt by clearing the SGONAK bit in
OTG_FS_DCTL.
1358/1751 RM0090 Rev 19

Application programming sequence

1. To stop receiving any kind of data in the receive FIFO, the application must set the
Global OUT NAK bit by programming the following field:
– SGONAK = 1 in OTG_FS_DCTL
2. Wait for the assertion of the GONAKEFF interrupt in OTG_FS_GINTSTS. When
asserted, this interrupt indicates that the core has stopped receiving any type of data
except SETUP packets.
3. The application can receive valid OUT packets after it has set SGONAK in
OTG_FS_DCTL and before the core asserts the GONAKEFF interrupt
(OTG_FS_GINTSTS).
4. The application can temporarily mask this interrupt by writing to the GINAKEFFM bit in
the OTG_FS_GINTMSK register.
– GINAKEFFM = 0 in the OTG_FS_GINTMSK register
5. Whenever the application is ready to exit the Global OUT NAK mode, it must clear the
SGONAK bit in OTG_FS_DCTL. This also clears the GONAKEFF interrupt
(OTG_FS_GINTSTS).
– OTG_FS_DCTL = 1 in CGONAK
6. If the application has masked this interrupt earlier, it must be unmasked as follows:
– GINAKEFFM = 1 in GINTMSK
• Disabling an OUT endpoint
The application must use this sequence to disable an OUT endpoint that it has enabled.
1. Before disabling any OUT endpoint, the application must enable Global OUT NAK
mode in the core.
2. Wait for the GONAKEFF interrupt (OTG_FS_GINTSTS)
3. Disable the required OUT endpoint by programming the following fields:
– EPDIS = 1 in OTG_FS_DOEPCTLx
– SNAK = 1 in OTG_FS_DOEPCTLx
4. Wait for the EPDISD interrupt (OTG_FS_DOEPINTx), which indicates that the OUT
endpoint is completely disabled. When the EPDISD interrupt is asserted, the core also
clears the following bits:
– EPDIS = 0 in OTG_FS_DOEPCTLx
– EPENA = 0 in OTG_FS_DOEPCTLx
5. The application must clear the Global OUT NAK bit to start receiving data from other
non-disabled OUT endpoints.
• Transfer Stop Programming for OUT endpoints
The application must use the following programing sequence to stop any transfers (because
of an interrupt from the host, typically a reset).
RM0090 Rev 19 1359/1751

1380
Sequence of operations:
1. Enable all OUT endpoints by setting
– EPENA = 1 in all OTG_FS_DOEPCTLx registers.
2. Flush the RxFIFO as follows
– Poll OTG_FS_GRSTCTL.AHBIDL until it is 1. This indicates that AHB master is
idle.
– Perform read modify write operation on OTG_FS_GRSTCTL.RXFFLSH =1
– Poll OTG_FS_GRSTCTL.RXFFLSH until it is 0, but also using a timeout of less
than 10 milli-seconds (corresponds to minimum reset signaling duration). If 0 is
seen before the timeout, then the RxFIFO flush is successful. If at the moment the
timeout occurs, there is still a 1, (this may be due to a packet on EP0 coming from
the host) then go back (once only) to the previous step (“Perform read modify write
operation”).
3. Before disabling any OUT endpoint, the application must enable Global OUT NAK
mode in the core, according to the instructions in “Setting the global OUT NAK on
page 1358”. This ensures that data in the RxFIFO is sent to the application
successfully. Set SGONAK = 1 in OTG_FS_DCTL
4. Wait for the GONAKEFF interrupt (OTG_FS_GINTSTS)
5. Disable all active OUT endpoints by programming the following register bits:
– EPDIS = 1 in registers OTG_FS_DOEPCTLx
– SNAK = 1 in registers OTG_FS_DOEPCTLx
6. Wait for the EPDIS interrupt in OTG_FS_DOEPINTx for each OUT endpoint
programmed in the previous step. The EPDIS interrupt in OTG_FS_DOEPINTx
indicates that the corresponding OUT endpoint is completely disabled. When the
EPDIS interrupt is asserted, the following bits are cleared:
– EPENA = 0 in registers OTG_FS_DOEPCTLx
– EPDIS = 0 in registers OTG_FS_DOEPCTLx
– SNAK = 0 in registers OTG_FS_DOEPCTLx
• Generic non-isochronous OUT data transfers
This section describes a regular non-isochronous OUT data transfer (control, bulk, or
interrupt).
Application requirements
1. Before setting up an OUT transfer, the application must allocate a buffer in the memory
to accommodate all data to be received as part of the OUT transfer.
2. For OUT transfers, the transfer size field in the endpoint’s transfer size register must be
a multiple of the maximum packet size of the endpoint, adjusted to the word boundary.
– transfer size[EPNUM] = n × (MPSIZ[EPNUM] + 4 – (MPSIZ[EPNUM] mod 4))
– packet count[EPNUM] = n
– n>0
3. On any OUT endpoint interrupt, the application must read the endpoint’s transfer size
register to calculate the size of the payload in the memory. The received payload size
can be less than the programmed transfer size.
– Payload size in memory = application programmed initial transfer size – core
updated final transfer size
1360/1751 RM0090 Rev 19

– Number of USB packets in which this payload was received = application

programmed initial packet count – core updated final packet count
Internal data flow
1. The application must set the transfer size and packet count fields in the endpoint-
specific registers, clear the NAK bit, and enable the endpoint to receive the data.
2. Once the NAK bit is cleared, the core starts receiving data and writes it to the receive
FIFO, as long as there is space in the receive FIFO. For every data packet received on
the USB, the data packet and its status are written to the receive FIFO. Every packet
(maximum packet size or short packet) written to the receive FIFO decrements the
packet count field for that endpoint by 1.
– OUT data packets received with bad data CRC are flushed from the receive FIFO
automatically.
– After sending an ACK for the packet on the USB, the core discards non-
isochronous OUT data packets that the host, which cannot detect the ACK, re-
sends. The application does not detect multiple back-to-back data OUT packets
on the same endpoint with the same data PID. In this case the packet count is not
decremented.
– If there is no space in the receive FIFO, isochronous or non-isochronous data
packets are ignored and not written to the receive FIFO. Additionally, non-
isochronous OUT tokens receive a NAK handshake reply.
– In all the above three cases, the packet count is not decremented because no data
are written to the receive FIFO.
3. When the packet count becomes 0 or when a short packet is received on the endpoint,
the NAK bit for that endpoint is set. Once the NAK bit is set, the isochronous or non-
isochronous data packets are ignored and not written to the receive FIFO, and non-
isochronous OUT tokens receive a NAK handshake reply.
4. After the data are written to the receive FIFO, the application reads the data from the
receive FIFO and writes it to external memory, one packet at a time per endpoint.
5. At the end of every packet write on the AHB to external memory, the transfer size for
the endpoint is decremented by the size of the written packet.
6. The OUT data transfer completed pattern for an OUT endpoint is written to the receive
FIFO on one of the following conditions:
– The transfer size is 0 and the packet count is 0
– The last OUT data packet written to the receive FIFO is a short packet
(0 ≤ packet size < maximum packet size)
7. When either the application pops this entry (OUT data transfer completed), a transfer
completed interrupt is generated for the endpoint and the endpoint enable is cleared.
RM0090 Rev 19 1361/1751

1380

1. Program the OTG_FS_DOEPTSIZx register for the transfer size and the corresponding
packet count.
2. Program the OTG_FS_DOEPCTLx register with the endpoint characteristics, and set
the EPENA and CNAK bits.
– EPENA = 1 in OTG_FS_DOEPCTLx
– CNAK = 1 in OTG_FS_DOEPCTLx
3. Wait for the RXFLVL interrupt (in OTG_FS_GINTSTS) and empty the data packets
from the receive FIFO.
– This step can be repeated many times, depending on the transfer size.
4. Asserting the XFRC interrupt (OTG_FS_DOEPINTx) marks a successful completion of
the non-isochronous OUT data transfer.
5. Read the OTG_FS_DOEPTSIZx register to determine the size of the received data
payload.
• Generic isochronous OUT data transfer
This section describes a regular isochronous OUT data transfer.
Application requirements
1. All the application requirements for non-isochronous OUT data transfers also apply to
isochronous OUT data transfers.
2. For isochronous OUT data transfers, the transfer size and packet count fields must
always be set to the number of maximum-packet-size packets that can be received in a
single frame and no more. Isochronous OUT data transfers cannot span more than 1
frame.
3. The application must read all isochronous OUT data packets from the receive FIFO
(data and status) before the end of the periodic frame (EOPF interrupt in
OTG_FS_GINTSTS).
4. To receive data in the following frame, an isochronous OUT endpoint must be enabled
after the EOPF (OTG_FS_GINTSTS) and before the SOF (OTG_FS_GINTSTS).
Internal data flow
1. The internal data flow for isochronous OUT endpoints is the same as that for non-
isochronous OUT endpoints, but for a few differences.
2. When an isochronous OUT endpoint is enabled by setting the Endpoint Enable and
clearing the NAK bits, the Even/Odd frame bit must also be set appropriately. The core
receives data on an isochronous OUT endpoint in a particular frame only if the
following condition is met:
– EONUM (in OTG_FS_DOEPCTLx) = SOFFN[0] (in OTG_FS_DSTS)
3. When the application completely reads an isochronous OUT data packet (data and
status) from the receive FIFO, the core updates the RXDPID field in
OTG_FS_DOEPTSIZx with the data PID of the last isochronous OUT data packet read
from the receive FIFO.
1362/1751 RM0090 Rev 19


1. Program the OTG_FS_DOEPTSIZx register for the transfer size and the corresponding
packet count
2. Program the OTG_FS_DOEPCTLx register with the endpoint characteristics and set
the Endpoint Enable, ClearNAK, and Even/Odd frame bits.
– EPENA = 1
– CNAK = 1
– EONUM = (0: Even/1: Odd)
3. Wait for the RXFLVL interrupt (in OTG_FS_GINTSTS) and empty the data packets
from the receive FIFO
– This step can be repeated many times, depending on the transfer size.
4. The assertion of the XFRC interrupt (in OTG_FS_DOEPINTx) marks the completion of
the isochronous OUT data transfer. This interrupt does not necessarily mean that the
data in memory are good.
5. This interrupt cannot always be detected for isochronous OUT transfers. Instead, the
application can detect the IISOOXFRM interrupt in OTG_FS_GINTSTS.
6. Read the OTG_FS_DOEPTSIZx register to determine the size of the received transfer
and to determine the validity of the data received in the frame. The application must
treat the data received in memory as valid only if one of the following conditions is met:
– RXDPID = D0 (in OTG_FS_DOEPTSIZx) and the number of USB packets in
which this payload was received = 1
The number of USB packets in which this payload was received =
Application programmed initial packet count – Core updated final packet count
The application can discard invalid data packets.
• Incomplete isochronous OUT data transfers
This section describes the application programming sequence when isochronous OUT data
packets are dropped inside the core.
Internal data flow
1. For isochronous OUT endpoints, the XFRC interrupt (in OTG_FS_DOEPINTx) may not
always be asserted. If the core drops isochronous OUT data packets, the application
could fail to detect the XFRC interrupt (OTG_FS_DOEPINTx) under the following
circumstances:
– When the receive FIFO cannot accommodate the complete ISO OUT data packet,
the core drops the received ISO OUT data
– When the isochronous OUT data packet is received with CRC errors
– When the isochronous OUT token received by the core is corrupted
– When the application is very slow in reading the data from the receive FIFO
2. When the core detects an end of periodic frame before transfer completion to all
isochronous OUT endpoints, it asserts the incomplete Isochronous OUT data interrupt
(IISOOXFRM in OTG_FS_GINTSTS), indicating that an XFRC interrupt (in
OTG_FS_DOEPINTx) is not asserted on at least one of the isochronous OUT
RM0090 Rev 19 1363/1751

1380
endpoints. At this point, the endpoint with the incomplete transfer remains enabled, but
no active transfers remain in progress on this endpoint on the USB.
1. Asserting the IISOOXFRM interrupt (OTG_FS_GINTSTS) indicates that in the current
frame, at least one isochronous OUT endpoint has an incomplete transfer.
2. If this occurs because isochronous OUT data is not completely emptied from the
endpoint, the application must ensure that the application empties all isochronous OUT
data (data and status) from the receive FIFO before proceeding.
– When all data are emptied from the receive FIFO, the application can detect the
XFRC interrupt (OTG_FS_DOEPINTx). In this case, the application must re-
enable the endpoint to receive isochronous OUT data in the next frame.
3. When it receives an IISOOXFRM interrupt (in OTG_FS_GINTSTS), the application
must read the control registers of all isochronous OUT endpoints
(OTG_FS_DOEPCTLx) to determine which endpoints had an incomplete transfer in the
current microframe. An endpoint transfer is incomplete if both the following conditions
are met:
– EONUM bit (in OTG_FS_DOEPCTLx) = SOFFN[0] (in OTG_FS_DSTS)
– EPENA = 1 (in OTG_FS_DOEPCTLx)
4. The previous step must be performed before the SOF interrupt (in OTG_FS_GINTSTS)
is detected, to ensure that the current frame number is not changed.
5. For isochronous OUT endpoints with incomplete transfers, the application must discard
the data in the memory and disable the endpoint by setting the EPDIS bit in
OTG_FS_DOEPCTLx.
6. Wait for the EPDIS interrupt (in OTG_FS_DOEPINTx) and enable the endpoint to
receive new data in the next frame.
– Because the core can take some time to disable the endpoint, the application may
not be able to receive the data in the next frame after receiving bad isochronous
data.
• Stalling a non-isochronous OUT endpoint
This section describes how the application can stall a non-isochronous endpoint.
1. Put the core in the Global OUT NAK mode.
2. Disable the required endpoint
– When disabling the endpoint, instead of setting the SNAK bit in
OTG_FS_DOEPCTL, set STALL = 1 (in OTG_FS_DOEPCTL).
The STALL bit always takes precedence over the NAK bit.
3. When the application is ready to end the STALL handshake for the endpoint, the
STALL bit (in OTG_FS_DOEPCTLx) must be cleared.
4. If the application is setting or clearing a STALL for an endpoint due to a
SetFeature.Endpoint Halt or ClearFeature.Endpoint Halt command, the STALL bit must
be set or cleared before the application sets up the Status stage transfer on the control
endpoint.
Examples
This section describes and depicts some fundamental transfer types and scenarios.
• Bulk OUT transaction
1364/1751 RM0090 Rev 19

Figure 404 depicts the reception of a single Bulk OUT Data packet from the USB to the AHB
and describes the events involved in the process.
Figure 404. Bulk OUT transaction
Host USB Device Application
init_ out_ ep
XFRSIZ = 64 bytes
1 PKTCNT = 1
2 wr_reg (DOEPTSIZx)
O UT EPENA= 1
CNAK = 1
3 wr_reg(D OEPCTLx)
64 bytes
4 6
xact _1
AC K RXFLVL iintr
D OE P C idle until intr
T L x.N A
5 PKTCN K = 1
T0
XFRSIZ
r =0 rcv_out _pkt()
On new xfer
OU T XF or RxFIFO
int r RC not empty
7
NA K
idle until intr
8
ai15679b
After a SetConfiguration/SetInterface command, the application initializes all OUT endpoints

by setting CNAK = 1 and EPENA = 1 (in OTG_FS_DOEPCTLx), and setting a suitable
XFRSIZ and PKTCNT in the OTG_FS_DOEPTSIZx register.
1. host attempts to send data (OUT token) to an endpoint.
2. When the core receives the OUT token on the USB, it stores the packet in the RxFIFO
because space is available there.
3. After writing the complete packet in the RxFIFO, the core then asserts the RXFLVL
interrupt (in OTG_FS_GINTSTS).
4. On receiving the PKTCNT number of USB packets, the core internally sets the NAK bit
for this endpoint to prevent it from receiving any more packets.
5. The application processes the interrupt and reads the data from the RxFIFO.
6. When the application has read all the data (equivalent to XFRSIZ), the core generates
an XFRC interrupt (in OTG_FS_DOEPINTx).
7. The application processes the interrupt and uses the setting of the XFRC interrupt bit
(in OTG_FS_DOEPINTx) to determine that the intended transfer is complete.
IN data transfers
• Packet write
This section describes how the application writes data packets to the endpoint FIFO when
dedicated transmit FIFOs are enabled.
1. The application can either choose the polling or the interrupt mode.
RM0090 Rev 19 1365/1751

1380
– In polling mode, the application monitors the status of the endpoint transmit data
FIFO by reading the OTG_FS_DTXFSTSx register, to determine if there is enough
space in the data FIFO.
– In interrupt mode, the application waits for the TXFE interrupt (in
OTG_FS_DIEPINTx) and then reads the OTG_FS_DTXFSTSx register, to
determine if there is enough space in the data FIFO.
– To write a single non-zero length data packet, there must be space to write the
entire packet in the data FIFO.
– To write zero length packet, the application must not look at the FIFO space.
2. Using one of the above mentioned methods, when the application determines that
there is enough space to write a transmit packet, the application must first write into the
endpoint control register, before writing the data into the data FIFO. Typically, the
application, must do a read modify write on the OTG_FS_DIEPCTLx register to avoid
modifying the contents of the register, except for setting the Endpoint Enable bit.
The application can write multiple packets for the same endpoint into the transmit FIFO, if
space is available. For periodic IN endpoints, the application must write packets only for one
microframe. It can write packets for the next periodic transaction only after getting transfer
complete for the previous transaction.
• Setting IN endpoint NAK
Internal data flow
1. When the application sets the IN NAK for a particular endpoint, the core stops
transmitting data on the endpoint, irrespective of data availability in the endpoint’s
transmit FIFO.
2. Non-isochronous IN tokens receive a NAK handshake reply
– Isochronous IN tokens receive a zero-data-length packet reply
3. The core asserts the INEPNE (IN endpoint NAK effective) interrupt in
OTG_FS_DIEPINTx in response to the SNAK bit in OTG_FS_DIEPCTLx.
4. Once this interrupt is seen by the application, the application can assume that the
endpoint is in IN NAK mode. This interrupt can be cleared by the application by setting
the CNAK bit in OTG_FS_DIEPCTLx.
1. To stop transmitting any data on a particular IN endpoint, the application must set the
IN NAK bit. To set this bit, the following field must be programmed.
– SNAK = 1 in OTG_FS_DIEPCTLx
2. Wait for assertion of the INEPNE interrupt in OTG_FS_DIEPINTx. This interrupt
indicates that the core has stopped transmitting data on the endpoint.
3. The core can transmit valid IN data on the endpoint after the application has set the
NAK bit, but before the assertion of the NAK Effective interrupt.
4. The application can mask this interrupt temporarily by writing to the INEPNEM bit in
DIEPMSK.
– INEPNEM = 0 in DIEPMSK
5. To exit Endpoint NAK mode, the application must clear the NAK status bit (NAKSTS) in
OTG_FS_DIEPCTLx. This also clears the INEPNE interrupt (in OTG_FS_DIEPINTx).
– CNAK = 1 in OTG_FS_DIEPCTLx
6. If the application masked this interrupt earlier, it must be unmasked as follows:
1366/1751 RM0090 Rev 19

– INEPNEM = 1 in DIEPMSK
• IN endpoint disable
Use the following sequence to disable a specific IN endpoint that has been previously
enabled.
1. The application must stop writing data on the AHB for the IN endpoint to be disabled.
2. The application must set the endpoint in NAK mode.
– SNAK = 1 in

STM32 RM

Uploaded by

Copyright:

Available Formats

STM32 RM

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

STM32 RM

Uploaded by

Copyright:

Available Formats

RM0090

February 2021 RM0090 Rev 19 1/1751

2 Memory and bus architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

3 Embedded Flash memory interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

2/1751 RM0090 Rev 19

3.5.2 Adaptive real-time memory accelerator (ART Accelerator™) . . . . . . . . 82

4 CRC calculation unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113

RM0090 Rev 19 3/1751

4.1 CRC introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113

5 Power controller (PWR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116

4/1751 RM0090 Rev 19

5.6 PWR register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

6 Reset and clock control for

RM0090 Rev 19 5/1751

7 Reset and clock control for

6/1751 RM0090 Rev 19

7.3.4 RCC clock interrupt register (RCC_CIR) . . . . . . . . . . . . . . . . . . . . . . . 230

8 General-purpose I/Os (GPIO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267

RM0090 Rev 19 7/1751

8.3.8 External interrupt/wakeup lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275

9 System configuration controller (SYSCFG) . . . . . . . . . . . . . . . . . . . . 289

8/1751 RM0090 Rev 19

9.2.7 Compensation cell control register (SYSCFG_CMPCR) . . . . . . . . . . . 293

10 DMA controller (DMA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302

RM0090 Rev 19 9/1751

10.3.18 Error management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323

11 Chrom-Art Accelerator™ controller (DMA2D) . . . . . . . . . . . . . . . . . . 339

10/1751 RM0090 Rev 19

11.5.1 DMA2D control register (DMA2D_CR) . . . . . . . . . . . . . . . . . . . . . . . . 352

12 Interrupts and events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371

RM0090 Rev 19 11/1751

12.3.2 Event mask register (EXTI_EMR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384

13 Analog-to-digital converter (ADC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388

12/1751 RM0090 Rev 19

13.10 Temperature sensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412

14 Digital-to-analog converter (DAC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433

RM0090 Rev 19 13/1751

14.4 Dual DAC channel conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440

15 Digital camera interface (DCMI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455

14/1751 RM0090 Rev 19

15.1 DCMI introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455

16 LCD-TFT controller (LTDC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480

RM0090 Rev 19 15/1751

16.3.1 LTDC block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 481

16/1751 RM0090 Rev 19

16.7.23 LTDC Layerx Color Frame Buffer Length Register (LTDC_LxCFBLR)

17 Advanced-control timers (TIM1 and TIM8) . . . . . . . . . . . . . . . . . . . . . 515

RM0090 Rev 19 17/1751

17.4.5 TIM1 and TIM8 status register (TIMx_SR) . . . . . . . . . . . . . . . . . . . . . . 569

18 General-purpose timers (TIM2 to TIM5) . . . . . . . . . . . . . . . . . . . . . . . . 589

18/1751 RM0090 Rev 19

18.3.16 Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626

19 General-purpose timers (TIM9 to TIM14) . . . . . . . . . . . . . . . . . . . . . . . 650

RM0090 Rev 19 19/1751

19.3.8 Output compare mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664

20 Basic timers (TIM6 and TIM7) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 696

20/1751 RM0090 Rev 19

20.3 TIM6 and TIM7 functional description . . . . . . . . . . . . . . . . . . . . . . . . . . 697

21 Independent watchdog (IWDG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708