28
A 83-dB SFDR 10-MHz Bandwidth A 83-dB SFDR 10-MHz Bandwidth Continuous-Time Delta-Sigma Mdlt E l i O El t Modulator Employing a One-Element- Shifting Dynamic Element Matching Hong Phuc Ninh , Masaya Miyahara, and Akira Matsuzawa Akira Matsuzawa Tokyo Institute of Technology, Japan Matsuzawa & Okada Lab. Matsuzawa & Okada Lab.

A83A 83-dB SFDR 10dB SFDR 10-MHz BandwidthMHz Bandwidth

  • Upload
    others

  • View
    8

  • Download
    0

Embed Size (px)

Citation preview

A 83-dB SFDR 10-MHz BandwidthA 83-dB SFDR 10-MHz Bandwidth Continuous-Time Delta-Sigma

M d l t E l i O El tModulator Employing a One-Element-Shifting Dynamic Element Matching

Hong Phuc Ninh, Masaya Miyahara, and Akira MatsuzawaAkira Matsuzawa

Tokyo Institute of Technology, Japan

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.

1Outline

• BackgroundBackground• Proposed one-element-shifting

(OES) DEM method(OES) DEM method• Implementation and measurement

ltresults• Conclusion

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

2Outline

• BackgroundBackground• Proposed one-element-shifting

(OES) DEM method(OES) DEM method• Implementation and measurement

ltresults• Conclusion

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

3Receiver architectureTV tuner2G/3G cellularWLAN …

*10 MHz bandwidth (our target design)*High Dynamic Range (DR)*High Spurious-Free Dynamic Range (SFDR)

ΣΔ ADC is a hopeful solution

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

to achieve high DR & SFDR

4ΣΔ ADC architecture

N: Quantizer resolution

L=4

N: Quantizer resolutionOSR: oversampling ratio (=Fs/2/BW)L: filter order

L=3

L=4

N=1,2,3,4150

of F

sContinuous-time ΣΔ ADC with multi-bit quantizer & DAC

SNR

L=1

L=2100

DAC linearity i iat

ion

oOur design

S L=150 is an issue

Low resolutionLi

mita

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

1 10 1000

OSR

L

5

Ci

Linearity issues of feedback DACI t t f ΣΔ ADC

VipRin

Cin

Vom

V+ -N

Input stage of ΣΔ ADCUnity cell

Rin

Cin

VopVim - +P

Static error inStatic errorMismatch

Inp<

0>In

n<0>

Inp<

1>In

n<1>

Inp<

7>In

n<7>

1 05 0 981 02Ex: 1.05 0.981.02Ex:

Mismatch deviation Transitor size

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

For simplicity, a 3bit DAC is considered

6Linearity issues of feedback DACCiI t t f ΣΔ ADC

VipRin

Cin

Vom

V+ -N

Input stage of ΣΔ ADC

Dynamic error

Rin

Cin

VopVim - +P

Dynamic error Glitch

in

Inp<

0>In

n<0>

Inp<

1>In

n<1>

Inp<

7>In

n<7>

P iti itNormalized glitch energy

Parasitic capacitanceNon-ideal switching

F hi h d ti d iMatsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

For high speed operation, dynamic error becomes more critical

7What is glitch energy?Switching asymmetry

I (uA)

T (ns)

Glitch energy(Glitch area)

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

Glitch energy: average of 8192points

8Requirements of DAC linearityStatic errorMismatch

Dynamic error Glitch

80

90

60

70

40

50

60

0.3% 1.6%0 1 2 3

404 5

R i t f SNR 70dB

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

Requirement for SNR>70dB(BW=10MHz, Fs=500MHz)

9Outline

• MotivationMotivation• Proposed one-element-shifting

(OES) DEM method(OES) DEM method• Implementation and measurement

ltresults• Conclusion

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

10DEM topology summaryProp.

OESDWA-group

(ADWA, Bi-DWA)TC-group(RTC, RSTC)

Glitch Good Bad ExcellentMismatch Good Excellent BadMismatch Good Excellent Bad

(DEM: to improve DAC linearity)*Data Weighted Averaging (DWA) [1]*Advanced Random DWA (ADWA) [2]*Bi-directional DWA (Bi-DWA) [3]*Th t C di (TC / DEM)*Thermometer Coding (TC, w/o DEM)*Randomized Thermometer Coding (RTC) [4]*Restricted Swapping Thermometer Coding (RSTC) [5][1] R T Baird et al IEEE Trans Circuits Syst II Dec 1995

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

[1] R. T. Baird et al., IEEE Trans. Circuits Syst. II,, Dec. 1995.[2] I. Fujimori et al., IEEE J. Solid-State Circuits, Dec. 2000.[3] D. H. Lee et al., IEEE Trans. Circuits Syst. II, Oct. 2007.[4] D. H. Lee et al., IEEE Trans. Circuits Syst. II, Feb. 2009.[5] M. H. Shen et al., IEEE Trans. Circuits Syst. II, May. 2010.

11OES: Eliminating Effect of GlitchBy reducing the number of switched elements g(n)(w/ same other glitch conditions)

g(n) Glitch energy

⎩⎨⎧

>−+−−−≤−+−+

=NnxnxnxnxN

Nnxnxnxnxng

)1()(),1()(2)1()(),1()(

)(⎩⎨⎧

−<−−

−≥−−+=

)1()(,)1()()1()(),1()(2

)(nxnxnxnx

nxnxnxnxng

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

)1()()( −−= nxnxng

12Glitch Energyg(n) Glitch energy

Sim condition*tfb tf trb tr 20ps*tfb-tf=trb-tr=20ps*Cp=10fF*3bit DAC (w/o mismatch)( )

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

Requirement for SNR>70dB(BW=10MHz, Fs=500MHz)

13OES: Preserve Reduction of Mismatch Effect By reducing the mismatch error spectrum in the interesting bandwidth(w/ same mismatch deviation)(w/ same mismatch deviation)

OES(Good)

ADWA(Excellent)

RSTC(Bad)

pect

rum

[dB

]

-40

-20

0

pect

rum

[dB

]

-40

-20

0

pect

rum

[dB

]

(Good) (Excellent) (Bad)A

C m

ism

atch

sp

120

-100

-80

-60

AC

mis

mat

ch sp

120

-100

-80

-60

AC

mis

mat

ch sp

DA

Frequency [MHz]50 100 150 200 250

-120

DA 0

Frequency [MHz]50 100 150 200 2500

-120

DA

(1%mismatch input: 1MHz@ 30dBFS)

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

(1%mismatch, input: 1MHz@-30dBFS)

14Mismatch requirementMismatch DAC area

MismatchSim condition*tfb tf trb tr 0psMismatch

Relaxation*tfb-tf=trb-tr=0ps*Cp=10fF*3bit DAC (w/o glitch)( g )

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

Requirement for SNR>70dB(BW=10MHz, Fs=500MHz)

15With Both of Glitch and Mismatch 90

80OES

w/o DEMw/glitch w/glitch

70RTCRSTC

50

60

ADWA

Bi-DWA

OES achieves better SNDR & SFDR

50

mismatch

0 1 2 3

OES achieves better SNDR & SFDRperformance over the published DEM methods

Sim condition

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

Sim condition*tfb-tf=trb-tr=20ps *Cp=10fF*3bit DAC (w/ mismatch)

16Outline

• BackgroundBackground• Proposed one-element-shifting

(OES) DEM method(OES) DEM method• Implementation and measurement

ltresults• Conclusion

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

17System architecture

M d l t SModulator Spec

FF+FB, 3rd order4bit AD/DABW: 10MHzFs: 500MHzFs: 500MHzSNDRreq: 70dB90nm CMOS process

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

18OES DEM architecture

OES DEM

Example for 4 elements DAC

OES DEM*Simplicity (no extra

i t i t )pointer, no register)*Relax timing requirementf f db k DAC

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

for feedback DAC

19Modulator layout

OES DEM Core area: 9%

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

Power consumption: 6%

20Measurement Results

w/o DEM

OES DEM

Remove by digital filter

OES-DEMg

BW

W/o DEM OES DEMSNDR 62.8 63.3

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

SFDR 71.8 82.6

21Measurement ResultsSNDR-w/o DEMSFDR-w/o DEMSNDR-OES DEMSFDR-OES DEM

Average of 10dB SFDR improvement are achieved

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

22Performance ComparisonUnit This work [6] [7] [8]

Type/ DEM CT/OES CT/DWA DT/DEM CT/DWAB d idth MH 10 20 5 10Bandwidth MHz 10 20 5 10Samp. freq. MHz 500 640 80 300SFDR dB 83 77* 85 64*

SNDR dB 65 63.9 75.4 62.5DR dB 66 68 - 70.2P W 15 7 58 36 5 31Power mW 15.7 58 36 5.31CMOS proc. nm 90 130 180 110FoM fJ/conv 530 1130 750 240

*Better SFDR (compared with conv. DEM method)*Less power (w/ same SFDR)

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

Less power (w/ same SFDR)[6]J. G. Jo et al., ASSCC Dig. Tech. Papers, Nov. 2010.[7]O. Rajaee et al., IEEE J. Solid-State Circuits, Apr. 2010.[8]K. Matsukawa et al., IEEE Symp. on VLSI Circuits, Jun. 2009.

23Outline

• BackgroundBackground• Proposed one-element-shifting

(OES) DEM method(OES) DEM method• Implementation and measurement

ltresults• Conclusion

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

24Conclusion

• Proposed OES DEM method substantially suppresses the both effects of mismatchsuppresses the both effects of mismatch and glitch.

• ΣΔ modulator using OES DEM achieves• ΣΔ modulator using OES DEM achieves 83dB SFDR and 10dB improvementcompared to no DEMcompared to no DEM.

• Simplicity and effectiveness of the OES technique makes it very attractive andtechnique makes it very attractive and prefer for cost and power considerations.

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

25Acknowledgments

This work was supported by CREST, JST, pp yVLSI Design and Education Center (VDEC), the University of Tokyo in collaboration with y yCadence Design Systems. The authors also acknowledge Berkeley Design Automation for g y gthe use of the Analog FastSPICE (AFS) Platform.

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

26Reference[1] R. T. Baird and T. S. Fiez, “Linearity enhancement of multibit ∆Σ A/D and D/A converters using

data weighted averaging,” IEEE Trans. Circuits Syst. II, Analog Digit. Signal Process., vol. 42, no. 12, pp. 753–762, Dec. 1995.

[2] I. Fujimori, L. Longo, A. Hairapetian, K. Seiyama, S. Kosic, J. Cao, and S. L. Chan, “A 90-dB SNR 2.5-MHz output-rate ADC using cascaded multibit delta-sigma modulation at 8X oversampling ratio,” IEEE J. Solid-State Circuits, vol. 35, no. 12, pp. 1820–1828, Dec. 2000.

[3] D. H. Lee, and T. H. Kuo, “Advancing data weighted averaging technique for multi-bit sigma–delta modulators ,” IEEE Trans. Circuits Syst. II, Exp. Briefs, vol. 54, no. 10, pp. 838–842, Oct. 2007.

[4] D. H. Lee, T. H. Kuo, and K. L. Wen, “Low-cost 14-bit current-steering DAC with a randomized thermometer-coding method,” IEEE Trans. Circuits Syst. II, Exp. Briefs, vol. 56, no. 2, pp. 137–141 Feb 2009141, Feb. 2009.

[5] M. H. Shen, J. H. Tsai, and P. C. Huang, “Random swapping dynamic element matching technique for glitch energy minimization in current-steering DAC,” IEEE Trans. Circuits Syst. II, Exp. Briefs, vol. 57, no. 5, pp. 369–373, May. 2010.

[6] J. G. Jo, J. Noh, and C. Yoo, “A 20MHz bandwidth continuous-time Σ∆ modulator with jitter immunity improved full-clock period SCR (FSCR) DAC and high speed DWA ” ASSCC Dig Techimmunity improved full-clock period SCR (FSCR) DAC and high speed DWA, ASSCC Dig. Tech. Papers, pp. 1-4, Nov. 2010.

[7] O. Rajaee, T. Musah, N. Maghari, S. Takeuchi, M. Aniya, K. Hamashita, and U. K. Moon, “Design of a 79 dB 80 MHz 8X-OSR Hybrid Delta-Sigma Pipelined ADC,” IEEE J. Solid-State Circuits, Vol. 45, No. 4, pp. 719-730, Apr. 2010.

[8] K. Matsukawa, Y. Mitani, M. Takayama, K. Obata, S. Dosho and A. Matsuzawa, “A 5th-Order[8] K. Matsukawa, Y. Mitani, M. Takayama, K. Obata, S. Dosho and A. Matsuzawa, A 5th Order Delta-Sigma Modulator with Single-Opamp Resonator,” IEEE Symp. on VLSI Circuits, pp. 68-69, Jun. 2009.

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.

27

Thank you!!!

Matsuzawa& Okada Lab.Matsuzawa& Okada Lab.2011/11/30 H.P. Ninh, Tokyo Tech.