ATmega 128

Features
• High-performance, Low-power AVR® 8-bit Microcontroller

• Advanced RISC Architecture
– 133 Powerful Instructions – Most Single Clock Cycle Execution
– 32 x 8 General Purpose Working Registers + Peripheral Control Registers
– Fully Static Operation
– Up to 16 MIPS Throughput at 16 MHz
– On-chip 2-cycle Multiplier
• Nonvolatile Program and Data Memories
– 128K Bytes of In-System Reprogrammable Flash
Endurance: 1,000 Write/Erase Cycles
– Optional Boot Code Section with Independent Lock Bits
8-bit
In-System Programming by On-chip Boot Program
True Read-While-Write Operation
Microcontroller
– 4K Bytes EEPROM
Endurance: 100,000 Write/Erase Cycles with 128K Bytes
– 4K Bytes Internal SRAM
– Up to 64K Bytes Optional External Memory Space In-System
– Programming Lock for Software Security
– SPI Interface for In-System Programming Programmable
• JTAG (IEEE std. 1149.1 Compliant) Interface
– Boundary-scan Capabilities According to the JTAG Standard
– Extensive On-chip Debug Support
Flash
– Programming of Flash, EEPROM, Fuses and Lock Bits through the JTAG Interface
• Peripheral Features
– Two 8-bit Timer/Counters with Separate Prescalers and Compare Modes
– Two Expanded 16-bit Timer/Counters with Separate Prescaler, Compare Mode and
ATmega128
Capture Mode
– Real Time Counter with Separate Oscillator
ATmega128L
– Two 8-bit PWM Channels
– 6 PWM Channels with Programmable Resolution from 2 to 16 Bits
– Output Compare Modulator
– 8-channel, 10-bit ADC Preliminary
8 Single-ended Channels
7 Differential Channels
2 Differential Channels with Programmable Gain at 1x, 10x, or 200x
– Byte-oriented Two-wire Serial Interface
– Dual Programmable Serial USARTs
– Master/Slave SPI Serial Interface
– Programmable Watchdog Timer with On-chip Oscillator
– On-chip Analog Comparator
• Special Microcontroller Features
– Power-on Reset and Programmable Brown-out Detection
– Internal Calibrated RC Oscillator
– External and Internal Interrupt Sources
– Six Sleep Modes: Idle, ADC Noise Reduction, Power-save, Power-down, Standby,
and Extended Standby
– Software Selectable Clock Frequency
– ATmega103 Compatibility Mode Selected by a Fuse
– Global Pull-up Disable
• I/O and Packages
– 53 Programmable I/O Lines
– 64-lead TQFP
• Operating Voltages
– 2.7 - 5.5V for ATmega128L
– 4.5 - 5.5V for ATmega128
• Speed Grades
– 0 - 8 MHz for ATmega128L Rev. 2467E–AVR–05/02
– 0 - 16 MHz for ATmega128
1
Pin Configurations Figure 1. Pinout ATmega128
PF5 (ADC5/TMS)
PF6 (ADC6/TDO)
PF4 (ADC4/TCK)
PF7 (ADC7/TDI)
PF0 (ADC0)
PF1 (ADC1)
PF2 (ADC2)
PF3 (ADC3)
PA0 (AD0)
PA1 (AD1)
PA2 (AD2)
AVCC
AREF
GND
GND
VCC
64
63
62
61
60
59
58
57
56
55
54
53
52
51
50
49
PEN 1 48 PA3 (AD3)
RXD0/(PDI) PE0 2 47 PA4 (AD4)
(TXD0/PDO) PE1 3 46 PA5 (AD5)
(XCK0/AIN0) PE2 4 45 PA6 (AD6)
(OC3A/AIN1) PE3 5 44 PA7 (AD7)
(OC3B/INT4) PE4 6 43 PG2(ALE)
(OC3C/INT5) PE5 7 42 PC7 (A15)
(T3/INT6) PE6 8 41 PC6 (A14)
(IC3/INT7) PE7 9 40 PC5 (A13)
(SS) PB0 10 39 PC4 (A12)
(SCK) PB1 11 38 PC3 (A11)
(MOSI) PB2 12 37 PC2 (A10)
(MISO) PB3 13 36 PC1 (A9)
(OC0) PB4 14 35 PC0 (A8)
(OC1A) PB5 15 34 PG1(RD)
(OC1B) PB6 16 33 PG0(WR)
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
(OC2/OC1C) PB7
TOSC2/PG3
TOSC1/1PG4
RESET
VCC
GND
XTAL2
XTAL1
(SCL/INT0) PD0
(SDA/INT1) PD1
(RXD1/INT2) PD2
(TXD1/INT3) PD3
(IC1) PD4
(XCK1) PD5
(T1) PD6
(T2) PD7
Overview The ATmega128 is a low-power CMOS 8-bit microcontroller based on the AVR
enhanced RISC architecture. By executing powerful instructions in a single clock cycle,
the ATmega128 achieves throughputs approaching 1 MIPS per MHz allowing the sys-
tem designer to optimize power consumption versus processing speed.
2 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Block Diagram
Figure 2. Block Diagram
RESET
XTAL2
XTAL1
PF0 - PF7 PA0 - PA7 PC0 - PC7
VCC
GND
PORTF DRIVERS PORTA DRIVERS PORTC DRIVERS
DATA REGISTER DATA DIR. DATA REGISTER DATA DIR. DATA REGISTER DATA DIR.
PORTF REG. PORTF PORTA REG. PORTA PORTC REG. PORTC
8-BIT DATA BUS
AVCC
CALIB. OSC
ADC INTERNAL
AGND OSCILLATOR
AREF
OSCILLATOR
PROGRAM STACK WATCHDOG

JTAG TAP
COUNTER POINTER TIMER
OSCILLATOR
PROGRAM SRAM MCU CONTROL TIMING AND

ON-CHIP DEBUG FLASH REGISTER CONTROL
BOUNDARY-
INSTRUCTION TIMER/
SCAN GENERAL
REGISTER COUNTERS
PURPOSE
REGISTERS
X
PROGRAMMING
PEN INSTRUCTION Y INTERRUPT
LOGIC
DECODER Z UNIT
CONTROL
LINES ALU EEPROM
STATUS
REGISTER
USART0 SPI TWO-WIRE SERIAL

USART1
INTERFACE
COMPARATOR
ANALOG
DATA REGISTER DATA DIR. DATA REGISTER DATA DIR. DATA REGISTER DATA DIR. DATA REG. DATA DIR.
PORTE REG. PORTE PORTB REG. PORTB PORTD REG. PORTD PORTG REG. PORTG
+
-
PORTE DRIVERS PORTB DRIVERS PORTD DRIVERS PORTG DRIVERS
PE0 - PE7 PB0 - PB7 PD0 - PD7 PG0 - PG4
3
2467E–AVR–05/02
The AVR core combines a rich instruction set with 32 general purpose working registers.
All the 32 registers are directly connected to the Arithmetic Logic Unit (ALU), allowing
two independent registers to be accessed in one single instruction executed in one clock
cycle. The resulting architecture is more code efficient while achieving throughputs up to
ten times faster than conventional CISC microcontrollers.
The ATmega128 provides the following features: 128K bytes of In-System Programma-
ble Flash with Read-While-Write capabilities, 4K bytes EEPROM, 4K bytes SRAM, 53
general purpose I/O lines, 32 general purpose working registers, Real Time Counter
(RTC), four flexible Timer/Counters with compare modes and PWM, 2 USARTs, a byte
oriented Two-wire Serial Interface, an 8-channel, 10-bit ADC with optional differential
input stage with programmable gain, programmable Watchdog Timer with Internal Oscil-
lator, an SPI serial port, IEEE std. 1149.1 compliant JTAG test interface, also used for
accessing the On-chip Debug system and programming and six software selectable
power saving modes. The Idle mode stops the CPU while allowing the SRAM,
Timer/Counters, SPI port, and interrupt system to continue functioning. The Power-
down mode saves the register contents but freezes the OscillatorOscillator, disabling all
other chip functions until the next interrupt or Hardware Reset. In Power-save mode, the
asynchronous timer continues to run, allowing the user to maintain a timer base while
the rest of the device is sleeping. The ADC Noise Reduction mode stops the CPU and
all I/O modules except Asynchronous Timer and ADC, to minimize switching noise dur-
ing ADC conversions. In Standby mode, the Crystal/Resonator Oscillator is running
while the rest of the device is sleeping. This allows very fast start-up combined with low
power consumption. In Extended Standby mode, both the main Oscillator and the Asyn-
chronous Timer continue to run.
The device is manufactured using Atmel’s high-density nonvolatile memory technology.
The On-chip ISP Flash allows the program memory to be reprogrammed in-system
through an SPI serial interface, by a conventional nonvolatile memory programmer, or
by an On-chip Boot program running on the AVR core. The boot program can use any
interface to download the application program in the application Flash memory. Soft-
ware in the Boot Flash section will continue to run while the Application Flash section is
updated, providing true Read-While-Write operation. By combining an 8-bit RISC CPU
with In-System Self-Programmable Flash on a monolithic chip, the Atmel ATmega128 is
a powerful microcontroller that provides a highly flexible and cost effective solution to
many embedded control applications.
The ATmega128 AVR is supported with a full suite of program and system development
tools including: C compilers, macro assemblers, program debugger/simulators, in-circuit
emulators, and evaluation kits.
ATmega103 and The ATmega128 is a highly complex microcontroller where the number of I/O locations
ATmega128 supersedes the 64 I/O locations reserved in the AVR instruction set. To ensure back-
Compatibility ward compatibility with the ATmega103, all I/O locations present in ATmega103 have
the same location in ATmega128. Most additional I/O locations are added in an
Extended I/O space starting from $60 to $FF, (i.e., in the ATmega103 internal RAM
space). These locations can be reached by using LD/LDS/LDD and ST/STS/STD
instructions only, not by using IN and OUT instructions. The relocation of the internal
RAM space may still be a problem for ATmega103 users. Also, the increased number of
interrupt vectors might be a problem if the code uses absolute addresses. To solve
these problems, an ATmega103 compatibility mode can be selected by programming
the fuse M103C. In this mode, none of the functions in the Extended I/O space are in
use, so the internal RAM is located as in ATmega103. Also, the Extended Interrupt vec-
tors are removed.
4 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
The ATmega128 is 100% pin compatible with ATmega103, and can replace the
ATmega103 on current Printed Circuit Boards. The application note “Replacing
ATmega103 by ATmega128” describes what the user should be aware of replacing the
ATmega103 by an ATmega128.
ATmega103 Compatibility By programming the M103C fuse, the ATmega128 will be compatible with the
Mode ATmega103 regards to RAM, I/O pins and interrupt vectors as described above. How-
ever, some new features in ATmega128 are not available in this compatibility mode,
these features are listed below:
• One USART instead of two, Asynchronous mode only. Only the eight least
significant bits of the Baud Rate Register is available.
• One 16 bits Timer/Counter with two compare registers instead of two 16-bit
Timer/Counters with three compare registers.
• Two-wire serial interface is not supported.
• Port C is output only.
• Port G serves alternate functions only (not a general I/O port).
• Port F serves as digital input only in addition to analog input to the ADC.
• Boot Loader capabilities is not supported.
• It is not possible to adjust the frequency of the internal calibrated RC Oscillator.
• The External Memory Interface can not release any Address pins for general I/O,
neither configure different wait-states to different External Memory Address
sections.
In addition, there are some other minor differences to make it more compatible to
ATmega103:
• Only EXTRF and PORF exists in MCUCSR.
• Timed sequence not required for Watchdog Time-out change.
• External Interrupt pins 3 - 0 serve as level interrupt only.
• USART has no FIFO buffer, so data overrun comes earlier.
Unused I/O bits in ATmega103 should be written to 0 to ensure same operation in
ATmega128.
Pin Descriptions
VCC Digital supply voltage.
GND Ground.
Port A (PA7..PA0) Port A is an 8-bit bi-directional I/O port with internal pull-up resistors (selected for each
bit). The Port A output buffers have symmetrical drive characteristics with both high sink
and source capability. As inputs, Port A pins that are externally pulled low will source
current if the pull-up resistors are activated. The Port A pins are tri-stated when a reset
condition becomes active, even if the clock is not running.
Port A also serves the functions of various special features of the ATmega128 as listed
on page 67.
Port B (PB7..PB0) Port B is an 8-bit bi-directional I/O port with internal pull-up resistors (selected for each
bit). The Port B output buffers have symmetrical drive characteristics with both high sink
and source capability. As inputs, Port B pins that are externally pulled low will source
5
2467E–AVR–05/02
current if the pull-up resistors are activated. The Port B pins are tri-stated when a reset
Port B also serves the functions of various special features of the ATmega128 as listed
on page 68.
Port C (PC7..PC0) Port C is an 8-bit bi-directional I/O port with internal pull-up resistors (selected for each
bit). The Port C output buffers have symmetrical drive characteristics with both high sink
and source capability. As inputs, Port C pins that are externally pulled low will source
current if the pull-up resistors are activated. The Port C pins are tri-stated when a reset
Port C also serves the functions of special features of the ATmega128 as listed on page
71. In ATmega103 compatibility mode, Port C is output only, and the port C pins are not
tri-stated when a reset condition becomes active.
Port D (PD7..PD0) Port D is an 8-bit bi-directional I/O port with internal pull-up resistors (selected for each
bit). The Port D output buffers have symmetrical drive characteristics with both high sink
and source capability. As inputs, Port D pins that are externally pulled low will source
current if the pull-up resistors are activated. The Port D pins are tri-stated when a reset
Port D also serves the functions of various special features of the ATmega128 as listed
on page 72.
Port E (PE7..PE0) Port E is an 8-bit bi-directional I/O port with internal pull-up resistors (selected for each
bit). The Port E output buffers have symmetrical drive characteristics with both high sink
and source capability. As inputs, Port E pins that are externally pulled low will source
current if the pull-up resistors are activated. The Port E pins are tri-stated when a reset
Port E also serves the functions of various special features of the ATmega128 as listed
on page 75.
Port F (PF7..PF0) Port F serves as the analog inputs to the A/D Converter.
Port F also serves as an 8-bit bi-directional I/O port, if the A/D Converter is not used.
Port pins can provide internal pull-up resistors (selected for each bit). The Port F output
buffers have symmetrical drive characteristics with both high sink and source capability.
As inputs, Port F pins that are externally pulled low will source current if the pull-up
resistors are activated. The Port F pins are tri-stated when a reset condition becomes
active, even if the clock is not running. If the JTAG interface is enabled, the pull-up resis-
tors on pins PF7(TDI), PF5(TMS), and PF4(TCK) will be activated even if a Reset
occurs.
Port F also serves the functions of the JTAG interface.
In ATmega103 compatibility mode, Port F is an input Port only.
Port G (PG4..PG0) Port G is a 5-bit bi-directional I/O port with internal pull-up resistors (selected for each
bit). The Port G output buffers have symmetrical drive characteristics with both high sink
and source capability. As inputs, Port G pins that are externally pulled low will source
current if the pull-up resistors are activated. The Port G pins are tri-stated when a reset
Port G also serves the functions of various special features.
6 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
The port G pins are tri-stated when a reset condition becomes active, even if the clock is
not running.
In ATmega103 compatibility mode, these pins only serves as strobes signals to the
external memory as well as input to the 32 kHz Oscillator, and the pins are initialized to
PG0 = 1, PG1 = 1, and PG2 = 0 asynchronously when a reset condition becomes active,
even if the clock is not running. PG3 and PG4 are oscillator pins.
RESET Reset input. A low level on this pin for longer than the minimum pulse length will gener-
ate a reset, even if the clock is not running. The minimum pulse length is given in Table
19 on page 46. Shorter pulses are not guaranteed to generate a reset.
XTAL1 Input to the inverting Oscillator amplifier and input to the internal clock operating circuit.
XTAL2 Output from the inverting Oscillator amplifier.
AVCC AVCC is the supply voltage pin for Port F and the A/D Converter. It should be externally
connected to VCC, even if the ADC is not used. If the ADC is used, it should be con-
nected to VCC through a low-pass filter.
AREF AREF is the analog reference pin for the A/D Converter.
PEN PEN is a programming enable pin for the SPI Serial Programming mode. By holding this
pin low during a Power-on Reset, the device will enter the SPI Serial Programming
mode. PEN has no function during normal operation.
About Code This datasheet contains simple code examples that briefly show how to use various
parts of the device. These code examples assume that the part specific header file is
Examples
included before compilation. Be aware that not all C compiler vendors include bit defini-
tions in the header files and interrupt handling in C is compiler dependent. Please
confirm with the C compiler documentation for more details.
7
2467E–AVR–05/02
Edited by Foxit Reader
Copyright(C) by Foxit Corporation,2005-2009
For Evaluation Only.
AVR CPU Core
Introduction This section discusses the AVR core architecture in general. The main function of the
CPU core is to ensure correct program execution. The CPU must therefore be able to
access memories, perform calculations, control peripherals and handle interrupts.
Architectural Overview Figure 3. Block Diagram of the AVR Architecture
Data Bus 8-bit
Program Status
Flash
Counter and Control
Program
Memory
Interrupt
32 x 8 Unit
Instruction General
Register Purpose SPI
Registrers Unit
Instruction Watchdog
Decoder Timer
Indirect Addressing
Direct Addressing
ALU Analog
Control Lines Comparator
I/O Module1
Data I/O Module 2

SRAM
I/O Module n
EEPROM
I/O Lines
In order to maximize performance and parallelism, the AVR uses a Harvard architecture
– with separate memories and buses for program and data. Instructions in the program
memory are executed with a single level pipelining. While one instruction is being exe-
cuted, the next instruction is pre-fetched from the program memory. This concept
enables instructions to be executed in every clock cycle. The program memory is In-
System Reprogrammable Flash memory.
The fast-access Register file contains 32 x 8-bit general purpose working registers with
a single clock cycle access time. This allows single-cycle Arithmetic Logic Unit (ALU)
operation. In a typical ALU operation, two operands are output from the Register file, the
operation is executed, and the result is stored back in the Register file – in one clock
cycle.
Six of the 32 registers can be used as three 16-bit indirect address register pointers for
Data Space addressing – enabling efficient address calculations. One of the these
address pointers can also be used as an address pointer for look up tables in Flash Pro-
gram memory. These added function registers are the 16-bit X-register, Y-register and
Z-register, described later in this section.
The ALU supports arithmetic and logic operations between registers or between a con-
stant and a register. Single register operations can also be executed in the ALU. After
8 ATmega128(L)
2467E–AVR–05/02
Edited by Foxit Reader
ATmega128(L)
Copyright(C) by Foxit Corporation,2005-2009
For Evaluation Only.
an arithmetic operation, the Status Register is updated to reflect information about the
result of the operation.
Program flow is provided by conditional and unconditional jump and call instructions,
able to directly address the whole address space. Most AVR instructions have a single
16-bit word format. Every program memory address contains a 16- or 32-bit instruction.
Program Flash memory space is divided in two sections, the Boot Program section and
the Application Program section. Both sections have dedicated Lock bits for write and
read/write protection. The SPM instruction that writes into the Application Flash Memory
section must reside in the Boot Program section.
During interrupts and subroutine calls, the return address Program Counter (PC) is
stored on the Stack. The Stack is effectively allocated in the general data SRAM, and
consequently the stack size is only limited by the total SRAM size and the usage of the
SRAM. All user programs must initialize the SP in the reset routine (before subroutines
or interrupts are executed). The Stack Pointer – SP – is read/write accessible in the I/O
space. The data SRAM can easily be accessed through the five different addressing
modes supported in the AVR architecture.
The memory spaces in the AVR architecture are all linear and regular memory maps.
A flexible interrupt module has its control registers in the I/O space with an additional
global interrupt enable bit in the Status Register. All interrupts have a separate interrupt
vector in the interrupt vector table. The interrupts have priority in accordance with their
interrupt vector position. The lower the interrupt vector address, the higher the priority.
The I/O memory space contains 64 addresses which can be accessed directly, or as the
Data Space locations following those of the Register file, $20 - $5F. In addition, the
ATmega128 has Extended I/O space from $60 - $FF in SRAM where only the
ST/STS/STD and LD/LDS/LDD instructions can be used.
ALU – Arithmetic Logic The high-performance AVR ALU operates in direct connection with all the 32 general
Unit purpose working registers. Within a single clock cycle, arithmetic operations between
general purpose registers or between a register and an immediate are executed. The
ALU operations are divided into three main categories – arithmetic, logical, and bit-func-
tions. Some implementations of the architecture also provide a powerful multiplier
supporting both signed/unsigned multiplication and fractional format. See the “Instruc-
tion Set” section for a detailed description.
Status Register The Status Register contains information about the result of the most recently executed
arithmetic instruction. This information can be used for altering program flow in order to
perform conditional operations. Note that the Status Register is updated after all ALU
operations, as specified in the Instruction Set Reference. This will in many cases
remove the need for using the dedicated compare instructions, resulting in faster and
more compact code.
The status register is not automatically stored when entering an interrupt routine and
restored when returning from an interrupt. This must be handled by software.
The AVR status Register – SREG – is defined as:
Bit 7 6 5 4 3 2 1 0
I T H S V N Z C SREG
Read/Write R/W R/W R/W R/W R/W R/W R/W R/W
Initial Value 0 0 0 0 0 0 0 0
9
2467E–AVR–05/02
• Bit 7 – I: Global Interrupt Enable
The Global Interrupt Enable bit must be set for the interrupts to be enabled. The individ-
ual interrupt enable control is then performed in separate control registers. If the Global
Interrupt Enable Register is cleared, none of the interrupts are enabled independent of
the individual interrupt enable settings. The I-bit is cleared by hardware after an interrupt
has occurred, and is set by the RETI instruction to enable subsequent interrupts. The I-
bit can also be set and cleared in software with the SEI and CLI instructions, as
described in the instruction set reference.
• Bit 6 – T: Bit Copy Storage

The Bit Copy instructions BLD (Bit LoaD) and BST (Bit STore) use the T-bit as source or
destination for the operated bit. A bit from a register in the Register file can be copied
into T by the BST instruction, and a bit in T can be copied into a bit in a register in the
Register file by the BLD instruction.
• Bit 5 – H: Half Carry Flag

The Half Carry Flag H indicates a half carry in some arithmetic operations. Half carry is
useful in BCD arithmetic. See the “Instruction Set Description” for detailed information.
• Bit 4 – S: Sign Bit, S = N ⊕ V

The S-bit is always an exclusive or between the negative flag N and the two’s comple-
ment overflow flag V. See the “Instruction Set Description” for detailed information.
• Bit 3 – V: Two’s Complement Overflow Flag

The Two’s Complement Overflow Flag V supports two’s complement arithmetics. See
the “Instruction Set Description” for detailed information.
• Bit 2 – N: Negative Flag

The Negative Flag N indicates a negative result in an arithmetic or logic operation. See
the “Instruction Set Description” for detailed information.
• Bit 1 – Z: Zero Flag

The Zero Flag Z indicates a zero result in an arithmetic or logic operation. See the
“Instruction Set Description” for detailed information.
• Bit 0 – C: Carry Flag

The Carry Flag C indicates a carry in an arithmetic or logic operation. See the “Instruc-
tion Set Description” for detailed information.
General Purpose The Register file is optimized for the AVR Enhanced RISC instruction set. In order to
Register File achieve the required performance and flexibility, the following input/output schemes are
supported by the Register file:
• One 8-bit output operand and one 8-bit result input
• Two 8-bit output operands and one 8-bit result input
• Two 8-bit output operands and one 16-bit result input
• One 16-bit output operand and one 16-bit result input
Figure 4 shows the structure of the 32 general purpose working registers in the CPU.
10 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 4. AVR CPU General Purpose Working Registers
7 0 Addr.
R0 $00
R1 $01
R2 $02
…
R13 $0D
General R14 $0E
Purpose R15 $0F
Working R16 $10
Registers R17 $11
…
R26 $1A X-register Low Byte
R27 $1B X-register High Byte
R28 $1C Y-register Low Byte
R29 $1D Y-register High Byte
R30 $1E Z-register Low Byte
R31 $1F Z-register High Byte
Most of the instructions operating on the Register file have direct access to all registers,
and most of them are single cycle instructions.
As shown in Figure 4, each register is also assigned a data memory address, mapping
them directly into the first 32 locations of the user Data Space. Although not being phys-
ically implemented as SRAM locations, this memory organization provides great
flexibility in access of the registers, as the X-, Y-, and Z-pointer Registers can be set to
index any register in the file.
X-register, Y-register, and Z- The registers R26..R31 have some added functions to their general purpose usage.
register These registers are 16-bit address pointers for indirect addressing of the Data Space.
The three indirect address registers X, Y, and Z are described in Figure 5.
Figure 5. The X-, Y-, and Z-registers
15 XH XL 0
X - register 7 0 7 0
R27 ($1B) R26 ($1A)
15 YH YL 0
Y - register 7 0 7 0
R29 ($1D) R28 ($1C)
15 ZH ZL 0
Z - register 7 0 7 0
R31 ($1F) R30 ($1E)
In the different addressing modes these address registers have functions as fixed dis-
placement, automatic increment, and automatic decrement (see the Instruction Set
Reference for details).
11
2467E–AVR–05/02
Stack Pointer The Stack is mainly used for storing temporary data, for storing local variables and for
storing return addresses after interrupts and subroutine calls. The Stack Pointer Regis-
ter always points to the top of the Stack. Note that the Stack is implemented as growing
from higher memory locations to lower memory locations. This implies that a Stack
PUSH command decreases the Stack Pointer.
The Stack Pointer points to the data SRAM stack area where the Subroutine and Inter-
rupt Stacks are located. This Stack space in the data SRAM must be defined by the
program before any subroutine calls are executed or interrupts are enabled. The Stack
Pointer must be set to point above $60. The Stack Pointer is decremented by one when
data is pushed onto the Stack with the PUSH instruction, and it is decremented by two
when the return address is pushed onto the Stack with subroutine call or interrupt. The
Stack Pointer is incremented by one when data is popped from the Stack with the POP
instruction, and it is incremented by two when data is popped from the Stack with return
from subroutine RET or return from interrupt RETI.
The AVR Stack Pointer is implemented as two 8-bit registers in the I/O space. The num-
ber of bits actually used is implementation dependent. Note that the data space in some
implementations of the AVR architecture is so small that only SPL is needed. In this
case, the SPH Register will not be present.
Bit 15 14 13 12 11 10 9 8
SP15 SP14 SP13 SP12 SP11 SP10 SP9 SP8 SPH
SP7 SP6 SP5 SP4 SP3 SP2 SP1 SP0 SPL
7 6 5 4 3 2 1 0
R/W R/W R/W R/W R/W R/W R/W R/W
0 0 0 0 0 0 0 0
RAM Page Z Select Register –

RAMPZ Bit 7 6 5 4 3 2 1 0
– –– – – – – – RAMPZ0 RAMPZ
Read/Write R R R R R R R R/W
• Bits 7..2 – Res: Reserved Bits

These are reserved bits and will always read as zero. When writing to this address loca-
tion, write these bits to zero for compatibility with future devices.
• Bit 1 – RAMPZ0: Extended RAM Page Z-pointer

The RAMPZ Register is normally used to select which 64K RAM Page is accessed by
the Z-pointer. As the ATmega128 does not support more than 64K of SRAM memory,
this register is used only to select which page in the program memory is accessed when
the ELPM/SPM instruction is used. The different settings of the RAMPZ0 bit have the
following effects:
RAMPZ0 = 0: Program memory address $0000 - $7FFF (lower 64K bytes) is

accessed by ELPM/SPM
RAMPZ0 = 1: Program memory address $8000 - $FFFF (higher 64K bytes) is
accessed by ELPM/SPM
Note that LPM is not affected by the RAMPZ setting.
12 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Instruction Execution This section describes the general access timing concepts for instruction execution. The
Timing AVR CPU is driven by the CPU clock clkCPU, directly generated from the selected clock
source for the chip. No internal clock division is used.
Figure 6 shows the parallel instruction fetches and instruction executions enabled by the
Harvard architecture and the fast-access Register file concept. This is the basic pipelin-
ing concept to obtain up to 1 MIPS per MHz with the corresponding unique results for
functions per cost, functions per clocks, and functions per power-unit.
Figure 6. The Parallel Instruction Fetches and Instruction Executions

T1 T2 T3 T4
clkCPU
1st Instruction Fetch
1st Instruction Execute
2nd Instruction Fetch
2nd Instruction Execute
3rd Instruction Fetch
3rd Instruction Execute
4th Instruction Fetch
Figure 7 shows the internal timing concept for the Register file. In a single clock cycle an
ALU operation using two register operands is executed, and the result is stored back to
the destination register.
Figure 7. Single Cycle ALU Operation

T1 T2 T3 T4
clkCPU
Total Execution Time
Register Operands Fetch
ALU Operation Execute
Result Write Back
Reset and Interrupt The AVR provides several different interrupt sources. These interrupts and the separate
Handling reset vector each have a separate program vector in the program memory space. All
interrupts are assigned individual enable bits which must be written logic one together
with the Global Interrupt Enable bit in the Status Register in order to enable the interrupt.
Depending on the Program Counter value, interrupts may be automatically disabled
when Boot Lock bits BLB02 or BLB12 are programmed. This feature improves software
security. See the section “Memory Programming” on page 282 for details.
The lowest addresses in the program memory space are by default defined as the Reset
and Interrupt vectors. The complete list of vectors is shown in “Interrupts” on page 54.
The list also determines the priority levels of the different interrupts. The lower the
address the higher is the priority level. RESET has the highest priority, and next is INT0
– the External Interrupt Request 0. The interrupt vectors can be moved to the start of the
boot Flash section by setting the IVSEL bit in the MCU Control Register (MCUCR).
Refer to “Interrupts” on page 54 for more information. The Reset vector can also be
moved to the start of the boot Flash section by programming the BOOTRST fuse, see
“Boot Loader Support – Read-While-Write Self-Programming” on page 269.
13
2467E–AVR–05/02
When an interrupt occurs, the Global Interrupt Enable I-bit is cleared and all interrupts
are disabled. The user software can write logic one to the I-bit to enable nested inter-
rupts. All enabled interrupts can then interrupt the current interrupt routine. The I-bit is
automatically set when a Return from Interrupt instruction – RETI – is executed.
There are basically two types of interrupts. The first type is triggered by an event that
sets the interrupt flag. For these interrupts, the Program Counter is vectored to the
actual interrupt vector in order to execute the interrupt handling routine, and hardware
clears the corresponding interrupt flag. Interrupt flags can also be cleared by writing a
logic one to the flag bit position(s) to be cleared. If an interrupt condition occurs while the
corresponding interrupt enable bit is cleared, the interrupt flag will be set and remem-
bered until the interrupt is enabled, or the flag is cleared by software. Similarly, if one or
more interrupt conditions occur while the global interrupt enable bit is cleared, the corre-
sponding interrupt flag(s) will be set and remembered until the global interrupt enable bit
is set, and will then be executed by order of priority.
The second type of interrupts will trigger as long as the interrupt condition is present.
These interrupts do not necessarily have interrupt flags. If the interrupt condition disap-
pears before the interrupt is enabled, the interrupt will not be triggered.
When the AVR exits from an interrupt, it will always return to the main program and exe-
cute one more instruction before any pending interrupt is served.
Note that the Status Register is not automatically stored when entering an interrupt rou-
tine, nor restored when returning from an interrupt routine. This must be handled by
software.
When using the CLI instruction to disable interrupts, the interrupts will be immediately
disabled. No interrupt will be executed after the CLI instruction, even if it occurs simulta-
neously with the CLI instruction. The following example shows how this can be used to
avoid interrupts during the timed EEPROM write sequence.
Assembly Code Example
in r16, SREG ; store SREG value
cli ; disable interrupts during timed sequence
sbi EECR, EEMWE ; start EEPROM write
sbi EECR, EEWE
out SREG, r16 ; restore SREG value (I-bit)
C Code Example
char cSREG;
cSREG = SREG; /* store SREG value */
/* disable interrupts during timed sequence */
_CLI();
EECR |= (1<<EEMWE); /* start EEPROM write */
EECR |= (1<<EEWE);
SREG = cSREG; /* restore SREG value (I-bit) */
14 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
When using the SEI instruction to enable interrupts, the instruction following SEI will be
executed before any pending interrupts, as shown in this example.
sei ; set global interrupt enable
sleep ; enter sleep, waiting for interrupt
; note: will enter sleep before any pending
; interrupt(s)
C Code Example
_SEI(); /* set global interrupt enable */
_SLEEP(); /* enter sleep, waiting for interrupt */
/* note: will enter sleep before any pending interrupt(s) */
Interrupt Response Time The interrupt execution response for all the enabled AVR interrupts is four clock cycles
minimum. After four clock cycles, the program vector address for the actual interrupt
handling routine is executed. During this 4-clock cycle period, the Program Counter is
pushed onto the Stack. The vector is normally a jump to the interrupt routine, and this
jump takes three clock cycles. If an interrupt occurs during execution of a multi-cycle
instruction, this instruction is completed before the interrupt is served. If an interrupt
occurs when the MCU is in Sleep mode, the interrupt execution response time is
increased by four clock cycles. This increase comes in addition to the start-up time from
the selected sleep mode.
A return from an interrupt handling routine takes four clock cycles. During these 4-clock
cycles, the Program Counter (two bytes) is popped back from the Stack, the Stack
Pointer is incremented by two, and the I-bit in SREG is set.
15
2467E–AVR–05/02
AVR ATmega128 This section describes the different memories in the ATmega128. The AVR architecture
has two main memory spaces, the Data Memory and the Program Memory space. In
Memories
addition, the ATmega128 features an EEPROM Memory for data storage. All three
memory spaces are linear and regular.
In-System The ATmega128 contains 128K bytes On-chip In-System Reprogrammable Flash mem-
Reprogrammable Flash ory for program storage. Since all AVR instructions are 16 or 32 bits wide, the Flash is
Program Memory organized as 64K x 16. For software security, the Flash Program memory space is
divided into two sections, Boot Program section and Application Program section.
The Flash memory has an endurance of at least 1,000 write/erase cycles. The
ATmega128 Program Counter (PC) is 16 bits wide, thus addressing the 64K program
memory locations. The operation of Boot Program section and associated Boot Lock
bits for software protection are described in detail in “Boot Loader Support – Read-
While-Write Self-Programming” on page 269. “Memory Programming” on page 282 con-
tains a detailed description on Flash data serial downloading using the SPI pins or the
JTAG interface.
Constant tables can be allocated within the entire program memory address space (see
the LPM – Load Program Memory and ELPM – Extended Load Program Memory
instruction description).
Timing diagrams for instruction fetch and execution are presented in “Instruction Execu-
tion Timing” on page 13.
Figure 8. Program Memory Map

Program Memory
$0000
Application Flash Section
Boot Flash Section

$FFFF
16 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
SRAM Data Memory The ATmega128 supports two different configurations for the SRAM data memory as
listed in Table 1.
Table 1. Memory Configurations

Configuration Internal SRAM Data Memory External SRAM Data Memory
Normal mode 4096 up to 64K
ATmega103 Compatibility
4000 up to 64K
mode
Figure 9 shows how the ATmega128 SRAM Memory is organized.

The ATmega128 is a complex microcontroller with more peripheral units than can be
supported within the 64 location reserved in the Opcode for the IN and OUT instructions.
For the Extended I/O space from $60 - $FF in SRAM, only the ST/STS/STD and
LD/LDS/LDD instructions can be used. The Extended I/O space does not exist when the
ATmega128 is in the ATmega103 compatibility mode.
In normal mode, the first 4352 Data Memory locations address both the Register file, the
I/O Memory, Extended I/O Memory, and the internal data SRAM. The first 32 locations
address the Register file, the next 64 location the standard I/O memory, then 160 loca-
tions of Extended I/O memory, and the next 4096 locations address the internal data
SRAM.
In ATmega103 compatibility mode, the first 4096 Data Memory locations address both
the Register file, the I/O Memory and the internal data SRAM. The first 32 locations
address the Register file, the next 64 location the standard I/O memory, and the next
4000 locations address the internal data SRAM.
An optional external data SRAM can be used with the ATmega128. This SRAM will
occupy an area in the remaining address locations in the 64K address space. This area
starts at the address following the internal SRAM. The Register file, I/O, Extended I/O
and Internal SRAM occupies the lowest 4352 bytes in normal mode, and the lowest
4096 bytes in the ATmega103 compatibility mode (Extended I/O not present), so when
using 64KB (65536 bytes) of External Memory, 61184 Bytes of External Memory are
available in normal mode, and 61440 Bytes in ATmega103 compatibility mode. See
“External Memory Interface” on page 24 for details on how to take advantage of the
external memory map.
When the addresses accessing the SRAM memory space exceeds the internal data
memory locations, the external data SRAM is accessed using the same instructions as
for the internal data memory access. When the internal data memories are accessed,
the read and write strobe pins (PG0 and PG1) are inactive during the whole access
cycle. External SRAM operation is enabled by setting the SRE bit in the MCUCR
Register.
Accessing external SRAM takes one additional clock cycle per byte compared to access
of the internal SRAM. This means that the commands LD, ST, LDS, STS, LDD, STD,
PUSH, and POP take one additional clock cycle. If the Stack is placed in external
SRAM, interrupts, subroutine calls and returns take three clock cycles extra because the
two-byte program counter is pushed and popped, and external memory access does not
take advantage of the internal pipe-line memory access. When external SRAM interface
is used with wait-state, one-byte external access takes two, three, or four additional
clock cycles for one, two, and three wait-states respectively. Interrupts, subroutine calls
and returns will need five, seven, or nine clock cycles more than specified in the instruc-
tion set manual for one, two, and three wait-states.
17
2467E–AVR–05/02
The five different addressing modes for the data memory cover: Direct, Indirect with Dis-
placement, Indirect, Indirect with Pre-decrement, and Indirect with Post-increment. In
the Register file, registers R26 to R31 feature the indirect addressing pointer registers.
The direct addressing reaches the entire data space.
The Indirect with Displacement mode reaches 63 address locations from the base
address given by the Y- or Z-register.
When using register indirect addressing modes with automatic pre-decrement and post-
increment, the address registers X, Y, and Z are decremented or incremented.
The 32 general purpose working registers, 64 I/O registers, and the 4096 bytes of inter-
nal data SRAM in the ATmega128 are all accessible through all these addressing
modes. The Register file is described in “General Purpose Register File” on page 10.
Figure 9. Data Memory Map
Memory Configuration A Memory Configuration B
Data Memory Data Memory

32 Registers $0000 - $001F 32 Registers $0000 - $001F
64 I/O Registers $0020 - $005F 64 I/O Registers $0020 - $005F
160 Ext I/O Reg. $0060 - $00FF $0060
$0100 Internal SRAM
Internal SRAM (4000 x 8)
(4096 x 8) $0FFF
$1000
$10FF
$1100
External SRAM External SRAM

(0 - 64K x 8) (0 - 64K x 8)
$FFFF $FFFF
18 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Data Memory Access Times This section describes the general access timing concepts for internal memory access.
The internal data SRAM access is performed in two clkCPU cycles as described in Figure
10.
Figure 10. On-chip Data SRAM Access Cycles

T1 T2 T3
clkCPU
Address Compute Address Address valid
Data
Write
WR
Data
Read
RD
Memory access instruction Next instruction
EEPROM Data Memory The ATmega128 contains 4K bytes of data EEPROM memory. It is organized as a sep-
arate data space, in which single bytes can be read and written. The EEPROM has an
endurance of at least 100,000 write/erase cycles. The access between the EEPROM
and the CPU is described in the following, specifying the EEPROM Address Registers,
the EEPROM Data Register, and the EEPROM Control Register.
For a detailed description of SPI and JTAG data downloading to the EEPROM, see
page 297 and page 302 respectively.
EEPROM Read/Write Access The EEPROM access registers are accessible in the I/O space.
The write access time for the EEPROM is given in Table 2. A self-timing function, how-
ever, lets the user software detect when the next byte can be written. If the user code
contains instructions that write the EEPROM, some precautions must be taken. In
heavily filtered power supplies, VCC is likely to rise or fall slowly on Power-up/down. This
causes the device for some period of time to run at a voltage lower than specified as
minimum for the clock frequency used. See “Preventing EEPROM Corruption” on page
23. for details on how to avoid problems in these situations.
In order to prevent unintentional EEPROM writes, a specific write procedure must be fol-
lowed. Refer to the description of the EEPROM Control Register for details on this.
When the EEPROM is read, the CPU is halted for four clock cycles before the next
instruction is executed. When the EEPROM is written, the CPU is halted for two clock
cycles before the next instruction is executed.
EEPROM Address Register –

EEARH and EEARL Bit 15 14 13 12 11 10 9 8
– – – – EEAR11 EEAR10 EEAR9 EEAR8 EEARH
EEAR7 EEAR6 EEAR5 EEAR4 EEAR3 EEAR2 EEAR1 EEAR0 EEARL
7 6 5 4 3 2 1 0
Read/Write R R R R R/W R/W R/W R/W
Initial Value 0 0 0 0 X X X X
X X X X X X X X
19
2467E–AVR–05/02
• Bits 11..0 – EEAR11..0: EEPROM Address

The EEPROM Address Registers – EEARH and EEARL – specify the EEPROM
address in the 4K bytes EEPROM space. The EEPROM data bytes are addressed lin-
early between 0 and 4096. The initial value of EEAR is undefined. A proper value must
be written before the EEPROM may be accessed.
EEPROM Data Register –

EEDR Bit 7 6 5 4 3 2 1 0
MSB LSB EEDR
• Bits 7..0 – EEDR7.0: EEPROM Data

For the EEPROM write operation, the EEDR Register contains the data to be written to
the EEPROM in the address given by the EEAR Register. For the EEPROM read oper-
ation, the EEDR contains the data read out from the EEPROM at the address given by
EEAR.
EEPROM Control Register –

EECR Bit 7 6 5 4 3 2 1 0
– – – – EERIE EEMWE EEWE EERE EECR
Initial Value 0 0 0 0 0 0 X 0

These bits are reserved bits in the ATmega128 and will always read as zero.
• Bit 3 – EERIE: EEPROM Ready Interrupt Enable

Writing EERIE to one enables the EEPROM Ready Interrupt if the I-bit in SREG is set.
Writing EERIE to zero disables the interrupt. The EEPROM Ready interrupt generates a
constant interrupt when EEWE is cleared.
• Bit 2 – EEMWE: EEPROM Master Write Enable

The EEMWE bit determines whether setting EEWE to one causes the EEPROM to be
written. When EEMWE is written to one, writing EEWE to one within four clock cycles
will write data to the EEPROM at the selected address. If EEMWE is zero, writing EEWE
to one will have no effect. When EEMWE has been written to one by software, hardware
clears the bit to zero after four clock cycles. See the description of the EEWE bit for an
EEPROM write procedure.
• Bit 1 – EEWE: EEPROM Write Enable

The EEPROM Write Enable Signal EEWE is the write strobe to the EEPROM. When
address and data are correctly set up, the EEWE bit must be set to write the value into
the EEPROM. The EEMWE bit must be set when the logical one is written to EEWE,
otherwise no EEPROM write takes place. The following procedure should be followed
when writing the EEPROM (the order of steps 3 and 4 is not essential):
20 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
1. Wait until EEWE becomes zero.

2. Wait until SPMEN in SPMCR becomes zero.
3. Write new EEPROM address to EEAR (optional).
4. Write new EEPROM data to EEDR (optional).
5. Write a logical one to the EEMWE bit while writing a zero to EEWE in EECR.
6. Within four clock cycles after setting EEMWE, write a logical one to EEWE.
The EEPROM can not be programmed during a CPU write to the Flash memory. The
software must check that the Flash programming is completed before initiating a new
EEPROM write. Step 2 is only relevant if the software contains a boot loader allowing
the CPU to program the Flash. If the Flash is never being updated by the CPU, step 2
can be omitted. See “Boot Loader Support – Read-While-Write Self-Programming” on
page 269 for details about boot programming.
Caution: An interrupt between step 5 and step 6 will make the write cycle fail, since the
EEPROM Master Write Enable will time-out. If an interrupt routine accessing the
EEPROM is interrupting another EEPROM access, the EEAR or EEDR Register will be
modified, causing the interrupted EEPROM access to fail. It is recommended to have
the global interrupt flag cleared during the four last steps to avoid these problems.
When the write access time has elapsed, the EEWE bit is cleared by hardware. The
user software can poll this bit and wait for a zero before writing the next byte. When
EEWE has been set, the CPU is halted for two cycles before the next instruction is
executed.
• Bit 0 – EERE: EEPROM Read Enable

The EEPROM Read Enable Signal EERE is the read strobe to the EEPROM. When the
correct address is set up in the EEAR Register, the EERE bit must be written to a logic
one to trigger the EEPROM read. The EEPROM read access takes one instruction, and
the requested data is available immediately. When the EEPROM is read, the CPU is
halted for four cycles before the next instruction is executed.
The user should poll the EEWE bit before starting the read operation. If a write operation
is in progress, it is neither possible to read the EEPROM, nor to change the EEAR
Register.
The calibrated Oscillator is used to time the EEPROM accesses. Table 2 lists the typical
programming time for EEPROM access from the CPU.
Table 2. EEPROM Programming Time

Number of Calibrated RC
Symbol Oscillator Cycles(1) Typ Programming Time
EEPROM Write (from CPU) 8448 8.5 ms
Note: 1. Uses 1 MHz clock, independent of CKSEL-fuse settings.
21
2467E–AVR–05/02
The following code examples show one assembly and one C function for writing to the
EEPROM. The examples assume that interrupts are controlled (e.g., by disabling inter-
rupts globally) so that no interrupts will occur during execution of these functions. The
examples also assume that no flash boot loader is present in the software. If such code
is present, the EEPROM write function must also wait for any ongoing SPM command to
finish.
EEPROM_write:
; Wait for completion of previous write
sbic EECR,EEWE
rjmp EEPROM_write
; Set up address (r18:r17) in address register
out EEARH, r18
out EEARL, r17
; Write data (r16) to data register
out EEDR,r16
; Write logical one to EEMWE
sbi EECR,EEMWE
; Start eeprom write by setting EEWE
sbi EECR,EEWE
ret
C Code Example
void EEPROM_write(unsigned int uiAddress, unsigned char ucData)
{
/* Wait for completion of previous write */
while(EECR & (1<<EEWE))
;
/* Set up address and data registers */
EEAR = uiAddress;
EEDR = ucData;
/* Write logical one to EEMWE */
EECR |= (1<<EEMWE);
/* Start eeprom write by setting EEWE */
EECR |= (1<<EEWE);
}
22 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
The next code examples show assembly and C functions for reading the EEPROM. The
examples assume that interrupts are controlled so that no interrupts will occur during
execution of these functions.
EEPROM_read:
; Wait for completion of previous write
sbic EECR,EEWE
rjmp EEPROM_read
; Set up address (r18:r17) in address register
out EEARH, r18
out EEARL, r17
; Start eeprom read by writing EERE
sbi EECR,EERE
; Read data from data register
in r16,EEDR
ret
C Code Example
unsigned char EEPROM_read(unsigned int uiAddress)
{
/* Wait for completion of previous write */
while(EECR & (1<<EEWE))
;
/* Set up address register */
EEAR = uiAddress;
/* Start eeprom read by writing EERE */
EECR |= (1<<EERE);
/* Return data from data register */
return EEDR;
}
Preventing EEPROM During periods of low VCC, the EEPROM data can be corrupted because the supply volt-
Corruption age is too low for the CPU and the EEPROM to operate properly. These issues are the
same as for board level systems using EEPROM, and the same design solutions should
be applied.
An EEPROM data corruption can be caused by two situations when the voltage is too
low. First, a regular write sequence to the EEPROM requires a minimum voltage to
operate correctly. Secondly, the CPU itself can execute instructions incorrectly, if the
supply voltage is too low.
E E P R O M d a t a c o r r u p t io n c a n e a si ly b e a vo id e d b y fo llo w in g th is d e s ig n
recommendation:
Keep the AVR RESET active (low) during periods of insufficient power supply voltage.
This can be done by enabling the internal Brown-out Detector (BOD). If the detection
level of the internal BOD does not match the needed detection level, an external low
VCC Reset Protection circuit can be used. If a reset occurs while a write operation is in
progress, the write operation will be completed provided that the power supply voltage is
sufficient.
23
2467E–AVR–05/02
I/O Memory The I/O space definition of the ATmega128 is shown in “Register Summary” on page
350.
All ATmega128 I/Os and peripherals are placed in the I/O space. All I/O locations may
be accessed by the LD/LDS/LDD and ST/STS/STD instructions, transferring data
between the 32 general purpose working registers and the I/O space. I/O registers
within the address range $00 - $1F are directly bit-accessible using the SBI and CBI
instructions. In these registers, the value of single bits can be checked by using the
SBIS and SBIC instructions. Refer to the instruction set section for more details. When
using the I/O specific commands IN and OUT, the I/O addresses $00 - $3F must be
used. When addressing I/O registers as data space using LD and ST instructions, $20
must be added to these addresses. The ATmega128 is a complex microcontroller with
more peripheral units than can be supported within the 64 location reserved in Opcode
for the IN and OUT instructions. For the Extended I/O space from $60 - $FF in SRAM,
only the ST/STS/STD and LD/LDS/LDD instructions can be used. The Extended I/O
space is replaced with SRAM locations when the ATmega128 is in the ATmega103
compatibility mode.
For compatibility with future devices, reserved bits should be written to zero if accessed.
Reserved I/O memory addresses should never be written.
Some of the status flags are cleared by writing a logical one to them. Note that the CBI
and SBI instructions will operate on all bits in the I/O register, writing a one back into any
flag read as set, thus clearing the flag. The CBI and SBI instructions work with registers
$00 to $1F only.
The I/O and peripherals control registers are explained in later sections.
External Memory With all the features the External Memory Interface provides, it is well suited to operate
Interface as an interface to memory devices such as External SRAM and Flash, and peripherals
such as LCD-display, A/D, and D/A. The main features are:
• Four different wait-state settings (including no wait-state).
• Independent wait-state setting for different extErnal Memory sectors (configurable
sector size).
• The number of bits dedicated to address high byte is selectable.
• Bus keepers on data lines to minimize current consumption (optional).
Overview When the eXternal MEMory (XMEM) is enabled, address space outside the internal
SRAM becomes available using the dedicated External Memory pins (see Figure 1 on
page 2, Table 27 on page 67, Table 33 on page 71, and Table 45 on page 79). The
memory configuration is shown in Figure 11.
24 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 11. External Memory with Sector Select

Memory Configuration A Memory Configuration B
0x0000 0x0000
Internal memory Internal memory
0x0FFF
0x1000
0x10FF
0x1100
Lower sector
SRW01
SRW00
SRL[2..0] SRW10
External Memory Upper sector External Memory

(0-60K x 8) (0-60K x 8)
SRW11
SRW10
0xFFFF 0xFFFF
Note: ATmega128 in non ATmega103 compatibility mode: Memory Configuration A is available

(Memory Configuration B N/A)
ATmega128 in ATmega103 compatibility mode: Memory Configuration B is available
(Memory Configuration A N/A)
ATmega103 Compatibility Both External Memory Control Registers (XMCRA and XMCRB) are placed in Extended
I/O space. In ATmega103 compatibility mode, these registers are not available, and the
features selected by these registers are not available. The device is still ATmega103
compatible, as these features did not exist in ATmega103. The limitations in
ATmega103 compatibility mode are:
• Only two wait-states settings are available (SRW1n = 0b00 and SRW1n = 0b01).
• The number of bits that are assigned to address high byte are fixed.
• The External Memory section can not be divided into sectors with different wait-
state settings.
• Bus-keeper is not available.
• RD, WR and ALE pins are output only (Port G in ATmega128).
Using the External Memory The interface consists of:

Interface • AD7:0: Multiplexed low-order address bus and data bus.
• A15:8: High-order address bus (configurable number of bits).
• ALE: Address latch enable.
• RD: Read strobe.
• WR: Write strobe.
25
2467E–AVR–05/02
The control bits for the External Memory Interface are located in three registers, the
MCU Control Register – MCUCR, the External Memory Control Register A – XMCRA,
and the External Memory Control Register B – XMCRB.
When the XMEM interface is enabled, the XMEM interface will override the setting in the
data direction registers that corresponds to the ports dedicated to the XMEM interface.
For details about the port override, see the alternate functions in section “I/O Ports” on
page 60. The XMEM interface will auto-detect whether an access is internal or external.
If the access is external, the XMEM interface will output address, data, and the control
signals on the ports according to Figure 13 (this figure shows the wave forms without
wait-states). When ALE goes from high-to-low, there is a valid address on AD7:0. ALE is
low during a data transfer. When the XMEM interface is enabled, also an internal access
will cause activity on address, data and ALE ports, but the RD and WR strobes will not
toggle during internal access. When the External Memory Interface is disabled, the nor-
mal pin and data direction settings are used. Note that when the XMEM interface is
disabled, the address space above the internal SRAM boundary is not mapped into the
internal SRAM. Figure 12 illustrates how to connect an external SRAM to the AVR using
an octal latch (typically “74 x 573” or equivalent) which is transparent when G is high.
Address Latch Requirements Due to the high-speed operation of the XRAM interface, the address latch must be
selected with care for system frequencies above 8 MHz @ 4V and 4 MHz @ 2.7V.
When operating at conditions above these frequencies, the typical old style 74HC series
latch becomes inadequate. The External Memory Interface is designed in compliance to
the 74AHC series latch. However, most latches can be used as long they comply with
the main timing parameters. The main parameters for the address latch are:
• D to Q propagation delay (tPD).
• Data setup time before G low (tSU).
• Data (address) hold time after G low ( TH).
The External Memory Interface is designed to guaranty minimum address hold time
after G is asserted low of th = 5 ns. Refer to tLAXX_LD/tLLAXX_ST in “External Data Memory
Timing” Tables 137 through Tables 144 on pages 322 - 324. The D-to-Q propagation
delay (tPD) must be taken into consideration when calculating the access time require-
ment of the external component. The data setup time before G low (tSU ) must not
exceed address valid to ALE low (tAVLLC) minus PCB wiring delay (dependent on the
capacitive load).
Figure 12. External SRAM Connected to the AVR
D[7:0]
AD7:0 D Q A[7:0]
ALE G
SRAM
AVR
A15:8 A[15:8]
RD RD
WR WR
26 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Pull-up and Bus-keeper The pull-ups on the AD7:0 ports may be activated if the corresponding Port register is
written to one. To reduce power consumption in sleep mode, it is recommended to dis-
able the pull-ups by writing the Port register to zero before entering sleep.
The XMEM interface also provides a bus-keeper on the AD7:0 lines. The bus-keeper
can be disabled and enabled in software as described in “External Memory Control Reg-
ister B – XMCRB” on page 31. When enabled, the bus-keeper will keep the previous
value on the AD7:0 bus while these lines are tri-stated by the XMEM interface.
If neither bus-keeper nor pull-ups are enabled, the XMEM interface will leave the AD7:0
tri-stated during a read access until the next RAM access (internal or external) appears.
Timing External Memory devices have different timing requirements. To meet these require-
ments, the ATmega128 XMEM interface provides four different wait-states as shown in
Table 4. It is important to consider the timing specification of the External Memory
device before selecting the wait-state. The most important parameters are the access
time for the external memory compared to the set-up requirement of the ATmega128.
The access time for the External Memory is defined to be the time from receiving the
chip select/address until the data of this address actually is driven on the bus. The
access time cannot exceed the time from the ALE pulse must be asserted low until data
is stable during a read sequence (See tLLRL+ tRLRH - tDVRH in Tables 137 through Tables
144 on pages 322 - 324). The different wait-states are set up in software. As an addi-
tional feature, it is possible to divide the external memory space in two sectors with
individual wait-state settings. This makes it possible to connect two different memory
devices with different timing requirements to the same XMEM interface. For XMEM
interface timing details, please refer to Table 137 to Table 144 and Figure 156 to Figure
159 in the “External Data Memory Timing” on page 322.
Note that the XMEM interface is asynchronous and that the waveforms in the following
figures are related to the internal system clock. The skew between the internal and
external clock (XTAL1) is not guarantied (varies between devices temperature, and sup-
ply voltage). Consequently, the XMEM interface is not suited for synchronous operation.
Figure 13. External Data Memory Cycles without Wait-state (SRWn1=0 and SRWn0=0)
T1 T2 T3 T4
System Clock (CLKCPU )
ALE
A15:8 Prev. addr. Address
Write
DA7:0 Prev. data Address XX Data
WR
DA7:0 (XMBK = 0) Prev. data Address Data

Read
RD
Note: 1. SRWn1 = SRW11 (upper sector) or SRW01 (lower sector), SRWn0 = SRW10 (upper
sector) or SRW00 (lower sector). The ALE pulse in period T4 is only present if the
next instruction accesses the RAM (internal or external).
27
2467E–AVR–05/02
Figure 14. External Data Memory Cycles with SRWn1 = 0 and SRWn0 = 1(1)
T1 T2 T3 T4 T5
ALE
Write
WR
Read
RD
sector) or SRW00 (lower sector).
The ALE pulse in period T5 is only present if the next instruction accesses the RAM
(internal or external).
T1 T2 T3 T4 T5 T6
ALE
Write
WR
Read
RD
28 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
T1 T2 T3 T4 T5 T6 T7
ALE
Write
WR
Read
RD
XMEM Register Description
MCU Control Register –

MCUCR Bit 7 6 5 4 3 2 1 0
SRE SRW10 SE SM1 SM0 SM2 IVSEL IVCE MCUCR
• Bit 7 – SRE: External SRAM/XMEM Enable

Writing SRE to one enables the External Memory Interface.The pin functions AD7:0,
A15:8, ALE, WR, and RD are activated as the alternate pin functions. The SRE bit over-
rides any pin direction settings in the respective data direction registers. Writing SRE to
zero, disables the External Memory Interface and the normal pin and data direction set-
tings are used.
• Bit 6 – SRW10: Wait-state Select Bit

For a detailed description in non-ATmega103 compatibility mode, see common descrip-
tion for the SRWn bits below (XMCRA description). In ATmega103 compatibility mode,
writing SRW10 to one enables the wait-state and one extra cycle is added during
read/write strobe as shown in Figure 14.
External Memory Control

Register A – XMCRA Bit 7 6 5 4 3 2 1 0
– SRL2 SRL1 SRL0 SRW01 SRW00 SRW11 – XMCRA
Read/Write R R/W R/W R/W R/W R/W R/W R
• Bit 7 – Res: Reserved Bit

This is a reserved bit and will always read as zero. When writing to this address location,
write this bit to zero for compatibility with future devices.
29
2467E–AVR–05/02
• Bit 6..4 – SRL2, SRL1, SRL0: Wait-state Sector Limit
It is possible to configure different wait-states for different External Memory addresses.
The external memory address space can be divided in two sectors that have separate
wait-state bits. The SRL2, SRL1, and SRL0 bits select the split of the sectors, see Table
3 and Figure 11. By default, the SRL2, SRL1, and SRL0 bits are set to zero and the
entire external memory address space is treated as one sector. When the entire SRAM
address space is configured as one sector, the wait-states are configured by the
SRW11 and SRW10 bits.
Table 3. Sector limits with different settings of SRL2..0

SRL2 SRL1 SRL0 Sector Limits
0 0 0 Lower sector = N/A
Upper sector = 0x1100 - 0xFFFF
0 0 1 Lower sector = 0x1100 - 0x1FFF
Upper sector = 0xA000 - 0xFFFF
1 1 0 Lower sector = 0x1100 - 0xBFFF
Upper sector = 0xC000 - 0xFFFF
1 1 1 Lower sector = 0x1100 - 0xDFFF
Upper sector = 0xE000 - 0xFFFF
• Bit 1 and Bit 6 MCUCR – SRW11, SRW10: Wait-state Select Bits for Upper
Sector
The SRW11 and SRW10 bits control the number of wait-states for the upper sector of
the external memory address space, see Table 4.
• Bit 3..2 – SRW01, SRW00: Wait-state Select Bits for Lower Sector
The SRW01 and SRW00 bits control the number of wait-states for the lower sector of
the external memory address space, see Table 4.
Table 4. Wait States(1)

SRWn1 SRWn0 Wait States
0 0 No wait-states
0 1 Wait one cycle during read/write strobe
1 0 Wait two cycles during read/write strobe
1 1 Wait two cycles during read/write and wait one cycle before driving out
new address
Note: 1. n = 0 or 1 (lower/upper sector).
For further details of the timing and wait-states of the External Memory Interface, see
Figures 13 through Figures 16 for how the setting of the SRW bits affects the timing.
30 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)

This is a reserved bit and will always read as zero. When writing to this address location,
write this bit to zero for compatibility with future devices.
External Memory Control

Register B – XMCRB Bit 7 6 5 4 3 2 1 0
XMBK – – – – XMM2 XMM1 XMM0 XMCRB
Read/Write R/W R R R R R/W R/W R/W
• Bit 7– XMBK: External Memory Bus-keeper Enable

Writing XMBK to one enables the bus keeper on the AD7:0 lines. When the bus keeper
is enabled, AD7:0 will keep the last driven value on the lines even if the XMEM interface
has tri-stated the lines. Writing XMBK to zero disables the bus keeper. XMBK is not
qualified with SRE, so even if the XMEM interface is disabled, the bus keepers are still
activated as long as XMBK is one.
• Bit 6..4 – Res: Reserved Bits

• Bit 2..0 – XMM2, XMM1, XMM0: External Memory High Mask

When the External Memory is enabled, all Port C pins are default used for the high
address byte. If the full 60KB address space is not required to access the External Mem-
ory, some, or all, Port C pins can be released for normal Port Pin function as described
in Table 5. As described in “Using all 64KB Locations of External Memory” on page 32,
it is possible to use the XMMn bits to access all 64KB locations of the External Memory.
Table 5. Port C Pins Released as Normal Port Pins when the External Memory is
Enabled
XMM2 XMM1 XMM0 # Bits for External Memory Address Released Port Pins
0 0 0 8 (Full 60 KB space) None
0 0 1 7 PC7
0 1 0 6 PC7 - PC6
0 1 1 5 PC7 - PC5
1 0 0 4 PC7 - PC4
1 0 1 3 PC7 - PC3
1 1 0 2 PC7 - PC2
1 1 1 No Address high bits Full Port C
31
2467E–AVR–05/02
Using all 64KB Locations of Since the External Memory is mapped after the Internal Memory as shown in Figure 11,
External Memory only 60KB of External Memory is available by default (address space 0x0000 to 0x10FF
is reserved for internal memory). However, it is possible to take advantage of the entire
External Memory by masking the higher address bits to zero. This can be done by using
the XMMn bits and control by software the most significant bits of the address. By set-
ting Port C to output 0x00, and releasing the most significant bits for normal Port Pin
operation, the Memory Interface will address 0x0000 - 0x1FFF. See the following code
examples.
Assembly Code Example(1)
; OFFSET is defined to 0x2000 to ensure
; external memory access
; Configure Port C (address high byte) to
; output 0x00 when the pins are released
; for normal Port Pin operation
ldi r16, 0xFF
out DDRC, r16
ldi r16, 0x00
out PORTC, r16
; release PC7:5
ldi r16, (1<<XMM1)|(1<<XMM0)
sts XMCRB, r16
; write 0xAA to address 0x0001 of external
; memory
ldi r16, 0xaa
sts 0x0001+OFFSET, r16
; re-enable PC7:5 for external memory
ldi r16, (0<<XMM1)|(0<<XMM0)
sts XMCRB, r16
; store 0x55 to address (OFFSET + 1) of
; external memory
ldi r16, 0x55
sts 0x0001+OFFSET, r16
C Code Example(1)
#define OFFSET 0x2000
void XRAM_example(void)
{
unsigned char *p = (unsigned char *) (OFFSET + 1);
DDRC = 0xFF;
PORTC = 0x00;
XMCRB = (1<<XMM1) | (1<<XMM0);
*p = 0xaa;
XMCRB = 0x00;
*p = 0x55;
}
Note: 1. The example code assumes that the part specific header file is included.
Care must be exercised using this option as most of the memory is masked away.
32 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
System Clock and

Clock Options
Clock Systems and their Figure 17 presents the principal clock systems in the AVR and their distribution. All of
Distribution the clocks need not be active at a given time. In order to reduce power consumption, the
clocks to modules not being used can be halted by using different sleep modes, as
described in “Power Management and Sleep Modes” on page 41. The clock systems
are detailed below.
Figure 17. Clock Distribution
Asynchronous General I/O Flash and

ADC CPU Core RAM
Timer/Counter modules EEPROM
clkADC
clkI/O AVR Clock clkCPU

Control Unit
clkASY clkFLASH
Reset Logic Watchdog Timer
Source clock Watchdog clock
Clock Watchdog
Multiplexer Oscillator
Timer/Counter External RC Crystal Low-Frequency Calibrated RC

Oscillator Oscillator External clock Oscillator Crystal Oscillator Oscillator
CPU Clock – clkCPU The CPU clock is routed to parts of the system concerned with operation of the AVR
core. Examples of such modules are the General Purpose Register File, the Status Reg-
ister and the data memory holding the Stack Pointer. Halting the CPU clock inhibits the
core from performing general operations and calculations.
I/O Clock – clkI/O The I/O clock is used by the majority of the I/O modules, like Timer/Counters, SPI, and
USART. The I/O clock is also used by the External Interrupt module, but note that some
external interrupts are detected by asynchronous logic, allowing such interrupts to be
detected even if the I/O clock is halted. Also note that address recognition in the TWI
module is carried out asynchronously when clkI/O is halted, enabling TWI address recep-
tion in all sleep modes.
Flash Clock – clkFLASH The Flash clock controls operation of the Flash interface. The Flash clock is usually
active simultaneously with the CPU clock.
33
2467E–AVR–05/02
Asynchronous Timer Clock – The Asynchronous Timer clock allows the Asynchronous Timer/Counter to be clocked
clkASY directly from an external 32 kHz clock crystal. The dedicated clock domain allows using
this Timer/Counter as a real-time counter even when the device is in sleep mode.
ADC Clock – clkADC The ADC is provided with a dedicated clock domain. This allows halting the CPU and
I/O clocks in order to reduce noise generated by digital circuitry. This gives more accu-
rate ADC conversion results.
Clock Sources The device has the following clock source options, selectable by Flash fuse bits as
shown below. The clock from the selected source is input to the AVR clock generator,
and routed to the appropriate modules.
Table 6. Device Clocking Options Select

Device Clocking Option CKSEL3..0(1)
External Crystal/Ceramic Resonator 1111 - 1010
External Low-frequency Crystal 1001
External RC Oscillator 1000 - 0101
Calibrated Internal RC Oscillator 0100 - 0001
External Clock 0000
Note: 1. For all fuses “1” means unprogrammed while “0” means programmed.
The various choices for each clocking option is given in the following sections. When the
CPU wakes up from Power-down or Power-save, the selected clock source is used to
time the start-up, ensuring stable Oscillator operation before instruction execution starts.
When the CPU starts from reset, there is as an additional delay allowing the power to
reach a stable level before commencing normal operation. The Watchdog Oscillator is
used for timing this real-time part of the start-up time. The number of WDT Oscillator
cycles used for each time-out is shown in Table 7. The frequency of the Watchdog Oscil-
lator is voltage dependent as shown in the “ATmega128 Typical Characteristics –
Preliminary Data” on page 327. The device is shipped with CKSEL = “0001” and SUT =
“10” (1 MHz Internal RC Oscillator, slowly rising power).
Table 7. Number of Watchdog Oscillator Cycles

Typical Time-out (VCC = 5.0V) Typical Time-Out (VCC = 3.0V) Number of Cycles
4.1 ms 4.3 ms 4K (4,096)
65 ms 69 ms 64K (65,536)
Crystal Oscillator XTAL1 and XTAL2 are input and output, respectively, of an inverting amplifier which can
be configured for use as an On-chip Oscillator, as shown in Figure 18. Either a quartz
crystal or a ceramic resonator may be used. The CKOPT fuse selects between two dif-
ferent Oscillator Amplifier modes. When CKOPT is programmed, the Oscillator output
will oscillate will a full rail-to-rail swing on the output. This mode is suitable when operat-
ing in a very noisy environment or when the output from XTAL2 drives a second clock
buffer. This mode has a wide frequency range. When CKOPT is unprogrammed, the
Oscillator has a smaller output swing. This reduces power consumption considerably.
This mode has a limited frequency range and it can not be used to drive other clock
buffers.
For resonators, the maximum frequency is 8 MHz with CKOPT unprogrammed and 16
MHz with CKOPT programmed. C1 and C2 should always be equal for both crystals and
34 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
resonators. The optimal value of the capacitors depends on the crystal or resonator in
use, the amount of stray capacitance, and the electromagnetic noise of the environ-
ment. Some initial guidelines for choosing capacitors for use with crystals are given in
Table 8. For ceramic resonators, the capacitor values given by the manufacturer should
be used. For more information on how to choose capacitors and other details on Oscilla-
tor operation, refer to the Multi-purpose Oscillator application note.
Figure 18. Crystal Oscillator Connections
C2
XTAL2
C1
XTAL1
GND
The Oscillator can operate in three different modes, each optimized for a specific fre-
quency range. The operating mode is selected by the fuses CKSEL3..1 as shown in
Table 8.
Table 8. Crystal Oscillator Operating Modes

Frequency Range(1) Recommended Range for Capacitors
CKOPT CKSEL3..1 (MHz) C1 and C2 for Use with Crystals
1 101(2) 0.4 - 0.9 –
1 110 0.9 - 3.0 12 pF - 22 pF
1 111 3.0 - 8.0 12 pF - 22 pF
0 101, 110, 111 1.0 - 12 pF - 22 pF
Notes: 1. The frequency ranges are preliminary values. Actual values are TBD.
2. This option should not be used with crystals, only with ceramic resonators.
The CKSEL0 fuse together with the SUT1..0 fuses select the start-up times as shown in
Table 9.
35
2467E–AVR–05/02
Table 9. Start-up Times for the Crystal Oscillator Clock Selection
Start-up Time from Additional Delay
Power-down and from Reset (VCC
CKSEL0 SUT1..0 Power-save = 5.0V) Recommended Usage
(1)
00 258 CK 4.1 ms Ceramic resonator, fast
0
rising power
01 258 CK(1) 65 ms Ceramic resonator,
0
slowly rising power
10 1K CK(2) – Ceramic resonator,
0
BOD enabled
11 1K CK(2) 4.1 ms Ceramic resonator, fast
0
rising power
00 1K CK(2) 65 ms Ceramic resonator,
1
slowly rising power
01 16K CK – Crystal Oscillator, BOD
1
enabled
10 16K CK 4.1 ms Crystal Oscillator, fast
1
rising power
11 16K CK 65 ms Crystal Oscillator,
1
slowly rising power
Notes: 1. These options should only be used when not operating close to the maximum fre-
quency of the device, and only if frequency stability at start-up is not important for the
application. These options are not suitable for crystals.
2. These options are intended for use with ceramic resonators and will ensure fre-
quency stability at start-up. They can also be used with crystals when not operating
close to the maximum frequency of the device, and if frequency stability at start-up is
not important for the application.
Low-frequency Crystal To use a 32.768 kHz watch crystal as the clock source for the device, the Low-fre-
Oscillator quency Crystal Oscillator must be selected by setting the CKSEL fuses to “1001”. The
crystal should be connected as shown in Figure 18. By programming the CKOPT fuse,
the user can enable internal capacitors on XTAL1 and XTAL2, thereby removing the
need for external capacitors. The internal capacitors have a nominal value of 36 pF.
Refer to the 32 kHz Crystal Oscillator application note for details on Oscillator operation
and how to choose appropriate values for C1 and C2.
When this Oscillator is selected, start-up times are determined by the SUT fuses as
shown in Table 10.
Table 10. Start-up Times for the Low-frequency Crystal Oscillator Clock Selection
Power-down and from Reset (VCC =
SUT1..0 Power-save 5.0V) Recommended Usage
(1)
00 1K CK 4.1 ms Fast rising power or BOD enabled
(1)
01 1K CK 65 ms Slowly rising power
10 32K CK 65 ms Stable frequency at start-up
11 Reserved
36 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Note: 1. These options should only be used if frequency stability at start-up is not important for
the application.
External RC Oscillator For timing insensitive applications, the External RC configuration shown in Figure 19
can be used. The frequency is roughly estimated by the equation f = 1/(3RC). C should
be at least 22 pF. By programming the CKOPT fuse, the user can enable an internal
36 pF capacitor between XTAL1 and GND, thereby removing the need for an external
capacitor. For more information on Oscillator operation and details on how to choose R
and C, refer to the External RC Oscillator application note.
Figure 19. External RC Configuration

VCC
R NC XTAL2
XTAL1
C
GND
The Oscillator can operate in four different modes, each optimized for a specific fre-
quency range. The operating mode is selected by the fuses CKSEL3..0 as shown in
Table 11.
Table 11. External RC Oscillator Operating Modes

CKSEL3..0 Frequency Range (MHz)
0101 - 0.9
0110 0.9 - 3.0
0111 3.0 - 8.0
1000 8.0 - 12.0
shown in Table 12.
Table 12. Start-Up Times for the External RC Oscillator Clock Selection
Power-down and from Reset
SUT1..0 Power-save (VCC = 5.0V) Recommended Usage
00 18 CK – BOD enabled
01 18 CK 4.1 ms Fast rising power
10 18 CK 65 ms Slowly rising power
(1)
11 6 CK 4.1 ms Fast rising power or BOD enabled
Note: 1. This option should not be used when operating close to the maximum frequency of
the device.
37
2467E–AVR–05/02
Calibrated Internal RC The Calibrated Internal RC Oscillator provides a fixed 1.0, 2.0, 4.0, or 8.0 MHz clock. All
Oscillator frequencies are nominal values at 5V and 25°C. This clock may be selected as the sys-
tem clock by programming the CKSEL fuses as shown in Table 13. If selected, it will
operate with no external components. The CKOPT fuse should always be unpro-
grammed when using this clock option. During Reset, hardware loads the calibration
byte into the OSCCAL Register and thereby automatically calibrates the RC Oscillator.
At 5V, 25°C and 1.0 MHz Oscillator frequency selected, this calibration gives a fre-
quency within ± 1% of the nominal frequency. When this Oscillator is used as the chip
clock, the Watchdog Oscillator will still be used for the Watchdog Timer and for the
Reset Time-out. For more information on the pre-programmed calibration value, see the
section “Calibration Byte” on page 285.
Table 13. Internal Calibrated RC Oscillator Operating Modes

CKSEL3..0 Nominal Frequency (MHz)
0001(1) 1.0
0010 2.0
0011 4.0
0100 8.0
Note: 1. The device is shipped with this option selected.
shown in Table 14. XTAL1 and XTAL2 should be left unconnected (NC).
Table 14. Start-up Times for the Internal Calibrated RC Oscillator Clock Selection
Start-up Time from Power- Additional Delay from
SUT1..0 down and Power-save Reset (V CC = 5.0V) Recommended Usage
(1)
11 Reserved
Note: 1. The device is shipped with this option selected.
Oscillator Calibration Register

– OSCCAL Bit 7 6 5 4 3 2 1 0
CAL7 CAL6 CAL5 CAL4 CAL3 CAL2 CAL1 CAL0 OSCCAL
Initial Value Device Specific Calibration Value
Note: OSCCAL Register is not available in ATmega103 compatibility mode.
• Bits 7..0 – CAL7..0: Oscillator Calibration Value

Writing the calibration byte to this address will trim the Internal Oscillator to remove pro-
cess variations from the Oscillator frequency. During Reset, the 1 MHz calibration value
which is located in the signature row high byte (address 0x00) is automatically loaded
into the OSCCAL Register. If the internal RC is used at other frequencies, the calibration
values must be loaded manually. This can be done by first reading the signature row by
a programmer, and then store the calibration values in the Flash or EEPROM. Then the
value can be read by software and loaded into the OSCCAL Register. When OSCCAL is
zero, the lowest available frequency is chosen. Writing non-zero values to this register
38 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
will increase the frequency of the Internal Oscillator. Writing $FF to the register gives the
highest available frequency. The calibrated Oscillator is used to time EEPROM and
Flash access. If EEPROM or Flash is written, do not calibrate to more than 10% above
the nominal frequency. Otherwise, the EEPROM or Flash write may fail. Note that the
Oscillator is intended for calibration to 1.0, 2.0, 4.0, or 8.0 MHz. Tuning to other values is
not guaranteed, as indicated in Table 15.
Table 15. Internal RC Oscillator Frequency Range.

Min Frequency in Percentage of Max Frequency in Percentage of
OSCCAL Value Nominal Frequency (%) Nominal Frequency (%)
$00 50 100
$7F 75 150
$FF 100 200
External Clock To drive the device from an external clock source, XTAL1 should be driven as shown in
Figure 20. To run the device on an external clock, the CKSEL fuses must be pro-
grammed to “0000”. By programming the CKOPT fuse, the user can enable an internal
36 pF capacitor between XTAL1 and GND.
Figure 20. External Clock Drive Configuration
EXTERNAL
CLOCK
SIGNAL
When this clock source is selected, start-up times are determined by the SUT fuses as
shown in Table 16.
Table 16. Start-up Times for the External Clock Selection

Start-up Time from Power- Additional Delay from
SUT1..0 down and Power-save Reset (V CC = 5.0V) Recommended Usage
11 Reserved
39
2467E–AVR–05/02
Timer/Counter Oscillator For AVR microcontrollers with Timer/Counter Oscillator pins (TOSC1 and TOSC2), the
crystal is connected directly between the pins. No external capacitors are needed. The
Oscillator is optimized for use with a 32.768 kHz watch crystal. Applying an external
clock source to TOSC1 is not recommended.
XTAL Divide Control Register The XTAL Divide Control Register is used to divide the Source clock frequency by a
– XDIV number in the range 2 - 129. This feature can be used to decrease power consumption
when the requirement for processing power is low.
Bit 7 6 5 4 3 2 1 0
XDIVEN XDIV6 XDIV5 XDIV4 XDIV3 XDIV2 XDIV1 XDIV0 XDIV
• Bit 7 – XDIVEN: XTAL Divide Enable

When the XDIVEN bit is written one, the clock frequency of the CPU and all peripherals
(clkI/O, clkADC, clkCPU, clkFLASH) is divided by the factor defined by the setting of XDIV6 -
XDIV0. This bit can be written run-time to vary the clock frequency as suitable to the
application.
• Bits 6..0 – XDIV6..XDIV0: XTAL Divide Select Bits 6 - 0

These bits define the division factor that applies when the XDIVEN bit is set (one). If the
value of these bits is denoted d, the following formula defines the resulting CPU and
peripherals clock frequency fCLK:
f CLK = Source clock-

---------------------------------
129 – d
The value of these bits can only be changed when XDIVEN is zero. When XDIVEN is
written to one, the value written simultaneously into XDIV6..XDIV0 is taken as the divi-
sion factor. When XDIVEN is written to zero, the value written simultaneously into
XDIV6..XDIV0 is rejected. As the divider divides the master clock input to the MCU, the
speed of all peripherals is reduced when a division factor is used.
Note: Timer/Counter0 should not be used when the system clock is divided. Since the
Timer/Counter0 also serves as an asynchronous Timer/Counter, the clock will not be
divided for this module according to the setting in the XDIV Register, even if the
Timer/Counter is run synchronously. As a consequence, interrupts may be lost and
accessing the Timer/Counter0 Registers may fail.
40 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Power Management Sleep modes enable the application to shut down unused modules in the MCU, thereby
saving power. The AVR provides various sleep modes allowing the user to tailor the
and Sleep Modes
power consumption to the application’s requirements.
To enter any of the six sleep modes, the SE bit in MCUCR must be written to logic one
and a SLEEP instruction must be executed. The SM2, SM1, and SM0 bits in the
MCUCR Register select which sleep mode (Idle, ADC Noise Reduction, Power-down,
Power-save, Standby, or Extended Standby) will be activated by the SLEEP instruction.
See Table 17 for a summary. If an enabled interrupt occurs while the MCU is in a sleep
mode, the MCU wakes up. The MCU is then halted for four cycles in addition to the
start-up time, it executes the interrupt routine, and resumes execution from the instruc-
tion following SLEEP. The contents of the register file and SRAM are unaltered when
the device wakes up from sleep. If a reset occurs during sleep mode, the MCU wakes
up and executes from the Reset Vector.
Figure 17 on page 33 presents the different clock systems in the ATmega128, and their
distribution. The figure is helpful in selecting an appropriate sleep mode.
MCU Control Register – The MCU Control Register contains control bits for power management.
MCUCR
Bit 7 6 5 4 3 2 1 0
• Bit 5 – SE: Sleep Enable

The SE bit must be written to logic one to make the MCU enter the Sleep mode when
the SLEEP instruction is executed. To avoid the MCU entering the Sleep mode unless it
is the programmers purpose, it is recommended to write the Sleep Enable (SE) bit to
one just before the execution of the SLEEP instruction and to clear it immediately after
waking up.
• Bits 4..2 – SM2..0: Sleep Mode Select Bits 2, 1, and 0

These bits select between the six available sleep modes as shown in Table 17.
Table 17. Sleep Mode Select

SM2 SM1 SM0 Sleep Mode
0 0 0 Idle
0 0 1 ADC Noise Reduction
0 1 0 Power-down
0 1 1 Power-save
1 0 0 Reserved
1 0 1 Reserved
1 1 0 Standby(1)
1 1 1 Extended Standby(1)
Note: 1. Standby mode and Extended Standby mode are only available with external crystals
or resonators.
41
2467E–AVR–05/02
Idle Mode When the SM2..0 bits are written to 000, the SLEEP instruction makes the MCU enter
Idle mode, stopping the CPU but allowing SPI, USART, Analog Comparator, ADC, Two-
wire Serial Interface, Timer/Counters, Watchdog, and the interrupt system to continue
operating. This sleep mode basically halts clkCPU and clkFLASH, while allowing the other
clocks to run.
Idle mode enables the MCU to wake up from external triggered interrupts as well as
internal ones like the Timer Overflow and USART Transmit Complete interrupts. If
wake-up from the Analog Comparator interrupt is not required, the Analog Comparator
can be powered down by setting the ACD bit in the Analog Comparator Control and Sta-
tus Register – ACSR. This will reduce power consumption in Idle mode. If the ADC is
enabled, a conversion starts automatically when this mode is entered.
ADC Noise Reduction When the SM2..0 bits are written to 001, the SLEEP instruction makes the MCU enter
Mode ADC Noise Reduction mode, stopping the CPU but allowing the ADC, the External
Interrupts, the Two-wire Serial Interface address watch, Timer/Counter0 and the
Watchdog to continue operating (if enabled). This sleep mode basically halts clkI/O,
clkCPU, and clkFLASH, while allowing the other clocks to run.
This improves the noise environment for the ADC, enabling higher resolution measure-
ments. If the ADC is enabled, a conversion starts automatically when this mode is
entered. Apart form the ADC Conversion Complete interrupt, only an External Reset, a
Watchdog Reset, a Brown-out Reset, a Two-wire Serial Interface address match inter-
rupt, a Timer/Counter0 interrupt, an SPM/EEPROM ready interrupt, an External Level
Interrupt on INT7:4, or an External Interrupt on INT3:0 can wake up the MCU from ADC
Noise Reduction mode.
Power-down Mode When the SM2..0 bits are written to 010, the SLEEP instruction makes the MCU enter
Power-down mode. In this mode, the External Oscillator is stopped, while the External
Interrupts, the Two-wire Serial Interface address watch, and the Watchdog continue
operating (if enabled). Only an External Reset, a Watchdog Reset, a Brown-out Reset, a
Two-wire Serial Interface address match interrupt, an External Level Interrupt on
INT7:4, or an External Interrupt on INT3:0 can wake up the MCU. This sleep mode basi-
cally halts all generated clocks, allowing operation of asynchronous modules only.
Note that if a level triggered interrupt is used for wake-up from Power-down mode, the
changed level must be held for some time to wake up the MCU. Refer to “External Inter-
rupts” on page 84 for details.
When waking up from Power-down mode, there is a delay from the wake-up condition
occurs until the wake-up becomes effective. This allows the clock to restart and become
stable after having been stopped. The wake-up period is defined by the same CKSEL
fuses that define the Reset Time-out period, as described in “Clock Sources” on page
34.
Power-save Mode When the SM2..0 bits are written to 011, the SLEEP instruction makes the MCU enter
Power-save mode. This mode is identical to Power-down, with one exception:
If Timer/Counter0 is clocked asynchronously, i.e., the AS0 bit in ASSR is set,
Timer/Counter0 will run during sleep. The device can wake up from either Timer Over-
f lo w o r O u t p u t C o m p a r e e ve n t fr o m Ti me r / C o u n te r 0 if t h e c o r r e sp o n d in g
Timer/Counter0 interrupt enable bits are set in TIMSK, and the global interrupt enable
bit in SREG is set.
If the Asynchronous Timer is NOT clocked asynchronously, Power-down mode is rec-
ommended instead of Power-save mode because the contents of the registers in the
42 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
asynchronous timer should be considered undefined after wake-up in Power-save mode

if AS0 is 0.
This sleep mode basically halts all clocks except clkASY, allowing operation only of asyn-
chronous modules, including Timer/Counter0 if clocked asynchronously.
Standby Mode When the SM2..0 bits are 110 and an External Crystal/Resonator clock option is
selected, the SLEEP instruction makes the MCU enter Standby mode. This mode is
identical to Power-down with the exception that the Oscillator is kept running. From
Standby mode, the device wakes up in 6 clock cycles.
Extended Standby Mode When the SM2..0 bits are 111 and an external crystal/resonator clock option is selected,
the SLEEP instruction makes the MCU enter Extended Standby mode. This mode is
identical to Power-save mode with the exception that the Oscillator is kept running.
From Extended Standby mode, the device wakes up in six clock cycles.
Table 18. Active Clock Domains and Wake Up Sources in the Different Sleep Modes
Active Clock Domains Oscillators Wake Up Sources
Main Clock Timer TWI SPM/
Sleep Source Osc Address EEPROM Other
Mode clkCPU clkFLASH clkIO clkADC clkASY Enabled Enabled INT7:0 Match Timer 0 Ready ADC I/O
Idle X X X X X(2) X X X X X X
ADC
Noise X X X X(2) X(3) X X X X
Reduction
Power-
X(3) X
down
Power-
X(2) X(2) X(3) X X(2)
save
Standby(1) X X(3) X
Extended
X(2) X X(2) X(3) X X(2)
Standby(1)
Notes: 1. External Crystal or resonator selected as clock source
2. If AS0 bit in ASSR is set
3. Only INT3:0 or level interrupt INT7:4
43
2467E–AVR–05/02
Minimizing Power There are several issues to consider when trying to minimize the power consumption in
Consumption an AVR controlled system. In general, sleep modes should be used as much as possi-
ble, and the sleep mode should be selected so that as few as possible of the device’s
functions are operating. All functions not needed should be disabled. In particular, the
following modules may need special consideration when trying to achieve the lowest
possible power consumption.
Analog to Digital Converter If enabled, the ADC will be enabled in all sleep modes. To save power, the ADC should
be disabled before entering any sleep mode. When the ADC is turned off and on again,
the next conversion will be an extended conversion. Refer to “Analog to Digital Con-
verter” on page 225 for details on ADC operation.
Analog Comparator When entering Idle mode, the Analog Comparator should be disabled if not used. When
entering ADC Noise Reduction mode, the Analog Comparator should be disabled. In the
other sleep modes, the Analog Comparator is automatically disabled. However, if the
Analog Comparator is set up to use the Internal Voltage Reference as input, the Analog
Comparator should be disabled in all sleep modes. Otherwise, the Internal Voltage Ref-
erence will be enabled, independent of sleep mode. Refer to “Analog Comparator” on
page 222 for details on how to configure the Analog Comparator.
Brown-out Detector If the Brown-out Detector is not needed in the application, this module should be turned
off. If the Brown-out Detector is enabled by the BODEN fuse, it will be enabled in all
sleep modes, and hence, always consume power. In the deeper sleep modes, this will
contribute significantly to the total current consumption. Refer to “Brown-out Detector”
on page 44 for details on how to configure the Brown-out Detector.
Internal Voltage Reference The Internal Voltage Reference will be enabled when needed by the Brown-out Detec-
tor, the Analog Comparator or the ADC. If these modules are disabled as described in
the sections above, the internal voltage reference will be disabled and it will not be con-
suming power. When turned on again, the user must allow the reference to start up
before the output is used. If the reference is kept on in sleep mode, the output can be
used immediately. Refer to “Internal Voltage Reference” on page 50 for details on the
start-up time.
Watchdog Timer If the Watchdog Timer is not needed in the application, this module should be turned off.
If the Watchdog Timer is enabled, it will be enabled in all sleep modes, and hence,
always consume power. In the deeper sleep modes, this will contribute significantly to
the total current consumption. Refer to “Watchdog Timer” on page 50 for details on how
to configure the Watchdog Timer.
Port Pins When entering a sleep mode, all port pins should be configured to use minimum power.
The most important thing is then to ensure that no pins drive resistive loads. In sleep
modes where the both the I/O clock (clkI/O) and the ADC clock (clkADC) are stopped, the
input buffers of the device will be disabled. This ensures that no power is consumed by
the input logic when not needed. In some cases, the input logic is needed for detecting
wake-up conditions, and it will then be enabled. Refer to the section “Digital Input
Enable and Sleep Modes” on page 64 for details on which pins are enabled. If the input
buffer is enabled and the input signal is left floating or have an analog signal level close
to VCC/2, the input buffer will use excessive power.
44 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
System Control and

Reset
Resetting the AVR During Reset, all I/O registers are set to their initial values, and the program starts exe-
cution from the Reset Vector. The instruction placed at the Reset Vector must be a JMP
– absolute jump – instruction to the reset handling routine. If the program never enables
an interrupt source, the interrupt vectors are not used, and regular program code can be
placed at these locations. This is also the case if the Reset Vector is in the Application
section while the interrupt vectors are in the Boot section or vice versa. The circuit dia-
gram in Figure 21 shows the reset logic. Table 19 defines the electrical parameters of
the reset circuitry.
The I/O ports of the AVR are immediately reset to their initial state when a reset source
goes active. This does not require any clock source to be running.
After all reset sources have gone inactive, a delay counter is invoked, stretching the
internal reset. This allows the power to reach a stable level before normal operation
starts. The time-out period of the delay counter is defined by the user through the
CKSEL fuses. The different selections for the delay period are presented in “Clock
Sources” on page 34.
Reset Sources The ATmega128 has five sources of reset:

• Power-on Reset. The MCU is reset when the supply voltage is below the Power-on
Reset threshold (VPOT).
• External Reset. The MCU is reset when a low level is present on the RESET pin for
longer than the minimum pulse length.
• Watchdog Reset. The MCU is reset when the Watchdog Timer period expires and
the Watchdog is enabled.
• Brown-out Reset. The MCU is reset when the supply voltage VCC is below the
Brown-out Reset threshold (VBOT) and the Brown-out Detector is enabled.
• JTAG AVR Reset. The MCU is reset as long as there is a logic one in the Reset
Register, one of the scan chains of the JTAG system. Refer to the section “IEEE
1149.1 (JTAG) Boundary-scan” on page 248 for details.
45
2467E–AVR–05/02
Figure 21. Reset Logic
DATA BUS
PEN D Q MCU Control and Status

Register (MCUCSR)
L Q
PORF
BORF
EXTRF
WDRF
JTRF
Pull-up Resistor
Power-On Reset
Circuit
Brown-Out
BODEN
BODLEVEL Reset Circuit
Pull-up Resistor
SPIKE
RESET FILTER Reset Circuit
COUNTER RESET
JTAG Reset Watchdog
Register Timer
Watchdog
Oscillator
Clock CK Delay Counters

Generator TIMEOUT
CKSEL[3:0]
SUT[1:0]
Table 19. Reset Characteristics

Symbol Parameter Condition Min Typ Max Units
Power-on Reset
Threshold Voltage 1.4 2.3 V
(rising)
VPOT
Power-on Reset
Threshold Voltage 1.3 2.3 V
(falling)(1)
RESET Pin Threshold
VRST 0.2 V CC 0.85 V CC V
Voltage
Minimum pulse width on
tRST 50 ns
RESET Pin
Brown-out Reset BODLEVEL = 1 2.5 2.7 3.2
VBOT Threshold Voltage(2) V
BODLEVEL = 0 3.7 4.0 4.5
Minimum low voltage BODLEVEL = 1 2 µs
tBOD period for Brown-out
Detection BODLEVEL = 0 2 µs
Brown-out Detector
VHYST 50 mV
hysteresis
Notes: 1. The Power-on Reset will not work unless the supply voltage has been below VPOT
(falling)
46 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
2. VBOT may be below nominal minimum operating voltage for some devices. For
devices where this is the case, the device is tested down to V CC = VBOT during the
production test. This guarantees that a Brown-out Reset will occur before V CC drops
to a voltage where correct operation of the microcontroller is no longer guaranteed.
The test is performed using BODLEVEL=1 for ATmega128L and BODLEVEL=0 for
ATmega128. BODLEVEL=1 is not applicable for ATmega128.
Power-on Reset A Power-on Reset (POR) pulse is generated by an On-chip detection circuit. The detec-
tion level is defined in Table 19. The POR is activated whenever V CC is below the
detection level. The POR circuit can be used to trigger the Start-up Reset, as well as to
detect a failure in supply voltage.
A Power-on Reset (POR) circuit ensures that the device is reset from Power-on. Reach-
ing the Power-on Reset threshold voltage invokes the delay counter, which determines
how long the device is kept in RESET after VCC rise. The RESET signal is activated
again, without any delay, when VCC decreases below the detection level.
Figure 22. MCU Start-up, RESET Tied to VCC.

VPOT
VCC
VRST
RESET
tTOUT
TIME-OUT
INTERNAL
RESET
Figure 23. MCU Start-up, RESET Extended Externally
VPOT
VCC
VRST
RESET
tTOUT
TIME-OUT
INTERNAL
RESET
External Reset An External Reset is generated by a low level on the RESET pin. Reset pulses longer
than the minimum pulse width (see Table 19) will generate a reset, even if the clock is
not running. Shorter pulses are not guaranteed to generate a reset. When the applied
signal reaches the Reset Threshold Voltage – VRST on its positive edge, the delay
counter starts the MCU after the Time-out period tTOUT has expired.
47
2467E–AVR–05/02
Figure 24. External Reset During Operation
CC
Brown-out Detection ATmega128 has an On-chip Brown-out Detection (BOD) circuit for monitoring the VCC
level during operation by comparing it to a fixed trigger level. The trigger level for the
BOD can be selected by the fuse BODLEVEL to be 2.7V (BODLEVEL unprogrammed),
or 4.0V (BODLEVEL programmed). The trigger level has a hysteresis to ensure spike
free Brown-out Detection. The hysteresis on the detection level should be interpreted as
VBOT+ = VBOT + VHYST/2 and VBOT- = VBOT - VHYST/2.
The BOD circuit can be enabled/disabled by the fuse BODEN. When the BOD is
enabled (BODEN programmed), and VCC decreases to a value below the trigger level
(VBOT- in Figure 25), the Brown-out Reset is immediately activated. When VCC increases
above the trigger level (VBOT+ in Figure 25), the delay counter starts the MCU after the
time-out period tTOUT has expired.
The BOD circuit will only detect a drop in VCC if the voltage stays below the trigger level
for longer than tBOD given in Table 19.
Figure 25. Brown-out Reset During Operation
VCC VBOT+
VBOT-
RESET
TIME-OUT tTOUT
INTERNAL
RESET
48 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Watchdog Reset When the Watchdog times out, it will generate a short reset pulse of 1 CK cycle dura-
tion. On the falling edge of this pulse, the delay timer starts counting the Time-out period
tTOUT. Refer to page 50 for details on operation of the Watchdog Timer.
Figure 26. Watchdog Reset During Operation

CC
CK
MCU Control and Status The MCU Control and Status Register provides information on which reset source
Register – MCUCSR caused an MCU reset.
Bit 7 6 5 4 3 2 1 0
JTD – – JTRF WDRF BORF EXTRF PORF MCUCSR
Read/Write R/W R R R/W R/W R/W R/W R/W
Initial Value 0 0 0 See Bit Description
Note that only EXTRF and PORF are available in ATmega103 compatibility mode.
• Bit 4 – JTRF: JTAG Reset Flag

This bit is set if a reset is being caused by a logic one in the JTAG Reset Register
selected by the JTAG instruction AVR_RESET. This bit is reset by a Power-on Reset, or
by writing a logic zero to the flag.
• Bit 3 – WDRF: Watchdog Reset Flag

This bit is set if a Watchdog Reset occurs. The bit is reset by a Power-on Reset, or by
writing a logic zero to the flag.
• Bit 2 – BORF: Brown-out Reset Flag

This bit is set if a Brown-out Reset occurs. The bit is reset by a Power-on Reset, or by
• Bit 1 – EXTRF: External Reset Flag

This bit is set if an External Reset occurs. The bit is reset by a Power-on Reset, or by
• Bit 0 – PORF: Power-On Reset Flag

This bit is set if a Power-on Reset occurs. The bit is reset only by writing a logic zero to
the flag.
To make use of the reset flags to identify a reset condition, the user should read and
then reset the MCUCSR as early as possible in the program. If the register is cleared
49
2467E–AVR–05/02
before another reset occurs, the source of the reset can be found by examining the reset
flags.
Internal Voltage ATmega128 features an internal bandgap reference. This reference is used for Brown-
Reference out Detection, and it can be used as an input to the Analog Comparator or the ADC. The
2.56V reference to the ADC is generated from the internal bandgap reference.
Voltage Reference Enable The voltage reference has a start-up time that may influence the way it should be used.
Signals and Start-up Time The start-up time is given in Table 20. To save power, the reference is not always turned
on. The reference is on during the following situations:
1. When the BOD is enabled (by programming the BODEN fuse).
2. When the bandgap reference is connected to the Analog Comparator (by setting
the ACBG bit in ACSR).
3. When the ADC is enabled.
Thus, when the BOD is not enabled, after setting the ACBG bit or enabling the ADC, the
user must always allow the reference to start up before the output from the Analog Com-
parator or ADC is used. To reduce power consumption in Power-down mode, the user
can avoid the three conditions above to ensure that the reference is turned off before
entering Power-down mode.
Table 20. Internal Voltage Reference Characteristics

Symbol Parameter Min Typ Max Units
VBG Bandgap reference voltage 1.15 1.23 1.40 V
tBG Bandgap reference start-up time 40 70 µs
IBG Bandgap reference current consumption 10 µA
Watchdog Timer The Watchdog Timer is clocked from a separate On-chip Oscillator which runs at 1 Mhz.
This is the typical value at VCC = 5V. See characterization data for typical values at other
VCC levels. By controlling the Watchdog Timer prescaler, the Watchdog Reset interval
can be adjusted as shown in Table 22 on page 52. The WDR – Watchdog Reset –
instruction resets the Watchdog Timer. The Watchdog Timer is also reset when it is dis-
abled and when a Chip Reset occurs. Eight different clock cycle periods can be selected
to determine the reset period. If the reset period expires without another Watchdog
Reset, the ATmega128 resets and executes from the Reset Vector. For timing details on
the Watchdog Reset, refer to page 49.
To prevent unintentional disabling of the Watchdog or unintentional change of time-out
period, 3 different safety levels are selected by the Fuses M103C and WDTON as
shown in Table 21. Safety level 0 corresponds to the setting in ATmega103. There is no
restriction on enabling the WDT in any of the safety levels. Refer to “Timed Sequences
for Changing the Configuration of the Watchdog Timer” on page 53 for details.
50 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 21. WDT Configuration as a Function of the Fuse Settings of M103C and
WDTON.
How to
Safety WDT Initial How to Disable Change
M103C WDTON Level State the WDT Time-out
Unprogrammed Unprogrammed 1 Disabled Timed Timed
sequence sequence
Unprogrammed Programmed 2 Enabled Always enabled Timed
sequence
Programmed Unprogrammed 0 Disabled Timed No
sequence restriction
Programmed Programmed 2 Enabled Always enabled Timed
sequence
Figure 27. Watchdog Timer
WATCHDOG
OSCILLATOR
Watchdog Timer Control

Register – WDTCR Bit 7 6 5 4 3 2 1 0
– – – WDCE WDE WDP2 WDP1 WDP0 WDTCR
Read/Write R R R R/W R/W R/W R/W R/W

• Bit 4 – WDCE: Watchdog Change Enable

This bit must be set when the WDE bit is written to logic zero. Otherwise, the Watchdog
will not be disabled. Once written to one, hardware will clear this bit after four clock
cycles. Refer to the description of the WDE bit for a Watchdog disable procedure. In
Safety Level 1 and 2, this bit must also be set when changing the prescaler bits. See
“Timed Sequences for Changing the Configuration of the Watchdog Timer” on page 53.
• Bit 3 – WDE: Watchdog Enable

When the WDE is written to logic one, the Watchdog Timer is enabled, and if the WDE is
written to logic zero, the Watchdog Timer function is disabled. WDE can only be cleared
51
2467E–AVR–05/02
if the WDCE bit has logic level one. To disable an enabled Watchdog Timer, the follow-
ing procedure must be followed:
1. In the same operation, write a logic one to WDCE and WDE. A logic one must be
written to WDE even though it is set to one before the disable operation starts.
2. Within the next four clock cycles, write a logic 0 to WDE. This disables the
Watchdog.
In safety level 2, it is not possible to disable the Watchdog Timer, even with the algo-
rithm described above. See “Timed Sequences for Changing the Configuration of the
Watchdog Timer” on page 53.
• Bits 2..0 – WDP2, WDP1, WDP0: Watchdog Timer Prescaler 2, 1, and 0

The WDP2, WDP1, and WDP0 bits determine the Watchdog Timer prescaling when the
Watchdog Timer is enabled. The different prescaling values and their corresponding
Timeout Periods are shown in Table 22.
Table 22. Watchdog Timer Prescale Select

Number of WDT Typical Time-out Typical Time-out
WDP2 WDP1 WDP0 Oscillator Cycles at V CC = 3.0V at VCC = 5.0V
0 0 0 16K (16,384) 17.1 ms 16.3 ms
0 0 1 32K (32,768) 34.3 ms 32.5 ms
0 1 0 64K (65,536) 68.5 ms 65 ms
0 1 1 128K (131,072) 0.14 s 0.13 s
1 0 0 256K (262,144) 0.27 s 0.26 s
1 0 1 512K (524,288) 0.55 s 0.52 s
1 1 0 1,024K (1,048,576) 1.1 s 1.0 s
1 1 1 2,048K (2,097,152) 2.2 s 2.1 s
52 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
The following code example shows one assembly and one C function for turning off the
WDT. The example assumes that interrupts are controlled (e.g. by disabling interrupts
globally) so that no interrupts will occur during execution of these functions.
WDT_off:
; Write logical one to WDCE and WDE
ldi r16, (1<<WDCE)|(1<<WDE)
out WDTCR, r16
; Turn off WDT
ldi r16, (0<<WDE)
out WDTCR, r16
ret
C Code Example
void WDT_off(void)
{
/* Write logical one to WDCE and WDE */
WDTCR = (1<<WDCE) | (1<<WDE);
/* Turn off WDT */
WDTCR = 0x00;
}
Timed Sequences for Changing the Configuration of the Watchdog Timer

The sequence for changing configuration differs slightly between the three safety levels.
Separate procedures are described for each level.
Safety Level 0 This mode is compatible with the Watchdog operation found in ATmega103. The Watch-
dog Timer is initially disabled, but can be enabled by writing the WDE bit to 1 without
any restriction. The time-out period can be changed at any time without restriction. To
disable an enabled Watchdog Timer, the procedure described on page 51 (WDE bit
description) must be followed.
Safety Level 1 In this mode, the Watchdog Timer is initially disabled, but can be enabled by writing the
WDE bit to 1 without any restriction. A timed sequence is needed when changing the
Watchdog Time-out period or disabling an enabled Watchdog Timer. To disable an
enabled Watchdog Timer, and/or changing the watch Og Time-out, the following proce-
dure must be followed:
1. In the same operation, write a logic one to WDCE and WDE. A logic one must be
written to WDE regardless of the previous value of the WDE bit.
2. Within the next four clock cycles, in the same operation, write the WDE and
WDP bits as desired, but with the WDCE bit cleared.
Safety Level 2 In this mode, the Watchdog Timer is always enabled, and the WDE bit will always read
as one. A timed sequence is needed when changing the Watchdog Time-out period. To
change the Watchdog Time-out, the following procedure must be followed:
1. In the same operation, write a logical one to WDCE and WDE. Even though the
WDE always is set, the WDE must be written to one to start the timed sequence.
2. Within the next four clock cycles, in the same operation, write the WDP bits as
desired, but with the WDCE bit cleared. The value written to the WDE bit is
irrelevant.
53
2467E–AVR–05/02
Interrupts This section describes the specifics of the interrupt handling as performed in
ATmega128. For a general explanation of the AVR interrupt handling, refer to “Reset
and Interrupt Handling” on page 13.
Interrupt Vectors in
ATmega128 Table 23. Reset and Interrupt Vectors
Vector Program
No. Address(2) Source Interrupt Definition
External Pin, Power-on Reset, Brown-out Reset,
1 $0000(1) RESET Watchdog Reset, and JTAG AVR Reset
2 $0002 INT0 External Interrupt Request 0
6 $000A INT4 External Interrupt Request 4
7 $000C INT5 External Interrupt Request 5
8 $000E INT6 External Interrupt Request 6
10 $0012 TIMER2 COMP Timer/Counter2 Compare Match
11 $0014 TIMER2 OVF Timer/Counter2 Overflow
12 $0016 TIMER1 CAPT Timer/Counter1 Capture Event
13 $0018 TIMER1 COMPA Timer/Counter1 Compare Match A
14 $001A TIMER1 COMPB Timer/Counter1 Compare Match B
15 $001C TIMER1 OVF Timer/Counter1 Overflow
16 $001E TIMER0 COMP Timer/Counter0 Compare Match
17 $0020 TIMER0 OVF Timer/Counter0 Overflow
18 $0022 SPI, STC SPI Serial Transfer Complete
19 $0024 USART0, RX USART0, Rx Complete
20 $0026 USART0, UDRE USART0 Data Register Empty
21 $0028 USART0, TX USART0, Tx Complete
22 $002A ADC ADC Conversion Complete
23 $002C EE READY EEPROM Ready
24 $002E ANALOG COMP Analog Comparator
25 $0030(3) TIMER1 COMPC Timer/Countre1 Compare Match C
(3)
26 $0032 TIMER3 CAPT Timer/Counter3 Capture Event
(3)
27 $0034 TIMER3 COMPA Timer/Counter3 Compare Match A
(3)
28 $0036 TIMER3 COMPB Timer/Counter3 Compare Match B
29 $0038(3) TIMER3 COMPC Timer/Counter3 Compare Match C
(3)
30 $003A TIMER3 OVF Timer/Counter3 Overflow
54 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 23. Reset and Interrupt Vectors (Continued)

Vector Program
No. Address(2) Source Interrupt Definition
31 $003C(3) USART1, RX USART1, Rx Complete
(3)
32 $003E USART1, UDRE USART1 Data Register Empty
(3)
33 $0040 USART1, TX USART1, Tx Complete
(3)
34 $0042 TWI Two-wire Serial Interface
35 $0044(3) SPM READY Store Program Memory Ready
Notes: 1. When the BOOTRST fuse is programmed, the device will jump to the Boot Loader
address at reset, see “Boot Loader Support – Read-While-Write Self-Programming”
on page 269.
2. When the IVSEL bit in MCUCR is set, interrupt vectors will be moved to the start of
the Boot Flash section. The address of each interrupt vector will then be address in
this table added to the start address of the boot Flash section.
3. The Interrupts on address $0030 - $0044 do not exist in ATmega103 compatibility
mode.
Table 24 shows Reset and interrupt vectors placement for the various combinations of
BOOTRST and IVSEL settings. If the program never enables an interrupt source, the
interrupt vectors are not used, and regular program code can be placed at these loca-
tions. This is also the case if the Reset Vector is in the Application section while the
interrupt vectors are in the Boot section or vice versa.
Table 24. Reset and Interrupt Vectors Placement

BOOTRST IVSEL Reset Address Interrupt Vectors Start Address
1 0 $0000 $0002
1 1 $0000 Boot Reset Address + $0002
0 0 Boot Reset Address $0002
0 1 Boot Reset Address Boot Reset Address + $0002
Note: The Boot Reset Address is shown in Table 112 on page 280. For the BOOTRST fuse “1”
means unprogrammed while “0” means programmed.
55
2467E–AVR–05/02
The most typical and general program setup for the Reset and Interrupt Vector
Addresses in ATmega128 is:
Address LabelsCode Comments
$0000 jmp RESET ; Reset Handler
$0002 jmp EXT_INT0 ; IRQ0 Handler
$000A jmp EXT_INT4 ; IRQ4 Handler
$000C jmp EXT_INT5 ; IRQ5 Handler
$000E jmp EXT_INT6 ; IRQ6 Handler
$0012 jmp TIM2_COMP ; Timer2 Compare Handler
$0014 jmp TIM2_OVF ; Timer2 Overflow Handler
$0016 jmp TIM1_CAPT ; Timer1 Capture Handler
$0018 jmp TIM1_COMPA ; Timer1 CompareA Handler
$001A jmp TIM1_COMPB ; Timer1 CompareB Handler
$001C jmp TIM1_OVF ; Timer1 Overflow Handler
$001E jmp TIM0_COMP ; Timer0 Compare Handler
$0020 jmp TIM0_OVF ; Timer0 Overflow Handler
$0022 jmp SPI_STC ; SPI Transfer Complete Handler
$0024 jmp USART0_RXC ; USART0 RX Complete Handler
$0026 jmp USART0_DRE ; USART0,UDR Empty Handler
$0028 jmp USART0_TXC ; USART0 TX Complete Handler
$002A jmp ADC ; ADC Conversion Complete Handler
$002C jmp EE_RDY ; EEPROM Ready Handler
$002E jmp ANA_COMP ; Analog Comparator Handler
$0030 jmp TIM1_COMPC ; Timer1 CompareC Handler
$0032 jmp TIM3_CAPT ; Timer3 Capture Handler
$0034 jmp TIM3_COMPA ; Timer3 CompareA Handler
$0036 jmp TIM3_COMPB ; Timer3 CompareB Handler
$0038 jmp TIM3_COMPC ; Timer3 CompareC Handler
$003A jmp TIM3_OVF ; Timer3 Overflow Handler
$003C jmp USART1_RXC ; USART1 RX Complete Handler
$003E jmp USART1_DRE; USART1,UDR Empty Handler
$0040 jmp USART1_TXC ; USART1 TX Complete Handler
$0042 jmp TWI ; Two-wire Serial Interface Interrupt
Handler
$0044 jmp SPM_RDY ; SPM Ready Handler
;
$0046 RESET:ldir16, high(RAMEND); Main program start
$0047 out SPH,r16 ; Set stack pointer to top of RAM
$0048 ldi r16, low(RAMEND)
$0049 out SPL,r16
$004A sei ; Enable interrupts
$004B <instr> xxx
... ... ... ...
56 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
When the BOOTRST fuse is unprogrammed, the Boot section size set to 8K bytes and
the IVSEL bit in the MCUCR Register is set before any interrupts are enabled, the most
typical and general program setup for the Reset and Interrupt Vector Addresses is:
$0000 RESET:ldi r16,high(RAMEND) ; Main program start
$0001 out SPH,r16 ; Set stack pointer to top of RAM
$0002 ldi r16,low(RAMEND)
$0003 out SPL,r16
$0004 sei ; Enable interrupts
$0005 <instr> xxx
;
.org $F002
$F002 jmp EXT_INT0 ; IRQ0 Handler
... ... ... ;
$F044 jmp SPM_RDY ; Store Program Memory Ready Handler
When the BOOTRST fuse is programmed and the Boot section size set to 8K bytes, the
most typical and general program setup for the Reset and Interrupt Vector Addresses is:
.org $0002
... ... ... ;
$0044 jmp SPM_RDY ; Store Program Memory Ready Handler
;
.org $F000
$F000 RESET: ldi r16,high(RAMEND) ; Main program start
$F001 out SPH,r16 ; Set stack pointer to top of RAM
$F002 ldi r16,low(RAMEND)
$F003 out SPL,r16
$F004 sei ; Enable interrupts
$F005 <instr> xxx
When the BOOTRST fuse is programmed, the Boot section size set to 8K bytes and the
IVSEL bit in the MCUCR Register is set before any interrupts are enabled, the most typ-
ical and general program setup for the Reset and Interrupt Vector Addresses is:
Address Labels Code Comments
;
.org $F000
$F000 jmp RESET ; Reset handler
... ... ... ;
$F044 jmp SPM_RDY ; Store Program Memory Ready Handler
$F046 RESET: ldi r16,high(RAMEND) ; Main program start
$F047 out SPH,r16 ; Set stack pointer to top of RAM
$F048 ldi r16,low(RAMEND)
$F049 out SPL,r16
$F04A sei ; Enable interrupts
$F04B <instr> xxx
57
2467E–AVR–05/02
Moving Interrupts Between The General Interrupt Control Register controls the placement of the interrupt vector
Application and Boot Space table.
MCU Control Register –

MCUCR Bit 7 6 5 4 3 2 1 0
• Bit 1 – IVSEL: Interrupt Vector Select

When the IVSEL bit is cleared (zero), the interrupt vectors are placed at the start of the
Flash memory. When this bit is set (one), the interrupt vectors are moved to the begin-
ning of the Boot Loader section of the flash. The actual address of the start of the Boot
Flash section is determined by the BOOTSZ fuses. Refer to the section “Boot Loader
Support – Read-While-Write Self-Programming” on page 269 for details. To avoid unin-
tentional changes of interrupt vector tables, a special write procedure must be followed
to change the IVSEL bit:
1. Write the Interrupt Vector Change Enable (IVCE) bit to one.
2. Within four cycles, write the desired value to IVSEL while writing a zero to IVCE.
Interrupts will automatically be disabled while this sequence is executed. Interrupts are
disabled in the cycle IVCE is set, and they remain disabled until after the instruction
following the write to IVSEL. If IVSEL is not written, interrupts remain disabled for four
cycles. The I-bit in the Status Register is unaffected by the automatic disabling.
Note: If interrupt vectors are placed in the Boot Loader section and Boot Lock bit BLB02 is pro-
grammed, interrupts are disabled while executing from the Application section. If
interrupt vectors are placed in the Application section and Boot Lock bit BLB12 is pro-
gramed, interrupts are disabled while executing from the Boot Loader section. Refer to
the section “Boot Loader Support – Read-While-Write Self-Programming” on page 269
for details on Boot Lock bits.
58 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
• Bit 0 – IVCE: Interrupt Vector Change Enable

The IVCE bit must be written to logic one to enable change of the IVSEL bit. IVCE is
cleared by hardware four cycles after it is written or when IVSEL is written. Setting the
IVCE bit will disable interrupts, as explained in the IVSEL description above. See Code
Example below.
Move_interrupts:
; Enable change of interrupt vectors
ldi r16, (1<<IVCE)
out MCUCR, r16
; Move interrupts to boot flash section
ldi r16, (1<<IVSEL)
out MCUCR, r16
ret
C Code Example
void Move_interrupts(void)
{
/* Enable change of interrupt vectors */
MCUCR = (1<<IVCE);
/* Move interrupts to boot flash section */
MCUCR = (1<<IVSEL);
}
59
2467E–AVR–05/02
I/O Ports
Introduction All AVR ports have true Read-Modify-Write functionality when used as general digital
I/O ports. This means that the direction of one port pin can be changed without uninten-
tionally changing the direction of any other pin with the SBI and CBI instructions. The
same applies when changing drive value (if configured as output) or enabling/disabling
of pull-up resistors (if configured as input). Each output buffer has symmetrical drive
characteristics with both high sink and source capability. The pin driver is strong enough
to drive LED displays directly. All port pins have individually selectable pull-up resistors
with a supply-voltage invariant resistance. All I/O pins have protection diodes to both
VCC and Ground as indicated in Figure 28. Refer to “Electrical Characteristics” on page
315 for a complete list of parameters.
Figure 28. I/O Pin Equivalent Schematic
RPU
Pxn Logic
CPIN
See Figure
"General Digital I/O" for
Details
All registers and bit references in this section are written in general form. A lower case
“x” represents the numbering letter for the port, and a lower case “n” represents the bit
number. However, when using the register or bit defines in a program, the precise form
must be used. For example, PORTB3 for bit no. 3 in Port B, here documented generally
as PORTxn. The physical I/O registers and bit locations are listed in “Register Descrip-
tion for I/O Ports” on page 81.
Three I/O memory address locations are allocated for each port, one each for the Data
Register – PORTx, Data Direction Register – DDRx, and the Port Input Pins – PINx. The
Port Input Pins I/O location is read only, while the Data Register and the Data Direction
Register are read/write. In addition, the Pull-up Disable – PUD bit in SFIOR disables the
pull-up function for all pins in all ports when set.
Using the I/O port as General Digital I/O is described in “Ports as General Digital I/O” on
page 61. Most port pins are multiplexed with alternate functions for the peripheral fea-
tures on the device. How each alternate function interferes with the port pin is described
in “Alternate Port Functions” on page 65. Refer to the individual module sections for a
full description of the alternate functions.
Note that enabling the alternate function of some of the port pins does not affect the use
of the other pins in the port as General Digital I/O.
60 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Ports as General Digital The ports are bi-directional I/O ports with optional internal pull-ups. Figure 29 shows a
I/O functional description of one I/O port pin, here generically called Pxn.
Figure 29. General Digital I/O(1)
PUD
Q D
DDxn
Q CLR
WDx
RESET
RDx
DATA BUS
Pxn Q D
PORTxn
Q CLR
WPx
RESET
SLEEP RRx
SYNCHRONIZER
RPx
D Q D Q
PINxn
L Q Q
clk I/O
WDx: WRITE DDRx

PUD: PULLUP DISABLE RDx: READ DDRx
SLEEP: SLEEP CONTROL WPx: WRITE PORTx
clkI/O: I/O CLOCK RRx: READ PORTx REGISTER
RPx: READ PORTx PIN
Note: 1. WPx, WDx, RRx, RPx, and RDx are common to all pins within the same port. clkI/O,
SLEEP, and PUD are common to all ports.
Configuring the Pin Each port pin consists of three Register bits: DDxn, PORTxn, and PINxn. As shown in
“Register Description for I/O Ports” on page 81, the DDxn bits are accessed at the DDRx
I/O address, the PORTxn bits at the PORTx I/O address, and the PINxn bits at the PINx
I/O address.
The DDxn bit in the DDRx Register selects the direction of this pin. If DDxn is written
logic one, Pxn is configured as an output pin. If DDxn is written logic zero, Pxn is config-
ured as an input pin.
If PORTxn is written logic one when the pin is configured as an input pin, the pull-up
resistor is activated. To switch the pull-up resistor off, PORTxn has to be written logic
zero or the pin has to be configured as an output pin. The port pins are tri-stated when a
Reset condition becomes active, even if no clocks are running.
If PORTxn is written logic one when the pin is configured as an output pin, the port pin is
driven high (one). If PORTxn is written logic zero when the pin is configured as an out-
put pin, the port pin is driven low (zero).
61
2467E–AVR–05/02
When switching between tri-state ({DDxn, PORTxn} = 0b00) and output high ({DDxn,
PORTxn} = 0b11), an intermediate state with either pull-up enabled ({DDxn, PORTxn} =
0b01) or output low ({DDxn, PORTxn} = 0b10) must occur. Normally, the pull-up
enabled state is fully acceptable, as a high-impedant environment will not notice the dif-
ference between a strong high driver and a pull-up. If this is not the case, the PUD bit in
the SFIOR Register can be written to one to disable all pull-ups in all ports.
Switching between input with pull-up and output low generates the same problem. The
user must use either the tri-state ({DDxn, PORTxn} = 0b00) or the output high state
({DDxn, PORTxn} = 0b10) as an intermediate step.
Table 25 summarizes the control signals for the pin value.
Table 25. Port Pin Configurations

PUD
DDxn PORTxn (in SFIOR) I/O Pull-up Comment
0 0 X Input No Tri-state (Hi-Z)
Pxn will source current if ext. pulled
0 1 0 Input Yes low.
0 1 1 Input No Tri-state (Hi-Z)
1 0 X Output No Output Low (Sink)
1 1 X Output No Output High (Source)
Reading the Pin Value Independent of the setting of Data Direction bit DDxn, the port pin can be read through
the PINxn Register bit. As shown in Figure 29, the PINxn Register bit and the preceding
latch constitute a synchronizer. This is needed to avoid metastability if the physical pin
changes value near the edge of the internal clock, but it also introduces a delay. Figure
30 shows a timing diagram of the synchronization when reading an externally applied
pin value. The maximum and minimum propagation delays are denoted tpd,max and tpd,min
respectively.
Figure 30. Synchronization when Reading an Externally Applied Pin Value
SYSTEM CLK
INSTRUCTIONS XXX XXX in r17, PINx
SYNC LATCH
PINxn
r17 0x00 0xFF
tpd, max
tpd, min
62 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Consider the clock period starting shortly after the first falling edge of the system clock.
The latch is closed when the clock is low, and goes transparent when the clock is high,
as indicated by the shaded region of the “SYNC LATCH” signal. The signal value is
latched when the system clock goes low. It is clocked into the PINxn Register at the suc-
ceeding positive clock edge. As indicated by the two arrows tpd,max and tpd,min, a single
signal transition on the pin will be delayed between ½ and 1½ system clock period
depending upon the time of assertion.
When reading back a software assigned pin value, a nop instruction must be inserted as
indicated in Figure 31. The out instruction sets the “SYNC LATCH” signal at the positive
edge of the clock. In this case, the delay tpd through the synchronizer is one system
clock period.
Figure 31. Synchronization when Reading a Software Assigned Pin Value
SYSTEM CLK
r16 0xFF
INSTRUCTIONS out PORTx, r16 nop in r17, PINx
SYNC LATCH
PINxn
r17 0x00 0xFF
tpd
63
2467E–AVR–05/02
The following code example shows how to set port B pins 0 and 1 high, 2 and 3 low, and
define the port pins from 4 to 7 as input with pull-ups assigned to port pins 6 and 7. The
resulting pin values are read back again, but as previously discussed, a nop instruction
is included to be able to read back the value recently assigned to some of the pins.
...
; Define pull-ups and set outputs high
; Define directions for port pins
ldi r16,(1<<PB7)|(1<<PB6)|(1<<PB1)|(1<<PB0)
ldi r17,(1<<DDB3)|(1<<DDB2)|(1<<DDB1)|(1<<DDB0)
out PORTB,r16
out DDRB,r17
; Insert nop for synchronization
nop
; Read port pins
in r16,PINB
...
C Code Example(1)
unsigned char i;
...
/* Define pull-ups and set outputs high */
/* Define directions for port pins */
PORTB = (1<<PB7)|(1<<PB6)|(1<<PB1)|(1<<PB0);
DDRB = (1<<DDB3)|(1<<DDB2)|(1<<DDB1)|(1<<DDB0);
/* Insert nop for synchronization*/
_NOP();
/* Read port pins */
i = PINB;
...
Note: 1. For the assembly program, two temporary registers are used to minimize the time
from pull-ups are set on pins 0, 1, 6, and 7, until the direction bits are correctly set,
defining bit 2 and 3 as low and redefining bits 0 and 1 as strong high drivers.
Digital Input Enable and Sleep As shown in Figure 29, the digital input signal can be clamped to ground at the input of
Modes the schmitt-trigger. The signal denoted SLEEP in the figure, is set by the MCU Sleep
Controller in Power-down mode, Power-save mode, Standby mode, and Extended
Standby mode to avoid high power consumption if some input signals are left floating, or
have an analog signal level close to VCC/2.
SLEEP is overridden for port pins enabled as External Interrupt pins. If the External
Interrupt Request is not enabled, SLEEP is active also for these pins. SLEEP is also
overridden by various other alternate functions as described in “Alternate Port Func-
tions” on page 65.
If a logic high level (“one”) is present on an Asynchronous External Interrupt pin config-
ured as “Interrupt on Any Logic Change on Pin” while the External Interrupt is not
enabled, the corresponding External Interrupt Flag will be set when resuming from the
above mentioned sleep modes, as the clamping in these sleep modes produces the
requested logic change.
64 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Alternate Port Functions Most port pins have alternate functions in addition to being general digital I/Os. Figure
32 shows how the port pin control signals from the simplified Figure 29 can be overrid-
den by alternate functions. The overriding signals may not be present in all port pins, but
the figure serves as a generic description applicable to all port pins in the AVR micro-
controller family.
Figure 32. Alternate Port Functions(1)
PUOExn
PUOVxn
1
0
PUD
DDOExn
DDOVxn
1
0 Q D
DDxn
Q CLR
WDx
PVOExn
RESET
PVOVxn RDx
DATA BUS
1
Pxn
0 Q D
PORTxn
Q CLR
DIEOExn
WPx
DIEOVxn RESET
1
RRx
0 SLEEP
SYNCHRONIZER
RPx
SET
D Q D Q
PINxn
L CLR Q CLR Q
clk I/O
DIxn
AIOxn
PUOExn: Pxn PULL-UP OVERRIDE ENABLE PUD: PULLUP DISABLE

PUOVxn: Pxn PULL-UP OVERRIDE VALUE WDx: WRITE DDRx
DDOExn: Pxn DATA DIRECTION OVERRIDE ENABLE RDx: READ DDRx
DDOVxn: Pxn DATA DIRECTION OVERRIDE VALUE RRx: READ PORTx REGISTER
PVOExn: Pxn PORT VALUE OVERRIDE ENABLE WPx: WRITE PORTx
PVOVxn: Pxn PORT VALUE OVERRIDE VALUE RPx: READ PORTx PIN
DIEOExn: Pxn DIGITAL INPUT-ENABLE OVERRIDE ENABLE clkI/O: I/O CLOCK
DIEOVxn: Pxn DIGITAL INPUT-ENABLE OVERRIDE VALUE DIxn: DIGITAL INPUT PIN n ON PORTx
SLEEP: SLEEP CONTROL AIOxn: ANALOG INPUT/OUTPUT PIN n ON PORTx
Note: 1. WPx, WDx, RLx, RPx, and RDx are common to all pins within the same port. clkI/O,
SLEEP, and PUD are common to all ports. All other signals are unique for each pin.
Table 26 summarizes the function of the overriding signals. The pin and port indexes
from Figure 32 are not shown in the succeeding tables. The overriding signals are gen-
erated internally in the modules having the alternate function.
65
2467E–AVR–05/02
Table 26. Generic Description of Overriding Signals for Alternate Functions.
Signal
Name Full Name Description
PUOE Pull-up If this signal is set, the pull-up enable is controlled by the
Override Enable PUOV signal. If this signal is cleared, the pull-up is
enabled when {DDxn, PORTxn, PUD} = 0b010.
PUOV Pull-up If PUOE is set, the pull-up is enabled/disabled when
Override Value PUOV is set/cleared, regardless of the setting of the
DDxn, PORTxn, and PUD Register bits.
DDOE Data Direction If this signal is set, the Output Driver Enable is controlled
Override Enable by the DDOV signal. If this signal is cleared, the Output
driver is enabled by the DDxn Register bit.
DDOV Data Direction If DDOE is set, the Output Driver is enabled/disabled
Override Value when DDOV is set/cleared, regardless of the setting of
the DDxn Register bit.
PVOE Port Value If this signal is set and the Output Driver is enabled, the
Override Enable port value is controlled by the PVOV signal. If PVOE is
cleared, and the Output Driver is enabled, the port Value
is controlled by the PORTxn Register bit.
PVOV Port Value If PVOE is set, the port value is set to PVOV, regardless
Override Value of the setting of the PORTxn Register bit.
DIEOE Digital Input If this bit is set, the Digital Input Enable is controlled by
Enable Override the DIEOV signal. If this signal is cleared, the Digital Input
Enable Enable is determined by MCU-state (Normal mode, Sleep
modes).
DIEOV Digital Input If DIEOE is set, the Digital Input is enabled/disabled when
Enable Override DIEOV is set/cleared, regardless of the MCU state
Value (Normal mode, Sleep modes).
DI Digital Input This is the Digital Input to alternate functions. In the
figure, the signal is connected to the output of the schmitt
trigger but before the synchronizer. Unless the Digital
Input is used as a clock source, the module with the
alternate function will use its own synchronizer.
AIO Analog This is the Analog Input/output to/from alternate functions.
Input/output The signal is connected directly to the pad, and can be
used bi-directionally.
The following subsections shortly describes the alternate functions for each port, and
relates the overriding signals to the alternate function. Refer to the alternate function
description for further details.
Special Function IO Register –

SFIOR Bit 7 6 5 4 3 2 1 0
TSM – – ADHSM ACME PUD PSR0 PSR321 SFIOR
66 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
• Bit 2 – PUD: Pull-up disable

When this bit is written to one, the pull-ups in the I/O ports are disabled even if the DDxn
and PORTxn Registers are configured to enable the pull-ups ({DDxn, PORTxn} = 0b01).
See “Configuring the Pin” on page 61 for more details about this feature.
Alternate Functions of Port A The Port A has an alternate function as the address low byte and data lines for the
External Memory Interface.
Table 27. Port A Pins Alternate Functions

Port Pin Alternate Function
PA7 AD7 (External memory interface address and data bit 7)
Table 28 and Table 29 relates the alternate functions of Port A to the overriding signals
shown in Figure 32 on page 65.
Table 28. Overriding Signals for Alternate Functions in PA7..PA4

Signal
Name PA7/AD7 PA6/AD6 PA5/AD5 PA4/AD4
PUOE SRE SRE SRE SRE
(1)
PUOV ~(WR | ADA ) • ~(WR | ADA) • ~(WR | ADA) • ~(WR | ADA) •
PORTA7 • PUD PORTA6 • PUD PORTA5 • PUD PORTA4 • PUD
DDOE SRE SRE SRE SRE
DDOV WR | ADA WR | ADA WR | ADA WR | ADA
PVOE SRE SRE SRE SRE
PVOV A7 • ADA | D7 A6 • ADA | D6 A5 • ADA | D5 A4 • ADA | D4
OUTPUT • WR OUTPUT • WR OUTPUT • WR OUTPUT • WR
DIEOE 0 0 0 0
DIEOV 0 0 0 0
DI D7 INPUT D6 INPUT D5 INPUT D4 INPUT
AIO – – – –
Note: 1. ADA is short for ADdress Active and represents the time when address is output. See
“External Memory Interface” on page 24 for details.
67
2467E–AVR–05/02
Table 29. Overriding Signals for Alternate Functions in PA3..PA0
Signal
Name PA3/AD3 PA2/AD2 PA1/AD1 PA0/AD0
PUOE SRE SRE SRE SRE
PUOV ~(WR | ADA) • ~(WR | ADA) • ~(WR | ADA) • ~(WR | ADA) •
PORTA3 • PUD PORTA2 • PUD PORTA1 • PUD PORTA0 • PUD
DDOE SRE SRE SRE SRE
DDOV WR | ADA WR | ADA WR | ADA WR | ADA
PVOE SRE SRE SRE SRE
PVOV A3 • ADA | D3 A2• ADA | D2 A1 • ADA | D1 A0 • ADA | D0
OUTPUT • WR OUTPUT • WR OUTPUT • WR OUTPUT • WR
DIEOE 0 0 0 0
DIEOV 0 0 0 0
DI D3 INPUT D2 INPUT D1 INPUT D0 INPUT
AIO – – – –
Alternate Functions of Port B The Port B pins with alternate functions are shown in Table 30.
Table 30. Port B Pins Alternate Functions

Port Pin Alternate Functions
OC2/OC1C(1) (Output Compare and PWM Output for Timer/Counter2 or Output
PB7
Compare and PWM Output C for Timer/Counter1)
PB6 OC1B (Output Compare and PWM Output B for Timer/Counter1)
PB5 OC1A (Output Compare and PWM Output A for Timer/Counter1)
PB4 OC0 (Output Compare and PWM Output for Timer/Counter0)
PB3 MISO (SPI Bus Master Input/Slave Output)
PB2 MOSI (SPI Bus Master Output/Slave Input)
PB1 SCK (SPI Bus Serial Clock)
PB0 SS (SPI Slave Select input)
Note: 1. OC1C not applicable in ATmega103 compatibility mode.
The alternate pin configuration is as follows:
• OC2/OC1C, Bit 7
OC2, Output Compare Match output: The PB7 pin can serve as an external output for
the Timer/Counter2 Output Compare. The pin has to be configured as an output (DDB7
set “one”) to serve this function. The OC2 pin is also the output pin for the PWM mode
timer function.
OC1C, Output Compare Match C output: The PB7 pin can serve as an external output
for the Timer/Counter1 Output Compare C. The pin has to be configured as an output
(DDB7 set (one)) to serve this function. The OC1C pin is also the output pin for the
PWM mode timer function.
68 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
• OC1B, Bit 6
OC1B, Output Compare Match B output: The PB6 pin can serve as an external output
for the Timer/Counter1 Output Compare B. The pin has to be configured as an output
(DDB6 set (one)) to serve this function. The OC1B pin is also the output pin for the PWM
mode timer function.
• OC1A, Bit 5
OC1A, Output Compare Match A output: The PB5 pin can serve as an external output
for the Timer/Counter1 Output Compare A. The pin has to be configured as an output
(DDB5 set (one)) to serve this function. The OC1A pin is also the output pin for the PWM
• OC0, Bit 4
OC0, Output Compare Match output: The PB4 pin can serve as an external output for
the Timer/Counter0 Output Compare. The pin has to be configured as an output (DDB4
set (one)) to serve this function. The OC0 pin is also the output pin for the PWM mode
timer function.
• MISO – Port B, Bit 3

MISO: Master Data input, Slave Data output pin for SPI channel. When the SPI is
enabled as a master, this pin is configured as an input regardless of the setting of
DDB3. When the SPI is enabled as a slave, the data direction of this pin is controlled by
DDB3. When the pin is forced to be an input, the pull-up can still be controlled by the
PORTB3 bit.
• MOSI – Port B, Bit 2

MOSI: SPI Master Data output, Slave Data input for SPI channel. When the SPI is
enabled as a slave, this pin is configured as an input regardless of the setting of DDB2.
When the SPI is enabled as a master, the data direction of this pin is controlled by
PORTB2 bit.
• SCK – Port B, Bit 1

SCK: Master Clock output, Slave Clock input pin for SPI channel. When the SPI is
enabled as a slave, this pin is configured as an input regardless of the setting of DDB1.
When the SPI is enabled as a master, the data direction of this pin is controlled by
PORTB1 bit.
• SS – Port B, Bit 0
SS: Slave Port Select input. When the SPI is enabled as a slave, this pin is configured
as an input regardless of the setting of DDB0. As a slave, the SPI is activated when this
pin is driven low. When the SPI is enabled as a master, the data direction of this pin is
controlled by DDB0. When the pin is forced to be an input, the pull-up can still be con-
trolled by the PORTB0 bit.
Table 31 and Table 32 relate the alternate functions of Port B to the overriding signals
shown in Figure 32 on page 65. SPI MSTR INPUT and SPI SLAVE OUTPUT constitute
the MISO signal, while MOSI is divided into SPI MSTR OUTPUT and SPI SLAVE
INPUT.
69
2467E–AVR–05/02
Table 31. Overriding Signals for Alternate Functions in PB7..PB4
Signal Name PB7/OC2/OC1C PB6/OC1B PB5/OC1A PB4/OC0
PUOE 0 0 0 0
PUOV 0 0 0 0
DDOE 0 0 0 0
DDOV 0 0 0 0
(1)
PVOE OC2/OC1C ENABLE OC1B ENABLE OC1A ENABLE OC0 ENABLE
PVOV OC2/OC1C(1) OC1B OC1A OC0B
DIEOE 0 0 0 0
DIEOV 0 0 0 0
DI – – – –
AIO – – – –
Note: 1. See “Output Compare Modulator (OCM1C2)” on page 156 for details. OC1C does
not exist in ATmega103 compatibility mode.
Table 32. Overriding Signals for Alternate Functions in PB3..PB0

Signal
Name PB3/MISO PB2/MOSI PB1/SCK PB0/SS
PUOE SPE • MSTR SPE • MSTR SPE • MSTR SPE • MSTR
PUOV PORTB3 • PUD PORTB2 • PUD PORTB1 • PUD PORTB0 • PUD
DDOE SPE • MSTR SPE • MSTR SPE • MSTR SPE • MSTR
DDOV 0 0 0 0
PVOE SPE • MSTR SPE • MSTR SPE • MSTR 0
PVOV SPI SLAVE OUTPUT SPI MSTR OUTPUT SCK OUTPUT 0
DIEOE 0 0 0 0
DIEOV 0 0 0 0
DI SPI MSTR INPUT SPI SLAVE INPUT SCK INPUT SPI SS
AIO – – – –
70 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Alternate Functions of Port C In ATmega103 compatibility mode, Port C is output only. The Port C has an alternate
function as the address high byte for the External Memory Interface.
Table 33. Port C Pins Alternate Functions

PC7 A15
PC6 A14
PC5 A13
PC4 A12
PC3 A11
PC2 A10
PC1 A9
PC0 A8
Table 34 and Table 35 relate the alternate functions of Port C to the overriding signals
Table 34. Overriding Signals for Alternate Functions in PC7..PC4

Signal
Name PC7/A15 PC6/A14 PC5/A13 PC4/A12
(1)
PUOE SRE • (XMM <1) SRE • (XMM<2) SRE • (XMM<3) SRE • (XMM<4)
PUOV 0 0 0 0
DDOE SRE • (XMM<1) SRE • (XMM<2) SRE • (XMM<3) SRE • (XMM<4)
DDOV 1 1 1 1
PVOE SRE • (XMM<1) SRE • (XMM<2) SRE • (XMM<3) SRE • (XMM<4)
PVOV A11 A10 A9 A8
DIEOE 0 0 0 0
DIEOV 0 0 0 0
DI – – – –
AIO – – – –
Note: 1. XMM = 0 in ATmega103 compatibility mode.
71
2467E–AVR–05/02
Table 35. Overriding Signals for Alternate Functions in PC3..PC0(1)
Signal
Name PC3/A11 PC2/A10 PC1/A9 PC0/A8
PUOE SRE • (XMM<5) SRE • (XMM<6) SRE • (XMM<7) SRE • (XMM<7)
PUOV 0 0 0 0
DDOE SRE • (XMM<5) SRE • (XMM<6) SRE • (XMM<7) SRE • (XMM<7)
DDOV 1 1 1 1
PVOE SRE • (XMM<5) SRE • (XMM<6) SRE • (XMM<7) SRE • (XMM<7)
PVOV A11 A10 A9 A8
DIEOE 0 0 0 0
DIEOV 0 0 0 0
DI – – – –
AIO – – – –
Note: 1. XMM = 0 in ATmega103 compatibility mode.
Alternate Functions of Port D The Port D pins with alternate functions are shown in Table 36.
Table 36. Port D Pins Alternate Functions

PD7 T2 (Timer/Counter2 Clock Input)
PD6 T1 (Timer/Counter1 Clock Input)
PD5 XCK1(1) (USART1 External Clock Input/Output)
PD4 IC1 (Timer/Counter1 Input Capture Trigger)
PD3 INT3/TXD1(1) (External Interrupt3 Input or UART1 Transmit Pin)
PD2 INT2/RXD1(1) (External Interrupt2 Input or UART1 Receive Pin)
PD1 INT1/SDA(1) (External Interrupt1 Input or TWI Serial DAta)
PD0 INT0/SCL(1) (External Interrupt0 Input or TWI Serial CLock)
Note: 1. XCK1, TXD1, RXD1, SDA, and SCL not applicable in ATmega103 compatibility
mode.
The alternate pin configuration is as follows:
• T2 – Port D, Bit 7
T2, Timer/Counter2 counter source.
• T1 – Port D, Bit 6
• XCK1 – Port D, Bit 4

XCK1, USART1 External clock. The Data Direction Register (DDD4) controls whether
the clock is output (DDD4 set) or input (DDD4 cleared). The XCK1 pin is active only
when the USART1 operates in Synchronous mode.
72 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
• IC1 – Port D, Bit 4

IC 1 – In pu t Ca ptu re Pin 1: Th e P D4 pin can ac t as an in pu t ca ptu re pin fo r
Timer/Counter1.
• INT3/TXD1 – Port D, Bit 3

INT3, External Interrupt source 3: The PD3 pin can serve as an external interrupt source
to the MCU.
TXD1, Transmit Data (Data output pin for the USART1). When the USART1 Transmitter
is enabled, this pin is configured as an output regardless of the value of DDD3.
• INT2/RXD1 – Port D, Bit 2

INT2, External Interrupt source 2. The PD2 pin can serve as an External Interrupt
source to the MCU.
RXD1, Receive Data (Data input pin for the USART1). When the USART1 receiver is
enabled this pin is configured as an input regardless of the value of DDD2. When the
USART forces this pin to be an input, the pull-up can still be controlled by the PORTD2
bit.
• INT1/SDA – Port D, Bit 1

INT1, External Interrupt source 1. The PD1 pin can serve as an external interrupt source
to the MCU.
SDA, Two-wire Serial Interface Data: When the TWEN bit in TWCR is set (one) to
enable the Two-wire Serial Interface, pin PD1 is disconnected from the port and
becomes the Serial Data I/O pin for the Two-wire Serial Interface. In this mode, there is
a spike filter on the pin to suppress spikes shorter than 50 ns on the input signal, and the
pin is driven by an open drain driver with slew-rate limitation.
• INT0/SCL – Port D, Bit 0

INT0, External Interrupt source 0. The PD0 pin can serve as an external interrupt source
to the MCU.
SCL, Two-wire Serial Interface Clock: When the TWEN bit in TWCR is set (one) to
enable the Two-wire Serial Interface, pin PD0 is disconnected from the port and
becomes the Serial Clock I/O pin for the Two-wire Serial Interface. In this mode, there is
a spike filter on the pin to suppress spikes shorter than 50 ns on the input signal, and the
pin is driven by an open drain driver with slew-rate limitation.
Table 37 and Table 38 relates the alternate functions of Port D to the overriding signals
73
2467E–AVR–05/02
Table 37. Overriding Signals for Alternate Functions PD7..PD4
Signal Name PD7/T2 PD6/T1 PD5/XCK1 PD4/IC1
PUOE 0 0 0 0
PUOV 0 0 0 0
DDOE 0 0 0 0
DDOV 0 0 0 0
PVOE 0 0 UMSEL1 0
PVOV 0 0 XCK1 OUTPUT 0
DIEOE 0 0 0 0
DIEOV 0 0 0 0
DI T2 INPUT T1 INPUT XCK1 INPUT IC1 INPUT
AIO – – – –
Table 38. Overriding Signals for Alternate Functions in PD3..PD0 (1)

Signal Name PD3/INT3/TXD1 PD2/INT2/RXD1 PD1/INT1/SDA PD0/INT0/SCL
PUOE TXEN1 RXEN1 TWEN TWEN
PUOV 0 PORTD2 • PUD PORTD1 • PUD PORTD0 • PUD
DDOE TXEN1 RXEN1 TWEN TWEN
DDOV 1 0 SDA_OUT SCL_OUT
PVOE TXEN1 0 TWEN TWEN
PVOV TXD1 0 0 0
DIEOE INT3 ENABLE INT2 ENABLE INT1 ENABLE INT0 ENABLE
DIEOV 1 1 1 1
DI INT3 INPUT INT2 INPUT/RXD1 INT1 INPUT INT0 INPUT
AIO – – SDA INPUT SCL INPUT
Note: 1. When enabled, the Two-wire Serial Interface enables Slew-Rate controls on the out-
put pins PD0 and PD1. This is not shown in this table. In addition, spike filters are
connected between the AIO outputs shown in the port figure and the digital logic of
the TWI module.
74 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Alternate Functions of Port E The Port E pins with alternate functions are shown in Table 39.
Table 39. Port E Pins Alternate Functions

PE7 INT7/IC3(1) (External Interrupt 7 Input or Timer/Counter3 Input Capture Trigger)
PE6 INT6/ T3(1) (External Interrupt 6 Input or Timer/Counter3 Clock Input)
INT5/OC3C(1) (External Interrupt 5 Input or Output Compare and PWM Output C
PE5
for Timer/Counter3)
INT4/OC3B(1) (External Interrupt4 Input or Output Compare and PWM Output B for
PE4
Timer/Counter3)
AIN1/OC3A (1) (Analog Comparator Negative Input or Output Compare and PWM
PE3
Output A for Timer/Counter3)
AIN0/XCK0(1) (Analog Comparator Positive Input or USART0 external clock
PE2
input/output)
PE1 PDO/TXD0 (Programming Data Output or UART0 Transmit Pin)
PE0 PDI/RXD0 (Programming Data Input or UART0 Receive Pin)
Note: 1. IC3, T3, OC3C, OC3B, OC3B, OC3A, and XCK0 not applicable in ATmega103 com-
patibility mode.
• INT7/IC3 – Port E, Bit 7

INT7, External Interrupt source 7: The PE7 pin can serve as an external interrupt
source.
IC 3 – In pu t C ap ture Pin 3: Th e P E7 pin c an a ct as an inp ut cap tu re pin fo r
Timer/Counter3.
• INT6/T3 – Port E, Bit 6

INT6, External Interrupt source 6: The PE6 pin can serve as an external interrupt
source.
• INT5/OC3C – Port E, Bit 5

INT5, External Interrupt source 5: The PE5 pin can serve as an External Interrupt
source.
OC3C, Output Compare Match C output: The PE5 pin can serve as an External output
for the Timer/Counter3 Output Compare C. The pin has to be configured as an output
(DDE5 set “one”) to serve this function. The OC3C pin is also the output pin for the
PWM mode timer function.
• INT4/OC3B – Port E, Bit 4

INT4, External Interrupt source 4: The PE4 pin can serve as an External Interrupt
source.
OC3B, Output Compare Match B output: The PE4 pin can serve as an External output
for the Timer/Counter3 Output Compare B. The pin has to be configured as an output
(DDE4 set (one)) to serve this function. The OC3B pin is also the output pin for the PWM
75
2467E–AVR–05/02
• AIN1/OC3A – Port E, Bit 3
AIN1 – Analog Comparator Negative input. This pin is directly connected to the negative
input of the Analog Comparator.
OC3A, Output Compare Match A output: The PE3 pin can serve as an External output
for the Timer/Counter3 Output Compare A. The pin has to be configured as an output
(DDE3 set “one”) to serve this function. The OC3A pin is also the output pin for the PWM
• AIN0/XCK0 – Port E, Bit 2

AIN0 – Analog Comparator Positive input. This pin is directly connected to the positive
input of the Analog Comparator.
XCK0, USART0 External clock. The Data Direction Register (DDE2) controls whether
the clock is output (DDE2 set) or input (DDE2 cleared). The XCK0 pin is active only
when the USART0 operates in Synchronous mode.
• PDO/TXD0 – Port E, Bit 1

PDO, SPI Serial Programming Data Output. During Serial Program Downloading, this
pin is used as data output line for the ATmega128.
TXD0, UART0 Transmit pin.
• PDI/RXD0 – Port E, Bit 0

PDI, SPI Serial Programming Data Input. During Serial Program Downloading, this pin
is used as data input line for the ATmega128.
RXD0, USART0 Receive Pin. Receive Data (Data input pin for the USART0). When the
USART0 receiver is enabled this pin is configured as an input regardless of the value of
DDRE0. When the USART0 forces this pin to be an input, a logical one in PORTE0 will
turn on the internal pull-up.
Table 40 and Table 41 relates the alternate functions of Port E to the overriding signals
Table 40. Overriding Signals for Alternate Functions PE7..PE4

Signal
Name PE7/INT7/IC3 PE6/INT6/T3 PE5/INT5/OC3C PE4/INT4/OC3B
PUOE 0 0 0 0
PUOV 0 0 0 0
DDOE 0 0 0 0
DDOV 0 0 0 0
PVOE 0 0 OC3C ENABLE OC3B ENABLE
PVOV 0 0 OC3C OC3B
DIEOE INT7 ENABLE INT6 ENABLE INT5 ENABLE INT4 ENABLE
DIEOV 1 1 1 1
DI INT7 INPUT/IC3 INT7 INPUT/T3 INT5 INPUT INT4 INPUT
INPUT INPUT
AIO – – – –
76 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 41. Overriding Signals for Alternate Functions in PE3..PE0

Signal Name PE3/AIN1/OC3A PE2/AIN0/XCK0 PE1/PDO/TXD0 PE0/PDI/RXD0
PUOE 0 0 TXEN0 RXEN0
PUOV 0 0 0 PORTE0 • PUD
DDOE 0 0 TXEN0 RXEN0
DDOV 0 0 1 0
PVOE OC3B ENABLE UMSEL0 TXEN0 0
PVOV OC3B XCK0 OUTPUT TXD0 0
DIEOE 0 0 0 0
DIEOV 0 0 0 0
DI 0 XCK0 INPUT – RXD0
AIO AIN1 INPUT AIN0 INPUT – –
Alternate Functions of Port F The Port F has an alternate function as analog input for the ADC as shown in Table 42.
If some Port F pins are configured as outputs, it is essential that these do not switch
when a conversion is in progress. This might corrupt the result of the conversion. In
ATmega103 compatibility mode Port F is input only. If the JTAG interface is enabled, the
pull-up resistors on pins PF7(TDI), PF5(TMS), and PF4(TCK) will be activated even if a
Reset occurs.
Table 42. Port F Pins Alternate Functions

PF7 ADC7/TDI (ADC input channel 7 or JTAG Test Data Input)
PF6 ADC6/TDO (ADC input channel 6 or JTAG Test Data Output)
PF5 ADC5/TMS (ADC input channel 5 or JTAG Test Mode Select)
PF4 ADC4/TCK (ADC input channel 4 or JTAG Test ClocK)
PF3 ADC3 (ADC input channel 3)
• TDI, ADC7 – Port F, Bit 7

ADC7, Analog to Digital Converter, Channel 7.
TDI, JTAG Test Data In: Serial input data to be shifted in to the Instruction Register or
Data Register (scan chains). When the JTAG interface is enabled, this pin can not be
used as an I/O pin.
• TCK, ADC6 – Port F, Bit 6

TDO, JTAG Test Data Out: Serial output data from Instruction Register or Data Register.
When the JTAG interface is enabled, this pin can not be used as an I/O pin.
77
2467E–AVR–05/02
• TMS, ADC5 – Port F, Bit 5
TMS, JTAG Test Mode Select: This pin is used for navigating through the TAP-controller
state machine. When the JTAG interface is enabled, this pin can not be used as an I/O
pin.
• TDO, ADC4 – Port F, Bit 4

TCK, JTAG Test Clock: JTAG operation is synchronous to TCK. When the JTAG inter-
face is enabled, this pin can not be used as an I/O pin.
• ADC3 – ADC0 – Port F, Bit 3..0

Analog to Digital Converter, Channel 3..0.
Table 43. Overriding Signals for Alternate Functions in PF7..PF4

Signal
Name PF7/ADC7/TDI PF6/ADC6/TDO PF5/ADC5/TMS PF4/ADC4/TCK
PUOE JTAGEN JTAGEN JTAGEN JTAGEN
PUOV 1 0 1 1
DDOE JTAGEN JTAGEN JTAGEN JTAGEN
DDOV 0 SHIFT_IR + 0 0
SHIFT_DR
PVOE 0 JTAGEN 0 0
PVOV 0 TDO 0 0
DIEOE JTAGEN JTAGEN JTAGEN JTAGEN
DIEOV 0 0 0 0
DI – – – –
AIO TDI/ADC7 INPUT ADC6 INPUT TMS/ADC5 TCKADC4 INPUT
INPUT
78 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 44. Overriding Signals for Alternate Functions in PF3..PF0

Signal Name PF3/ADC3 PF2/ADC2 PF1/ADC1 PF0/ADC0
PUOE 0 0 0 0
PUOV 0 0 0 0
DDOE 0 0 0 0
DDOV 0 0 0 0
PVOE 0 0 0 0
PVOV 0 0 0 0
DIEOE 0 0 0 0
DIEOV 0 0 0 0
DI – – – –
AIO ADC3 INPUT ADC2 INPUT ADC1 INPUT ADC0 INPUT
Alternate Functions of Port G In ATmega103 compatibility mode, only the alternate functions are the defaults for Port
G, and Port G cannot be used as General Digital Port Pins. The alternate pin configura-
tion is as follows:
Table 45. Port G Pins Alternate Functions

PG4 TOSC1 (RTC Oscillator Timer/Counter0)
PG3 TOSC2 (RTC Oscillator Timer/Counter0)
PG2 ALE (Address Latch Enable to external memory)
PG1 RD (Read strobe to external memory)
PG0 WR (Write strobe to external memory)
• TOSC1 – Port G, Bit 4

TOSC2, Timer Oscillator pin 1: When the AS0 bit in ASSR is set (one) to enable asyn-
chronous clocking of Timer/Counter0, pin PG4 is disconnected from the port, and
becomes the input of the inverting Oscillator amplifier. In this mode, a Crystal Oscillator
is connected to this pin, and the pin can not be used as an I/O pin.
• TOSC2 – Port G, Bit 3

TOSC2, Timer Oscillator pin 2: When the AS0 bit in ASSR is set (one) to enable asyn-
chronous clocking of Timer/Counter0, pin PG3 is disconnected from the port, and
becomes the inverting output of the Oscillator amplifier. In this mode, a Crystal Oscillator
is connected to this pin, and the pin can not be used as an I/O pin.
• ALE – Port G, Bit 2

ALE is the external data memory Address Latch Enable signal.
• RD – Port G, Bit 1
RD is the external data memory read control strobe.
79
2467E–AVR–05/02
• WR – Port G, Bit 0
WR is the external data memory write control strobe.
Table 46 and Table 47 relates the alternate functions of Port G to the overriding signals
Table 46. Overriding Signals for Alternate Functions in PG4..PG1

Signal Name PG4/TOSC1 PG3/TOSC2 PG2/ALE PG1/RD
PUOE AS0 AS0 SRE SRE
PUOV 0 0 0 0
DDOE AS0 AS0 SRE SRE
DDOV 0 0 1 1
PVOE 0 0 SRE SRE
PVOV 0 0 ALE RD
DIEOE AS0 AS0 0 0
DIEOV 0 0 0 0
DI – – – –
AIO T/C0 OSC INPUT T/C0 OSC OUTPUT – –
Table 47. Overriding Signals for Alternate Functions in PG0

Signal Name PG0/WR
PUOE SRE
PUOV 0
DDOE SRE
DDOV 1
PVOE SRE
PVOV WR
DIEOE 0
DIEOV 0
DI –
AIO –
80 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Register Description for

I/O Ports
Port A Data Register – PORTA

Bit 7 6 5 4 3 2 1 0
PORTA7 PORTA6 PORTA5 PORTA4 PORTA3 PORTA2 PORTA1 PORTA0 PORTA
Port A Data Direction Register

Bit 7 6 5 4 3 2 1 0
– DDRA
DDA7 DDA6 DDA5 DDA4 DDA3 DDA2 DDA1 DDA0 DDRA
Port A Input Pins Address –

Bit 7 6 5 4 3 2 1 0
PINA
PINA7 PINA6 PINA5 PINA4 PINA3 PINA2 PINA1 PINA0 PINA
Read/Write R R R R R R R R
Initial Value N/A N/A N/A N/A N/A N/A N/A N/A
Port B Data Register – PORTB

Bit 7 6 5 4 3 2 1 0
PORTB7 PORTB6 PORTB5 PORTB4 PORTB3 PORTB2 PORTB1 PORTB0 PORTB
Port B Data Direction Register

Bit 7 6 5 4 3 2 1 0
– DDRB
DDB7 DDB6 DDB5 DDB4 DDB3 DDB2 DDB1 DDB0 DDRB
Port B Input Pins Address –

Bit 7 6 5 4 3 2 1 0
PINB
PINB7 PINB6 PINB5 PINB4 PINB3 PINB2 PINB1 PINB0 PINB
Port C Data Register – PORTC

Bit 7 6 5 4 3 2 1 0
PORTC7 PORTC6 PORTC5 PORTC4 PORTC3 PORTC2 PORTC1 PORTC0 PORTC
Port C Data Direction Register

Bit 7 6 5 4 3 2 1 0
– DDRC
DDC7 DDC6 DDC5 DDC4 DDC3 DDC2 DDC1 DDC0 DDRC
81
2467E–AVR–05/02
Port C Input Pins Address –
Bit 7 6 5 4 3 2 1 0
PINC
PINC7 PINC6 PINC5 PINC4 PINC3 PINC2 PINC1 PINC0 PINC
In ATmega103 compatibility mode, DDRC and PINC Registers are initialized to being
Push-Pull Zero Output. The port pins assumes their initial value, even if the clock is not
running. Note that the DDRC and PINC Registers are available in ATmega103 compat-
ibility mode, and should not be used for 100% back-ward compatibility.
Port D Data Register – PORTD

Bit 7 6 5 4 3 2 1 0
PORTD7 PORTD6 PORTD5 PORTD4 PORTD3 PORTD2 PORTD1 PORTD0 PORTD
Port D Data Direction Register

Bit 7 6 5 4 3 2 1 0
– DDRD
DDD7 DDD6 DDD5 DDD4 DDD3 DDD2 DDD1 DDD0 DDRD
Port D Input Pins Address –

Bit 7 6 5 4 3 2 1 0
PIND
PIND7 PIND6 PIND5 PIND4 PIND3 PIND2 PIND1 PIND0 PIND
Port E Data Register – PORTE

Bit 7 6 5 4 3 2 1 0
PORTE7 PORTE6 PORTE5 PORTE4 PORTE3 PORTE2 PORTE1 PORTE0 PORTE
Port E Data Direction Register

Bit 7 6 5 4 3 2 1 0
– DDRE
DDE7 DDE6 DDE5 DDE4 DDE3 DDE2 DDE1 DDE0 DDRE
Port E Input Pins Address –

Bit 7 6 5 4 3 2 1 0
PINE
PINE7 PINE6 PINE5 PINE4 PINE3 PINE2 PINE1 PINE0 PINF
Port F Data Register – PORTF

Bit 7 6 5 4 3 2 1 0
PORTF7 PORTF6 PORTF5 PORTF4 PORTF3 PORTF2 PORTF1 PORTF0 PORTF
82 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Port F Data Direction Register

Bit 7 6 5 4 3 2 1 0
– DDRF
DDF7 DDF6 DDF5 DDF4 DDF3 DDF2 DDF1 DDF0 DDRF
Port F Input Pins Address –

Bit 7 6 5 4 3 2 1 0
PINF
PINF7 PINF6 PINF5 PINF4 PINF3 PINF2 PINF1 PINF0 PINF
Note that PORTF and DDRF Registers are not available in ATmega103 compatibility
mode where Port F serves as digital input only.
Port G Data Register – PORTG

Bit 7 6 5 4 3 2 1 0
– – – PORTG4 PORTG3 PORTG2 PORTG1 PORTG0 PORTG
Port G Data Direction Register

Bit 7 6 5 4 3 2 1 0
– DDRG
– – – DDG4 DDG3 DDG2 DDG1 DDG0 DDRG
Port G Input Pins Address –

Bit 7 6 5 4 3 2 1 0
PING
– – – PING4 PING3 PING2 PING1 PING0 PING
Initial Value 0 0 0 N/A N/A N/A N/A N/A
Note that PORTG, DDRG, and PING are not available in ATmega103 compatibility
mode. In the ATmega103 compatibility mode Port G serves its alternate functions only
(TOSC1, TOSC2, WR, RD and ALE).
83
2467E–AVR–05/02
External Interrupts The External Interrupts are triggered by the INT7:0 pins. Observe that, if enabled, the
interrupts will trigger even if the INT7:0 pins are configured as outputs. This feature pro-
vides a way of generating a software interrupt. The External Interrupts can be triggered
by a falling or rising edge or a low level. This is set up as indicated in the specification for
the External Interrupt Control Registers – EICRA (INT3:0) and EICRB (INT7:4). When
the external interrupt is enabled and is configured as level triggered, the interrupt will
trigger as long as the pin is held low. Note that recognition of falling or rising edge inter-
rupts on INT7:4 requires the presence of an I/O clock, described in “Clock Systems and
their Distribution” on page 33. Low level interrupts and the edge interrupt on INT3:0 are
detected asynchronously. This implies that these interrupts can be used for waking the
part also from sleep modes other than Idle mode. The I/O clock is halted in all sleep
modes except Idle mode.
Note that if a level triggered interrupt is used for wake-up from Power-down mode, the
changed level must be held for some time to wake up the MCU. This makes the MCU
less sensitive to noise. The changed level is sampled twice by the Watchdog Oscillator
clock. The period of the Watchdog Oscillator is 1 µs (nominal) at 5.0V and 25°C. The
frequency of the Watchdog Oscillator is voltage dependent as shown in the “Electrical
Characteristics” on page 315. The MCU will wake up if the input has the required level
during this sampling or if it is held until the end of the start-up time. The start-up time is
defined by the SUT fuses as described in “Clock Systems and their Distribution” on
page 33. If the level is sampled twice by the Watchdog Oscillator clock but disappears
before the end of the start-up time, the MCU will still wake up, but no interrupt will be
generated. The required level must be held long enough for the MCU to complete the
wake up to trigger the level interrupt.
External Interrupt Control

Register A – EICRA Bit 7 6 5 4 3 2 1 0
ISC31 ISC30 ISC21 ISC20 ISC11 ISC10 ISC01 ISC00 EICRA
This Register can not be reached in ATmega103 compatibility mode, but the initial value
defines INT3:0 as low level interrupts, as in ATmega103.
• Bits 7..0 – ISC31, ISC30 – ISC00, ISC00: External Interrupt 3 - 0 Sense Control
Bits
The External Interrupts 3 - 0 are activated by the external pins INT3:0 if the SREG I-flag
and the corresponding interrupt mask in the EIMSK is set. The level and edges on the
external pins that activate the interrupts are defined in Table 48. Edges on INT3..INT0
are registered asynchronously. Pulses on INT3:0 pins wider than the minimum pulse
width given in Table 49 will generate an interrupt. Shorter pulses are not guaranteed to
generate an interrupt. If low level interrupt is selected, the low level must be held until
the completion of the currently executing instruction to generate an interrupt. If enabled,
a level triggered interrupt will generate an interrupt request as long as the pin is held
low. When changing the ISCn bit, an interrupt can occur. Therefore, it is recommended
to first disable INTn by clearing its Interrupt Enable bit in the EIMSK Register. Then, the
ISCn bit can be changed. Finally, the INTn interrupt flag should be cleared by writing a
logical one to its Interrupt Flag bit (INTFn) in the EIFR Register before the interrupt is re-
enabled.
84 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 48. Interrupt Sense Control(1)

ISCn1 ISCn0 Description
0 0 The low level of INTn generates an interrupt request.
0 1 Reserved
1 0 The falling edge of INTn generates asynchronously an interrupt request.
1 1 The rising edge of INTn generates asynchronously an interrupt request.
Note: 1. n = 3, 2, 1or 0.
When changing the ISCn1/ISCn0 bits, the interrupt must be disabled by clearing its
Interrupt Enable bit in the EIMSK Register. Otherwise an interrupt can occur when
the bits are changed.
Table 49. Asynchronous External Interrupt Characteristics

Minimum pulse width for
tINT 50 ns
asynchronous external interrupt
External Interrupt Control

Register B – EICRB Bit 7 6 5 4 3 2 1 0
ISC71 ISC70 ISC61 ISC60 ISC51 ISC50 ISC41 ISC40 EICRB
• Bits 7..0 – ISC71, ISC70 - ISC41, ISC40: External Interrupt 7 - 4 Sense Control
Bits
The External Interrupts 7 - 4 are activated by the external pins INT7:4 if the SREG I-flag
and the corresponding interrupt mask in the EIMSK is set. The level and edges on the
external pins that activate the interrupts are defined in Table 50. The value on the
INT7:4 pins are sampled before detecting edges. If edge or toggle interrupt is selected,
pulses that last longer than one clock period will generate an interrupt. Shorter pulses
are not guaranteed to generate an interrupt. Observe that CPU clock frequency can be
lower than the XTAL frequency if the XTAL divider is enabled. If low level interrupt is
selected, the low level must be held until the completion of the currently executing
instruction to generate an interrupt. If enabled, a level triggered interrupt will generate an
interrupt request as long as the pin is held low.
Table 50. Interrupt Sense Control(1)

ISCn1 ISCn0 Description
0 0 The low level of INTn generates an interrupt request.
0 1 Any logical change on INTn generates an interrupt request
The falling edge between two samples of INTn generates an interrupt
1 0
request.
The rising edge between two samples of INTn generates an interrupt
1 1
request.
Note: 1. n = 7, 6, 5 or 4.
When changing the ISCn1/ISCn0 bits, the interrupt must be disabled by clearing its
Interrupt Enable bit in the EIMSK Register. Otherwise an interrupt can occur when
the bits are changed.
85
2467E–AVR–05/02
External Interrupt Mask
Register – EIMSK Bit 7 6 5 4 3 2 1 0
INT7 INT6 INT5 INT4 INT3 INT2 INT1 IINT0 EIMSK
• Bits 7..0 – INT7 – INT0: External Interrupt Request 7 - 0 Enable

When an INT7 – INT0 bit is written to one and the I-bit in the Status Register (SREG) is
set (one), the corresponding external pin interrupt is enabled. The Interrupt Sense Con-
trol bits in the External Interrupt Control Registers – EICRA and EICRB – defines
whether the external interrupt is activated on rising or falling edge or level sensed. Activ-
ity on any of these pins will trigger an interrupt request even if the pin is enabled as an
output. This provides a way of generating a software interrupt.
External Interrupt Flag

Register – EIFR Bit 7 6 5 4 3 2 1 0
INTF7 INTF6 INTF5 INTF4 INTF3 INTF2 INTF1 IINTF0 EIFR
• Bits 7..0 – INTF7 - INTF0: External Interrupt Flags 7 - 0

When an edge or logic change on the INT7:0 pin triggers an interrupt request, INTF7:0
becomes set (one). If the I-bit in SREG and the corresponding interrupt enable bit,
INT7:0 in EIMSK, are set (one), the MCU will jump to the interrupt vector. The flag is
cleared when the interrupt routine is executed. Alternatively, the flag can be cleared by
writing a logical one to it. These flags are always cleared when INT7:0 are configured as
level interrupt. Note that when entering sleep mode with the INT3:0 interrupts disabled,
the input buffers on these pins will be disabled. This may cause a logic change in inter-
nal signals which will set the INTF3:0 flags. See “Digital Input Enable and Sleep Modes”
on page 64 for more information.
86 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
8-bit Timer/Counter0 Timer/Counter0 is a general purpose, single channel, 8-bit Timer/Counter module. The
main features are:
with PWM and • Single Channel Counter
Asynchronous • Clear Timer on Compare Match (Auto Reload)
• Glitch-free, Phase Correct Pulse Width Modulator (PWM)
Operation • Frequency Generator
• 10-bit Clock Prescaler
• Overflow and Compare Match Interrupt Sources (TOV0 and OCF0)
• Allows Clocking from External 32 kHz Watch Crystal Independent of the I/O Clock
Overview A simplified block diagram of the 8-bit Timer/Counter is shown in Figure 33. For the
actual placement of I/O pins, refer to “Pin Configurations” on page 2. CPU accessible
I/O registers, including I/O bits and I/O pins, are shown in bold. The device-specific I/O
register and bit locations are listed in the “8-bit Timer/Counter Register Description” on
page 98.
Figure 33. 8-bit Timer/Counter Block Diagram
TCCRn
count TOVn
clear (Int.Req.)
Control Logic
direction clkTn
TOSC1
BOTTOM TOP T/C
Prescaler Oscillator
TOSC2
Timer/Counter
TCNTn
=0 = 0xFF
OCn clk I/O
(Int.Req.)
Waveform
= Generation
OCn
OCRn
DATABUS
clk I/O
Synchronized Status flags
Synchronization Unit
clk ASY
Status flags
ASSRn
asynchronous mode
select (ASn)
Registers The Timer/Counter (TCNT0) and Output Compare Register (OCR0) are 8-bit registers.
Interrupt request (shorten as Int.Req.) signals are all visible in the Timer Interrupt Flag
Register (TIFR). All interrupts are individually masked with the Timer Interrupt Mask
Register (TIMSK). TIFR and TIMSK are not shown in the figure since these registers are
shared by other timer units.
The Timer/Counter can be clocked internally, via the prescaler, or asynchronously
clocked from the TOSC1/2 pins, as detailed later in this section. The asynchronous
87
2467E–AVR–05/02
operation is controlled by the Asynchronous Status Register (ASSR). The Clock Select
logic block controls which clock source the Timer/Counter uses to increment (or decre-
ment) its value. The Timer/Counter is inactive when no clock source is selected. The
output from the clock select logic is referred to as the timer clock (clkT0).
The double buffered Output Compare Register (OCR0) is compared with the
Timer/Counter value at all times. The result of the compare can be used by the wave-
form generator to generate a PWM or variable frequency output on the Output Compare
Pin (OC0). See “Output Compare Unit” on page 89. for details. The compare match
event will also set the compare flag (OCF0) which can be used to generate an output
compare interrupt request.
Definitions Many register and bit references in this document are written in general form. A lower
case “n” replaces the Timer/Counter number, in this case 0. However, when using the
register or bit defines in a program, the precise form must be used (i.e., TCNT0 for
accessing Timer/Counter0 counter value and so on).
The definitions in Table 51 are also used extensively throughout the document.
Table 51. Definitions

BOTTOM The counter reaches the BOTTOM when it becomes zero (0x00).
MAX The counter reaches its MAXimum when it becomes 0xFF (decimal 255).
TOP The counter reaches the TOP when it becomes equal to the highest
value in the count sequence. The TOP value can be assigned to be the
fixed value 0xFF (MAX) or the value stored in the OCR0 Register. The
assignment is dependent on the mode of operation.
Timer/Counter Clock The Timer/Counter can be clocked by an internal synchronous or an external asynchro-
Sources nous clock source. The clock source clkT0 is by default equal to the MCU clock, clkI/O .
When the AS0 bit in the ASSR Register is written to logic one, the clock source is taken
from the Timer/Counter Oscillator connected to TOSC1 and TOSC2. For details on
asynchronous operation, see “Asynchronous Status Register – ASSR” on page 101. For
details on clock sources and prescaler, see “Timer/Counter Prescaler” on page 104.
Counter Unit The main part of the 8-bit Timer/Counter is the programmable bi-directional counter unit.
Figure 34 shows a block diagram of the counter and its surrounding environment.
Figure 34. Counter Unit Block Diagram

DATA BUS
TOVn
(Int.Req.)
TOSC1
count
T/C
clear clk Tn
TCNTn Control Logic Prescaler Oscillator
direction
TOSC2
bottom top clkI/O
88 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Signal description (internal signals):

count Increment or decrement TCNT0 by 1.
direction Selects between increment and decrement.
clear Clear TCNT0 (set all bits to zero).
clkT0 Timer/Counter clock.
top Signalizes that TCNT0 has reached maximum value.
bottom Signalizes that TCNT0 has reached minimum value (zero).
Depending on the mode of operation used, the counter is cleared, incremented, or dec-
remented at each timer clock (clkT0). clkT0 can be generated from an external or internal
clock source, selected by the clock select bits (CS02:0). When no clock source is
selected (CS02:0 = 0) the timer is stopped. However, the TCNT0 value can be accessed
by the CPU, regardless of whether clkT0 is present or not. A CPU write overrides (has
priority over) all counter clear or count operations.
The counting sequence is determined by the setting of the WGM01 and WGM00 bits
located in the Timer/Counter Control Register (TCCR0). There are close connections
between how the counter behaves (counts) and how waveforms are generated on the
output compare output OC0. For more details about advanced counting sequences and
waveform generation, see “Modes of Operation” on page 92.
The Timer/Counter overflow (TOV0) flag is set according to the mode of operation
selected by the WGM01:0 bits. TOV0 can be used for generating a CPU interrupt.
Output Compare Unit The 8-bit comparator continuously compares TCNT0 with the Output Compare Register
(OCR0). Whenever TCNT0 equals OCR0, the comparator signals a match. A match will
set the output compare flag (OCF0) at the next timer clock cycle. If enabled (OCIE0 = 1),
the output compare flag generates an output compare interrupt. The OCF0 flag is auto-
matically cleared when the interrupt is executed. Alternatively, the OCF0 flag can be
cleared by software by writing a logical one to its I/O bit location. The waveform genera-
tor uses the match signal to generate an output according to operating mode set by the
WGM01:0 bits and compare output mode (COM01:0) bits. The max and bottom signals
are used by the waveform generator for handling the special cases of the extreme val-
ues in some modes of operation (“Modes of Operation” on page 92). Figure 35 shows a
block diagram of the output compare unit.
89
2467E–AVR–05/02
Figure 35. Output Compare Unit, Block Diagram
DATA BUS
OCRn TCNTn
= (8-bit Comparator )
OCFn (Int.Req.)
top
bottom
Waveform Generator OCxy
FOCn
WGMn1:0 COMn1:0
The OCR0 Register is double buffered when using any of the Pulse Width Modulation
(PWM) modes. For the normal and Clear Timer on Compare (CTC) modes of operation,
the double buffering is disabled. The double buffering synchronizes the update of the
OCR0 Compare Register to either top or bottom of the counting sequence. The synchro-
nization prevents the occurrence of odd-length, non-symmetrical PWM pulses, thereby
making the output glitch-free.
The OCR0 Register access may seem complex, but this is not case. When the double
buffering is enabled, the CPU has access to the OCR0 buffer Register, and if double
buffering is disabled the CPU will access the OCR0 directly.
Force Output Compare In non-PWM waveform generation modes, the match output of the comparator can be
forced by writing a one to the Force Output Compare (FOC0) bit. Forcing compare
match will not set the OCF0 flag or reload/clear the timer, but the OC0 pin will be
updated as if a real compare match had occurred (the COM01:0 bits settings define
whether the OC0 pin is set, cleared or toggled).
Compare Match Blocking by All CPU write operations to the TCNT0 Register will block any compare match that
TCNT0 Write occurs in the next timer clock cycle, even when the timer is stopped. This feature allows
OCR0 to be initialized to the same value as TCNT0 without triggering an interrupt when
the Timer/Counter clock is enabled.
Using the Output Compare Since writing TCNT0 in any mode of operation will block all compare matches for one
Unit timer clock cycle, there are risks involved when changing TCNT0 when using the output
compare channel, independently of whether the Timer/Counter is running or not. If the
value written to TCNT0 equals the OCR0 value, the compare match will be missed,
resulting in incorrect waveform generation. Similarly, do not write the TCNT0 value
equal to BOTTOM when the counter is downcounting.
90 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
The setup of the OC0 should be performed before setting the Data Direction Register for
the port pin to output. The easiest way of setting the OC0 value is to use the force output
compare (FOC0) strobe bit in normal mode. The OC0 Register keeps its value even
when changing between waveform generation modes.
Be aware that the COM01:0 bits are not double buffered together with the compare
value. Changing the COM01:0 bits will take effect immediately.
Compare Match Output The Compare Output mode (COM01:0) bits have two functions. The waveform genera-
Unit tor uses the COM01:0 bits for defining the Output Compare (OC0) state at the next
compare match. Also, the COM01:0 bits control the OC0 pin output source. Figure 36
shows a simplified schematic of the logic affected by the COM01:0 bit setting. The I/O
registers, I/O bits, and I/O pins in the figure are shown in bold. Only the parts of the Gen-
eral I/O Port Control Registers (DDR and PORT) that are affected by the COM01:0 bits
are shown. When referring to the OC0 state, the reference is for the internal OC0 Regis-
ter, not the OC0 pin.
Figure 36. Compare Match Output Unit, Schematic
COMn1
COMn0 Waveform
D Q
FOCn Generator
1
OCn
OCn Pin
0
D Q
DATA BUS
PORT
D Q
DDR
clk I/O
The general I/O port function is overridden by the output compare (OC0) from the wave-
form generator if either of the COM01:0 bits are set. However, the OC0 pin direction
(input or output) is still controlled by the Data Direction Register (DDR) for the port pin.
The Data Direction Register bit for the OC0 pin (DDR_OC0) must be set as output
before the OC0 value is visible on the pin. The port override function is independent of
the waveform generation mode.
The design of the output compare pin logic allows initialization of the OC0 state before
the output is enabled. Note that some COM01:0 bit settings are reserved for certain
modes of operation. See “8-bit Timer/Counter Register Description” on page 98.
Compare Output Mode and The waveform generator uses the COM01:0 bits differently in normal, CTC, and PWM
Waveform Generation modes. For all modes, setting the COM01:0 = 0 tells the Waveform Generator that no
action on the OC0 Register is to be performed on the next compare match. For compare
output actions in the non-PWM modes refer to Table 53 on page 99. For fast PWM
mode, refer to Table 54 on page 99, and for phase correct PWM refer to Table 55 on
page 100.
91
2467E–AVR–05/02
A change of the COM01:0 bits state will have effect at the first compare match after the
bits are written. For non-PWM modes, the action can be forced to have immediate effect
by using the FOC0 strobe bits.
Modes of Operation The mode of operation, i.e., the behavior of the Timer/Counter and the output compare
pins, is defined by the combination of the Waveform Generation mode (WGM01:0) and
Compare Output mode (COM01:0) bits. The Compare Output mode bits do not affect
the counting sequence, while the Waveform Generation mode bits do. The COM01:0
bits control whether the PWM output generated should be inverted or not (inverted or
non-inverted PWM). For non-PWM modes the COM01:0 bits control whether the output
should be set, cleared, or toggled at a compare match (See “Compare Match Output
Unit” on page 91.).
For detailed timing information refer to “Timer/Counter Timing Diagrams” on page 96.
Normal Mode The simplest mode of operation is the normal mode (WGM01:0 = 0). In this mode the
counting direction is always up (incrementing), and no counter clear is performed. The
counter simply overruns when it passes its maximum 8-bit value (TOP = 0xFF) and then
restarts from the bottom (0x00). In normal operation the Timer/Counter overflow flag
(TOV0) will be set in the same timer clock cycle as the TCNT0 becomes zero. The TOV0
flag in this case behaves like a ninth bit, except that it is only set, not cleared. However,
combined with the timer overflow interrupt that automatically clears the TOV0 flag, the
timer resolution can be increased by software. There are no special cases to consider in
the normal mode, a new counter value can be written anytime.
The output compare unit can be used to generate interrupts at some given time. Using
the output compare to generate waveforms in normal mode is not recommended, since
this will occupy too much of the CPU time.
Clear Timer on Compare In Clear Timer on Compare or CTC mode (WGM01:0 = 2), the OCR0 Register is used to
Match (CTC) Mode manipulate the counter resolution. In CTC mode the counter is cleared to zero when the
counter value (TCNT0) matches the OCR0. The OCR0 defines the top value for the
counter, hence also its resolution. This mode allows greater control of the compare
match output frequency. It also simplifies the operation of counting external events.
The timing diagram for the CTC mode is shown in Figure 37. The counter value
(TCNT0) increases until a compare match occurs between TCNT0 and OCR0, and then
counter (TCNT0) is cleared.
Figure 37. CTC Mode, Timing Diagram
OCn Interrupt Flag Set
TCNTn
OCn
(COMn1:0 = 1)
(Toggle)
Period 1 2 3 4
92 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
An interrupt can be generated each time the counter value reaches the TOP value by
using the OCF0 flag. If the interrupt is enabled, the interrupt handler routine can be used
for updating the TOP value. However, changing the TOP to a value close to BOTTOM
when the counter is running with none or a low prescaler value must be done with care
since the CTC mode does not have the double buffering feature. If the new value written
to OCR0 is lower than the current value of TCNT0, the counter will miss the compare
match. The counter will then have to count to its maximum value (0xFF) and wrap
around starting at 0x00 before the compare match can occur.
For generating a waveform output in CTC mode, the OC0 output can be set to toggle its
logical level on each compare match by setting the Compare Output mode bits to Tog-
gle mode (COM01:0 = 1). The OC0 value will not be visible on the port pin unless the
data direction for the pin is set to output. The waveform generated will have a maximum
frequency of fOC0 = fclk_I/O/2 when OCR0 is set to zero (0x00). The waveform frequency
is defined by the following equation:
f clk_I/O
f OCn = ----------------------------------------------
-
2 ⋅ N ⋅ ( 1 + OCRn )
The N variable represents the prescale factor (1, 8, 32, 64, 128, 256, or 1024).
As for the normal mode of operation, the TOV0 flag is set in the same timer clock cycle
that the counter counts from MAX to 0x00.
Fast PWM Mode The fast Pulse Width Modulation or fast PWM mode (WGM01:0 = 1) provides a high fre-
quency PWM waveform generation option. The fast PWM differs from the other PWM
option by its single-slope operation. The counter counts from BOTTOM to MAX then
restarts from BOTTOM. In non-inverting Compare Output mode, the output compare
(OC0) is cleared on the compare match between TCNT0 and OCR0, and set at BOT-
TOM. In inverting Compare Output mode, the output is set on compare match and
cleared at BOTTOM. Due to the single-slope operation, the operating frequency of the
fast PWM mode can be twice as high as the phase correct PWM mode that uses dual-
slope operation. This high frequency makes the fast PWM mode well suited for power
regulation, rectification, and DAC applications. High frequency allows physically small
sized external components (coils, capacitors), and therefore reduces total system cost.
In fast PWM mode, the counter is incremented until the counter value matches the MAX
value. The counter is then cleared at the following timer clock cycle. The timing diagram
for the fast PWM mode is shown in Figure 38. The TCNT0 value is in the timing diagram
shown as a histogram for illustrating the single-slope operation. The diagram includes
non-inverted and inverted PWM outputs. The small horizontal line marks on the TCNT0
slopes represent compare matches between OCR0 and TCNT0.
93
2467E–AVR–05/02
Figure 38. Fast PWM Mode, Timing Diagram
OCRn Interrupt Flag Set
OCRn Update
and
TOVn Interrupt Flag Set
TCNTn
OCn (COMn1:0 = 2)
OCn (COMn1:0 = 3)
Period 1 2 3 4 5 6 7
The Timer/Counter overflow flag (TOV0) is set each time the counter reaches Max If the
interrupt is enabled, the interrupt handler routine can be used for updating the compare
value.
In fast PWM mode, the compare unit allows generation of PWM waveforms on the OC0
pin. Setting the COM01:0 bits to 2 will produce a non-inverted PWM and an inverted
PWM output can be generated by setting the COM01:0 to 3 (See Table 54 on page 99).
The actual OC0 value will only be visible on the port pin if the data direction for the port
pin is set as output. The PWM waveform is generated by setting (or clearing) the OC0
Register at the compare match between OCR0 and TCNT0, and clearing (or setting) the
OC0 Register at the timer clock cycle the counter is cleared (changes from MAX to
BOTTOM).
The PWM frequency for the output can be calculated by the following equation:
f clk_I/O
f OCnPWM = -----------------
-
N ⋅ 256
The extreme values for the OCR0 Register represent special cases when generating a
PWM waveform output in the fast PWM mode. If the OCR0 is set equal to BOTTOM, the
output will be a narrow spike for each MAX+1 timer clock cycle. Setting the OCR0 equal
to MAX will result in a constantly high or low output (depending on the polarity of the out-
put set by the COM01:0 bits.)
A frequency (with 50% duty cycle) waveform output in fast PWM mode can be achieved
by setting OC0 to toggle its logical level on each compare match (COM01:0 = 1). The
waveform generated will have a maximum frequency of foc0 = fclk_I/O/2 when OCR0 is set
to zero. This feature is similar to the OC0 toggle in CTC mode, except the double buffer
feature of the output compare unit is enabled in the fast PWM mode.
94 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Phase Correct PWM Mode The phase correct PWM mode (WGM01:0 = 3) provides a high resolution phase correct
PWM waveform generation option. The phase correct PWM mode is based on a dual-
slope operation. The counter counts repeatedly from BOTTOM to MAX and then from
MAX to BOTTOM. In non-inverting Compare Output mode, the output compare (OC0) is
cleared on the compare match between TCNT0 and OCR0 while upcounting, and set on
the compare match while downcounting. In inverting Output Compare mode, the opera-
tion is inverted. The dual-slope operation has lower maximum operation frequency than
single slope operation. However, due to the symmetric feature of the dual-slope PWM
modes, these modes are preferred for motor control applications.
The PWM resolution for the phase correct PWM mode is fixed to 8 bits. In phase correct
PWM mode the counter is incremented until the counter value matches Max When the
counter reaches MAX, it changes the count direction. The TCNT0 value will be equal to
MAX for one timer clock cycle. The timing diagram for the phase correct PWM mode is
shown on Figure 39. The TCNT0 value is in the timing diagram shown as a histogram
for illustrating the dual-slope operation. The diagram includes non-inverted and inverted
PWM outputs. The small horizontal line marks on the TCNT0 slopes represent compare
matches between OCR0 and TCNT0.
Figure 39. Phase Correct PWM Mode, Timing Diagram
OCRn Update
TCNTn
OCn (COMn1:0 = 2)
OCn (COMn1:0 = 3)
Period 1 2 3
The Timer/Counter Overflow Flag (TOV0 ) is set each time the counter reaches BOT-
TOM. The interrupt flag can be used to generate an interrupt each time the counter
reaches the BOTTOM value.
In phase correct PWM mode, the compare unit allows generation of PWM waveforms on
the OC0 pin. Setting the COM01:0 bits to 2 will produce a non-inverted PWM. An
inverted PWM output can be generated by setting the COM01:0 to 3 (See Table 55 on
page 100). The actual OC0 value will only be visible on the port pin if the data direction
for the port pin is set as output. The PWM waveform is generated by clearing (or setting)
the OC0 Register at the compare match between OCR0 and TCNT0 when the counter
increments, and setting (or clearing) the OC0 Register at compare match between
95
2467E–AVR–05/02
OCR0 and TCNT0 when the counter decrements. The PWM frequency for the output
when using phase correct PWM can be calculated by the following equation:
f clk_I/O
f OCn PCPW M = -----------------
-
N ⋅ 510
PWM waveform output in the phase correct PWM mode. If the OCR0 is set equal to
BOTTOM, the output will be continuously low and if set equal to MAX the output will be
continuously high for non-inverted PWM mode. For inverted PWM the output will have
the opposite logic values.
Timer/Counter Timing Figure 40 and Figure 41 contain timing data for the Timer/Counter operation. The
Diagrams Timer/Counter is a synchronous design and the timer clock (clkT0) is therefore shown as
a clock enable signal. The figure shows the count sequence close to the MAX value.
Figure 42 and Figure 43 show the same timing data, but with the prescaler enabled. The
figures illustrate when interrupt flags are set.
The following figures show the Timer/Counter in Synchronous mode, and the timer clock
(clkT0) is therefore shown as a clock enable signal. In asynchronous mode, clkI/O should
be replaced by the Timer/Counter Oscillator clock. The figures include information on
when interrupt flags are set. Figure 40 contains timing data for basic Timer/Counter
operation. The figure shows the count sequence close to the MAX value in all modes
other than phase correct PWM mode.
Figure 40. Timer/Counter Timing Diagram, No Prescaling
clkI/O
clkTn
(clkI/O /1)
TCNTn MAX - 1 MAX BOTTOM BOTTOM + 1
TOVn
Figure 41 shows the same timing data, but with the prescaler enabled.
96 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 41. Timer/Counter Timing Diagram, with Prescaler (fclk_I/O/8)
clkI/O
clkTn
(clkI/O /8)
TOVn
Figure 42 shows the setting of OCF0 in all modes except CTC mode.
Figure 42. Timer/Counter Timing Diagram, Setting of OCF0, with Prescaler (fclk_I/O/8)
clkI/O
clkTn
(clkI/O /8)
TCNTn OCRn - 1 OCRn OCRn + 1 OCRn + 2
OCRn OCRn Value
OCFn
Figure 43 shows the setting of OCF0 and the clearing of TCNT0 in CTC mode.
97
2467E–AVR–05/02
Figure 43. Timer/Counter Timing Diagram, Clear Timer on Compare Match Mode, with
Prescaler (fclk_I/O/8)
clkI/O
clkTn
(clkI/O /8)
TCNTn
TOP - 1 TOP BOTTOM BOTTOM + 1
(CTC)
OCRn TOP
OCFn
8-bit Timer/Counter
Register Description
Timer/Counter Control
Register – TCCR0 Bit 7 6 5 4 3 2 1 0
FOC0 WGM00 COM01 COM00 WGM01 CS02 CS01 CS00 TCCR0
Read/Write W R/W R/W R/W R/W R/W R/W R/W
• Bit 7 – FOC0: Force Output Compare

The FOC0 bit is only active when the WGM bits specify a non-PWM mode. However, for
ensuring compatibility with future devices, this bit must be set to zero when TCCR0 is
written when operating in PWM mode. When writing a logical one to the FOC0 bit, an
immediate compare match is forced on the waveform generation unit. The OC0 output is
changed according to its COM01:0 bits setting. Note that the FOC0 bit is implemented
as a strobe. Therefore it is the value present in the COM01:0 bits that determines the
effect of the forced compare.
A FOC0 strobe will not generate any interrupt, nor will it clear the timer in CTC mode
using OCR0 as TOP.
The FOC0 bit is always read as zero.
• Bit 6, 3 – WGM01:0: Waveform Generation Mode

These bits control the counting sequence of the counter, the source for the maximum
(TOP) counter value, and what type of waveform generation to be used. Modes of oper-
ation supported by the Timer/Counter unit are: Normal mode, Clear Timer on Compare
match (CTC) mode, and two types of Pulse Width Modulation (PWM) modes. See Table
52 and “Modes of Operation” on page 92.
98 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 52. Waveform Generation Mode Bit Description

WGM01(1) WGM00(1) Timer/Counter Update of TOV0 Flag
Mode (CTC0) (PWM0) Mode of Operation TOP OCR0 at Set on
0 0 0 Normal 0xFF Immediate MAX
1 0 1 PWM, Phase 0xFF TOP BOTTOM
Correct
2 1 0 CTC OCR0 Immediate MAX
3 1 1 Fast PWM 0xFF TOP MAX
Note: 1. The CTC0 and PWM0 bit definition names are now obsolete. Use the WGM01:0 def-
initions. However, the functionality and location of these bits are compatible with
previous versions of the timer.
• Bit 5:4 – COM01:0: Compare Match Output Mode

These bits control the output compare pin (OC0) behavior. If one or both of the
COM01:0 bits are set, the OC0 output overrides the normal port functionality of the I/O
pin it is connected to. However, note that the Data Direction Register (DDR) bit corre-
sponding to OC0 pin must be set in order to enable the output driver.
When OC0 is connected to the pin, the function of the COM01:0 bits depends on the
WGM01:0 bit setting. Table 53 shows the COM01:0 bit functionality when the WGM01:0
bits are set to a normal or CTC mode (non-PWM).
Table 53. Compare Output Mode, non-PWM Mode
COM01 COM00 Description
0 0 Normal port operation, OC0 disconnected.
0 1 Toggle OC0 on compare match
1 0 Clear OC0 on compare match
1 1 Set OC0 on compare match
Table 54 shows the COM01:0 bit functionality when the WGM01:0 bits are set to fast
PWM mode.
Table 54. Compare Output Mode, Fast PWM Mode(1)

0 1 Reserved
1 0 Clear OC0 on compare match, set OC0 at TOP
1 1 Set OC0 on compare match, clear OC0 at TOP
Note: 1. A special case occurs when OCR0 equals TOP and COM01 is set. In this case, the
compare match is ignored, but the set or clear is done at TOP. See “Fast PWM Mode”
on page 93 for more details.
Table 55 shows the COM01:0 bit functionality when the WGM01:0 bits are set to phase
correct PWM mode.
99
2467E–AVR–05/02
Table 55. Compare Output Mode, Phase Correct PWM Mode(1)
0 1 Reserved
1 0 Clear OC0 on compare match when up-counting. Set OC0 on compare
match when downcounting.
1 1 Set OC0 on compare match when up-counting. Clear OC0 on compare
compare match is ignored, but the set or clear is done at TOP. See “Phase Correct
PWM Mode” on page 95 for more details.
• Bit 2:0 – CS02:0: Clock Select

The three clock select bits select the clock source to be used by the Timer/Counter, see
Table 56.
Table 56. Clock Select Bit Description
CS02 CS01 CS00 Description
0 0 0 No clock source (Timer/Counter stopped)
0 0 1 clkT0S/(No prescaling)
0 1 0 clkT0S/8 (From prescaler)
Timer/Counter Register –
TCNT0 Bit 7 6 5 4 3 2 1 0
TCNT0[7:0] TCNT0
The Timer/Counter Register gives direct access, both for read and write operations, to
the Timer/Counter unit 8-bit counter. Writing to the TCNT0 Register blocks (removes)
the compare match on the following timer clock. Modifying the counter (TCNT0) while
the counter is running, introduces a risk of missing a compare match between TCNT0
and the OCR0 Register.
Output Compare Register –

OCR0 Bit 7 6 5 4 3 2 1 0
OCR0[7:0] OCR0
The Output Compare Register contains an 8-bit value that is continuously compared
with the counter value (TCNT0). A match can be used to generate an output compare
interrupt, or to generate a waveform output on the OC0 pin.
100 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Asynchronous Operation
of the Timer/Counter
Asynchronous Status
Register – ASSR Bit 7 6 5 4 3 2 1 0
– – – – AS0 TCN0UB OCR0UB TCR0UB ASSR
Read/Write R R R R R/W R R R
• Bit 3 – AS0: Asynchronous Timer/Counter0

When AS0 is written to zero, Timer/Counter0 is clocked from the I/O clock, clkI/O. When
AS0 is written to one, Timer/Counter is clocked from a crystal Oscillator connected to
the Timer Oscillator 1 (TOSC1) pin. When the value of AS0 is changed, the contents of
TCNT0, OCR0, and TCCR0 might be corrupted.
• Bit 2 – TCN0UB: Timer/Counter0 Update Busy

When Timer/Counter0 operates asynchronously and TCNT0 is written, this bit becomes
set. When TCNT0 has been updated from the temporary storage register, this bit is
cleared by hardware. A logical zero in this bit indicates that TCNT0 is ready to be
updated with a new value.
• Bit 1 – OCR0UB: Output Compare Register0 Update Busy

When Timer/Counter0 operates asynchronously and OCR0 is written, this bit becomes
set. When OCR0 has been updated from the temporary storage register, this bit is
cleared by hardware. A logical zero in this bit indicates that OCR0 is ready to be
• Bit 0 – TCR0UB: Timer/Counter Control Register0 Update Busy

When Timer/Counter0 operates asynchronously and TCCR0 is written, this bit becomes
set. When TCCR0 has been updated from the temporary storage register, this bit is
cleared by hardware. A logical zero in this bit indicates that TCCR0 is ready to be
If a write is performed to any of the three Timer/Counter0 Registers while its update
busy flag is set, the updated value might get corrupted and cause an unintentional inter-
rupt to occur.
The mechanisms for reading TCNT0, OCR0, and TCCR0 are different. When reading
TCNT0, the actual timer value is read. When reading OCR0 or TCCR0, the value in the
temporary storage register is read.
Asynchronous Operation of When Timer/Counter0 operates asynchronously, some considerations must be taken.
Timer/Counter0 • Warning: When switching between asynchronous and synchronous clocking of
Timer/Counter0, the Timer Registers TCNT0, OCR0, and TCCR0 might be
corrupted. A safe procedure for switching clock source is:
1. Disable the Timer/Counter0 interrupts by clearing OCIE0 and TOIE0.
2. Select clock source by setting AS0 as appropriate.
3. Write new values to TCNT0, OCR0, and TCCR0.
4. To switch to asynchronous operation: Wait for TCN0UB, OCR0UB, and
TCR0UB.
5. Clear the Timer/Counter0 interrupt flags.
101
2467E–AVR–05/02
6. Enable interrupts, if needed.
• The Oscillator is optimized for use with a 32.768 kHz watch crystal. Applying an
external clock to the TOSC1 pin may result in incorrect Timer/Counter0 operation.
The CPU main clock frequency must be more than four times the Oscillator
frequency.
• When writing to one of the registers TCNT0, OCR0, or TCCR0, the value is
transferred to a temporary register, and latched after two positive edges on TOSC1.
The user should not write a new value before the contents of the Temporary
Register have been transferred to its destination. Each of the three mentioned
registers have their individual temporary register, which means that e.g., writing to
TCNT0 does not disturb an OCR0 write in progress. To detect that a transfer to the
destination register has taken place, the Asynchronous Status Register – ASSR has
been implemented.
• When entering Power-save or Extended Standby mode after having written to
TCNT0, OCR0, or TCCR0, the user must wait until the written register has been
updated if Timer/Counter0 is used to wake up the device. Otherwise, the MCU will
enter sleep mode before the changes are effective. This is particularly important if
the Output Compare0 interrupt is used to wake up the device, since the output
compare function is disabled during writing to OCR0 or TCNT0. If the write cycle is
not finished, and the MCU enters sleep mode before the OCR0UB bit returns to
zero, the device will never receive a compare match interrupt, and the MCU will not
wake up.
• If Timer/Counter0 is used to wake the device up from Power-save or Extended
Standby mode, precautions must be taken if the user wants to re-enter one of these
modes: The interrupt logic needs one TOSC1 cycle to be reset. If the time between
wake-up and re-entering sleep mode is less than one TOSC1 cycle, the interrupt will
not occur, and the device will fail to wake up. If the user is in doubt whether the time
before re-entering Power-save or Extended Standby mode is sufficient, the
following algorithm can be used to ensure that one TOSC1 cycle has elapsed:
1. Write a value to TCCR0, TCNT0, or OCR0.
2. Wait until the corresponding Update Busy flag in ASSR returns to zero.
3. Enter Power-save or Extended Standby mode.
• When the asynchronous operation is selected, the 32.768 kHZ Oscillator for
Timer/Counter0 is always running, except in Power-down and Standby modes. After
a Power-up Reset or wake-up from Power-down or Standby mode, the user should
be aware of the fact that this Oscillator might take as long as one second to
stabilize. The user is advised to wait for at least one second before using
Timer/Counter0 after power-up or wake-up from Power-down or Standby mode. The
contents of all Timer/Counter0 Registers must be considered lost after a wake-up
from Power-down or Standby mode due to unstable clock signal upon start-up, no
matter whether the Oscillator is in use or a clock signal is applied to the TOSC1 pin.
• Description of wake up from Power-save or Extended Standby mode when the timer
is clocked asynchronously: When the interrupt condition is met, the wake up
process is started on the following cycle of the timer clock, that is, the timer is
always advanced by at least one before the processor can read the counter value.
After wake-up, the MCU is halted for four cycles, it executes the interrupt routine,
and resumes execution from the instruction following SLEEP.
• Reading of the TCNT0 Register shortly after wake-up from Power-save may give an
incorrect result. Since TCNT0 is clocked on the asynchronous TOSC clock, reading
TCNT0 must be done through a register synchronized to the internal I/O clock
domain. Synchronization takes place for every rising TOSC1 edge. When waking up
102 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
from Power-save mode, and the I/O clock (clk I/O) again becomes active, TCNT0 will
read as the previous value (before entering sleep) until the next rising TOSC1 edge.
The phase of the TOSC clock after waking up from Power-save mode is essentially
unpredictable, as it depends on the wake-up time. The recommended procedure for
reading TCNT0 is thus as follows:
1. Write any value to either of the registers OCR0 or TCCR0.
2. Wait for the corresponding Update Busy Flag to be cleared.
3. Read TCNT0.
• During asynchronous operation, the synchronization of the interrupt flags for the
asynchronous timer takes three processor cycles plus one timer cycle. The timer is
therefore advanced by at least one before the processor can read the timer value
causing the setting of the interrupt flag. The output compare pin is changed on the
timer clock and is not synchronized to the processor clock.
Timer/Counter Interrupt Mask

Register – TIMSK Bit 7 6 5 4 3 2 1 0
OCIE2 TOIE2 TICIE1 OCIE1A OCIE1B TOIE1 OCIE0 TOIE0 TIMSK
• Bit 1 – OCIE0: Timer/Counter0 Output Compare Match Interrupt Enable

When the OCIE0 bit is written to one, and the I-bit in the Status Register is set (one), the
Timer/Counter0 Compare Match interrupt is enabled. The corresponding interrupt is
executed if a compare match in Timer/Counter0 occurs, i.e., when the OCF0 bit is set in
the Timer/Counter Interrupt Flag Register – TIFR.
• Bit 0 – TOIE0: Timer/Counter0 Overflow Interrupt Enable

When the TOIE0 bit is written to one, and the I-bit in the Status Register is set (one), the
Timer/Counter0 Overflow interrupt is enabled. The corresponding interrupt is executed if
a n ove rflow in Time r/C oun te r0 o ccurs, i.e., whe n the TO V0 bit is set in th e
Timer/Counter Interrupt Flag Register – TIFR.
Timer/Counter Interrupt Flag

Register – TIFR Bit 7 6 5 4 3 2 1 0
OCF2 TOV2 ICF1 OCF1A OCF1B TOV1 OCF0 TOV0 TIFR
• Bit 1 – OCF0: Output Compare Flag 0

The OCF0 bit is set (one) when a compare match occurs between the Timer/Counter0
and the data in OCR0 – Output Compare Register0. OCF0 is cleared by hardware when
executing the corresponding interrupt handling vector. Alternatively, OCF0 is cleared by
writing a logic one to the flag. When the I-bit in SREG, OCIE0 (Timer/Counter0 Com-
pare Match Interrupt Enable), and OCF0 are set (one), the Timer/Counter0 Compare
Match Interrupt is executed.
• Bit 0 – TOV0: Timer/Counter0 Overflow Flag

The bit TOV0 is set (one) when an overflow occurs in Timer/Counter0. TOV0 is cleared
by hardware when executing the corresponding interrupt handling vector. Alternatively,
TOV0 is cleared by writing a logic one to the flag. When the SREG I-bit, TOIE0
103
2467E–AVR–05/02
( Time r /C o u n te r0 O ve r flo w In te r ru p t En a b le ), a n d TO V 0 a r e se t ( o ne ) , th e
Timer/Counter0 Overflow Interrupt is executed. In PWM mode, this bit is set when
Timer/Counter0 changes counting direction at $00.
Timer/Counter Prescaler Figure 44. Prescaler for Timer/Counter0
clkI/O clkT0S
10-BIT T/C PRESCALER
Clear
TOSC1
clkT0S/8
clkT0S/32
clkT0S/64
clkT0S/128
clkT0S/256
clkT0S/1024
AS0
PSR0 0
CS00
CS01
CS02
TIMER/COUNTER0 CLOCK SOURCE

clkT0
The clock source for Timer/Counter0 is named clkT0. clkT0 is by default connected to the
main system clock clkI/O. By setting the AS0 bit in ASSR, Timer/Counter0 is asynchro-
nously clocked from the TOSC1 pin. This enables use of Timer/Counter0 as a Real
Time Counter (RTC). When AS0 is set, pins TOSC1 and TOSC2 are disconnected from
Port C. A crystal can then be connected between the TOSC1 and TOSC2 pins to serve
as an independent clock source for Timer/Counter0. The Oscillator is optimized for use
with a 32.768 kHz crystal. Applying an external clock source to TOSC1 is not
recommended.
For Timer/Counter0, the possible prescaled selections are: clkT0S/8, clkT0S/32, clkT0S/64,
clkT0S/128, clkT0S/256, and clk T0S/1024. Additionally, clkT0S as well as 0 (stop) may be
selected. Setting the PSR0 bit in SFIOR resets the prescaler. This allows the user to
operate with a predictable prescaler.

SFIOR Bit 7 6 5 4 3 2 1 0
• Bit 7 – TSM: Timer/Counter Synchronization Mode

Writing TSM to one, PSR0 and PSR321 becomes registers that hold their value until
rewritten, or the TSM bit is written zero. This mode is useful for synchronizing
Timer/Counters. By setting both TSM and the appropriate PSR bit(s), the appropriate
Timer/Counters are halted, and can be configured to same value without the risk of one
104 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
of the m advancin g d uring configuration. W hen the TSM bit written zero, the
Timer/Counters start counting simultaneously.
• Bit 1 – PSR0: Prescaler Reset Timer/Counter0

When this bit is written to one, the Timer/Counter0 prescaler will be reset. The bit will be
cleared by hardware after the operation is performed. Writing a zero to this bit will have
no effect. This bit will always be read as zero if Timer/Counter0 is clocked by the internal
CPU clock. If this bit is written when Timer/Counter0 is operating in Asynchronous
mode, the bit will remain one until the prescaler has been reset.
105
2467E–AVR–05/02
16-bit Timer/Counter The 16-bit Timer/Counter unit allows accurate program execution timing (event man-
agement), wave generation, and signal timing measurement. The main features are:
(Timer/Counter1 and • True 16-bit Design (i.e. ,Allows 16-bit PWM)
Timer/Counter3) • Three Independent Output Compare Units
• Double Buffered Output Compare Registers
• One Input Capture Unit
• Input Capture Noise Canceler
• Clear Timer on Compare Match (Auto Reload)
• Glitch-free, Phase Correct Pulse width Modulator (PWM)
• Variable PWM Period
• Frequency Generator
• External Event Counter
• Ten Independent Interrupt Sources (TOV1, OCF1A, OCF1B, OCF1C, ICF1, TOV3, OCF3A,
OCF3B, OCF3C, and ICF3)
Restrictions in ATmega103 Note that in ATmega103 compatibility mode, only one 16-bit Timer/Counter is available
Compatibility Mode (Timer/Counter1). Also note that in ATmega103 compatibility mode, the Timer/Counter1
has two Compare Registers (Compare A and Compare B) only.
Overview Most register and bit references in this document are written in general form. A lower
case “n” replaces the Timer/Counter number, and a lower case “x” replaces the output
compare unit channel. However, when using the register or bit defines in a program, the
precise form must be used, i.e., TCNT1 for accessing Timer/Counter1 counter value
and so on. The physical I/O register and bit locations for ATmega128 are listed in the
“16-bit Timer/Counter Register Description” on page 127.
A simplified block diagram of the 16-bit Timer/Counter is shown in Figure 45. CPU
accessible I/O registers, including I/O bits and I/O pins, are shown in bold.
106 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 45. 16-bit Timer/Counter Block Diagram

Count TOVx
Clear (Int.Req.)
Control Logic
Direction TCLK Clock Select
Edge
Tx
Detector
TOP BOTTOM
( From Prescaler )
Timer/Counter
TCNTx
= =0
OCFxA
(Int.Req.)
Waveform
= Generation
OCxA
OCRxA
Fixed OCFxB
TOP (Int.Req.)
Values
Waveform
= Generation
OCxB
DATABUS
OCRxB
OCFxC
(Int.Req.)
Waveform
= Generation
OCxC
OCRxC ( From Analog

Comparator Ouput )
ICFx (Int.Req.)
Edge Noise
ICRx
Detector Canceler
ICPx
TCCRxA TCCRxB TCCRxC
Note: Refer to Figure 1 on page 2, Table 30 on page 68, and Table 39 on page 75 for
Timer/Counter1 and 3 pin placement and description.
Registers The Timer/Counter (TCNTn), Output Compare Registers (OCRnA/B/C), and Input Cap-
ture Register (ICRn) are all 16-bit registers. Special procedures must be followed when
accessing the 16-bit registers. These procedures are described in the section “Access-
ing 16-bit Registers” on page 109. The Timer/Counter Control Registers (TCCRnA/B/C)
are 8-bit registers and have no CPU access restrictions. Interrupt requests (shorten as
Int.Req.) signals are all visible in the Timer Interrupt Flag Register (TIFR) and Extended
Timer Interrupt Flag Register (ETIFR). All interrupts are individually masked with the
Timer Interrupt Mask Register (TIMSK) and Extended Timer Interrupt Mask Register
(ETIMSK). (E)TIFR and (E)TIMSK are not shown in the figure since these registers are
shared by other timer units.
The Timer/Counter can be clocked internally, via the prescaler, or by an external clock
source on the Tn pin. The Clock Select logic block controls which clock source and edge
the Timer/Counter uses to increment (or decrement) its value. The Timer/Counter is
inactive when no clock source is selected. The output from the clock select logic is
referred to as the timer clock (clkTn).
107
2467E–AVR–05/02
The double buffered Output Compare Registers (OCRnA/B/C) are compared with the
Timer/Counter value at all time. The result of the compare can be used by the waveform
generator to generate a PWM or variable frequency output on the Output Compare Pin
(OCnA/B/C). See “Output Compare Units” on page 115.. The compare match event will
also set the compare match flag (OCFnA/B/C) which can be used to generate an output
The Input Capture Register can capture the Timer/Counter value at a given external
(edge triggered) event on either the Input Capture Pin (ICPn) or on the Analog Compar-
ator pins (See “Analog Comparator” on page 222.) The input capture unit includes a
digital filtering unit (Noise Canceler) for reducing the chance of capturing noise spikes.
The TOP value, or maximum Timer/Counter value, can in some modes of operation be
defined by either the OCRnA Register, the ICRn Register, or by a set of fixed values.
When using OCRnA as TOP value in a PWM mode, the OCRnA Register can not be
used for generating a PWM output. However, the TOP value will in this case be double
buffered allowing the TOP value to be changed in run time. If a fixed TOP value is
required, the ICRn Register can be used as an alternative, freeing the OCRnA to be
used as PWM output.
Definitions The following definitions are used extensively throughout the document:
BOTTOM The counter reaches the BOTTOM when it becomes 0x0000.
MAX The counter reaches its MAX imum when it becomes 0xFFFF (decimal 65535).
TOP The counter reaches the TOP when it becomes equal to the highest value in the
count sequence. The TOP value can be assigned to be one of the fixed values:
0x00FF, 0x01FF, or 0x03FF, or to the value stored in the OCRnA or ICRn
Register. The assignment is dependent of the mode of operation.
Compatibility The 16-bit Timer/Counter has been updated and improved from previous versions of the
16-bit AVR Timer/Counter. This 16-bit Timer/Counter is fully compatible with the earlier
version regarding:
• All 16-bit Timer/Counter related I/O register address locations, including timer
interrupt registers.
• Bit locations inside all 16-bit Timer/Counter Registers, including Timer Interrupt
Registers.
• Interrupt vectors.
The following control bits have changed name, but have same functionality and register
location:
• PWMn0 is changed to WGMn0.
• PWMn1 is changed to WGMn1.
• CTCn is changed to WGMn2.
The following registers are added to the 16-bit Timer/Counter:
• Timer/Counter Control Register C (TCCRnC).
• Output Compare Register C, OCRnCH and OCRnCL, combined OCRnC.
The following bits are added to the 16-bit Timer/Counter Control Registers:
• COM1C1:0 are added to TCCR1A.
• FOCnA, FOCnB, and FOCnC are added in the new TCCRnC Register.
108 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
• WGMn3 is added to TCCRnB.

Interrupt flag and mask bits for output compare unit C are added.
The 16-bit Timer/Counter has improvements that will affect the compatibility in some
special cases.
Accessing 16-bit The TCNTn, OCRnA/B/C, and ICRn are 16-bit registers that can be accessed by the
Registers AVR CPU via the 8-bit data bus. The 16-bit register must be byte accessed using two
read or write operations. Each 16-bit timer has a single 8-bit register for temporary stor-
ing of the high byte of the 16-bit access. The same Temporary Register is shared
between all 16-bit registers within each 16-bit timer. Accessing the low byte triggers the
16-bit read or write operation. When the low byte of a 16-bit register is written by the
CPU, the high byte stored in the Temporary Register, and the low byte written are both
copied into the 16-bit register in the same clock cycle. When the low byte of a 16-bit reg-
ister is read by the CPU, the high byte of the 16-bit register is copied into the Temporary
Register in the same clock cycle as the low byte is read.
Not all 16-bit accesses uses the Temporary Register for the high byte. Reading the
OCRnA/B/C 16-bit registers does not involve using the Temporary Register.
To do a 16-bit write, the high byte must be written before the low byte. For a 16-bit read,
the low byte must be read before the high byte.
The following code examples show how to access the 16-bit timer registers assuming
that no interrupts updates the temporary register. The same principle can be used
directly for accessing the OCRnA/B/C and ICRn Registers. Note that when using “C”,
the compiler handles the 16-bit access.
Assembly Code Examples(1)
...
; Set TCNTn to 0x01FF
ldi r17,0x01
ldi r16,0xFF
out TCNTnH,r17
out TCNTnL,r16
; Read TCNTn into r17:r16
in r16,TCNTnL
in r17,TCNTnH
...
C Code Examples(1)
unsigned int i;
...
/* Set TCNTn to 0x01FF */
TCNTn = 0x1FF;
/* Read TCNTn into i */
i = TCNTn;
...
For I/O registers located in extended I/O map, “IN”, “OUT”, “SBIS”, “SBIC”, “CBI”, and
“SBI” instructions must be replaced with instructions that allow access to extended
I/O. Typically “LDS” and “STS” combined with “SBRS”, “SBRC”, “SBR”, and “CBR”.
The assembly code example returns the TCNTn value in the r17:r16 register pair.
109
2467E–AVR–05/02
It is important to notice that accessing 16-bit registers are atomic operations. If an inter-
rupt occurs between the two instructions accessing the 16-bit register, and the interrupt
code updates the temporary register by accessing the same or any other of the 16-bit
Timer Registers, then the result of the access outside the interrupt will be corrupted.
Therefore, when both the main code and the interrupt code update the temporary regis-
ter, the main code must disable the interrupts during the 16-bit access.
The following code examples show how to do an atomic read of the TCNTn Register
contents. Reading any of the OCRnA/B/C or ICRn Registers can be done by using the
same principle.
TIM16_ReadTCNTn:
; Save global interrupt flag
in r18,SREG
; Disable interrupts
cli
; Read TCNTn into r17:r16
in r16,TCNTnL
in r17,TCNTnH
; Restore global interrupt flag
out SREG,r18
ret
C Code Example(1)
unsigned int TIM16_ReadTCNTn( void )
{
unsigned char sreg;
unsigned int i;
/* Save global interrupt flag */
sreg = SREG;
/* Disable interrupts */
_CLI();
/* Read TCNTn into i */
i = TCNTn;
/* Restore global interrupt flag */
SREG = sreg;
return i;
}
The assembly code example returns the TCNTn value in the r17:r16 register pair.
110 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
The following code examples show how to do an atomic write of the TCNTn Register
contents. Writing any of the OCRnA/B/C or ICRn Registers can be done by using the
same principle.
TIM16_WriteTCNTn:
; Save global interrupt flag
in r18,SREG
; Disable interrupts
cli
; Set TCNTn to r17:r16
out TCNTnH,r17
out TCNTnL,r16
; Restore global interrupt flag
out SREG,r18
ret
C Code Example(1)
void TIM16_WriteTCNTn( unsigned int i )
{
unsigned char sreg;
unsigned int i;
/* Save global interrupt flag */
sreg = SREG;
/* Disable interrupts */
_CLI();
/* Set TCNTn to i */
TCNTn = i;
/* Restore global interrupt flag */
SREG = sreg;
}
The assembly code example requires that the r17:r16 register pair contains the value to
be written to TCNTn.
Reusing the Temporary High If writing to more than one 16-bit register where the high byte is the same for all registers
Byte Register written, then the high byte only needs to be written once. However, note that the same
rule of atomic operation described previously also applies in this case.
111
2467E–AVR–05/02
Timer/Counter Clock The Timer/Counter can be clocked by an internal or an external clock source. The clock
Sources source is selected by the clock select logic which is controlled by the Clock Select
(CSn2:0) bits located in the Timer/Counter Control Register B (TCCRnB). For details on
c lo ck s o u r c e s a n d p r e sc a l e r , s e e “ T im e r /C o u n t e r 3 , T im e r /C o u n t e r 2 , a n d
Timer/Counter1 Prescalers” on page 139.
Counter Unit The main part of the 16-bit Timer/Counter is the programmable 16-bit bi-directional
counter unit. Figure 46 shows a block diagram of the counter and its surroundings.

DATA BUS (8-bit)
TOVn
(Int.Req.)
TEMP (8-bit)
Clock Select
Count Edge
Tn
TCNTnH (8-bit) TCNTnL (8-bit) Clear clkTn Detector
Control Logic
Direction
TCNTn (16-bit Counter)
( From Prescaler )
TOP BOTTOM

Count Increment or decrement TCNTn by 1.
Direction Select between increment and decrement.
Clear Clear TCNTn (set all bits to zero).
clkTn Timer/Counter clock.
TOP Signalize that TCNTn has reached maximum value.
BOTTOM Signalize that TCNTn has reached minimum value (zero).
The 16-bit counter is mapped into two 8-bit I/O memory locations: Counter High
(TCNTnH) containing the upper 8 bits of the counter, and Counter Low (TCNTnL) con-
taining the lower 8 bits. The TCNTnH Register can only be indirectly accessed by the
CPU. When the CPU does an access to the TCNTnH I/O location, the CPU accesses
the high byte Temporary Register (TEMP). The Temporary Register is updated with the
TCNTnH value when the TCNTnL is read, and TCNTnH is updated with the Temporary
Register value when TCNTnL is written. This allows the CPU to read or write the entire
16-bit counter value within one clock cycle via the 8-bit data bus. It is important to notice
that there are special cases of writing to the TCNTn Register when the counter is count-
ing that will give unpredictable results. The special cases are described in the sections
where they are of importance.
Depending on the mode of operation used, the counter is cleared, incremented, or dec-
remented at each Timer Clock (clk Tn). The clkTn can be generated from an external or
internal clock source, selected by the Clock Select bits (CSn2:0). When no clock source
is selected (CSn2:0 = 0) the timer is stopped. However, the TCNTn value can be
accessed by the CPU, independent of whether clkTn is present or not. A CPU write over-
rides (has priority over) all counter clear or count operations.
The counting sequence is determined by the setting of the Waveform Generation mode
bits (WGMn3:0) located in the Timer/Counter Control Registers A and B (TCCRnA and
112 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
TCCRnB). There are close connections between how the counter behaves (counts) and
how waveforms are generated on the output compare outputs OCnx. For more details
about advanced counting sequences and waveform generation, see “Modes of Opera-
tion” on page 118.
The Timer/Counter Overflow (TOVn) flag is set according to the mode of operation
selected by the WGMn3:0 bits. TOVn can be used for generating a CPU interrupt.
Input Capture Unit The Timer/Counter incorporates an input capture unit that can capture external events
and give them a time-stamp indicating time of occurrence. The external signal indicating
an event, or multiple events, can be applied via the ICPn pin or alternatively, for the
Timer/Counter1 only, via the Analog Comparator unit. The time-stamps can then be
used to calculate frequency, duty-cycle, and other features of the signal applied. Alter-
natively the time-stamps can be used for creating a log of the events.
The Input Capture unit is illustrated by the block diagram shown in Figure 47. The ele-
ments of the block diagram that are not directly a part of the input capture unit are gray
shaded. The small “n” in register and bit names indicates the Timer/Counter number.
Figure 47. Input Capture Unit Block Diagram

DATA BUS (8-bit)
TEMP (8-bit)
ICRnH (8-bit) ICRnL (8-bit) TCNTnH (8-bit) TCNTnL (8-bit)
WRITE ICRn (16-bit Register) TCNTn (16-bit Counter)
ACO* ACIC* ICNC ICES
Analog
Comparator Noise Edge
ICFn (Int.Req.)
Canceler Detector
ICPn
Note: The Analog Comparator Output (ACO) can only trigger the Timer/Counter1 ICP – not
Timer/Counter3.
When a change of the logic level (an event) occurs on the Input Capture Pin (ICPn),
alternatively on the analog Comparator output (ACO), and this change confirms to the
setting of the edge detector, a capture will be triggered. When a capture is triggered, the
16-bit value of the counter (TCNTn) is written to the Input Capture Register (ICRn). The
Input Capture Flag (ICFn) is set at the same system clock as the TCNTn value is copied
into ICRn Register. If enabled (TICIEn = 1), the input capture flag generates an input
capture interrupt. The ICFn flag is automatically cleared when the interrupt is executed.
Alternatively the ICFn flag can be cleared by software by writing a logical one to its I/O
bit location.
113
2467E–AVR–05/02
Reading the 16-bit value in the Input Capture Register (ICRn) is done by first reading the
low byte (ICRnL) and then the high byte (ICRnH). When the low byte is read the high
byte is copied into the high byte Temporary Register (TEMP). When the CPU reads the
ICRnH I/O location it will access the TEMP Register.
The ICRn Register can only be written when using a Waveform Generation mode that
utilizes the ICRn Register for defining the counter’s TOP value. In these cases the
Waveform Generation mode (WGMn3:0) bits must be set before the TOP value can be
written to the ICRn Register. When writing the ICRn Register the high byte must be writ-
ten to the ICRnH I/O location before the low byte is written to ICRnL.
For more information on how to access the 16-bit registers refer to “Accessing 16-bit
Registers” on page 109.
Input Capture Trigger Source The main trigger source for the input capture unit is the Input Capture Pin (ICPn).
Timer/Counter1 can alternatively use the analog comparator output as trigger source for
the input capture unit. The Analog Comparator is selected as trigger source by setting
the analog Comparator Input Capture (ACIC) bit in the Analog Comparator Control and
Status Register (ACSR). Be aware that changing trigger source can trigger a capture.
The input capture flag must therefore be cleared after the change.
Both the Input Capture Pin (ICPn) and the Analog Comparator output (ACO) inputs are
sampled using the same technique as for the Tn pin (Figure 58 on page 139). The edge
detector is also identical. However, when the noise canceler is enabled, additional logic
is inserted before the edge detector, which increases the delay by four system clock
cycles. Note that the input of the noise canceler and edge detector is always enabled
unless the Timer/Counter is set in a Waveform Generation mode that uses ICRn to
define TOP.
An input capture can be triggered by software by controlling the port of the ICPn pin.
Noise Canceler The noise canceler improves noise immunity by using a simple digital filtering scheme.
The noise canceler input is monitored over four samples, and all four must be equal for
changing the output that in turn is used by the edge detector.
The noise canceler is enabled by setting the Input Capture Noise Canceler (ICNCn) bit
in Timer/Counter Control Register B (TCCRnB). When enabled the noise canceler intro-
duces additional four system clock cycles of delay from a change applied to the input, to
the update of the ICRn Register. The noise canceler uses the system clock and is there-
fore not affected by the prescaler.
Using the Input Capture Unit The main challenge when using the input capture unit is to assign enough processor
capacity for handling the incoming events. The time between two events is critical. If the
processor has not read the captured value in the ICRn Register before the next event
occurs, the ICRn will be overwritten with a new value. In this case the result of the cap-
ture will be incorrect.
When using the input capture interrupt, the ICRn Register should be read as early in the
interrupt handler routine as possible. Even though the input capture interrupt has rela-
tively high priority, the maximum interrupt response time is dependent on the maximum
number of clock cycles it takes to handle any of the other interrupt requests.
Using the input capture unit in any mode of operation when the TOP value (resolution) is
actively changed during operation, is not recommended.
Measurement of an external signal’s duty cycle requires that the trigger edge is changed
after each capture. Changing the edge sensing must be done as early as possible after
the ICRn Register has been read. After a change of the edge, the input capture flag
114 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
(ICFn) must be cleared by software (writing a logical one to the I/O bit location). For
measuring frequency only, the clearing of the ICFn flag is not required (if an interrupt
handler is used).
Output Compare Units The 16-bit comparator continuously compares TCNTn with the Output Compare Regis-
ter (OCRnx). If TCNT equals OCRnx the comparator signals a match. A match will set
the Output Compare Flag (OCFnx) at the next timer clock cycle. If enabled (OCIEnx =
1), the output compare flag generates an output compare interrupt. The OCFnx flag is
automatically cleared when the interrupt is executed. Alternatively the OCFnx flag can
be cleared by software by writing a logical one to its I/O bit location. The Waveform Gen-
erator uses the match signal to generate an output according to operating mode set by
the Waveform Generation mode (WGMn3:0) bits and Compare Output mode
(COMnx1:0) bits. The TOP and BOTTOM signals are used by the waveform generator
for handling the special cases of the extreme values in some modes of operation (See
“Modes of Operation” on page 118.)
A special feature of output compare unit A allows it to define the Timer/Counter TOP
value (i.e., counter resolution). In addition to the counter resolution, the TOP value
defines the period time for waveforms generated by the waveform generator.
Figure 48 shows a block diagram of the output compare unit. The small “n” in the regis-
ter and bit names indicates the device number (n = n for Timer/Counter n), and the “x”
indicates output compare unit (A/B/C). The elements of the block diagram that are not
directly a part of the output compare unit are gray shaded.
DATABUS (8-bit)
TEMP (8-bit)
OCRnxH Buf. (8-bit) OCRnxL Buf. (8-bit) TCNTnH (8-bit) TCNTnL (8-bit)
OCRnx Buffer (16-bit Register) TCNTn (16-bit Counter)
OCRnxH (8-bit) OCRnxL (8-bit)
OCRnx (16-bit Register)
OCFnx (Int.Req.)
TOP
Waveform Generator OCnx
BOTTOM
WGMn3:0 COMnx1:0
115
2467E–AVR–05/02
The OCRnx Register is double buffered when using any of the twelve Pulse Width Mod-
ulation (PWM) modes. For the normal and Clear Timer on Compare (CTC) modes of
operation, the double buffering is disabled. The double buffering synchronizes the
update of the OCRnx Compare Register to either TOP or BOTTOM of the counting
sequence. The synchronization prevents the occurrence of odd-length, non-symmetrical
PWM pulses, thereby making the output glitch-free.
The OCRnx Register access may seem complex, but this is not case. When the double
buffering is enabled, the CPU has access to the OCRnx buffer register, and if double
buffering is disabled the CPU will access the OCRnx directly. The content of the OCR1x
(buffer or compare) register is only changed by a write operation (the Timer/Counter
does not update this register automatically as the TCNTn- and ICRn Register). There-
fore OCRnx is not read via the high byte Temporary Register (TEMP). However, it is a
good practice to read the low byte first as when accessing other 16-bit registers. Writing
the OCRnx registers must be done via the TEMP Register since the compare of all 16
bits is done continuously. The high byte (OCRnxH) has to be written first. When the high
byte I/O location is written by the CPU, the TEMP Register will be updated by the value
written. Then when the low byte (OCRnxL) is written to the lower 8 bits, the high byte will
be copied into the upper 8 bits of either the OCRnx buffer or OCRnx Compare Register
in the same system clock cycle.
For more information of how to access the 16-bit registers refer to “Accessing 16-bit
Registers” on page 109.
Force Output Compare In non-PWM Waveform Generation modes, the match output of the comparator can be
forced by writing a one to the Force Output Compare (FOCnx) bit. Forcing compare
match will not set the OCFnx flag or reload/clear the timer, but the OCnx pin will be
updated as if a real compare match had occurred (the COMn1:0 bits settings define
whether the OCnx pin is set, cleared or toggled).
Compare Match Blocking by All CPU writes to the TCNTn Register will block any compare match that occurs in the
TCNTn Write next timer clock cycle, even when the timer is stopped. This feature allows OCRnx to be
initialized to the same value as TCNTn without triggering an interrupt when the
Timer/Counter clock is enabled.
Using the Output Compare Since writing TCNTn in any mode of operation will block all compare matches for one
Unit timer clock cycle, there are risks involved when changing TCNTn when using any of the
output compare channels, independent of whether the Timer/Counter is running or not.
If the value written to TCNTn equals the OCRnx value, the compare match will be
missed, resulting in incorrect waveform generation. Do not write the TCNTn equal to
TOP in PWM modes with variable TOP values. The compare match for the TOP will be
ignored and the counter will continue to 0xFFFF. Similarly, do not write the TCNTn value
The setup of the OCnx should be performed before setting the Data Direction Register
for the port pin to output. The easiest way of setting the OCnx value is to use the force
output compare (FOCnx) strobe bits in normal mode. The OCnx Register keeps its
value even when changing between waveform generation modes.
Be aware that the COMnx1:0 bits are not double buffered together with the compare
value. Changing the COMnx1:0 bits will take effect immediately.
116 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Compare Match Output The Compare Output mode (COMnx1:0) bits have two functions. The waveform genera-
Unit tor uses the COMnx1:0 bits for defining the output compare (OCnx) state at the next
compare match. Secondly the COMnx1:0 bits control the OCnx pin output source. Fig-
ure 49 shows a simplified schematic of the logic affected by the COMnx1:0 bit setting.
The I/O registers, I/O bits, and I/O pins in the figure are shown in bold. Only the parts of
the general I/O port control registers (DDR and PORT) that are affected by the
COMnx1:0 bits are shown. When referring to the OCnx state, the reference is for the
internal OCnx Register, not the OCnx pin. If a system Reset occur, the OCnx Register is
reset to “0”.
COMnx1
COMnx0 Waveform
D Q
FOCnx Generator
1
OCnx
OCnx Pin
0
D Q
DATA BUS
PORT
D Q
DDR
clk I/O
The general I/O port function is overridden by the output compare (OCnx) from the
Waveform Generator if either of the COMnx1:0 bits are set. However, the OCnx pin
direction (input or output) is still controlled by the Data Direction Register (DDR) for the
port pin. The data direction register bit for the OCnx pin (DDR_OCnx) must be set as
output before the OCnx value is visible on the pin. The port override function is generally
independent of the waveform generation mode, but there are some exceptions. Refer to
Table 58, Table 59 and Table 60 for details.
The design of the output compare pin logic allows initialization of the OCnx state before
the output is enabled. Note that some COMnx1:0 bit settings are reserved for certain
The COMnx1:0 bits have no effect on the input capture unit.
Compare Output Mode and The waveform generator uses the COMnx1:0 bits differently in normal, CTC, and PWM
Waveform Generation modes. For all modes, setting the COMnx1:0 = 0 tells the waveform generator that no
action on the OCnx Register is to be performed on the next compare match. For com-
pare output actions in the non-PWM modes refer to Table 58 on page 128. For fast
PWM mode refer to Table 59 on page 128, and for phase correct and phase and fre-
quency correct PWM refer to Table 60 on page 129.
117
2467E–AVR–05/02
A change of the COMnx1:0 bits state will have effect at the first compare match after the
by using the FOCnx strobe bits.
Modes of Operation The mode of operation, i.e., the behavior of the Timer/Counter and the output compare
pins, is defined by the combination of the Waveform Generation mode (WGMn3:0) and
Compare Output mode (COMnx1:0) bits. The Compare Output mode bits do not affect
the counting sequence, while the waveform generation mode bits do. The COMnx1:0
non-inverted PWM). For non-PWM modes the COMnx1:0 bits control whether the
output should be set, cleared or toggle at a compare match (See “Compare Match
Output Unit” on page 117.)
For detailed timing information refer to “Timer/Counter Timing Diagrams” on page 125.
Normal Mode The simplest mode of operation is the normal mode (WGMn3:0 = 0). In this mode the
counter simply overruns when it passes its maximum 16-bit value (MAX = 0xFFFF) and
then restarts from the BOTTOM (0x0000). In normal operation the Timer/Counter Over-
flow Flag (TOVn) will be set in the same timer clock cycle as the TCNTn becomes zero.
The TOVn flag in this case behaves like a 17th bit, except that it is only set, not cleared.
However, combined with the timer overflow interrupt that automatically clears the TOVn
flag, the timer resolution can be increased by software. There are no special cases to
consider in the normal mode, a new counter value can be written anytime.
The input capture unit is easy to use in normal mode. However, observe that the maxi-
mum interval between the external events must not exceed the resolution of the counter.
If the interval between events are too long, the timer overflow interrupt or the prescaler
must be used to extend the resolution for the capture unit.
The output compare units can be used to generate interrupts at some given time. Using
Clear Timer on Compare In Clear Timer on Compare or CTC mode (WGMn3:0 = 4 or 12), the OCRnA or ICRn
Match (CTC) Mode Register are used to manipulate the counter resolution. In CTC mode the counter is
cleared to zero when the counter value (TCNTn) matches either the OCRnA (WGMn3:0
= 4) or the ICRn (WGMn3:0 = 12). The OCRnA or ICRn define the top value for the
(TCNTn) increases until a compare match occurs with either OCRnA or ICRn, and then
counter (TCNTn) is cleared.
118 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
OCnA Interrupt Flag Set

or ICFn Interrupt Flag Set
(Interrupt on TOP)
TCNTn
OCnA
(COMnA1:0 = 1)
(Toggle)
Period 1 2 3 4
An interrupt can be generated at each time the counter value reaches the TOP value by
either using the OCFnA or ICFn flag according to the register used to define the TOP
value. If the interrupt is enabled, the interrupt handler routine can be used for updating
the TOP value. However, changing the TOP to a value close to BOTTOM when the
counter is running with none or a low prescaler value must be done with care since the
CTC mode does not have the double buffering feature. If the new value written to
OCRnA or ICRn is lower than the current value of TCNTn, the counter will miss the com-
pare match. The counter will then have to count to its maximum value (0xFFFF) and
wrap around starting at 0x0000 before the compare match can occur. In many cases
this feature is not desirable. An alternative will then be to use the fast PWM mode using
OCRnA for defining TOP (WGMn3:0 = 15) since the OCRnA then will be double
buffered.
For generating a waveform output in CTC mode, the OCnA output can be set to toggle
its logical level on each compare match by setting the compare output mode bits to tog-
gle mode (COMnA1:0 = 1). The OCnA value will not be visible on the port pin unless the
data direction for the pin is set to output (DDR_OCnA = 1). The waveform generated will
have a maximum frequency of fOCnA = fclk_I/O/2 when OCRnA is set to zero (0x0000). The
waveform frequency is defined by the following equation:
f clk_I/O
f OCn A = --------------------------------------------------
-
2 ⋅ N ⋅ ( 1 + OCRnA )
The N variable represents the prescaler factor (1, 8, 64, 256, or 1024).
As for the normal mode of operation, the TOVn flag is set in the same timer clock cycle
Fast PWM Mode The fast Pulse Width Modulation or fast PWM mode (WGMn3:0 = 5,6,7,14, or 15) pro-
vides a high frequency PWM waveform generation option. The fast PWM differs from
the other PWM options by its single-slope operation. The counter counts from BOTTOM
to TOP then restarts from BOTTOM. In non-inverting Compare Output mode, the output
compare (OCnx) is set on the compare match between TCNTn and OCRnx, and cleared
at TOP. In inverting compare output mode output is cleared on compare match and set
at TOP. Due to the single-slope operation, the operating frequency of the fast PWM
mode can be twice as high as the phase correct and phase and frequency correct PWM
modes that use dual-slope operation. This high frequency makes the fast PWM mode
well suited for power regulation, rectification, and DAC applications. High frequency
allows physically small sized external components (coils, capacitors), hence reduces
total system cost.
119
2467E–AVR–05/02
The PWM resolution for fast PWM can be fixed to 8-, 9-, or 10-bit, or defined by either
ICRn or OCRnA. The minimum resolution allowed is 2-bit (ICRn or OCRnA set to
0x0003), and the maximum resolution is 16-bit (ICRn or OCRnA set to MAX). The PWM
resolution in bits can be calculated by using the following equation:
log ( TOP + 1 )
R FPWM = -----------------------------------
log ( 2 )
In fast PWM mode the counter is incremented until the counter value matches either
one of the fixed values 0x00FF, 0x01FF, or 0x03FF (WGMn3:0 = 5, 6, or 7), the value in
ICRn (WGMn3:0 = 14), or the value in OCRnA (WGMn3:0 = 15). The counter is then
cleared at the following timer clock cycle. The timing diagram for the fast PWM mode is
shown in Figure 51. The figure shows fast PWM mode when OCRnA or ICRn is used to
define TOP. The TCNTn value is in the timing diagram shown as a histogram for illus-
trating the single-slope operation. The diagram includes non-inverted and inverted PWM
outputs. The small horizontal line marks on the TCNTn slopes represent compare
matches between OCRnx and TCNTn. The OCnx interrupt flag will be set when a com-
pare match occurs.

OCRnx / TOP Update
and
and
(Interrupt on TOP)
TCNTn
OCnx (COMnx1:0 = 2)
OCnx (COMnx1:0 = 3)
Period 1 2 3 4 5 6 7 8
The Timer/Counter Overflow Flag (TOVn) is set each time the counter reaches TOP. In
addition the OCnA or ICFn flag is set at the same timer clock cycle as TOVn is set when
either OCRnA or ICRn is used for defining the TOP value. If one of the interrupts are
enabled, the interrupt handler routine can be used for updating the TOP and compare
values.
When changing the TOP value the program must ensure that the new TOP value is
higher or equal to the value of all of the compare registers. If the TOP value is lower
than any of the compare registers, a compare match will never occur between the
TCNTn and the OCRnx. Note that when using fixed TOP values the unused bits are
masked to zero when any of the OCRnx Registers are written.
The procedure for updating ICRn differs from updating OCRnA when used for defining
the TOP value. The ICRn Register is not double buffered. This means that if ICRn is
changed to a low value when the counter is running with none or a low prescaler value,
there is a risk that the new ICRn value written is lower than the current value of TCNTn.
120 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
The result will then be that the counter will miss the compare match at the TOP value.
The counter will then have to count to the MAX value (0xFFFF) and wrap around start-
ing at 0x0000 before the compare match can occur. The OCRnA Register however, is
double buffered. This feature allows the OCRnA I/O location to be written anytime.
When the OCRnA I/O location is written the value written will be put into the OCRnA
buffer Register. The OCRnA Compare Register will then be updated with the value in
the buffer register at the next timer clock cycle the TCNTn matches TOP. The update is
done at the same timer clock cycle as the TCNTn is cleared and the TOVn flag is set.
Using the ICRn Register for defining TOP works well when using fixed TOP values. By
using ICRn, the OCRnA Register is free to be used for generating a PWM output on
OCnA. However, if the base PWM frequency is actively changed (by changing the TOP
value), using the OCRnA as TOP is clearly a better choice due to its double buffer
feature.
In fast PWM mode, the compare units allow generation of PWM waveforms on the
OCnx pins. Setting the COMnx1:0 bits to 2 will produce a non-inverted PWM and an
inverted PWM output can be generated by setting the COMnx1:0 to 3 (See Table 59 on
page 128). The actual OCnx value will only be visible on the port pin if the data direction
for the port pin is set as output (DDR_OCnx). The PWM waveform is generated by
setting (or clearing) the OCnx Register at the compare match between OCRnx and
TCNTn, and clearing (or setting) the OCnx Register at the timer clock cycle the counter
is cleared (changes from TOP to BOTTOM).
f clk_I/O
f OCnxPWM = ----------------------------------
-
N ⋅ ( 1 + TOP )
The N variable represents the prescaler divider (1, 8, 64, 256, or 1024).
The extreme values for the OCRnx Register represents special cases when generating
a PWM waveform output in the fast PWM mode. If the OCRnx is set equal to BOTTOM
(0x0000) the output will be a narrow spike for each TOP+1 timer clock cycle. Setting the
OCRnx equal to TOP will result in a constant high or low output (depending on the polar-
ity of the output set by the COMnx1:0 bits.)
by setting OCnA to toggle its logical level on each compare match (COMnA1:0 = 1). The
waveform generated will have a maximum frequency of fOCnA = fclk_I/O/2 when OCRnA is
set to zero (0x0000). This feature is similar to the OCnA toggle in CTC mode, except the
double buffer feature of the output compare unit is enabled in the fast PWM mode.
Phase Correct PWM Mode The phase correct Pulse Width Modulation or phase correct PWM mode (WGMn3:0 = 1,
2, 3, 10, or 11) provides a high resolution phase correct PWM waveform generation
option. The phase correct PWM mode is, like the phase and frequency correct PWM
mode, based on a dual-slope operation. The counter counts repeatedly from BOTTOM
(0x0000) to TOP and then from TOP to BOTTOM. In non-inverting compare output
mode, the output compare (OCnx) is cleared on the compare match between TCNTn
and OCRnx while upcounting, and set on the compare match while downcounting. In
inverting Output Compare mode, the operation is inverted. The dual-slope operation has
lower maximum operation frequency than single slope operation. However, due to the
symmetric feature of the dual-slope PWM modes, these modes are preferred for motor
control applications.
The PWM resolution for the phase correct PWM mode can be fixed to 8-, 9-, or 10-bit, or
defined by either ICRn or OCRnA. The minimum resolution allowed is 2 bit (ICRn or
121
2467E–AVR–05/02
OCRnA set to 0x0003), and the maximum resolution is 16 bit (ICRn or OCRnA set to
MAX). The PWM resolution in bits can be calculated by using the following equation:
log ( TOP + 1 )
R PCPWM = -----------------------------------
log ( 2 )
In phase correct PWM mode the counter is incremented until the counter value matches
either one of the fixed values 0x00FF, 0x01FF, or 0x03FF (WGMn3:0 = 1, 2, or 3), the
value in ICRn (WGMn3:0 = 10), or the value in OCRnA (WGMn3:0 = 11). The counter
has then reached the TOP and changes the count direction. The TCNTn value will be
equal to TOP for one timer clock cycle. The timing diagram for the phase correct PWM
mode is shown on Figure 52. The figure shows phase correct PWM mode when OCRnA
or ICRn is used to define TOP. The TCNTn value is in the timing diagram shown as a
histogram for illustrating the dual-slope operation. The diagram includes non-inverted
and inverted PWM outputs. The small horizontal line marks on the TCNTn slopes repre-
sent compare matches between OCRnx and TCNTn. The OCnx interrupt flag will be set
when a compare match occurs.
OCRnx / TOP Update

and
(Interrupt on TOP)

(Interrupt on Bottom)
TCNTn
OCnx (COMnx1:0 = 2)
OCnx (COMnx1:0 = 3)
Period 1 2 3 4
The Timer/Counter Overflow Flag (TOVn) is set each time the counter reaches BOT-
TOM. When either OCRnA or ICRn is used for defining the TOP value, the OCnA or
ICFn flag is set accordingly at the same timer clock cycle as the OCRnx Registers are
updated with the double buffer value (at TOP). The interrupt flags can be used to gener-
ate an interrupt each time the counter reaches the TOP or BOTTOM value.
TCNTn and the OCRnx. Note that when using fixed TOP values, the unused bits are
masked to zero when any of the OCRnx Registers are written. As the third period shown
in Figure 52 illustrates, changing the TOP actively while the Timer/Counter is running in
the phase correct mode can result in an unsymmetrical output. The reason for this can
be found in the time of update of the OCRnx Register. Since the OCRnx update occurs
122 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
at TOP, the PWM period starts and ends at TOP. This implies that the length of the fall-
ing slope is determined by the previous TOP value, while the length of the rising slope is
determined by the new TOP value. When these two values differ the two slopes of the
period will differ in length. The difference in length gives the unsymmetrical result on the
output.
It is recommended to use the phase and frequency correct mode instead of the phase
correct mode when changing the TOP value while the Timer/Counter is running. When
using a static TOP value there are practically no differences between the two modes of
operation.
In phase correct PWM mode, the compare units allow generation of PWM waveforms on
the OCnx pins. Setting the COMnx1:0 bits to 2 will produce a non-inverted PWM and an
inverted PWM output can be generated by setting the COMnx1:0 to 3 (See Table 60 on
page 129). The actual OCnx value will only be visible on the port pin if the data direction
for the port pin is set as output (DDR_OCnx). The PWM waveform is generated by set-
ting (or clearing) the OCnx Register at the compare match between OCRnx and TCNTn
when the counter increments, and clearing (or setting) the OCnx Register at compare
match between OCRnx and TCNTn when the counter decrements. The PWM frequency
for the output when using phase correct PWM can be calculated by the following
equation:
f clk_I/O
f OCn xPCPWM = ---------------------------
-
2 ⋅ N ⋅ TOP
The extreme values for the OCRnx Register represents special cases when generating
a PWM waveform output in the phase correct PWM mode. If the OCRnx is set equal to
BOTTOM the output will be continuously low and if set equal to TOP the output will be
Phase and Frequency Correct The phase and frequency correct Pulse Width Modulation, or phase and frequency cor-
PWM Mode rect PWM mode (WGMn3:0 = 8 or 9) provides a high resolution phase and frequency
correct PWM waveform generation option. The phase and frequency correct PWM
mode is, like the phase correct PWM mode, based on a dual-slope operation. The
counter counts repeatedly from BOTTOM (0x0000) to TOP and then from TOP to BOT-
TOM. In non-inverting Compare Output mode, the output compare (OCnx) is cleared on
the compare match between TCNTn and OCRnx while upcounting, and set on the com-
pare match while downcounting. In inverting Compare Output mode, the operation is
inverted. The dual-slope operation gives a lower maximum operation frequency com-
pared to the single-slope operation. However, due to the symmetric feature of the dual-
slope PWM modes, these modes are preferred for motor control applications.
The main difference between the phase correct, and the phase and frequency correct
PWM mode is the time the OCRnx Register is updated by the OCRnx buffer Register,
(see Figure 52 and Figure 53).
The PWM resolution for the phase and frequency correct PWM mode can be defined by
either ICRn or OCRnA. The minimum resolution allowed is 2-bit (ICRn or OCRnA set to
0x0003), and the maximum resolution is 16-bit (ICRn or OCRnA set to MAX). The PWM
resolution in bits can be calculated using the following equation:
log ( TOP + 1 )
R PFCPW M = -----------------------------------
log ( 2 )
123
2467E–AVR–05/02
In phase and frequency correct PWM mode the counter is incremented until the counter
value matches either the value in ICRn (WGMn3:0 = 8), or the value in OCRnA
(WGMn3:0 = 9). The counter has then reached the TOP and changes the count direc-
tion. The TCNTn value will be equal to TOP for one timer clock cycle. The timing
diagram for the phase correct and frequency correct PWM mode is shown on Figure 53.
The figure shows phase and frequency correct PWM mode when OCRnA or ICRn is
used to define TOP. The TCNTn value is in the timing diagram shown as a histogram for
illustrating the dual-slope operation. The diagram includes non-inverted and inverted
PWM outputs. The small horizontal line marks on the TCNTn slopes represent compare
matches between OCRnx and TCNTn. The OCnx interrupt flag will be set when a com-
pare match occurs.
Figure 53. Phase and Frequency Correct PWM Mode, Timing Diagram

(Interrupt on TOP)
OCRnx / TOP Update

and
(Interrupt on Bottom)
TCNTn
OCnx (COMnx1:0 = 2)
OCnx (COMnx1:0 = 3)
Period 1 2 3 4
The Timer/Counter Overflow Flag (TOVn) is set at the same timer clock cycle as the
OCRnx Registers are updated with the double buffer value (at BOTTOM). When either
OCRnA or ICRn is used for defining the TOP value, the OCnA or ICFn flag set when
TCNTn has reached TOP. The interrupt flags can then be used to generate an interrupt
each time the counter reaches the TOP or BOTTOM value.
TCNTn and the OCRnx.
As Figure 53 shows the output generated is, in contrast to the phase correct mode, sym-
metrical in all periods. Since the OCRnx Registers are updated at BOTTOM, the length
of the rising and the falling slopes will always be equal. This gives symmetrical output
pulses and is therefore frequency correct.
Using the ICRn Register for defining TOP works well when using fixed TOP values. By
using ICRn, the OCRnA Register is free to be used for generating a PWM output on
OCnA. However, if the base PWM frequency is actively changed by changing the TOP
value, using the OCRnA as TOP is clearly a better choice due to its double buffer
feature.
124 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
In phase and frequency correct PWM mode, the compare units allow generation of
PWM waveforms on the OCnx pins. Setting the COMnx1:0 bits to 2 will produce a non-
inverted PWM and an inverted PWM output can be generated by setting the COMnx1:0
to 3 (See Table 60 on page 129). The actual OCnx value will only be visible on the port
pin if the data direction for the port pin is set as output (DDR_OCnx). The PWM
waveform is generated by setting (or clearing) the OCnx Register at the compare match
between OCRnx and TCNTn when the counter increments, and clearing (or setting) the
OCnx Register at compare match between OCRnx and TCNTn when the counter
decrements. The PWM frequency for the output when using phase and frequency
correct PWM can be calculated by the following equation:
f clk_I/O
f OCnxPFCPWM = ---------------------------
-
2 ⋅ N ⋅ TOP
The extreme values for the OCRnx Register represent special cases when generating a
PWM waveform output in the phase correct PWM mode. If the OCRnx is set equal to
BOTTOM the output will be continuously low and if set equal to TOP the output will be
set to high for non-inverted PWM mode. For inverted PWM the output will have the
opposite logic values.
Timer/Counter Timing The Timer/Counter is a synchronous design and the timer clock (clkTn) is therefore
Diagrams shown as a clock enable signal in the following figures. The figures include information
on when interrupt flags are set, and when the OCRnx Register is updated with the
OCRnx buffer value (only for modes utilizing double buffering). Figure 54 shows a timing
diagram for the setting of OCFnx.
Figure 54. Timer/Counter Timing Diagram, Setting of OCFnx, no Prescaling
clkI/O
clkTn
(clkI/O /1)
TCNTn OCRnx - 1 OCRnx OCRnx + 1 OCRnx + 2
OCRnx OCRnx Value
OCFnx
125
2467E–AVR–05/02
Figure 55. Timer/Counter Timing Diagram, Setting of OCFnx, with Prescaler (fclk_I/O/8)
clkI/O
clkTn
(clkI/O /8)
TCNTn OCRnx - 1 OCRnx OCRnx + 1 OCRnx + 2
OCRnx OCRnx Value
OCFnx
Figure 56 shows the count sequence close to TOP in various modes. When using phase
and frequency correct PWM mode the OCRnx Register is updated at BOTTOM. The
timing diagrams will be the same, but TOP should be replaced by BOTTOM, TOP-1 by
BOTTOM+1 and so on. The same renaming applies for modes that set the TOVn flag at
BOTTOM.
Figure 56. Timer/Counter Timing Diagram, no Prescaling
clkI/O
clkTn
(clkI/O /1)
TCNTn
(CTC and FPWM)
TCNTn
TOP - 1 TOP TOP - 1 TOP - 2
(PC and PFC PWM)
TOVn (FPWM)
and ICFn (if used
as TOP)
OCRnx
Old OCRnx Value New OCRnx Value
(Update at TOP)
126 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
clkI/O
clkTn
(clkI/O /8)
TCNTn
(CTC and FPWM)
TCNTn
TOP - 1 TOP TOP - 1 TOP - 2
(PC and PFC PWM)
TOVn (FPWM)
and ICFn (if used
as TOP)
OCRnx
Old OCRnx Value New OCRnx Value
(Update at TOP)
16-bit Timer/Counter
Timer/Counter1 Control
Register A – TCCR1A Bit 7 6 5 4 3 2 1 0
COM1A1 COM1A0 COM1B1 COM1B0 COM1C1 COM1C0 WGM11 WGM10 TCCR1A
Register A – TCCR3A Bit 7 6 5 4 3 2 1 0
COM3A1 COM3A0 COM3B1 COM3B0 COM3C1 COM3C0 WGM31 WGM30 TCCR3A
• Bit 7:6 – COMnA1:0: Compare Output Mode for Channel A

• Bit 5:4 – COMnB1:0: Compare Output Mode for Channel B
• Bit 3:2 – COMnC1:0: Compare Output Mode for Channel C
The COMnA1:0, COMnB1:0, and COMnC1:0 control the output compare pins (OCnA,
OCnB, and OCnC respectively) behavior. If one or both of the COMnA1:0 bits are writ-
ten to one, the OCnA output overrides the normal port functionality of the I/O pin it is
connected to. If one or both of the COMnB1:0 bits are written to one, the OCnB output
overrides the normal port functionality of the I/O pin it is connected to. If one or both of
the COMnC1:0 bits are written to one, the OCnC output overrides the normal port func-
tionality of the I/O pin it is connected to. However, note that the Data Direction Register
(DDR) bit corresponding to the OCnA, OCnB or OCnC pin must be set in order to
enable the output driver.
When the OCnA, OCnB or OCnC is connected to the pin, the function of the COMnx1:0
bits is dependent of the WGMn3:0 bits setting. Table 58 shows the COMnx1:0 bit func-
tionality when the WGMn3:0 bits are set to a normal or a CTC mode (non-PWM).
127
2467E–AVR–05/02
Table 58. Compare Output Mode, non-PWM
COMnA1/COMnB1/ COMnA0/COMnB0/
COMnC1 COMnC0 Description
0 0 Normal port operation, OCnA/OCnB/OCnC
disconnected.
0 1 Toggle OCnA/OCnB/OCnC on compare
match
1 0 Clear OCnA/OCnB/OCnC on compare
match (Set output to low level)
1 1 Set OCnA/OCnB/OCnC on compare match
(Set output to high level)
Table 59 shows the COMnx1:0 bit functionality when the WGMn3:0 bits are set to the
fast PWM mode
Table 59. Compare Output Mode, Fast PWM

COMnA1/COMnB1/ COMnA0/COMnB0/
disconnected.
0 1 WGMn3=0: Normal port operation,
OCnA/OCnB/OCnC disconnected.
WGMn3=1: Toggle OCnA on compare
match, OCnB/OCnC reserved.
match, set OCnA/OCnB/OCnC at TOP
1 1 Set OCnA/OCnB/OCnC on compare match,
clear OCnA/OCnB/OCnC at TOP
Note: A special case occurs when OCRnA/OCRnB/OCRnC equals TOP and
COMnA1/COMnB1/COMnC1 is set. In this case the compare match is ignored, but the
set or clear is done at TOP. See “Fast PWM Mode” on page 119. for more details.
Table 59 shows the COMnx1:0 bit functionality when the WGMn3:0 bits are set to the
phase correct and frequency correct PWM mode.
128 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 60. Compare Output Mode, Phase Correct and Phase and Frequency Correct
PWM
COMnA1/COMnB/ COMnA0/COMnB0/
disconnected.
0 1 WGMn3=0: Normal port operation,
OCnA/OCnB/OCnC disconnected.
WGMn3=1: Toggle OCnA on compare
match, OCnB/OCnC reserved.
match when up-counting. Set
OCnA/OCnB/OCnC on compare match
when downcounting.
1 1 Set OCnA/OCnB/OCnC on compare match
when up-counting. Clear
OCnA/OCnB/OCnC on compare match
when downcounting.
Note: A special case occurs when OCRnA/OCRnB/OC RnC equals TOP and
COMnA1/COMnB1//COMnC1 is set. See “Phase Correct PWM Mode” on page 121. for
more details.
• Bit 1:0 – WGMn1:0: Waveform Generation Mode

Combined with the WGMn3:2 bits found in the TCCRnB Register, these bits control the
counting sequence of the counter, the source for maximum (TOP) counter value, and
what type of waveform generation to be used, see Table 61. Modes of operation sup-
ported by the Timer/Counter unit are: Normal mode (counter), Clear Timer on Compare
match (CTC) mode, and three types of Pulse Width Modulation (PWM) modes. (See
“Modes of Operation” on page 118.)
129
2467E–AVR–05/02
WGMn2 WGMn1 WGMn0 Timer/Counter Mode of Update of TOVn Flag
Mode WGMn3 (CTCn) (PWMn1) (PWMn0) Operation(1) TOP OCRnx at Set on
0 0 0 0 0 Normal 0xFFFF Immediate MAX
1 0 0 0 1 PWM, Phase Correct, 8-bit 0x00FF TOP BOTTOM
4 0 1 0 0 CTC OCRnA Immediate MAX
5 0 1 0 1 Fast PWM, 8-bit 0x00FF TOP TOP
8 1 0 0 0 PWM, Phase and Frequency
ICRn BOTTOM BOTTOM
Correct
9 1 0 0 1 PWM, Phase and Frequency
OCRnA BOTTOM BOTTOM
Correct
10 1 0 1 0 PWM, Phase Correct ICRn TOP BOTTOM
11 1 0 1 1 PWM, Phase Correct OCRnA TOP BOTTOM
12 1 1 0 0 CTC ICRn Immediate MAX
13 1 1 0 1 (Reserved) – – –
14 1 1 1 0 Fast PWM ICRn TOP TOP
15 1 1 1 1 Fast PWM OCRnA TOP TOP
Note: 1. The CTCn and PWMn1:0 bit definition names are obsolete. Use the WGMn2:0 definitions. However, the functionality and
location of these bits are compatible with previous versions of the timer.
Register B – TCCR1B Bit 7 6 5 4 3 2 1 0
ICNC1 ICES1 – WGM13 WGM12 CS12 CS11 CS10 TCCR1B
Read/Write R/W R/W R R/W R/W R/W R/W R/W
Register B – TCCR3B Bit 7 6 5 4 3 2 1 0
ICNC3 ICES3 – WGM33 WGM32 CS32 CS31 CS30 TCCR3B
• Bit 7 – ICNCn: Input Capture Noise Canceler

Setting this bit (to one) activates the Input Capture Noise Canceler. When the Noise
Canceler is activated, the input from the Input Capture Pin (ICPn) is filtered. The filter
function requires four successive equal valued samples of the ICPn pin for changing its
output. The input capture is therefore delayed by four Oscillator cycles when the noise
canceler is enabled.
130 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
• Bit 6 – ICESn: Input Capture Edge Select

This bit selects which edge on the Input Capture Pin (ICPn) that is used to trigger a cap-
ture event. When the ICESn bit is written to zero, a falling (negative) edge is used as
trigger, and when the ICESn bit is written to one, a rising (positive) edge will trigger the
capture.
When a capture is triggered according to the ICESn setting, the counter value is copied
into the Input Capture Register (ICRn). The event will also set the Input Capture Flag
(ICFn), and this can be used to cause an Input Capture Interrupt, if this interrupt is
enabled.
When the ICRn is used as TOP value (see description of the WGMn3:0 bits located in
the TCCRnA and the TCCRnB Register), the ICPn is disconnected and consequently
the input capture function is disabled.
• Bit 5 – Reserved Bit

This bit is reserved for future use. For ensuring compatibility with future devices, this bit
must be written to zero when TCCRnB is written.
• Bit 4:3 – WGMn3:2: Waveform Generation Mode

See TCCRnA Register description.
• Bit 2:0 – CSn2:0: Clock Select

The three clock select bits select the clock source to be used by the Timer/Counter, see
Figure 54 and Figure 55.

CSn2 CSn1 CSn0 Description
0 0 0 No clock source. (Timer/Counter stopped)
0 0 1 clkI/O/1 (No prescaling
0 1 0 clkI/O/8 (From prescaler)
1 1 0 External clock source on Tn pin. Clock on falling edge
1 1 1 External clock source on Tn pin. Clock on rising edge
If external pin modes are used for the Timer/Countern, transitions on the Tn pin will
clock the counter even if the pin is configured as an output. This feature allows software
control of the counting.
Register C – TCCR1C Bit 7 6 5 4 3 2 1 0
FOC1A FOC1B FOC1C – – – – – TCCR1C
Read/Write W W W R R R R R
131
2467E–AVR–05/02
Register C – TCCR3C Bit 7 6 5 4 3 2 1 0
FOC3A FOC3B FOC3C – – – – – TCCR3C
Read/Write W W W R R R R R
• Bit 7 – FOCnA: Force Output Compare for Channel A

• Bit 6 – FOCnB: Force Output Compare for Channel B
• Bit 5 – FOCnC: Force Output Compare for Channel C
The FOCnA/FOCnB/FOCnC bits are only active when the WGMn3:0 bits specifies a
non-PWM mode. When writing a logical one to the FOCnA/FOCnB/FOCnC bit, an
imme d ia te co mp ar e m atc h is for ce d on th e wa ve for m g en e ra tio n un it. Th e
OCnA/OCnB/OCnC output is changed according to its COMnx1:0 bits setting. Note that
the FOCnA/FOCnB/FOCnC bits are implemented as strobes. Therefore it is the value
present in the COMnx1:0 bits that determine the effect of the forced compare.
A FOCnA/FOCnB/FOCnC strobe will not generate any interrupt nor will it clear the timer
in Clear Timer on Compare Match (CTC) mode using OCRnA as TOP.
The FOCnA/FOCnB/FOCnB bits are always read as zero.
• Bit 4:0 – Reserved Bits
These bits are reserved for future use. For ensuring compatibility with future devices,
these bits must be written to zero when TCCRnC is written.
Timer/Counter1 – TCNT1H and

TCNT1L Bit 7 6 5 4 3 2 1 0
TCNT1[15:8] TCNT1H
TCNT1[7:0] TCNT1L
Timer/Counter3 – TCNT3H and

TCNT3L Bit 7 6 5 4 3 2 1 0
TCNT3[15:8] TCNT3H
TCNT3[7:0] TCNT3L
The two Timer/Counter I/O locations (TCNTnH and TCNTnL, combined TCNTn) give
direct access, both for read and for write operations, to the Timer/Counter unit 16-bit
counter. To ensure that both the high and low bytes are read and written simultaneously
when the CPU accesses these registers, the access is performed using an 8-bit tempo-
rary High Byte Register (TEMP). This Temporary Register is shared by all the other 16-
bit registers. See “Accessing 16-bit Registers” on page 109.
Modifying the counter (TCNTn) while the counter is running introduces a risk of missing
a compare match between TCNTn and one of the OCRnx Registers.
Writing to the TCNTn Register blocks (removes) the compare match on the following
timer clock for all compare units.
132 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Output Compare Register 1 A

– OCR1AH and OCR1AL Bit 7 6 5 4 3 2 1 0
OCR1A[15:8] OCR1AH
OCR1A[7:0] OCR1AL
Output Compare Register 1 B

– OCR1BH and OCR1BL Bit 7 6 5 4 3 2 1 0
OCR1B[15:8] OCR1BH
OCR1B[7:0] OCR1BL
Output Compare Register 1 C

– OCR1CH and OCR1CL Bit 7 6 5 4 3 2 1 0
OCR1C[15:8] OCR1CH
OCR1C[7:0] OCR1CL
Output Compare Register 3 A

– OCR3AH and OCR3AL Bit 7 6 5 4 3 2 1 0
OCR3A[15:8] OCR3AH
OCR3A[7:0] OCR3AL
Output Compare Register 3 B

– OCR3BH and OCR3BL Bit 7 6 5 4 3 2 1 0
OCR3B[15:8] OCR3BH
OCR3B[7:0] OCR3BL
Output Compare Register 3 C

– OCR3CH and OCR3CL Bit 7 6 5 4 3 2 1 0
OCR3C[15:8] OCR3CH
OCR3C[7:0] OCR3CL
The Output Compare Registers contain a 16-bit value that is continuously compared
with the counter value (TCNTn). A match can be used to generate an output compare
interrupt, or to generate a waveform output on the OCnx pin.
The Output Compare Registers are 16-bit in size. To ensure that both the high and low
bytes are written simultaneously when the CPU writes to these registers, the access is
performed using an 8-bit temporary High Byte Register (TEMP). This Temporary Regis-
ter is shared by all the other 16-bit registers. See “Accessing 16-bit Registers” on page
109.
133
2467E–AVR–05/02
Input Capture Register 1 –
ICR1H and ICR1L Bit 7 6 5 4 3 2 1 0
ICR1[15:8] ICR1H
ICR1[7:0] ICR1L
Input Capture Register 3 –

ICR3H and ICR3L Bit 7 6 5 4 3 2 1 0
ICR3[15:8] ICR3H
ICR3[7:0] ICR3L
The Input Capture is updated with the counter (TCNTn) value each time an event occurs
on the ICPn pin (or optionally on the Analog Comparator Output for Timer/Counter1).
The input capture can be used for defining the counter TOP value.
The Input Capture Register is 16-bit in size. To ensure that both the high and low bytes
are read simultaneously when the CPU accesses these registers, the access is per-
formed using an 8-bit temporary High Byte Register (TEMP). This Temporary Register
is shared by all the other 16-bit registers. See “Accessing 16-bit Registers” on page 109.

Note: This register contains interrupt control bits for several Timer/Counters, but only Timer1
bits are described in this section. The remaining bits are described in their respective
timer sections.
• Bit 5 – TICIE1: Timer/Counter1, Input Capture Interrupt Enable

When this bit is written to one, and the I-flag in the Status Register is set (interrupts glo-
b ally e nab le d), the Timer/Co unte r1 Inp ut Ca pture inte rrup t is en ab led . Th e
corresponding interrupt vector (See “Interrupts” on page 54.) is executed when the ICF1
flag, located in TIFR, is set.
• Bit 4 – OCIE1A: Timer/Counter1, Output Compare A Match Interrupt Enable

bally enabled), the Timer/Counter1 Output Compare A Match Interrupt is enabled. The
corresponding interrupt vector (see “Interrupts” on page 54) is executed when the
OCF1A flag, located in TIFR, is set.
• Bit 3 – OCIE1B: Timer/Counter1, Output Compare B Match Interrupt Enable

bally enabled), the Timer/Counter1 Output Compare B Match Interrupt is enabled. The
OCF1B flag, located in TIFR, is set.
134 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
• Bit 2 – TOIE1: Timer/Counter1, Overflow Interrupt Enable

bally enabled), the Timer/Counter1 overflow interrupt is enabled. The corresponding
interrupt vector (see “Interrupts” on page 54) is executed when the TOV1 flag, located in
TIFR, is set.
Extended Timer/Counter
Interrupt Mask Register – Bit 7 6 5 4 3 2 1 0
ETIMSK – – TICIE3 OCIE3A OCIE3B TOIE3 OCIE3C OCIE1C ETIMSK
Read/Write R R R/W R/W R/W R/W R/W R/W
Note: This register is not available in ATmega103 compatibility mode.

these bits must be set to zero when ETIMSK is written.
• Bit 5 – TICIE3: Timer/Counter3, Input Capture Interrupt Enable

b ally en abled ), the Timer/Co unte r3 Inpu t Cap tu re Interru pt is e nab le d. Th e
corresponding interrupt vector (see “Interrupts” on page 54) is executed when the ICF3
flag, located in ETIFR, is set.
• Bit 4 – OCIE3A: Timer/Counter3, Output Compare A Match Interrupt Enable

bally enabled), the Timer/Counter3 Output Compare A Match Interrupt is enabled. The
OCF3A flag, located in ETIFR, is set.
• Bit 3 – OCIE3B: Timer/Counter3, Output Compare B Match Interrupt Enable

bally enabled), the Timer/Counter3 Output Compare B Match Interrupt is enabled. The
OCF3B flag, located in ETIFR, is set.
• Bit 2 – TOIE3: Timer/Counter3, Overflow Interrupt Enable

bally enabled), the Timer/Counter3 Overflow Interrupt is enabled. The corresponding
interrupt vector (see “Interrupts” on page 54) is executed when the TOV3 flag, located in
ETIFR, is set.
• Bit 1 – OCIE3C: Timer/Counter3, Output Compare C Match Interrupt Enable

bally enabled), the Timer/Counter3 Output Compare C Match Interrupt is enabled. The
OCF3C flag, located in ETIFR, is set.
135
2467E–AVR–05/02
• Bit 0 – OCIE1C: Timer/Counter1, Output Compare C Match Interrupt Enable
bally enabled), the Timer/Counter1 Output Compare C Match Interrupt is enabled. The
OCF1C flag, located in ETIFR, is set.

Note: This register contains flag bits for several Timer/Counters, but only timer 1 bits are
described in this section. The remaining bits are described in their respective timer
sections.
• Bit 5 – ICF1: Timer/Counter1, Input Capture Flag

This flag is set when a capture event occurs on the ICP1 pin. When the Input Capture
Register (ICR1) is set by the WGMn3:0 to be used as the TOP value, the ICF1 flag is set
when the counter reaches the TOP value.
ICF1 is automatically cleared when the Input Capture Interrupt vector is executed. Alter-
natively, ICF1 can be cleared by writing a logic one to its bit location.
• Bit 4 – OCF1A: Timer/Counter1, Output Compare A Match Flag

This flag is set in the timer clock cycle after the counter (TCNT1) value matches the Out-
put Compare Register A (OCR1A).
Note that a forced output compare (FOC1A) strobe will not set the OCF1A flag.
OCF1A is automatically cleared when the Output Compare Match A interrupt vector is
executed. Alternatively, OCF1A can be cleared by writing a logic one to its bit location.
• Bit 3 – OCF1B: Timer/Counter1, Output Compare B Match Flag

put Compare Register B (OCR1B).
Note that a forced output compare (FOC1B) strobe will not set the OCF1B flag.
OCF1B is automatically cleared when the Output Compare Match B interrupt vector is
executed. Alternatively, OCF1B can be cleared by writing a logic one to its bit location.
• Bit 2 – TOV1: Timer/Counter1, Overflow Flag

The setting of this flag is dependent of the WGMn3:0 bits setting. In normal and CTC
modes, the TOV1 flag is set when the timer overflows. Refer to Table 61 on page 130
for the TOV1 flag behavior when using another WGMn3:0 bit setting.
TOV1 is automatically cleared when the Timer/Counter1 Overflow interrupt vector is
executed. Alternatively, TOV1 can be cleared by writing a logic one to its bit location.
136 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Extended Timer/Counter
Interrupt Flag Register – Bit 7 6 5 4 3 2 1 0
ETIFR – – ICF3 OCF3A OCF3B TOV3 OCF3C OCF1C ETIFR

these bits must be set to zero when ETIFR is written.
• Bit 5 – ICF3: Timer/Counter3, Input Capture Flag

This flag is set when a capture event occurs on the ICP3 pin. When the Input Capture
Register (ICR3) is set by the WGM3:0 to be used as the TOP value, the ICF3 flag is set
when the counter reaches the TOP value.
ICF3 is automatically cleared when the Input Capture 3 interrupt vector is executed.
Alternatively, ICF3 can be cleared by writing a logic one to its bit location.
• Bit 4 – OCF3A: Timer/Counter3, Output Compare A Match Flag

put Compare Register A (OCR3A).
Note that a forced output compare (FOC3A) strobe will not set the OCF3A flag.
OCF3A is automatically cleared when the Output Compare Match 3 A interrupt vector is
executed. Alternatively, OCF3A can be cleared by writing a logic one to its bit location.
• Bit 3 – OCF3B: Timer/Counter3, Output Compare B Match Flag

put Compare Register B (OCR3B).
Note that a forced output compare (FOC3B) strobe will not set the OCF3B flag.
OCF3B is automatically cleared when the Output Compare Match 3 B interrupt vector is
• Bit 2 – TOV3: Timer/Counter3, Overflow Flag

The setting of this flag is dependent of the WGM3:0 bits setting. In normal and CTC
modes, the TOV3 flag is set when the timer overflows. Refer to Table 52 on page 99 for
the TOV3 flag behavior when using another WGM3:0 bit setting.
TOV3 is automatically cleared when the Timer/Counter3 Overflow interrupt vector is
• Bit 1 – OCF3C: Timer/Counter3, Output Compare C Match Flag

put Compare Register C (OCR3C).
Note that a forced output compare (FOC3C) strobe will not set the OCF3C flag.
OCF3C is automatically cleared when the Output Compare Match 3 C interrupt vector is
executed. Alternatively, OCF3C can be cleared by writing a logic one to its bit location.
137
2467E–AVR–05/02
• Bit 0 – OCF1C: Timer/Counter1, Output Compare C Match Flag
put Compare Register C (OCR1C).
Note that a forced output compare (FOC1C) strobe will not set the OCF1C flag.
OCF1C is automatically cleared when the Output Compare Match 1 C interrupt vector is
executed. Alternatively, OCF1C can be cleared by writing a logic one to its bit location.
138 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Timer/Counter3, Timer/Counter3, Timer/Counter1, and Timer/Counter2 share the same prescaler mod-
ule, but the Timer/Counters can have different prescaler settings. The description below
Timer/Counter2, and
applies to all of the mentioned Timer/Counters.
Timer/Counter1
Prescalers
Internal Clock Source The Timer/Counter can be clocked directly by the System Clock (by setting the CSn2:0
= 1). This provides the fastest operation, with a maximum Timer/Counter clock fre-
quency equal to system clock frequency (fCLK_I/O). Alternatively, one of four taps from
the prescaler can be used as a clock source. The prescaled clock has a frequency of
either fCLK_I/O/8, fCLK_I/O/64, fCLK_I/O/256, or fCLK_I/O/1024.
Prescaler Reset The prescaler is free running, i.e., operates independently of the clock select logic of the
T ime r /C o u n te r , a n d it is s h a r e d b y T ime r /C o u n te r 1 , Tim er / Co u n t e r2 , a n d
Timer/Counter3. Since the prescaler is not affected by the Timer/Counter’s clock select,
the state of the prescaler will have implications for situations where a prescaled clock is
used. One example of prescaling artifacts occurs when the timer is enabled and clocked
by the prescaler (6 > CSn2:0 > 1). The number of system clock cycles from when the
timer is enabled to the first count occurs can be from 1 to N+1 system clock cycles,
where N equals the prescaler divisor (8, 64, 256, or 1024).
It is possible to use the Prescaler Reset for synchronizing the Timer/Counter to program
execution. However, care must be taken if the other Timer/Counter that shares the
same prescaler also use prescaling. A Prescaler Reset will affect the prescaler period
for all Timer/Counters it is connected to.
External Clock Source An external clock source applied to the Tn pin can be used as Timer/Counter clock
(clkT1/clkT2/clkT3). The Tn pin is sampled once every system clock cycle by the pin syn-
chronization logic. The synchronized (sampled) signal is then passed through the edge
detector. Figure 58 shows a functional equivalent block diagram of the Tn synchroniza-
tion and edge detector logic. The registers are clocked at the positive edge of the
internal system clock (clkI/O). The latch is transparent in the high period of the internal
system clock.
The edge detector generates one clkT1/clkT2/clkT3 pulse for each positive (CSn2:0 = 7) or
negative (CSn2:0 = 6) edge it detects.
Figure 58. Tn Pin Sampling
Tn D Q D Q D Q Tn_sync
(To Clock
Select Logic)
LE
clk I/O
Synchronization Edge Detector
The synchronization and edge detector logic introduces a delay of 2.5 to 3.5 system
clock cycles from an edge has been applied to the Tn pin to the counter is updated.
Enabling and disabling of the clock input must be done when Tn has been stable for at
least one system clock cycle, otherwise it is a risk that a false Timer/Counter clock pulse
is generated.
139
2467E–AVR–05/02
Each half period of the external clock applied must be longer than one system clock
cycle to ensure correct sampling. The external clock must be guaranteed to have less
than half the system clock frequency (fExtClk < fclk_I/O/2) given a 50/50% duty cycle. Since
the edge detector uses sampling, the maximum frequency of an external clock it can
detect is half the sampling frequency (Nyquist sampling theorem). However, due to vari-
ation of the system clock frequency and duty cycle caused by Oscillator source (crystal,
resonator, and capacitors) tolerances, it is recommended that maximum frequency of an
external clock source is less than fclk_I/O/2.5.
An external clock source can not be prescaled.
Figure 59. Prescaler for Timer/Counter1, Timer/Counter2, and Timer/Counter3
CK 10-BIT T/C PRESCALER

Clear
CK/256
CK/1024
CK/8
CK/64
PSR321
T3 T2 T1
0 0 0
CS30 CS20 CS10

CS31 CS21 CS11
CS32 CS22 CS12
TIMER/COUNTER3 CLOCK SOURCE TIMER/COUNTER2 CLOCK SOURCE TIMER/COUNTER1 CLOCK SOURCE

clkT3 clkT2 clkT1
Note: The synchronization logic on the input pins (T3/T2/T1) is shown in Figure 58.

SFIOR Bit 7 6 5 4 3 2 1 0
• Bit 7 – TSM: Timer/Counter Synchronization Mode

Writing TSM to one, PSR0 and PSR321 becomes registers that hold their value until
rewritten, or the TSM bit is written zero. This mode is useful for synchronizing
Timer/Counters. By setting both TSM and the appropriate PSR bit(s), the appropriate
Timer/Counters are halted, and can be configured to same value without the risk of one
of the m advancin g d uring configuration. W hen the TSM bit written zero, the
Timer/Counters start counting simultaneously.
• Bit 0 – PSR321: Prescaler Reset Timer/Counter3, Timer/Counter2, and

Timer/Counter1
Writing PSR321 to one resets the Prescaler for Timer/Counter3, Timer/Counter2, and
Timer/Counter1. The bit will be cleared by hardware after the operation is performed.
Writing a zero to this bit will have no effect. Note that Timer/Counter3 Timer/Counter2,
and Timer/Counter1 share the same prescaler and a Reset of this prescaler will affect
both timers. This bit will always be read as zero.
140 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
8-bit Timer/Counter2 Timer/Counter2 is a general purpose, single channel, 8-bit Timer/Counter module. The
main features are:
with PWM • Single Channel Counter
• Clear Timer on Compare Match (Auto Reload)
• Glitch-free, Phase Correct Pulse width Modulator (PWM)
• Frequency Generator
• External Event Counter
• 10-bit Clock Prescaler
• Overflow and Compare Match Interrupt Sources (TOV2 and OCF2)
Overview A simplified block diagram of the 8-bit Timer/Counter is shown in Figure 60. For the
actual placement of I/O pins, refer to “Pin Configurations” on page 2. CPU accessible
I/O registers, including I/O bits and I/O pins, are shown in bold. The device-specific I/O
register and bit locations are listed in the “8-bit Timer/Counter Register Description” on
page 152.
Figure 60. 8-Bit Timer/Counter Block Diagram
TCCRn
count TOVn
clear (Int.Req.)
Control Logic
direction clk Tn Clock Select
Edge
Tn
Detector
BOTTOM TOP
( From Prescaler )
DATA BUS
Timer/Counter
TCNTn
=0 = 0xFF OCn
(Int.Req.)
Waveform
= Generation
OCn
OCRn
Registers The Timer/Counter (TCNT2) and Output Compare Register (OCR2) are 8-bit registers.
Interrupt request (abbreviated to Int.Req. in the figure) signals are all visible in the Timer
Interrupt Flag Register (TIFR). All interrupts are individually masked with the Timer
Interrupt Mask Register (TIMSK). TIFR and TIMSK are not shown in the figure since
these registers are shared by other timer units.
The Timer/Counter can be clocked internally, via the prescaler, or by an external clock
source on the T2 pin. The Clock Select logic block controls which clock source and edge
the Timer/Counter uses to increment (or decrement) its value. The Timer/Counter is
141
2467E–AVR–05/02
inactive when no clock source is selected. The output from the clock select logic is
referred to as the timer clock (clkT2).
The double buffered Output Compare Register (OCR2) is compared with the
Timer/Counter value at all times. The result of the compare can be used by the wave-
form generator to generate a PWM or variable frequency output on the Output Compare
Pin (OC2). See “Output Compare Unit” on page 143. for details. The compare match
event will also set the compare flag (OCF2) which can be used to generate an output
Definitions Many register and bit references in this document are written in general form. A lower
case “n” replaces the Timer/Counter number, in this case 2. However, when using the
register or bit defines in a program, the precise form must be used (i.e., TCNT2 for
accessing Timer/Counter2 counter value and so on).
The definitions in Table 63 are also used extensively throughout the document.
BOTTOM The counter reaches the BOTTOM when it becomes 0x00.
MAX The counter reaches its MAXimum when it becomes 0xFF (decimal 255).
TOP The counter reaches the TOP when it becomes equal to the highest
value in the count sequence. The TOP value can be assigned to be the
fixed value 0xFF (MAX) or the value stored in the OCR2 Register. The
assignment is dependent on the mode of operation.
Timer/Counter Clock The Timer/Counter can be clocked by an internal or an external clock source. The clock
Sources source is selected by the clock select logic which is controlled by the clock select
(CS22:0) bits located in the Timer/Counter Control Register (TCCR2). For details on
c lo ck s o u r c e s a n d p r e sc a l e r , s e e “ T im e r /C o u n t e r 3 , T im e r /C o u n t e r 2 , a n d
Timer/Counter1 Prescalers” on page 139.
Counter Unit The main part of the 8-bit Timer/Counter is the programmable bi-directional counter unit.
Figure 61 shows a block diagram of the counter and its surroundings.

TOVn
DATA BUS (Int.Req.)
Clock Select
count Edge
Tn
clkTn Detector
clear
TCNTn Control Logic
direction
( From Prescaler )
bottom top

count Increment or decrement TCNT2 by 1.
direction Select between increment and decrement.
142 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
clear Clear TCNT2 (set all bits to zero).

clkTn Timer/Counter clock, referred to as clkT0 in the following.
top Signalize that TCNT2 has reached maximum value.
bottom Signalize that TCNT2 has reached minimum value (zero).
Depending of the mode of operation used, the counter is cleared, incremented, or dec-
remented at each timer clock (clkT2). clkT2 can be generated from an external or internal
clock source, selected by the clock select bits (CS22:0). When no clock source is
selected (CS22:0 = 0) the timer is stopped. However, the TCNT2 value can be accessed
by the CPU, regardless of whether clkT2 is present or not. A CPU write overrides (has
priority over) all counter clear or count operations.
The counting sequence is determined by the setting of the WGM01 and WGM00 bits
located in the Timer/Counter Control Register (TCCR2). There are close connections
between how the counter behaves (counts) and how waveforms are generated on the
output compare output OC2. For more details about advanced counting sequences and
waveform generation, see “Modes of Operation” on page 146.
The Timer/Counter overflow (TOV2) flag is set according to the mode of operation
selected by the WGM21:0 bits. TOV2 can be used for generating a CPU interrupt.
Output Compare Unit The 8-bit comparator continuously compares TCNT2 with the Output Compare Register
(OCR2). Whenever TCNT2 equals OCR2, the comparator signals a match. A match will
set the output compare flag (OCF2) at the next timer clock cycle. If enabled (OCIE2 = 1
and global interrupt flag in SREG is set), the output compare flag generates an output
compare interrupt. The OCF2 flag is automatically cleared when the interrupt is exe-
cuted. Alternatively, the OCF2 flag can be cleared by software by writing a logical one to
its I/O bit location. The waveform generator uses the match signal to generate an output
according to operating mode set by the WGM21:0 bits and compare output mode
(COM21:0) bits. The max and bottom signals are used by the waveform generator for
handling the special cases of the extreme values in some modes of operation (see
“Modes of Operation” on page 146). Figure 62 shows a block diagram of the output
compare unit.
143
2467E–AVR–05/02
DATA BUS
OCRn TCNTn
OCFn (Int.Req.)
top
bottom
Waveform Generator OCn
FOCn
WGMn1:0 COMn1:0
The OCR2 Register is double buffered when using any of the pulse width modulation
(PWM) modes. For the normal and Clear Timer on Compare (CTC) modes of operation,
the double buffering is disabled. The double buffering synchronizes the update of the
OCR2 Compare Register to either top or bottom of the counting sequence. The synchro-
nization prevents the occurrence of odd-length, non-symmetrical PWM pulses, thereby
making the output glitch-free.
The OCR2 Register access may seem complex, but this is not case. When the double
buffering is enabled, the CPU has access to the OCR2 buffer Register, and if double
buffering is disabled the CPU will access the OCR2 directly.
Force Output Compare In non-PWM Waveform Generation modes, the match output of the comparator can be
forced by writing a one to the force output compare (FOC2) bit. Forcing compare match
will not set the OCF2 flag or reload/clear the timer, but the OC2 pin will be updated as if
a real compare match had occurred (the COM21:0 bits settings define whether the OC2
pin is set, cleared or toggled).
Compare Match Blocking by All CPU write operations to the TCNT2 Register will block any compare match that
TCNT2 Write occur in the next timer clock cycle, even when the timer is stopped. This feature allows
OCR2 to be initialized to the same value as TCNT2 without triggering an interrupt when
the Timer/Counter clock is enabled.
Using the Output Compare Since writing TCNT2 in any mode of operation will block all compare matches for one
Unit timer clock cycle, there are risks involved when changing TCNT2 when using the output
compare channel, independently of whether the Timer/Counter is running or not. If the
value written to TCNT2 equals the OCR2 value, the compare match will be missed,
resulting in incorrect waveform generation. Similarly, do not write the TCNT2 value
The setup of the OC2 should be performed before setting the Data Direction Register for
the port pin to output. The easiest way of setting the OC2 value is to use the Force Out-
144 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
put Compare (FOC2) strobe bits in normal mode. The OC2 Register keeps its value
even when changing between waveform generation modes.
Be aware that the COM21:0 bits are not double buffered together with the compare
value. Changing the COM21:0 bits will take effect immediately.
Compare Match Output The Compare Output mode (COM21:0) bits have two functions. The waveform genera-
Unit tor uses the COM21:0 bits for defining the output compare (OC2) state at the next
compare match. Also, the COM21:0 bits control the OC2 pin output source. Figure 63
shows a simplified schematic of the logic affected by the COM21:0 bit setting. The I/O
registers, I/O bits, and I/O pins in the figure are shown in bold. Only the parts of the gen-
eral I/O Port Control Registers (DDR and PORT) that are affected by the COM21:0 bits
are shown. When referring to the OC2 state, the reference is for the internal OC2 Regis-
ter, not the OC2 pin. If a System Reset occur, the OC2 Register is reset to “0”.
COMn1
COMn0 Waveform
D Q
FOCn Generator
1
OCn
OCn Pin
0
D Q
DATA BUS
PORT
D Q
DDR
clk I/O
The general I/O port function is overridden by the output compare (OC2) from the wave-
form generator if either of the COM21:0 bits are set. However, the OC2 pin direction
(input or output) is still controlled by the Data Direction Register (DDR) for the port pin.
The Data Direction Register bit for the OC2 pin (DDR_OC2) must be set as output
before the OC2 value is visible on the pin. The port override function is independent of
the Waveform Generation mode.
The design of the output compare pin logic allows initialization of the OC2 state before
the output is enabled. Note that some COM21:0 bit settings are reserved for certain
Compare Output Mode and The waveform generator uses the COM21:0 bits differently in normal, CTC, and PWM
Waveform Generation modes. For all modes, setting the COM21:0 = 0 tells the waveform generator that no
action on the OC2 Register is to be performed on the next compare match. For compare
output actions in the non-PWM modes refer to Table 65 on page 153. For fast PWM
mode, refer to Table 66 on page 153, and for phase correct PWM refer to Table 67 on
page 153.
145
2467E–AVR–05/02
A change of the COM21:0 bits state will have effect at the first compare match after the
by using the FOC2 strobe bits.
Modes of Operation The mode of operation, i.e., the behavior of the Timer/Counter and the Output Compare
pins, is defined by the combination of the Waveform Generation mode (WGM21:0) and
Compare Output mode (COM21:0) bits. The Compare Output mode bits do not affect
the counting sequence, while the Waveform Generation mode bits do. The COM21:0
non-inverted PWM). For non-PWM modes the COM21:0 bits control whether the output
should be set, cleared, or toggled at a compare match (see “Compare Match Output
Unit” on page 145).
For detailed timing information refer to Figure 67, Figure 68, Figure 69, and Figure 70 in
“Timer/Counter Timing Diagrams” on page 150.
Normal Mode The simplest mode of operation is the normal mode (WGM21:0 = 0). In this mode the
counter simply overruns when it passes its maximum 8-bit value (TOP = 0xFF) and then
restarts from the bottom (0x00). In normal operation the Timer/Counter overflow flag
(TOV2) will be set in the same timer clock cycle as the TCNT2 becomes zero. The TOV2
flag in this case behaves like a ninth bit, except that it is only set, not cleared. However,
combined with the timer overflow interrupt that automatically clears the TOV2 flag, the
timer resolution can be increased by software. There are no special cases to consider in
the normal mode, a new counter value can be written anytime.
The output compare unit can be used to generate interrupts at some given time. Using
Clear Timer on Compare In Clear Timer on Compare or CTC mode (WGM21:0 = 2), the OCR2 Register is used to
Match (CTC) Mode manipulate the counter resolution. In CTC mode the counter is cleared to zero when the
counter value (TCNT2) matches the OCR2. The OCR2 defines the top value for the
(TCNT2) increases until a compare match occurs between TCNT2 and OCR2 and then
counter (TCNT2) is cleared.
TCNTn
OCn
(COMn1:0 = 1)
(Toggle)
Period 1 2 3 4
146 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
An interrupt can be generated each time the counter value reaches the TOP value by
using the OCF2 flag. If the interrupt is enabled, the interrupt handler routine can be used
for updating the TOP value. However, changing the TOP to a value close to BOTTOM
when the counter is running with none or a low prescaler value must be done with care
since the CTC mode does not have the double buffering feature. If the new value written
to OCR2 is lower than the current value of TCNT2, the counter will miss the compare
match. The counter will then have to count to its maximum value (0xFF) and wrap
around starting at 0x00 before the compare match can occur.
For generating a waveform output in CTC mode, the OC2 output can be set to toggle its
logical level on each compare match by setting the compare output mode bits to toggle
mode (COM21:0 = 1). The OC2 value will not be visible on the port pin unless the data
direction for the pin is set to output. The waveform generated will have a maximum fre-
quency of fOC2 = fclk_I/O/2 when OCR2 is set to zero (0x00). The waveform frequency is
defined by the following equation:
f clk_I/O
f OCn = ----------------------------------------------
-
2 ⋅ N ⋅ ( 1 + OCRn )
The N variable represents the prescale factor (1, 8, 64, 256, or 1024).
As for the normal mode of operation, the TOV2 flag is set in the same timer clock cycle
Fast PWM Mode The fast Pulse Width Modulation or fast PWM mode (WGM21:0 = 1) provides a high fre-
quency PWM waveform generation option. The fast PWM differs from the other PWM
option by its single-slope operation. The counter counts from BOTTOM to MAX then
restarts from BOTTOM. In non-inverting Compare Output mode, the output compare
(OC2) is cleared on the compare match between TCNT2 and OCR2, and set at BOT-
TOM. In inverting Compare Output mode, the output is set on compare match and
cleared at BOTTOM. Due to the single-slope operation, the operating frequency of the
fast PWM mode can be twice as high as the phase correct PWM mode that use dual-
slope operation. This high frequency makes the fast PWM mode well suited for power
regulation, rectification, and DAC applications. High frequency allows physically small
sized external components (coils, capacitors), and therefore reduces total system cost.
In fast PWM mode, the counter is incremented until the counter value matches the MAX
value. The counter is then cleared at the following timer clock cycle. The timing diagram
for the fast PWM mode is shown in Figure 65. The TCNT2 value is in the timing diagram
shown as a histogram for illustrating the single-slope operation. The diagram includes
non-inverted and inverted PWM outputs. The small horizontal line marks on the TCNT2
slopes represent compare matches between OCR2 and TCNT2.
147
2467E–AVR–05/02
OCRn Interrupt Flag Set
OCRn Update
and
TCNTn
OCn (COMn1:0 = 2)
OCn (COMn1:0 = 3)
Period 1 2 3 4 5 6 7
The Timer/Counter overflow flag (TOV2) is set each time the counter reaches Max If the
interrupt is enabled, the interrupt handler routine can be used for updating the compare
value.
In fast PWM mode, the compare unit allows generation of PWM waveforms on the OC2
pin. Setting the COM21:0 bits to 2 will produce a non-inverted PWM and an inverted
PWM output can be generated by setting the COM21:0 to 3 (see Table 66 on page 153).
The actual OC2 value will only be visible on the port pin if the data direction for the port
pin is set as output. The PWM waveform is generated by setting (or clearing) the OC2
Register at the compare match between OCR2 and TCNT2, and clearing (or setting) the
OC2 Register at the timer clock cycle the counter is cleared (changes from MAX to
BOTTOM).
f clk_I/O
f OCnPWM = -----------------
-
N ⋅ 256
The extreme values for the OCR2 Register represents special cases when generating a
PWM waveform output in the fast PWM mode. If the OCR2 is set equal to BOTTOM, the
output will be a narrow spike for each MAX+1 timer clock cycle. Setting the OCR2 equal
to MAX will result in a constantly high or low output (depending on the polarity of the out-
put set by the COM21:0 bits.)
by setting OC2 to toggle its logical level on each compare match (COM21:0 = 1). The
waveform generated will have a maximum frequency of fOC2 = fclk_I/O/2 when OCR2 is
set to zero. This feature is similar to the OC2 toggle in CTC mode, except the double
buffer feature of the output compare unit is enabled in the fast PWM mode.
Phase Correct PWM Mode The phase correct PWM mode (WGM21:0 = 3) provides a high resolution phase correct
PWM waveform generation option. The phase correct PWM mode is based on a dual-
slope operation. The counter counts repeatedly from BOTTOM to MAX and then from
MAX to BOTTOM. In non-inverting Compare Output mode, the output compare (OC2) is
148 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
cleared on the compare match between TCNT2 and OCR2 while upcounting, and set on
the compare match while downcounting. In inverting Output Compare mode, the opera-
tion is inverted. The dual-slope operation has lower maximum operation frequency than
single slope operation. However, due to the symmetric feature of the dual-slope PWM
modes, these modes are preferred for motor control applications.
The PWM resolution for the phase correct PWM mode is fixed to 8 bits. In phase correct
PWM mode the counter is incremented until the counter value matches Max When the
counter reaches MAX, it changes the count direction. The TCNT2 value will be equal to
MAX for one timer clock cycle. The timing diagram for the phase correct PWM mode is
shown on Figure 66. The TCNT2 value is in the timing diagram shown as a histogram
for illustrating the dual-slope operation. The diagram includes non-inverted and inverted
PWM outputs. The small horizontal line marks on the TCNT2 slopes represent compare
matches between OCR2 and TCNT2.
OCRn Update
TCNTn
OCn (COMn1:0 = 2)
OCn (COMn1:0 = 3)
Period 1 2 3
The Timer/Counter Overflow Flag (TOV2) is set each time the counter reaches BOT-
TOM. The interrupt flag can be used to generate an interrupt each time the counter
reaches the BOTTOM value.
In phase correct PWM mode, the compare unit allows generation of PWM waveforms on
the OC2 pin. Setting the COM21:0 bits to 2 will produce a non-inverted PWM. An
inverted PWM output can be generated by setting the COM21:0 to 3 (see Table 67 on
page 153). The actual OC2 value will only be visible on the port pin if the data direction
for the port pin is set as output. The PWM waveform is generated by clearing (or setting)
the OC2 Register at the compare match between OCR2 and TCNT2 when the counter
increments, and setting (or clearing) the OC2 Register at compare match between
OCR2 and TCNT2 when the counter decrements. The PWM frequency for the output
when using phase correct PWM can be calculated by the following equation:
f clk_I/O
f OCn PCPW M = -----------------
-
N ⋅ 510
149
2467E–AVR–05/02
PWM waveform output in the phase correct PWM mode. If the OCR2 is set equal to
BOTTOM, the output will be continuously low and if set equal to MAX the output will be
Timer/Counter Timing The Timer/Counter is a synchronous design and the timer clock (clkT2) is therefore
Diagrams shown as a clock enable signal in the following figures. The figures include information
on when interrupt flags are set. Figure 67 contains timing data for basic Timer/Counter
operation. The figure shows the count sequence close to the MAX value in all modes
other than phase correct PWM mode.
Figure 67. Timer/Counter Timing Diagram, no Prescaling
clkI/O
clkTn
(clkI/O /1)
TOVn
clkI/O
clkTn
(clkI/O /8)
TOVn
Figure 69 shows the setting of OCF2 in all modes except CTC mode.
150 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 69. Timer/Counter Timing Diagram, Setting of OCF2, with Prescaler (fclk_I/O/8)
clkI/O
clkTn
(clkI/O /8)
TCNTn OCRn - 1 OCRn OCRn + 1 OCRn + 2
OCRn OCRn Value
OCFn
Figure 70 shows the setting of OCF2 and the clearing of TCNT2 in CTC mode.
Figure 70. Timer/Counter Timing Diagram, Clear Timer on Compare Match Mode, with
Prescaler (fclk_I/O/8)
clkI/O
clkTn
(clkI/O /8)
TCNTn
(CTC)
OCRn TOP
OCFn
151
2467E–AVR–05/02
8-bit Timer/Counter
Timer/Counter Control
Register – TCCR2 Bit 7 6 5 4 3 2 1 0
FOC2 WGM20 COM21 COM20 WGM21 CS22 CS21 CS20 TCCR2
Read/Write W R/W R/W R/W R/W R/W R/W R/W
• Bit 7 – FOC2: Force Output Compare

The FOC2 bit is only active when the WGM20 bit specifies a non-PWM mode. However,
for ensuring compatibility with future devices, this bit must be set to zero when TCCR2 is
written when operating in PWM mode. When writing a logical one to the FOC2 bit, an
immediate compare match is forced on the waveform generation unit. The OC2 output is
changed according to its COM21:0 bits setting. Note that the FOC2 bit is implemented
as a strobe. Therefore it is the value present in the COM21:0 bits that determines the
effect of the forced compare.
A FOC2 strobe will not generate any interrupt, nor will it clear the Timer in CTC mode
using OCR2 as TOP.
The FOC2 bit is always read as zero.
• Bit 6, 3 – WGM21:0: Waveform Generation Mode

These bits control the counting sequence of the counter, the source for the maximum
(TOP) counter value, and what type of waveform generation to be used. Modes of oper-
ation supported by the Timer/Counter unit are: Normal mode, Clear Timer on Compare
match (CTC) mode, and two types of Pulse Width Modulation (PWM) modes. See Table
64 and “Modes of Operation” on page 146.

WGM21 WGM20 Timer/Counter Mode Update of TOV2 Flag
Mode (CTC2) (PWM2) of Operation TOP OCR2 at Set on
0 0 0 Normal 0xFF Immediate MAX
1 0 1 PWM, Phase Correct 0xFF TOP BOTTOM
2 1 0 CTC OCR2 Immediate MAX
3 1 1 Fast PWM 0xFF TOP MAX
Note: The CTC2 and PWM2 bit definition names are now obsolete. Use the WGM21:0 defini-
tions. However, the functionality and location of these bits are compatible with previous
versions of the timer.
• Bit 5:4 – COM21:0: Compare Match Output Mode

These bits control the Output Compare Pin (OC2) behavior. If one or both of the
COM21:0 bits are set, the OC2 output overrides the normal port functionality of the I/O
pin it is connected to. However, note that the Data Direction Register (DDR) bit corre-
sponding to the OC2 pin must be set in order to enable the output driver.
152 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
When OC2 is connected to the pin, the function of the COM21:0 bits depends on the
WGM21:0 bit setting. Table 65 shows the COM21:0 bit functionality when the WGM21:0
bits are set to a normal or CTC mode (non-PWM).
Table 65. Compare Output Mode, Non-PWM Mode

0 1 Toggle OC2 on compare match
1 0 Clear OC2 on compare match
1 1 Set OC2 on compare match
Table 66 shows the COM21:0 bit functionality when the WGM21:0 bits are set to fast
PWM mode.
Table 66. Compare Output Mode, Fast PWM Mode(1)

0 1 Reserved
1 0 Clear OC2 on compare match, set OC2 at TOP
1 1 Set OC2 on compare match, clear OC2 at TOP
compare match is ignored, but the set or clear is done at TOP. See “Fast PWM Mode”
on page 147 for more details.
Table 67 shows the COM21:0 bit functionality when the WGM21:0 bits are set to phase
correct PWM mode.
Table 67. Compare Output Mode, Phase Correct PWM Mode(1)

0 1 Reserved
1 0 Clear OC2 on compare match when up-counting. Set OC2 on compare
1 1 Set OC2 on compare match when up-counting. Clear OC2 on compare
compare match is ignored, but the set or clear is done at TOP. See “Phase Correct
PWM Mode” on page 148 for more details.
• Bit 2:0 – CS22:0: Clock Select

The three clock select bits select the clock source to be used by the Timer/Counter.
153
2467E–AVR–05/02
CS22 CS21 CS20 Description
0 0 0 No clock source (Timer/Counter stopped)
0 0 1 clkI/O/(No prescaling)
1 1 0 External clock source on T2 pin. Clock on falling edge
1 1 1 External clock source on T2 pin. Clock on rising edge
If external pin modes are used for the Timer/Counter2, transitions on the T2 pin will
clock the counter even if the pin is configured as an output. This feature allows software
control of the counting.
Timer/Counter Register –
TCNT2 Bit 7 6 5 4 3 2 1 0
TCNT2[7:0] TCNT2
The Timer/Counter Register gives direct access, both for read and write operations, to
the Timer/Counter unit 8-bit counter. Writing to the TCNT2 Register blocks (removes)
the compare match on the following timer clock. Modifying the counter (TCNT2) while
the counter is running, introduces a risk of missing a compare match between TCNT2
and the OCR2 Register.
Output Compare Register –

OCR2 Bit 7 6 5 4 3 2 1 0
OCR2[7:0] OCR2
The Output Compare Register contains an 8-bit value that is continuously compared
with the counter value (TCNT2). A match can be used to generate an output compare
interrupt, or to generate a waveform output on the OC2 pin.

• Bit 7 – OCIE2: Timer/Counter2 Output Compare Match Interrupt Enable

When the OCIE2 bit is written to one, and the I-bit in the Status Register is set (one), the
Timer/Counter2 Compare Match interrupt is enabled. The corresponding interrupt is
executed if a compare match in Timer/Counter2 occurs, i.e., when the OCF2 bit is set in
the Timer/Counter Interrupt Flag Register – TIFR.
154 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
• Bit 6 – TOIE2: Timer/Counter2 Overflow Interrupt Enable

When the TOIE2 bit is written to one, and the I-bit in the Status Register is set (one), the
Timer/Counter2 Overflow interrupt is enabled. The corresponding interrupt is executed if
a n ove rflow in Time r/C oun te r2 o ccurs, i.e., whe n the TO V2 bit is set in th e
Timer/Counter Interrupt Flag Register – TIFR.

• Bit 7 – OCF2: Output Compare Flag 2

The OCF2 bit is set (one) when a compare match occurs between the Timer/Counter2
and the data in OCR2 – Output Compare Register2. OCF2 is cleared by hardware when
executing the corresponding interrupt handling vector. Alternatively, OCF2 is cleared by
writing a logic one to the flag. When the I-bit in SREG, OCIE2 (Timer/Counter2 Com-
pare Match Interrupt Enable), and OCF2 are set (one), the Timer/Counter2 Compare
Match Interrupt is executed.
• Bit 6 – TOV2: Timer/Counter2 Overflow Flag

The bit TOV2 is set (one) when an overflow occurs in Timer/Counter2. TOV2 is cleared
by hardware when executing the corresponding interrupt handling vector. Alternatively,
TOV2 is cleared by writing a logic one to the flag. When the SREG I-bit, TOIE2
( Time r /C o u n te r2 O ve r flo w In te r ru p t En a b le ), a n d TO V 2 a r e se t ( o ne ) , th e
Timer/Counter2 Overflow interrupt is executed. In PWM mode, this bit is set when
Timer/Counter2 changes counting direction at $00.
155
2467E–AVR–05/02
Output Compare
Modulator (OCM1C2)
Overview The Output Compare Modulator (OCM) allows generation of waveforms modulated with
a carrier frequency. The modulator uses the outputs from the Output Compare Unit C of
the 16-bit Timer/Counter1 and the Output Compare Unit of the 8-bit Timer/Counter2. For
more details about these Timer/Counters see “16-bit Timer/Counter (Timer/Counter1
and Timer/Counter3)” on page 106 and “8-bit Timer/Counter2 with PWM” on page 141.
Note that this feature is not available in ATmega103 compatibility mode.
Figure 71. Output Compare Modulator, Block Diagram
Timer/Counter 1 OC1C
Pin
OC1C /
Timer/Counter 2 OC2 OC2 / PB7
When the modulator is enabled, the two output compare channels are modulated
together as shown in the block diagram (Figure 71).
Description The Output Compare unit 1C and Output Compare unit 2 shares the PB7 port pin for
output. The outputs of the Output Compare units (OC1C and OC2) overrides the normal
PORTB7 Register when one of them is enabled (i.e., when COMnx1:0 is not equal to
zero). When both OC1C and OC2 are enabled at the same time, the modulator is auto-
matically enabled.
The functional equivalent schematic of the modulator is shown on Figure 72. The sche-
matic includes part of the Timer/Counter units and the port B pin 7 output driver circuit.
Figure 72. Output Compare Modulator, Schematic

COM21 Vcc
COM20
COM1C1 Modulator
COM1C0
0
1
( From Waveform Generator ) D Q
1
OC1C
Pin
0
OC1C /
( From Waveform Generator ) D Q OC2 / PB7
OC2
D Q D Q
PORTB7 DDRB7
DATABUS
156 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
When the modulator is enabled the type of modulation (logical AND or OR) can be
selected by the PORTB7 Register. Note that the DDRB7 controls the direction of the
port independent of the COMnx1:0 bit setting.
Timing Example Figure 73 illustrates the modulator in action. In this example the Timer/Counter1 is set to
operate in fast PWM mode (non-inverted) and Timer/Counter2 uses CTC waveform
mode with toggle Compare Output mode (COMnx1:0 = 1).
Figure 73. Output Compare Modulator, Timing Diagram
clk I/O
OC1C
(FPWM Mode)
OC2
(CTC Mode)
PB7
(PORTB7 = 0)
PB7
(PORTB7 = 1)
1 2 3
(Period)
In this example, Timer/Counter2 provides the carrier, while the modulating signal is gen-
erated by the Output Compare unit C of the Timer/Counter1.
The resolution of the PWM signal (OC1C) is reduced by the modulation. The reduction
factor is equal to the number of system clock cycles of one period of the carrier (OC2).
In this example the resolution is reduced by a factor of two. The reason for the reduction
is illustrated in Figure 73 at the second and third period of the PB7 output when
PORTB7 equals zero. The period 2 high time is one cycle longer than the period 3 high
time, but the result on the PB7 output is equal in both periods.
157
2467E–AVR–05/02
Serial Peripheral The Serial Peripheral Interface (SPI) allows high-speed synchronous data transfer
between the ATmega128 and peripheral devices or between several AVR devices. The
Interface – SPI
ATmega128 SPI includes the following features:
• Full-duplex, Three-wire Synchronous Data Transfer
• Master or Slave Operation
• LSB First or MSB First Data Transfer
• Seven Programmable Bit Rates
• End of Transmission Interrupt Flag
• Write Collision Flag Protection
• Wake-up from Idle Mode
• Double Speed (CK/2) Master SPI Mode
Figure 74. SPI Block Diagram
DIVIDER
/2/4/8/16/32/64/128
SPI2X
SPI2X
Note: Refer to Figure 1 on page 2 and Table 30 on page 68 for SPI pin placement.
The interconnection between Master and Slave CPUs with SPI is shown in Figure 75.
The system consists of two Shift Registers, and a Master clock generator. The SPI Mas-
ter initiates the communication cycle when pulling low the Slave Select SS pin of the
desired Slave. Master and Slave prepare the data to be sent in their respective Shift
Registers, and the Master generates the required clock pulses on the SCK line to inter-
change data. Data is always shifted from Master to Slave on the Master Out – Slave In,
MOSI, line, and from Slave to Master on the Master In – Slave Out, MISO, line. After
each data packet, the Master will synchronize the Slave by pulling high the Slave Select,
SS, line.
158 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
When configured as a Master, the SPI interface has no automatic control of the SS line.
This must be handled by user software before communication can start. When this is
done, writing a byte to the SPI Data Register starts the SPI clock generator, and the
hardware shifts the 8 bits into the Slave. After shifting one byte, the SPI clock generator
stops, setting the end of transmission flag (SPIF). If the SPI interrupt enable bit (SPIE) in
the SPCR Register is set, an interrupt is requested. The Master may continue to shift the
next byte by writing it into SPDR, or signal the end of packet by pulling high the Slave
Select, SS line. The last incoming byte will be kept in the buffer register for later use.
When configured as a Slave, the SPI interface will remain sleeping with MISO tri-stated
as long as the SS pin is driven high. In this state, software may update the contents of
the SPI Data Register, SPDR, but the data will not be shifted out by incoming clock
pulses on the SCK pin until the SS pin is driven low. As one byte has been completely
shifted, the end of transmission flag, SPIF is set. If the SPI interrupt enable bit, SPIE, in
the SPCR Register is set, an interrupt is requested. The Slave may continue to place
new data to be sent into SPDR before reading the incoming data. The last incoming byte
will be kept in the buffer register for later use.
Figure 75. SPI Master-Slave Interconnection
SHIFT
ENABLE
The system is single buffered in the transmit direction and double buffered in the receive
direction. This means that bytes to be transmitted cannot be written to the SPI Data
Register before the entire shift cycle is completed. When receiving data, however, a
received character must be read from the SPI Data Register before the next character
has been completely shifted in. Otherwise, the first byte is lost.
In SPI Slave mode, the control logic will sample the incoming signal of the SCK pin. To
ensure correct sampling of the clock signal, the frequency of the SPI clock should never
exceed fosc/4.
When the SPI is enabled, the data direction of the MOSI, MISO, SCK, and SS pins is
overridden according to Table 69. For more details on automatic port overrides, refer to
“Alternate Port Functions” on page 65.
Table 69. SPI Pin Overrides(1)
Pin Direction, Master SPI Direction, Slave SPI
MOSI User Defined Input
MISO Input User Defined
SCK User Defined Input
SS User Defined Input
159
2467E–AVR–05/02
Note: 1. See “Alternate Functions of Port B” on page 68 for a detailed description of how to
define the direction of the user defined SPI pins.
The following code examples show how to initialize the SPI as a Master and how to per-
form a simple transmission. DDR_SPI in the examples must be replaced by the actual
data direction register controlling the SPI pins. DD_MOSI, DD_MISO and DD_SCK
must be replaced by the actual data direction bits for these pins. For example, if MOSI is
placed on pin PB5, replace DD_MOSI with DDB5 and DDR_SPI with DDRB.
SPI_MasterInit:
; Set MOSI and SCK output, all others input
ldi r17,(1<<DD_MOSI)|(1<<DD_SCK)
out DDR_SPI,r17
; Enable SPI, Master, set clock rate fck/16
ldi r17,(1<<SPE)|(1<<MSTR)|(1<<SPR0)
out SPCR,r17
ret
SPI_MasterTransmit:
; Start transmission of data (r16)
out SPDR,r16
Wait_Transmit:
; Wait for transmission complete
sbis SPSR,SPIF
rjmp Wait_Transmit
ret
C Code Example(1)
void SPI_MasterInit(void)
{
/* Set MOSI and SCK output, all others input */
DDR_SPI = (1<<DD_MOSI)|(1<<DD_SCK);
/* Enable SPI, Master, set clock rate fck/16 */
SPCR = (1<<SPE)|(1<<MSTR)|(1<<SPR0);
}
void SPI_MasterTransmit(char cData)

{
/* Start transmission */
SPDR = cData;
/* Wait for transmission complete */
while(!(SPSR & (1<<SPIF)))
;
}
160 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
The following code examples show how to initialize the SPI as a slave and how to per-
form a simple reception.
SPI_SlaveInit:
; Set MISO output, all others input
ldi r17,(1<<DD_MISO)
out DDR_SPI,r17
; Enable SPI
ldi r17,(1<<SPE)
out SPCR,r17
ret
SPI_SlaveReceive:
; Wait for reception complete
sbis SPSR,SPIF
rjmp SPI_SlaveReceive
; Read received data and return
in r16,SPDR
ret
C Code Example(1)
void SPI_SlaveInit(void)
{
/* Set MISO output, all others input */
DDR_SPI = (1<<DD_MISO);
/* Enable SPI */
SPCR = (1<<SPE);
}
char SPI_SlaveReceive(void)
{
/* Wait for reception complete */
while(!(SPSR & (1<<SPIF)))
;
/* Return data register */
return SPDR;
}
161
2467E–AVR–05/02
SS Pin Functionality
Slave Mode When the SPI is configured as a slave, the Slave Select (SS) pin is always input. When
SS is held low, the SPI is activated, and MISO becomes an output if configured so by
the user. All other pins are inputs. When SS is driven high, all pins are inputs, and the
SPI is passive, which means that it will not receive incoming data. Note that the SPI
logic will be reset once the SS pin is driven high.
The SS pin is useful for packet/byte synchronization to keep the slave bit counter syn-
chronous with the master clock generator. When the SS pin is driven high, the SPI slave
will immediately reset the send and receive logic, and drop any partially received data in
the Shift Register.
Master Mode When the SPI is configured as a master (MSTR in SPCR is set), the user can determine
the direction of the SS pin.
If SS is configured as an output, the pin is a general output pin which does not affect the
SPI system. Typically, the pin will be driving the SS pin of the SPI slave.
If SS is configured as an input, it must be held high to ensure Master SPI operation. If
the SS pin is driven low by peripheral circuitry when the SPI is configured as a master
with the SS pin defined as an input, the SPI system interprets this as another master
selecting the SPI as a slave and starting to send data to it. To avoid bus contention, the
SPI system takes the following actions:
1. The MSTR bit in SPCR is cleared and the SPI system becomes a slave. As a
result of the SPI becoming a slave, the MOSI and SCK pins become inputs.
2. The SPIF flag in SPSR is set, and if the SPI interrupt is enabled, and the I-bit in
SREG is set, the interrupt routine will be executed.
Thus, when interrupt-driven SPI transmission is used in master mode, and there exists a
possibility that SS is driven low, the interrupt should always check that the MSTR bit is
still set. If the MSTR bit has been cleared by a slave select, it must be set by the user to
re-enable SPI master mode.
SPI Control Register – SPCR

Bit 7 6 5 4 3 2 1 0
SPIE SPE DORD MSTR CPOL CPHA SPR1 SPR0 SPCR
• Bit 7 – SPIE: SPI Interrupt Enable

This bit causes the SPI interrupt to be executed if SPIF bit in the SPSR Register is set
and the if the global interrupt enable bit in SREG is set.
• Bit 6 – SPE: SPI Enable

When the SPE bit is written to one, the SPI is enabled. This bit must be set to enable
any SPI operations.
• Bit 5 – DORD: Data Order

When the DORD bit is written to one, the LSB of the data word is transmitted first.
When the DORD bit is written to zero, the MSB of the data word is transmitted first.
162 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
• Bit 4 – MSTR: Master/Slave Select

This bit selects Master SPI mode when written to one, and Slave SPI mode when written
logic zero. If SS is configured as an input and is driven low while MSTR is set, MSTR will
be cleared, and SPIF in SPSR will become set. The user will then have to set MSTR to
re-enable SPI Master mode.
• Bit 3 – CPOL: Clock Polarity

When this bit is written to one, SCK is high when idle. When CPOL is written to zero,
SCK is low when idle. Refer to Figure 76 and Figure 77 for an example. The CPOL func-
tionality is summarized below:
Table 70. CPOL functionality

CPOL Leading edge Trailing edge
0 Rising Falling
1 Falling Rising
• Bit 2 – CPHA: Clock Phase

The settings of the clock phase bit (CPHA) determine if data is sampled on the leading
(first) or trailing (last) edge of SCK. Refer to Figure 76 and Figure 77 for an example.
The CPOL functionality is summarized below:
Table 71. CPHA functionality

CPHA Leading edge Trailing edge
0 Sample Setup
1 Setup Sample
• Bits 1, 0 – SPR1, SPR0: SPI Clock Rate Select 1 and 0

These two bits control the SCK rate of the device configured as a master. SPR1 and
SPR0 have no effect on the slave. The relationship between SCK and the Oscillator
Clock frequency fosc is shown in the following table:
Table 72. Relationship Between SCK and the Oscillator Frequency

SPI2X SPR1 SPR0 SCK Frequency
0 0 0 fosc /4
0 0 1 fosc /16
0 1 0 fosc /64
0 1 1 fosc /128
1 0 0 fosc /2
1 0 1 fosc /8
1 1 0 fosc /32
1 1 1 fosc /64
163
2467E–AVR–05/02
SPI Status Register – SPSR
Bit 7 6 5 4 3 2 1 0
SPIF WCOL – – – – – SPI2X SPSR
Read/Write R R R R R R R R/W
• Bit 7 – SPIF: SPI Interrupt Flag

When a serial transfer is complete, the SPIF flag is set. An interrupt is generated if SPIE
in SPCR is set and global interrupts are enabled. If SS is an input and is driven low
when the SPI is in Master mode, this will also set the SPIF flag. SPIF is cleared by hard-
ware when executing the corresponding interrupt handling vector. Alternatively, the
SPIF bit is cleared by first reading the SPI Status Register with SPIF set, then accessing
the SPI Data Register (SPDR).
• Bit 6 – WCOL: Write COLlision flag

The WCOL bit is set if the SPI Data Register (SPDR) is written during a data transfer.
The WCOL bit (and the SPIF bit) are cleared by first reading the SPI Status Register
with WCOL set, and then accessing the SPI Data Register.
• Bit 5..1 – Res: Reserved Bits

• Bit 0 – SPI2X: Double SPI Speed Bit

When this bit is written logic one the SPI speed (SCK Frequency) will be doubled when
the SPI is in Master mode (see Table 72). This means that the minimum SCK period will
be 2 CPU clock periods. When the SPI is configured as Slave, the SPI is only guaran-
teed to work at fosc /4 or lower.
The SPI interface on the ATmega128 is also used for program memory and EEPROM
downloading or uploading. See page 297 for SPI Serial Programming and verification.
SPI Data Register – SPDR

Bit 7 6 5 4 3 2 1 0
MSB LSB SPDR
Initial Value X X X X X X X X Undefined
The SPI Data Register is a Read/Write Register used for data transfer between the reg-
ister file and the SPI Shift Register. Writing to the register initiates data transmission.
Reading the register causes the Shift Register Receive buffer to be read.
164 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Data Modes There are four combinations of SCK phase and polarity with respect to serial data,
which are determined by control bits CPHA and CPOL. The SPI data transfer formats
are shown in Figure 76 and Figure 77. Data bits are shifted out and latched in on oppo-
site edges of the SCK signal, ensuring sufficient time for data signals to stabilize. This is
clearly seen by summarizing Table 70 and Table 71, as done below:
Table 73. CPOL Functionality

Leading edge Trailing edge SPI mode
CPOL = 0, CPHA = 0 Sample (Rising) Setup (Falling) 0
CPOL = 0, CPHA = 1 Setup (Rising) Sample (Falling) 1
CPOL = 1, CPHA = 0 Sample (Falling) Setup (Rising) 2
CPOL = 1, CPHA = 1 Setup (Falling) Sample (Rising) 3
Figure 76. SPI Transfer Format with CPHA = 0

SCK (CPOL = 0)
mode 0
SCK (CPOL = 1)
mode 2
SAMPLE I
MOSI/MISO
CHANGE 0
MOSI PIN
CHANGE 0
MISO PIN
SS
MSB first (DORD = 0) MSB Bit 6 Bit 5 Bit 4 Bit 3 Bit 2 Bit 1 LSB
LSB first (DORD = 1) LSB Bit 1 Bit 2 Bit 3 Bit 4 Bit 5 Bit 6 MSB
Figure 77. SPI Transfer Format with CPHA = 1
SCK (CPOL = 0)
mode 1
SCK (CPOL = 1)
mode 3
SAMPLE I
MOSI/MISO
CHANGE 0
MOSI PIN
CHANGE 0
MISO PIN
SS
MSB first (DORD = 0) MSB Bit 6 Bit 5 Bit 4 Bit 3 Bit 2 Bit 1 LSB
LSB first (DORD = 1) LSB Bit 1 Bit 2 Bit 3 Bit 4 Bit 5 Bit 6 MSB
165
2467E–AVR–05/02
USART The Universal Synchronous and Asynchronous serial Receiver and Transmitter
(USART) is a highly flexible serial communication device. The main features are:
• Full Duplex Operation (Independent Serial Receive and Transmit Registers)
• Asynchronous or Synchronous Operation
• Master or Slave Clocked Synchronous Operation
• High Resolution Baud Rate Generator
• Supports Serial Frames with 5, 6, 7, 8, or 9 Data Bits and 1 or 2 Stop Bits
• Odd or Even Parity Generation and Parity Check Supported by Hardware
• Data OverRun Detection
• Framing Error Detection
• Noise Filtering Includes False Start Bit Detection and Digital Low Pass Filter
• Three Separate Interrupts on TX Complete, TX Data Register Empty, and RX Complete
• Multi-processor Communication Mode
• Double Speed Asynchronous Communication Mode
Dual USART The ATmega128 has two USART’s, USART0 and USART1. The functionality for both
USART’s is described below. USART0 and USART1 have different I/O registers as
shown in “Register Summary” on page 350. Note that in ATmega103 compatibility
mode, USART1 is not available, neither is the UBRR0H or UCRS0C Registers. This
means that in ATmega103 compatibility mode, the ATmega128 supports asynchronous
operation of USART0 only.
Overview A simplified block diagram of the USART transmitter is shown in Figure 78. CPU acces-
sible I/O registers and I/O pins are shown in bold.
166 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 78. USART Block Diagram
Clock Generator
UBRR[H:L]
OSC
BAUD RATE GENERATOR
SYNC LOGIC PIN

XCK
CONTROL
Transmitter
TX
UDR (Transmit)
CONTROL
PARITY
GENERATOR
DATABUS
PIN
TRANSMIT SHIFT REGISTER TxD
CONTROL
Receiver
CLOCK RX
RECOVERY CONTROL
DATA PIN
RECEIVE SHIFT REGISTER RxD
RECOVERY CONTROL
PARITY
UDR (Receive)
CHECKER
UCSRA UCSRB UCSRC
Note: Refer to Figure 1 on page 2, Table 36 on page 72, and Table 39 on page 75 for USART
pin placement.
The dashed boxes in the block diagram separate the three main parts of the USART
(listed from the top): Clock Generator, Transmitter, and Receiver. Control registers are
shared by all units. The clock generation logic consists of synchronization logic for exter-
nal clock input used by synchronous slave operation, and the baud rate generator. The
XCK (Transfer Clock) pin is only used by Synchronous Transfer mode. The Transmitter
consists of a single write buffer, a serial Shift Register, parity generator and control logic
for handling different serial frame formats. The write buffer allows a continuous transfer
of data without any delay between frames. The Receiver is the most complex part of the
USART module due to its clock and data recovery units. The recovery units are used for
asynchronous data reception. In addition to the recovery units, the receiver includes a
parity checker, control logic, a Shift Register and a two level receive buffer (UDR). The
receiver supports the same frame formats as the Transmitter, and can detect frame
error, data overrun and parity errors.
167
2467E–AVR–05/02
AVR USART vs. AVR UART – The USART is fully compatible with the AVR UART regarding:
Compatibility • Bit locations inside all USART registers
• Baud Rate Generation
• Transmitter Operation
• Transmit Buffer Functionality
• Receiver Operation
However, the receive buffering has two improvements that will affect the compatibility in
some special cases:
• A second buffer register has been added. The two buffer registers operate as a
circular FIFO buffer. Therefore the UDR must only be read once for each incoming
data! More important is the fact that the error flags (FE and DOR) and the ninth data
bit (RXB8) are buffered with the data in the receive buffer. Therefore the status bits
must always be read before the UDR Register is read. Otherwise the error status
will be lost since the buffer state is lost.
• The receiver Shift Register can now act as a third buffer level. This is done by
allowing the received data to remain in the serial Shift Register (see Figure 78) if the
buffer registers are full, until a new start bit is detected. The USART is therefore
more resistant to Data OverRun (DOR) error conditions.
The following control bits have changed name, but have same functionality and register
location:
• CHR9 is changed to UCSZ2
• OR is changed to DOR
Clock Generation The clock generation logic generates the base clock for the transmitter and receiver.
The USART supports four modes of clock operation: Normal Asynchronous, Double
Speed Asynchronous, Master Synchronous, and Slave Synchronous mode. The
UMSEL bit in USART Control and Status Register C (UCSRC) selects between asyn-
chronous and synchronous operation. Double speed (Asynchronous mode only) is
controlled by the U2X found in the UCSRA Register. When using Synchronous mode
(UMSEL = 1), the Data Direction Register for the XCK pin (DDR_XCK) controls whether
the clock source is internal (Master mode) or external (Slave mode). The XCK pin is only
active when using Synchronous mode.
Figure 79 shows a block diagram of the clock generation logic.
Figure 79. Clock Generation Logic, Block Diagram

UBRR
U2X
fosc
Prescaling UBRR+1
/2 /4 /2
Down-Counter 0
1
0
OSC txclk
1
DDR_XCK
Sync Edge
xcki Register Detector 0
XCK UMSEL
xcko 1
Pin
DDR_XCK UCPOL 1
rxclk
0
168 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Signal description:
txclk Transmitter clock. (Internal Signal)
rxclk Receiver base clock. (Internal Signal)
xcki Input from XCK pin (internal Signal). Used for synchronous slave operation.
xcko Clock output to XCK pin (Internal Signal). Used for synchronous master
operation.
fosc XTAL pin frequency (System Clock).
Internal Clock Generation – Internal clock generation is used for the asynchronous and the synchronous master
The Baud Rate Generator modes of operation. The description in this section refers to Figure 79.
The USART Baud Rate Register (UBRR) and the down-counter connected to it function
as a programmable prescaler or baud rate generator. The down-counter, running at sys-
tem clock (fosc), is loaded with the UBRR value each time the counter has counted
down to zero or when the UBRRL Register is written. A clock is generated each time the
cou nter re aches zero. This clock is the b aud rate gen erator clock output (=
fosc/(UBRR+1)). The transmitter divides the baud rate generator clock output by 2, 8, or
16 depending on mode. The baud rate generator output is used directly by the receiver’s
clock and data recovery units. However, the recovery units use a state machine that
uses 2, 8, or 16 states depending on mode set by the state of the UMSEL, U2X and
DDR_XCK bits.
Table 74 contains equations for calculating the baud rate (in bits per second) and for
calculating the UBRR value for each mode of operation using an internally generated
clock source.
Table 74. Equations for Calculating Baud Rate Register Setting

Equation for Calculating Equation for Calculating
Operating Mode Baud Rate(1) UBRR Value
Asynchronous Normal Mode f OSC f OS C
(U2X = 0) BAUD = --------------------------------------
- -–1
UBRR = -----------------------
16 ( UBRR + 1 ) 16BAUD
Asynchronous Double Speed f OSC f OSC
Mode (U2X = 1) BAUD = ----------------------------------
- UBRR = -------------------- –1
8 ( UBRR + 1 ) 8BAUD
Synchronous Master Mode f OSC f OSC
BAUD = ----------------------------------
- UBRR = -------------------- –1
2 ( UBRR + 1 ) 2BAUD
Note: 1. The baud rate is defined to be the transfer rate in bit per second (bps).
BAUD Baud rate (in bits per second, bps)
fOSC System Oscillator clock frequency
UBRR Contents of the UBRRH and UBRRL Registers, (0 - 4095)
Some examples of UBRR values for some system clock frequencies are found in Table
82 (see page 189).
Double Speed Operation The transfer rate can be doubled by setting the U2X bit in UCSRA. Setting this bit only
(U2X) has effect for the asynchronous operation. Set this bit to zero when using synchronous
operation.
Setting this bit will reduce the divisor of the baud rate divider from 16 to 8, effectively
doubling the transfer rate for asynchronous communication. Note however that the
169
2467E–AVR–05/02
receiver will in this case only use half the number of samples (reduced from 16 to 8) for
data sampling and clock recovery, and therefore a more accurate baud rate setting and
system clock are required when this mode is used. For the Transmitter, there are no
downsides.
External Clock External clocking is used by the synchronous slave modes of operation. The description
in this section refers to Figure 79 for details.
External clock input from the XCK pin is sampled by a synchronization register to mini-
mize the chance of meta-stability. The output from the synchronization register must
then pass through an edge detector before it can be used by the transmitter and
receiver. This process introduces a two CPU clock period delay and therefore the maxi-
mum external XCK clock frequency is limited by the following equation:
f OS C
f XCK < -----------
4
Note that fosc depends on the stability of the system clock source. It is therefore recom-
mended to add some margin to avoid possible loss of data due to frequency variations.
Synchronous Clock Operation When Synchronous mode is used (UMSEL = 1), the XCK pin will be used as either clock
input (slave) or clock output (master). The dependency between the clock edges and
data sampling or data change is the same. The basic principle is that data input (on
RxD) is sampled at the opposite XCK clock edge of the edge the data output (TxD) is
changed.
Figure 80. Synchronous Mode XCK Timing.
UCPOL = 1 XCK
RxD / TxD
Sample
UCPOL = 0 XCK
RxD / TxD
Sample
The UCPOL bit UCRSC selects which XCK clock edge is used for data sampling and
which is used for data change. As Figure 80 shows, when UCPOL is zero the data will
be changed at rising XCK edge and sampled at falling XCK edge. If UCPOL is set, the
data will be changed at falling XCK edge and sampled at rising XCK edge.
170 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Frame Formats A serial frame is defined to be one character of data bits with synchronization bits (start
and stop bits), and optionally a parity bit for error checking. The USART accepts all 30
combinations of the following as valid frame formats:
• 1 start bit
• 5, 6, 7, 8, or 9 data bits
• no, even or odd parity bit
• 1 or 2 stop bits
A frame starts with the start bit followed by the least significant data bit. Then the next
data bits, up to a total of nine, are succeeding, ending with the most significant bit. If
enabled, the parity bit is inserted after the data bits, before the stop bits. When a com-
plete frame is transmitted, it can be directly followed by a new frame, or the
communication line can be set to an idle (high) state. Figure 81 illustrates the possible
combinations of the frame formats. Bits inside brackets are optional.
Figure 81. Frame Formats

FRAME
(IDLE) St 0 1 2 3 4 [5] [6] [7] [8] [P] Sp1 [Sp2] (St / IDLE)
St Start bit, always low.

(n) Data bits (0 to 8).
P Parity bit. Can be odd or even.
Sp Stop bit, always high.
IDLE No transfers on the communication line (RxD or TxD). An IDLE line must be
high.
The frame format used by the USART is set by the UCSZ2:0, UPM1:0 and USBS bits in
UCSRB and UCSRC. The receiver and transmitter use the same setting. Note that
changing the setting of any of these bits will corrupt all ongoing communication for both
the receiver and transmitter.
The USART Character SiZe (UCSZ2:0) bits select the number of data bits in the frame.
The USART Parity mode (UPM1:0) bits enable and set the type of parity bit. The selec-
tion between one or two stop bits is done by the USART Stop Bit Select (USBS) bit. The
receiver ignores the second stop bit. An FE (Frame Error) will therefore only be detected
in the cases where the first stop bit is zero.
Parity Bit Calculation The parity bit is calculated by doing an exclusive-or of all the data bits. If odd parity is
used, the result of the exclusive or is inverted. The relation between the parity bit and
data bits is as follows::
P ev en = d n – 1 ⊕ … ⊕ d 3 ⊕ d 2 ⊕ d 1 ⊕ d 0 ⊕ 0
P odd = d n – 1 ⊕ … ⊕ d 3 ⊕ d 2 ⊕ d 1 ⊕ d 0 ⊕ 1
Peven Parity bit using even parity

Podd Parity bit using odd parity
dn Data bit n of the character
171
2467E–AVR–05/02
If used, the parity bit is located between the last data bit and first stop bit of a serial
frame.
USART Initialization The USART has to be initialized before any communication can take place. The initial-
ization process normally consists of setting the baud rate, setting frame format and
enabling the Transmitter or the Receiver depending on the usage. For interrupt driven
USART operation, the global interrupt flag should be cleared (and interrupts globally dis-
abled) when doing the initialization.
Before doing a re-initialization with changed baud rate or frame format, be sure that
there are no ongoing transmissions during the period the registers are changed. The
TXC flag can be used to check that the Transmitter has completed all transfers, and the
RXC flag can be used to check that there are no unread data in the receive buffer. Note
that the TXC flag must be cleared before each transmission (before UDR is written) if it
is used for this purpose.
The following simple USART initialization code examples show one assembly and one
C function that are equal in functionality. The examples assume asynchronous opera-
tion using polling (no interrupts enabled) and a fixed frame format. The baud rate is
given as a function parameter. For the assembly code, the baud rate parameter is
assumed to be stored in the r17:r16 registers.
USART_Init:
; Set baud rate
out UBRRH, r17
out UBRRL, r16
; Enable receiver and transmitter
ldi r16, (1<<RXEN)|(1<<TXEN)
out UCSRB,r16
; Set frame format: 8data, 2stop bit
ldi r16, (1<<USBS)|(3<<UCSZ0)
out UCSRC,r16
ret
C Code Example(1)
void USART_Init( unsigned int baud )
{
/* Set baud rate */
UBRRH = (unsigned char)(baud>>8);
UBRRL = (unsigned char)baud;
/* Enable receiver and transmitter */
UCSRB = (1<<RXEN)|(1<<TXEN);
/* Set frame format: 8data, 2stop bit */
UCSRC = (1<<USBS)|(3<<UCSZ0);
}
More advanced initialization routines can be made that include frame format as parame-
ters, disable interrupts and so on. However, many applications use a fixed setting of the
172 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
baud and control registers, and for these types of applications the initialization code can
be placed directly in the main routine, or be combined with initialization code for other
I/O modules.
Data Transmission – The The USART Transmitter is enabled by setting the Transmit Enable (TXEN) bit in the
USART Transmitter UCSRB Register. When the Transmitter is enabled, the normal port operation of the
TxD pin is overridden by the USART and given the function as the transmitter’s serial
output. The baud rate, mode of operation and frame format must be set up once before
doing any transmissions. If synchronous operation is used, the clock on the XCK pin will
be overridden and used as transmission clock.
Sending Frames with 5 to 8 A data transmission is initiated by loading the transmit buffer with the data to be trans-
Data Bit mitted. The CPU can load the transmit buffer by writing to the UDR I/O location. The
buffered data in the transmit buffer will be moved to the Shift Register when the Shift
Register is ready to send a new frame. The Shift Register is loaded with new data if it is
in idle state (no ongoing transmission) or immediately after the last stop bit of the previ-
ous frame is transmitted. When the Shift Register is loaded with new data, it will transfer
one complete frame at the rate given by the baud register, U2X bit or by XCK depending
on mode of operation.
The following code examples show a simple USART transmit function based on polling
of the Data Register Empty (UDRE) flag. When using frames with less than eight bits,
the most significant bits written to the UDR are ignored. The USART has to be initialized
before the function can be used. For the assembly code, the data to be sent is assumed
to be stored in Register R16
USART_Transmit:
; Wait for empty transmit buffer
sbis UCSRA,UDRE
rjmp USART_Transmit
; Put data (r16) into buffer, sends the data
out UDR,r16
ret
C Code Example(1)
void USART_Transmit( unsigned char data )
{
/* Wait for empty transmit buffer */
while ( !( UCSRA & (1<<UDRE)) )
;
/* Put data into buffer, sends the data */
UDR = data;
}
The function simply waits for the transmit buffer to be empty by checking the UDRE flag,
before loading it with new data to be transmitted. If the data register empty interrupt is
utilized, the interrupt routine writes the data into the buffer.
173
2467E–AVR–05/02
Sending Frames with 9 Data If 9-bit characters are used (UCSZ = 7), the ninth bit must be written to the TXB8 bit in
Bit UCSRB before the low byte of the character is written to UDR. The following code
examples show a transmit function that handles 9 bit characters. For the assembly
code, the data to be sent is assumed to be stored in Registers R17:R16.
USART_Transmit:
; Wait for empty transmit buffer
sbis UCSRA,UDRE
rjmp USART_Transmit
; Copy 9th bit from r17 to TXB8
cbi UCSRB,TXB8
sbrc r17,0
sbi UCSRB,TXB8
; Put LSB data (r16) into buffer, sends the data
out UDR,r16
ret
C Code Example
void USART_Transmit( unsigned int data )
{
/* Wait for empty transmit buffer */
while ( !( UCSRA & (1<<UDRE))) )
;
/* Copy 9th bit to TXB8 */
UCSRB &= ~(1<<TXB8);
if ( data & 0x0100 )
UCSRB |= (1<<TXB8);
/* Put data into buffer, sends the data */
UDR = data;
}
Note: 1. These transmit functions are written to be general functions. They can be optimized if
the contents of the UCSRB is static. I.e., only the TXB8 bit of the UCSRB Register is
used after initialization.
The ninth bit can be used for indicating an address frame when using multi processor
communication mode or for other protocol handling as for example synchronization.
Transmitter Flags and The USART Transmitter has two flags that indicate its state: USART Data Register
Interrupts Empty (UDRE) and Transmit Complete (TXC). Both flags can be used for generating
interrupts.
The Data Register Empty (UDRE) flag indicates whether the transmit buffer is ready to
receive new data. This bit is set when the transmit buffer is empty, and cleared when the
transmit buffer contains data to be transmitted that has not yet been moved into the Shift
Register. For compatibility with future devices, always write this bit to zero when writing
the UCSRA Register.
When the Data Register empty Interrupt Enable (UDRIE) bit in UCSRB is written to one,
the USART Data Register Empty Interrupt will be executed as long as UDRE is set (pro-
vided that global interrupts are enabled). UDRE is cleared by writing UDR. When
174 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
interrupt-driven data transmission is used, the data register empty Interrupt routine must
either write new data to UDR in order to clear UDRE or disable the data register empty
interrupt, otherwise a new interrupt will occur once the interrupt routine terminates.
The Transmit Complete (TXC) flag bit is set one when the entire frame in the Transmit
Shift Register has been shifted out and there are no new data currently present in the
transmit buffer. The TXC flag bit is automatically cleared when a transmit complete inter-
rupt is executed, or it can be cleared by writing a one to its bit location. The TXC flag is
useful in half-duplex communication interfaces (like the RS485 standard), where a
transmitting application must enter receive mode and free the communication bus
immediately after completing the transmission.
When the Transmit Compete Interrupt Enable (TXCIE) bit in UCSRB is set, the USART
Transmit Complete Interrupt will be executed when the TXC flag becomes set (provided
that global interrupts are enabled). When the transmit complete interrupt is used, the
interrupt handling routine does not have to clear the TXC flag, this is done automatically
when the interrupt is executed.
Parity Generator The parity generator calculates the parity bit for the serial frame data. When parity bit is
enabled (UPM1 = 1), the transmitter control logic inserts the parity bit between the last
data bit and the first stop bit of the frame that is sent.
Disabling the Transmitter The disabling of the Transmitter (setting the TXEN to zero) will not become effective
until ongoing and pending transmissions are completed, i.e., when the Transmit Shift
Register and Transmit Buffer register do not contain data to be transmitted. When dis-
abled, the Transmitter will no longer override the TxD pin.
Data Reception – The The USART Receiver is enabled by writing the Receive Enable (RXEN) bit in the
USART Receiver UCSRB Register to one. When the receiver is enabled, the normal pin operation of the
RxD pin is overridden by the USART and given the function as the receiver’s serial
input. The baud rate, mode of operation and frame format must be set up once before
any serial reception can be done. If synchronous operation is used, the clock on the
XCK pin will be used as transfer clock.
Receiving Frames with 5 to 8 The Receiver starts data reception when it detects a valid start bit. Each bit that follows
Data Bits the start bit will be sampled at the baud rate or XCK clock, and shifted into the Receive
Shift Register until the first stop bit of a frame is received. A second stop bit will be
ignored by the receiver. When the first stop bit is received, i.e., a complete serial frame
is present in the Receive Shift Register, the contents of the Shift Register will be moved
into the receive buffer. The receive buffer can then be read by reading the UDR I/O
location.
175
2467E–AVR–05/02
The following code example shows a simple USART receive function based on polling
of the Receive Complete (RXC) flag. When using frames with less than eight bits the
most significant bits of the data read from the UDR will be masked to zero. The USART
has to be initialized before the function can be used.
USART_Receive:
; Wait for data to be received
sbis UCSRA, RXC
rjmp USART_Receive
; Get and return received data from buffer
in r16, UDR
ret
C Code Example(1)
unsigned char USART_Receive( void )
{
/* Wait for data to be received */
while ( !(UCSRA & (1<<RXC)) )
;
/* Get and return received data from buffer */
return UDR;
}
The function simply waits for data to be present in the receive buffer by checking the
RXC flag, before reading the buffer and returning the value.
176 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Receiving Frames with 9 Data If 9-bit characters are used (UCSZ=7) the ninth bit must be read from the RXB8 bit in
Bits UCSRB before reading the low bits from the UDR. This rule applies to the FE, DOR and
UPE status flags as well. Read status from UCSRA, then data from UDR. Reading the
UDR I/O location will change the state of the receive buffer FIFO and consequently the
TXB8, FE, DOR, and UPE bits, which all are stored in the FIFO, will change.
The following code example shows a simple USART receive function that handles both
nine bit characters and the status bits.
USART_Receive:
; Wait for data to be received
sbis UCSRA, RXC
rjmp USART_Receive
; Get status and 9th bit, then data from buffer
in r18, UCSRA
in r17, UCSRB
in r16, UDR
; If error, return -1
andi r18,(1<<FE)|(1<<DOR)|(1<<UPE)
breq USART_ReceiveNoError
ldi r17, HIGH(-1)
ldi r16, LOW(-1)
USART_ReceiveNoError:
; Filter the 9th bit, then return
lsr r17
andi r17, 0x01
ret
C Code Example(1)
unsigned int USART_Receive( void )
{
unsigned char status, resh, resl;
/* Wait for data to be received */
while ( !(UCSRA & (1<<RXC)) )
;
/* Get status and 9th bit, then data */
/* from buffer */
status = UCSRA;
resh = UCSRB;
resl = UDR;
/* If error, return -1 */
if ( status & (1<<FE)|(1<<DOR)|(1<<UPE) )
return -1;
/* Filter the 9th bit, then return */
resh = (resh >> 1) & 0x01;
return ((resh << 8) | resl);
}
177
2467E–AVR–05/02
The receive function example reads all the I/O registers into the register file before any
computation is done. This gives an optimal receive buffer utilization since the buffer
location read will be free to accept new data as early as possible.
Receive Compete Flag and The USART Receiver has one flag that indicates the receiver state.
Interrupt
The Receive Complete (RXC) flag indicates if there are unread data present in the
receive buffer. This flag is one when unread data exist in the receive buffer, and zero
when the receive buffer is empty (i.e., does not contain any unread data). If the Receiver
is disabled (RXEN = 0), the receive buffer will be flushed and consequently the RXC bit
will become zero.
When the Receive Complete Interrupt Enable (RXCIE) in UCSRB is set, the USART
Receive Complete Interrupt will be executed as long as the RXC flag is set (provided
that global interrupts are enabled). When interrupt-driven data reception is used, the
receive complete routine must read the received data from UDR in order to clear the
RXC flag, otherwise a new interrupt will occur once the interrupt routine terminates.
Receiver Error Flags The USART receiver has three error flags: Frame Error (FE), Data OverRun (DOR) and
Parity Error (UPE). All can be accessed by reading UCSRA. Common for the error flags
is that they are located in the receive buffer together with the frame for which they indi-
cate the error status. Due to the buffering of the error flags, the UCSRA must be read
before the receive buffer (UDR), since reading the UDR I/O location changes the buffer
read location. Another equality for the error flags is that they can not be altered by soft-
ware doing a write to the flag location. However, all flags must be set to zero when the
UCSRA is written for upward compatibility of future USART implementations. None of
the error flags can generate interrupts.
The Frame Error (FE) flag indicates the state of the first stop bit of the next readable
frame stored in the receive buffer. The FE flag is zero when the stop bit was correctly
read (as one), and the FE flag will be one when the stop bit was incorrect (zero). This
flag can be used for detecting out-of-sync conditions, detecting break conditions and
protocol handling. The FE flag is not affected by the setting of the USBS bit in UCSRC
since the receiver ignores all, except for the first, stop bits. For compatibility with future
devices, always set this bit to zero when writing to UCSRA.
The Data OverRun (DOR) flag indicates data loss due to a receiver buffer full condition.
A data overrun occurs when the receive buffer is full (two characters), it is a new charac-
ter waiting in the Receive Shift Register, and a new start bit is detected. If the DOR flag
is set there was one or more serial frame lost between the frame last read from UDR,
and the next frame read from UDR. For compatibility with future devices, always write
this bit to zero when writing to UCSRA. The DOR flag is cleared when the frame
received was successfully moved from the Shift Register to the receive buffer.
The Parity Error (UPE) flag indicates that the next frame in the receive buffer had a par-
ity error when received. If parity check is not enabled the UPE bit will always be read
zero. For compatibility with future devices, always set this bit to zero when writing to
UCSRA. For more details see “Parity Bit Calculation” on page 171 and “Parity Checker”
on page 178.
Parity Checker The parity checker is active when the high USART Parity mode (UPM1) bit is set. Type
of parity check to be performed (odd or even) is selected by the UPM0 bit. When
enabled, the parity checker calculates the parity of the data bits in incoming frames and
compares the result with the parity bit from the serial frame. The result of the check is
178 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
stored in the receive buffer together with the received data and stop bits. The Parity
Error (UPE) flag can then be read by software to check if the frame had a Parity Error.
The UPE bit is set if the next character that can be read from the receive buffer had a
parIty Error when received and the parity checking was enabled at that point (UPM1 =
1). This bit is valid until the Receive buffer (UDR) is read.
Disabling the Receiver In contrast to the Transmitter, disabling of the Receiver will be immediate. Data from
ongoing receptions will therefore be lost. When disabled (i.e., the RXEN is set to zero)
the receiver will no longer override the normal function of the RxD port pin. The receiver
buffer FIFO will be flushed when the receiver is disabled. Remaining data in the buffer
will be lost
Flushing the Receive Buffer The receiver buffer FIFO will be flushed when the receiver is disabled, i.e. the buffer will
be emptied of its contents. Unread data will be lost. If the buffer has to be flushed during
normal operation, due to for instance an error condition, read the UDR I/O location until
the RXC flag is cleared. The following code example shows how to flush the receive
buffer.
USART_Flush:
sbis UCSRA, RXC
ret
in r16, UDR
rjmp USART_Flush
C Code Example(1)
void USART_Flush( void )
{
unsigned char dummy;
while ( UCSRA & (1<<RXC) ) dummy = UDR;
}
Asynchronous Data The USART includes a clock recovery and a data recovery unit for handling asynchro-
Reception nous data reception. The clock recovery logic is used for synchronizing the internally
generated baud rate clock to the incoming asynchronous serial frames at the RxD pin.
The data recovery logic samples and low pass filters each incoming bit, thereby improv-
ing the noise immunity of the receiver. The asynchronous reception operational range
depends on the accuracy of the internal baud rate clock, the rate of the incoming
frames, and the frame size in number of bits.
Asynchronous Clock The clock recovery logic synchronizes internal clock to the incoming serial frames. Fig-
Recovery ure 82 illustrates the sampling process of the start bit of an incoming frame. The sample
rate is 16 times the baud rate for normal mode, and 8 times the baud rate for Double
Speed mode. The horizontal arrows illustrate the synchronization variation due to the
sampling process. Note the larger time variation when using the double speed mode
(U2X = 1) of operation. Samples denoted zero are samples done when the RxD line is
idle (i.e., no communication activity).
179
2467E–AVR–05/02
Figure 82. Start Bit Sampling
RxD IDLE START BIT 0
Sample
(U2X = 0) 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 2 3
Sample
(U2X = 1) 0 1 2 3 4 5 6 7 8 1 2
When the clock recovery logic detects a high (idle) to low (start) transition on the RxD
line, the start bit detection sequence is initiated. Let sample 1 denote the first zero-sam-
ple as shown in the figure. The clock recovery logic then uses samples 8, 9, and 10 for
normal mode, and samples 4, 5, and 6 for Double Speed mode (indicated with sample
numbers inside boxes on the figure), to decide if a valid start bit is received. If two or
more of these three samples have logical high levels (the majority wins), the start bit is
rejected as a noise spike and the receiver starts looking for the next high to low-transi-
tion. If however, a valid start bit is detected, the clock recovery logic is synchronized and
the data recovery can begin. The synchronization process is repeated for each start bit.
Asynchronous Data Recovery When the receiver clock is synchronized to the start bit, the data recovery can begin.
The data recovery unit uses a state machine that has 16 states for each bit in normal
mode and 8 states for each bit in Double Speed mode. Figure 83 shows the sampling of
the data bits and the parity bit. Each of the samples is given a number that is equal to
the state of the recovery unit.
Figure 83. Sampling of Data and Parity Bit
RxD BIT n
Sample
(U2X = 0) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1
Sample
(U2X = 1) 1 2 3 4 5 6 7 8 1
The decision of the logic level of the received bit is taken by doing a majority voting of
the logic value to the three samples in the center of the received bit. The center samples
are emphasized on the figure by having the sample number inside boxes. The majority
voting process is done as follows: If two or all three samples have high levels, the
received bit is registered to be a logic 1. If two or all three samples have low levels, the
received bit is registered to be a logic 0. This majority voting process acts as a low pass
filter for the incoming signal on the RxD pin. The recovery process is then repeated until
a complete frame is received. Including the first stop bit. Note that the receiver only uses
the first stop bit of a frame. Figure 84 shows the sampling of the stop bit and the earliest
possible beginning of the start bit of the next frame.
180 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 84. Stop Bit Sampling and Next Start Bit Sampling
RxD STOP 1 (A) (B) (C)
Sample
(U2X = 0) 1 2 3 4 5 6 7 8 9 10 0/1 0/1 0/1
Sample
(U2X = 1) 1 2 3 4 5 6 0/1
The same majority voting is done to the stop bit as done for the other bits in the frame. If
the stop bit is registered to have a logic 0 value, the Frame Error (FE) flag will be set.
A new high to low transition indicating the start bit of a new frame can come right after
the last of the bits used for majority voting. For Normal Speed mode, the first low level
sample can be at point marked (A) in Figure 84. For Double Speed mode the first low
level must be delayed to (B). (C) marks a stop bit of full length. The early start bit detec-
tion influences the operational range of the receiver.
Asynchronous Operational The operational range of the Receiver is dependent on the mismatch between the
Range received bit rate and the internally generated baud rate. If the Transmitter is sending
frames at too fast or too slow bit rates, or the internally generated baud rate of the
receiver does not have a similar (see Table 75) base frequency, the Receiver will not be
able to synchronize the frames to the start bit.
The following equations can be used to calculate the ratio of the incoming data rate and
internal receiver baud rate.
( D + 1 )S ( D + 2 )S
R fast = -----------------------------------
R slow = -------------------------------------------
S – 1 + D ⋅ S + SF ( D + 1 )S + SM
D Sum of character size and parity size (D = 5 to 10-bit)

S Samples per bit. S = 16 for Normal Speed mode and S = 8 for Double Speed
mode.
SF First sample number used for majority voting. SF = 8 for Normal Speed and SF = 4
for Double Speed mode.
SM Middle sample number used for majority voting. SM = 9 for Normal Speed and
SM = 5 for Double Speed mode.
Rslow is the ratio of the slowest incoming data rate that can be accepted in relation to the
receiver baud rate. R fast is the ratio of the fastest incoming data rate that can be
accepted in relation to the receiver baud rate.
Table 75 and Table 76 list the maximum receiver baud rate error that can be tolerated.
Note that normal speed mode has higher toleration of baud rate variations.
181
2467E–AVR–05/02
Table 75. Recommended Maximum Receiver Baud Rate Error for Normal Speed Mode
(U2X = 0)
D Max Total Recommended Max
# (Data+Parity Bit) Rslow % Rfast % Error % Receiver Error %
5 93,20 106,67 +6.67/-6.8 ± 3.0
6 94,12 105,79 +5.79/-5.88 ± 2.5
7 94,81 105,11 +5.11/-5.19 ± 2.0
8 95,36 104,58 +4.58/-4.54 ± 2.0
9 95,81 104,14 +4.14/-4.19 ± 1.5
10 96,17 103,78 % +3.78/-3.83 ± 1.5
Table 76. Recommended Maximum Receiver Baud Rate Error for Double Speed Mode
(U2X = 1)
D Max Total Recommended Max
# (Data+Parity Bit) Rslow % Rfast % Error % Receiver Error %
5 94,12 105,66 +5.66/-5.88 ± 2.5
6 94,92 104,92 +4.92/-5.08 ± 2.0
7 95,52 104,35 +4.35/-4.48 ± 1.5
8 96,00 103,90 +3.90/-4.00 ± 1.5
9 96,39 103,53 +3.53/-3.61 ± 1.5
10 96,70 103,23 +3.23/-3.30 ± 1.0
The recommendations of the maximum receiver baud rate error was made under the
assumption that the receiver and transmitter equally divides the maximum total error.
There are two possible sources for the receivers baud rate error. The receiver’s system
clock (XTAL) will always have some minor instability over the supply voltage range and
the temperature range. When using a crystal to generate the system clock, this is rarely
a problem, but for a resonator the system clock may differ more than 2% depending of
the resonators tolerance. The second source for the error is more controllable. The baud
rate generator can not always do an exact division of the system frequency to get the
baud rate wanted. In this case an UBRR value that gives an acceptable low error can be
used if possible.
Multi-processor Setting the Multi-processor Communication mode (MPCM) bit in UCSRA enables a fil-
Communication Mode tering function of incoming frames received by the USART receiver. Frames that do not
contain address information will be ignored and not put into the receive buffer. This
effectively reduces the number of incoming frames that has to be handled by the CPU,
in a system with multiple MCUs that communicate via the same serial bus. The transmit-
ter is unaffected by the MPCM setting, but has to be used differently when it is a part of
a system utilizing the Multi-processor Communication mode.
If the receiver is set up to receive frames that contain 5 to 8 data bits, then the first stop
bit indicates if the frame contains data or address information. If the receiver is set up for
frames with 9 data bits, then the ninth bit (RXB8) is used for identifying address and
182 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
data frames. When the frame type bit (the first stop or the 9th bit) is one, the frame con-
tains an address. When the frame type bit is zero the frame is a data frame.
The Multi-processor Communication mode enables several slave MCUs to receive data
from a master MCU. This is done by first decoding an address frame to find out which
MCU has been addressed. If a particular slave MCU has been addressed, it will receive
the following data frames as normal, while the other slave MCUs will ignore the received
frames until another address frame is received.
Using MPCM For an MCU to act as a master MCU, it can use a 9-bit character frame format (UCSZ =
7). The ninth bit (TXB8) must be set when an address frame (TXB8 = 1) or cleared when
a data frame (TXB = 0) is being transmitted. The slave MCUs must in this case be set to
use a 9-bit character frame format.
The following procedure should be used to exchange data in Multi-processor Communi-
cation mode:
1. All slave MCUs are in Multi-processor Communication mode (MPCM in UCSRA
is set).
2. The master MCU sends an address frame, and all slaves receive and read this
frame. In the slave MCUs, the RXC flag in UCSRA will be set as normal.
3. Each slave MCU reads the UDR Register and determines if it has been selected.
If so, it clears the MPCM bit in UCSRA, otherwise it waits for the next address
byte and keeps the MPCM setting.
4. The addressed MCU will receive all data frames until a new address frame is
received. The other slave MCUs, which still have the MPCM bit set, will ignore
the data frames.
5. When the last data frame is received by the addressed MCU, the addressed
MCU sets the MPCM bit and waits for a new address frame from master. The
process then repeats from 2.
Using any of the 5- to 8-bit character frame formats is possible, but impractical since the
receiver must change between using n and n+1 character frame formats. This makes
full-duplex operation difficult since the transmitter and receiver uses the same character
size setting. If 5- to 8-bit character frames are used, the transmitter must be set to use
two stop bit (USBS = 1) since the first stop bit is used for indicating the frame type.
Do not use read-modify-write instructions (SBI and CBI) to set or clear the MPCM bit.
The MPCM bit shares the same I/O location as the TXC flag and this might accidentally
be cleared when using SBI or CBI instructions.
USART Register
Description
USART I/O Data Register –

UDR Bit 7 6 5 4 3 2 1 0
RXB[7:0] UDR (Read)
TXB[7:0] UDR (Write)
The USART Transmit Data Buffer Register and USART Receive Data Buffer Registers
share the same I/O address referred to as USART Data Register or UDR. The Transmit
Data Buffer Register (TXB) will be the destination for data written to the UDR Register
183
2467E–AVR–05/02
location. Reading the UDR Register location will return the contents of the receive data
buffer register (RXB).
For 5-, 6-, or 7-bit characters the upper unused bits will be ignored by the Transmitter
and set to zero by the Receiver.
The transmit buffer can only be written when the UDRE flag in the UCSRA Register is
set. Data written to UDR when the UDRE flag is not set, will be ignored by the USART
Transmitter. When data is written to the transmit buffer, and the Transmitter is enabled,
the Transmitter will load the data into the Transmit Shift Register when the Shift Register
is empty. Then the data will be serially transmitted on the TxD pin.
The receive buffer consists of a two level FIFO. The FIFO will change its state whenever
the receive buffer is accessed. Due to this behavior of the receive buffer, do not use
read modify write instructions (SBI and CBI) on this location. Be careful when using bit
test instructions (SBIC and SBIS), since these also will change the state of the FIFO.
USART Control and Status

Register A – UCSRA Bit 7 6 5 4 3 2 1 0
RXC TXC UDRE FE DOR UPE U2X MPCM UCSRA
Read/Write R R/W R R R R R/W R/W
• Bit 7 – RXC: USART Receive Complete

This flag bit is set when there are unread data in the receive buffer and cleared when the
receive buffer is empty (i.e., does not contain any unread data). If the receiver is dis-
abled, the receive buffer will be flushed and consequently the RXC bit will become zero.
The RXC flag can be used to generate a Receive Complete interrupt (see description of
the RXCIE bit).
• Bit 6 – TXC: USART Transmit Complete

This flag bit is set when the entire frame in the Transmit Shift Register has been shifted
out and there are no new data currently present in the transmit buffer (UDR). The TXC
flag bit is automatically cleared when a transmit complete interrupt is executed, or it can
be cleared by writing a one to its bit location. The TXC flag can generate a Transmit
Complete interrupt (see description of the TXCIE bit).
• Bit 5 – UDRE: USART Data Register Empty

The UDRE flag indicates if the transmit buffer (UDR) is ready to receive new data. If
UDRE is one, the buffer is empty, and therefore ready to be written. The UDRE flag can
generate a Data Register Empty interrupt (see description of the UDRIE bit).
UDRE is set after a reset to indicate that the Transmitter is ready.
• Bit 4 – FE: Frame Error

This bit is set if the next character in the receive buffer had a Frame Error when
received. I.e. when the first stop bit of the next character in the receive buffer is zero.
This bit is valid until the receive buffer (UDR) is read. The FE bit is zero when the stop
bit of received data is one. Always set this bit to zero when writing to UCSRA.
• Bit 3 – DOR: Data OverRun

This bit is set if a Data OverRun condition is detected. A data overrun occurs when the
receive buffer is full (two characters), it is a new character waiting in the Receive Shift
184 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Register, and a new start bit is detected. This bit is valid until the receive buffer (UDR) is
read. Always set this bit to zero when writing to UCSRA.
• Bit 2 – UPE: Parity Error

This bit is set if the next character in the receive buffer had a Parity Error when received
and the parity checking was enabled at that point (UPM1 = 1). This bit is valid until the
receive buffer (UDR) is read. Always set this bit to zero when writing to UCSRA.
• Bit 1 – U2X: Double the USART Transmission Speed

This bit only has effect for the asynchronous operation. Write this bit to zero when using
synchronous operation.
Writing this bit to one will reduce the divisor of the baud rate divider from 16 to 8 effec-
tively doubling the transfer rate for asynchronous communication.
• Bit 0 – MPCM: Multi-Processor Communication Mode

This bit enables the Multi-processor Communication mode. When the MPCM bit is writ-
ten to one, all the incoming frames received by the USART receiver that do not contain
address information will be ignored. The transmitter is unaffected by the MPCM setting.
For more detailed information see “Multi-processor Communication Mode” on page 182.

Register B – UCSRB Bit 7 6 5 4 3 2 1 0
RXCIE TXCIE UDRIE RXEN TXEN UCSZ2 RXB8 TXB8 UCSRB
Read/Write R/W R/W R/W R/W R/W R/W R R/W
• Bit 7 – RXCIE: RX Complete Interrupt Enable

Writing this bit to one enables interrupt on the RXC flag. A USART Receive Complete
interrupt will be generated only if the RXCIE bit is written to one, the global interrupt flag
in SREG is written to one and the RXC bit in UCSRA is set.
• Bit 6 – TXCIE: TX Complete Interrupt Enable

Writing this bit to one enables interrupt on the TXC flag. A USART Transmit Complete
interrupt will be generated only if the TXCIE bit is written to one, the global interrupt flag
in SREG is written to one and the TXC bit in UCSRA is set.
• Bit 5 – UDRIE: USART Data Register Empty Interrupt Enable

Writing this bit to one enables interrupt on the UDRE flag. A Data Register Empty inter-
rupt will be generated only if the UDRIE bit is written to one, the global interrupt flag in
SREG is written to one and the UDRE bit in UCSRA is set.
• Bit 4 – RXEN: Receiver Enable

Writing this bit to one enables the USART Receiver. The Receiver will override normal
port operation for the RxD pin when enabled. Disabling the Receiver will flush the
receive buffer invalidating the FE, DOR and UPE flags.
• Bit 3 – TXEN: Transmitter Enable

Writing this bit to one enables the USART Transmitter. The Transmitter will override nor-
mal port operation for the TxD pin when enabled. The disabling of the Transmitter
185
2467E–AVR–05/02
(writing TXEN to zero) will not become effective until ongoing and pending transmis-
sions are completed, i.e., when the Transmit Shift Register and transmit buffer register
do not contain data to be transmitted. When disabled, the transmitter will no longer over-
ride the TxD port.
• Bit 2 – UCSZ2: Character Size

The UCSZ2 bits combined with the UCSZ1:0 bit in UCSRC sets the number of data bits
(character size) in a frame the Receiver and Transmitter use.
• Bit 1 – RXB8: Receive Data Bit 8

RXB8 is the ninth data bit of the received character when operating with serial frames
with 9-data bits. Must be read before reading the low bits from UDR.
• Bit 0 – TXB8: Transmit Data Bit 8

TXB8 is the 9th data bit in the character to be transmitted when operating with serial
frames with 9 data bits. Must be written before writing the low bits to UDR.

Register C – UCSRC Bit 7 6 5 4 3 2 1 0
– UMSEL UPM1 UPM0 USBS UCSZ1 UCSZ0 UCPOL UCSRC
Note that this register is not available in ATmega103 compatibility mode.
• Bit 7 – Reserved Bit

This bit is reserved for future use. For compatibility with future devices, these bit must be
written to zero when UCSRC is written.
• Bit 6 – UMSEL: USART Mode Select

This bit selects between Asynchronous and Synchronous mode of operation.
Table 77. UMSEL Bit Settings

UMSEL Mode
0 Asynchronous Operation
1 Synchronous Operation
• Bit 5:4 – UPM1:0: Parity Mode

These bits enable and set type of parity generation and check. If enabled, the Transmit-
ter will automatically generate and send the parity of the transmitted data bits within
each frame. The Receiver will generate a parity value for the incoming data and com-
pare it to the UPM0 setting. If a mismatch is detected, the UPE flag in UCSRA will be
set.
186 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 78. UPM Bits Settings

UPM1 UPM0 Parity Mode
0 0 Disabled
0 1 (Reserved)
1 0 Enabled, Even Parity
1 1 Enabled, Odd Parity
• Bit 3 – USBS: Stop Bit Select

This bit selects the number of stop bits to be inserted by the Transmitter. The Receiver
ignores this setting.
Table 79. USBS Bit Settings

USBS Stop Bit(s)
0 1-bit
1 2-bits
• Bit 2:1 – UCSZ1:0: Character Size

The UCSZ1:0 bits combined with the UCSZ2 bit in UCSRB sets the number of data bits
(character size) in a frame the Receiver and Transmitter use.
Table 80. UCSZ Bits Settings

UCSZ2 UCSZ1 UCSZ0 Character Size
0 0 0 5-bit
0 0 1 6-bit
0 1 0 7-bit
0 1 1 8-bit
1 0 0 Reserved
1 0 1 Reserved
1 1 0 Reserved
1 1 1 9-bit
• Bit 0 – UCPOL: Clock Polarity

This bit is used for synchronous mode only. Write this bit to zero when Asynchronous
mode is used. The UCPOL bit sets the relationship between data output change and
data input sample, and the synchronous clock (XCK).
Table 81. UCPOL Bit Settings

Transmitted Data Changed (Output of Received Data Sampled (Input on
UCPOL TxD Pin) RxD Pin)
0 Falling XCK Edge Rising XCK Edge
1 Rising XCK Edge Falling XCK Edge
187
2467E–AVR–05/02
USART Baud Rate Registers –
UBRRL and UBRRH Bit 15 14 13 12 11 10 9 8
– – – – UBRR[11:8] UBRRH
UBRR[7:0] UBRRL
7 6 5 4 3 2 1 0
0 0 0 0 0 0 0 0
UBRRH is not available in mega103 compatibility mode

These bits are reserved for future use. For compatibility with future devices, these bit
must be written to zero when UBRRH is written.
• Bit 11:0 – UBRR11:0: USART Baud Rate Register

This is a 12-bit register which contains the USART baud rate. The UBRRH contains the
4 most significant bits, and the UBRRL contains the 8 least significant bits of the USART
baud rate. Ongoing transmissions by the transmitter and receiver will be corrupted if the
baud rate is changed. Writing UBRRL will trigger an immediate update of the baud rate
prescaler.
188 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Examples of Baud Rate For standard crystal and resonator frequencies, the most commonly used baud rates for
Setting asynchronous operation can be generated by using the UBRR settings in Table 82.
UBRR values which yield an actual baud rate differing less than 0.5% from the target
baud rate, are bold in the table. Higher error ratings are acceptable, but the receiver will
have less noise resistance when the error ratings are high, especially for large serial
frames (see “Asynchronous Operational Range” on page 181). The error values are cal-
culated using the following equation:
BaudRate Closest Match

æ -------------------------------------------------------
Error[%] = - – 1ö • 100%
è BaudRate ø
Table 82. Examples of UBRR Settings for Commonly Used Oscillator Frequencies
fosc = 1.0000 MHz fosc = 1.8432 MHz fosc = 2.0000 MHz
Baud
U2X = 0 U2X = 1 U2X = 0 U2X = 1 U2X = 0 U2X = 1
Rate
(bps) UBRR Error UBRR Error UBRR Error UBRR Error UBRR Error UBRR Error
2400 25 0.2% 51 0.2% 47 0.0% 95 0.0% 51 0.2% 103 0.2%
4800 12 0.2% 25 0.2% 23 0.0% 47 0.0% 25 0.2% 51 0.2%
9600 6 -7.0% 12 0.2% 11 0.0% 23 0.0% 12 0.2% 25 0.2%
14.4k 3 8.5% 8 -3.5% 7 0.0% 15 0.0% 8 -3.5% 16 2.1%
19.2k 2 8.5% 6 -7.0% 5 0.0% 11 0.0% 6 -7.0% 12 0.2%
28.8k 1 8.5% 3 8.5% 3 0.0% 7 0.0% 3 8.5% 8 -3.5%
38.4k 1 -18.6% 2 8.5% 2 0.0% 5 0.0% 2 8.5% 6 -7.0%
57.6k 0 8.5% 1 8.5% 1 0.0% 3 0.0% 1 8.5% 3 8.5%
76.8k – – 1 -18.6% 1 -25.0% 2 0.0% 1 -18.6% 2 8.5%
115.2k – – 0 8.5% 0 0.0% 1 0.0% 0 8.5% 1 8.5%
230.4k – – – – – – 0 0.0% – – – –
250k – – – – – – – – – – 0 0.0%
(1)
Max 62.5 kbps 125 kbps 115.2 kbps 230.4 Mbps 125 kpbs 250 kbps
1. UBRR = 0, Error = 0.0%
189
2467E–AVR–05/02
Baud
U2X = 0 U2X = 1 U2X = 0 U2X = 1 U2X = 0 U2X = 1
Rate
2400 95 0.0% 191 0.0% 103 0.2% 207 0.2% 191 0.0% 383 0.0%
4800 47 0.0% 95 0.0% 51 0.2% 103 0.2% 95 0.0% 191 0.0%
9600 23 0.0% 47 0.0% 25 0.2% 51 0.2% 47 0.0% 95 0.0%
14.4k 15 0.0% 31 0.0% 16 2.1% 34 -0.8% 31 0.0% 63 0.0%
19.2k 11 0.0% 23 0.0% 12 0.2% 25 0.2% 23 0.0% 47 0.0%
28.8k 7 0.0% 15 0.0% 8 -3.5% 16 2.1% 15 0.0% 31 0.0%
38.4k 5 0.0% 11 0.0% 6 -7.0% 12 0.2% 11 0.0% 23 0.0%
57.6k 3 0.0% 7 0.0% 3 8.5% 8 -3.5% 7 0.0% 15 0.0%
76.8k 2 0.0% 5 0.0% 2 8.5% 6 -7.0% 5 0.0% 11 0.0%
115.2k 1 0.0% 3 0.0% 1 8.5% 3 8.5% 3 0.0% 7 0.0%
230.4k 0 0.0% 1 0.0% 0 8.5% 1 8.5% 1 0.0% 3 0.0%
250k 0 -7.8% 1 -7.8% 0 0.0% 1 0.0% 1 -7.8% 3 -7.8%
0.5M – – 0 -7.8% – – 0 0.0% 0 -7.8% 1 -7.8%
1M – – – – – – – – – – 0 -7.8%
(1)
Max 230.4 kbps 460.8 kbps 250 kbps 0.5 Mbps 460.8 kpbs 921.6 kbps
1. UBRR = 0, Error = 0.0%
190 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Baud
U2X = 0 U2X = 1 U2X = 0 U2X = 1 U2X = 0 U2X = 1
Rate
2400 207 0.2% 416 -0.1% 287 0.0% 575 0.0% 383 0.0% 767 0.0%
4800 103 0.2% 207 0.2% 143 0.0% 287 0.0% 191 0.0% 383 0.0%
9600 51 0.2% 103 0.2% 71 0.0% 143 0.0% 95 0.0% 191 0.0%
14.4k 34 -0.8% 68 0.6% 47 0.0% 95 0.0% 63 0.0% 127 0.0%
19.2k 25 0.2% 51 0.2% 35 0.0% 71 0.0% 47 0.0% 95 0.0%
28.8k 16 2.1% 34 -0.8% 23 0.0% 47 0.0% 31 0.0% 63 0.0%
38.4k 12 0.2% 25 0.2% 17 0.0% 35 0.0% 23 0.0% 47 0.0%
57.6k 8 -3.5% 16 2.1% 11 0.0% 23 0.0% 15 0.0% 31 0.0%
76.8k 6 -7.0% 12 0.2% 8 0.0% 17 0.0% 11 0.0% 23 0.0%
115.2k 3 8.5% 8 -3.5% 5 0.0% 11 0.0% 7 0.0% 15 0.0%
230.4k 1 8.5% 3 8.5% 2 0.0% 5 0.0% 3 0.0% 7 0.0%
250k 1 0.0% 3 0.0% 2 -7.8% 5 -7.8% 3 -7.8% 6 5.3%
0.5M 0 0.0% 1 0.0% – – 2 -7.8% 1 -7.8% 3 -7.8%
1M – – 0 0.0% – – – – 0 -7.8% 1 -7.8%
(1)
Max 0.5 Mbps 1 Mbps 691.2 kbps 1.3824 Mbps 921.6 kbps 1.8432 Mbps
1. UBRR = 0, Error = 0.0%
191
2467E–AVR–05/02
Baud
U2X = 0 U2X = 1 U2X = 0 U2X = 1 U2X = 0 U2X = 1
Rate
2400 416 -0.1% 832 0.0% 479 0.0% 959 0.0% 520 0.0% 1041 0.0%
4800 207 0.2% 416 -0.1% 239 0.0% 479 0.0% 259 0.2% 520 0.0%
9600 103 0.2% 207 0.2% 119 0.0% 239 0.0% 129 0.2% 259 0.2%
14.4k 68 0.6% 138 -0.1% 79 0.0% 159 0.0% 86 -0.2% 173 -0.2%
19.2k 51 0.2% 103 0.2% 59 0.0% 119 0.0% 64 0.2% 129 0.2%
28.8k 34 -0.8% 68 0.6% 39 0.0% 79 0.0% 42 0.9% 86 -0.2%
38.4k 25 0.2% 51 0.2% 29 0.0% 59 0.0% 32 -1.4% 64 0.2%
57.6k 16 2.1% 34 -0.8% 19 0.0% 39 0.0% 21 -1.4% 42 0.9%
76.8k 12 0.2% 25 0.2% 14 0.0% 29 0.0% 15 1.7% 32 -1.4%
115.2k 8 -3.5% 16 2.1% 9 0.0% 19 0.0% 10 -1.4% 21 -1.4%
230.4k 3 8.5% 8 -3.5% 4 0.0% 9 0.0% 4 8.5% 10 -1.4%
250k 3 0.0% 7 0.0% 4 -7.8% 8 2.4% 4 0.0% 9 0.0%
0.5M 1 0.0% 3 0.0% – – 4 -7.8% – – 4 0.0%
1M 0 0.0% 1 0.0% – – – – – – – –
(1)
Max 1 Mbps 2 Mbps 1.152 Mbps 2.304 Mbps 1.25 Mbps 2.5 Mbps
1. UBRR = 0, Error = 0.0%
192 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Two-wire Serial
Interface
Features • Simple yet Powerful and Flexible Communication Interface, only Two Bus Lines Needed
• Both Master and Slave Operation Supported
• Device can Operate as Transmitter or Receiver
• 7-bit Address Space allows up to 128 Different Slave Addresses
• Multi-master Arbitration Support
• Up to 400 kHz Data Transfer Speed
• Slew-rate Limited Output Drivers
• Noise Suppression Circuitry Rejects Spikes on Bus Lines
• Fully Programmable Slave Address with General Call Support
• Address Recognition Causes Wake-up when AVR is in Sleep Mode
Two-wire Serial Interface The Two-wire Serial Interface (TWI) is ideally suited for typical microcontroller applica-
Bus Definition tions. The TWI protocol allows the systems designer to interconnect up to 128 different
devices using only two bi-directional bus lines, one for clock (SCL) and one for data
(SDA). The only external hardware needed to implement the bus is a single pull-up
resistor for each of the TWI bus lines. All devices connected to the bus have individual
addresses, and mechanisms for resolving bus contention are inherent in the TWI
protocol.
Figure 85. TWI Bus Interconnection

VCC
Device 1 Device 2 Device 3 ........ Device n R1 R2
SDA
SCL
TWI Terminology The following definitions are frequently encountered in this section.
Table 86. TWI Terminology

Term Description
Master The device that initiates and terminates a transmission. The master also
generates the SCL clock
Slave The device addressed by a master
Transmitter The device placing data on the bus
Receiver The device reading data from the bus
Electrical Interconnection As depicted in Figure 85, both bus lines are connected to the positive supply voltage
through pull-up resistors. The bus drivers of all TWI-compliant devices are open-drain or
open-collector. This implements a wired-AND function which is essential to the opera-
tion of the interface. A low level on a TWI bus line is generated when one or more TWI
193
2467E–AVR–05/02
devices output a zero. A high level is output when all TWI devices tri-state their outputs,
allowing the pull-up resistors to pull the line high. Note that all AVR devices connected to
the TWI bus must be powered in order to allow any bus operation.
The number of devices that can be connected to the bus is only limited by the bus
capacitance limit of 400 pF and the 7-bit slave address space. A detailed specification of
the electrical characteristics of the TWI is given in “Two-wire Serial Interface Character-
istics” on page 318. Two different sets of specifications are presented there, one
relevant for bus speeds below 100 kHz, and one valid for bus speeds up to 400 kHz.
Data Transfer and Frame

Format
Transferring Bits Each data bit transferred on the TWI bus is accompanied by a pulse on the clock line.
The level of the data line must be stable when the clock line is high. The only exception
to this rule is for generating start and stop conditions.
Figure 86. Data Validity
SDA
SCL
Data Stable Data Stable
Data Change
START and STOP Conditions The master initiates and terminates a data transmission. The transmission is initiated
when the master issues a START condition on the bus, and it is terminated when the
master issues a STOP condition. Between a START and a STOP condition, the bus is
considered busy, and no other master should try to seize control of the bus. A special
case occurs when a new START condition is issued between a START and STOP con-
dition. This is referred to as a REPEATED START condition, and is used when the
master wishes to initiate a new transfer without relinquishing control of the bus. After a
REPEATED START, the bus is considered busy until the next STOP. This is identical to
the START behaviour, and therefore START is used to describe both START and
REPEATED START for the remainder of this datasheet, unless otherwise noted. As
depicted below, START and STOP conditions are signalled by changing the level of the
SDA line when the SCL line is high.
Figure 87. START, REPEATED START and STOP Conditions
SDA
SCL
START STOP START REPEATED START STOP
194 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Address Packet Format All address packets transmitted on the TWI bus are 9 bits long, consisting of 7 address
bits, one READ/WRITE control bit and an acknowledge bit. If the READ/WRITE bit is
set, a read operation is to be performed, otherwise a write operation should be per-
formed. When a slave recognizes that it is being addressed, it should acknowledge by
pulling SDA low in the ninth SCL (ACK) cycle. If the addressed slave is busy, or for
some other reason can not service the master’s request, the SDA line should be left
high in the ACK clock cycle. The master can then transmit a STOP condition, or a
REPEATED START condition to initiate a new transmission. An address packet consist-
ing of a slave address and a READ or a WRITE bit is called SLA+R or SLA+W,
respectively.
The MSB of the address byte is transmitted first. Slave addresses can freely be allo-
cated by the designer, but the address 0000 000 is reserved for a general call.
When a general call is issued, all slaves should respond by pulling the SDA line low in
the ACK cycle. A general call is used when a master wishes to transmit the same mes-
sage to several slaves in the system. When the general call address followed by a Write
bit is transmitted on the bus, all slaves set up to acknowledge the general call will pull
the SDA line low in the ack cycle. The following data packets will then be received by all
the slaves that acknowledged the general call. Note that transmitting the general call
address followed by a Read bit is meaningless, as this would cause contention if several
slaves started transmitting different data.
All addresses of the format 1111 xxx should be reserved for future purposes.
Figure 88. Address Packet Format
Addr MSB Addr LSB R/W ACK
SDA
SCL
1 2 7 8 9
START
Data Packet Format All data packets transmitted on the TWI bus are 9 bits long, consisting of one data byte
and an acknowledge bit. During a data transfer, the master generates the clock and the
START and STOP conditions, while the receiver is responsible for acknowledging the
reception. An Acknowledge (ACK) is signalled by the receiver pulling the SDA line low
during the ninth SCL cycle. If the receiver leaves the SDA line high, a NACK is signalled.
When the receiver has received the last byte, or for some reason cannot receive any
more bytes, it should inform the transmitter by sending a NACK after the final byte. The
MSB of the data byte is transmitted first.
195
2467E–AVR–05/02
Figure 89. Data Packet Format
Data MSB Data LSB ACK

Aggregate
SDA
SDA from
Transmitter
SDA from
Receiver
SCL from
Master
1 2 7 8 9
STOP, REPEATED
SLA+R/W Data Byte START or Next
Data Byte
Combining Address and Data A transmission basically consists of a START condition, a SLA+R/W, one or more data
Packets Into a Transmission packets and a STOP condition. An empty message, consisting of a START followed by
a STOP condition, is illegal. Note that the Wired-ANDing of the SCL line can be used to
implement handshaking between the master and the slave. The slave can extend the
SCL low period by pulling the SCL line low. This is useful if the clock speed set up by the
master is too fast for the slave, or the slave needs extra time for processing between the
data transmissions. The slave extending the SCL low period will not affect the SCL high
period, which is determined by the master. As a consequence, the slave can reduce the
TWI data transfer speed by prolonging the SCL duty cycle.
Figure 90 shows a typical data transmission. Note that several data bytes can be trans-
mitted between the SLA+R/W and the STOP condition, depending on the software
protocol implemented by the application software.
Figure 90. Typical Data Transmission
Addr MSB Addr LSB R/W ACK Data MSB Data LSB ACK
SDA
SCL
1 2 7 8 9 1 2 7 8 9
START SLA+R/W Data Byte STOP
Multi-master Bus The TWI protocol allows bus systems with several masters. Special concerns have
Systems, Arbitration and been taken in order to ensure that transmissions will proceed as normal, even if two or
Synchronization more masters initiate a transmission at the same time. Two problems arise in multi-mas-
ter systems:
• An algorithm must be implemented allowing only one of the masters to complete the
transmission. All other masters should cease transmission when they discover that
they have lost the selection process. This selection process is called arbitration.
When a contending master discovers that it has lost the arbitration process, it
should immediately switch to slave mode to check whether it is being addressed by
the winning master. The fact that multiple masters have started transmission at the
same time should not be detectable to the slaves, i.e., the data being transferred on
the bus must not be corrupted.
196 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
• Different masters may use different SCL frequencies. A scheme must be devised to
synchronize the serial clocks from all masters, in order to let the transmission
proceed in a lockstep fashion. This will facilitate the arbitration process.
The wired-ANDing of the bus lines is used to solve both these problems. The serial
clocks from all masters will be wired-ANDed, yielding a combined clock with a high
period equal to the one from the master with the shortest high period. The low period of
the combined clock is equal to the low period of the master with the longest low period.
Note that all masters listen to the SCL line, effectively starting to count their SCL high
and low time-out periods when the combined SCL line goes high or low, respectively.
Figure 91. SCL Synchronization between Multiple Masters

TA low TA high
SCL from
master A
SCL from
master B
SCL Bus
Line
TBlow TBhigh
Masters Start Masters Start

Counting Low Period Counting High Period
Arbitration is carried out by all masters continuously monitoring the SDA line after out-
putting data. If the value read from the SDA line does not match the value the master
had output, it has lost the arbitration. Note that a master can only lose arbitration when it
outputs a high SDA value while another master outputs a low value. The losing master
should immediately go to slave mode, checking if it is being addressed by the winning
master. The SDA line should be left high, but losing masters are allowed to generate a
clock signal until the end of the current data or address packet. Arbitration will continue
until only one master remains, and this may take many bits. If several masters are trying
to address the same slave, arbitration will continue into the data packet.
197
2467E–AVR–05/02
Figure 92. Arbitration Between two Masters
START Master A loses
Arbitration, SDAA SDA
SDA from
Master A
SDA from
Master B
SDA Line
Synchronized
SCL Line
Note that arbitration is not allowed between:

• A REPEATED START condition and a data bit
• A STOP condition and a data bit
• A REPEATED START and a STOP condition
It is the user software’s responsibility to ensure that these illegal arbitration conditions
never occur. This implies that in multi-master systems, all data transfers must use the
same composition of SLA+R/W and data packets. In other words: All transmissions
must contain the same number of data packets, otherwise the result of the arbitration is
undefined.
198 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Overview of the TWI The TWI module is comprised of several submodules, as shown in Figure 93. All regis-
Module ters drawn in a thick line are accessible through the AVR data bus.
Figure 93. Overview of the TWI Module

SCL SDA
Slew-rate Spike Slew-rate Spike
Control Filter Control Filter
Bus Interface Unit Bit Rate Generator

START / STOP
Spike Suppression Prescaler
Control
Address/Data Shift Bit Rate Register

Arbitration detection Ack
Register (TWDR) (TWBR)
Address Match Unit Control Unit
TWI Unit
Address Register Status Register Control Register
(TWAR) (TWSR) (TWCR)
State Machine and

Address Comparator
Status control
Scl and SDA Pins These pins interface the AVR TWI with the rest of the MCU system. The output drivers
contain a slew-rate limiter in order to conform to the TWI specification. The input stages
contain a spike suppression unit removing spikes shorter than 50 ns. Note that the inter-
nal pullups in the AVR pads can be enabled by setting the PORT bits corresponding to
the SCL and SDA pins, as explained in the I/O Port section. The internal pull-ups can in
some systems eliminate the need for external ones.
Bit Rate Generator Unit This unit controls the period of SCL when operating in a Master mode. The SCL period
is controlled by settings in the TWI Bit Rate Register (TWBR) and the Prescaler bits in
the TWI Status Register (TWSR). Slave operation does not depend on Bit Rate or Pres-
caler settings, but the CPU clock frequency in the slave must be at least 16 times higher
than the SCL frequency. Note that slaves may prolong the SCL low period, thereby
reducing the average TWI bus clock period. The SCL frequency is generated according
to the following equation:
CPU Clock frequency
SCL frequency = -----------------------------------------------------------
TWPS
16 + 2(TWBR) ⋅ 4
• TWBR = Value of the TWI Bit Rate Register

• TWPS = Value of the prescaler bits in the TWI Status Register
Note: TWBR should be 10 or higher if the TWI operates in Master mode. If TWBR is lower than
10, the master may produce an incorrect output on SDA and SCL for the reminder of the
byte. The problem occurs when operating the TWI in Master mode, sending Start + SLA
+ R/W to a slave (a slave does not need to be connected to the bus for the condition to
happen).
199
2467E–AVR–05/02
Bus Interface Unit This unit contains the Data and Address Shift Register (TWDR), a START/STOP Con-
troller and Arbitration detection hardware. The TWDR contains the address or data
bytes to be transmitted, or the address or data bytes received. In addition to the 8-bit
TWDR, the Bus Interface Unit also contains a register containing the (N)ACK bit to be
transmitted or received. This (N)ACK Register is not directly accessible by the applica-
tion software. However, when receiving, it can be set or cleared by manipulating the
TWI Control Register (TWCR). When in Transmitter mode, the value of the received
(N)ACK bit can be determined by the value in the TWSR.
The START/STOP Controller is responsible for generation and detection of START,
REPEATED START, and STOP conditions. The START/STOP controller is able to
detect START and STOP conditions even when the AVR MCU is in one of the sleep
modes, enabling the MCU to wake up if addressed by a master.
If the TWI has initiated a transmission as master, the Arbitration Detection hardware
continuously monitors the transmission trying to determine if arbitration is in process. If
the TWI has lost an arbitration, the Control Unit is informed. Correct action can then be
taken and appropriate status codes generated.
Address Match Unit The Address Match unit checks if received address bytes match the 7-bit address in the
TWI Address Register (TWAR). If the TWI General Call Recognition Enable (TWGCE)
bit in the TWAR is written to one, all incoming address bits will also be compared
against the General Call address. Upon an address match, the Control Unit is informed,
allowing correct action to be taken. The TWI may or may not acknowledge its address,
depending on settings in the TWCR. The Address Match unit is able to compare
addresses even when the AVR MCU is in sleep mode, enabling the MCU to wake up if
addressed by a master. If another interrupt (e.g., INT0) occurs during TWI Power-down
address match and wakes up the CPU, the TWI aborts operation and return to it’s idle
state. If this cause any problems, ensure that TWI Address Match is the only enabled
interrupt when entering Power-down.
Control Unit The Control unit monitors the TWI bus and generates responses corresponding to set-
tings in the TWI Control Register (TWCR). When an event requiring the attention of the
application occurs on the TWI bus, the TWI Interrupt Flag (TWINT) is asserted. In the
next clock cycle, the TWI Status Register (TWSR) is updated with a status code identify-
ing the event. The TWSR only contains relevant status information when the TWI
Interrupt Flag is asserted. At all other times, the TWSR contains a special status code
indicating that no relevant status information is available. As long as the TWINT flag is
set, the SCL line is held low. This allows the application software to complete its tasks
before allowing the TWI transmission to continue.
The TWINT flag is set in the following situations:
• After the TWI has transmitted a START/REPEATED START condition
• After the TWI has transmitted SLA+R/W
• After the TWI has transmitted an address byte
• After the TWI has lost arbitration
• After the TWI has been addressed by own slave address or general call
• After the TWI has received a data byte
• After a STOP or REPEATED START has been received while still addressed as a
slave
• When a bus error has occurred due to an illegal START or STOP condition
200 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
TWI Register Description
TWI Bit Rate Register – TWBR

Bit 7 6 5 4 3 2 1 0
TWBR7 TWBR6 TWBR5 TWBR4 TWBR3 TWBR2 TWBR1 TWBR0 TWBR
• Bits 7..0 – TWI Bit Rate Register

TWBR selects the division factor for the bit rate generator. The bit rate generator is a
frequency divider which generates the SCL clock frequency in the Master modes. See
“Bit Rate Generator Unit” on page 199 for calculating bit rates.
TWI Control Register – TWCR

Bit 7 6 5 4 3 2 1 0
TWINT TWEA TWSTA TWSTO TWWC TWEN – TWIE TWCR
Read/Write R/W R/W R/W R/W R R/W R R/W
The TWCR is used to control the operation of the TWI. It is used to enable the TWI, to
initiate a master access by applying a START condition to the bus, to generate a
receiver acknowledge, to generate a stop condition, and to control halting of the bus
while the data to be written to the bus are written to the TWDR. It also indicates a write
collision if data is attempted written to TWDR while the register is inaccessible.
• Bit 7 – TWINT: TWI Interrupt Flag

This bit is set by hardware when the TWI has finished its current job and expects appli-
cation software response. If the I-bit in SREG and TWIE in TWCR are set, the MCU will
jump to the TWI interrupt vector. While the TWINT flag is set, the SCL low period is
stretched. The TWINT flag must be cleared by software by writing a logic one to it. Note
that this flag is not automatically cleared by hardware when executing the interrupt rou-
tine. Also note that clearing this flag starts the operation of the TWI, so all accesses to
the TWI Address Register (TWAR), TWI Status Register (TWSR), and TWI Data Regis-
ter (TWDR) must be complete before clearing this flag.
• Bit 6 – TWEA: TWI Enable Acknowledge Bit

The TWEA bit controls the generation of the acknowledge pulse. If the TWEA bit is writ-
ten to one, the ACK pulse is generated on the TWI bus if the following conditions are
met:
1. The device’s own slave address has been received.
2. A general call has been received, while the TWGCE bit in the TWAR is set.
3. A data byte has been received in Master Receiver or Slave Receiver mode.
By writing the TWEA bit to zero, the device can be virtually disconnected from the Two-
wire Serial Bus temporarily. Address recognition can then be resumed by writing the
TWEA bit to one again.
• Bit 5 – TWSTA: TWI START Condition Bit

The application writes the TWSTA bit to one when it desires to become a master on the
Two-wire Serial Bus. The TWI hardware checks if the bus is available, and generates a
START condition on the bus if it is free. However, if the bus is not free, the TWI waits
201
2467E–AVR–05/02
until a STOP condition is detected, and then generates a new START condition to claim
the bus Master status. TWSTA is cleared by the TWI hardware when the START condi-
tion has been transmitted.
• Bit 4 – TWSTO: TWI STOP Condition Bit

Writing the TWSTO bit to one in Master mode will generate a STOP condition on the
Two-wire Serial Bus. When the STOP condition is executed on the bus, the TWSTO bit
is cleared automatically. In slave mode, setting the TWSTO bit can be used to recover
from an error condition. This will not generate a STOP condition, but the TWI returns to
a well-defined unaddressed Slave mode and releases the SCL and SDA lines to a high
impedance state.
• Bit 3 – TWWC: TWI Write Collision Flag

The TWWC bit is set when attempting to write to the TWI Data Register – TWDR when
TWINT is low. This flag is cleared by writing the TWDR Register when TWINT is high.
• Bit 2 – TWEN: TWI Enable Bit

The TWEN bit enables TWI operation and activates the TWI interface. When TWEN is
written to one, the TWI takes control over the I/O pins connected to the SCL and SDA
pins, enabling the slew-rate limiters and spike filters. If this bit is written to zero, the TWI
is switched off and all TWI transmissions are terminated, regardless of any ongoing
operation.

This bit is a reserved bit and will always read as zero.
• Bit 0 – TWIE: TWI Interrupt Enable

When this bit is written to one, and the I-bit in SREG is set, the TWI interrupt request will
be activated for as long as the TWINT flag is high.
TWI Status Register – TWSR

Bit 7 6 5 4 3 2 1 0
TWS7 TWS6 TWS5 TWS4 TWS3 – TWPS1 TWPS0 TWSR
Read/Write R R R R R R R/W R/W
• Bits 7..3 – TWS: TWI Status

These 5 bits reflect the status of the TWI logic and the Two-wire Serial Bus. The differ-
ent status codes are described later in this section. Note that the value read from TWSR
contains both the 5-bit status value and the 2-bit prescaler value. The application
designer should mask the prescaler bits to zero when checking the Status bits. This
makes status checking independent of prescaler setting. This approach is used in this
datasheet, unless otherwise noted.

This bit is reserved and will always read as zero.
202 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
• Bits 1..0 – TWPS: TWI Prescaler Bits

These bits can be read and written, and control the bit rate prescaler. See “Bit Rate
Generator Unit” on page 199 for calculating bit rates.
TWI Data Register – TWDR

Bit 7 6 5 4 3 2 1 0
TWD7 TWD6 TWD5 TWD4 TWD3 TWD2 TWD1 TWD0 TWDR
In Transmit mode, TWDR contains the next byte to be transmitted. In receive mode, the
TWDR contains the last byte received. It is writable while the TWI is not in the process of
shifting a byte. This occurs when the TWI interrupt flag (TWINT) is set by hardware.
Note that the Data Register cannot be initialized by the user before the first interrupt
occurs. The data in TWDR remains stable as long as TWINT is set. While data is shifted
out, data on the bus is simultaneously shifted in. TWDR always contains the last byte
present on the bus, except after a wake up from a sleep mode by the TWI interrupt. In
this case, the contents of TWDR is undefined. In the case of a lost bus arbitration, no
data is lost in the transition from Master to Slave. Handling of the ACK bit is controlled
automatically by the TWI logic, the CPU cannot access the ACK bit directly.
• Bits 7..0 – TWD: TWI Data Register

These eight bits constitute the next data byte to be transmitted, or the latest data byte
received on the Two-wire Serial Bus.
TWI (Slave) Address Register

– TWAR Bit 7 6 5 4 3 2 1 0
TWA6 TWA5 TWA4 TWA3 TWA2 TWA1 TWA0 TWGCE TWAR
The TWAR should be loaded with the 7-bit slave address (in the seven most significant
bits of TWAR) to which the TWI will respond when programmed as a slave transmitter or
receiver, and not needed in the master modes. In multimaster systems, TWAR must be
set in masters which can be addressed as slaves by other masters.
The LSB of TWAR is used to enable recognition of the general call address ($00). There
is an associated address comparator that looks for the slave address (or general call
address if enabled) in the received serial address. If a match is found, an interrupt
request is generated.
• Bits 7..1 – TWA: TWI (Slave) Address Register

These seven bits constitute the slave address of the TWI unit.
• Bit 0 – TWGCE: TWI General Call Recognition Enable Bit

If set, this bit enables the recognition of a General Call given over the Two-wire Serial
Bus.
Using the TWI The AVR TWI is byte-oriented and interrupt based. Interrupts are issued after all bus
events, like reception of a byte or transmission of a START condition. Because the TWI
is interrupt-based, the application software is free to carry on other operations during a
TWI byte transfer. Note that the TWI Interrupt Enable (TWIE) bit in TWCR together with
203
2467E–AVR–05/02
the Global Interrupt Enable bit in SREG allow the application to decide whether or not
assertion of the TWINT flag should generate an interrupt request. If the TWIE bit is
cleared, the application must poll the TWINT flag in order to detect actions on the TWI
bus.
When the TWINT flag is asserted, the TWI has finished an operation and awaits applica-
tion response. In this case, the TWI Status Register (TWSR) contains a value indicating
the current state of the TWI bus. The application software can then decide how the TWI
should behave in the next TWI bus cycle by manipulating the TWCR and TWDR
Registers.
Figure 94 is a simple example of how the application can interface to the TWI hardware.
In this example, a master wishes to transmit a single data byte to a slave. This descrip-
tion is quite abstract, a more detailed explanation follows later in this section. A simple
code example implementing the desired behaviour is also presented.
Figure 94. Interfacing the Application to the TWI in a Typical Transmission
1. Application 3. Check TWSR to see if 5. Check TWSR to see if SLA+W 7. Check TWSR to see if data
Application
writes to TWCR START was sent. Application was sent and ACK received. was sent and ACK received.
to initiate loads SLA+W into TWDR, and Application loads data into TWDR, Application loads appropriate
Action
transmission of loads appropriate control signals and loads appropriate control signals control signals to send STOP
START into TWCR, making sure that into TWCR, making sure that TWINT into TWCR, making sure that
TWINT is written to one. is written to one. TWINT is written to one
TWI bus START SLA+W A Data A STOP
Indicates
2. TWINT set. 4. TWINT set. 6. TWINT set.
TWINT set
Status code indicates Status code indicates Status code indicates
TWI
START condition sent SLA+W sendt, ACK data sent, ACK received
Hardware
received
Action
1. The first step in a TWI transmission is to transmit a START condition. This is

done by writing a specific value into TWCR, instructing the TWI hardware to
transmit a START condition. Which value to write is described later on. However,
it is important that the TWINT bit is set in the value written. Writing a one to
TWINT clears the flag. The TWI will not start any operation as long as the
TWINT bit in TWCR is set. Immediately after the application has cleared TWINT,
the TWI will initiate transmission of the START condition.
2. When the START condition has been transmitted, the TWINT flag in TWCR is
set, and TWSR is updated with a status code indicating that the START condition
has successfully been sent.
3. The application software should now examine the value of TWSR, to make sure
that the START condition was successfully transmitted. If TWSR indicates other-
wise, the application software might take some special action, like calling an
error routine. Assuming that the status code is as expected, the application must
load SLA+W into TWDR. Remember that TWDR is used both for address and
data. After TWDR has been loaded with the desired SLA+W, a specific value
must be written to TWCR, instructing the TWI hardware to transmit the SLA+W
204 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
present in TWDR. Which value to write is described later on. However, it is

important that the TWINT bit is set in the value written. Writing a one to TWINT
clears the flag. The TWI will not start any operation as long as the TWINT bit in
TWCR is set. Immediately after the application has cleared TWINT, the TWI will
initiate transmission of the address packet.
4. When the address packet has been transmitted, the TWINT flag in TWCR is set,
and TWSR is updated with a status code indicating that the address packet has
successfully been sent. The status code will also reflect whether a slave
acknowledged the packet or not.
that the address packet was successfully transmitted, and that the value of the
ACK bit was as expected. If TWSR indicates otherwise, the application software
might take some special action, like calling an error routine. Assuming that the
status code is as expected, the application must load a data packet into TWDR.
Subsequently, a specific value must be written to TWCR, instructing the TWI
hardware to transmit the data packet present in TWDR. Which value to write is
described later on. However, it is important that the TWINT bit is set in the value
written. Writing a one to TWINT clears the flag. The TWI will not start any opera-
tion as long as the TWINT bit in TWCR is set. Immediately after the application
has cleared TWINT, the TWI will initiate transmission of the data packet.
6. When the data packet has been transmitted, the TWINT flag in TWCR is set, and
TWSR is updated with a status code indicating that the data packet has success-
fully been sent. The status code will also reflect whether a slave acknowledged
the packet or not.
that the data packet was successfully transmitted, and that the value of the ACK
bit was as expected. If TWSR indicates otherwise, the application software might
take some special action, like calling an error routine. Assuming that the status
code is as expected, the application must write a specific value to TWCR,
instructing the TWI hardware to transmit a STOP condition. Which value to write
is described later on. However, it is important that the TWINT bit is set in the
value written. Writing a one to TWINT clears the flag. The TWI will not start any
operation as long as the TWINT bit in TWCR is set. Immediately after the appli-
cation has cleared TWINT, the TWI will initiate transmission of the STOP
condition. Note that TWINT is NOT set after a STOP condition has been sent.
Even though this example is simple, it shows the principles involved in all TWI transmis-
sions. These can be summarized as follows:
• When the TWI has finished an operation and expects application response, the
TWINT flag is set. The SCL line is pulled low until TWINT is cleared.
• When the TWINT flag is set, the user must update all TWI Registers with the value
relevant for the next TWI bus cycle. As an example, TWDR must be loaded with the
value to be transmitted in the next bus cycle.
• After all TWI Register updates and other pending application software tasks have
been completed, TWCR is written. When writing TWCR, the TWINT bit should be
set. Writing a one to TWINT clears the flag. The TWI will then commence executing
whatever operation was specified by the TWCR setting.
In the following an assembly and C implementation of the example is given. Note that
the code below assumes that several definitions have been made for example by using
include-files.
205
2467E–AVR–05/02
Assembly Code Example C Example Comments
1 ldi r16, (1<<TWINT)|(1<<TWSTA)| TWCR = (1<<TWINT)|(1<<TWSTA)|
(1<<TWEN) (1<<TWEN) Send START condition
out TWCR, r16
2 wait1: while (!(TWCR & (1<<TWINT)))
in r16,TWCR ; Wait for TWINT flag set. This indicates that
sbrs r16,TWINT the START condition has been transmitted
rjmp wait1
3 in r16,TWSR if ((TWSR & 0xF8) != START)
andi r16, 0xF8 ERROR(); Check value of TWI Status Register. Mask
prescaler bits. If status different from START
cpi r16, START go to ERROR
brne ERROR
ldi r16, SLA_W TWDR = SLA_W;
out TWDR, r16 TWCR = (1<<TWINT) | (1<<TWEN); Load SLA_W into TWDR Register. Clear
TWINT bit in TWCR to start transmission of
ldi r16, (1<<TWINT) | (1<<TWEN) address
out TWCR, r16
the SLA+W has been transmitted, and
sbrs r16,TWINT ACK/NACK has been received.
rjmp wait2
5 in r16,TWSR if ((TWSR & 0xF8) != MT_SLA_ACK)
prescaler bits. If status different from
cpi r16, MT_SLA_ACK MT_SLA_ACK go to ERROR
brne ERROR
ldi r16, DATA TWDR = DATA;
out TWDR, r16 TWCR = (1<<TWINT) | (1<<TWEN); Load DATA into TWDR Register. Clear TWINT
ldi r16, (1<<TWINT) | (1<<TWEN) bit in TWCR to start transmission of address
out TWCR, r16
the DATA has been transmitted, and
sbrs r16,TWINT ACK/NACK has been received.
rjmp wait3
7 in r16,TWSR if ((TWSR & 0xF8) != MT_DATA_ACK)
prescaler bits. If status different from
cpi r16, MT_DATA_ACK MT_DATA_ACK go to ERROR
brne ERROR
ldi r16, (1<<TWINT)|(1<<TWEN)| TWCR = (1<<TWINT)|(1<<TWEN)|
(1<<TWSTO) (1<<TWSTO); Transmit STOP condition
out TWCR, r16
Note: For I/O registers located in extended I/O map, “IN”, “OUT”, “SBIS”, “SBIC”, “CBI”, and “SBI” instructions must be replaced with
instructions that allow access to extended I/O. Typically “LDS” and “STS” combined with “SBRS”, “SBRC”, “SBR”, and “CBR”.
206 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Transmission Modes The TWI can operate in one of four major modes. These are named Master Transmitter
(MT), Master Receiver (MR), Slave Transmitter (ST) and Slave Receiver (SR). Several
of these modes can be used in the same application. As an example, the TWI can use
MT mode to write data into a TWI EEPROM, MR mode to read the data back from the
EEPROM. If other masters are present in the system, some of these might transmit data
to the TWI, and then SR mode would be used. It is the application software that decides
which modes are legal.
The following sections describe each of these modes. Possible status codes are
described along with figures detailing data transmission in each of the modes. These fig-
ures contain the following abbreviations:
S: START condition
Rs: REPEATED START condition
R: Read bit (high level at SDA)
W: Write bit (low level at SDA)
A: Acknowledge bit (low level at SDA)
A: Not acknowledge bit (high level at SDA)
Data: 8-bit data byte
P: STOP condition
SLA: Slave Address
In Figure 96 to Figure 102, circles are used to indicate that the TWINT flag is set. The
numbers in the circles show the status code held in TWSR, with the prescaler bits
masked to zero. At these points, actions must be taken by the application to continue or
complete the TWI transfer. The TWI transfer is suspended until the TWINT flag is
cleared by software.
When the TWINT flag is set, the status code in TWSR is used to determine the appropri-
ate software action. For each status code, the required software action and details of the
following serial transfer are given in Table 87 to Table 90. Note that the prescaler bits
are masked to zero in these tables.
Master Transmitter Mode In the Master Transmitter mode, a number of data bytes are transmitted to a slave
receiver (see Figure 95). In order to enter a Master mode, a START condition must be
transmitted. The format of the following address packet determines whether Master
Transmitter or Master Receiver mode is to be entered. If SLA+W is transmitted, MT
mode is entered, if SLA+R is transmitted, MR mode is entered. All the status codes
mentioned in this section assume that the prescaler bits are zero or are masked to zero.
207
2467E–AVR–05/02
Figure 95. Data Transfer in Master Transmitter Mode
VCC
Device 1 Device 2
MASTER SLAVE Device 3 ........ Device n R1 R2
TRANSMITTER RECEIVER
SDA
SCL
A START condition is sent by writing the following value to TWCR:

TWCR TWINT TWEA TWSTA TWSTO TWWC TWEN – TWIE
value 1 X 1 0 X 1 0 X
TWEN must be set to enable the Two-wire Serial Interface, TWSTA must be written to
one to transmit a START condition and TWINT must be written to one to clear the
TWINT flag. The TWI will then test the Two-wire Serial Bus and generate a START con-
dition as soon as the bus becomes free. After a START condition has been transmitted,
the TWINT flag is set by hardware, and the status code in TWSR will be $08 (See Table
87). In order to enter MT mode, SLA+W must be transmitted. This is done by writing
SLA+W to TWDR. Thereafter the TWINT bit should be cleared (by writing it to one) to
continue the transfer. This is accomplished by writing the following value to TWCR:
value 1 X 0 0 X 1 0 X
When SLA+W have been transmitted and an acknowledgment bit has been received,
TWINT is set again and a number of status codes in TWSR are possible. Possible sta-
tus codes in Master mode are $18, $20, or $38. The appropriate action to be taken for
each of these status codes is detailed in Table 87.
When SLA+W has been successfully transmitted, a data packet should be transmitted.
This is done by writing the data byte to TWDR. TWDR must only be written when
TWINT is high. If not, the access will be discarded, and the Write Collision bit (TWWC)
will be set in the TWCR Register. After updating TWDR, the TWINT bit should be
cleared (by writing it to one) to continue the transfer. This is accomplished by writing the
following value to TWCR:
value 1 X 0 0 X 1 0 X
This scheme is repeated until the last byte has been sent and the transfer is ended by
generating a STOP condition or a repeated START condition. A STOP condition is gen-
erated by writing the following value to TWCR:
value 1 X 0 1 X 1 0 X
A REPEATED START condition is generated by writing the following value to TWCR:

value 1 X 1 0 X 1 0 X
208 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
After a repeated START condition (state $10) the Two-wire Serial Interface can access
the same slave again, or a new slave without transmitting a STOP condition. Repeated
START enables the master to switch between slaves, Master Transmitter mode and
Master Receiver mode without losing control of the bus.
Table 87. Status Codes for Master Transmitter Mode

Status Code Application Software Response
(TWSR) Status of the Two-wire Serial To TWCR
Prescaler Bits Bus and Two-wire Serial Inter- To/from TWDR
are 0 face Hardware STA STO TWINT TWEA Next Action Taken by TWI Hardware
$08 A START condition has been Load SLA+W X 0 1 X SLA+W will be transmitted;
transmitted ACK or NOT ACK will be received
$10 A repeated START condition Load SLA+W or X 0 1 X SLA+W will be transmitted;
has been transmitted ACK or NOT ACK will be received
Load SLA+R X 0 1 X SLA+R will be transmitted;
Logic will switch to master receiver mode
$18 SLA+W has been transmitted; Load data byte or 0 0 1 X Data byte will be transmitted and ACK or NOT ACK will
ACK has been received be received
No TWDR action or 1 0 1 X Repeated START will be transmitted
No TWDR action or 0 1 1 X STOP condition will be transmitted and
TWSTO flag will be reset
No TWDR action 1 1 1 X STOP condition followed by a START condition will be
transmitted and TWSTO flag will be reset
$20 SLA+W has been transmitted; Load data byte or 0 0 1 X Data byte will be transmitted and ACK or NOT ACK will
NOT ACK has been received be received
$28 Data byte has been transmitted; Load data byte or 0 0 1 X Data byte will be transmitted and ACK or NOT ACK will
ACK has been received be received
$30 Data byte has been transmitted; Load data byte or 0 0 1 X Data byte will be transmitted and ACK or NOT ACK will
NOT ACK has been received be received
$38 Arbitration lost in SLA+W or data No TWDR action or 0 0 1 X Two-wire Serial Bus will be released and not addressed
bytes slave mode entered
No TWDR action 1 0 1 X A START condition will be transmitted when the bus be-
comes free
209
2467E–AVR–05/02
Figure 96. Formats and States in the Master Transmitter Mode
MT
Successfull
transmission S SLA W A DATA A P
to a slave
receiver
$08 $18 $28
Next transfer
started with a RS SLA W
repeated start
condition
$10
Not acknowledge R
received after the A P
slave address
$20
MR
Not acknowledge
received after a data A P
byte
$30
Arbitration lost in slave Other master Other master

address or data byte A or A continues A or A continues
$38 $38
Arbitration lost and Other master

addressed as slave A continues
To corresponding
$68 $78 $B0 states in slave mode
Any number of data bytes

From master to slave DATA A and their associated acknowledge bits
From slave to master This number (contained in TWSR) corresponds

n to a defined state of the Two-wire Serial Bus. The
prescaler bits are zero or masked to zero
210 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Master Receiver Mode In the Master Receiver Mode, a number of data bytes are received from a slave trans-
mitter (see Figure 97). In order to enter a Master mode, a START condition must be
transmitted. The format of the following address packet determines whether Master
Transmitter or Master Receiver mode is to be entered. If SLA+W is transmitted, MT
mode is entered, if SLA+R is transmitted, MR mode is entered. All the status codes
mentioned in this section assume that the prescaler bits are zero or are masked to zero.
Figure 97. Data Transfer in Master Receiver Mode

VCC
Device 1 Device 2
MASTER SLAVE Device 3 ........ Device n R1 R2
RECEIVER TRANSMITTER
SDA
SCL
A START condition is sent by writing the following value to TWCR:

value 1 X 1 0 X 1 0 X
TWEN must be written to one to enable the Two-wire Serial Interface, TWSTA must be
written to one to transmit a START condition and TWINT must be set to clear the TWINT
flag. The TWI will then test the Two-wire Serial Bus and generate a START condition as
soon as the bus becomes free. After a START condition has been transmitted, the
TWINT flag is set by hardware, and the status code in TWSR will be $08 (See Table 87).
In order to enter MR mode, SLA+R must be transmitted. This is done by writing SLA+R
to TWDR. Thereafter the TWINT bit should be cleared (by writing it to one) to continue
the transfer. This is accomplished by writing the following value to TWCR:
value 1 X 0 0 X 1 0 X
When SLA+R have been transmitted and an acknowledgment bit has been received,
TWINT is set again and a number of status codes in TWSR are possible. Possible sta-
tus codes in Master mode are $38, $40, or $48. The appropriate action to be taken for
each of these status codes is detailed in Table 96. Received data can be read from the
TWDR Register when the TWINT flag is set high by hardware. This scheme is repeated
until the last byte has been received. After the last byte has been received, the MR
should inform the ST by sending a NACK after the last received data byte. The transfer
is ended by generating a STOP condition or a repeated START condition. A STOP con-
dition is generated by writing the following value to TWCR:
value 1 X 0 1 X 1 0 X
A REPEATED START condition is generated by writing the following value to TWCR:

value 1 X 1 0 X 1 0 X
After a repeated START condition (state $10) the Two-wire Serial Interface can access
the same slave again, or a new slave without transmitting a STOP condition. Repeated
211
2467E–AVR–05/02
START enables the master to switch between slaves, Master Transmitter mode and
Master Receiver mode without losing control over the bus.
Figure 98. Formats and States in the Master Receiver Mode

MR
Successfull
reception S SLA R A DATA A DATA A P
from a slave
receiver
$08 $40 $50 $58
Next transfer
started with a RS SLA R
repeated start
condition
$10
Not acknowledge W
received after the A P
slave address
$48
MT
Arbitration lost in slave Other master Other master
address or data byte A or A continues A continues
$38 $38
Arbitration lost and Other master

addressed as slave A continues
To corresponding
$68 $78 $B0 states in slave mode


212 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 88. Status Codes for Master Receiver Mode

$08 A START condition has been Load SLA+R X 0 1 X SLA+R will be transmitted
transmitted ACK or NOT ACK will be received
$10 A repeated START condition Load SLA+R or X 0 1 X SLA+R will be transmitted
has been transmitted ACK or NOT ACK will be received
Load SLA+W X 0 1 X SLA+W will be transmitted
Logic will switch to master transmitter mode
$38 Arbitration lost in SLA+R or NOT No TWDR action or 0 0 1 X Two-wire Serial Bus will be released and not addressed
ACK bit slave mode will be entered
No TWDR action 1 0 1 X A START condition will be transmitted when the bus
becomes free
$40 SLA+R has been transmitted; No TWDR action or 0 0 1 0 Data byte will be received and NOT ACK will be
ACK has been received returned
No TWDR action 0 0 1 1 Data byte will be received and ACK will be returned
$48 SLA+R has been transmitted; No TWDR action or 1 0 1 X Repeated START will be transmitted
NOT ACK has been received No TWDR action or 0 1 1 X STOP condition will be transmitted and TWSTO flag will
be reset
$50 Data byte has been received; Read data byte or 0 0 1 0 Data byte will be received and NOT ACK will be
ACK has been returned returned
Read data byte 0 0 1 1 Data byte will be received and ACK will be returned
$58 Data byte has been received; Read data byte or 1 0 1 X Repeated START will be transmitted
NOT ACK has been returned Read data byte or 0 1 1 X STOP condition will be transmitted and TWSTO flag will
be reset
Read data byte 1 1 1 X STOP condition followed by a START condition will be
Slave Receiver Mode In the Slave Receiver mode, a number of data bytes are received from a master trans-
mitter (see Figure 99). All the status codes mentioned in this section assume that the
prescaler bits are zero or are masked to zero.
Figure 99. Data Transfer in Slave Receiver Mode

VCC
Device 1 Device 2
SLAVE MASTER Device 3 ........ Device n R1 R2
RECEIVER TRANSMITTER
SDA
SCL
To initiate the Slave Receiver mode, TWAR and TWCR must be initialized as follows:
TWAR TWA6 TWA5 TWA4 TWA3 TWA2 TWA1 TWA0 TWGCE
value Device’s Own Slave Address
213
2467E–AVR–05/02
The upper seven bits are the address to which the Two-wire Serial Interface will respond
when addressed by a master. If the LSB is set, the TWI will respond to the general call
address ($00), otherwise it will ignore the general call address.
value 0 1 0 0 0 1 0 X
TWEN must be written to one to enable the TWI. The TWEA bit must be written to one
to enable the acknowledgment of the device’s own slave address or the general call
address. TWSTA and TWSTO must be written to zero.
When TWAR and TWCR have been initialized, the TWI waits until it is addressed by its
own slave address (or the general call address if enabled) followed by the data direction
bit. If the direction bit is “0” (write), the TWI will operate in SR mode, otherwise ST mode
is entered. After its own slave address and the write bit have been received, the TWINT
flag is set and a valid status code can be read from TWSR. The status code is used to
determine the appropriate software action. The appropriate action to be taken for each
status code is detailed in Table 89. The slave receiver mode may also be entered if arbi-
tration is lost while the TWI is in the master mode (see states $68 and $78).
If the TWEA bit is reset during a transfer, the TWI will return a “Not Acknowledge” (“1”)
to SDA after the next received data byte. This can be used to indicate that the slave is
not able to receive any more bytes. While TWEA is zero, the TWI does not acknowledge
its own slave address. However, the Two-wire Serial Bus is still monitored and address
recognition may resume at any time by setting TWEA. This implies that the TWEA bit
may be used to temporarily isolate the TWI from the Two-wire Serial Bus.
In all sleep modes other than Idle mode, the clock system to the TWI is turned off. If the
TWEA bit is set, the interface can still acknowledge its own slave address or the general
call address by using the Two-wire Serial Bus clock as a clock source. The part will then
wake up from sleep and the TWI will hold the SCL clock low during the wake up and
until the TWINT flag is cleared (by writing it to one). Further data reception will be car-
ried out as normal, with the AVR clocks running as normal. Observe that if the AVR is
set up with a long start-up time, the SCL line may be held low for a long time, blocking
other data transmissions.
Note that the Two-wire Serial Interface Data Register – TWDR does not reflect the last
byte present on the bus when waking up from these sleep modes.
214 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 89. Status Codes for Slave Receiver Mode

(TWSR) Status of the Two-wire Serial Bus To TWCR
Prescaler Bits and Two-wire Serial Interface To/from TWDR
are 0 Hardware STA STO TWINT TWEA Next Action Taken by TWI Hardware
$60 Own SLA+W has been received; No TWDR action or X 0 1 0 Data byte will be received and NOT ACK will be
ACK has been returned returned
No TWDR action X 0 1 1 Data byte will be received and ACK will be returned
$68 Arbitration lost in SLA+R/W as No TWDR action or X 0 1 0 Data byte will be received and NOT ACK will be
master; own SLA+W has been returned
received; ACK has been returned No TWDR action X 0 1 1 Data byte will be received and ACK will be returned
$70 General call address has been No TWDR action or X 0 1 0 Data byte will be received and NOT ACK will be
received; ACK has been returned returned
No TWDR action X 0 1 1 Data byte will be received and ACK will be returned
$78 Arbitration lost in SLA+R/W as No TWDR action or X 0 1 0 Data byte will be received and NOT ACK will be
master; General call address has returned
been received; ACK has been No TWDR action X 0 1 1 Data byte will be received and ACK will be returned
returned
$80 Previously addressed with own Read data byte or X 0 1 0 Data byte will be received and NOT ACK will be
SLA+W; data has been received; returned
ACK has been returned Read data byte X 0 1 1 Data byte will be received and ACK will be returned
$88 Previously addressed with own Read data byte or 0 0 1 0 Switched to the not addressed slave mode;
SLA+W; data has been received; no recognition of own SLA or GCA
NOT ACK has been returned Read data byte or 0 0 1 1 Switched to the not addressed slave mode;
own SLA will be recognized;
GCA will be recognized if TWGCE = “1”
Read data byte or 1 0 1 0 Switched to the not addressed slave mode;
no recognition of own SLA or GCA;
a START condition will be transmitted when the bus
becomes free
Read data byte 1 0 1 1 Switched to the not addressed slave mode;
GCA will be recognized if TWGCE = “1”;
becomes free
$90 Previously addressed with Read data byte or X 0 1 0 Data byte will be received and NOT ACK will be
general call; data has been re- returned
ceived; ACK has been returned Read data byte X 0 1 1 Data byte will be received and ACK will be returned
$98 Previously addressed with Read data byte or 0 0 1 0 Switched to the not addressed slave mode;
general call; data has been no recognition of own SLA or GCA
received; NOT ACK has been Read data byte or 0 0 1 1 Switched to the not addressed slave mode;
returned own SLA will be recognized;
becomes free
becomes free
$A0 A STOP condition or repeated Read data byte or 0 0 1 0 Switched to the not addressed slave mode;
START condition has been no recognition of own SLA or GCA
received while still addressed as Read data byte or 0 0 1 1 Switched to the not addressed slave mode;
slave own SLA will be recognized;
becomes free
becomes free
215
2467E–AVR–05/02
Figure 100. Formats and States in the Slave Receiver Mode
Reception of the
own slave address S SLA W A DATA A DATA A P or S
and one or more
data bytes. All are
acknowledged
$60 $80 $80 $A0
Last data byte received

is not acknowledged A P or S
$88
Arbitration lost as master

and addressed as slave A
$68
Reception of the general call

address and one or more data General Call A DATA A DATA A P or S
bytes
$70 $90 $90 $A0
Last data byte received is

not acknowledged A P or S
$98
Arbitration lost as master and

addressed as slave by general call A
$78


Slave Transmitter Mode In the Slave Transmitter mode, a number of data bytes are transmitted to a master
receiver (see Figure 101). All the status codes mentioned in this section assume that the
prescaler bits are zero or are masked to zero.
Figure 101. Data Transfer in Slave Transmitter Mode

VCC
Device 1 Device 2
SLAVE MASTER Device 3 ........ Device n R1 R2
TRANSMITTER RECEIVER
SDA
SCL
To initiate the Slave Transmitter mode, TWAR and TWCR must be initialized as follows:
216 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
TWAR TWA6 TWA5 TWA4 TWA3 TWA2 TWA1 TWA0 TWGCE

value Device’s Own Slave Address
The upper seven bits are the address to which the Two-wire Serial Interface will respond
when addressed by a master. If the LSB is set, the TWI will respond to the general call
address ($00), otherwise it will ignore the general call address.
value 0 1 0 0 0 1 0 X
TWEN must be written to one to enable the TWI. The TWEA bit must be written to one
to enable the acknowledgment of the device’s own slave address or the general call
address. TWSTA and TWSTO must be written to zero.
When TWAR and TWCR have been initialized, the TWI waits until it is addressed by its
own slave address (or the general call address if enabled) followed by the data direction
bit. If the direction bit is “1” (read), the TWI will operate in ST mode, otherwise SR mode
is entered. After its own slave address and the write bit have been received, the TWINT
flag is set and a valid status code can be read from TWSR. The status code is used to
determine the appropriate software action. The appropriate action to be taken for each
status code is detailed in Table 90. The Slave Transmitter mode may also be entered if
arbitration is lost while the TWI is in the Master mode (see state $B0).
If the TWEA bit is written to zero during a transfer, the TWI will transmit the last byte of
the transfer. State $C0 or state $C8 will be entered, depending on whether the master
receiver transmits a NACK or ACK after the final byte. The TWI is switched to the not
addressed slave mode, and will ignore the master if it continues the transfer. Thus the
master receiver receives all “1” as serial data. State $C8 is entered if the master
demands additional data bytes (by transmitting ACK), even though the slave has trans-
mitted the last byte (TWEA zero and expecting NACK from the master).
While TWEA is zero, the TWI does not respond to its own slave address. However, the
Two-wire Serial Bus is still monitored and address recognition may resume at any time
by setting TWEA. This implies that the TWEA bit may be used to temporarily isolate the
TWI from the Two-wire Serial Bus.
In all sleep modes other than Idle mode, the clock system to the TWI is turned off. If the
TWEA bit is set, the interface can still acknowledge its own slave address or the general
call address by using the Two-wire Serial Bus clock as a clock source. The part will then
wake up from sleep and the TWI will hold the SCL clock will low during the wake up and
until the TWINT flag is cleared (by writing it to one). Further data transmission will be
carried out as normal, with the AVR clocks running as normal. Observe that if the AVR is
set up with a long start-up time, the SCL line may be held low for a long time, blocking
other data transmissions.
Note that the Two-wire Serial Interface Data Register – TWDR does not reflect the last
byte present on the bus when waking up from these sleep modes.
217
2467E–AVR–05/02
Table 90. Status Codes for Slave Transmitter Mode
(TWSR) Status of the Two-wire Serial Bus To TWCR
Prescaler Bits and Two-wire Serial Interface To/from TWDR
are 0 Hardware STA STO TWINT TWEA Next Action Taken by TWI Hardware
$A8 Own SLA+R has been received; Load data byte or X 0 1 0 Last data byte will be transmitted and NOT ACK should
ACK has been returned be received
Load data byte X 0 1 1 Data byte will be transmitted and ACK should be re-
ceived
$B0 Arbitration lost in SLA+R/W as Load data byte or X 0 1 0 Last data byte will be transmitted and NOT ACK should
master; own SLA+R has been be received
received; ACK has been returned Load data byte X 0 1 1 Data byte will be transmitted and ACK should be re-
ceived
$B8 Data byte in TWDR has been Load data byte or X 0 1 0 Last data byte will be transmitted and NOT ACK should
transmitted; ACK has been be received
received Load data byte X 0 1 1 Data byte will be transmitted and ACK should be re-
ceived
$C0 Data byte in TWDR has been No TWDR action or 0 0 1 0 Switched to the not addressed slave mode;
transmitted; NOT ACK has been no recognition of own SLA or GCA
received No TWDR action or 0 0 1 1 Switched to the not addressed slave mode;
No TWDR action or 1 0 1 0 Switched to the not addressed slave mode;
becomes free
No TWDR action 1 0 1 1 Switched to the not addressed slave mode;
becomes free
$C8 Last data byte in TWDR has been No TWDR action or 0 0 1 0 Switched to the not addressed slave mode;
transmitted (TWEA = “0”); ACK no recognition of own SLA or GCA
has been received No TWDR action or 0 0 1 1 Switched to the not addressed slave mode;
No TWDR action or 1 0 1 0 Switched to the not addressed slave mode;
becomes free
No TWDR action 1 0 1 1 Switched to the not addressed slave mode;
becomes free
Figure 102. Formats and States in the Slave Transmitter Mode

Reception of the
own slave address S SLA R A DATA A DATA A P or S
and one or
more data bytes
$A8 $B8 $C0
Arbitration lost as master

and addressed as slave A
$B0
Last data byte transmitted.

Switched to not addressed A All 1's P or S
slave (TWEA = '0')
$C8


218 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Miscellaneous States There are two status codes that do not correspond to a defined TWI state, see Table 91.
Status $F8 indicates that no relevant information is available because the TWINT flag is
not set. This occurs between other states, and when the TWI is not involved in a serial
transfer.
Status $00 indicates that a bus error has occurred during a Two-wire Serial Bus trans-
fer. A bus error occurs when a START or STOP condition occurs at an illegal position in
the format frame. Examples of such illegal positions are during the serial transfer of an
address byte, a data byte, or an acknowledge bit. When a bus error occurs, TWINT is
set. To recover from a bus error, the TWSTO flag must set and TWINT must be cleared
by writing a logic one to it. This causes the TWI to enter the not addressed slave mode
and to clear the TWSTO flag (no other bits in TWCR are affected). The SDA and SCL
lines are released, and no STOP condition is transmitted.
Table 91. Miscellaneous States

$F8 No relevant state information No TWDR action No TWCR action Wait or proceed current transfer
available; TWINT = “0”
$00 Bus error due to an illegal No TWDR action 0 1 1 X Only the internal hardware is affected, no STOP condi-
START or STOP condition tion is sent on the bus. In all cases, the bus is released
and TWSTO is cleared.
Combining Several TWI In some cases, several TWI modes must be combined in order to complete the desired
Modes action. Consider for example reading data from a serial EEPROM. Typically, such a
transfer involves the following steps:
1. The transfer must be initiated
2. The EEPROM must be instructed what location should be read
3. The reading must be performed
4. The transfer must be finished
Note that data is transmitted both from master to slave and vice versa. The master must
instruct the slave what location it wants to read, requiring the use of the MT mode. Sub-
sequently, data must be read from the slave, implying the use of the MR mode. Thus,
the transfer direction must be changed. The master must keep control of the bus during
all these steps, and the steps should be carried out as an atomical operation. If this prin-
ciple is violated in a multimaster system, another master can alter the data pointer in the
EEPROM between steps 2 and 3, and the master will read the wrong data location.
Such a change in transfer direction is accomplished by transmitting a REPEATED
START between the transmission of the address byte and reception of the data. After a
REPEATED START, the master keeps ownership of the bus. The following figure shows
the flow in this transfer.
Figure 103. Combining Several TWI Modes to Access a Serial EEPROM

Master Transmitter Master Receiver
S SLA+W A ADDRESS A Rs SLA+R A DATA A P
S = START Rs = REPEATED START P = STOP
Transmitted from master to slave Transmitted from slave to master
219
2467E–AVR–05/02
Multi-master Systems If multiple masters are connected to the same bus, transmissions may be initiated simul-
and Arbitration taneously by one or more of them. The TWI standard ensures that such situations are
handled in such a way that one of the masters will be allowed to proceed with the trans-
fer, and that no data will be lost in the process. An example of an arbitration situation is
depicted below, where two masters are trying to transmit data to a slave receiver.
Figure 104. An Arbitration Example
VCC
Device 1 Device 2 Device 3

MASTER MASTER SLAVE ........ Device n R1 R2
TRANSMITTER TRANSMITTER RECEIVER
SDA
SCL
Several different scenarios may arise during arbitration, as described below:

• Two or more masters are performing identical communication with the same slave.
In this case, neither the slave nor any of the masters will know about the bus
contention.
• Two or more masters are accessing the same slave with different data or direction
bit. In this case, arbitration will occur, either in the READ/WRITE bit or in the data
bits. The masters trying to output a one on SDA while another master outputs a zero
will lose the arbitration. Losing masters will switch to not addressed slave mode or
wait until the bus is free and transmit a new START condition, depending on
application software action.
• Two or more masters are accessing different slaves. In this case, arbitration will
occur in the SLA bits. Masters trying to output a one on SDA while another master
outputs a zero will lose the arbitration. Masters losing arbitration in SLA will switch to
slave mode to check if they are being addressed by the winning master. If
addressed, they will switch to SR or ST mode, depending on the value of the
READ/WRITE bit. If they are not being addressed, they will switch to not addressed
slave mode or wait until the bus is free and transmit a new START condition,
depending on application software action.
This is summarized in Figure 105. Possible status values are given in circles.
220 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 105. Possible Status Codes Caused by Arbitration
START SLA Data STOP
Arbitration lost in SLA Arbitration lost in Data
Own No 38 TWI bus will be released and not addressed slave mode will be entered
Address / General Call
A START condition will be transmitted when the bus becomes free
received
Yes
Direction
Write 68/78 Data byte will be received and NOT ACK will be returned
Data byte will be received and ACK will be returned
Read Last data byte will be transmitted and NOT ACK should be received
B0 Data byte will be transmitted and ACK should be received
221
2467E–AVR–05/02
Analog Comparator The Analog Comparator compares the input values on the positive pin AIN0 and nega-
tive pin AIN1. When the voltage on the positive pin AIN0 is higher than the voltage on
the negative pin AIN1, the Analog Comparator Output, ACO, is set. The comparator’s
output can be set to trigger the Timer/Counter1 Input Capture function. In addition, the
comparator can trigger a separate interrupt, exclusive to the Analog Comparator. The
user can select Interrupt triggering on comparator output rise, fall or toggle. A block dia-
gram of the comparator and its surrounding logic is shown in Figure 106.
Figure 106. Analog Comparator Block Diagram

BANDGAP
REFERENCE
ACBG
ACME
ADEN
ADC MULTIPLEXER
OUTPUT1)
Notes: 1. See Table 93 on page 224.

2. Refer to Figure 1 on page 2 and Table 30 on page 68 for Analog Comparator pin
placement.

SFIOR Bit 7 6 5 4 3 2 1 0
• Bit 3 – ACME: Analog Comparator Multiplexer Enable

When this bit is written logic one and the ADC is switched off (ADEN in ADCSRA is
zero), the ADC multiplexer selects the negative input to the Analog Comparator. When
this bit is written logic zero, AIN1 is applied to the negative input of the Analog Compar-
ator. For a detailed description of this bit, see “Analog Comparator Multiplexed Input” on
page 224.
Analog Comparator Control

and Status Register – ACSR Bit 7 6 5 4 3 2 1 0
ACD ACBG ACO ACI ACIE ACIC ACIS1 ACIS0 ACSR
Initial Value 0 0 N/A 0 0 0 0 0
222 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
• Bit 7 – ACD: Analog Comparator Disable

When this bit is written logic one, the power to the Analog Comparator is switched off.
This bit can be set at any time to turn off the Analog Comparator. This will reduce power
consumption in Active and Idle mode. When changing the ACD bit, the Analog Compar-
ator Interrupt must be disabled by clearing the ACIE bit in ACSR. Otherwise an interrupt
can occur when the bit is changed.
• Bit 6 – ACBG: Analog Comparator Bandgap Select

When this bit is set, a fixed bandgap reference voltage replaces the positive input to the
Analog Comparator. When this bit is cleared, AIN0 is applied to the positive input of the
Analog Comparator. See “Internal Voltage Reference” on page 50.
• Bit 5 – ACO: Analog Comparator Output

The output of the Analog Comparator is synchronized and then directly connected to
ACO. The synchronization introduces a delay of 1 – 2 clock cycles.
• Bit 4 – ACI: Analog Comparator Interrupt Flag

This bit is set by hardware when a comparator output event triggers the interrupt mode
defined by ACIS1 and ACIS0. The Analog Comparator Interrupt routine is executed if
the ACIE bit is set and the I-bit in SREG is set. ACI is cleared by hardware when execut-
ing the corresponding interrupt handling vector. Alternatively, ACI is cleared by writing a
logic one to the flag.
• Bit 3 – ACIE: Analog Comparator Interrupt Enable

When the ACIE bit is written logic one and the I-bit in the Status Register is set, the Ana-
log Comparator interrupt is activated. When written logic zero, the interrupt is disabled.
• Bit 2 – ACIC: Analog Comparator Input Capture Enable

When written logic one, this bit enables the Input Capture function in Timer/Counter1 to
be triggered by the Analog Comparator. The comparator output is in this case directly
connected to the Input Capture front-end logic, making the comparator utilize the noise
canceler and edge select features of the Timer/Counter1 Input Capture interrupt. When
written logic zero, no connection between the analog comparator and the Input Capture
function exists. To make the comparator trigger the Timer/Counter1 Input Capture inter-
rupt, the TICIE1 bit in the Timer Interrupt Mask Register (TIMSK) must be set.
• Bits 1, 0 – ACIS1, ACIS0: Analog Comparator Interrupt Mode Select

These bits determine which comparator events that trigger the Analog Comparator inter-
rupt. The different settings are shown in Table 92.
Table 92. ACIS1/ACIS0 Settings

ACIS1 ACIS0 Interrupt Mode
0 0 Comparator Interrupt on Output Toggle
0 1 Reserved
1 0 Comparator Interrupt on Falling Output Edge
1 1 Comparator Interrupt on Rising Output Edge
223
2467E–AVR–05/02
When changing the ACIS1/ACIS0 bits, the Analog Comparator Interrupt must be dis-
abled by clearing its Interrupt Enable bit in the ACSR Register. Otherwise an interrupt
can occur when the bits are changed.
Analog Comparator It is possible to select any of the ADC7..0 pins to replace the negative input to the Ana-
Multiplexed Input log Comparator. The ADC multiplexer is used to select this input, and consequently, the
ADC must be switched off to utilize this feature. If the Analog Comparator Multiplexer
Enable bit (ACME in SFIOR) is set and the ADC is switched off (ADEN in ADCSRA is
zero), MUX2..0 in ADMUX select the input pin to replace the negative input to the Ana-
log Comparator, as shown in Table 93. If ACME is cleared or ADEN is set, AIN1 is
applied to the negative input to the Analog Comparator.
Table 93. Analog Comparator Multiplexed Input

ACME ADEN MUX2..0 Analog Comparator Negative Input
0 x xxx AIN1
1 1 xxx AIN1
1 0 000 ADC0
1 0 001 ADC1
1 0 010 ADC2
1 0 011 ADC3
1 0 100 ADC4
1 0 101 ADC5
1 0 110 ADC6
1 0 111 ADC7
224 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Analog to Digital
Converter
Features • 10-bit Resolution

• 0.5 LSB Integral Non-linearity
• ±2 LSB Absolute Accuracy
• 65 - 260 µs Conversion Time
• Up to 15 kSPS at Maximum Resolution
• 8 Multiplexed Single Ended Input Channels
• 7 Differential Input Channels
• 2 Differential Input Channels with Optional Gain of 10x and 200x
• Optional Left Adjustment for ADC Result Readout
• 0 - VCC ADC Input Voltage Range
• Selectable 2.56 V ADC Reference Voltage
• Free Running or Single Conversion Mode
• Interrupt on ADC Conversion Complete
• Sleep Mode Noise Canceler
The ATmega128 features a 10-bit successive approximation ADC. The ADC is con-
nected to an 8-channel Analog Multiplexer which allows 8 single-ended voltage inputs
constructed from the pins of Port F. The single-ended voltage inputs refer to 0V (GND).
The device also supports 16 differential voltage input combinations. Two of the differen-
tial inputs (ADC1, ADC0 and ADC3, ADC2) are equipped with a programmable gain
stage, providing amplification steps of 0 dB (1x), 20 dB (10x), or 46 dB (200x) on the dif-
ferential input voltage before the A/D conversion. Seven differential analog input
channels share a common negative terminal (ADC1), while any other ADC input can be
selected as the positive input terminal. If 1x or 10x gain is used, 8-bit resolution can be
expected. If 200x gain is used, 7-bit resolution can be expected.
The ADC contains a Sample and Hold circuit which ensures that the input voltage to the
ADC is held at a constant level during conversion. A block diagram of the ADC is shown
in Figure 107.
The ADC has a separate analog supply voltage pin, AVCC. AVCC must not differ more
than ± 0.3V from VCC. See the paragraph “ADC Noise Canceler” on page 231 on how to
connect this pin.
Internal reference voltages of nominally 2.56V or AVCC are provided On-chip. The volt-
age reference may be externally decoupled at the AREF pin by a capacitor for better
noise performance.
225
2467E–AVR–05/02
Figure 107. Analog to Digital Converter Block Schematic
ADC CONVERSION
COMPLETE IRQ
8-BIT DATA BUS
ADIE
ADIF
15 0
ADC MULTIPLEXER ADC CTRL. & STATUS ADC DATA REGISTER
SELECT (ADMUX) REGISTER (ADCSRA) (ADCH/ADCL)
REFS1
ADLAR
MUX4
MUX3
MUX2
MUX1
MUX0
REFS0
ADPS2
ADPS1
ADPS0
ADEN
ADSC
ADFR
ADIF
ADC[9:0]
PRESCALER
MUX DECODER
CHANNEL SELECTION
GAIN SELECTION
CONVERSION LOGIC
AVCC
INTERNAL 2.56V
REFERENCE SAMPLE & HOLD
COMPARATOR
AREF
10-BIT DAC -
+
ADHSM
AGND
BANDGAP
REFERENCE
ADC7
SINGLE ENDED / DIFFERENTIAL SELECTION
ADC6
POS. ADC MULTIPLEXER

ADC5 INPUT OUTPUT
MUX
ADC4
ADC3 GAIN
AMPLIFIER
+
ADC2
-
ADC1
ADC0
NEG.
INPUT
MUX
226 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Operation The ADC converts an analog input voltage to a 10-bit digital value through successive
approximation. The minimum value represents GND and the maximum value represents
the voltage on the AREF pin minus 1 LSB. Optionally, AVCC or an internal 2.56V refer-
ence voltage may be connected to the AREF pin by writing to the REFSn bits in the
ADMUX Register. The internal voltage reference may thus be decoupled by an external
capacitor at the AREF pin to improve noise immunity.
The analog input channel and differential gain are selected by writing to the MUX bits in
ADMUX. Any of the ADC input pins, as well as GND and a fixed bandgap voltage refer-
ence, can be selected as single ended inputs to the ADC. A selection of ADC input pins
can be selected as positive and negative inputs to the differential gain amplifier.
If differential channels are selected, the differential gain stage amplifies the voltage dif-
ference between the selected input channel pair by the selected gain factor. This
amplified value then becomes the analog input to the ADC. If single ended channels are
used, the gain amplifier is bypassed altogether.
The ADC is enabled by setting the ADC Enable bit, ADEN in ADCSRA. Voltage refer-
ence and input channel selections will not go into effect until ADEN is set. The ADC
does not consume power when ADEN is cleared, so it is recommended to switch off the
ADC before entering power saving sleep modes.
The ADC generates a 10-bit result which is presented in the ADC Data Registers,
ADCH and ADCL. By default, the result is presented right adjusted, but can optionally
be presented left adjusted by setting the ADLAR bit in ADMUX.
If the result is left adjusted and no more than 8-bit precision is required, it is sufficient to
read ADCH. Otherwise, ADCL must be read first, then ADCH, to ensure that the content
of the data registers belongs to the same conversion. Once ADCL is read, ADC access
to data registers is blocked. This means that if ADCL has been read, and a conversion
completes before ADCH is read, neither register is updated and the result from the con-
version is lost. When ADCH is read, ADC access to the ADCH and ADCL Registers is
re-enabled.
The ADC has its own interrupt which can be triggered when a conversion completes.
When ADC access to the data registers is prohibited between reading of ADCH and
ADCL, the interrupt will trigger even if the result is lost.
Starting a Conversion A single conversion is started by writing a logical one to the ADC Start Conversion bit,
ADSC. This bit stays high as long as the conversion is in progress and will be cleared by
hardware when the conversion is completed. If a different data channel is selected while
a conversion is in progress, the ADC will finish the current conversion before performing
the channel change.
In Free Running mode, the ADC is constantly sampling and updating the ADC Data
Register. Free Running mode is selected by writing the ADFR bit in ADCSRA to one.
The first conversion must be started by writing a logical one to the ADSC bit in ADC-
SRA. In this mode the ADC will perform successive conversions independently of
whether the ADC Interrupt Flag, ADIF is cleared or not.
227
2467E–AVR–05/02
Prescaling and Figure 108. ADC Prescaler
Conversion Timing
ADEN Reset
7-BIT ADC PRESCALER
CK
CK/128
CK/64
CK/32
CK/16
CK/2
CK/4
CK/8
ADPS0
ADPS1
ADPS2
ADC CLOCK SOURCE
By default, the successive approximation circuitry requires an input clock frequency

between 50 kHz and 200 kHz to get maximum resolution. If a lower resolution than 10
bits is needed, the input clock frequency to the ADC can be higher than 200 kHz to get a
higher sample rate. Alternatively, setting the ADHSM bit in SFIOR allows an increased
ADC clock frequency at the expense of higher power consumption.
The ADC module contains a prescaler, which generates an acceptable ADC clock fre-
quency from any CPU frequency above 100 kHz. The prescaling is set by the ADPS bits
in ADCSRA. The prescaler starts counting from the moment the ADC is switched on by
setting the ADEN bit in ADCSRA. The prescaler keeps running for as long as the ADEN
bit is set, and is continuously reset when ADEN is low.
When initiating a single ended conversion by setting the ADSC bit in ADCSRA, the con-
version starts at the following rising edge of the ADC clock cycle. See “Differential Gain
Channels” on page 230 for details on differential conversion timing.
A normal conversion takes 13 ADC clock cycles. The first conversion after the ADC is
switched on (ADEN in ADCSRA is set) takes 25 ADC clock cycles in order to initialize
the analog circuitry.
The actual sample-and-hold takes place 1.5 ADC clock cycles after the start of a normal
conversion and 13.5 ADC clock cycles after the start of an first conversion. When a con-
version is complete, the result is written to the ADC data registers, and ADIF is set. In
single conversion mode, ADSC is cleared simultaneously. The software may then set
ADSC again, and a new conversion will be initiated on the first rising ADC clock edge.
In Free Running mode, a new conversion will be started immediately after the conver-
sion completes, while ADSC remains high. For a summary of conversion times, see
Table 94.
228 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 109. ADC Timing Diagram, First Conversion (Single Conversion Mode)
Next
First Conversion Conversion
Cycle Number 1 2 12 13 14 15 16 17 18 19 20 21 22 23 24 25 1 2 3
ADC Clock
ADEN
ADSC
ADIF
ADCH MSB of Result
ADCL LSB of Result
MUX and REFS Conversion

Update Sample &Hold Complete MUX and REFS
Update
Figure 110. ADC Timing Diagram, Single Conversion

One Conversion Next Conversion
Cycle Number 1 2 3 4 5 6 7 8 9 10 11 12 13 1 2 3
ADC Clock
ADSC
ADIF
ADCH MSB of Result
ADCL LSB of Result
Sample & Hold Conversion

MUX and REFS Complete MUX and REFS
Update Update
Figure 111. ADC Timing Diagram, Free Running Conversion

One Conversion Next Conversion
11 12 13 1 2 3 4
Cycle Number
ADC Clock
ADSC
ADIF
ADCH MSB of Result
ADCL LSB of Result
Conversion Sample & Hold

Complete MUX and REFS
Update
229
2467E–AVR–05/02
Table 94. ADC Conversion Time
Sample & Hold (Cycles from Conversion Time
Condition Start of Conversion) (Cycles)
First conversion 14.5 25
Normal conversions, single ended 1.5 13
Normal conversions, differential 1.5/2.5 13/14
Differential Gain Channels When using differential gain channels, certain aspects of the conversion need to be
taken into consideration.
Differential conversions are synchronized to the internal clock CKADC2 equal to half the
ADC clock. This synchronization is done automatically by the ADC interface in such a
way that the sample-and-hold occurs at a specific edge of CKADC2. A conversion initi-
ated by the user (i.e., all single conversions, and the first free running conversion) when
CKADC2 is low will take the same amount of time as a single ended conversion (13 ADC
clock cycles from the next prescaled clock cycle). A conversion initiated by the user
when CKADC2 is high will take 14 ADC clock cycles due to the synchronization mecha-
nism. In free running mode, a new conversion is initiated immediately after the previous
conversion completes, and since CKADC2 is high at this time, all automatically started
(i.e., all but the first) free running conversions will take 14 ADC clock cycles.
The gain stage is optimized for a bandwidth of 4 kHz at all gain settings. Higher frequen-
cies may be subjected to non-linear amplification. An external low-pass filter should be
used if the input signal contains higher frequency components than the gain stage band-
width. Note that the ADC clock frequency is independent of the gain stage bandwidth
limitation. E.g. the ADC clock period may be 6 µs, allowing a channel to be sampled at
12 kSPS, regardless of the bandwidth of this channel.
Changing Channel or The MUXn and REFS1:0 bits in the ADMUX Register are single buffered through a tem-
Reference Selection porary register to which the CPU has random access. This ensures that the channels
and reference selection only takes place at a safe point during the conversion. The
channel and reference selection is continuously updated until a conversion is started.
Once the conversion starts, the channel and reference selection is locked to ensure a
sufficient sampling time for the ADC. Continuous updating resumes in the last ADC
clock cycle before the conversion completes (ADIF in ADCSRA is set). Note that the
conversion starts on the following rising ADC clock edge after ADSC is written. The user
is thus advised not to write new channel or reference selection values to ADMUX until
one ADC clock cycle after ADSC is written.
Special care should be taken when changing differential channels. Once a differential
channel has been selected, the gain stage may take as much as 125 µs to stabilize to
the new value. Thus conversions should not be started within the first 125 µs after
selecting a new differential channel. Alternatively, conversion results obtained within this
period should be discarded.
The same settling time should be observed for the first differential conversion after
changing ADC reference (by changing the REFS1:0 bits in ADMUX).
The settling time and gain stage bandwidth is independent of the ADHSM bit setting.
ADC Input Channels When changing channel selections, the user should observe the following guidelines to
ensure that the correct channel is selected:
230 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
In Single Conversion mode, always select the channel before starting the conversion.
The channel selection may be changed one ADC clock cycle after writing one to ADSC.
However, the simplest method is to wait for the conversion to complete before changing
the channel selection.
In Free Running mode, always select the channel before starting the first conversion.
The channel selection may be changed one ADC clock cycle after writing one to ADSC.
However, the simplest method is to wait for the first conversion to complete, and then
change the channel selection. Since the next conversion has already started automati-
cally, the next result will reflect the previous channel selection. Subsequent conversions
will reflect the new channel selection.
When switching to a differential gain channel, the first conversion result may have a
poor accuracy due to the required settling time for the automatic offset cancellation cir-
cuitry. The user should preferably disregard the first conversion result.
ADC Voltage Reference The reference voltage for the ADC (VREF) indicates the conversion range for the ADC.
Single ended channels that exceed VREF will result in codes close to 0x3FF. VREF can be
selected as either AVCC, internal 2.56V reference, or external AREF pin.
AVCC is connected to the ADC through a passive switch. The internal 2.56V reference
is generated from the internal bandgap reference (VBG) through an internal amplifier. In
either case, the external AREF pin is directly connected to the ADC, and the reference
voltage can be made more immune to noise by connecting a capacitor between the
AREF pin and ground. VREF can also be measured at the AREF pin with a high impedant
voltmeter. Note that VREF is a high impedant source, and only a capacitive load should
be connected in a system.
If the user has a fixed voltage source connected to the AREF pin, the user may not use
the other reference voltage options in the application, as they will be shorted to the
external voltage. If no external voltage is applied to the AREF pin, the user may switch
between AVCC and 2.56 V as reference selection. The first ADC conversion result after
switching reference voltage source may be inaccurate, and the user is advised to dis-
card this result.
If differential channels are used, the selected reference should not be closer to AVCC
than indicated in Table 136 on page 321.
ADC Noise Canceler The ADC features a noise canceler that enables conversion during sleep mode to
reduce noise induced from the CPU core and other I/O peripherals. The noise canceler
can be used with ADC Noise Reduction and Idle mode. To make use of this feature, the
following procedure should be used:
1. Make sure that the ADC is enabled and is not busy converting. Single Con-
version mode must be selected and the ADC conversion complete interrupt
must be enabled.
2. Enter ADC Noise Reduction mode (or Idle mode). The ADC will start a con-
version once the CPU has been halted.
3. If no other interrupts occur before the ADC conversion completes, the ADC
interrupt will wake up the CPU and execute the ADC Conversion Complete
interrupt routine. If another interrupt wakes up the CPU before the ADC con-
version is complete, that interrupt will be executed, and an ADC Conversion
Complete interrupt request will be generated when the ADC conversion
completes. The CPU will remain in active mode until a new sleep command
is executed.
231
2467E–AVR–05/02
Note that the ADC will not be automatically turned off when entering other sleep modes
than Idle mode and ADC Noise Reduction mode. The user is advised to write zero to
ADEN before entering such sleep modes to avoid excessive power consumption. If the
ADC is enabled in such sleep modes and the user wants to perform differential conver-
sions, the user is advised to switch the ADC off and on after waking up from sleep to
prompt an extended conversion to get a valid result.
Analog Input Circuitry The Analog Input circuitry for single ended channels is illustrated in Figure 112. An ana-
log source applied to ADCn is subjected to the pin capacitance and input leakage of that
pin, regardless of whether that channel is selected as input for the ADC. When the chan-
nel is selected, the source must drive the S/H capacitor through the series resistance
(combined resistance in the input path).
The ADC is optimized for analog signals with an output impedance of approximately
10 kΩ or less. If such a source is used, the sampling time will be negligible. If a source
with higher impedance is used, the sampling time will depend on how long time the
source needs to charge the S/H capacitor, with can vary widely. The user is recom-
mended to only use high impedant sources with slowly varying signals, since this
minimizes the required charge transfer to the S/H capacitor.
If differential gain channels are used, the input circuitry looks somewhat different,
although source impedances of a few hundred kΩ or less is recommended.
Signal components higher than the Nyquist frequency (fADC / 2) should not be present
for either kind of channels, to avoid distortion from unpredictable signal convolution. The
user is advised to remove high frequency components with a low-pass filter before
applying the signals as inputs to the ADC.
Figure 112. Analog Input Circuitry
IIH
ADCn
1..100 kΩ
CS/H= 14 pF
IIL
VCC/2
Analog Noise Canceling Digital circuitry inside and outside the device generates EMI which might affect the
Techniques accuracy of analog measurements. If conversion accuracy is critical, the noise level can
be reduced by applying the following techniques:
1. Keep analog signal paths as short as possible. Make sure analog tracks run
over the analog ground plane, and keep them well away from high-speed
switching digital tracks.
2. The AVCC pin on the device should be connected to the digital VCC supply
voltage via an LC network as shown in Figure 113.
3. Use the ADC noise canceler function to reduce induced noise from the CPU.
4. If any ADC port pins are used as digital outputs, it is essential that these do
not switch while a conversion is in progress.
232 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 113. ADC Power Connections
(AD0) PA0 51
VCC
52
GND 53
(ADC7) PF7 54
(ADC6) PF6 55
(ADC5) PF5 56
(ADC4) PF4 57
(ADC3) PF3 58
(ADC2) PF2 59
(ADC1) PF1 60
(ADC0) PF0 61
10µΗ
AREF 62
GND
63
AVCC
64
100nF
1
Analog Ground Plane
PEN
Offset Compensation The gain stage has a built-in offset cancellation circuitry that nulls the offset of differen-
Schemes tial measurements as much as possible. The remaining offset in the analog path can be
measured directly by selecting the same channel for both differential inputs. This offset
residue can be then subtracted in software from the measurement results. Using this
kind of software based offset correction, offset on any channel can be reduced below
one LSB.
ADC Accuracy Definitions An n-bit single-ended ADC converts a voltage linearly between GND and VREF in 2n
steps (LSBs). The lowest code is read as 0, and the highest code is read as 2n-1.
Several parameters describe the deviation from the ideal behavior:
• Offset: The deviation of the first transition (0x000 to 0x001) compared to the ideal
transition (at 0.5 LSB). Ideal value: 0 LSB.
233
2467E–AVR–05/02
Figure 114. Offset Error
Output Code
Ideal ADC
Actual ADC
Offset
Error
VREF Input Voltage
• Gain Error: After adjusting for offset, the gain error is found as the deviation of the
last transition (0x3FE to 0x3FF) compared to the ideal transition (at 1.5 LSB below
maximum). Ideal value: 0 LSB
Figure 115. Gain Error

Output Code Gain
Error
Ideal ADC
Actual ADC
VREF Input Voltage
234 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
• Integral Non-linearity (INL): After adjusting for offset and gain error, the INL is the
maximum deviation of an actual transition compared to an ideal transition for any
code. Ideal value: 0 LSB.
Figure 116. Integral Non-linearity (INL)

Output Code
INL
Ideal ADC
Actual ADC
VREF Input Voltage
• Differential Non-linearity (DNL): The maximum deviation of the actual code width
(the interval between two adjacent transitions) from the ideal code width (1 LSB).
Ideal value: 0 LSB.
Figure 117. Differential Non-linearity (DNL)

Output Code
0x3FF
1 LSB
DNL
0x000
0 VREF Input Voltage
• Quantization Error: Due to the quantization of the input voltage into a finite number
of codes, a range of input voltages (1 LSB wide) will code to the same value. Always
±0.5 LSB.
• Absolute Accuracy: The maximum deviation of an actual (unadjusted) transition
compared to an ideal transition for any code. This is the compound effect of offset,
gain error, differential error, non-linearity, and quantization error. Ideal value: ±0.5
LSB.
235
2467E–AVR–05/02
ADC Conversion Result After the conversion is complete (ADIF is high), the conversion result can be found in
the ADC Result Registers (ADCL, ADCH).
For single ended conversion, the result is
V IN ⋅ 1024
ADC = --------------------------
V REF
where VIN is the voltage on the selected input pin and VREF the selected voltage refer-
ence (see Table 96 on page 238 and Table 97 on page 238). 0x000 represents analog
ground, and 0x3FF represents the selected reference voltage minus one LSB.
If differential channels are used, the result is
( V POS – V NEG ) ⋅ GAIN ⋅ 512

ADC = -----------------------------------------------------------------------
-
V REF
where VPOS is the voltage on the positive input pin, V NEG the voltage on the negative
input pin, GAIN the selected gain factor, and VREF the selected voltage reference. The
result is presented in two’s complement form, from 0x200 (-512d) through 0x1FF
(+511d). Note that if the user wants to perform a quick polarity check of the results, it is
sufficient to read the MSB of the result (ADC9 in ADCH). If this bit is one, the result is
negative, and if this bit is zero, the result is positive. Figure 118 shows the decoding of
the differential input range.
Table 95 shows the resulting output codes if the differential input channel pair (ADCn -
ADCm) is selected with a gain of GAIN and a reference voltage of VREF.
236 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 118. Differential Measurement Range
Output Code
0x1FF
0x000
- V REF/GAIN 0 VREF/GAIN Differential Input

0x3FF
Voltage (Volts)
0x200
Table 95. Correlation Between Input Voltage and Output Codes

VADCn Read code Corresponding decimal value
VADCm + V REF /GAIN 0x1FF 511
VADCm + 0.999 VREF /GAIN 0x1FF 511
VADCm + 0.998 VREF /GAIN 0x1FE 510
... ... ...
VADCm + 0.001 VREF /GAIN 0x001 1
VADCm 0x000 0
VADCm - 0.001 VREF /GAIN 0x3FF -1
... ... ...
VADCm - 0.999 VREF /GAIN 0x201 -511
VADCm - VREF /GAIN 0x200 -512
Example:
ADMUX = 0xED (ADC3 - ADC2, 10x gain, 2.56V reference, left adjusted result)
Voltage on ADC3 is 300 mV, voltage on ADC2 is 500 mV.
ADCR = 512 * 10 * (300 - 500) / 2560 = -400 = 0x270
237
2467E–AVR–05/02
ADCL will thus read 0x00, and ADCH will read 0x9C. Writing zero to ADLAR right
adjusts the result: ADCL = 0x70, ADCH = 0x02.
ADC Multiplexer Selection

Register – ADMUX Bit 7 6 5 4 3 2 1 0
REFS1 REFS0 ADLAR MUX4 MUX3 MUX2 MUX1 MUX0 ADMUX
• Bit 7:6 – REFS1:0: Reference Selection Bits

These bits select the voltage reference for the ADC, as shown in Table 96. If these bits
are changed during a conversion, the change will not go in effect until this conversion is
complete (ADIF in ADCSRA is set). The internal voltage reference options may not be
used if an external reference voltage is being applied to the AREF pin.
Table 96. Voltage Reference Selections for ADC
REFS1 REFS0 Voltage Reference Selection
0 0 AREF, Internal Vref turned off
0 1 AVCC with external capacitor at AREF pin
1 0 Reserved
1 1 Internal 2.56V Voltage Reference with external capacitor at AREF pin
• Bit 5 – ADLAR: ADC Left Adjust Result

The ADLAR bit affects the presentation of the ADC conversion result in the ADC Data
Register. Write one to ADLAR to left adjust the result. Otherwise, the result is right
adjusted. Changing the ADLAR bit will affect the ADC Data Register immediately,
regardless of any ongoing conversions. For a complete description of this bit, see “The
ADC Data Register – ADCL and ADCH” on page 241.
• Bits 4:0 – MUX4:0: Analog Channel and Gain Selection Bits

The value of these bits selects which combination of analog inputs are connected to the
ADC. These bits also select the gain for the differential channels. See Table 97 for
details. If these bits are changed during a conversion, the change will not go in effect
until this conversion is complete (ADIF in ADCSRA is set).
Table 97. Input Channel and Gain Selections
Single Ended Positive Differential Negative Differential
MUX4..0 Input Input Input Gain
00000 ADC0
00001 ADC1
00010 ADC2
00011 ADC3 N/A
00100 ADC4
00101 ADC5
00110 ADC6
00111 ADC7
01000 ADC0 ADC0 10x
238 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 97. Input Channel and Gain Selections (Continued)

Single Ended Positive Differential Negative Differential
MUX4..0 Input Input Input Gain
01001 ADC1 ADC0 10x
01010 ADC0 ADC0 200x
01011 ADC1 ADC0 200x
01100 ADC2 ADC2 10x
01101 ADC3 ADC2 10x
01110 ADC2 ADC2 200x
01111 ADC3 ADC2 200x
10000 ADC0 ADC1 1x
10001 ADC1 ADC1 1x
10010 N/A ADC2 ADC1 1x
10011 ADC3 ADC1 1x
10100 ADC4 ADC1 1x
10101 ADC5 ADC1 1x
10110 ADC6 ADC1 1x
10111 ADC7 ADC1 1x
11000 ADC0 ADC2 1x
11001 ADC1 ADC2 1x
11010 ADC2 ADC2 1x
11011 ADC3 ADC2 1x
11100 ADC4 ADC2 1x
11101 ADC5 ADC2 1x
11110 1.22 V (VBG) N/A
11111 0 V (GND)
ADC Control and Status

Register A – ADCSRA Bit 7 6 5 4 3 2 1 0
ADEN ADSC ADFR ADIF ADIE ADPS2 ADPS1 ADPS0 ADCSRA
• Bit 7 – ADEN: ADC Enable

Writing this bit to one enables the ADC. By writing it to zero, the ADC is turned off. Turn-
ing the ADC off while a conversion is in progress, will terminate this conversion.
• Bit 6 – ADSC: ADC Start Conversion

In Single Conversion mode, write this bit to one to start each conversion. In Free Run-
ning mode, write this bit to one to start the first conversion. The first conversion after
ADSC has been written after the ADC has been enabled, or if ADSC is written at the
239
2467E–AVR–05/02
same time as the ADC is enabled, will take 25 ADC clock cycles instead of the normal
13. This first conversion performs initialization of the ADC.
ADSC will read as one as long as a conversion is in progress. When the conversion is
complete, it returns to zero. Writing zero to this bit has no effect.
• Bit 5 – ADFR: ADC Free Running Select

When this bit is written to one, the ADC operates in Free Running mode. In this mode,
the ADC samples and updates the data registers continuously. Writing zero to this bit
will terminate Free Running mode.
• Bit 4 – ADIF: ADC Interrupt Flag

This bit is set when an ADC conversion completes and the data registers are updated.
The ADC Conversion Complete Interrupt is executed if the ADIE bit and the I-bit in
SREG are set. ADIF is cleared by hardware when executing the corresponding interrupt
handling vector. Alternatively, ADIF is cleared by writing a logical one to the flag.
Beware that if doing a read-modify-write on ADCSRA, a pending interrupt can be dis-
abled. This also applies if the SBI and CBI instructions are used.
• Bit 3 – ADIE: ADC Interrupt Enable

When this bit is written to one and the I-bit in SREG is set, the ADC Conversion Com-
plete Interrupt is activated.
• Bits 2:0 – ADPS2:0: ADC Prescaler Select Bits

These bits determine the division factor between the XTAL frequency and the input
clock to the ADC.
Table 98. ADC Prescaler Selections
ADPS2 ADPS1 ADPS0 Division Factor
0 0 0 2
0 0 1 2
0 1 0 4
0 1 1 8
1 0 0 16
1 0 1 32
1 1 0 64
1 1 1 128
240 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
The ADC Data Register –

ADCL and ADCH
ADLAR = 0:
Bit 15 14 13 12 11 10 9 8
– – – – – – ADC9 ADC8 ADCH
ADC7 ADC6 ADC5 ADC4 ADC3 ADC2 ADC1 ADC0 ADCL
7 6 5 4 3 2 1 0
R R R R R R R R
0 0 0 0 0 0 0 0
ADLAR = 1:
Bit 15 14 13 12 11 10 9 8
ADC9 ADC8 ADC7 ADC6 ADC5 ADC4 ADC3 ADC2 ADCH
ADC1 ADC0 – – – – – – ADCL
7 6 5 4 3 2 1 0
R R R R R R R R
0 0 0 0 0 0 0 0
When an ADC conversion is complete, the result is found in these two registers. If differ-
ential channels are used, the result is presented in two’s complement form.
When ADCL is read, the ADC Data Register is not updated until ADCH is read. Conse-
quently, if the result is left adjusted and no more than 8-bit precision is required, it is
sufficient to read ADCH. Otherwise, ADCL must be read first, then ADCH.
The ADLAR bit in ADMUX, and the MUXn bits in ADMUX affect the way the result is
read from the registers. If ADLAR is set, the result is left adjusted. If ADLAR is cleared
(default), the result is right adjusted.
• ADC9:0: ADC Conversion Result

These bits represent the result from the conversion, as detailed in “ADC Conversion
Result” on page 236.

SFIOR Bit 7 6 5 4 3 2 1 0
• Bit 4 – ADHSM: ADC High Speed Mode

Writing this bit to one enables the ADC High Speed mode. This mode enables higher
conversion rate at the expense of higher power consumption.
241
2467E–AVR–05/02
JTAG Interface and
On-chip Debug
System
Features • JTAG (IEEE std. 1149.1 Compliant) Interface

• Boundary-scan Capabilities According to the IEEE std. 1149.1 (JTAG) Standard
• Debugger Access to:
– All Internal Peripheral Units
– Internal and External RAM
– The Internal Register File
– Program Counter
– EEPROM and Flash Memories
• Extensive On-chip Debug Support for Break Conditions, Including
– AVR Break Instruction
– Break on Change of Program Memory Flow
– Single Step Break
– Program Memory Breakpoints on Single Address or Address Range
– Data Memory Breakpoints on Single Address or Address Range
• Programming of Flash, EEPROM, Fuses, and Lock Bits through the JTAG Interface
• On-chip Debugging Supported by AVR Studio
Overview The AVR IEEE std. 1149.1 compliant JTAG interface can be used for
• Testing PCBs by using the JTAG Boundary-scan capability
• Programming the non-volatile memories, Fuses and Lock bits
• On-chip debugging
A brief description is given in the following sections. Detailed descriptions for Program-
ming via the JTAG interface, and using the Boundary-scan Chain can be found in the
sections “Programming Via the JTAG Interface” on page 302 and “IEEE 1149.1 (JTAG)
Boundary-scan” on page 248, respectively. The On-chip Debug support is considered
being private JTAG instructions, and distributed within ATMEL and to selected third
party vendors only.
Figure 119 shows a block diagram of the JTAG interface and the On-chip Debug sys-
tem. The TAP Controller is a state machine controlled by the TCK and TMS signals. The
TAP Controller selects either the JTAG Instruction Register or one of several Data Reg-
isters as the scan chain (Shift Register) between the TDI – input and TDO – output. The
Instruction Register holds JTAG instructions controlling the behavior of a Data Register.
The ID-Register, Bypass Register, and the Boundary-scan Chain are the data registers
used for board-level testing. The JTAG Programming Interface (actually consisting of
several physical and virtual Data Registers) is used for serial programming via the JTAG
interface. The Internal Scan Chain and Break Point Scan Chain are used for On-chip
debugging only.
Test Access Port – TAP The JTAG interface is accessed through four of the AVR’s pins. In JTAG terminology,
these pins constitute the Test Access Port –– TAP. These pins are:
• TMS: Test mode select. This pin is used for navigating through the TAP-controller
state machine.
• TCK: Test clock. JTAG operation is synchronous to TCK.
• TDI: Test Data In. Serial input data to be shifted in to the Instruction Register or Data
Register (Scan Chains).
242 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
• TDO: Test Data Out. Serial output data from Instruction Register or Data Register.
The IEEE std. 1149.1 also specifies an optional TAP signal; TRST – Test ReSeT –
which is not provided.
When the JTAGEN fuse is unprogrammed, these four TAP pins are normal port pins,
and the TAP controller is in reset. When programmed, the input TAP signals are inter-
nally pulled high and the JTAG is enabled for Boundary-scan and programming. The
device is shipped with this fuse programmed.
For the On-chip Debug system, in addition to the JTAG interface pins, the RESET pin is
monitored by the debugger to be able to detect External Reset sources. The debugger
can also pull the RESET pin low to reset the whole system, assuming only open collec-
tors on the Reset line are used in the application.
Figure 119. Block Diagram

I/O PORT 0
DEVICE BOUNDARY
BOUNDARY SCAN CHAIN
TDI
JTAG PROGRAMMING
TDO TAP INTERFACE
TCK CONTROLLER
TMS
AVR CPU
INTERNAL
FLASH Address SCAN PC
INSTRUCTION MEMORY Data CHAIN Instruction
REGISTER
ID
REGISTER BREAKPOINT
UNIT
M FLOW CONTROL
BYPASS
PERIPHERIAL
Analog inputs
U UNIT
REGISTER
ANALOG
X DIGITAL
UNITS
PERIPHERAL
UNITS
BREAKPOINT
SCAN CHAIN
JTAG / AVR CORE
COMMUNICATION
ADDRESS INTERFACE
OCD STATUS
Control & Clock lines

DECODER
AND CONTROL
I/O PORT n
243
2467E–AVR–05/02
Figure 120. TAP Controller State Diagram
1 Test-Logic-Reset
0 1 1 1
Run-Test/Idle Select-DR Scan Select-IR Scan
0 0
1 1
Capture-DR Capture-IR
0 0
Shift-DR 0 Shift-IR 0
1 1
1 1
Exit1-DR Exit1-IR
0 0
Pause-DR 0 Pause-IR 0
1 1
0 0
Exit2-DR Exit2-IR
1 1
Update-DR Update-IR
1 0 1 0
TAP Controller The TAP controller is a 16-state finite state machine that controls the operation of the
Boundary-scan circuitry, JTAG programming circuitry, or On-chip Debug system. The
state transitions depicted in Figure 120 depend on the signal present on TMS (shown
adjacent to each state transition) at the time of the rising edge at TCK. The initial state
after a Power-on Reset is Test-Logic-Reset.
As a definition in this document, the LSB is shifted in and out first for all Shift Registers.
Assuming Run-Test/Idle is the present state, a typical scenario for using the JTAG inter-
face is:
• At the TMS input, apply the sequence 1, 1, 0, 0 at the rising edges of TCK to enter
the Shift Instruction Register – Shift-IR state. While in this state, shift the 4 bits of the
JTAG instructions into the JTAG instruction register from the TDI input at the rising
edge of TCK. The TMS input must be held low during input of the 3 LSBs in order to
remain in the Shift-IR state. The MSB of the instruction is shifted in when this state
244 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
is left by setting TMS high. While the instruction is shifted in from the TDI pin, the
captured IR-state 0x01 is shifted out on the TDO pin. The JTAG Instruction selects a
particular Data Register as path between TDI and TDO and controls the circuitry
surrounding the selected Data Register.
• Apply the TMS sequence 1, 1, 0 to re-enter the Run-Test/Idle state. The instruction
is latched onto the parallel output from the Shift Register path in the Update-IR
state. The Exit-IR, Pause-IR, and Exit2-IR states are only used for navigating the
state machine.
• At the TMS input, apply the sequence 1, 0, 0 at the rising edges of TCK to enter the
Shift Data Register – Shift-DR state. While in this state, upload the selected Data
Register (selected by the present JTAG instruction in the JTAG Instruction Register)
from the TDI input at the rising edge of TCK. In order to remain in the Shift-DR state,
the TMS input must be held low during input of all bits except the MSB. The MSB of
the data is shifted in when this state is left by setting TMS high. While the Data
Register is shifted in from the TDI pin, the parallel inputs to the Data Register
captured in the Capture-DR state is shifted out on the TDO pin.
• Apply the TMS sequence 1, 1, 0 to re-enter the Run-Test/Idle state. If the selected
Data Register has a latched parallel-output, the latching takes place in the Update-
DR state. The Exit-DR, Pause-DR, and Exit2-DR states are only used for navigating
the state machine.
As shown in the state diagram, the Run-Test/Idle state need not be entered between
selecting JTAG instruction and using Data Registers, and some JTAG instructions may
select certain functions to be performed in the Run-Test/Idle, making it unsuitable as an
Idle state.
Note: Independent of the initial state of the TAP Controller, the Test-Logic-Reset state can
always be entered by holding TMS high for 5 TCK clock periods.
For detailed information on the JTAG specification, refer to the literature listed in “Bibli-
ography” on page 247.
Using the Boundary- A complete description of the Boundary-scan capabilities are given in the section “IEEE
scan Chain 1149.1 (JTAG) Boundary-scan” on page 248.
Using the On-chip Debug As shown in Figure 119, the hardware support for On-chip Debugging consists mainly of
System • A scan chain on the interface between the internal AVR CPU and the internal
peripheral units
• Break point unit
• Communication interface between the CPU and JTAG system
All read or modify/write operations needed for implementing the Debugger are done by
applying AVR instructions via the internal AVR CPU Scan Chain. The CPU sends the
result to an I/O memory mapped location which is part of the communication interface
between the CPU and the JTAG system.
The Break point Unit implements Break on Change of Program Flow, Single Step Break,
two Program Memory Breakpoints, and two combined break points. Together, the four
break points can be configured as either:
• 4 single Program Memory break points
• 3 Single Program Memory break point + 1 single Data Memory break point
• 2 single Program Memory break points + 2 single Data Memory break points
• 2 single Program Memory break points + 1 Program Memory break point with mask
(“range break point”)
245
2467E–AVR–05/02
• 2 single Program Memory break points + 1 Data Memory break point with mask
“range break point”.
A debugger, like the AVR Studio®, may however use one or more of these resources for
its internal purpose, leaving less flexibility to the end-user.
A list of the On-chip Debug specific JTAG instructions is given in “On-chip Debug Spe-
cific JTAG Instructions” on page 246.
The JTAGEN fuse must be programmed to enable the JTAG Test Access Port. In addi-
tion, the OCDEN fuse must be programmed and no Lock bits must be set for the On-
chip Debug system to work. As a security feature, the On-chip Debug system is disabled
when any Lock bits are set. Otherwise, the On-chip Debug system would have provided
a back-door into a secured device.
The AVR Studio enables the user to fully control execution of programs on an AVR
device with On-chip Debug capability, AVR In-Circuit Emulator, or the built-in AVR
Instruction Set Simulator. AVR Studio supports source level execution of Assembly pro-
grams assembled with Atmel Corporation’s AVR Assembler and C programs compiled
with third party vendors’ compilers.
AVR Studio runs under Microsoft® Windows® 95/98/2000 and Windows NT®.
For a full description of the AVR Studio, please refer to the AVR Studio User Guide.
Only highlights are presented in this document.
All necessary execution commands are available in AVR Studio, both on source level
and on disassembly level. The user can execute the program, single step through the
code either by tracing into or stepping over functions, step out of functions, place the
cursor on a statement and execute until the statement is reached, stop the execution,
and reset the execution target. In addition, the user can have an unlimited number of
code break points (using the BREAK instruction) and up to two data memory break
points, alternatively combined as a mask (range) break point.
On-chip Debug Specific The On-chip debug support is considered being private JTAG instructions, and distrib-
JTAG Instructions uted within ATMEL and to selected third-party vendors only. Instruction opcodes are
listed for reference.
PRIVATE0; $8 Private JTAG instruction for accessing On-chip Debug system.
PRIVATE1; $9 Private JTAG instruction for accessing On-chip Debug system.
PRIVATE2; $A Private JTAG instruction for accessing On-chip Debug system.
PRIVATE3; $B Private JTAG instruction for accessing On-chip Debug system.
246 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
On-chip Debug Related

Register in I/O Memory
On-chip Debug Register –

OCDR Bit 7 6 5 4 3 2 1 0
MSB/IDRD LSB OCDR
The OCDR Register provides a communication channel from the running program in the
microcontroller to the debugger. The CPU can transfer a byte to the debugger by writing
to this location. At the same time, an internal flag; I/O Debug Register Dirty – IDRD – is
set to indicate to the debugger that the register has been written. When the CPU reads
the OCDR Register the 7 LSB will be from the OCDR Register, while the MSB is the
IDRD bit. The debugger clears the IDRD bit when it has read the information.
In some AVR devices, this register is shared with a standard I/O location. In this case,
the OCDR Register can only be accessed if the OCDEN fuse is programmed, and the
debugger enables access to the OCDR Register. In all other cases, the standard I/O
location is accessed.
Refer to the debugger documentation for further information on how to use this register.
Using the JTAG Programming of AVR parts via JTAG is performed via the four-pin JTAG port, TCK,
Programming TMS, TDI, and TDO. These are the only pins that need to be controlled/observed to per-
Capabilities form JTAG programming (in addition to power pins). It is not required to apply 12V
externally. The JTAGEN fuse must be programmed and the JTD bit in the MCUSR Reg-
ister must be cleared to enable the JTAG Test Access Port.
The JTAG programming capability supports:
• Flash programming and verifying
• EEPROM programming and verifying
• Fuse programming and verifying
• Lock bit programming and verifying
The Lock bit security is exactly as in Parallel Programming mode. If the Lock bits LB1 or
LB2 are programmed, the OCDEN Fuse cannot be programmed unless first doing a
chip erase. This is a security feature that ensures no back-door exists for reading out the
content of a secured device.
The details on programming through the JTAG interface and programming specific
JTAG instructions are given in the section “Programming Via the JTAG Interface” on
page 302.
Bibliography For more information about general Boundary-scan, the following literature can be
consulted:
• IEEE: IEEE Std 1149.1-1990. IEEE Standard Test Access Port and Boundary-scan
Architecture, IEEE, 1993
• Colin Maunder: The Board Designers Guide to Testable Logic Circuits, Addison-
Wesley, 1992
247
2467E–AVR–05/02
IEEE 1149.1 (JTAG)
Boundary-scan
Features • JTAG (IEEE std. 1149.1 Compliant) Interface

• Boundary-scan Capabilities According to the JTAG Standard
• Full Scan of all Port Functions as well as Analog Circuitry having Off-chip Connections
• Supports the Optional IDCODE Instruction
• Additional Public AVR_RESET Instruction to Reset the AVR
System Overview The Boundary-scan Chain has the capability of driving and observing the logic levels on
the digital I/O pins, as well as the boundary between digital and analog logic for analog
circuitry having off-chip connections. At system level, all ICs having JTAG capabilities
are connected serially by the TDI/TDO signals to form a long Shift Register. An external
controller sets up the devices to drive values at their output pins, and observe the input
values received from other devices. The controller compares the received data with the
expected result. In this way, Boundary-scan provides a mechanism for testing intercon-
nections and integrity of components on Printed Circuits Boards by using the four TAP
signals only.
The four IEEE 1149.1 defined mandatory JTAG instructions IDCODE, BYPASS, SAM-
PLE/PRELOAD, and EXTEST, as well as the AVR specific public JTAG instruction
AVR_RESET can be used for testing the Printed Circuit Board. Initial scanning of the
data register path will show the ID-code of the device, since IDCODE is the default
JTAG instruction. It may be desirable to have the AVR device in reset during Test mode.
If not reset, inputs to the device may be determined by the scan operations, and the
internal software may be in an undetermined state when exiting the Test mode. Entering
Reset, the outputs of any Port Pin will instantly enter the high impedance state, making
the HIGHZ instruction redundant. If needed, the BYPASS instruction can be issued to
make the shortest possible scan chain through the device. The device can be set in the
Reset state either by pulling the external RESET pin low, or issuing the AVR_RESET
instruction with appropriate setting of the Reset Data Register.
The EXTEST instruction is used for sampling external pins and loading output pins with
data. The data from the output latch will be driven out on the pins as soon as the
EXTEST instruction is loaded into the JTAG IR-register. Therefore, the SAMPLE/PRE-
LOAD should also be used for setting initial values to the scan ring, to avoid damaging
the board when issuing the EXTEST instruction for the first time. SAMPLE/PRELOAD
can also be used for taking a snapshot of the external pins during normal operation of
the part.
The JTAGEN fuse must be programmed and the JTD bit in the I/O register MCUCSR
must be cleared to enable the JTAG Test Access Port.
When using the JTAG interface for Boundary-scan, using a JTAG TCK clock frequency
higher than the internal chip frequency is possible. The chip clock is not required to run.
Data Registers The data registers relevant for Boundary-scan operations are:
• Bypass Register
• Device Identification Register
• Reset Register
• Boundary-scan Chain
248 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Bypass Register The Bypass Register consists of a single Shift Register stage. When the Bypass Regis-
ter is selected as path between TDI and TDO, the register is reset to 0 when leaving the
Capture-DR controller state. The Bypass Register can be used to shorten the scan
chain on a system when the other devices are to be tested.
Device Identification Register Figure 121 shows the structure of the Device Identification Register.
Figure 121. The Format of the Device Identification Register
MSB LSB
Bit 31 28 27 12 11 1 0
Device ID Version Part Number Manufacturer ID 1
4 bits 16 bits 11 bits 1 bit
• Version
Version is a 4-bit number identifying the revision of the component. The relevant version
number is shown in Table 99.
Table 99. JTAG Version Numbers

Version JTAG Version number (Hex)
ATmega128 revision C 0x3
ATmega128 revision F 0x5
ATmega128 revision G 0x6
• Part Number
The part number is a 16-bit code identifying the component. The JTAG Part Number for
ATmega128 is listed in Table 100.
Table 100. AVR JTAG Part Number

Part number JTAG Part Number (Hex)
ATmega128 0x9702
• Manufacturer ID
The Manufacturer ID is a 11-bit code identifying the manufacturer. The JTAG manufac-
turer ID for ATMEL is listed in Table 101.
Table 101. Manufacturer ID

Manufacturer JTAG Man. ID (Hex)
ATMEL 0x01F
Reset Register The Reset Register is a Test Data Register used to reset the part. Since the AVR tri-
states Port Pins when reset, the Reset Register can also replace the function of the
unimplemented optional JTAG instruction HIGHZ.
A high value in the Reset Register corresponds to pulling the External Reset low. The
part is reset as long as there is a high value present in the Reset Register. Depending
on the Fuse settings for the clock options, the part will remain reset for a Reset Time-
Out Period (refer to “Clock Sources” on page 34) after releasing the Reset Register. The
output from this Data Register is not latched, so the Reset will take place immediately,
as shown in Figure 122.
249
2467E–AVR–05/02
Figure 122. Reset Register
To
TDO
From Other Internal and

External Reset Sources
From Internal Reset

D Q
TDI
ClockDR · AVR_RESET
Boundary-scan Chain The Boundary-scan Chain has the capability of driving and observing the logic levels on
circuitry having off-chip connections.
See “Boundary-scan Chain” on page 252 for a complete description.
Boundary-scan Specific The Instruction Register is 4-bit wide, supporting up to 16 instructions. Listed below are
JTAG Instructions the JTAG instructions useful for Boundary-scan operation. Note that the optional HIGHZ
instruction is not implemented, but all outputs with tri-state capability can be set in high-
impedant state by using the AVR_RESET instruction, since the initial state for all port
pins is tri-state.
As a definition in this data sheet, the LSB is shifted in and out first for all Shift Registers.
The OPCODE for each instruction is shown behind the instruction name in hex format.
The text describes which data register is selected as path between TDI and TDO for
each instruction.
EXTEST; $0 Mandatory JTAG instruction for selecting the Boundary-scan Chain as Data Register for
testing circuitry external to the AVR package. For port-pins, Pull-up Disable, Output
Control, Output Data, and Input Data are all accessible in the scan chain. For Analog cir-
cuits having off-chip connections, the interface between the analog and the digital logic
is in the scan chain. The contents of the latched outputs of the Boundary-scan chain is
driven out as soon as the JTAG IR-register is loaded with the EXTEST instruction.
The active states are:
• Capture-DR: Data on the external pins are sampled into the Boundary-scan Chain.
• Shift-DR: The Internal Scan Chain is shifted by the TCK input.
• Update-DR: Data from the scan chain is applied to output pins.
IDCODE; $1 Optional JTAG instruction selecting the 32-bit ID Register as Data Register. The ID Reg-
ister consists of a version number, a device number and the manufacturer code chosen
by JEDEC. This is the default instruction after power-up.
• Capture-DR: Data in the IDCODE Register is sampled into the Boundary-scan
Chain.
250 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
• Shift-DR: The IDCODE scan chain is shifted by the TCK input.
SAMPLE_PRELOAD; $2 Mandatory JTAG instruction for pre-loading the output latches and taking a snap-shot of
the input/output pins without affecting the system operation. However, the output
latched are not connected to the pins. The Boundary-scan Chain is selected as Data
Register.
• Capture-DR: Data on the external pins are sampled into the Boundary-scan Chain.
• Shift-DR: The Boundary-scan Chain is shifted by the TCK input.
• Update-DR: Data from the Boundary-scan Chain is applied to the output latches.
However, the output latches are not connected to the pins.
AVR_RESET; $C The AVR specific public JTAG instruction for forcing the AVR device into the Reset
mode or releasing the JTAG Reset source. The TAP controller is not reset by this
instruction. The one bit Reset Register is selected as Data Register. Note that the Reset
will be active as long as there is a logic 'one' in the Reset Chain. The output from this
chain is not latched.
• Shift-DR: The Reset Register is shifted by the TCK input.
BYPASS; $F Mandatory JTAG instruction selecting the Bypass Register for Data Register.
• Capture-DR: Loads a logic “0” into the Bypass Register.
• Shift-DR: The Bypass Register cell between TDI and TDO is shifted.
Boundary-scan Related
Register in I/O Memory
MCU Control and Status The MCU Control and Status Register contains control bits for general MCU functions,
Register – MCUCSR and provides information on which reset source caused an MCU Reset.
Bit 7 6 5 4 3 2 1 0
JTD – – JTRF WDRF BORF EXTRF PORF MCUCSR
Initial Value 0 0 0 See Bit Description
• Bits 7 – JTD: JTAG Interface Disable

When this bit is zero, the JTAG interface is enabled if the JTAGEN fuse is programmed.
If this bit is one, the JTAG interface is disabled. In order to avoid unintentional disabling
or enabling of the JTAG interface, a timed sequence must be followed when changing
this bit: The application software must write this bit to the desired value twice within four
cycles to change its value.
• Bit 4 – JTRF: JTAG Reset Flag

This bit is set if a Reset is being caused by a logic one in the JTAG Reset Register
selected by the JTAG instruction AVR_RESET. This bit is reset by a Power-on Reset, or
by writing a logic zero to the flag.
251
2467E–AVR–05/02
Boundary-scan Chain The Boundary-scan chain has the capability of driving and observing the logic levels on
circuitry having off-chip connection.
Scanning the Digital Port Pins Figure 123 shows the Boundary-scan Cell for a bi-directional port pin with pull-up func-
tion. The cell consists of a standard Boundary-scan cell for the Pull-up Enable – PUExn
– function, and a bi-directional pin cell that combines the three signals Output Control –
OCxn, Output Data – ODxn, and Input Data – IDxn, into only a two-stage Shift Register.
The port and pin indexes are not used in the following description
The Boundary-scan logic is not included in the figures in the Data Sheet. Figure 124
shows a simple digital Port Pin as described in the section “I/O Ports” on page 60. The
Boundary-scan details from Figure 123 replaces the dashed box in Figure 124.
When no alternate port function is present, the Input Data – ID corresponds to the PINxn
Register value (but ID has no synchronizer), Output Data corresponds to the PORT
Register, Output Control corresponds to the Data Direction – DD Register, and the Pull-
up Enable – PUExn – corresponds to logic expression PUD · DDxn · PORTxn.
Digital alternate port functions are connected outside the dotted box in Figure 124 to
make the scan chain read the actual pin value. For Analog function, there is a direct
connection from the external pin to the analog circuit, and a scan chain is inserted on
the interface between the digital logic and the analog circuitry.
Figure 123. Boundary-scan Cell for Bi-directional Port Pin with Pull-Up Function.
ShiftDR To Next Cell EXTEST Vcc
Pullup Enable (PUE)

0
FF2 LD2 1
0
D Q D Q
1
G
Output Control (OC)
FF1 LD1 0
0
D Q D Q 1
1
G
Output Data (OD)
0 FF0 LD0 0
0 Port Pin (PXn)
1 D Q D Q 1
1
G
Input Data (ID)
From Last Cell ClockDR UpdateDR
252 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 124. General Port Pin Schematic diagram
See Boundary-Scan description

for details!
PUExn PUD
Q D
DDxn
Q CLR
WDx
RESET
OCxn
RDx
DATA BUS
Pxn Q D
ODxn PORTxn
Q CLR
IDxn WPx
RESET
SLEEP RRx
SYNCHRONIZER
RPx
D Q D Q
PINxn
L Q Q
CLK I/O
PUD: PULLUP DISABLE WDx: WRITE DDRx

PUExn: PULLUP ENABLE for pin Pxn RDx: READ DDRx
OCxn: OUTPUT CONTROL for pin Pxn WPx: WRITE PORTx
ODxn: OUTPUT DATA to pin Pxn RRx: READ PORTx REGISTER
IDxn: INPUT DATA from pin Pxn RPx: READ PORTx PIN
SLEEP: SLEEP CONTROL CLK I/O : I/O CLOCK
Boundary-scan and the Two- The two Two-wire Interface pins SCL and SDA have one additional control signal in the
wire Interface scan-chain; Two-wire Interface Enable – TWIEN. As shown in Figure 125, the TWIEN
signal enables a tri-state buffer with slew-rate control in parallel with the ordinary digital
port pins. A general scan cell as shown in Figure 129 is attached to the TWIEN signal.
Notes: 1. A separate scan chain for the 50 ns spike filter on the input is not provided. The ordi-
nary scan support for digital port pins suffice for connectivity tests. The only reason
for having TWIEN in the scan path, is to be able to disconnect the slew-rate control
buffer when doing boundary-scan.
2. Make sure the OC and TWIEN signals are not asserted simultaneously, as this will
lead to drive contention.
253
2467E–AVR–05/02
Figure 125. Additional Scan Signal for the Two-wire Interface
PUExn
OCxn
ODxn
Pxn TWIEN
SRC
Slew-rate limited
IDxn
Scanning the RESET Pin The RESET pin accepts 5V active low logic for standard Reset operation, and 12V
active high logic for High Voltage Parallel programming. An observe-only cell as shown
in Figure 126 is inserted both for the 5V Reset signal; RSTT, and the 12V Reset signal;
RSTHV.
Figure 126. Observe-only Cell

To
next
ShiftDR cell
From system pin To system logic
FF1
0
D Q
1
From ClockDR
previous
cell
Scanning the Clock Pins The AVR devices have many clock options selectable by fuses. These are: Internal RC
Oscillator, External RC, External Clock, (High Frequency) Crystal Oscillator, Low-fre-
quency Crystal Oscillator, and Ceramic Resonator.
Figure 127 shows how each Oscillator with external connection is supported in the scan
chain. The Enable signal is supported with a general boundary-scan cell, while the
Oscillator/Clock output is attached to an observe-only cell. In addition to the main clock,
the Timer Oscillator is scanned in the same way. The output from the internal RC Oscil-
lator is not scanned, as this Oscillator does not have external connections.
254 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 127. Boundary-scan Cells for Oscillators and Clock Options

XTAL1/TOSC1 XTAL2/TOSC2
To
Next To
ShiftDR Cell EXTEST Oscillator next
ShiftDR cell
From Digital Logic

0
ENABLE OUTPUT To System Logic
1
0 FF1
D Q D Q 0
1 D Q
G 1
From ClockDR UpdateDR

Previous From ClockDR
Cell Previous
Cell
Table 102 summaries the scan registers for the external clock pin XTAL1, oscillators
with XTAL1/XTAL2 connections as well as 32 kHz Timer Oscillator.
Table 102. Scan Signals for the Oscillator(1)(2)(3)

Scanned Clock Scanned Clock Line
Enable signal Line Clock Option when not Used
EXTCLKEN EXTCLK (XTAL1) External Clock 0
OSCON OSCCK External Crystal 0
External Ceramic Resonator
RCOSCEN RCCK External RC 1
OSC32EN OSC32CK Low Freq. External Crystal 0
TOSKON TOSCK 32 kHz Timer Oscillator 0
Notes: 1. Do not enable more than one clock source as main clock at a time.
2. Scanning an Oscillator output gives unpredictable results as there is a frequency drift
between the Internal Oscillator and the JTAG TCK clock. If possible, scanning an
external clock is preferred.
3. The clock configuration is programmed by fuses. As a fuse is not changed run-time,
the clock configuration is considered fixed for a given application. The user is advised
to scan the same clock option as to be used in the final system. The enable signals
are supported in the scan chain because the system logic can disable clock options
in sleep modes, thereby disconnecting the Oscillator pins from the scan path if not
provided. The INTCAP fuses are not supported in the scan-chain, so the boundary
scan chain can not make a XTAL Oscillator requiring internal capacitors to run unless
the fuse is correctly programmed.
Scanning the Analog The relevant Comparator signals regarding Boundary-scan are shown in Figure 128.
Comparator The Boundary-scan cell from Figure 129 is attached to each of these signals. The sig-
nals are described in Table 103.
The Comparator need not be used for pure connectivity testing, since all analog inputs
are shared with a digital port pin as well.
255
2467E–AVR–05/02
Figure 128. Analog comparator
BANDGAP
REFERENCE
ACBG
ACO
AC_IDLE
ACME
ADCEN
ADC MULTIPLEXER
OUTPUT
Figure 129. General Boundary-scan Cell used for Signals for Comparator and ADC
To
Next
ShiftDR Cell EXTEST
From Digital Logic/

0
From Analog Ciruitry To Analog Circuitry/
1 To Digital Logic
0
D Q D Q
1
G
From ClockDR UpdateDR

Previous
Cell
256 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 103. Boundary-scan Signals for the Analog Comparator

Direction as Recommended Output values when
Signal Seen from the Input when not Recommended
Name Comparator Description in Use Inputs are Used
AC_IDLE Input Turns off Analog 1 Depends upon µC
comparator when code being executed
true
ACO Output Analog Will become 0
Comparator input to µC code
Output being executed
ACME Input Uses output 0 Depends upon µC
signal from ADC code being executed
mux when true
ACBG Input Bandgap 0 Depends upon µC
Reference enable code being executed
Scanning the ADC Figure 130 shows a block diagram of the ADC with all relevant control and observe sig-
nals. The Boundary-scan cell from Figure 126 is attached to each of these signals. The
ADC need not be used for pure connectivity testing, since all analog inputs are shared
with a digital port pin as well.
Figure 130. Analog to Digital Converter
VCCREN
AREF
IREFEN
2.56V
To Comparator ref
PASSEN
MUXEN_7
ADC_7
MUXEN_6
ADC_6
MUXEN_5
ADC_5
MUXEN_4 ADCBGEN
SCTEST
ADC_4
ADHSM
ADHSM
EXTCH 1.22V PRECH PRECH

MUXEN_3 ref AREF
AREF
ADC_3
DACOUT
MUXEN_2
ADC_2 DAC_9..0
10-bit DAC +
MUXEN_1 COMP COMP
ADC_1 G10 G20 -
ADCEN
MUXEN_0
ACTEN
ADC_0 + +
NEGSEL_2 10x 20x

ADC_2 - - HOLD
NEGSEL_1
ADC_1 GNDEN
ST
NEGSEL_0
ADC_0
ACLK
AMPEN
The signals are described briefly in Table 104.
257
2467E–AVR–05/02
Table 104. Boundary-scan Signals for the ADC
Direction Recommen- Output Values when
as Seen ded Input Recommended Inputs
Signal from the when not are Used, and CPU is
Name ADC Description in Use not Using the ADC
COMP Output Comparator Output 0 0
ACLK Input Clock signal to gain 0 0
stages implemented
as Switch-cap filters
ACTEN Input Enable path from gain 0 0
stages to the
comparator
ADHSM Input Increases speed of 0 0
comparator at the
sacrifice of higher
power consumption
ADCBGEN Input Enable Band-gap 0 0
reference as negative
input to comparator
ADCEN Input Power-on signal to the 0 0
ADC
AMPEN Input Power-on signal to the 0 0
gain stages
DAC_9 Input Bit 9 of digital value to 1 1
DAC
DAC
DAC
DAC
DAC
DAC
DAC
DAC
DAC
DAC
EXTCH Input Connect ADC 1 1
channels 0 - 3 to by-
pass path around gain
stages
258 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 104. Boundary-scan Signals for the ADC (Continued)

G10 Input Enable 10x gain 0 0
G20 Input Enable 20x gain 0 0
GNDEN Input Ground the negative 0 0
input to comparator
when true
HOLD Input Sample & Hold signal. 1 1
Sample analog signal
when low. Hold signal
when high. If gain
stages are used, this
signal must go active
when ACLK is high.
IREFEN Input Enables Band-gap 0 0
reference as AREF
signal to DAC
MUXEN_7 Input Input Mux bit 7 0 0
NEGSEL_2 Input Input Mux for negative 0 0
input for differential
signal, bit 2
signal, bit 1
signal, bit 0
PASSEN Input Enable pass-gate of 1 1
gain stages.
PRECH Input Precharge output latch 1 1
of comparator. (Active
low)
259
2467E–AVR–05/02
Table 104. Boundary-scan Signals for the ADC (Continued)
SCTEST Input Switch-cap TEST 0 0
enable. Output from
x10 gain stage send
out to Port Pin having
ADC_4
ST Input Output of gain stages 0 0
will settle faster if this
signal is high first two
ACLK periods after
AMPEN goes high.
VCCREN Input Selects Vcc as the 0 0
ACC reference
voltage.
Note: Incorrect setting of the switches in Figure 130 will make signal contention and may dam-
age the part. There are several input choices to the S&H circuitry on the negative input of
the output comparator in Figure 130. Make sure only one path is selected from either one
ADC pin, Bandgap reference source, or Ground.
If the ADC is not to be used during scan, the recommended input values from Table 104
should be used. The user is recommended not to use the Differential Gain stages dur-
ing scan. Switch-Cap based gain stages require fast operation and accurate timing
which is difficult to obtain when used in a scan chain. Details concerning operations of
the differential gain stage is therefore not provided. For the same reason, the ADC High
Speed mode (ADHSM) bit does not make any sense during boundary-scan operation.
The AVR ADC is based on the analog circuitry shown in Figure 130 with a successive
approximation algorithm implemented in the digital logic. When used in Boundary-scan,
the problem is usually to ensure that an applied analog voltage is measured within some
limits. This can easily be done without running a successive approximation algorithm:
apply the lower limit on the digital DAC[9:0] lines, make sure the output from the com-
parator is low, then apply the upper limit on the digital DAC[9:0] lines, and verify the
output from the comparator to be high.
The ADC need not be used for pure connectivity testing, since all analog inputs are
shared with a digital port pin as well.
When using the ADC, remember the following
• The Port Pin for the ADC channel in use must be configured to be an input with pull-
up disabled to avoid signal contention.
• In normal mode, a dummy conversion (consisting of 10 comparisons) is performed
when enabling the ADC. The user is advised to wait at least 200ns after enabling
the ADC before controlling/observing any ADC signal, or perform a dummy
conversion before using the first result.
• The DAC values must be stable at the midpoint value 0x200 when having the HOLD
signal low (Sample mode).
260 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
As an example, consider the task of verifying a 1.5V ± 5% input signal at ADC channel 3
when the power supply is 5.0V and AREF is externally connected to VCC.
The lower limit is: 1024 ⋅ 1.5V ⋅ 0,95 ⁄ 5V = 291 = 0x123

The upper limit is: 1024 ⋅ 1.5V ⋅ 1.05 ⁄ 5V = 323 = 0x143
The recommended values from Table 104 are used unless other values are given in the
algorithm in Table 105. Only the DAC and Port Pin values of the Scan Chain are shown.
The column “Actions” describes what JTAG instruction to be used before filling the
Boundary-scan Register with the succeeding columns. The verification should be done
on the data scanned out when scanning in the data on the same row in the table.
Table 105. Algorithm for Using the ADC

PA3.
PA3. PA3. Pullup_
Step Actions ADCEN DAC MUXEN HOLD PRECH Data Control Enable
SAMPLE_
1 1 0x200 0x08 1 1 0 0 0
PRELOAD
2 EXTEST 1 0x200 0x08 0 1 0 0 0
3 1 0x200 0x08 1 1 0 0 0
4 1 0x123 0x08 1 1 0 0 0
5 1 0x123 0x08 1 0 0 0 0
Verify the
COMP bit
6 1 0x200 0x08 1 1 0 0 0
scanned
out to be 0
7 1 0x200 0x08 0 1 0 0 0
8 1 0x200 0x08 1 1 0 0 0
9 1 0x143 0x08 1 1 0 0 0
10 1 0x143 0x08 1 0 0 0 0
Verify the
COMP bit
11 1 0x200 0x08 1 1 0 0 0
scanned
out to be 1
Using this algorithm, the timing constraint on the HOLD signal constrains the TCK clock
frequency. As the algorithm keeps HOLD high for five steps, the TCK clock frequency
has to be at least five times the number of scan bits divided by the maximum hold time,
thold,max
261
2467E–AVR–05/02
ATmega128 Boundary- Table 106 shows the Scan order between TDI and TDO when the Boundary-scan Chain
scan Order is selected as data path. Bit 0 is the LSB; the first bit scanned in, and the first bit
scanned out. The scan order follows the pin-out order as far as possible. Therefore, the
bits of Port A is scanned in the opposite bit order of the other ports. Exceptions from the
rules are the Scan chains for the analog circuits, which constitute the most significant
bits of the scan chain regardless of which physical pin they are connected to. In Figure
123, PXn. Data corresponds to FF0, PXn. Control corresponds to FF1, and PXn.
Pullup_enable corresponds to FF2. Bit 2, 3, 4, and 5 of Port C is not in the scan chain,
since these pins constitute the TAP pins when the JTAG is enabled.
Table 106. ATmega128 Boundary-scan Order

Bit Number Signal Name Module
204 AC_IDLE Comparator
203 ACO
202 ACME
201 AINBG
200 COMP ADC
(1)
199 PRIVATE_SIGNAL1
198 ACLK
197 ACTEN
196 ADHSM
195 ADCBGEN
194 ADCEN
193 AMPEN
192 DAC_9
191 DAC_8
190 DAC_7
189 DAC_6
188 DAC_5
187 DAC_4
262 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 106. ATmega128 Boundary-scan Order (Continued)

186 DAC_3 ADC
185 DAC_2
184 DAC_1
183 DAC_0
182 EXTCH
181 G10
180 G20
179 GNDEN
178 HOLD
177 IREFEN
176 MUXEN_7
175 MUXEN_6
174 MUXEN_5
173 MUXEN_4
172 MUXEN_3
171 MUXEN_2
170 MUXEN_1
169 MUXEN_0
168 NEGSEL_2
167 NEGSEL_1
166 NEGSEL_0
165 PASSEN
164 PRECH
163 SCTEST
162 ST
161 VCCREN
160 PEN Programming enable (observe only)
263
2467E–AVR–05/02
159 PE0.Data Port E
158 PE0.Control
157 PE0.Pullup_Enable
156 PE1.Data
155 PE1.Control
153 PE2.Data
152 PE2.Control
150 PE3.Data
149 PE3.Control
147 PE4.Data
146 PE4.Control
144 PE5.Data
143 PE5.Control
141 PE6.Data
140 PE6.Control
138 PE7.Data
137 PE7.Control
135 PB0.Data Port B
134 PB0.Control
133 PB0.Pullup_Enable
132 PB1.Data
131 PB1.Control
129 PB2.Data
128 PB2.Control
126 PB3.Data
125 PB3.Control
264 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)

123 PB4.Data Port B
122 PB4.Control
120 PB5.Data
119 PB5.Control
117 PB6.Data
116 PB6.Control
114 PB7.Data
113 PB7.Control
111 PG3.Data Port G
110 PG3.Control
109 PG3.Pullup_Enable
108 PG4.Data
107 PG4.Control
105 TOSC 32 kHz Timer Oscillator
104 TOSCON
103 RSTT Reset Logic
102 RSTHV (Observe-only)
101 EXTCLKEN Enable signals for main Clock/Oscillators

100 OSCON
99 RCOSCEN
98 OSC32EN
97 EXTCLK (XTAL1) Clock input and Osillators for the main clock
96 OSCCK (Observe-only)
95 RCCK
94 OSC32CK
93 TWIEN TWI
265
2467E–AVR–05/02
92 PD0.Data Port D
91 PD0.Control
90 PD0.Pullup_Enable
89 PD1.Data
88 PD1.Control
86 PD2.Data
85 PD2.Control
83 PD3.Data
82 PD3.Control
80 PD4.Data
79 PD4.Control
77 PD5.Data
76 PD5.Control
74 PD6.Data
73 PD6.Control
71 PD7.Data
70 PD7.Control
68 PG0.Data Port G
67 PG0.Control
65 PG1.Data
64 PG1.Control
266 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)

62 PC0.Data Port C
61 PC0.Control
60 PC0.Pullup_Enable
59 PC1.Data
58 PC1.Control
56 PC2.Data
55 PC2.Control
53 PC3.Data
52 PC3.Control
50 PC4.Data
49 PC4.Control
47 PC5.Data
46 PC5.Control
44 PC6.Data
43 PC6.Control
41 PC7.Data
40 PC7.Control
38 PG2.Data Port G
37 PG2.Control
35 PA7.Data Port A
34 PA7.Control
33 PA7.Pullup_Enable
32 PA6.Data
31 PA6.Control
29 PA5.Data
28 PA5.Control
267
2467E–AVR–05/02
26 PA4.Data Port A
25 PA4.Control
23 PA3.Data
22 PA3.Control
20 PA2.Data
19 PA2.Control
17 PA1.Data
16 PA1.Control
14 PA0.Data
13 PA0.Control
11 PF3.Data Port F
10 PF3.Control
9 PF3.Pullup_Enable
8 PF2.Data
7 PF2.Control
6 PF2.Pullup_Enable
5 PF1.Data
4 PF1.Control
3 PF1.Pullup_Enable
2 PF0.Data
1 PF0.Control
0 PF0.Pullup_Enable
Note: 1. PRIVATE_SIGNAL1 should always scanned in as zero.
Boundary-scan Boundary-scan Description Language (BSDL) files describe Boundary-scan capable

Description Language devices in a standard format used by automated test-generation software. The order
Files and function of bits in the Boundary-scan Data Register are included in this description.
268 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Boot Loader Support The Boot Loader Support provides a real Read-While-Write Self-Programming mecha-
nism for downloading and uploading program code by the MCU itself. This feature
– Read-While-Write
allows flexible application software updates controlled by the MCU using a Flash-resi-
Self-Programming dent Boot Loader program. The Boot Loader program can use any available data
interface and associated protocol to read code and write (program) that code into the
Flash memory, or read the code from the program memory. The program code within
the Boot Loader section has the capability to write into the entire Flash, including the
Boot Loader memory. The Boot Loader can thus even modify itself, and it can also
erase itself from the code if the feature is not needed anymore. The size of the Boot
Loader memory is configurable with fuses and the Boot Loader has two separate sets of
Boot Lock bits which can be set independently. This gives the user a unique flexibility to
select different levels of protection.
Boot Loader Features • Read-While-Write Self-Programming

• Flexible Boot Memory Size
• High Security (Separate Boot Lock Bits for a Flexible Protection)
• Separate Fuse to Select Reset Vector
• Optimized Page(1) Size
• Code Efficient Algorithm
• Efficient Read-Modify-Write Support
Note: 1. A page is a section in the flash consisting of several bytes (see Table 123 on page
287) used during programming. The page organization does not affect normal
operation.
Application and Boot The Flash memory is organized in two main sections, the Application section and the
Loader Flash Sections Boot Loader section (see Figure 132). The size of the different sections is configured by
the BOOTSZ fuses as shown in Table on page 280 and Figure 132. These two sections
can have different level of protection since they have different sets of Lock bits.
Application Section The application section is the section of the Flash that is used for storing the application
code. The protection level for the application section can be selected by the application
Boot Lock bits (Boot Lock bits 0), see Table on page 271. The application section can
never store any Boot Loader code since the SPM instruction is disabled when executed
from the application section.
Boot Loader Section – BLS While the application section is used for storing the application code, the The Boot
Loader software must be located in the BLS since the SPM instruction can initiate a pro-
gramming when executing from the BLS only. The SPM instruction can access the
entire Flash, including the BLS itself. The protection level for the Boot Loader section
can be selected by the Boot Loader Lock bits (Boot Lock bits 1), see Table 109 on page
272.
Read-While-Write and No Whether the CPU supports Read-While-Write or if the CPU is halted during a Boot
Read-While-Write Flash Loader software update is dependent on which address that is being programmed. In
Sections addition to the two sections that are configurable by the BOOTSZ fuses as described
above, the Flash is also divided into two fixed sections, the Read-While-Write (RWW)
section and the No Read-While-Write (NRWW) section. The limit between the RWW-
and NRWW sections is given in Table Note: on page 280 and Figure 132 on page 271.
The main difference between the two sections is:
• When erasing or writing a page located inside the RWW section, the NRWW section
can be read during the operation.
• When erasing or writing a page located inside the NRWW section, the CPU is halted
during the entire operation.
269
2467E–AVR–05/02
Note that the user software can never read any code that is located inside the RWW
section during a Boot Loader software operation. The syntax “Read-While-Write sec-
tion” refers to which section that is being programmed (erased or written), not which
section that actually is being read during a Boot Loader software update.
Read-While-Write Section – If a Boot Loader software update is programming a page inside the RWW section, it is
RWW possible to read code from the Flash, but only code that is located in the NRWW sec-
tion. During an on-going programming, the software must ensure that the RWW section
never is being read. If the user software is trying to read code that is located inside the
RWW section (i.e., by a call/jmp/lpm or an interrupt) during programming, the software
might end up in an unknown state. To avoid this, the interrupts should either be disabled
or moved to the Boot Loader Section. The Boot Loader section is always located in the
NRWW section. The RWW Section Busy bit (RWWSB) in the Store Program Memory
Control Register (SPMCR) will be read as logical one as long as the RWW section is
blocked for reading. After a programming is completed, the RWWSB must be cleared by
software before reading code located in the RWW section. See “Store Program Memory
Control Register – SPMCR” on page 273. for details on how to clear RWWSB.
No Read-While-Write Section The code located in the NRWW section can be read when the Boot Loader software is
– NRWW updating a page in the RWW section. When the Boot Loader code updates the NRWW
section, the CPU is halted during the entire page erase or page write operation.
Table 107. Read-While-Write Features
Which Section does the Z- Which Section can be Read-While-
pointer Address During the Read During Is the CPU Write
Programming? Programming? Halted? Supported?
RWW section NRWW section No Yes
NRWW section None Yes No
Figure 131. Read-While-Write vs. No Read-While-Write
Read-While-Write
(RWW) Section
Z-pointer
Addresses NRWW
Z-pointer Section
Addresses RWW No Read-While-Write
Section (NRWW) Section
CPU is Halted
During the Operation
Code Located in
NRWW Section
Can be Read During
the Operation
270 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 132. Memory Sections(1)

Program Memory Program Memory
BOOTSZ = '11' BOOTSZ = '10'
$0000 $0000
Read-While-Write Section
Application Flash Section Application Flash Section
No Read-While-Write Section
End RWW End RWW
Start NRWW Start NRWW
Application Flash Section Application Flash Section
End Application
End Application Start Boot Loader
Start Boot Loader Boot Loader Flash Section
Boot Loader Flash Section
Flashend Flashend
Program Memory Program Memory

BOOTSZ = '01' BOOTSZ = '00'
$0000 $0000
Application Flash Section Application flash Section
End RWW, End Application

End RWW
Start NRWW Start NRWW, Start Boot Loader
Application Flash Section
End Application
Start Boot Loader
Flashend Flashend
Note: 1. The parameters in the figure above are given in Table on page 280.
Boot Loader Lock Bits If no Boot Loader capability is needed, the entire Flash is available for application code.
The Boot Loader has two separate sets of Boot Lock bits which can be set indepen-
dently. This gives the user a unique flexibility to select different levels of protection.
The user can select:
• To protect the entire Flash from a software update by the MCU
• To protect only the Boot Loader Flash section from a software update by the MCU
• To protect only the Application Flash section from a software update by the MCU
• Allow software update in the entire Flash
See Table 108 and Table 109 for further details. The Boot Lock bits can be set in soft-
ware and in Serial or Parallel Programming mode, but they can be cleared by a chip
erase command only. The general Write Lock (Lock bit mode 2) does not control the
programming of the Flash memory by SPM instruction. Similarly, the general
Read/Write Lock (Lock bit mode 1) does not control reading nor writing by LPM/SPM, if
it is attempted.
271
2467E–AVR–05/02
Table 108. Boot Lock Bit0 Protection Modes (Application Section)(1)
BLB0 mode BLB02 BLB01 Protection
No restrictions for SPM or LPM accessing the Application
1 1 1
section.
2 1 0 SPM is not allowed to write to the Application section.
SPM is not allowed to write to the Application section, and
LPM executing from the Boot Loader section is not
3 0 0 allowed to read from the Application section. If interrupt
vectors are placed in the Boot Loader section, interrupts
are disabled while executing from the Application section.
LPM executing from the Boot Loader section is not
allowed to read from the Application section. If interrupt
4 0 1
Note: 1. “1” means unprogrammed, “0” means programmed
Table 109. Boot Lock Bit1 Protection Modes (Boot Loader Section)(1)
BLB1 mode BLB12 BLB11 Protection
No restrictions for SPM or LPM accessing the Boot Loader
1 1 1
section.
2 1 0 SPM is not allowed to write to the Boot Loader section.
SPM is not allowed to write to the Boot Loader section,
and LPM executing from the Application section is not
3 0 0 allowed to read from the Boot Loader section. If interrupt
vectors are placed in the Application section, interrupts are
disabled while executing from the Boot Loader section.
LPM executing from the Application section is not allowed
to read from the Boot Loader section. If interrupt vectors
4 0 1
are placed in the Application section, interrupts are
disabled while executing from the Boot Loader section.
Note: 1. “1” means unprogrammed, “0´means programmed
Entering the Boot Loader Entering the Boot Loader takes place by a jump or call from the application program.
Program This may be initiated by a trigger such as a command received via USART, or SPI inter-
face. Alternatively, the Boot Reset Fuse can be programmed so that the Reset Vector
Resetis pointing to the Boot Flash start address after a reset. In this case, the Boot
Loader is started after a reset. After the application code is loaded, the program can
start executing the application code. Note that the fuses cannot be changed by the MCU
itself. This means that once the Boot Reset Fuse is programmed, the Reset Vector will
always point to the Boot Loader Reset and the fuse can only be changed through the
serial or parallel programming interface.
Table 110. Boot Reset Fuse(1)
BOOTRST Reset Address
1 Reset Vector = Application Reset (address $0000)
0 Reset Vector = Boot Loader Reset (see Table 112 on page 280)
Note: 1. “1” means unprogrammed, “0´means programmed
272 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Store Program Memory The Store Program Memory Control Register contains the control bits needed to control
Control Register – SPMCR the Boot Loader operations.
Bit 7 6 5 4 3 2 1 0
SPMIE RWWSB – RWWSRE BLBSET PGWRT PGERS SPMEN SPMCR
• Bit 7 – SPMIE: SPM Interrupt Enable

When the SPMIE bit is written to one, and the I-bit in the Status Register is set (one), the
SPM ready interrupt will be enabled. The SPM ready Interrupt will be executed as long
as the SPMEN bit in the SPMCR Register is cleared.
• Bit 6 – RWWSB: Read-While-Write Section Busy

When a Self-Programming (page erase or page write) operation to the RWW section is
initiated, the RWWSB will be set (one) by hardware. When the RWWSB bit is set, the
RWW section cannot be accessed. The RWWSB bit will be cleared if the RWWSRE bit
is written to one after a self-programming operation is completed. Alternatively the
RWWSB bit will automatically be cleared if a page load operation is initiated.

This bit is a reserved bit in the ATmega128 and always read as zero.
• Bit 4 – RWWSRE: Read-While-Write Section Read Enable

When Programming (page erase or page write) to the RWW section, the RWW section
is blocked for reading (the RWWSB will be set by hardware). To re-enable the RWW
section, the user software must wait until the programming is completed (SPMEN will be
cleared). Then, if the RWWSRE bit is written to one at the same time as SPMEN, the
next SPM instruction within four clock cycles re-enables the RWW section. The RWW
section cannot be re-enabled while the Flash is busy with a page erase or a page write
(SPMEN is set). If the RWWSRE bit is written while the Flash is being loaded, the Flash
load operation will abort and the data loaded will be lost.
• Bit 3 – BLBSET: Boot Lock Bit Set

If this bit is written to one at the same time as SPMEN, the next SPM instruction within
four clock cycles sets Boot Lock bits, according to the data in R0. The data in R1 and
the address in the Z-pointer are ignored. The BLBSET bit will automatically be cleared
upon completion of the lock bit set, or if no SPM instruction is executed within four clock
cycles.
An LPM instruction within three cycles after BLBSET and SPMEN are set in the SPMCR
Register, will read either the Lock bits or the Fuse bits (depending on Z0 in the Z-
pointer) into the destination register. See “Reading the Fuse and Lock Bits from Soft-
ware” on page 277 for details.
• Bit 2 – PGWRT: Page Write

four clock cycles executes page write, with the data stored in the temporary buffer. The
page address is taken from the high part of the Z-pointer. The data in R1 and R0 are
ignored. The PGWRT bit will auto-clear upon completion of a page write, or if no SPM
instruction is executed within four clock cycles. The CPU is halted during the entire page
write operation if the NRWW section is addressed.
273
2467E–AVR–05/02
• Bit 1 – PGERS: Page Erase
four clock cycles executes page erase. The page address is taken from the high part of
the Z-pointer. The data in R1 and R0 are ignored. The PGERS bit will auto-clear upon
completion of a page erase, or if no SPM instruction is executed within four clock cycles.
The CPU is halted during the entire page write operation if the NRWW section is
addressed.
• Bit 0 – SPMEN: Store Program Memory Enable

This bit enables the SPM instruction for the next four clock cycles. If written to one
together with either RWWSRE, BLBSET, PGWRT’ or PGERS, the following SPM
instruction will have a special meaning, see description above. If only SPMEN is written,
the following SPM instruction will store the value in R1:R0 in the temporary page buffer
addressed by the Z-pointer. The LSB of the Z-pointer is ignored. The SPMEN bit will
auto-clear upon completion of an SPM instruction, or if no SPM instruction is executed
within four clock cycles. During page erase and page write, the SPMEN bit remains high
until the operation is completed.
Writing any other combination than “10001”, “01001”, “00101”, “00011” or “00001” in the
lower five bits will have no effect.
Addressing the Flash The Z-pointer together with RAMPZ are used to address the SPM commands. For
During Self- details on how to use th RAMPZ, see “RAM Page Z Select Register – RAMPZ” on page
Programming 12.
Bit 15 14 13 12 11 10 9 8
ZH (R31) Z15 Z14 Z13 Z12 Z11 Z10 Z9 Z8
ZL (R30) Z7 Z6 Z5 Z4 Z3 Z2 Z1 Z0
7 6 5 4 3 2 1 0
Since the Flash is organized in pages (see Table 123 on page 287), the program
counter can be treated as having two different sections. One section, consisting of the
least significant bits, is addressing the words within a page, while the most significant
bits are addressing the pages. This is shown in Figure 133. Note that the page erase
and page write operations are addressed independently. Therefore it is of major impor-
tance that the Boot Loader software addresses the same page in both the page erase
and page write operation. Once a programming operation is initiated, the address is
latched and the Z-pointer/RAMPZ can be used for other operations.
The only SPM operation that does not use the Z-pointer/RAMPZ is setting the Boot
Loader Lock bits. The content of the Z-pointer/RAMPZ is ignored and will have no effect
on the operation. The (E)LPM instruction does also use the Z-pointer/RAMPZ to store
the address. Since this instruction addresses the Flash byte by byte, also the LSB (bit
Z0) of the Z-pointer is used.
274 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 133. Addressing the Flash During SPM(1)

RAMPZ BIT 15 ZPCMSB ZPAGEMSB 1 0
0 Z - REGISTER
PCMSB PAGEMSB
PROGRAM
PCPAGE PCWORD
COUNTER
PAGE ADDRESS WORD ADDRESS

WITHIN THE FLASH WITHIN A PAGE
PROGRAM MEMORY PAGE PCWORD[PAGEMSB:0]:

PAGE 00
INSTRUCTION WORD
01
02
PAGEEND
Note: 1. The different variables used in Figure 133 are listed in Table 114 on page 281.
Self-Programming the The program memory is updated in a page by page fashion. Before programming a
Flash page with the data stored in the temporary page buffer, the page must be erased. The
temporary page buffer is filled one word at a time using SPM and the buffer can be filled
either before the page erase command or between a page erase and a page write
operation:
Alternative 1, fill the buffer before a page erase
• Fill temporary page buffer
• Perform a page erase
• Perform a page write
Alternative 2, fill the buffer after page erase
• Perform a page erase
• Fill temporary page buffer
• Perform a page write
If only a part of the page needs to be changed, the rest of the page must be stored (for
example in the temporary page buffer) before the erase, and then be rewritten. When
using alternative 1, the Boot Loader provides an effective Read-Modify-Write feature
which allows the user software to first read the page, do the necessary changes, and
then write back the modified data. If alternative 2 is used, it is not possible to read the
old data while loading since the page is already erased. The temporary page buffer can
be accessed in a random sequence. It is essential that the page address used in both
the page erase and page write operation is addressing the same page. See “Simple
Assembly Code Example for a Boot Loader” on page 278 for an assembly code
example.
275
2467E–AVR–05/02
Performing Page Erase by To execute page erase, set up the address in the Z-pointer and RAMPZ, write
SPM “X0000011” to SPMCR and execute SPM within four clock cycles after writing SPMCR.
The data in R1 and R0 is ignored. The page address must be written to PCPAGE in the
Z-register. Other bits in the Z-pointer will be ignored during this operation.
• Page Erase to the RWW section: The NRWW section can be read during the page
erase.
• Page Erase to the NRWW section: The CPU is halted during the operation.
Filling the Temporary Buffer To write an instruction word, set up the address in the Z-pointer and data in R1:R0, write
(Page Loading) “00000001” to SPMCR and execute SPM within four clock cycles after writing SPMCR.
The content of PCWORD in the Z-register is used to address the data in the temporary
buffer. The temporary buffer will auto-erase after a page write operation or by writing the
RWWSRE bit in SPMCR. It is also erased after a System Reset. Note that it is not pos-
sible to write more than one time to each address without erasing the temporary buffer.
Performing a Page Write To execute page write, set up the address in the Z-pointer and RAMPZ, write
“X0000101” to SPMCR and execute SPM within four clock cycles after writing SPMCR.
The data in R1 and R0 is ignored. The page address must be written to PCPAGE. Other
bits in the Z-pointer will be ignored during this operation.
• Page Write to the RWW section: The NRWW section can be read during the page
write.
• Page Write to the NRWW section: The CPU is halted during the operation.
Using the SPM Interrupt If the SPM interrupt is enabled, the SPM interrupt will generate a constant interrupt
when the SPMEN bit in SPMCR is cleared. This means that the interrupt can be used
instead of polling the SPMCR Register in software. When using the SPM interrupt, the
interrupt vectors should be moved to the BLS section to avoid that an interrupt is
accessing the RWW section when it is blocked for reading. How to move the interrupts
is described in “Interrupts” on page 54.
Consideration While Updating Special care must be taken if the user allows the Boot Loader section to be updated by
BLS leaving Boot Lock bit11 unprogrammed. An accidental write to the Boot Loader itself can
corrupt the entire Boot Loader, and further software updates might be impossible. If it is
not necessary to change the Boot Loader software itself, it is recommended to program
the Boot Lock bit11 to protect the Boot Loader software from any internal software
changes.
Prevent Reading the RWW During seLf-Programming (either page erase or page write), the RWW section is always
Section During Self- blocked for reading. The user software itself must prevent that this section is addressed
Programming during the Self-Programming operation. The RWWSB in the SPMCR will be set as long
as the RWW section is busy. During Self-Programming the interrupt vector table should
be moved to the BLS as described in “Interrupts” on page 54, or the interrupts must be
disabled. Before addressing the RWW section after the programming is completed, the
user software must clear the RWWSB by writing the RWWSRE. See “Simple Assembly
Code Example for a Boot Loader” on page 278 for an example.
Setting the Boot Loader Lock To set the Boot Loader Lock bits, write the desired data to R0, write “X0001001” to
Bits by SPM SPMCR and execute SPM within four clock cycles after writing SPMCR. The only
accessible lock bits are the Boot Lock bits that may prevent the Application and Boot
Loader section from any software update by the MCU.
Bit 7 6 5 4 3 2 1 0
R0 1 1 BLB12 BLB11 BLB02 BLB01 1 1
276 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
See Table 108 and Table 109 for how the different settings of the Boot Loader Bits
affect the Flash access.
If bits 5..2 in R0 are cleared (zero), the corresponding Boot Lock bit will be programmed
if an SPM instruction is executed within four cycles after BLBSET and SPMEN are set in
SPMCR. The Z-pointer is don’t care during this operation, but for future compatibility it is
recommended to load the Z-pointer with $0001 (same as used for reading the Lock
bits). For future compatibility It is also recommended to set bits 7, 6, 1, and 0 in R0 to “1”
when writing the lock-bits. When programming the Lock Bits the entire Flash can be
read during the operation.
EEPROM Write Prevents Note that an EEPROM write operation will block all software programming to Flash.
Writing to SPMCR Reading the Fuses and Lock bits from software will also be prevented during the
EEPROM write operation. It is recommended that the user checks the status bit (EEWE)
in the EECR Register and verifies that the bit is cleared before writing to the SPMCR
Register.
Reading the Fuse and Lock It is possible to read both the Fuse and Lock bits from software. To read the Lock bits,
Bits from Software load the Z-pointer with $0001 and set the BLBSET and SPMEN bits in SPMCR. When
an LPM instruction is executed within three CPU cycles after the BLBSET and SPMEN
bits are set in SPMCR, the value of the Lock bits will be loaded in the destination regis-
ter. The BLBSET and SPMEN bits will auto-clear upon completion of reading the Lock
bits or if no LPM instruction is executed within three CPU cycles or no SPM instruction is
executed within four CPU cycles. When BLBSET and SPMEN are cleared, LPM will
work as described in the Instruction set Manual.
Bit 7 6 5 4 3 2 1 0
Rd – – BLB12 BLB11 BLB02 BLB01 LB2 LB1
The algorithm for reading the Fuse Low bits is similar to the one described above for
reading the Lock bits. To read the Fuse Low bits, load the Z-pointer with $0000 and set
the BLBSET and SPMEN bits in SPMCR. When an LPM instruction is executed within
three cycles after the BLBSET and SPMEN bits are set in the SPMCR, the value of the
Fuse Low bits (FLB) will be loaded in the destination register as shown below. Refer to
Table 119 on page 284 for a detailed description and mapping of the Fuse Low bits.
Bit 7 6 5 4 3 2 1 0
Rd FLB7 FLB6 FLB5 FLB4 FLB3 FLB2 FLB1 FLB0
Similarly, when reading the Fuse High bits, load $0003 in the Z-pointer. When an LPM
instruction is executed within three cycles after the BLBSET and SPMEN bits are set in
the SPMCR, the value of the Fuse High bits (FHB) will be loaded in the destination reg-
ister as shown below. Refer to Table 118 on page 284 for detailed description and
mapping of the Fuse High bits.
Bit 7 6 5 4 3 2 1 0
Rd FHB7 FHB6 FHB5 FHB4 FHB3 FHB2 FHB1 FHB0
When reading the Extended Fuse bits, load $0002 in the Z-pointer. When an LPM
instruction is executed within three cycles after the BLBSET and SPMEN bits are set in
the SPMCR, the value of the Extended Fuse bits (EFB) will be loaded in the destination
register as shown below. Refer to Table 117 on page 283 for detailed description and
mapping of the Fuse High bits.
Bit 7 6 5 4 3 2 1 0
Rd – – – – – – EFB1 EFB0
277
2467E–AVR–05/02
Fuse and Lock bits that are programmed, will be read as zero. Fuse and Lock bits that
are unprogrammed, will be read as one.
Preventing Flash Corruption During periods of low VCC, the Flash program can be corrupted because the supply volt-
age is too low for the CPU and the Flash to operate properly. These issues are the same
as for board level systems using the Flash, and the same design solutions should be
applied.
A Flash program corruption can be caused by two situations when the voltage is too low.
First, a regular write sequence to the Flash requires a minimum voltage to operate cor-
rectly. Secondly, the CPU itself can execute instructions incorrectly, if the supply voltage
for executing instructions is too low.
Flash corruption can easily be avoided by following these design recommendations (one
is sufficient):
1. If there is no need for a Boot Loader update in the system, program the Boot
Loader Lock bits to prevent any Boot Loader software updates.
2. Keep the AVR RESET active (low) during periods of insufficient power supply
voltage. This can be done by enabling the internal Brown-out Detector (BOD) if
the operating voltage matches the detection level. If not, an external low VCC
Reset Protection circuit can be used. If a Reset occurs while a write operation is
in progress, the write operation will be completed provided that the power supply
voltage is sufficient.
3. Keep the AVR core in Power-down Sleep mode during periods of low VCC. This
will prevent the CPU from attempting to decode and execute instructions, effec-
tively protecting the SPMCR Register and thus the Flash from unintentional
writes.
Programming Time for Flash The calibrated RC Oscillator is used to time Flash accesses. Table 111 shows the typi-
when Using SPM cal programming time for Flash accesses from the CPU.
Table 111. SPM Programming Time.
Symbol Min Programming Time Max Programming Time
Flash write (page erase, page write,
3.7 ms 4.5 ms
and write lock bits by SPM)
Simple Assembly Code ;-the routine writes one page of data from RAM to Flash
Example for a Boot Loader ; the first data location in RAM is pointed to by the Y pointer
; the first data location in Flash is pointed to by the Z-pointer
;-error handling is not included
;-the routine must be placed inside the boot space
; (at least the Do_spm sub routine). Only code inside NRWW section can
; be read during self-programming (page erase and page write).
;-registers used: r0, r1, temp1 (r16), temp2 (r17), looplo (r24),
; loophi (r25), spmcrval (r20)
; storing and restoring of registers is not included in the routine
; register usage can be optimized at the expense of code size
;-It is assumed that either the interrupt table is moved to the Boot
; loader section or that the interrupts are disabled.
.equ PAGESIZEB = PAGESIZE*2 ;PAGESIZEB is page size in BYTES, not
words
.org SMALLBOOTSTART
Write_page:
; page erase
ldi spmcrval, (1<<PGERS) | (1<<SPMEN)
call Do_spm
278 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
; re-enable the RWW section

ldi spmcrval, (1<<RWWSRE) | (1<<SPMEN)
call Do_spm
; transfer data from RAM to Flash page buffer

ldi looplo, low(PAGESIZEB) ;init loop variable
ldi loophi, high(PAGESIZEB) ;not required for PAGESIZEB<=256
Wrloop:
ld r0, Y+
ld r1, Y+
ldi spmcrval, (1<<SPMEN)
call Do_spm
adiw ZH:ZL, 2
sbiw loophi:looplo, 2 ;use subi for PAGESIZEB<=256
brne Wrloop
; execute page write

subi ZL, low(PAGESIZEB) ;restore pointer
sbci ZH, high(PAGESIZEB) ;not required for PAGESIZEB<=256
ldi spmcrval, (1<<PGWRT) | (1<<SPMEN)
call Do_spm

call Do_spm
; read back and check, optional

ldi looplo, low(PAGESIZEB) ;init loop variable
ldi loophi, high(PAGESIZEB) ;not required for PAGESIZEB<=256
subi YL, low(PAGESIZEB) ;restore pointer
sbci YH, high(PAGESIZEB)
Rdloop:
lpm r0, Z+
ld r1, Y+
cpse r0, r1
jmp Error
sbiw loophi:looplo, 1 ;use subi for PAGESIZEB<=256
brne Rdloop
; return to RWW section

; verify that RWW section is safe to read
Return:
lds temp1, SPMCR
sbrs temp1, RWWSB ; If RWWSB is set, the RWW section is not ready
yet
ret
call Do_spm
rjmp Return
Do_spm:
; check for previous SPM complete
Wait_spm:
lds temp1, SPMCR
sbrc temp1, SPMEN
rjmp Wait_spm
; input: spmcrval determines SPM action
; disable interrupts if enabled, store status
in temp2, SREG
cli
; check that no EEPROM write access is present
Wait_ee:
sbic EECR, EEWE
rjmp Wait_ee
279
2467E–AVR–05/02
; SPM timed sequence
sts SPMCR, spmcrval
spm
; restore SREG (to enable interrupts if originally enabled)
out SREG, temp2
ret
ATmega128 Boot Loader In Table 112 through Table 114, the parameters used in the description of the self pro-
Parameters gramming are given.
Table 112. Boot Size Configuration

Boot
Reset
Boot Address
Application Loader End (start Boot
Boot Flash Flash Application Loader
BOOTSZ1 BOOTSZ0 Size Pages Section Section section Section)
512 $0000 - $FE00 -

1 1 4 $FDFF $FE00
words $FDFF $FFFF
1024 $0000 - $FC00 -

1 0 8 $FBFF $FC00
words $FBFF $FFFF
2048 $0000 - $F800 -

0 1 16 $F7FF $F800
words $F7FF $FFFF
4096 $0000 - $F000 -

0 0 32 $EFFF $F000
words $EFFF $FFFF
Note: The different BOOTSZ fuse configurations are shown in Figure 132
Table 113. Read-While-Write Limit(1)

Section Pages Address
Read-While-Write section (RWW) 480 $0000 - $EFFF
No Read-While-Write section (NRWW) 32 $F000 - $FFFF
Note: 1. For details about these two section, see “No Read-While-Write Section – NRWW” on
page 270 and “Read-While-Write Section – RWW” on page 270
280 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 114. Explanation of Different Variables Used in Figure 133 and the Mapping to
the Z-Pointer(3)
Corresponding
Variable Z-value Description(2)
15 Most significant bit in the program counter. (The
PCMSB
program counter is 16 bits PC[15:0])
6 Most significant bit which is used to address the
PAGEMSB words within one page (128 words in a page
requires 7 bits PC [6:0]).
Z16(1) Bit in Z-register that is mapped to PCMSB.
ZPCMSB Because Z0 is not used, the ZPCMSB equals
PCMSB + 1.
Z7 Bit in Z-register that is mapped to PCMSB.
ZPAGEMSB Because Z0 is not used, the ZPAGEMSB
equals PAGEMSB + 1.
PC[15:7] Z16(1):Z7 Program counter page address: Page select,
PCPAGE
for page erase and page write
PC[6:0] Z7:Z1 Program counter word address: Word select,
PCWORD for filling temporary buffer (must be zero during
page write operation)
Notes: 1. The Z-register is only 16 bits wide. Bit 16 is located in the RAMPZ register in the I/O
map.
2. Z0: should be zero for all SPM commands, byte select for the (E)LPM instruction.
3. See “Addressing the Flash During Self-Programming” on page 274 for details about
the use of Z-pointer during self-programming.
281
2467E–AVR–05/02
Memory
Programming
Program and Data The ATmega128 provides six Lock bits which can be left unprogrammed (“1”) or can be
Memory Lock Bits programmed (“0”) to obtain the additional features listed in Table 116. The Lock bits can
only be erased to “1” with the Chip Erase command.
Table 115. Lock Bit Byte

Lock Bit Byte Bit No. Description Default Value
7 – 1 (unprogrammed)
6 – 1 (unprogrammed)
BLB12 5 Boot lock bit 1 (unprogrammed)
LB2 1 Lock bit 1 (unprogrammed)
LB1 0 Lock bit 1 (unprogrammed)
Note: “1” means unprogrammed, “0´means programmed
Table 116. Lock Bit Protection Modes

Memory Lock Bits Protection Type
LB mode LB2 LB1
1 1 1 No memory lock features enabled.
Further programming of the Flash and EEPROM is
disabled in Parallel and SPI/JTAG Serial Programming
2 1 0
mode. The Fuse bits are locked in both Serial and Parallel
Programming mode.(1)
Further programming and verification of the Flash and
EEPROM is disabled in Parallel and SPI/JTAG Serial
3 0 0
Programming mode. The Fuse bits are locked in both
Serial and Parallel Programming mode.(1)
BLB0 mode BLB02 BLB01
No restrictions for SPM or (E)LPM accessing the
1 1 1
Application section.
2 1 0 SPM is not allowed to write to the Application section.
SPM is not allowed to write to the Application section, and
(E)LPM executing from the Boot Loader section is not
3 0 0 allowed to read from the Application section. If interrupt
(E)LPM executing from the Boot Loader section is not
allowed to read from the Application section. If interrupt
4 0 1
BLB1 mode BLB12 BLB11
282 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 116. Lock Bit Protection Modes (Continued)

Memory Lock Bits Protection Type
No restrictions for SPM or (E)LPM accessing the Boot
1 1 1
Loader section.
2 1 0 SPM is not allowed to write to the Boot Loader section.
SPM is not allowed to write to the Boot Loader section,
and (E)LPM executing from the Application section is not
allowed to read from the Boot Loader section. If interrupt
3 0 0
vectors are placed in the Application section, interrupts
are disabled while executing from the Boot Loader
section.
(E)LPM executing from the Application section is not
allowed to read from the Boot Loader section. If interrupt
4 0 1 vectors are placed in the Application section, interrupts
are disabled while executing from the Boot Loader
section.
Notes: 1. Program the fuse bits before programming the Lock bits.
2. “1” means unprogrammed, “0´means programmed
Fuse Bits The ATmega128 has three fuse bytes. Table 117 - Table 119 describe briefly the func-
tionality of all the fuses and how they are mapped into the fuse bytes. Note that the
fuses are read as logical zero, “0”, if they are programmed.
Table 117. Extended Fuse Byte

Extended Fuse Byte Bit No. Description Default Value
– 7 – 1
– 6 – 1
– 5 – 1
– 4 – 1
– 3 – 1
– 2 – 1
(1)
M103C 1 ATmega103 compatibility mode 0 (programmed)
(2)
WDTON 0 Watchdog Timer always on 1 (unprogrammed)
Notes: 1. See “ATmega103 and ATmega128 Compatibility” on page 4 for details.
2. See “Watchdog Timer Control Register – WDTCR” on page 51 for details.
283
2467E–AVR–05/02
Table 118. Fuse High Byte
Fuse High
Byte Bit No. Description Default Value
(4)
OCDEN 7 Enable OCD 1 (unprogrammed, OCD
disabled)
JTAGEN 6 Enable JTAG 0 (programmed, JTAG enabled)
(1)
SPIEN 5 Enable Serial Program and 0 (programmed, SPI prog.
Data Downloading enabled)
CKOPT(2) 4 Oscillator options 1 (unprogrammed)
EESAVE 3 EEPROM memory is preserved 1 (unprogrammed, EEPROM not
through the Chip Erase preserved)
BOOTSZ1 2 Select Boot Size (see Table 112 0 (programmed)(3)
for details)
BOOTSZ0 1 Select Boot Size (see Table 112 0 (programmed)(3)
for details)
BOOTRST 0 Select Reset Vector 1 (unprogrammed)
Notes: 1. The SPIEN fuse is not accessible in SPI Serial Programming mode.
2. The CKOPT fuse functionality depends on the setting of the CKSEL bits. See “Clock
Sources” on page 34 for details.
3. The default value of BOOTSZ1..0 results in maximum Boot Size. See Table 112 on
page 280
4. Never ship a product with the OCDEN Fuse programmed regardless of the setting of
lockbits and the JTAGEN Fuse. A programmed OCDEN Fuse enables some parts of
the clock system to be running in all sleep modes. This may increase the power
consumption.
Table 119. Fuse Low Byte
Fuse Low
Byte Bit No. Description Default Value
BODLEVEL 7 Brown out detector trigger level 1 (unprogrammed)
BODEN 6 Brown out detector enable 1 (unprogrammed, BOD disabled)
SUT1 5 Select start-up time 1 (unprogrammed)(1)
SUT0 4 Select start-up time 0 (programmed)(1)
CKSEL3 3 Select Clock source 0 (programmed)(2)
CKSEL0 0 Select Clock source 1 (unprogrammed)(2)
Notes: 1. The default value of SUT1..0 results in maximum start-up time. See Table 14 on page
38 for details.
2. The default setting of CKSEL3..0 results in Internal RC Oscillator @ 1 MHz. See
Table 6 on page 34 for details.
The status of the Fuse bits is not affected by Chip Erase. Note that the Fuse bits are
locked if Lock bit1 (LB1) is programmed. Program the Fuse bits before programming the
Lock bits.
284 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Latching of Fuses The Fuse values are latched when the device enters Programming mode and changes
of the fuse values will have no effect until the part leaves Programming mode. This does
not apply to the EESAVE fuse which will take effect once it is programmed. The fuses
are also latched on power-up in normal mode.
Signature Bytes All Atmel microcontrollers have a three-byte signature code which identifies the device.
This code can be read in both serial and parallel mode, also when the device is locked.
The three bytes reside in a separate address space.
For the ATmega128 the signature bytes are:
1. $000: $1E (indicates manufactured by Atmel)
2. $001: $97 (indicates 128KB Flash memory)
3. $002: $02 (indicates ATmega128 device when $001 is $97)
Calibration Byte The ATmega128 stores four different calibration values for the internal RC Oscillator.
These bytes resides in the signature row high byte of the addresses 0x000, 0x0001,
0x0002, and 0x0003 for 1, 2, 4, and 8 MHz respectively. During Reset, the 1 MHz value
is automatically loaded into the OSCCAL Register. If other frequencies are used, the
calibration value has to be loaded manually, see “Oscillator Calibration Register – OSC-
CAL” on page 38 for details.
285
2467E–AVR–05/02
Parallel Programming This section describes how to parallel program and verify Flash Program memory,
Parameters, Pin EEPROM Data memory, Memory Lock bits, and Fuse bits in the ATmega128. Pulses
Mapping, and are assumed to be at least 250 ns unless otherwise noted.
Commands
Signal Names In this section, some pins of the ATmega128 are referenced by signal names describing
their functionality during parallel programming, see Figure 134 and Table 120. Pins not
described in the following table are referenced by pin names.
The XA1/XA0 pins determine the action executed when the XTAL1 pin is given a posi-
tive pulse. The bit coding is shown in Table 122.
When pulsing WR or OE, the command loaded determines the action executed. The dif-
ferent Commands are shown in Table 123.
Figure 134. Parallel Programming
+5V
RDY/BSY PD1
VCC
OE PD2 +5V
WR PD3
AVCC
BS1 PD4
XA0 PD5 PB7 - PB0 DATA
XA1 PD6
PAGEL PD7
+12 V RESET
BS2 PA0
XTAL1
GND
Table 120. Pin Name Mapping

Signal Name in
Programming Mode Pin Name I/O Function
RDY/BSY 0: Device is busy programming, 1: Device is ready
PD1 O
for new command
OE PD2 I Output Enable (Active low)
WR PD3 I Write Pulse (Active low)
BS1 Byte Select 1 (“0” selects low byte, “1” selects high
PD4 I
byte)
XA0 PD5 I XTAL Action Bit 0
XA1 PD6 I XTAL Action Bit 1
286 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 120. Pin Name Mapping (Continued)

Signal Name in
Programming Mode Pin Name I/O Function
PAGEL PD7 I Program Memory and EEPROM data Page Load
BS2 Byte Select 2 (“0” selects low byte, “1” selects 2’nd
PA0 I
high byte)
DATA PB7-0 I/O Bi-directional Data bus (Output when OE is low)
Table 121. Pin Values Used to Enter Programming Mode

Pin Symbol Value
PAGEL Prog_enable[3] 0
XA1 Prog_enable[2] 0
XA0 Prog_enable[1] 0
BS1 Prog_enable[0] 0
Table 122. XA1 and XA0 Coding

XA1 XA0 Action when XTAL1 is Pulsed
Load Flash or EEPROM Address (High or low address byte determined by
0 0
BS1)
0 1 Load Data (High or Low data byte for Flash determined by BS1)
1 0 Load Command
1 1 No Action, Idle
Table 123. Command Byte Bit Coding

Command Byte Command Executed
1000 0000 Chip Erase
0100 0000 Write Fuse bits
0010 0000 Write Lock bits
0001 0000 Write Flash
0001 0001 Write EEPROM
0000 1000 Read Signature Bytes and Calibration byte
0000 0100 Read Fuse and Lock bits
0000 0010 Read Flash
0000 0011 Read EEPROM
Table 124. No. of Words in a Page and no. of Pages in the Flash
Flash Size Page Size PCWORD No. of Pages PCPAGE PCMSB
64K words (128K bytes) 128 words PC[6:0] 512 PC[15:7] 15
287
2467E–AVR–05/02
Table 125. No. of Words in a Page and no. of Pages in the EEPROM
EEPROM Size Page Size PCWORD No. of Pages PCPAGE EEAMSB
4K bytes 8 bytes EEA[2:0] 512 EEA[11:3] 8
Parallel Programming
Enter Programming Mode The following algorithm puts the device in parallel programming mode:
1. Apply 4.5 - 5.5 V between VCC and GND.
2. Set RESET to “0” and toggle XTAL1 at least SIX times.
3. Set the Prog_enable pins listed in Table 121 on page 287 to “0000” and wait at
least 100 ns.
4. Apply 11.5 - 12.5V to RESET. Any activity on Prog_enable pins within 100 ns
after +12V has been applied to RESET, will cause the device to fail entering pro-
gramming mode.
Note, if External Crystal or External RC configuration is selected, it may not be possible
to apply qualified XTAL1 pulses. In such cases, the following algorithm should be
followed:
1. Set Prog_enable pins listed in Table on page 287 to “0000”.
2. Apply 4.5 - 5.5V between VCC and GND simultanously as 11.5 - 12.5V is applied
to RESET.
3. Wait 100 ns.
4. Re-program the fuses to ensure that External Clock is selected as clock source
(CKSEL3:0 = 0b0000) If Lock bits are programmed, a Chip Erase command
must be executed before changing the fuses.
5. Exit Programming mode by power the device down or by bringing RESET pin to
0b0.
6. Entering Programming mode with the original algorithm, as described above.
Considerations for Efficient The loaded command and address are retained in the device during programming. For
Programming efficient programming, the following should be considered.
• The command needs only be loaded once when writing or reading multiple memory
locations.
• Skip writing the data value $FF, that is the contents of the entire EEPROM (unless
the EESAVE fuse is programmed) and Flash after a Chip Erase.
• Address high byte needs only be loaded before programming or reading a new 256
word window in Flash or 256-byte EEPROM. This consideration also applies to
Signature bytes reading.
Chip Erase The Chip Erase will erase the Flash and EEPROM(1) memories plus Lock bits. The Lock
bits are not reset until the program memory has been completely erased. The Fuse bits
are not changed. A Chip Erase must be performed before the Flash is reprogrammed.
Note: 1. The EEPRPOM memory is preserved during chip erase if the EESAVE fuse is
programmed.
Load Command “Chip Erase”
1. Set XA1, XA0 to “10”. This enables command loading.
2. Set BS1 to “0”.
288 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
3. Set DATA to “1000 0000”. This is the command for Chip Erase.
4. Give XTAL1 a positive pulse. This loads the command.
5. Give WR a negative pulse. This starts the Chip Erase. RDY/BSY goes low.
6. Wait until RDY/BSY goes high before loading a new command.
Programming the Flash The Flash is organized in pages, see Table 123 on page 287. When programming the
Flash, the program data is latched into a page buffer. This allows one page of program
data to be programmed simultaneously. The following procedure describes how to pro-
gram the entire Flash memory:
A. Load Command “Write Flash”
1. Set XA1, XA0 to “10”. This enables command loading.
2. Set BS1 to “0”.
3. Set DATA to “0001 0000”. This is the command for Write Flash.
4. Give XTAL1 a positive pulse. This loads the command.
B. Load Address Low byte
1. Set XA1, XA0 to “00”. This enables address loading.
2. Set BS1 to “0”. This selects low address.
3. Set DATA = Address low byte ($00 - $FF).
4. Give XTAL1 a positive pulse. This loads the address low byte.
C. Load Data Low Byte
1. Set XA1, XA0 to “01”. This enables data loading.
2. Set DATA = Data low byte ($00 - $FF).
3. Give XTAL1 a positive pulse. This loads the data byte.
D. Load Data High Byte
1. Set BS1 to “1”. This selects high data byte.
2. Set XA1, XA0 to “01”. This enables data loading.
3. Set DATA = Data high byte ($00 - $FF).
4. Give XTAL1 a positive pulse. This loads the data byte.
E. Latch Data
1. Set BS1 to “1”. This selects high data byte.
2. Give PAGEL a positive pulse. This latches the data bytes. (See Figure 136 for
signal waveforms)
F. Repeat B through E until the entire buffer is filled or until all data within the page is
loaded.
While the lower bits in the address are mapped to words within the page, the higher bits
address the pages within the FLASH. This is illustrated in Figure 135 on page 290. Note
that if less than 8 bits are required to address words in the page (pagesize < 256), the
most significant bit(s) in the address low byte are used to address the page when per-
forming a page write.
G. Load Address High byte
1. Set XA1, XA0 to “00”. This enables address loading.
2. Set BS1 to “1”. This selects high address.
3. Set DATA = Address high byte ($00 - $FF).
4. Give XTAL1 a positive pulse. This loads the address high byte.
289
2467E–AVR–05/02
H. Program Page
1. Set BS1 = “0”
2. Give WR a negative pulse. This starts programming of the entire page of data.
RDY/BSYgoes low.
3. Wait until RDY/BSY goes high. (See Figure 136 for signal waveforms)
I. Repeat B through H until the entire Flash is programmed or until all data has been
programmed.
J. End Page Programming
1. 1. Set XA1, XA0 to “10”. This enables command loading.
2. Set DATA to “0000 0000”. This is the command for No Operation.
3. Give XTAL1 a positive pulse. This loads the command, and the internal write sig-
nals are reset.
Figure 135. Addressing the Flash which is Organized in Pages

PCMSB PAGEMSB
PROGRAM
PCPAGE PCWORD
COUNTER
PAGE ADDRESS WORD ADDRESS

WITHIN THE FLASH WITHIN A PAGE
PROGRAM MEMORY PAGE PCWORD[PAGEMSB:0]:

PAGE 00
INSTRUCTION WORD
01
02
PAGEEND
Note: 1. PCPAGE and PCWORD are listed in Table 124 on page 287.
290 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 136. Programming the Flash Waveforms

F
A B C D E B C D E G H
$10 ADDR. LOW DATA LOW DATA HIGH XX ADDR. LOW DATA LOW DATA HIGH XX ADDR. HIGH XX
DATA
XA1
XA0
BS1
XTAL1
WR
RDY/BSY
RESET +12V
OE
PAGEL
BS2
Note: “XX” is don’t care. The letters refer to the programming description above.
Programming the EEPROM The EEPROM is organized in pages, see Table 124 on page 287. When programming
the EEPROM, the program data is latched into a page buffer. This allows one page of
data to be programmed simultaneously. The programming algorithm for the EEPROM
data memory is as follows (refer to “Programming the Flash” on page 289 for details on
Command, Address and Data loading):
1. A: Load Command “0001 0001”.
2. G: Load Address High Byte ($00 - $FF).
3. B: Load Address Low Byte ($00 - $FF).
4. C: Load Data ($00 - $FF).
5. E: Latch data (give PAGEL a positive pulse).
K: Repeat 3 through 5 until the entire buffer is filled.
L: Program EEPROM page:
1. Set BS1 to “0”.
2. Give WR a negative pulse. This starts programming of the EEPROM page.
RDY/BSY goes low.
3. Wait until to RDY/BSY goes high before programming the next page.
(See Figure 137 for signal waveforms.)
291
2467E–AVR–05/02
Figure 137. Programming the EEPROM Waveforms
K
A G B C E B C E L
$10 ADDR. HIGH ADDR. LOW DATA XX ADDR. LOW DATA XX
DATA
XA1
XA0
BS1
XTAL1
WR
RDY/BSY
RESET +12V
OE
PAGEL
BS2
Reading the Flash The algorithm for reading the Flash memory is as follows (refer to “Programming the
Flash” on page 289 for details on Command and Address loading):
1. A: Load Command “0000 0010”.
4. Set OE to “0”, and BS1 to “0”. The Flash word low byte can now be read at
DATA.
5. Set BS1 to “1”. The Flash word high byte can now be read at DATA.
6. Set OE to “1”.
Reading the EEPROM The algorithm for reading the EEPROM memory is as follows (refer to “Programming the
Flash” on page 289 for details on Command and Address loading):
1. A: Load Command “0000 0011”.
4. Set OE to “0”, and BS1 to “0”. The EEPROM Data byte can now be read at
DATA.
Programming the Fuse Low The algorithm for programming the Fuse Low bits is as follows (refer to “Programming
Bits the Flash” on page 289 for details on Command and Data loading):
1. A: Load Command “0100 0000”.
2. C: Load Data Low Byte. Bit n = “0” programs and bit n = “1” erases the Fuse bit.
3. Set BS1 to “0” and BS2 to “0”.
4. Give WR a negative pulse and wait for RDY/BSY to go high.
292 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Programming the Fuse High The algorithm for programming the Fuse High bits is as follows (refer to “Programming
Bits the Flash” on page 289 for details on Command and Data loading):
1. A: Load Command “0100 0000”.
3. Set BS1 to “1” and BS2 to “0”. This selects high data byte.
5. Set BS1 to “0”. This selects low data byte.
Programming the Extended The algorithm for programming the Extended Fuse bits is as follows (refer to “Program-
Fuse Bits ming the Flash” on page 289 for details on Command and Data loading):
1. A: Load Command “0100 0000”.
3. Set BS2 to “1” and BS1 to “0”. This selects extended data byte.
5. Set BS2 to “0”. This selects low data byte.
Figure 138. Programming the Fuses

Write Fuse Low byte Write Fuse high byte Write Extended Fuse byte
A C A C A C
$40 DATA XX $40 DATA XX $40 DATA XX
DATA
XA1
XA0
BS1
BS2
XTAL1
WR
RDY/BSY
RESET +12V
OE
PAGEL
Programming the Lock Bits The algorithm for programming the Lock bits is as follows (refer to “Programming the
Flash” on page 289 for details on Command and Data loading):
1. A: Load Command “0010 0000”.
2. C: Load Data Low Byte. Bit n = “0” programs the Lock bit.
The Lock bits can only be cleared by executing Chip Erase.
Reading the Fuse and Lock The algorithm for reading the Fuse and Lock bits is as follows (refer to “Programming
Bits the Flash” on page 289 for details on Command loading):
1. A: Load Command “0000 0100”.
2. Set OE to “0”, BS2 to “0”, and BS1 to “0”. The status of the Fuse Low bits can
now be read at DATA (“0” means programmed).
3. Set OE to “0”, BS2 to “1”, and BS1 to “1”. The status of the Fuse High bits can
now be read at DATA (“0” means programmed).
293
2467E–AVR–05/02
4. Set OE to “0”, BS2 to “1”, and BS1 to “0”. The status of the Extended Fuse bits
can now be read at DATA (“0” means programmed).
5. Set OE to “0”, BS2 to “0”, and BS1 to “1”. The status of the Lock bits can now be
read at DATA (“0” means programmed).
Figure 139. Mapping Between BS1, BS2 and the Fuse- and Lock Bits During Read
Fuse Low Byte 0
0
Extended Fuse byte 1
DATA
BS2
Lock bits 0
1
BS1
Fuse high byte 1
BS2
Reading the Signature Bytes The algorithm for reading the Signature bytes is as follows (refer to Programming the
Flash for details on Command and Address loading):
1. A: Load Command “0000 1000”.
2. B: Load Address Low Byte ($00 - $02).
3. Set OE to “0”, and BS1 to “0”. The selected Signature byte can now be read at
DATA.
Reading the Calibration Byte The algorithm for reading the Calibration byte is as follows (refer to Programming the
Flash for details on Command and Address loading):
1. A: Load Command “0000 1000”.
2. B: Load Address Low Byte.
3. Set OE to “0”, and BS1 to “1”. The Calibration byte can now be read at DATA.
294 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Parallel Programming Figure 1 40. Parallel Pro grammin g Timing , Inclu din g some Gene ral Timin g
Characteristics Requirements
t XLWL
t XHXL
XTAL1
t DVXH t XLDX
Data & Contol
(DATA, XA0/1, BS1, BS2)
t BVPH t PLBX t BVWL
t WLBX
PAGEL t PHPL
t WL WH
WR t PLWL
WLRL
RDY/BSY
t WLRH
Figure 14 1. Pa rallel Prog ra mmin g Timin g, L oa ding Seq ue nce with Timin g
Requirements
LOAD ADDRESS LOAD DATA LOAD DATA LOAD DATA LOAD ADDRESS
(LOW BYTE) (LOW BYTE) (HIGH BYTE) (LOW BYTE)
t XLXH tXLPH
tPLXH
XTAL1
BS1
PAGEL
DATA ADDR0 (Low Byte) DATA (Low Byte) DATA (High Byte) ADDR1 (Low Byte)
XA0
XA1
Note: The timing requirements shown in Figure 140 (i.e. tDVXH, tXHXL, and tXLDX) also apply to
loading operation.
295
2467E–AVR–05/02
Figure 142. Parallel Programming Timing, Reading Sequence (Within the Same Page)
with Timing Requirements
LOAD ADDRESS READ DATA READ DATA LOAD ADDRESS
(LOW BYTE) (LOW BYTE) (HIGH BYTE) (LOW BYTE)
tXLOL
XTAL1
tBHDV
BS1
tOLDV
OE
tOHDZ
DATA ADDR0 (Low Byte) DATA (Low Byte) DATA (High Byte) ADDR1 (Low Byte)
XA0
XA1
Note: The timing requirements shown in Figure 140 (i.e. tDVXH, tXHXL, and tXLDX) also apply to
reading operation.
Table 126. Parallel Programming Characteristics, VCC = 5 V ± 10%
VPP Programming Enable Voltage 11.5 12.5 V
IPP Programming Enable Current 250 µA
tDVXH Data and Control Valid before XTAL1 High 67 ns
tXLXH XTAL1 Low to XTAL1 High 200 ns
tXHXL XTAL1 Pulse Width High 150 ns
tXLDX Data and Control Hold after XTAL1 Low 67 ns
tXLWL XTAL1 Low to WR Low 0 ns
tXLPH XTAL1 Low to PAGEL high 0 ns
tPLXH PAGEL low to XTAL1 high 150 ns
tBVPH BS1 Valid before PAGEL High 67 ns
tPHPL PAGEL Pulse Width High 150 ns
tPLBX BS1 Hold after PAGEL Low 67 ns
tWLBX BS2/1 Hold after WR Low 67 ns
tPLWL PAGEL Low to WR Low 67 ns
tBVWL BS1 Valid to WR Low 67 ns
tWLWH WR Pulse Width Low 150 ns
tWLRL WR Low to RDY/BSY Low 0 1 µs
(1)
tWLRH WR Low to RDY/BSY High 3.7 4.5 ms
(2)
tWLRH_CE WR Low to RDY/BSY High for Chip Erase 7.5 9 ms
tXLOL XTAL1 Low to OE Low 0 ns
296 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 126. Parallel Programming Characteristics, VCC = 5 V ± 10% (Continued)

tBVDV BS1 Valid to DATA valid 0 250 ns
tOLDV OE Low to DATA Valid 250 ns
tOHDZ OE High to DATA Tri-stated 250 ns
Notes: 1. tWLRH is valid for the Write Flash, Write EEPROM, Write Fuse bits and Write Lock
bits commands.
2. tWLRH_CE is valid for the Chip Erase command.
Serial Downloading Both the Flash and EEPROM memory arrays can be programmed using the serial SPI
bus while RESET is pulled to GND. The serial interface consists of pins SCK, MOSI
(input) and MISO (output). After RESET is set low, the Programming Enable instruction
needs to be executed first before program/erase operations can be executed. NOTE, in
Table 127 on page 297, the pin mapping for SPI programming is listed. Not all parts use
the SPI pins dedicated for the internal SPI interface. Note that throughout the descrip-
tion about Serial downloading, MOSI and MISO are used to describe the serial data in
and serial data out respectively. For ATmega128 these pins are mapped to PDI and
PDO.
SPI Serial Programming

Pin Mapping Table 127. Pin Mapping SPI Serial Programming
Symbol Pins I/O Description
MOSI (PDI) PE0 I Serial data in
MISO (PDO) PE1 O Serial data out
SCK PB1 I Serial clock
297
2467E–AVR–05/02
Figure 143. SPI Serial Programming and Verify(1)
+2.7 - 5.5V
VCC
+2.7 - 5.5V(2)
PDI PE0
AVCC
PDO PE1
SCK PB1
XTAL1
RESET
GND
Notes: 1. If the device is clocked by the Internal Oscillator, it is no need to connect a clock
source to the XTAL1 pin.
2. VCC - 0.3V < AVCC < VCC + 0.3V, however, AVCC should always be within 2.7 - 5.5V.
When programming the EEPROM, an auto-erase cycle is built into the self-timed pro-
gramming operation (in the serial mode ONLY) and there is no need to first execute the
Chip Erase instruction. The Chip Erase operation turns the content of every memory
location in both the Program and EEPROM arrays into $FF.
Depending on CKSEL Fuses, a valid clock must be present. The minimum low and high
periods for the serial clock (SCK) input are defined as follows:
Low:> 2 CPU clock cycles for fck < 12 MHz, 3 CPU clock cycles for fck ≥ 12 MHz
High:> 2 CPU clock cycles for fck < 12 MHz, 3 CPU clock cycles for fck ≥ 12 MHz
SPI Serial Programming When writing serial data to the ATmega128, data is clocked on the rising edge of SCK.
Algorithm
When reading data from the ATmega128, data is clocked on the falling edge of SCK.
See Figure 144, Figure 145 and Table 145 for timing details.
To program and verify the ATmega128 in the SPI Serial Programming mode, the follow-
ing sequence is recommended (See four byte instruction formats in Table 144):
1. Power-up sequence:
Apply power between VCC and GND while RESET and SCK are set to “0”. In
some systems, the programmer can not guarantee that SCK is held low during
power-up. In this case, RESET must be given a positive pulse of at least two
CPU clock cycles duration after SCK has been set to “0”.
As an alternative to using the RESET signal, PEN can be held low during Power-
on Reset while SCK is set to “0”. In this case, only the PEN value at Power-on
Reset is important. If the programmer cannot guarantee that SCK is held low
during power-up, the PEN method cannot be used. The device must be powered
down in order to commence normal operation when using this method.
2. Wait for at least 20 ms and enable SPI Serial Programming by sending the Pro-
gramming Enable serial instruction to pin MOSI.
298 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
3. The SPI Serial Programming instructions will not work if the communication is
out of synchronization. When in sync. the second byte ($53), will echo back
when issuing the third byte of the Programming Enable instruction. Whether the
echo is correct or not, all FOUR bytes of the instruction must be transmitted. If
the $53 did not echo back, give RESET a positive pulse and issue a new Pro-
gramming Enable command.
4. The Flash is programmed one page at a time. The memory page is loaded one
byte at a time by supplying the 7 LSB of the address and data together with the
Load Program Memory Page instruction. To ensure correct loading of the page,
the data low byte must be loaded before data high byte is applied for given
address. The Program Memory Page is stored by loading the Write Program
Memory Page instruction with the 9 MSB of the address. If polling is not used,
the user must wait at least tWD_FLASH before issuing the next page. (See Table
128). Accessing the SPI Serial Programming interface before the Flash write
operation completes can result in incorrect programming.
5. The EEPROM array is programmed one byte at a time by supplying the address
and data together with the appropriate Write instruction. An EEPROM memory
location is first automatically erased before new data is written. If polling is not
used, the user must wait at least tWD_EEPROM before issuing the next byte. (See
Table 128). In a chip erased device, no $FFs in the data file(s) need to be
programmed.
6. Any memory location can be verified by using the Read instruction which returns
the content at the selected address at serial output MISO.
7. At the end of the programming session, RESET can be set high to commence
normal operation.
8. Power-off sequence (if needed):
Set RESET to “1”.
Turn VCC power off.
Data Polling Flash When a page is being programmed into the Flash, reading an address location within
the page being programmed will give the value $FF. At the time the device is ready for a
new page, the programmed value will read correctly. This is used to determine when the
next page can be written. Note that the entire page is written simultaneously and any
address within the page can be used for polling. Data polling of the Flash will not work
for the value $FF, so when programming this value, the user will have to wait for at least
tWD_FLASH before programming the next page. As a chip-erased device contains $FF in
all locations, programming of addresses that are meant to contain $FF, can be skipped.
See Table 128 for tWD_FLASH value
Data Polling EEPROM When a new byte has been written and is being programmed into EEPROM, reading the
address location being programmed will give the value $FF. At the time the device is
ready for a new byte, the programmed value will read correctly. This is used to deter-
mine when the next byte can be written. This will not work for the value $FF, but the user
should have the following in mind: As a chip-erased device contains $FF in all locations,
programming of addresses that are meant to contain $FF, can be skipped. This does
not apply if the EEPROM is re-programmed without chip-erasing the device. In this
case, data polling cannot be used for the value $FF, and the user will have to wait at
least t WD_EEPROM before programming the next byte. See Table 128 for tWD_EEPROM
value.
299
2467E–AVR–05/02
Table 128. Minimum Wait Delay before Writing the Next Flash or EEPROM Location
Symbol Minimum Wait Delay
tWD_FLASH 4.5 ms
tWD_EEPROM 9.0 ms
tWD_ERASE 9.0 ms
Figure 144. .SPI Serial Programming Waveforms
SERIAL DATA INPUT MSB LSB

(MOSI)
SERIAL DATA OUTPUT MSB LSB

(MISO)
SERIAL CLOCK INPUT

(SCK)
SAMPLE
300 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 129. SPI Serial Programming Instruction Set

Instruction Format
Instruction Byte 1 Byte 2 Byte 3 Byte4 Operation
Programming Enable 1010 1100 0101 0011 xxxx xxxx xxxx xxxx Enable SPI Serial Programming after RESET
goes low.
Chip Erase 1010 1100 100x xxxx xxxx xxxx xxxx xxxx Chip Erase EEPROM and Flash.
Read Program 0010 H000 aaaa aaaa bbbb bbbb oooo oooo Read H (high or low) data o from Program
Memory memory at word address a:b.
Load Program 0100 H000 xxxx xxxx xbbb bbbb iiii iiii Write H (high or low) data i to Program
Memory Page Memory page at word address b. Data low
byte must be loaded before data high byte is
applied within the same address.
Write Program 0100 1100 aaaa aaaa bxxx xxxx xxxx xxxx
Memory Page Write Program Memory Page at address a:b.
Read EEPROM 1010 0000 xxxx aaaa bbbb bbbb oooo oooo Read data o from EEPROM memory at
Memory address a:b.
Write EEPROM 1100 0000 xxxx aaaa bbbb bbbb iiii iiii Write data i to EEPROM memory at address
Memory a:b.
Read Lock bits 0101 1000 0000 0000 xxxx xxxx xxoo oooo Read Lock bits. “0” = programmed, “1” =
unprogrammed. See Table 115 on page
282 for details.
Write Lock bits 1010 1100 111x xxxx xxxx xxxx 11ii iiii Write Lock bits. Set bits = “0” to program Lock
bits. See Table 115 on page 282 for details.
Read Signature Byte 0011 0000 xxxx xxxx xxxx xxbb oooo oooo Read Signature Byte o at address b.
Write Fuse bits 1010 1100 1010 0000 xxxx xxxx iiii iiii Set bits = “0” to program, “1” to unprogram.
See Table 119 on page 284 for details.
Write Fuse High Bits 1010 1100 1010 1000 xxxx xxxx iiii iiii Set bits = “0” to program, “1” to unprogram.
See Table 118 on page 284 for details.
Write Extended Fuse 1010 1100 1010 0100 xxxx xxxx xxxx xxii Set bits = “0” to program, “1” to unprogram.
bits See Table 119 on page 284 for details.
Read Fuse bits 0101 0000 0000 0000 xxxx xxxx oooo oooo Read Fuse bits. “0” = programmed, “1” =
284 for details.
Read Extendend 0101 0000 0000 1000 xxxx xxxx oooo oooo Read Extended Fuse bits. “0” = pro-grammed,
Fuse bits “1” = unprogrammed. See Table 119 on
page 284 for details.
Read Fuse High Bits 0101 1000 0000 1000 xxxx xxxx oooo oooo Read Fuse high bits. “0” = pro-grammed, “1” =
284 for details.
Read Calibration Byte 0011 1000 xxxx xxxx 0000 00bb oooo oooo Read Calibration Byte o at address b.
Note: a = address high bits
b = address low bits
H = 0 - Low byte, 1 - High Byte
o = data out
i = data in
x = don’t care
301
2467E–AVR–05/02
SPI Serial Programming Figure 145. SPI Serial Programming Timing
Characteristics
MOSI
tOVSH tSHOX tSLSH
SCK
tSHSL
MISO
tSLIV
Table 130. SPI Serial Programming Characteristics, TA = -40°C to 85°C, VCC = 2.7V -
5.5V (Unless Otherwise Noted)
1/tCLCL Oscillator Frequency (VCC = 2.7 - 5.5 V) 0 8 MHz
tCLCL Oscillator Period (VCC = 2.7 - 5.5 V) 125 ns
1/tCLCL Oscillator Frequency (VCC = 4.5 - 5.5 V) 0 16 MHz
tCLCL Oscillator Period (VCC = 4.5 - 5.5 V) 62.5 ns
(1)
tSHSL SCK Pulse Width High 2 tCLCL ns
(1)
tSLSH SCK Pulse Width Low 2 tCLCL ns
tOVSH MOSI Setup to SCK High tCLCL ns
tSHOX MOSI Hold after SCK High 2 tCLCL ns
tSLIV SCK Low to MISO Valid 15 ns
Note: 1. 2 tCLCL for fck < 12 MHz, 3 tCLCL for fck >= 12 MHz
Programming Via the Programming through the JTAG interface requires control of the four JTAG specific
JTAG Interface pins: TCK, TMS, TDI, and TDO. Control of the Reset and clock pins is not required.
To be able to use the JTAG interface, the JTAGEN fuse must be programmed. The
device is default shipped with the Fuse programmed. In addition, the JTD bit in
MCUCSR must be cleared. Alternatively, if the JTD bit is set, the external reset can be
forced low. Then, the JTD bit will be cleared after two chip clocks, and the JTAG pins
are available for programming. This provides a means of using the JTAG pins as normal
port pins in running mode while still allowing In-System Programming via the JTAG
interface. Note that this technique can not be used when using the JTAG pins for
Boundary-scan or On-chip Debug. In these cases the JTAG pins must be dedicated for
this purpose.
As a definition in this data sheet, the LSB is shifted in and out first of all Shift Registers.
Programming Specific JTAG The instruction register is 4-bit wide, supporting up to 16 instructions. The JTAG instruc-
Instructions tions useful for Programming are listed below.
The OPCODE for each instruction is shown behind the instruction name in hex format.
The text describes which data register is selected as path between TDI and TDO for
each instruction.
The Run-Test/Idle state of the TAP controller is used to generate internal clocks. It can
also be used as an idle state between JTAG sequences. The state machine sequence
for changing the instruction word is shown in Figure 146.
302 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 146. State Machine Sequence for Changing the Instruction Word
1 Test-Logic-Reset
0 1 1 1
0 0
1 1
0 0
1 1
1 1
Exit1-DR Exit1-IR
0 0
1 1
0 0
Exit2-DR Exit2-IR
1 1
Update-DR Update-IR
1 0 1 0
AVR_RESET ($C) The AVR specific public JTAG instruction for setting the AVR device in the Reset mode
or taking the device out from the Reset mode. The TAP controller is not reset by this
instruction. The one bit Reset Register is selected as Data Register. Note that the reset
will be active as long as there is a logic 'one' in the Reset Chain. The output from this
chain is not latched.
• Shift-DR: The Reset Register is shifted by the TCK input.
PROG_ENABLE ($4) The AVR specific public JTAG instruction for enabling programming via the JTAG port.
The 16-bit Programming Enable Register is selected as data register. The active states
are the following:
• Shift-DR: the programming enable signature is shifted into the data register.
• Update-DR: the programming enable signature is compared to the correct value,
and Programming mode is entered if the signature is valid.
303
2467E–AVR–05/02
PROG_COMMANDS ($5) The AVR specific public JTAG instruction for entering programming commands via the
JTAG port. The 15-bit Programming Command Register is selected as data register.
The active states are the following:
• Capture-DR: the result of the previous command is loaded into the data register.
• Shift-DR: the data register is shifted by the TCK input, shifting out the result of the
previous command and shifting in the new command.
• Update-DR: the programming command is applied to the Flash inputs
• Run-Test/Idle: one clock cycle is generated, executing the applied command (not
always required, see Table 131 below).
PROG_PAGELOAD ($6) The AVR specific public JTAG instruction to directly load the Flash data page via the
JTAG port. The 2048-bit Virtual Flash Page Load Register is selected as data register.
This is a virtual scan chain with length equal to the number of bits in one Flash page.
Internally the Shift Register is 8-bit. Unlike most JTAG instructions, the Update-DR state
is not used to transfer data from the Shift Register. The data are automatically trans-
ferred to the Flash page buffer byte by byte in the Shift-DR state by an internal state
machine. This is the only active state:
• Shift-DR: Flash page data are shifted in from TDI by the TCK input, and
automatically loaded into the Flash page one byte at a time.
PROG_PAGEREAD ($7) The AVR specific public JTAG instruction to read one full Flash data page via the JTAG
port. The 2056-bit Virtual Flash Page Read Register is selected as data register. This is
a virtual scan chain with length equal to the number of bits in one Flash page plus 8.
Internally the Shift Register is 8-bit. Unlike most JTAG instructions, the Capture-DR
state is not used to transfer data to the Shift Register. The data are automatically trans-
ferred from the Flash page buffer byte by byte in the Shift-DR state by an internal state
machine. This is the only active state:
• Shift-DR: Flash data are automatically read one byte at a time and shifted out on
TDO by the TCK input. The TDI input is ignored.
Note: The JTAG instructions PROG_PAGELOAD and PROG_PAGEREAD can only be used if
the AVR devce is the first decive in JTAG scan chain. If the AVR cannot be the first
device in the scan chain, the byte-wise programming algorithm must be used.
Data Registers The data registers are selected by the JTAG instruction registers described in section
“Programming Specific JTAG Instructions” on page 302. The data registers relevant for
programming operations are:
• Reset Register
• Programming Enable Register
• Programming Command Register
• Virtual Flash Page Load Register
• Virtual Flash Page Read Register
Reset Register The Reset Register is a Test Data Register used to reset the part during programming. It
is required to reset the part before entering programming mode.
A high value in the Reset Register corresponds to pulling the external Reset low. The
part is reset as long as there is a high value present in the Reset Register. Depending
on the Fuse settings for the clock options, the part will remain reset for a Reset Time-
Out Period (refer to “Clock Sources” on page 34) after releasing the Reset Register. The
output from this Data Register is not latched, so the reset will take place immediately, as
304 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Programming Enable Register The Programming Enable Register is a 16-bit register. The contents of this register is
compared to the programming enable signature, binary code 1010_0011_0111_0000.
When the contents of the register is equal to the programming enable signature, pro-
gramming via the JTAG port is enabled. The Register is reset to 0 on Power-on Reset,
and should always be reset when leaving Programming mode.
Figure 147. Programming Enable Register

TDI
D $A370
A
T
= D Q Programming enable
ClockDR & PROG_ENABLE
TDO
305
2467E–AVR–05/02
Programming Command The Programming Command Register is a 15-bit register. This register is used to seri-
Register ally shift in programming commands, and to serially shift out the result of the previous
command, if any. The JTAG Programming Instruction Set is shown in Table 131. The
state sequence when shifting in the programming commands is illustrated in Figure 149.
Figure 148. Programming Command Register

TDI
S
T
R
O
B
E
S
Flash
EEPROM
A
Fuses
D
D
Lock Bits
R
E
S
S
/
D
A
T
A
TDO
306 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 131. JTAG Programming Instruction

Set a = address high bits, b = address low bits, H = 0 - Low byte, 1 - High Byte, o = data out, i = data in, x = don’t care
Instruction TDI sequence TDO sequence Notes
1a. Chip erase 0100011_10000000 xxxxxxx_xxxxxxxx
0110001_10000000 xxxxxxx_xxxxxxxx
1b. Poll for chip erase complete 0110011_10000000 xxxxxox_xxxxxxxx (2)
2a. Enter Flash Write 0100011_00010000 xxxxxxx_xxxxxxxx
2b. Load Address High Byte 0000111_aaaaaaaa xxxxxxx_xxxxxxxx (9)
2c. Load Address Low Byte 0000011_bbbbbbbb xxxxxxx_xxxxxxxx
2d. Load Data Low Byte 0010011_iiiiiiii xxxxxxx_xxxxxxxx
2e. Load Data High Byte 0010111_iiiiiiii xxxxxxx_xxxxxxxx
2f. Latch Data 0110111_00000000 xxxxxxx_xxxxxxxx (1)
2g. Write Flash Page 0110111_00000000 xxxxxxx_xxxxxxxx (1)
2h. Poll for Page Write complete 0110111_00000000 xxxxxox_xxxxxxxx (2)
3a. Enter Flash Read 0100011_00000010 xxxxxxx_xxxxxxxx
3d. Read Data Low and High Byte 0110010_00000000 xxxxxxx_xxxxxxxx
0110110_00000000 xxxxxxx_oooooooo low byte
0110111_00000000 xxxxxxx_oooooooo high byte
4a. Enter EEPROM Write 0100011_00010001 xxxxxxx_xxxxxxxx
4d. Load Data Byte 0010011_iiiiiiii xxxxxxx_xxxxxxxx
4e. Latch Data 0110111_00000000 xxxxxxx_xxxxxxxx (1)
4f. Write EEPROM Page 0110011_00000000 xxxxxxx_xxxxxxxx (1)
4g. Poll for Page Write complete 0110011_00000000 xxxxxox_xxxxxxxx (2)
5a. Enter EEPROM Read 0100011_00000011 xxxxxxx_xxxxxxxx
307
2467E–AVR–05/02
Table 131. JTAG Programming Instruction (Continued)
Set (Continued) a = address high bits, b = address low bits, H = 0 - Low byte, 1 - High Byte, o = data out, i = data in, x = don’t care
5d. Read Data Byte 0110011_bbbbbbbb xxxxxxx_xxxxxxxx
0110011_00000000 xxxxxxx_oooooooo
6a. Enter Fuse Write 0100011_01000000 xxxxxxx_xxxxxxxx
(6)
6b. Load Data Low Byte 0010011_iiiiiiii xxxxxxx_xxxxxxxx (3)
6c. Write Fuse Extended byte 0111011_00000000 xxxxxxx_xxxxxxxx (1)
6d. Poll for Fuse Write complete 0110111_00000000 xxxxxox_xxxxxxxx (2)
6e. Load Data Low Byte(7) 0010011_iiiiiiii xxxxxxx_xxxxxxxx (3)
6f. Write Fuse High byte 0110111_00000000 xxxxxxx_xxxxxxxx (1)
6g. Poll for Fuse Write complete 0110111_00000000 xxxxxox_xxxxxxxx (2)
(7)
6h. Load Data Low Byte 0010011_iiiiiiii xxxxxxx_xxxxxxxx (3)
6i. Write Fuse Low byte 0110011_00000000 xxxxxxx_xxxxxxxx (1)
6j. Poll for Fuse Write complete 0110011_00000000 xxxxxox_xxxxxxxx (2)
7a. Enter Lock bit Write 0100011_00100000 xxxxxxx_xxxxxxxx
(9)
7b. Load Data Byte 0010011_11iiiiii xxxxxxx_xxxxxxxx (4)
7c. Write Lock bits 0110011_00000000 xxxxxxx_xxxxxxxx (1)
7d. Poll for Lock bit Write complete 0110011_00000000 xxxxxox_xxxxxxxx (2)
8a. Enter Fuse/Lock bit Read 0100011_00000100 xxxxxxx_xxxxxxxx
8b. Read Extended Fuse Byte(6) 0111010_00000000 xxxxxxx_xxxxxxxx
8c. Read Fuse High Byte(7) 0111110_00000000 xxxxxxx_xxxxxxxx
8d. Read Fuse Low Byte(8) 0110010_00000000 xxxxxxx_xxxxxxxx
8e. Read Lock bits(9) 0110110_00000000 xxxxxxx_xxxxxxxx (5)
0110111_00000000 xxxxxxx_xxoooooo
308 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 131. JTAG Programming Instruction (Continued)

Set (Continued) a = address high bits, b = address low bits, H = 0 - Low byte, 1 - High Byte, o = data out, i = data in, x = don’t care
8f. Read Fuses and Lock bits 0111010_00000000 xxxxxxx_xxxxxxxx (5)
0111110_00000000 xxxxxxx_oooooooo fuse ext. byte
0110010_00000000 xxxxxxx_oooooooo fuse high byte
0110110_00000000 xxxxxxx_oooooooo fuse low byte
0110111_00000000 xxxxxxx_oooooooo lock bits
9a. Enter Signature Byte Read 0100011_00001000 xxxxxxx_xxxxxxxx
9b. Load Address Byte 0000011_bbbbbbbb xxxxxxx_xxxxxxxx
9c. Read Signature Byte 0110010_00000000 xxxxxxx_xxxxxxxx
10a. Enter Calibration Byte Read 0100011_00001000 xxxxxxx_xxxxxxxx
10b. Load Address Byte 0000011_bbbbbbbb xxxxxxx_xxxxxxxx
10c. Read Calibration Byte 0110110_00000000 xxxxxxx_xxxxxxxx
11a. Load No Operation Command 0100011_00000000 xxxxxxx_xxxxxxxx
Notes: 1. This command sequence is not required if the seven MSB are correctly set by the previous command sequence (which is
normally the case).
2. Repeat until o = “1”.
3. Set bits to “0” to program the corresponding fuse, “1” to unprogram the Fuse.
4. Set bits to “0” to program the corresponding lock bit, “1” to leave the Lock bit unchanged.
5. “0” = programmed, “1” = unprogrammed.
6. The bit mapping for Fuses Extended byte is listed in Table 117 on page 283
7. The bit mapping for Fuses High byte is listed in Table 118 on page 284
8. The bit mapping for Fuses Low byte is listed in Table 119 on page 284
9. The bit mapping for Lock bits byte is listed in Table 115 on page 282
10. Address bits exceeding PCMSB and EEAMSB (Table 123 and Table 124) are don’t care
309
2467E–AVR–05/02
Figure 149. State Machine Sequence for Changing/Reading the Data Word
1 Test-Logic-Reset
0 1 1 1
0 0
1 1
0 0
1 1
1 1
Exit1-DR Exit1-IR
0 0
1 1
0 0
Exit2-DR Exit2-IR
1 1
Update-DR Update-IR
1 0 1 0
Virtual Flash Page Load The Virtual Flash Page Load Register is a virtual scan chain with length equal to the
Register number of bits in one Flash page. Internally the Shift Register is 8-bit, and the data are
automatically transferred to the Flash page buffer byte by byte. Shift in all instruction
words in the page, starting with the LSB of the first instruction in the page and ending
with the MSB of the last instruction in the page. This provides an efficient way to load the
entire Flash page buffer before executing Page Write.
310 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 150. Virtual Flash Page Load Register
STROBES
State
machine
TDI
ADDRESS
Flash
EEPROM
Fuses
Lock Bits
D
A
T
A
TDO
Virtual Flash Page Read The Virtual Flash Page Read Register is a virtual scan chain with length equal to the
Register number of bits in one Flash page plus 8. Internally the Shift Register is 8-bit, and the
data are automatically transferred from the Flash data page byte by byte. The first eight
cycles are used to transfer the first byte to the internal Shift Register, and the bits that
are shifted out during these 8 cycles should be ignored. Following this initialization, data
are shifted out starting with the LSB of the first instruction in the page and ending with
the MSB of the last instruction in the page. This provides an efficient way to read one full
Flash page to verify programming.
Figure 151. Virtual Flash Page Read Register
STROBES
State
machine
TDI
ADDRESS
Flash
EEPROM
Fuses
Lock Bits
D
A
T
A
TDO
Programming Algorithm All references below of type “1a”, “1b”, and so on, refer to Table 131.
311
2467E–AVR–05/02
Entering Programming Mode 1. Enter JTAG instruction AVR_RESET and shift 1 in the Reset Register.
2. Enter instruction PROG_ENABLE and shift 1010_0011_0111_0000 in the Pro-
gramming Enable Register.
Leaving Programming Mode 1. Enter JTAG instruction PROG_COMMANDS.

2. Disable all programming instructions by usning no operation instruction 11a.
3. Enter instruction PROG_ENABLE and shift 0000_0000_0000_0000 in the pro-
gramming Enable Register.
4. Enter JTAG instruction AVR_RESET and shift 0 in the Reset Register.
Performing Chip Erase 1. Enter JTAG instruction PROG_COMMANDS.

2. Start chip erase using programming instruction 1a.
3. Poll for chip erase complete using programming instruction 1b, or wait for
tWLRH_CE (refer to Table Note: on page 296).
Programming the Flash 1. Enter JTAG instruction PROG_COMMANDS.

2. Enable Flash write using programming instruction 2a.
3. Load address high byte using programming instruction 2b.
4. Load address low byte using programming instruction 2c.
5. Load data using programming instructions 2d, 2e and 2f.
6. Repeat steps 4 and 5 for all instruction words in the page.
7. Write the page using programming instruction 2g.
8. Poll for Flash write complete using programming instruction 2h, or wait for tWLRH
(refer to Table Note: on page 296).
9. Repeat steps 3 to 7 until all data have been programmed.
A more efficient data transfer can be achieved using the PROG_PAGELOAD
instruction:
1. Enter JTAG instruction PROG_COMMANDS.
2. Enable Flash write using programming instruction 2a.
3. Load the page address using programming instructions 2b and 2c. PCWORD
(refer to Table 123 on page 287) is used to address within one page and must be
written as 0.
4. Enter JTAG instruction PROG_PAGELOAD.
5. Load the entire page by shifting in all instruction words in the page, starting with
the LSB of the first instruction in the page and ending with the MSB of the last
instruction in the page.
7. Write the page using programming instruction 2g.
8. Poll for Flash write complete using programming instruction 2h, or wait for tWLRH
Reading the Flash 1. Enter JTAG instruction PROG_COMMANDS.

2. Enable Flash read using programming instruction 3a.
3. Load address using programming instructions 3b and 3c.
4. Read data using programming instruction 3d.
312 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
5. Repeat steps 3 and 4 until all data have been read.

A more efficient data transfer can be achieved using the PROG_PAGEREAD
instruction:
2. Enable Flash read using programming instruction 3a.
3. Load the page address using programming instructions 3b and 3c. PCWORD
(refer to Table 123 on page 287) is used to address within one page and must be
written as 0.
4. Enter JTAG instruction PROG_PAGEREAD.
5. Read the entire page by shifting out all instruction words in the page, starting
with the LSB of the first instruction in the page and ending with the MSB of the
last instruction in the page. Remember that the first 8 bits shifted out should be
ignored.
7. Repeat steps 3 to 6 until all data have been read.
Programming the EEPROM 1. Enter JTAG instruction PROG_COMMANDS.

2. Enable EEPROM write using programming instruction 4a.
3. Load address high byte using programming instruction 4b.
4. Load address low byte using programming instruction 4c.
5. Load data using programming instructions 4d and 4e.
6. Repeat steps 4 and 5 for all data bytes in the page.
7. Write the data using programming instruction 4f.
8. Poll for EEPROM write complete using programming instruction 4g, or wait for
tWLRH (refer to Table Note: on page 296).
Note that the PROG_PAGELOAD instruction can not be used when programming the
EEPROM
Reading the EEPROM 1. Enter JTAG instruction PROG_COMMANDS.

2. Enable EEPROM read using programming instruction 5a.
3. Load address using programming instructions 5b and 5c.
4. Read data using programming instruction 5d.
5. Repeat steps 3 and 4 until all data have been read.
Note that the PROG_PAGEREAD instruction can not be used when reading the
EEPROM
Programming the Fuses 1. Enter JTAG instruction PROG_COMMANDS.

2. Enable Fuse write using programming instruction 6a.
3. Load data byte using programming instructions 6b. A bit value of “0” will program
the corresponding fuse, a “1” will unprogram the fuse.
4. Write Extended Fuse byte using programming instruction 6c.
5. Poll for Fuse write complete using programming instruction 6d, or wait for tWLRH
6. Load data byte using programming instructions 6e. A bit value of “0” will program
the corresponding fuse, a “1” will unprogram the fuse.
313
2467E–AVR–05/02
7. Write Fuse high byte using programming instruction 6f.
8. Poll for Fuse write complete using programming instruction 6g, or wait for tWLRH
9. Load data byte using programming instructions 6h. A “0” will program the fuse, a
“1” will unprogram the fuse.
10. Write Fuse low byte using programming instruction 6i.
11. Poll for Fuse write complete using programming instruction 6j, or wait for tWLRH
Programming the Lock Bits 1. Enter JTAG instruction PROG_COMMANDS.

2. Enable Lock bit write using programming instruction 7a.
3. Load data using programming instructions 7b. A bit value of “0” will program the
corresponding lock bit, a “1” will leave the lock bit unchanged.
4. Write Lock bits using programming instruction 7c.
5. Poll for Lock bit write complete using programming instruction 7d, or wait for
tWLRH (refer to Table Note: on page 296).
Reading the Fuses and Lock 1. Enter JTAG instruction PROG_COMMANDS.

Bits 2. Enable Fuse/Lock bit read using programming instruction 8a.
3. To read all Fuses and Lock bits, use programming instruction 8f.
To only read Extended Fuse byte, use programming instruction 8b.
To only read Fuse high byte, use programming instruction 8c.
To only read Fuse low byte, use programming instruction 8d.
To only read Lock bits, use programming instruction 8e.
Reading the Signature Bytes 1. Enter JTAG instruction PROG_COMMANDS.

2. Enable Signature byte read using programming instruction 9a.
3. Load address $00 using programming instruction 9b.
4. Read first signature byte using programming instruction 9c.
5. Repeat steps 3 and 4 with address $01 and address $02 to read the second and
third signature bytes, respectively.
Reading the Calibration Byte 1. Enter JTAG instruction PROG_COMMANDS.

2. Enable Calibration byte read using programming instruction 10a.
3. Load address $00 using programming instruction 10b.
4. Read the calibration byte using programming instruction 10c.
314 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Electrical Characteristics
Note: Typical values contained in this data sheet are based on simulations and characterization of other AVR microcontrollers manu-
factured on the same process technology. Min and Max values will be available after the device is characterized.
Absolute Maximum Ratings*

Operating Temperature.................................. -55°C to +125°C *NOTICE: Stresses beyond those listed under “Absolute
Maximum Ratings” may cause permanent dam-
Storage Temperature ..................................... -65°C to +150°C age to the device. This is a stress rating only and
functional operation of the device at these or
Voltage on any Pin except RESET other conditions beyond those indicated in the
with respect to Ground ................................-1.0V to VCC +0.5V operational sections of this specification is not
implied. Exposure to absolute maximum rating
Voltage on RESET with respect to Ground......-1.0V to +13.0V conditions for extended periods may affect
device reliability.
Maximum Operating Voltage ............................................ 6.0V
DC Current per I/O Pin ............................................... 40.0 mA
DC Current VCC and GND Pins................................ 200.0 mA
DC Characteristics
TA = -40°C to 85°C, VCC = 2.7V to 5.5V (unless otherwise noted)
Except XTAL1 and
VIL Input Low Voltage -0.5 0.3 V CC(1) V
RESET pins
XTAL1 pin, External
VIL1 Input Low Voltage -0.5 0.1 V CC(1) V
Clock Selected
VIL2 Input Low Voltage RESET pin -0.5 0.2 V CC(1) V
Except XTAL1 and
VIH Input High Voltage 0.6 VCC(2) VCC + 0.5 V
RESET pins
XTAL1 pin, External
VIH1 Input High Voltage 0.7 VCC(2) VCC + 0.5 V
Clock Selected
VIH2 Input High Voltage RESET pin 0.85 V CC(2) VCC + 0.5 V
(3)
Output Low Voltage IOL = 20 mA, V CC = 5V 0.7 V
VOL
(Ports A,B,C,D, E, F, G) IOL = 10 mA, V CC = 3V 0.5 V
Output High Voltage(4) IOH = -20 mA, VCC = 5V 4.0 V
VOH
(Ports A,B,C,D) IOH = -10 mA, VCC = 3V 2.2 V
Input Leakage Vcc = 5.5V, pin low
IIL 8.0 µA
Current I/O Pin (absolute value)
Input Leakage Vcc = 5.5V, pin high
IIH 8.0 µA
Current I/O Pin (absolute value)
RRST Reset Pull-up Resistor 30 100 kΩ
RPEN PEN Pull-up Resistor 25 100 kΩ
RPU I/O Pin Pull-up Resistor 33 122 kΩ
315
2467E–AVR–05/02
TA = -40°C to 85°C, VCC = 2.7V to 5.5V (unless otherwise noted) (Continued)
Active 4 MHz, VCC = 3V
5 mA
(ATmega128L)
Active 8 MHz, VCC = 5V
20 mA
(ATmega128)
Power Supply Current
Idle 4 MHz, VCC = 3V
ICC 2 mA
(ATmega128L)
Idle 8 MHz, VCC = 5V
12 mA
(ATmega128)
WDT enabled, VCC = 3V < 25 40 µA
Power-down mode(5)
WDT disabled, VCC = 3V < 10 25 µA
Analog Comparator VCC = 5V
VACIO 40 mV
Input Offset Voltage Vin = VCC/2
Analog Comparator VCC = 5V
IACLK -50 50 nA
Input Leakage Current Vin = VCC/2
Analog Comparator VCC = 2.7V 750
tACID ns
Initialization Delay VCC = 5.0V 500
Analog Comparator VCC = 2.7V 750
tACID ns
Propagation Delay VCC = 5.0V 500
Notes: 1. “Max” means the highest value where the pin is guaranteed to be read as low
2. “Min” means the lowest value where the pin is guaranteed to be read as high
3. Although each I/O port can sink more than the test conditions (20 mA at VCC = 5V, 10 mA at VCC = 3V) under steady state
conditions (non-transient), the following must be observed:
TQFP Package:
1] The sum of all IOL, for all ports, should not exceed 400 mA.
2] The sum of all IOL, for ports A0 - A7, G2, C3 - C7 should not exceed 300 mA.
3] The sum of all IOL, for ports C0 - C2, G0 - G1, D0 - D7, XTAL2 should not exceed 150 mA.
4] The sum of all IOL, for ports B0 - B7, G3 - G4, E0 - E7 should not exceed 150 mA.
5] The sum of all IOL, for ports F0 - F7, should not exceed 200 mA.
If IOL exceeds the test condition, VOL may exceed the related specification. Pins are not guaranteed to sink current greater
than the listed test condition.
4. Although each I/O port can source more than the test conditions (20 mA at Vcc = 5V, 10 mA at Vcc = 3V) under steady state
conditions (non-transient), the following must be observed:
TQFP Package:
1] The sum of all IOH, for all ports, should not exceed 400 mA.
2] The sum of all IOH, for ports A0 - A7, G2, C3 - C7 should not exceed 300 mA.
3] The sum of all IOH, for ports C0 - C2, G0 - G1, D0 - D7, XTAL2 should not exceed 150 mA.
4] The sum of all IOH, for ports B0 - B7, G3 - G4, E0 - E7 should not exceed 150 mA.
5] The sum of all IOH, for ports F0 - F7, should not exceed 200 mA.
If IOH exceeds the test condition, VOH may exceed the related specification. Pins are not guaranteed to source current
greater than the listed test condition.
5. Minimum VCC for Power-down is 2.5V.
316 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
External Clock Drive Figure 152. External Clock Drive Waveforms

Waveforms
V IH1
V IL1
External Clock Drive

Table 132. External Clock Drive
VCC = 2.7V to 5.5V VCC = 4.5V to 5.5V
Symbol Parameter Min Max Min Max Units
1/tCLCL Oscillator Frequency 0 8 0 16 MHz
tCLCL Clock Period 125 62.5 ns
tCHCX High Time 50 25 ns
tCLCX Low Time 50 25 ns
tCLCH Rise Time 1.6 0.5 µs
tCHCL Fall Time 1.6 0.5 µs
Table 133. External RC Oscillator, Typical Frequencies

R [kΩ] C [pF] f
100 70 TBD
31.5 20 TBD
6.5 20 TBD
Note: R should be in the range 3 kΩ - 100 kΩ, and C should be at least 20 pF. The C values
given in the table includes pin capacitance. This will vary with package type.
317
2467E–AVR–05/02
Two-wire Serial Interface Characteristics
Table 134 describes the requirements for devices connected to the Two-wire Serial Bus. The ATmega128 Two-wire Serial
Interface meets or exceeds these requirements under the noted conditions.
Timing symbols refer to Figure 153.
Table 134. Two-wire Serial Bus Requirements

Symbol Parameter Condition Min Max Units
VIL Input Low-voltage -0.5 0.3 VCC V
VIH Input High-voltage 0.7 VCC VCC + 0.5 V
(1) (2)
Vhys Hysteresis of Schmitt Trigger Inputs 0.05 VCC – V
(1)
VOL Output Low-voltage 3 mA sink current 0 0.4 V
(1) (3)(2)
tr Rise Time for both SDA and SCL 20 + 0.1Cb 300 ns
tof(1) Output Fall Time from VIHmin to VILmax 10 pF < Cb < 400 pF(3) 20 + 0.1Cb(3)(2) 250 ns
tSP(1) Spikes Suppressed by Input Filter 0 50 (2)
ns
Ii Input Current each I/O Pin 0.1 VCC < Vi < 0.9 VCC -10 10 µA
(1)
Ci Capacitance for each I/O Pin – 10 pF
fSCL SCL Clock Frequency fCK(4) > max(16fSCL, 250kHz)(5) 0 400 kHz
fSCL ≤ 100 kHz V CC – 0,4V 1000ns
---------------------------- ------------------- Ω
3mA Cb
Rp Value of Pull-up resistor
fSCL > 100 kHz V CC – 0,4V 300ns
---------------------------- ---------------- Ω
3mA Cb
fSCL ≤ 100 kHz 4.0 – µs
tHD;STA Hold Time (repeated) START Condition
fSCL > 100 kHz 0.6 – µs
fSCL ≤ 100 kHz (6)
4.7 – µs
tLOW Low Period of the SCL Clock
fSCL > 100 kHz(7) 1.3 – µs
fSCL ≤ 100 kHz 4.0 – µs
tHIGH High period of the SCL clock
fSCL > 100 kHz 0.6 – µs
fSCL ≤ 100 kHz 4.7 – µs
tSU;STA Set-up time for a repeated START condition
fSCL > 100 kHz 0.6 – µs
fSCL ≤ 100 kHz 0 3.45 µs
tHD;DAT Data hold time
fSCL > 100 kHz 0 0.9 µs
fSCL ≤ 100 kHz 250 – ns
tSU;DAT Data setup time
fSCL > 100 kHz 100 – ns
fSCL ≤ 100 kHz 4.0 – µs
tSU;STO Setup time for STOP condition
fSCL > 100 kHz 0.6 – µs
Bus free time between a STOP and START
tBUF fSCL ≤ 100 kHz 4.7 – µs
condition
Notes: 1. In ATmega128, this parameter is characterized and not 100% tested.
2. Required only for fSCL > 100 kHz.
3. Cb = capacitance of one bus line in pF.
4. fCK = CPU clock frequency
318 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
5. This requirement applies to all ATmega128 Two-wire Serial Interface operation.

Other devices connected to the Two-wire Serial Bus need only obey the general fSCL
requirement.
6. The actual low period generated by the ATmega128 Two-wire Serial Interface is
(1/fSCL - 2/fCK), thus fCK must be greater than 6 MHz for the low time requirement to be
strictly met at fSCL = 100 kHz.
7. The actual low period generated by the ATmega128 Two-wire Serial Interface is
(1/fSCL - 2/fCK), thus the low time requirement will not be strictly met for fSCL > 308 kHz
when fCK = 8 MHz. Still, ATmega128 devices connected to the bus may communicate
at full speed (400 kHz) with other ATmega128 devices, as well as any other device
with a proper tLOW acceptance margin.
Figure 153. Two-wire Serial Bus Timing

tof tHIGH tr
tLOW tLOW
SCL
tSU;STA tHD;STA tHD;DAT tSU;DAT
tSU;STO
SDA
tBUF
SPI Timing See Figure 154 and Figure 155 for details.
Characteristics
Table 135. SPI Timing Parameters
Description Mode Min Typ Max
1 SCK period Master See Table 72
2 SCK high/low Master 50% duty cycle
3 Rise/Fall time Master TBD
4 Setup Master 10
5 Hold Master 10
6 Out to SCK Master 0.5 • tsck
7 SCK to out Master 10
8 SCK to out high Master 10
9 SS low to out Slave 15
ns
10 SCK period Slave 4 • tck
11 SCK high/low Slave 2 • tck
12 Rise/Fall time Slave TBD
13 Setup Slave 10
14 Hold Slave 10
15 SCK to out Slave 15
16 SCK to SS high Slave 20
17 SS high to tri-state Slave 10
18 SS low to SCK Slave 20
319
2467E–AVR–05/02
Figure 154. SPI Interface Timing Requirements (Master Mode)
SS
6 1
SCK
(CPOL = 0)
2 2
SCK
(CPOL = 1)
4 5 3
MISO
MSB ... LSB
(Data Input)
7 8
MOSI
MSB ... LSB
(Data Output)
Figure 155. SPI Interface Timing Requirements (Slave Mode)

18
SS
10 16
9
SCK
(CPOL = 0)
11 11
SCK
(CPOL = 1)
13 14 12
MOSI
MSB ... LSB
(Data Input)
15 17
MISO
MSB ... LSB X
(Data Output)
320 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
ADC Characteristics – Preliminary Data

Table 136. ADC Characteristics
Symbol Parameter Condition Min(1) Typ(1) Max(1) Units
Resolution Single Ended Conversion 10 Bits
Differential Conversion
Gain = 1x or 20x 8 Bits
Differential Conversion
Gain = 200x 7 Bits
Single Ended Conversion
VREF = 4V
Absolute accuracy 1 TBD LSB
ADC clock = 200 kHz
ADHSM = 0
Single Ended Conversion
VREF = 4V
TBD TBD LSB
ADC clock = 1 MHz
ADHSM = 1
Integral Non-linearity VREF = 4V 0.5 LSB
Differential Non-linearity VREF = 4V 0.5 LSB
Zero Error (Offset) VREF = 4V 1 LSB
Conversion Time Free Running Conversion 65 260 µs
ADHSM = 0
Free Running Conversion 65 TBD µs
ADHSM = 1
Clock Frequency ADHSM = 0 50 200 kHz
ADHSM = 1 50 TBD kHz
(2) (3)
AVCC Analog Supply Voltage VCC - 0.3 VCC + 0.3 V
VREF Reference Voltage Single Ended Conversion 2.0 AVCC V
Differential Conversion 2.0 AVCC - 0.2 V
VIN Input voltage Single ended channels GND VREF
Differential channels TBD TBD
Input bandwidth Single ended channels TBD kHz
Differential channels 4 kHz
VINT Internal Voltage Reference 2.4 2.56 2.8 V
RREF Reference Input Resistance TBD TBD TBD kΩ
RAIN Analog Input Resistance TBD MΩ
IHSM Increased current consumption TBD µA
in High-Speed mode
(ADHSM=1)
Note: 1. Values are guidelines only. Actual values are TBD.
2. Minimum for AVCC is 2.7 V.
3. Maximum for AVCC is 5.5 V.
321
2467E–AVR–05/02
External Data Memory Timing
Table 137. External Data Memory Characteristics, 4.5 - 5.5 Volts, No Wait-state
8 MHz Oscillator Variable Oscillator
Symbol Parameter Min Max Min Max Unit
0 1/tCLCL Oscillator Frequency 0.0 16 MHz
1 tLHLL ALE Pulse Width 115 1.0tCLCL-10 ns
2 tAVLL Address Valid A to ALE Low 57.5 0.5tCLCL-5(1) ns
Address Hold After ALE Low,
5 5
3a tLLAX_ST write access ns
Address Hold after ALE Low,
5 5
3b tLLAX_LD read access ns
(1)
4 tAVLLC Address Valid C to ALE Low 57.5 0.5tCLCL-5 ns
5 tAVRL Address Valid to RD Low 115 1.0tCLCL-10 ns
6 tAVWL Address Valid to WR Low 115 1.0tCLCL-10 ns
7 tLLWL ALE Low to WR Low 47.5 67.5 0.5tCLCL-15(2) 0.5tCLCL+5(2) ns
(2) (2)
8 tLLRL ALE Low to RD Low 47.5 67.5 0.5tCLCL-15 0.5tCLCL+5 ns
9 tDVRH Data Setup to RD High 40 40 ns
10 tRLDV Read Low to Data Valid 75 1.0tCLCL-50 ns
11 tRHDX Data Hold After RD High 0 0 ns
12 tRLRH RD Pulse Width 115 1.0tCLCL-10 ns
(1)
13 tDVWL Data Setup to WR Low 42.5 0.5tCLCL-20 ns
14 tWHDX Data Hold After WR High 115 1.0tCLCL-10 ns
15 tDVWH Data Valid to WR High 125 1.0tCLCL ns
16 tWLWH WR Pulse Width 115 1.0tCLCL-10 ns
Notes: 1. This assumes 50% clock duty cycle. The half period is actually the high time of the external clock, XTAL1.
2. This assumes 50% clock duty cycle. The half period is actually the low time of the external clock, XTAL1.
Table 138. External Data Memory Characteristics, 4.5 - 5.5 Volts, 1 Cycle Wait-state
322 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table 139. External Data Memory Characteristics, 4.5 - 5.5 Volts, SRWn1 = 1, SRWn0 = 0
Table 141. External Data Memory Characteristics, 2.7 - 5.5 Volts, No Wait-state
1 tLHLL ALE Pulse Width 235 tCLCL-15 ns
(1)
2 tAVLL Address Valid A to ALE Low 115 0.5tCLCL-10 ns
Address Hold After ALE Low,
5 5
3a tLLAX_ST write access ns
Address Hold after ALE Low,
5 5
3b tLLAX_LD read access ns
(1)
4 tAVLLC Address Valid C to ALE Low 115 0.5tCLCL-10 ns
5 tAVRL Address Valid to RD Low 235 1.0tCLCL-15 ns
6 tAVWL Address Valid to WR Low 235 1.0tCLCL-15 ns
(2) (2)
7 tLLWL ALE Low to WR Low 115 130 0.5tCLCL-10 0.5tCLCL+5 ns
(2) (2)
8 tLLRL ALE Low to RD Low 115 130 0.5tCLCL-10 0.5tCLCL+5 ns
9 tDVRH Data Setup to RD High 45 45 ns
11 tRHDX Data Hold After RD High 0 0 ns
323
2467E–AVR–05/02
Table 141. External Data Memory Characteristics, 2.7 - 5.5 Volts, No Wait-state (Continued)
(1)
13 tDVWL Data Setup to WR Low 105 0.5tCLCL-20 ns
Notes: 1. This assumes 50% clock duty cycle. The half period is actually the high time of the external clock, XTAL1.
2. This assumes 50% clock duty cycle. The half period is actually the low time of the external clock, XTAL1.
324 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 156. External Memory Timing (SRWn1 = 0, SRWn0 = 0
T1 T2 T3 T4
ALE
4 7

15
2 3a 13
Write
14
6 16
WR
3b 9 11
DA7:0 (XMBK = 0) Address Data
Read
5 10
8 12
RD
Figure 157. External Memory Timing (SRWn1 = 0, SRWn0 = 1)

T1 T2 T3 T4 T5
ALE
4 7

15
2 3a 13
Write
14
6 16
WR
3b 9 11

Read
5 10
8 12
RD
325
2467E–AVR–05/02
Figure 158. External Memory Timing (SRWn1 = 1, SRWn0 = 0)
T1 T2 T3 T4 T5 T6
ALE
4 7
15
2 3a 13
Write
14
6 16
WR
3b 9 11
Read
5 10
8 12
RD
Figure 159. External Memory Timing (SRWn1 = 1, SRWn0 = 1)(1)

T1 T2 T3 T4 T5 T6 T7

1
ALE
4 7
15
2 3a 13
Write
14
6 16
WR
3b 9 11
Read
5 10
8 12
RD
Note: 1. The ALE pulse in the last period (T4-T7) is only present if the next instruction
accesses the RAM (internal or external).
326 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
ATmega128 Typical The following charts show typical behavior. These figures are not tested during manu-
facturing. All current consumption measurements are performed with all I/O pins
Characteristics –
configured as inputs and with internal pull-ups enabled. A sine wave generator with rail-
Preliminary Data to-rail output is used as clock source.
The power consumption in Power-down mode is independent of clock selection.
The current consumption is a function of several factors such as: operating voltage,
operating frequency, loading of I/O pins, switching rate of I/O pins, code executed and
ambient temperature. The dominating factors are operating voltage and frequency.
The current drawn from capacitive loaded pins may be estimated (for one pin) as
CL*VCC*f where C L = load capacitance, VCC = operating voltage and f = average switch-
ing frequency of I/O pin.
The parts are characterized at frequencies higher than test limits. Parts are not guaran-
teed to function properly at frequencies higher than the ordering code indicates.
The difference between current consumption in Power-down mode with Watchdog
Timer enabled and Power-down mode with Watchdog Timer disabled represents the dif-
ferential current drawn by the Watchdog Timer.
327
2467E–AVR–05/02
Figure 160. Active Supply Current vs. Frequency
ACTIVE SUPPLY CURRENT vs. FREQUENCY
50
45 5.5V
40 5.0V
35 4.5V
30
4.0V
ICC (mA)
25 3.6V
3.3V
20 3.0V
2.7V
15
10
0
0 2 4 6 8 10 12 14 16 18 20
Frequency (MHz)
Figure 161. Active Supply Current vs. VCC, Internal RC Oscillator 1 MHz
ACTIVE SUPPLY CURRENT vs. VCC

INTERNAL RC OSCILLATOR, 1 MHz
3,5
3 85oC
25oC
2,5
2
ICC (mA)
1,5
0,5
0
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)
328 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)

7
6
85oC
25oC
5
ICC (mA)
3
0
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)
Figure 163. Avtive Supply Current vs. VCC, Internal RC Oscillator 4 MHz

12
85oC
25oC
10
8
ICC (mA)
0
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)
329
2467E–AVR–05/02

25
85oC
20 25oC
15
ICC (mA)
10
0
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)
Figure 165. Active Supply Current vs. VCC, 32 kHz External Oscillator

32kHz EXTERNAL OSCILLATOR
200
85oC
180
25oC
160
140
120
ICC (uA)
100
80
60
40
20
0
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)
330 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 166. Idle Supply Current vs. Frequency
IDLE SUPPLY CURRENT vs. FREQUENCY
25
5.5V
20 5.0V
4.5V
15
4.0V
ICC (mA) 3.6V
3.3V
10 3.0V
2.7V
0
0 2 4 6 8 10 12 14 16 18 20
Frequency (MHz)
Figure 167. Idle Supply Current vs. VCC, Internal RC Oscillator 1 MHz
IDLE SUPPLY CURRENT vs. VCC

1,6
85oC
1,4
25oC
1,2
1
ICC (mA)
0,8
0,6
0,4
0,2
0
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)
331
2467E–AVR–05/02

3
85oC
25oC
2,5
ICC (mA)
1,5
0,5
0
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)

6
85oC
25oC
5
4
ICC (mA)
0
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)
332 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 170. Idle Supply Current vs VCC, Internal RC Oscillator 8 MHz

12
85oC
25oC
10
ICC (mA)
6
0
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)
Figure 171. Idle Supply Current vs VCC, 32 kHz External Oscillator

32kHz EXTERNAL OSCILLATOR
80
70 85oC
25oC
60
50
ICC (uA)
40
30
20
10
0
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)
333
2467E–AVR–05/02
Figure 172. Power-down Supply Current vs. VCC, Watchdog Timer Disabled
POWER-DOWN SUPPLY CURRENT vs. VCC

WATCHDOG TIMER DISABLED
7
6
85oC
ICC (uA) 3
2 25oC
0
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)
Figure 173. Power-down Supply Current vs. VCC, Watchdog Timer Enabled
POWER-DOWN SUPPLY CURRENT vs. VCC

WATCHDOG TIMER ENABLED
35
85oC
30
25oC
25
20
ICC (uA)
15
10
0
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)
334 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 174. Power-save Supply Current vs. VCC, Watchdog Timer Disabled
POWER-SAVE SUPPLY CURRENT vs. VCC

WATCHDOG TIMER DISABLED
20
85oC
18
16
14 25oC
12
ICC (uA)
10
0
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)
Figure 175. Power-save Supply Current vs. VCC, Watchdog Timer Enabled
POWER-SAVE SUPPLY CURRENT vs. VCC

WATCHDOG TIMER ENABLED
45
85oC
40
25oC
35
30
25
ICC (uA)
20
15
10
0
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)
335
2467E–AVR–05/02
Figure 176. Standby Supply Current vs. V CC, 2 MHz Resonator, Watchdog Timer
Disabled
STANDBY SUPPLY CURRENT vs. VCC

2MHz RESONATOR, WATCHDOG TIMER DISABLED
120
100 85oC
25oC
80
ICC (uA)
60
40
20
0
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)
Figure 177. Standby Supply Current vs. VCC, 2 MHz Xtal, Watchdog Timer Disabled

2MHz XTAL, WATCHDOG TIMER DISABLED
90
85oC
80
25oC
70
60
50
ICC (uA)
40
30
20
10
0
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)
336 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 178. Standby Supply Current vs. V CC, 4 MHz Resonator, Watchdog Timer
Disabled

4MHz RESONATOR, WATCHDOG TIMER DISABLED
160
140
85oC
120 25oC
100
ICC (uA)
80
60
40
20
0
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)
Figure 179. Standby Supply Current vs. VCC, 6 MHz Xtal, Watchdog Timer Disabled

6MHz XTAL, WATCHDOG TIMER DISABLED
250
85oC
200
25oC
150
ICC (uA)
100
50
0
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)
337
2467E–AVR–05/02
Figure 180. I/O Pin Pullup Resistor Current vs. Input Voltage, VCC = 5V
I/O PIN PULL-UP RESISTOR CURRENT vs. INPUT VOLTAGE

Vcc = 5V
160
140
25oC
120
85oC
100
IOP (uA)
80
60
40
20
VOP (V)
Figure 181. I/O Pin Pullup Resistor Current vs. Input Voltage, VCC = 2.7V
I/O PIN PULL-UP RESISTOR CURRENT vs. INPUT VOLTAGE

Vcc = 2.7V
80
25oC
70
60 85oC
50
IOP (uA)
40
30
20
10
0
0 0.5 1 1.5 2 2.5 3
VOP (V)
338 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 182. Reset Pin Pullup Resistor Current vs. Input Voltage, VCC = 5V
RESET PULL-UP RESISTOR CURRENT vs. PIN VOLTAGE

Vcc = 5V
120
25oC
100
85oC
80
IRESET (uA) 60
40
20
0
0 1 2 3 4 5 6
VRESET (V)
Figure 183. Reset Pin Pullup Resistor Current vs. Input Voltage, VCC = 2.7V
RESET PULL-UP RESISTOR CURRENT vs. PIN VOLTAGE

Vcc = 2.7V
60
25oC
50
40
85oC
IRESET (uA)
30
20
10
0
0 0.5 1 1.5 2 2.5 3
VRESET (V)
339
2467E–AVR–05/02
Figure 184. I/O Pin Source Current vs. Output Voltage, VCC = 5V
I/O PIN SOURCE CURRENT vs. OUTPUT VOLTAGE

Vcc = 5V
70
25oC
60
85oC
50
40
IOH (mA) 30
20
10
0
0 1 2 3 4 5 6
VOH (V)
Figure 185. I/O Pin Source Current vs. Output Voltage, VCC = 2.7V
I/O PIN SOURCE CURRENT vs. OUTPUT VOLTAGE

Vcc = 2.7V
30
25oC
25
85oC
20
IOH (mA)
15
10
0
0 0.5 1 1.5 2 2.5 3
VOH (V)
340 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 186. I/O Pin Sink Current vs. Output Voltage, VCC = 5V
I/O PIN SINK CURRENT vs. OUTPUT VOLTAGE

Vcc = 5V
80
70 25oC
60
85oC
50
IOL (mA)
40
30
20
10
0
0 0.5 1 1.5 2 2.5
VOL (V)
Figure 187. I/O Pin Sink Current vs. Output Voltage, VCC = 2.7V
I/O PIN SINK CURRENT vs. OUTPUT VOLTAGE

Vcc = 2.7V
30
25oC
25
85oC
20
IOL (mA)
15
10
0
0 0.5 1 1.5 2 2.5
VOL (V)
341
2467E–AVR–05/02
Figure 188. BOD Threshold vs. Temperature, BOD Level is 4V
BOD THRESHOLDS vs. TEMPERATURE

BOD LEVEL IS 4.0 V
5
4.5
Rising VCC
4
Falling VCC
Threshold (V)
3.5
2.5
2
0 10 20 30 40 50 60 70 80 90
Temperature (˚C)
Figure 189. BOD Threshold vs. Temperature, BOD Level is 2.7V
BOD THRESHOLDS vs. TEMPERATURE

BOD LEVEL IS 2.7 V
5
4.5
4
Threshold (V)
3.5
3 Rising VCC
2.5
Falling VCC
2
0 10 20 30 40 50 60 70 80 90
Temperature (˚C)
342 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 190. Analog Comparator Offset Voltage vs. Common Mode Voltage, VCC = 5V
ANALOG COMPARATOR OFFSET VOLTAGE vs. COMMON MODE VOLTAGE

Vcc = 5V
8
7 -40oC
Offset Voltage (mV)

5
25oC
3
85oC
2
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Common Mode Voltage (V)
Figure 191. Analog Comparator Offset Voltage vs. Common Mode Voltage, VCC = 2.7V
ANALOG COMPARATOR OFFSET VOLTAGE vs. COMMON MODE VOLTAGE

Vcc = 2.7V
-40oC
5
Offset Voltage (mV)
25oC
3
85oC
2
0
0 0.5 1 1.5 2 2.5 3
Common Mode Voltage (V)
343
2467E–AVR–05/02
Figure 192. Analog Comparator Current vs. VCC
ANALOG COMPARATOR CURRENT vs. VCC
100
85oC
90
80
25oC
70
60
ICC (uA)
50
40
30
20
10
0
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)
Figure 193. Bandgap Voltage vs. VCC
BANDGAP VOLTAGE vs. VCC
1,26
1,25
1,24
-40oC
Bandgap Voltage (V)
1,23 25oC
85oC
1,22
1,21
1,2
1,19
1,18
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)
344 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 194. Watchdog Oscillator Frequency vs. VCC
WATCHDOG OSCILLATOR FREQUENCY vs. VCC
1200
1150
25oC
1100 85oC
1050
FRC (kHz) 1000
950
900
850
800
2 2.5 3 3.5 4 4.5 5 5.5
VCC (V)
345
2467E–AVR–05/02
Figure 195. RC Oscillator Frequency vs. Temperature (the devices are calibrated to
1 MHz at Vcc = 5V, T=25c)
CALIBRATED 1MHz RC OSCILLATOR FREQUENCY

vs. TEMPERATURE
1.04
1.02
1 Vcc = 5.5V
Vcc = 5.0V
FRc (MHz)
0.98
Vcc = 4.5V
Vcc = 4.0V
0.96
Vcc = 3.6V
Vcc = 3.3V
0.94
Vcc = 3.0V
Vcc = 2.7V
0.92
-40 -20 0 20 40 60 80
Ta(˚C)
Figure 196. RC Oscillator Frequency vs. Operating Voltage (the devices are calibrated
to 1 MHz at Vcc = 5V, T=25c)

vs. OPERATING VOLTAGE
1.04
1.02 TA = -40˚C TA = -10˚C

TA = 25˚C
TA = 45˚C
1 TA = 70˚C
TA = 85˚C
FRc (MHz)
0.98
0.96
0.94
0.92
2.5 3 3.5 4 4.5 5 5.5
Vcc(V)
346 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 197. RC Oscillator Frequency vs Temperature (the devices are calibrated to


vs. TEMPERATURE
2.1
2.05
2
Vcc = 5.5V
FRc (MHz) 1.95

Vcc = 5.0V
Vcc = 4.5V
1.9 Vcc = 4.0V

Vcc = 3.6V
Vcc = 3.3V
1.85
Vcc = 3.0V
Vcc = 2.7V
1.8
-40 -20 0 20 40 60 80
Ta(˚C)

2.1
2.05 TA = -40˚C
TA = -10˚C
TA = 25˚C
TA = 45˚C
2 TA = 70˚C
TA = 85˚C
FRc (MHz)
1.95
1.9
1.85
1.8
2.5 3 3.5 4 4.5 5 5.5
Vcc(V)
347
2467E–AVR–05/02
Figure 199. RC Oscillator Frequency vs Temperature (the devices are calibrated to

vs. TEMPERATURE
4.1
4.05
3.95 Vcc = 5.5V

Vcc = 5.0V
3.9
Vcc = 4.5V
FRc (MHz) 3.85
Vcc = 4.0V
3.8 Vcc = 3.6V
3.75 Vcc = 3.3V
3.7 Vcc = 3.0V
3.65 Vcc = 2.7V
3.6
-40 -20 0 20 40 60 80
Ta(˚C)

4.1
TA = -40˚C
4.05 TA = -10˚C
TA = 25˚C
4 TA = 45˚C
TA = 70˚C
3.95
TA = 85˚C
3.9
FRc (MHz)
3.85
3.8
3.75
3.7
3.65
3.6
2.5 3 3.5 4 4.5 5 5.5
Vcc(V)
348 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Figure 201. RC Oscillator Frequency vs. Temperature (the devices are calibrated to 8
MHz at Vcc = 5V, T=25c)

vs. TEMPERATURE
8.4
8.2
8
Vcc = 5.5V
7.8 Vcc = 5.0V
FRc (MHz) 7.6

Vcc = 4.5V
Vcc = 4.0V
7.4
Vcc = 3.6V
Vcc = 3.3V
7.2
Vcc = 3.0V
7
Vcc = 2.7V
6.8
-40 -20 0 20 40 60 80
Ta(˚C)

8.4
TA = -40˚C
8.2 TA = -10˚C
TA = 25˚C
8 TA = 45˚C
TA = 70˚C
7.8
TA = 85˚C
FRc (MHz)
7.6
7.4
7.2
6.8
2.5 3 3.5 4 4.5 5 5.5
349
2467E–AVR–05/02
Register Summary
Address Name Bit 7 Bit 6 Bit 5 Bit 4 Bit 3 Bit 2 Bit 1 Bit 0 Page
($FF) Reserved – – – – – – – –
.. Reserved – – – – – – – –
($9E) Reserved – – – – – – – –
($9D) UCSR1C – UMSEL1 UPM11 UPM10 USBS1 UCSZ11 UCSZ10 UCPOL1 186
($9C) UDR1 USART1 I/O Data Register 183
($9B) UCSR1A RXC1 TXC1 UDRE1 FE1 DOR1 UPE1 U2X1 MPCM1 184
($9A) UCSR1B RXCIE1 TXCIE1 UDRIE1 RXEN1 TXEN1 UCSZ12 RXB81 TXB81 185
($99) UBRR1L USART1 Baud Rate Register Low 188
($98) UBRR1H – – – – USART1 Baud Rate Register High 188
($97) Reserved – – – – – – – –
($96) Reserved – – – – – – – –
($95) UCSR0C – UMSEL0 UPM01 UPM00 USBS0 UCSZ01 UCSZ00 UCPOL0 186
($94) Reserved – – – – – – – –
($93) Reserved – – – – – – – –
($92) Reserved – – – – – – – –
($91) Reserved – – – – – – – –
($90) UBRR0H – – – – USART0 Baud Rate Register High 188
($8F) Reserved – – – – – – – –
($8E) Reserved – – – – – – – –
($8D) Reserved – – – – – – – –
($8C) TCCR3C FOC3A FOC3B FOC3C – – – – – 132
($8B) TCCR3A COM3A1 COM3A0 COM3B1 COM3B0 COM3C1 COM3C0 WGM31 WGM30 127
($8A) TCCR3B ICNC3 ICES3 – WGM33 WGM32 CS32 CS31 CS30 130
($89) TCNT3H Timer/Counter3 – Counter Register High Byte 132
($88) TCNT3L Timer/Counter3 – Counter Register Low Byte 132
($87) OCR3AH Timer/Counter3 – Output Compare Register A High Byte 133
($86) OCR3AL Timer/Counter3 – Output Compare Register A Low Byte 133
($85) OCR3BH Timer/Counter3 – Output Compare Register B High Byte 133
($84) OCR3BL Timer/Counter3 – Output Compare Register B Low Byte 133
($83) OCR3CH Timer/Counter3 – Output Compare Register C High Byte 133
($82) OCR3CL Timer/Counter3 – Output Compare Register C Low Byte 133
($81) ICR3H Timer/Counter3 – Input Capture Register High Byte 134
($80) ICR3L Timer/Counter3 – Input Capture Register Low Byte 134
($7F) Reserved – – – – – – – –
($7E) Reserved – – – – – – – –
($7D) ETIMSK – – TICIE3 OCIE3A OCIE3B TOIE3 OCIE3C OCIE1C 135
($7C) ETIFR – – ICF3 OCF3A OCF3B TOV3 OCF3C OCF1C 136
($7B) Reserved – – – – – – – –
($7A) TCCR1C FOC1A FOC1B FOC1C – – – – – 131
($79) OCR1CH Timer/Counter1 – Output Compare Register C High Byte 133
($78) OCR1CL Timer/Counter1 – Output Compare Register C Low Byte 133
($77) Reserved – – – – – – – –
($76) Reserved – – – – – – – –
($75) Reserved – – – – – – – –
($74) TWCR TWINT TWEA TWSTA TWSTO TWWC TWEN – TWIE 201
($73) TWDR Two-wire Serial Interface Data Register 203
($72) TWAR TWA6 TWA5 TWA4 TWA3 TWA2 TWA1 TWA0 TWGCE 203
($671 TWSR TWS7 TWS6 TWS5 TWS4 TWS3 – TWPS1 TWPS0 202
($70) TWBR Two-wire Serial Interface Bit Rate Register 201
($6F) OSCCAL Oscillator Calibration Register 38
($6E) Reserved – – – – – – – –
($6D) XMCRA – SRL2 SRL1 SRL0 SRW01 SRW00 SRW11 29
($6C) XMCRB XMBK – – – – XMM2 XMM1 XMM0 31
($6B) Reserved – – – – – – – –
($6A) EICRA ISC31 ISC30 ISC21 ISC20 ISC11 ISC10 ISC01 ISC00 84
($69) Reserved – – – – – – – –
($68) SPMCSR SPMIE RWWSB – RWWSRE BLBSET PGWRT PGERS SPMEN 273
($67) Reserved – – – – – – – –
($66) Reserved – – – – – – – –
($65) PORTG – – – PORTG4 PORTG3 PORTG2 PORTG1 PORTG0 83
($64) DDRG – – – DDG4 DDG3 DDG2 DDG1 DDG0 83
($63) PING – – – PING4 PING3 PING2 PING1 PING0 83
($62) PORTF PORTF7 PORTF6 PORTF5 PORTF4 PORTF3 PORTF2 PORTF1 PORTF0 82
350 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Register Summary (Continued)

($61) DDRF DDF7 DDF6 DDF5 DDF4 DDF3 DDF2 DDF1 DDF0 83
($60) Reserved – – – – – – – –
$3F ($5F) SREG I T H S V N Z C 9
$3E ($5E) SPH SP15 SP14 SP13 SP12 SP11 SP10 SP9 SP8 12
$3D ($5D) SPL SP7 SP6 SP5 SP4 SP3 SP2 SP1 SP0 12
$3C ($5C) XDIV XDIVEN XDIV6 XDIV5 XDIV4 XDIV3 XDIV2 XDIV1 XDIV0 40
$3B ($5B) RAMPZ – – – – – – – RAMPZ0 12
$3A ($5A) EICRB ISC71 ISC70 ISC61 ISC60 ISC51 ISC50 ISC41 ISC40 85
$39 ($59) EIMSK INT7 INT6 INT5 INT4 INT3 INT2 INT1 INT0 86
$38 ($58) EIFR INTF7 INTF6 INTF5 INTF4 INTF3 INTF INTF1 INTF0 86
$37 ($57) TIMSK OCIE2 TOIE2 TICIE1 OCIE1A OCIE1B TOIE1 OCIE0 TOIE0 103, 134, 154
$36 ($56) TIFR OCF2 TOV2 ICF1 OCF1A OCF1B TOV1 OCF0 TOV0 103, 136, 155
$35 ($55) MCUCR SRE SRW10 SE SM1 SM0 SM2 IVSEL IVCE 29, 41, 58
$34 ($54) MCUCSR JTD – – JTRF WDRF BORF EXTRF PORF 49, 250
$33 ($53) TCCR0 FOC0 WGM00 COM01 COM00 WGM01 CS02 CS01 CS00 98
$32 ($52) TCNT0 Timer/Counter0 (8 Bit) 100
$31 ($51) OCR0 Timer/Counter0 Output Compare Register 100
$30 ($50) ASSR – – – – AS0 TCN0UB OCR0UB TCR0UB 101
$2F ($4F) TCCR1A COM1A1 COM1A0 COM1B1 COM1B0 COM1C1 COM1C0 WGM11 WGM10 127
$2E ($4E) TCCR1B ICNC1 ICES1 – WGM13 WGM12 CS12 CS11 CS10 130
$2D ($4D) TCNT1H Timer/Counter1 – Counter Register High Byte 132
$2C ($4C) TCNT1L Timer/Counter1 – Counter Register Low Byte 132
$2B ($4B) OCR1AH Timer/Counter1 – Output Compare Register A High Byte 133
$2A ($4A) OCR1AL Timer/Counter1 – Output Compare Register A Low Byte 133
$29 ($49) OCR1BH Timer/Counter1 – Output Compare Register B High Byte 133
$28 ($48) OCR1BL Timer/Counter1 – Output Compare Register B Low Byte 133
$27 ($47) ICR1H Timer/Counter1 – Input Capture Register High Byte 134
$26 ($46) ICR1L Timer/Counter1 – Input Capture Register Low Byte 134
$25 ($45) TCCR2 FOC2 WGM20 COM21 COM20 WGM21 CS22 CS21 CS20 152
$24 ($44) TCNT2 Timer/Counter2 (8 Bit) 154
$23 ($43) OCR2 Timer/Counter2 Output Compare Register 154
IDRD/
$22 ($42) OCDR OCDR6 OCDR5 OCDR4 OCDR3 OCDR2 OCDR1 OCDR0 247
OCDR7
$21 ($41) WDTCR – – – WDCE WDE WDP2 WDP1 WDP0 51
$20 ($40) SFIOR TSM – – ADHSM ACME PUD PSR0 PSR321 66, 104, 140, 241
$1F ($3F) EEARH – – – – EEPROM Address Register High 19
$1E ($3E) EEARL EEPROM Address Register Low Byte 19
$1D ($3D) EEDR EEPROM Data Register 20
$1C ($3C) EECR – – – – EERIE EEMWE EEWE EERE 20
$1B ($3B) PORTA PORTA7 PORTA6 PORTA5 PORTA4 PORTA3 PORTA2 PORTA1 PORTA0 81
$1A ($3A) DDRA DDA7 DDA6 DDA5 DDA4 DDA3 DDA2 DDA1 DDA0 81
$19 ($39) PINA PINA7 PINA6 PINA5 PINA4 PINA3 PINA2 PINA1 PINA0 81
$18 ($38) PORTB PORTB7 PORTB6 PORTB5 PORTB4 PORTB3 PORTB2 PORTB1 PORTB0 81
$17 ($37) DDRB DDB7 DDB6 DDB5 DDB4 DDB3 DDB2 DDB1 DDB0 81
$16 ($36) PINB PINB7 PINB6 PINB5 PINB4 PINB3 PINB2 PINB1 PINB0 81
$15 ($35) PORTC PORTC7 PORTC6 PORTC5 PORTC4 PORTC3 PORTC2 PORTC1 PORTC0 81
$14 ($34) DDRC DDC7 DDC6 DDC5 DDC4 DDC3 DDC2 DDC1 DDC0 81
$13 ($33) PINC PINC7 PINC6 PINC5 PINC4 PINC3 PINC2 PINC1 PINC0 82
$12 ($32) PORTD PORTD7 PORTD6 PORTD5 PORTD4 PORTD3 PORTD2 PORTD1 PORTD0 82
$11 ($31) DDRD DDD7 DDD6 DDD5 DDD4 DDD3 DDD2 DDD1 DDD0 82
$10 ($30) PIND PIND7 PIND6 PIND5 PIND4 PIND3 PIND2 PIND1 PIND0 82
$0F ($2F) SPDR SPI Data Register 164
$0E ($2E) SPSR SPIF WCOL – – – – – SPI2X 164
$0D ($2D) SPCR SPIE SPE DORD MSTR CPOL CPHA SPR1 SPR0 162
$0C ($2C) UDR0 USART0 I/O Data Register 183
$0B ($2B) UCSR0A RXC0 TXC0 UDRE0 FE0 DOR0 UPE0 U2X0 MPCM0 184
$0A ($2A) UCSR0B RXCIE0 TXCIE0 UDRIE0 RXEN0 TXEN0 UCSZ02 RXB80 TXB80 185
$09 ($29) UBRR0L USART0 Baud Rate Register Low 188
$08 ($28) ACSR ACD ACBG ACO ACI ACIE ACIC ACIS1 ACIS0 222
$07 ($27) ADMUX REFS1 REFS0 ADLAR MUX4 MUX3 MUX2 MUX1 MUX0 238
$06 ($26) ADCSRA ADEN ADSC ADRF ADIF ADIE ADPS2 ADPS1 ADPS0 239
$05 ($25) ADCH ADC Data Register High Byte 241
$04 ($24) ADCL ADC Data Register Low byte 241
$03 ($23) PORTE PORTE7 PORTE6 PORTE5 PORTE4 PORTE3 PORTE2 PORTE1 PORTE0 82
$02 ($22) DDRE DDE7 DDE6 DDE5 DDE4 DDE3 DDE2 DDE1 DDE0 82
351
2467E–AVR–05/02
Register Summary (Continued)
$01 ($21) PINE PINE7 PINE6 PINE5 PINE4 PINE3 PINE2 PINE1 PINE0 82
$00 ($20) PINF PINF7 PINF6 PINF5 PINF4 PINF3 PINF2 PINF1 PINF0 83
Notes: 1. For compatibility with future devices, reserved bits should be written to zero if accessed. Reserved I/O memory addresses
should never be written.
2. Some of the status flags are cleared by writing a logical one to them. Note that the CBI and SBI instructions will operate on
all bits in the I/O register, writing a one back into any flag read as set, thus clearing the flag. The CBI and SBI instructions
work with registers $00 to $1F only.
352 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Instruction Set Summary

Mnemonics Operands Description Operation Flags #Clocks
ARITHMETIC AND LOGIC INSTRUCTIONS
ADD Rd, Rr Add two Registers Rd ← Rd + Rr Z,C,N,V,H 1
ADC Rd, Rr Add with Carry two Registers Rd ← Rd + Rr + C Z,C,N,V,H 1
ADIW Rdl,K Add Immediate to Word Rdh:Rdl ← Rdh:Rdl + K Z,C,N,V,S 2
SUB Rd, Rr Subtract two Registers Rd ← Rd - Rr Z,C,N,V,H 1
SUBI Rd, K Subtract Constant from Register Rd ← Rd - K Z,C,N,V,H 1
SBC Rd, Rr Subtract with Carry two Registers Rd ← Rd - Rr - C Z,C,N,V,H 1
SBCI Rd, K Subtract with Carry Constant from Reg. Rd ← Rd - K - C Z,C,N,V,H 1
SBIW Rdl,K Subtract Immediate from Word Rdh:Rdl ← Rdh:Rdl - K Z,C,N,V,S 2
AND Rd, Rr Logical AND Registers Rd ← Rd • Rr Z,N,V 1
ANDI Rd, K Logical AND Register and Constant Rd ← Rd • K Z,N,V 1
OR Rd, Rr Logical OR Registers Rd ← Rd v Rr Z,N,V 1
ORI Rd, K Logical OR Register and Constant Rd ← Rd v K Z,N,V 1
EOR Rd, Rr Exclusive OR Registers Rd ← Rd ⊕ Rr Z,N,V 1
COM Rd One’s Complement Rd ← $FF − Rd Z,C,N,V 1
NEG Rd Two’s Complement Rd ← $00 − Rd Z,C,N,V,H 1
SBR Rd,K Set Bit(s) in Register Rd ← Rd v K Z,N,V 1
CBR Rd,K Clear Bit(s) in Register Rd ← Rd • ($FF - K) Z,N,V 1
INC Rd Increment Rd ← Rd + 1 Z,N,V 1
DEC Rd Decrement Rd ← Rd − 1 Z,N,V 1
TST Rd Test for Zero or Minus Rd ← Rd • Rd Z,N,V 1
CLR Rd Clear Register Rd ← Rd ⊕ Rd Z,N,V 1
SER Rd Set Register Rd ← $FF None 1
MUL Rd, Rr Multiply Unsigned R1:R0 ← Rd x Rr Z,C 2
MULS Rd, Rr Multiply Signed R1:R0 ← Rd x Rr Z,C 2
MULSU Rd, Rr Multiply Signed with Unsigned R1:R0 ← Rd x Rr Z,C 2
FMUL Rd, Rr Fractional Multiply Unsigned R1:R0 ← (Rd x Rr) << 1 Z,C 2
FMULS Rd, Rr Fractional Multiply Signed R1:R0 ← (Rd x Rr) << 1 Z,C 2
FMULSU Rd, Rr Fractional Multiply Signed with Unsigned R1:R0 ← (Rd x Rr) << 1 Z,C 2
BRANCH INSTRUCTIONS
RJMP k Relative Jump PC ← PC + k + 1 None 2
IJMP Indirect Jump to (Z) PC ← Z None 2
JMP k Direct Jump PC ← k None 3
RCALL k Relative Subroutine Call PC ← PC + k + 1 None 3
ICALL Indirect Call to (Z) PC ← Z None 3
CALL k Direct Subroutine Call PC ← k None 4
RET Subroutine Return PC ← STACK None 4
RETI Interrupt Return PC ← STACK I 4
CPSE Rd,Rr Compare, Skip if Equal if (Rd = Rr) PC ← PC + 2 or 3 None 1/2/3
CP Rd,Rr Compare Rd − Rr Z, N,V,C,H 1
CPC Rd,Rr Compare with Carry Rd − Rr − C Z, N,V,C,H 1
CPI Rd,K Compare Register with Immediate Rd − K Z, N,V,C,H 1
SBRC Rr, b Skip if Bit in Register Cleared if (Rr(b)=0) PC ← PC + 2 or 3 None 1/2/3
SBRS Rr, b Skip if Bit in Register is Set if (Rr(b)=1) PC ← PC + 2 or 3 None 1/2/3
SBIC P, b Skip if Bit in I/O Register Cleared if (P(b)=0) PC ← PC + 2 or 3 None 1/2/3
SBIS P, b Skip if Bit in I/O Register is Set if (P(b)=1) PC ← PC + 2 or 3 None 1/2/3
BRBS s, k Branch if Status Flag Set if (SREG(s) = 1) then PC←PC+k + 1 None 1/2
BRBC s, k Branch if Status Flag Cleared if (SREG(s) = 0) then PC←PC+k + 1 None 1/2
BREQ k Branch if Equal if (Z = 1) then PC ← PC + k + 1 None 1/2
BRNE k Branch if Not Equal if (Z = 0) then PC ← PC + k + 1 None 1/2
BRCS k Branch if Carry Set if (C = 1) then PC ← PC + k + 1 None 1/2
BRCC k Branch if Carry Cleared if (C = 0) then PC ← PC + k + 1 None 1/2
BRSH k Branch if Same or Higher if (C = 0) then PC ← PC + k + 1 None 1/2
BRLO k Branch if Lower if (C = 1) then PC ← PC + k + 1 None 1/2
BRMI k Branch if Minus if (N = 1) then PC ← PC + k + 1 None 1/2
BRPL k Branch if Plus if (N = 0) then PC ← PC + k + 1 None 1/2
BRGE k Branch if Greater or Equal, Signed if (N ⊕ V= 0) then PC ← PC + k + 1 None 1/2
BRLT k Branch if Less Than Zero, Signed if (N ⊕ V= 1) then PC ← PC + k + 1 None 1/2
BRHS k Branch if Half Carry Flag Set if (H = 1) then PC ← PC + k + 1 None 1/2
BRHC k Branch if Half Carry Flag Cleared if (H = 0) then PC ← PC + k + 1 None 1/2
BRTS k Branch if T Flag Set if (T = 1) then PC ← PC + k + 1 None 1/2
BRTC k Branch if T Flag Cleared if (T = 0) then PC ← PC + k + 1 None 1/2
BRVS k Branch if Overflow Flag is Set if (V = 1) then PC ← PC + k + 1 None 1/2
BRVC k Branch if Overflow Flag is Cleared if (V = 0) then PC ← PC + k + 1 None 1/2
353
2467E–AVR–05/02
Instruction Set Summary (Continued)
BRIE k Branch if Interrupt Enabled if ( I = 1) then PC ← PC + k + 1 None 1/2
BRID k Branch if Interrupt Disabled if ( I = 0) then PC ← PC + k + 1 None 1/2
DATA TRANSFER INSTRUCTIONS
MOV Rd, Rr Move Between Registers Rd ← Rr None 1
MOVW Rd, Rr Copy Register Word Rd+1:Rd ← Rr+1:Rr None 1
LDI Rd, K Load Immediate Rd ← K None 1
LD Rd, X Load Indirect Rd ← (X) None 2
LD Rd, X+ Load Indirect and Post-Inc. Rd ← (X), X ← X + 1 None 2
LD Rd, - X Load Indirect and Pre-Dec. X ← X - 1, Rd ← (X) None 2
LD Rd, Y Load Indirect Rd ← (Y) None 2
LD Rd, Y+ Load Indirect and Post-Inc. Rd ← (Y), Y ← Y + 1 None 2
LD Rd, - Y Load Indirect and Pre-Dec. Y ← Y - 1, Rd ← (Y) None 2
LDD Rd,Y+q Load Indirect with Displacement Rd ← (Y + q) None 2
LD Rd, Z Load Indirect Rd ← (Z) None 2
LD Rd, Z+ Load Indirect and Post-Inc. Rd ← (Z), Z ← Z+1 None 2
LD Rd, -Z Load Indirect and Pre-Dec. Z ← Z - 1, Rd ← (Z) None 2
LDD Rd, Z+q Load Indirect with Displacement Rd ← (Z + q) None 2
LDS Rd, k Load Direct from SRAM Rd ← (k) None 2
ST X, Rr Store Indirect (X) ← Rr None 2
ST X+, Rr Store Indirect and Post-Inc. (X) ← Rr, X ← X + 1 None 2
ST - X, Rr Store Indirect and Pre-Dec. X ← X - 1, (X) ← Rr None 2
ST Y, Rr Store Indirect (Y) ← Rr None 2
ST Y+, Rr Store Indirect and Post-Inc. (Y) ← Rr, Y ← Y + 1 None 2
ST - Y, Rr Store Indirect and Pre-Dec. Y ← Y - 1, (Y) ← Rr None 2
STD Y+q,Rr Store Indirect with Displacement (Y + q) ← Rr None 2
ST Z, Rr Store Indirect (Z) ← Rr None 2
ST Z+, Rr Store Indirect and Post-Inc. (Z) ← Rr, Z ← Z + 1 None 2
ST -Z, Rr Store Indirect and Pre-Dec. Z ← Z - 1, (Z) ← Rr None 2
STD Z+q,Rr Store Indirect with Displacement (Z + q) ← Rr None 2
STS k, Rr Store Direct to SRAM (k) ← Rr None 2
LPM Load Program Memory R0 ← (Z) None 3
LPM Rd, Z Load Program Memory Rd ← (Z) None 3
LPM Rd, Z+ Load Program Memory and Post-Inc Rd ← (Z), Z ← Z+1 None 3
ELPM Extended Load Program Memory R0 ← (RAMPZ:Z) None 3
ELPM Rd, Z Extended Load Program Memory Rd ← (RAMPZ:Z) None 3
ELPM Rd, Z+ Extended Load Program Memory and Post-Inc Rd ← (RAMPZ:Z), RAMPZ:Z ← RAMPZ:Z+1 None 3
SPM Store Program Memory (Z) ← R1:R0 None -
IN Rd, P In Port Rd ← P None 1
OUT P, Rr Out Port P ← Rr None 1
PUSH Rr Push Register on Stack STACK ← Rr None 2
POP Rd Pop Register from Stack Rd ← STACK None 2
BIT AND BIT-TEST INSTRUCTIONS
SBI P,b Set Bit in I/O Register I/O(P,b) ← 1 None 2
CBI P,b Clear Bit in I/O Register I/O(P,b) ← 0 None 2
LSL Rd Logical Shift Left Rd(n+1) ← Rd(n), Rd(0) ← 0 Z,C,N,V 1
LSR Rd Logical Shift Right Rd(n) ← Rd(n+1), Rd(7) ← 0 Z,C,N,V 1
ROL Rd Rotate Left Through Carry Rd(0)←C,Rd(n+1)← Rd(n),C←Rd(7) Z,C,N,V 1
ROR Rd Rotate Right Through Carry Rd(7)←C,Rd(n)← Rd(n+1),C←Rd(0) Z,C,N,V 1
ASR Rd Arithmetic Shift Right Rd(n) ← Rd(n+1), n=0..6 Z,C,N,V 1
SWAP Rd Swap Nibbles Rd(3..0)←Rd(7..4),Rd(7..4)←Rd(3..0) None 1
BSET s Flag Set SREG(s) ← 1 SREG(s) 1
BCLR s Flag Clear SREG(s) ← 0 SREG(s) 1
BST Rr, b Bit Store from Register to T T ← Rr(b) T 1
BLD Rd, b Bit load from T to Register Rd(b) ← T None 1
SEC Set Carry C←1 C 1
CLC Clear Carry C←0 C 1
SEN Set Negative Flag N←1 N 1
CLN Clear Negative Flag N←0 N 1
SEZ Set Zero Flag Z←1 Z 1
CLZ Clear Zero Flag Z←0 Z 1
SEI Global Interrupt Enable I←1 I 1
CLI Global Interrupt Disable I←0 I 1
SES Set Signed Test Flag S←1 S 1
CLS Clear Signed Test Flag S←0 S 1
354 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Instruction Set Summary (Continued)

SEV Set Twos Complement Overflow. V←1 V 1
CLV Clear Twos Complement Overflow V←0 V 1
SET Set T in SREG T←1 T 1
CLT Clear T in SREG T←0 T 1
SEH Set Half Carry Flag in SREG H←1 H 1
CLH Clear Half Carry Flag in SREG H←0 H 1
MCU CONTROL INSTRUCTIONS
NOP No Operation None 1
SLEEP Sleep (see specific descr. for Sleep function) None 1
WDR Watchdog Reset (see specific descr. for WDR/timer) None 1
BREAK Break For On-chip Debug Only None N/A
355
2467E–AVR–05/02
Ordering Information
Speed (MHz) Power Supply Ordering Code Package Operation Range
8 2.7 - 5.5V ATmega128L-8AC 64A Commercial
(0oC to 70oC)
ATmega128L-8AI 64A Industrial
(-40oC to 85oC)
16 4.5 - 5.5V ATmega128-16AC 64A Commercial
(0oC to 70oC)
ATmega128-16AI 64A Industrial
(-40oC to 85oC)
Package Type
64A 64-Lead, Thin (1.0 mm) Plastic Gull Wing Quad Flat Package (TQFP)
356 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Packaging Information
64A
64-lead, Thin (1.0 mm) Plastic Quad Flat Package
(TQFP), 14x14mm body, 2.0mm footprint, 0.8mm pitch.
Dimensions in Millimeters and (Inches)*
JEDEC STANDARD MS-026 AEB
16.25(0.640)
SQ
15.75(0.620)
PIN 1 ID
PIN 1
0.45(0.018)
0.30(0.012)
0.80(0.0315) BSC
14.10(0.555)
SQ 1.20 (0.047) MAX
13.90(0.547)
0.20(0.008)
0˚~7˚
0.09(0.004)
0.75(0.030) 0.15(0.006)
0.45(0.018) 0.05(0.002 )
*Controlliing dimension: millimeter
REV. A 04/11/2001
357
2467E–AVR–05/02
Data Sheet Change Please note that the referring page numbers in this section are referred to this docu-
ment. The referring revision in this section are referring to the document revision.
Log for ATmega128
Changes from Rev. 1. Corrected Description of Alternate Functions of Port G

2467B-09/01 to Rev. Corrected description of TOSC1 and TOSC2 in “Alternate Functions of Port G” on
2467C-02/02 page 79
2. Added JTAG Version Numbers for rev. F and rev. G

Updated Table 99 on page 249.
3 Added Some Preliminary Test Limits and Characterization Data

Removed some of the TBD's in the following tables and pages:
Table 19 on page 46, Table 20 on page 50, Table 130 on page 302, “DC Character-
istics” on page 315, Table 132 on page 317, Table 135 on page 319, and Table 136
on page 321.
4. Corrected “Ordering Information” on page 356.
5. Added some Characterization Data in Section “ATmega128 Typical Character-

istics – Preliminary Data” on page 327.
6. Removed Alternative Algortihm for Leaving JTAG Programming Mode.

See “Leaving Programming Mode” on page 312.
7. Added Description on How to Access the Extended Fuse Byte Through JTAG
Programming Mode.
See “Programming the Fuses” on page 313 and “Reading the Fuses and Lock Bits”
on page 314.
Changes from Rev. 1. Added more information about “ATmega103 Compatibility Mode” on page 5.
2467C-02/02 to Rev.
2467D-03/02 2. Updated Table 2, “EEPROM Programming Time,” on page 21.
3. Updated typical Start-up Time in Table 7 on page 34, Table 9 and Table 10 on
page 36, Table 12 on page 37, Table 14 on page 38, and Table 16 on page 39.
4. Updated Table 22 on page 52 with typical WDT Time-out
5. Corrected description of ADSC bit in “ADC Control and Status Register A –

ADCSRA” on page 239.
6. Improved description on how to do a polarity check of the ADC diff results in

“ADC Conversion Result” on page 236.
7. Corrected JTAG version numbers in “JTAG Version Numbers” on page 249.
8. Improved description of addressing during SPM (usage of RAMPZ) on

“Addressing the Flash During Self-Programming” on page 274, “Performing
Page Erase by SPM” on page 276, and “Performing a Page Write” on page
276.
358 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
9. Added not regarding OCDEN Fuse below Table 118 on page 284.
10. Updated Programming Figures:

Figure 134 on page 286 and Figure 143 on page 298 are updated to also reflect that
AVCC must be connected during Programming mode. Figure 138 on page 293
added to illustrate how to program the fuses.
11. Added a note regarding usage of the PROG_PAGELOAD and

PROG_PAGEREAD instructions on page 304.
12. Added Calibrated RC Oscillator characterization curves in section

“ATmega128 Typical Characteristics – Preliminary Data” on page 327.
13. Updated “Two-wire Serial Interface” section.

More details regarding use of the TWI Power-down operation and using the TWI as
master with low TWBRR values are added into the data sheet. Added the note at
the end of the “Bit Rate Generator Unit” on page 199. Added the description at the
end of “Address Match Unit” on page 200.
14. Added a note regarding usage of Timer/Counter0 combined with the clock.
See “XTAL Divide Control Register – XDIV” on page 40.
Changes from Rev. 1. Updated the Characterization Data in Section “ATmega128 Typical Character-
2467D-03/02 to Rev. istics – Preliminary Data” on page 327.
2467E-04/02
2. Updated the folowing tables: Table 19 on page 46, Table 20 on page 50, Table
68 on page 154, Table 102 on page 255, and Table 136 on page 321.
3. Updated Description of OSCCAL Calibration Byte.

In the data sheet, it was not explained how to take advantage of the calibration
bytes for 2, 4, and 8 MHz Oscillator selections. This is now added in the following
sections:
Improved description of “Oscillator Calibration Register – OSCCAL” on page 38 and
“Calibration Byte” on page 285.
359
2467E–AVR–05/02
Erratas The revision letter in this section refers to the revision of the ATmega128 device.
ATmega128 Rev. F There are no errata for this revision of ATmega128.
ATmega128 Rev. G There are no errata for this revision of ATmega128.
360 ATmega128(L)
2467E–AVR–05/02
ATmega128(L)
Table of Contents Features................................................................................................. 1
Pin Configurations................................................................................ 2
Overview................................................................................................ 2
Block Diagram ...................................................................................................... 3
ATmega103 and ATmega128 Compatibility ......................................................... 4
Pin Descriptions.................................................................................................... 5
About Code Examples.......................................................................... 7
AVR CPU Core ...................................................................................... 8

Introduction ........................................................................................................... 8
Architectural Overview .......................................................................................... 8
ALU – Arithmetic Logic Unit.................................................................................. 9
Status Register ..................................................................................................... 9
General Purpose Register File ........................................................................... 10
Stack Pointer ...................................................................................................... 12
Instruction Execution Timing............................................................................... 13
Reset and Interrupt Handling .............................................................................. 13
AVR ATmega128 Memories ............................................................... 16

In-System Reprogrammable Flash Program Memory ........................................ 16
SRAM Data Memory........................................................................................... 17
EEPROM Data Memory...................................................................................... 19
I/O Memory ......................................................................................................... 24
External Memory Interface.................................................................................. 24
System Clock and Clock Options ..................................................... 33

Clock Systems and their Distribution .................................................................. 33
Clock Sources..................................................................................................... 34
Crystal Oscillator................................................................................................. 34
Low-frequency Crystal Oscillator ........................................................................ 36
External RC Oscillator ........................................................................................ 37
Calibrated Internal RC Oscillator ........................................................................ 38
External Clock..................................................................................................... 39
Timer/Counter Oscillator ..................................................................................... 40
Power Management and Sleep Modes.............................................. 41

Idle Mode ............................................................................................................ 42
ADC Noise Reduction Mode ............................................................................... 42
Power-down Mode.............................................................................................. 42
Power-save Mode ............................................................................................... 42
Standby Mode..................................................................................................... 43
Extended Standby Mode .................................................................................... 43
Minimizing Power Consumption ......................................................................... 44
i
2467E–AVR–04/02
System Control and Reset ................................................................. 45
Internal Voltage Reference ................................................................................. 50
Watchdog Timer ................................................................................................. 50
Timed Sequences for Changing the Configuration of the Watchdog Timer ....... 53
Interrupts ............................................................................................. 54
Interrupt Vectors in ATmega128......................................................................... 54
I/O Ports............................................................................................... 60
Introduction ......................................................................................................... 60
Ports as General Digital I/O ................................................................................ 61
Alternate Port Functions ..................................................................................... 65
Register Description for I/O Ports ....................................................................... 81
External Interrupts.............................................................................. 84
8-bit Timer/Counter0 with PWM and Asynchronous Operation ..... 87

Overview ............................................................................................................. 87
Timer/Counter Clock Sources............................................................................. 88
Counter Unit........................................................................................................ 88
Output Compare Unit.......................................................................................... 89
Compare Match Output Unit ............................................................................... 91
Modes of Operation ............................................................................................ 92
Timer/Counter Timing Diagrams......................................................................... 96
8-bit Timer/Counter Register Description ........................................................... 98
Asynchronous Operation of the Timer/Counter ................................................ 101
Timer/Counter Prescaler................................................................................... 104
16-bit Timer/Counter (Timer/Counter1 and Timer/Counter3)........ 106

Overview ........................................................................................................... 106
Accessing 16-bit Registers ............................................................................... 109
Timer/Counter Clock Sources........................................................................... 112
Counter Unit...................................................................................................... 112
Input Capture Unit............................................................................................. 113
Output Compare Units ...................................................................................... 115
Compare Match Output Unit ............................................................................. 117
Modes of Operation .......................................................................................... 118
Timer/Counter Timing Diagrams....................................................................... 125
16-bit Timer/Counter Register Description ....................................................... 127
Timer/Counter3, Timer/Counter2, and Timer/Counter1 Prescalers ....

139
8-bit Timer/Counter2 with PWM....................................................... 141

Overview ........................................................................................................... 141
Timer/Counter Clock Sources........................................................................... 142
ii ATmega128(L)
2467E–AVR–04/02
ATmega128(L)
Counter Unit...................................................................................................... 142

Output Compare Unit........................................................................................ 143
Compare Match Output Unit ............................................................................. 145
Modes of Operation .......................................................................................... 146
Timer/Counter Timing Diagrams....................................................................... 150
8-bit Timer/Counter Register Description ...................................... 152
Output Compare Modulator (OCM1C2)........................................... 156

Overview ........................................................................................................... 156
Description........................................................................................................ 156
Serial Peripheral Interface – SPI...................................................... 158

SS Pin Functionality.......................................................................................... 162
Data Modes ...................................................................................................... 165
USART ............................................................................................... 166

Overview ........................................................................................................... 166
Clock Generation .............................................................................................. 168
Frame Formats ................................................................................................. 171
USART Initialization.......................................................................................... 172
Data Transmission – The USART Transmitter ................................................. 173
Data Reception – The USART Receiver .......................................................... 175
Asynchronous Data Reception ......................................................................... 179
Multi-processor Communication Mode ............................................................. 182
USART Register Description ............................................................................ 183
Examples of Baud Rate Setting ........................................................................ 189
Two-wire Serial Interface ................................................................. 193

Features............................................................................................................ 193
Two-wire Serial Interface Bus Definition........................................................... 193
Data Transfer and Frame Format ..................................................................... 194
Multi-master Bus Systems, Arbitration and Synchronization ............................ 196
Overview of the TWI Module ............................................................................ 199
TWI Register Description .................................................................................. 201
Using the TWI ................................................................................................... 203
Transmission Modes......................................................................................... 207
Multi-master Systems and Arbitration ............................................................... 220
Analog Comparator .......................................................................... 222

Analog Comparator Multiplexed Input .............................................................. 224
Analog to Digital Converter ............................................................. 225

Features............................................................................................................ 225
Operation .......................................................................................................... 227
Starting a Conversion ....................................................................................... 227
iii
2467E–AVR–04/02
Prescaling and Conversion Timing ................................................................... 228
Changing Channel or Reference Selection ...................................................... 230
ADC Noise Canceler......................................................................................... 231
ADC Conversion Result.................................................................................... 236
JTAG Interface and On-chip Debug System .................................. 242
Features............................................................................................. 242
Overview ........................................................................................................... 242
Test Access Port – TAP.................................................................................... 242
TAP Controller .................................................................................................. 244
Using the Boundary-scan Chain ....................................................................... 245
Using the On-chip Debug System .................................................................... 245
On-chip Debug Specific JTAG Instructions ...................................................... 246
On-chip Debug Related Register in I/O Memory .............................................. 247
Using the JTAG Programming Capabilities ...................................................... 247
Bibliography ...................................................................................................... 247
IEEE 1149.1 (JTAG) Boundary-scan................................................................ 248
Features............................................................................................................ 248
System Overview.............................................................................................. 248
Data Registers .................................................................................................. 248
Boundary-scan Specific JTAG Instructions ...................................................... 250
Boundary-scan Related Register in I/O Memory .............................................. 251
Boundary-scan Chain ....................................................................................... 252
ATmega128 Boundary-scan Order ................................................................... 262
Boundary-scan Description Language Files ..................................................... 268
Boot Loader Support – Read-While-Write Self-Programming...... 269

Boot Loader Features ....................................................................................... 269
Application and Boot Loader Flash Sections .................................................... 269
Read-While-Write and No Read-While-Write Flash Sections ........................... 269
Boot Loader Lock Bits....................................................................................... 271
Entering the Boot Loader Program ................................................................... 272
Addressing the Flash During Self-Programming .............................................. 274
Self-Programming the Flash ............................................................................. 275
Memory Programming...................................................................... 282

Program and Data Memory Lock Bits............................................................... 282
Fuse Bits........................................................................................................... 283
Signature Bytes ................................................................................................ 285
Calibration Byte ................................................................................................ 285
Parallel Programming Parameters, Pin Mapping, and Commands .................. 286
Parallel Programming ....................................................................................... 288
Serial Downloading ........................................................................................... 297
SPI Serial Programming Pin Mapping .............................................................. 297
Programming Via the JTAG Interface ............................................................... 302
iv ATmega128(L)
2467E–AVR–04/02
ATmega128(L)
Electrical Characteristics................................................................. 315

Absolute Maximum Ratings*............................................................................. 315
DC Characteristics............................................................................................ 315
External Clock Drive Waveforms ...................................................................... 317
External Clock Drive ......................................................................................... 317
Two-wire Serial Interface Characteristics ......................................................... 318
SPI Timing Characteristics ............................................................................... 319
ADC Characteristics – Preliminary Data........................................................... 321
External Data Memory Timing .......................................................................... 322
ATmega128 Typical Characteristics – Preliminary Data............... 327
Register Summary ............................................................................ 350
Instruction Set Summary ................................................................. 353
Ordering Information........................................................................ 356
Packaging Information ..................................................................... 357

64A ................................................................................................................... 357
Data Sheet Change Log for ATmega128 ........................................ 358

Changes from Rev. 2467B-09/01 to Rev. 2467C-02/02 ................................... 358
Changes from Rev. 2467C-02/02 to Rev. 2467D-03/02................................... 358
Changes from Rev. 2467D-03/02 to Rev. 2467E-04/02 ................................... 359
Erratas ............................................................................................... 360

ATmega128 Rev. F........................................................................................... 360
ATmega128 Rev. G .......................................................................................... 360
Table of Contents .................................................................................. i
v
2467E–AVR–04/02
vi ATmega128(L)
2467E–AVR–04/02
Atmel Headquarters Atmel Operations
Corporate Headquarters Memory RF/Automotive
2325 Orchard Parkway 2325 Orchard Parkway Theresienstrasse 2
San Jose, CA 95131 San Jose, CA 95131 Postfach 3535
TEL 1(408) 441-0311 TEL 1(408) 441-0311 74025 Heilbronn, Germany
FAX 1(408) 487-2600 FAX 1(408) 436-4314 TEL (49) 71-31-67-0
FAX (49) 71-31-67-2340
Europe Microcontrollers
Atmel Sarl 2325 Orchard Parkway 1150 East Cheyenne Mtn. Blvd.
Route des Arsenaux 41 San Jose, CA 95131 Colorado Springs, CO 80906
Case Postale 80 TEL 1(408) 441-0311 TEL 1(719) 576-3300
CH-1705 Fribourg FAX 1(408) 436-4314 FAX 1(719) 540-1759
Switzerland
TEL (41) 26-426-5555 La Chantrerie Biometrics/Imaging/Hi-Rel MPU/
FAX (41) 26-426-5500 BP 70602 High Speed Converters/RF Datacom
44306 Nantes Cedex 3, France Avenue de Rochepleine
Asia TEL (33) 2-40-18-18-18 BP 123
Room 1219 FAX (33) 2-40-18-19-60 38521 Saint-Egreve Cedex, France
Chinachem Golden Plaza TEL (33) 4-76-58-30-00
77 Mody Road Tsimhatsui ASIC/ASSP/Smart Cards FAX (33) 4-76-58-34-80
East Kowloon Zone Industrielle
Hong Kong 13106 Rousset Cedex, France
TEL (852) 2721-9778 TEL (33) 4-42-53-60-00
FAX (852) 2722-1369 FAX (33) 4-42-53-60-01
Japan 1150 East Cheyenne Mtn. Blvd.

9F, Tonetsu Shinkawa Bldg. Colorado Springs, CO 80906
1-24-8 Shinkawa TEL 1(719) 576-3300
Chuo-ku, Tokyo 104-0033 FAX 1(719) 540-1759
Japan
TEL (81) 3-3523-3551 Scottish Enterprise Technology Park
FAX (81) 3-3523-7581 Maxwell Building
East Kilbride G75 0QR, Scotland
TEL (44) 1355-803-000
FAX (44) 1355-242-743
e-mail
literature@atmel.com
Web Site
http://www.atmel.com
© Atmel Corporation 2002.

Atmel Corporation makes no warranty for the use of its products, other than those expressly contained in the Company’s standard warranty
which is detailed in Atmel’s Terms and Conditions located on the Company’s web site. The Company assumes no responsibility for any errors
which may appear in this document, reserves the right to change devices or specifications detailed herein at any time without notice, and does
not make any commitment to update the information contained herein. No licenses to patents or other intellectual property of Atmel are granted
by the Company in connection with the sale of Atmel products, expressly or by implication. Atmel’s products are not authorized for use as critical
components in life support devices or systems.
ATMEL ®, AVR ® and AVR Studio ® are the registered trademarks of Atmel.
Microsoft®, Windows® and Windows NT® are the registered trademarks of Microsoft Corporation.
Printed on recycled paper.
Other terms and product names may be the trademarks of others.
2467E–AVR–04/02 0M

ATmega 128

Uploaded by

Copyright:

Available Formats

ATmega 128

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

ATmega 128

Uploaded by

Copyright:

Available Formats

Features

• High-performance, Low-power AVR® 8-bit Microcontroller

Figure 2. Block Diagram

8-BIT DATA BUS

PROGRAM STACK WATCHDOG

PROGRAM SRAM MCU CONTROL TIMING AND

USART0 SPI TWO-WIRE SERIAL

PORTE DRIVERS PORTB DRIVERS PORTD DRIVERS PORTG DRIVERS

PE0 - PE7 PB0 - PB7 PD0 - PD7 PG0 - PG4

VCC Digital supply voltage.

XTAL2 Output from the inverting Oscillator amplifier.

AVR CPU Core

Architectural Overview Figure 3. Block Diagram of the AVR Architecture

Data Bus 8-bit

Data I/O Module 2

• Bit 6 – T: Bit Copy Storage

• Bit 5 – H: Half Carry Flag

• Bit 4 – S: Sign Bit, S = N ⊕ V

• Bit 3 – V: Two’s Complement Overflow Flag

• Bit 2 – N: Negative Flag

• Bit 1 – Z: Zero Flag

• Bit 0 – C: Carry Flag

Figure 4. AVR CPU General Purpose Working Registers

Figure 5. The X-, Y-, and Z-registers

RAM Page Z Select Register –

• Bits 7..2 – Res: Reserved Bits

• Bit 1 – RAMPZ0: Extended RAM Page Z-pointer

RAMPZ0 = 0: Program memory address $0000 - $7FFF (lower 64K bytes) is

Figure 6. The Parallel Instruction Fetches and Instruction Executions

Figure 7. Single Cycle ALU Operation

Register Operands Fetch

ALU Operation Execute

Result Write Back

Figure 8. Program Memory Map

Application Flash Section

Boot Flash Section

Table 1. Memory Configurations

Figure 9 shows how the ATmega128 SRAM Memory is organized.

Figure 9. Data Memory Map

Memory Configuration A Memory Configuration B

Data Memory Data Memory

External SRAM External SRAM

Figure 10. On-chip Data SRAM Access Cycles

Memory access instruction Next instruction

EEPROM Address Register –

• Bits 11..0 – EEAR11..0: EEPROM Address

EEPROM Data Register –

• Bits 7..0 – EEDR7.0: EEPROM Data

EEPROM Control Register –

• Bits 7..4 – Res: Reserved Bits

• Bit 3 – EERIE: EEPROM Ready Interrupt Enable

• Bit 2 – EEMWE: EEPROM Master Write Enable

• Bit 1 – EEWE: EEPROM Write Enable

1. Wait until EEWE becomes zero.

• Bit 0 – EERE: EEPROM Read Enable

Table 2. EEPROM Programming Time

Figure 11. External Memory with Sector Select

Internal memory Internal memory

External Memory Upper sector External Memory