2013 International Conference on Computer Communication and Informatics ( ICCCI -2013), Jan.

09 11, 2013, Coimbatore, INDIA

A 1-V 1-GS/s 6-bit Low-Power Flash ADC in 90-nm CMOS with 15.75 mW Power Consumption
Kirankumar Lad1 and M S Bhat2 Department of Electronics and Communication National Institute Of Technology Karnataka Surathkal-575025, India Email: kirankumarlad@gmail.com1, msbhat@ieee.org2
AbstractA 1-V 1-GS/s 6-bit low power flash ADC in 90 nm CMOS technology is presented. Proposed Flash ADC consists of reference generator, comparator array, 1-out-of N code generator, Fat tree encoder and output D-latches. This Flash ADC achieves 5.76 ENOB at Nyquist input frequency without calibration. The measured peak INL and DNL are 0.08LSB and 0.1LSB, respectively. The proposed ADC consumes 15.75 mW from 1V supply and yielding an energy efficiency of 0.291 pJ/conv while operating at 1 GS/s.

II.

FLASH ADC DESIGN

The proposed ADC block diagram is shown in Figure 1. Individual blocks are briefly explained below

Keywords- Flash ADC, Preamplifier based latch comparator, Low

I.

INTRODUCTION

Analog to digital converters are an essential part of devices that interact with real world. Flash ADC is one of the most preferred architectures for high speed analog to digital data conversion applications. Flash ADC is also used as a building block for many architectures like folding and interpolating, pipelined, sub-ranging ADCs etc. Although the exponential increase in area and power consumption with resolution limits its use in hand held applications, Flash ADCs are used in applications requiring high speed of operation at a relatively low resolution (up to 7 bits). In recently published research papers, inverter based comparators are frequently used in the design of low power flash ADCs. In [1], a 6-bit low power flash ADC using inverter based comparators is presented which consumes less than 300W of power at a sampling rate of 50MS/s. In [2], a differential clocked comparator architecture is used as a building block for high speed flash ADC applications. Although, the sampling speed in this design is reported to be 1GS/s, the comparator block alone consumes 2mW of power. The present work intends to fill the gap by proposing a 6 bit flash ADC for high speed applications (up to 5 GS/s) and consuming slightly higher power compared to the inverter based ADC. The proposed ADC employs a modified version of the comparator block presented in [2] to achieve low power and high speed of operation. The rest of the paper is organized as follows. Section II briefly describes the flash ADC architecture and design. Simulation results are discussed in section III. Section IV presents a comparison of this work with recently published flash ADCs and section V concludes the paper.

Figure 1. Flash ADC block diagram A. Reference Voltage Generator A resistor ladder comprising 2N equal resistors is used to generate 2N equally spaced reference voltages for the comparators. Total resistance Rtot and total input capacitance of all comparators Cintot is calculated using equations 1 and 2 as shown below [3] : (1) where N is ADC resolution and fin is maximum input frequency and, (2) where C is approximate input capacitance of each comparator. In this work, the input dynamic range is fixed as 0.6 V peakto-peak and the reference voltage ranges from 0.2 V to 0.8 V with a step size of 9.375 mV (LSB).

978-1-4673-2907-1/13/$31.00 2013 IEEE

Figure 2. Proposed Comparator B. Voltage Comparator The proposed voltage comparator circuit diagram is shown in Fig. 2. This comparator consists of three stages: a preamplifier stage, a latch and a post amplifier stage [4]. The detail operation of each stage is described in following subsections: 1) Preamplifier : The stage-1 of proposed comparator shown in Fig. 2 is working as a preamplifier. It consists of nmos input differential pair (M1 and M2) and diode connected pmos pair (M3 and M4). In order to overcome dynamic offset in regenerative latch, this preamplifier is needed. In this Flash ADC frontend sample and hold circuit is not used. Because of this, each preamplifier in the proposed comparator has to continuously track the input signal. The large difference in delays across different preamplifiers result in error (or bubble) code. Furthermore, due to different changing rate of input signal that each preamplifier sees, its delay is amplitude dependent. This input dependent delay gives rise to 3rd harmonic distortion. In order to suppress this distortion, one has to make the preamplifiers fast enough, or equivalently they should have large bandwidth. The output of preamplifier is given to the input of latch through transistor M5 and M6 as shown in Fig.2. 2) Latch: The latch stage (stage-2 of proposed comparator) consists of cross coupled pair of NMOS and PMOS transistors that are connected to the ground through the clock enable transistor M10. When CLK is low (and CLKB is high), the latch is in its reset state and its output voltages are at the midpoint of the rails. Regeneration of output in evaluation phase is faster in this case because the latch output is at the midpoint of VDD and GND. During the reset phase, the preamplifier translates the voltage difference between the inputs of the comparator into an unbalanced state in the latch stage. Then, in the evaluation phase (when CLK goes high), the latch stage is activated. Because of the positive feedback, this unbalanced state is amplified towards the rail voltages. Output of the latch stage is fed to the third stage of comparator, which converts the differential output into a single ended output. 3) Postamplifier: The postamplifier stage (stage-3 of the comparator) consists of a self-biased single-ended differential amplifier and two inverters. During the reset phase of latch, output of the postamplifier goes low. In the evaluation phase, it converts differential output of latch to single-ended output. The inverter at the output of the self-biased differential amplifier introduces additional gain and isolates any load capacitance from self-biased differential amplifier. Output of the postamplifier stage goes to 1-outof-N code generator. 4) 1- out of N code generator, Fat tree encoder and Dlatch: The output of the comparators form a thermometer code. It is converted to 1-outof-N code using an array of NOT and AND gates as shown in Fig. 1. This NOT-AND array is realized using NOT and NOR gates. This 1-outof-N code is fed to a fat tree encoder, which converts it into a binary code. The fat tree encoder performs better than ROM encoder in terms of speed and power in case of a 6-bit flash ADC and it also consumes less silicon area [5]. The OR gates of fat tree encoder are realized using NAND and NOR gates. The six output binary bits from the fat tree encoder are latched using a positive edge D-latch, which gives synchronized output and acts as an output register. III. SIMULATION RESULTS The proposed ADC is implemented in a 90 nm CMOS technology and simulated using Cadence Spectre simulator. The supply voltage and reference voltage used are 1 V and 0.6 V, respectively. The sampling rate of this ADC is 1 GHz. The DFT spectrum at input frequency of 496.09 MHz and sampling frequency of 1 GHz is shown in Fig. 5 and at the same input frequency, the SNDR and SFDR are found to be 36.45 dB and 45 dB respectively. The plot of input frequency versus SNDR and SFDR at sampling frequency of 1 GHz is shown in Fig. 6. In the full Nyquist input frequency range, this ADC achieves an ENOB of more than 5.5 bits at 1 GS/s. This ADC has an effective resolution bandwidth (ERBW) of 710

2013 International Conference on Computer Communication and Informatics ( ICCCI -2013), Jan. 09 11, 2013, Coimbatore, INDIA

MHz. The plot of measured DNL and INL are shown in Fig. 3 and Fig. 4 respectively. The measured peak DNL and INL are found to be 0.08LSB and 0.1LSB respectively. At 1 GS/s, the total power consumed by this ADC is 15.75 mW from a 1 V power supply. Figure of Merit (FoM) of this ADC is 291 fJ/conv. Where the FoM is defined as

(3)

The performance specification of this ADC are listed in TableI.

Figure 5. 256 point FFT of 496 MHz reconstructed signal

Figure 3. DNL plot of the ADC

Figure 6. SNDR and SFDR of the ADC for different input frequencies IV. COMPARISON WITH SOME RECENT FLASH ADCS Comparison of this work with recently reported high sampling speed ADCs is given in Table-II. From this table , it can be observed that the present work has lowest power among other ADCs except Komars work[1] which uses inverter based comparators. The ADC works at a sampling speed of 1 GS/s, a FoM of below 1 pJ/conv and has a good ENOB compared to other ADCs operating at at Nyquist frequency. Further, this ADC consumes lesser power, it can be used in high speed low power ADC applications. V. CONCLUSION

Figure 4. INL plot of the ADC TABLE I: Performance Summary of the ADC
Parameter Technology Supply Voltage Resolution Sampling Frequency Input Range SNDR ENOB Maximum DNL, INL Total Power FoM Value 90 nm CMOS 1V 6 bit 1 GS/s 600 mV peak to peak 36.45 dB @ 496 MHz, 37 dB @ 121 MHz 5.76 @ 496 MHz, 5.82 @ 121 MHz 0.08 LSB, 0.1 LSB 15.75 mW 0.291 pJ/conv @ 496 MHz

In this paper, the design and the simulation results of 1-V 1GS/s 6-bit low power Flash ADC are represented. This architecture can be extended to high speed applications because the comparator used in this ADC can work up to 5 GS/s. As the ADC has high input bandwidth, low power consumption and high linearity, this ADC is most suitable for high speed communication applications up to 1 GS/s.

TABLE II: Comparison with low voltage state of the art designs
Author/Year [6] 2005 [7] - 2006 [8] 2006 [9] 2008 [10] 2008 [11] 2008 [12] 2008 [13] 2009 [14] 2009 [15] 2010 [16] 2011 [1]-2012 This Work Core Architecture Flash Flash Flash Flash Flash Flash Flash Flash Two-Channel Two-Step Flash Flash Inverter Flash Flash VDD (V) 1.5 1.8 1.8 1.2 1 1.2 1.8 1.2 1.2 1 1.2 0.5 1 N (bit) 6 5 6 6 8 6 6 6 6 5 7 6 6 Speed (GS/s) 1.2 3.2 1.056 2.5 1.25 1.2 1 1 1 3 1.5 0.05 1 Power (mW) 160 227 98 30 207 40.5 550 112 49 36 204 0.3 15.75 ENOB@Nyquist (bit) 5.7 3.627@1G 5.0 4.0 6.9 5.2 5.1@7.1 MHz 5.1@100MHz 5.3 4.3 6.05 4.88@5.1MHz 5.76 Tech. (nm) 130 90 180 90 90 90 180 130 130 65 90 180 90 FoM (pJ/conv) 2.2 3.1 1.4 0.9 1.24 0.6 2.05 0.291 DNL/INL (LSB) 0.93/0.89 0.47/0.56 0.48/0.54 1.3/1.1 0.9/0.8 1/1.1 1.1/1.1 0.28/0.30 0.36/0.41 0.7/0.64 0.025/0.375 0.08/0.1

REFERENCES
[1] R. Komar, M. Bhat, and T. Laxminidhi, A 0.5 v 300w 50ms/s 180nm 6bit flash adc using inverter based comparators, in Advances in Engineering, Science and Management (ICAESM), 2012 International Conference on. IEEE, 2012, pp. 331335. [2] S. Sheikhaei, S. Mirabbasi, and A. Ivanov, A 0.35 m cmos comparator circuit for high-speed adc applications, in ISCAS-2005. IEEE International Symposium on Circuits and Systems . IEEE, 2005, pp. 61346137. [3] A. Venes and R. van-de Plassche, An 80-mhz, 80-mw, 8-b cmos folding a/d converter with distributed track-and-hold preprocessing, IEEE Journal of Solid-State Circuits, vol. 31, no. 12, pp. 18461853, 1996. [4] R. Baker, CMOS Mixed Signal Circuit Design. John Wiley & Sons, 2008. [5] D. Lee, J. Yoo, K. Choi, and J. Ghaznavi, Fat tree encoder design for ultra-high speed flash a/d converters, in MWSCAS-2002. The 2002 45th Midwest Symposium on Circuits and Systems, vol. 2. IEEE, 2002, pp. II87. [6] C. Sandner, M. Clara, A. Santner, T. Hartig, and F. Kuttner, A 6 -bit 1.2-gs/s low-power flash-adc in 0.13-m digital cmos, Solid-State Circuits, IEEE Journal of, vol. 40, no. 7, pp. 14991505, 2005. [7] S. Park, Y. Palaskas, A. Ravi, R. Bishop, and M. Flynn, A 3.5 gs/s 5-b flash adc in 90 nm cmos, in Custom Integrated Circuits Conference, 2006. CICC06. IEEE. IEEE, 2006, pp. 489492. [8] J. Ma, S. Sin, U. Seng-Pan, and R. Martins, A power-efficient 1.056 gs/s resolution-switchable 5-bit/6-bit flash adc for uwb applications,

[9] [10] [11]

[12]

[13]

[14] [15] [16]

in Circuits and Systems, 2006. ISCAS 2006. Proceedings. 2006 IEEE International Symposium on. IEEE, 2006, pp. 4pp. T. Sundstrom and A. Alvandpour, A 2.5 -gs/s 30-mw 4-bit flash adc in 90nm cmos, in NORCHIP, 2008. IEEE, 2008, pp. 264267. H. Yu and M. Chang, A 1-v 1.25-gs/s 8-bit self-calibrated flash adc in 90-nm digital cmos, Circuits and Systems II: Express Briefs, IEEE Transactions on, vol. 55, no. 7, pp. 668672, 2008. H. Wei, U. Chio, S. Sin, R. Martins, et al., A power scalable 6-bit 1.2 gs/s flash adc with power on/off track-and-hold and preamplifier, in Circuits and Systems, 2008. ISCAS 2008. IEEE International Symposium on. IEEE, 2008, pp. 58. C. Chang, C. Hsiao, and C. Yang, A 1-gs/s cmos 6-bit flash adc with an offset calibrating method, in VLSI Design, Automation and Test, 2008. VLSI-DAT 2008. IEEE International Symposium on. IEEE, 2008, pp. 232235. Y. Lien, Y. Lin, and S. Chang, A 6 -bit 1gs/s low-power flash adc, in VLSI Design, Automation and Test, 2009. VLSI-DAT09. International Symposium on. IEEE, 2009, pp. 211214. H. Chen, I. Chen, H. Tseng, and H. Chen, A 1 -gs/s 6-bit two-channel two-step adc in 0.13- formula formulatype=, Solid-State Circuits, IEEE Journal of, vol. 44, no. 11, pp. 30513059, 2009. T. Ito and T. Itakura, A 3-gs/s 5-bit 36-mw flash adc in 65-nm cmos, in Solid State Circuits Conference (A-SSCC), 2010 IEEE Asian. IEEE, 2010, pp. 14. J. Pernillo and M. Flynn, A 1.5-gs/s flash adc with 57.7-db sfdr and 6.4-bit enob in 90 nm digital cmos, Circuits and Systems II: Express Briefs, IEEE Transactions on, vol. 58, no. 12, pp. 837841, 2011.