75
System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4. Prakash Easwaran, C Srinivasan Cosmic Circuits

System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

Embed Size (px)

Citation preview

Page 1: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

System Aspects of ADC Design

SHANTHI PAVAN

Assistant Professor

Department of Electrical Engineering

Indian Institute of Technology, Madras

05/18/01 V4.3

Prakash Easwaran, C Srinivasan

Cosmic Circuits

Page 2: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

A-to-D Converters Terminology & Architectures

05/18/01 V4.3

Page 3: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

PRESENTATION OVERVIEW

• ADC System Overview

• ADC Metrics

• Flash & Folding Converters

• Two Step Flash Converters

• Pipeline & Delta-Sigma ADCs

• Power efficiency of ADCs

• Conclusions

Page 4: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

A GENERIC MIXED-SIGNAL SYSTEM

Page 5: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

THE A-D CONVERSION PROCESS

• Anti-alias filter limits input bandwidth to fs/2

Page 6: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

ADC OPERATIONS : SAMPLING

• Input typically stored

as charge on a

capacitor

• Tracking bandwidth

• Aperture

• Hold pedestal

Page 7: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

QUANTIZER BASICS

• Quant. error assumptions– uniformly distributed– uncorrelated with input !– white !

][2

,2

eq

12

22 eq

0eq

Page 8: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

OFFSET ERROR

• Benign when relative accuracy is desired

- Cancelled digitally

Ideal

Actual

Page 9: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

GAIN ERROR

• Benign when relative accuracy is desired

-Correct using AGC- in the analog/digital domains

Ideal

Actual

Page 10: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

DIFFERENTIAL NONLINEARITY (DNL)

• Nonmonotonicity & missing codes • Monotonic if |DNL| <

Page 11: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

INTEGRAL NONLINEARITY (INL)

• Measures deviation from a line• |INL| < 0.5 sufficient condition for a monotonic characteristic

Page 12: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

DNL & INL REMARKS

• DNL & INL should be measured with the best fit line

for good repeatability

• DNL - picture of local variations in quantizer

thresholds

• INL - picture of long range variations in quantizer

thresholds

Page 13: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

DYNAMIC PERFORMANCE METRICS

• Signal to Noise Ratio (SNR)– Signal to quantization noise

• (6N + 1.76) dB for a sine wave • 1/2 the step size means ¼ the noise power

• Signal to Noise + Distortion Ratio (SNDR)– Signal to everything else

• Spurious Free Dynamic Range (SFDR)– Signal to the largest spectral spur

• Effective Number of Bits (ENOB)–

02.6

76.1SNDR

Page 14: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

SPURIOUS-FREE DYNAMIC RANGE

Input Tone

Quantization Noise

Page 15: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

HARMONIC DISTORTION

Distortion is related to INL

Page 16: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

SFDR DEPENDENCE ON “N”

• Amplitude is • Frequency ~ f

• Amplitude is • Frequency ~ 2f

¼ the power over twice as many harmonicsPeak harmonic goes down by ¼ ½ = 9 dB SFDR ~ 9N dB

Page 17: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

FLASH A-D CONVERSION

• (+) Parallel technique - low latency

• (+) References – resistor ladder

• (-) Complexity - O(2N)

• (-) Excessive power/area for N > 6

Page 18: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

THE CLOCK SKEW ISSUE

• Sampling is distributed

• Problem : Clock skew causes different comparators to sample inputs at different instances

• Result : Poor performance at high input frequencies

• Solution : Make all comparator inputs see “held” inputs

Page 19: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

PRACTICAL FLASH ADC

• T/H for good

dynamic

performance.

• Offset correction

in comparators.

Page 20: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

BACKEND LOGIC

Page 21: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

THE FOLDING ADC PRINCIPLE

• Motivation : Reduce latches & back end logic

Page 22: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

WHITHER FLASH & FOLDING ?

• Disk drive read-channels• Low precision (6 bits)• Very high speed (Gbps)• Need very low latency (for timing recovery)

Page 23: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

TWO STEP FLASH ADC

• Motivation: – Reduce number of comparators in a flash ADC

• Idea:– In a flash ADC, only comparators “near” the input give useful information– Use a coarse ADC to estimate where the signal is, then use a fine ADC placed “around” the coarse estimate for better accuracy

Page 24: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

TWO STEP FLASH ADC

• Number of comparators : (2Nc + 2Nf – 2) • Resolution Nc + Nf bits• ADCs & DAC must be good to (Nc + Nf) bits

Page 25: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

RESIDUE PLOT : A CLOSER LOOK

Slope = 1 ADC thresholds

Code : -2 -1 0 1

Page 26: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

RESIDUE PLOT : A CLOSER LOOK

Code : -2 -1 0 1

DAC LSB

]DAC )5.0(Code [ V V LSBADCinr

]DAC )5.0(Code [ V V LSBADCrin

Page 27: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

ISSUE : ADC THRESHOLD OFFSET

Fine ADC Range

Fine ADC overload !

Page 28: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

ADC THRESHOLD OFFSET : FIX

• Extend the range of fine ADC by adding extra levels• Redundancy in fine ADC relaxes coarse ADC errors• Often called “Digital Error Correction”

Page 29: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

ISSUE : DAC INACCURACY

)](DAC )5.0(Code [ V V LSBADCinr

Slope = 1 DAC level error

Fine ADC range not exercised

• DAC level too small : Missing codes• DAC level too large : Non-monotonicity

Page 30: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

TWO-STEP FLASH SUMMARY

• Example : – 10-bit flash needs 1023 comparators– (5 + 5) bit 2-step flash needs only 62 comparators

• Lower hardware compared to a flash

• More latency (2 conversions, pipelined)

• Used for ~8-10 bit resolutions

• Coarse ADC resolution can be poor if extra levels are used in

the fine ADC

• DAC must be accurate to the resolution of the entire ADC

Page 31: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

IMPROVED TWO-STEP FLASH ADC

• The fine ADC operates on a small input– Offset requirements for the fine ADC can be relaxed if the

input signal swing was larger– Amplify the input to the fine ADC

Page 32: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

PIPELINE ADC PRINCIPLE

Recursive implementation of the fine ADC

Page 33: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

PIPELINE ADC PRINCIPLE

• Nc bits per stage

• Issue with ADC threshold error

- Digital error correction • DAC & Interstage amplifier implemented with

switch-capacitor circuitry

STAGE 1 STAGE 2

Page 34: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

OVERSAMPLING A-D CONVERSION

Oversampling ADC

)(2/

OSRRationgOversamplif

f

s

sig

Signal

Quantization Noise

Nyquist rate ADC

Page 35: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

OVERSAMPLING & NOISE SHAPING IDEA

• Quantization noise is pushed out of the signal band• Digital filter required to eliminate out of band noise• Very high SQNR possible with a poor quantizer• Also referred to as ADCs

Page 36: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

MODULATOR BLOCK DIAGRAM

• Discrete time input & output• Quantization modeled as additive noise

Page 37: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

MODULATOR BLOCK DIAGRAM

Noise Transfer Fn. (NTF)Signal Transfer Fn. (STF)

)()(1

1)(

1)(

)()( zE

zLzV

zL

zLzV inout

Make L(z) large in the signal band

Page 38: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

DELTA-SIGMA WAVEFORM EXAMPLE

Analog Input

Quantizer Output

Page 39: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

OUTPUT SPECTRUM

Signal Bandwidth

Signal Tone

Shaped Quantization Noise

Page 40: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

OTHER LOW SPEED ARCHITECTURES

• Algorithmic ADCs– Reuse a single stage of the pipeline

• Successive Approximation ADCs

• Dual-slope ADCs

Page 41: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

HIGH SPEED : “SYSTEM” IDEAS

• Time – interleaving

multiple ADCs– N slow converters to

get one fast ADC– Sample and hold of

each converter must be good up to fs/2

– Gain & Offset mismatch– N times speed for N

times power

Page 42: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

ADC FIGURE OF MERIT

• FOM Units : Joules per level resolved

• Small FOM means more power efficiency

• Higher speed – more power

• Higher resolution – more power

• ENOB : Effective number of bits for a Nyquist input

sfENOB2

Power FOM

Page 43: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

FIGURE OF MERIT : COMMENTS

• For the same ADC spec, FOM usually improves with

technology (& skill of the designer !)

• Ponder this : If “X” can design a 10-bit 500 Msps ADC in a

certain process with 400 mW, does this mean that he/she can

design a 10-bit, 1Gsps ADC with 800 mW ?

sfENOB2

Power FOM

Page 44: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

POWER EFFICIENCY OF ADCs

• Flash ADCs are least efficient– But unavoidable when latency cannot be

tolerated

• Pipelines more efficient than a flash– Reduced hardware at the expense of

latency

• Delta-Sigma more efficient than a pipeline– but only suitable for low bandwidth signals

Page 45: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

REFERENCES

• M. Gustavsson, J. Wikner & N. Tan, “CMOS Data Converters for Communications”, Kluwer, 2000

• R. van de Plassche, “Integrated Analog-to-Digital & Digital-to-Analog Converters”, Kluwer, 1994

• R. Schreier, S. Norsworthy & G. Temes, “ Delta-Sigma Data Converters – Principles, Design & Applications”, Wiley, 1998

Page 46: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

A-to-D Converters Specifications from

System Aspects

05/18/01 V4.3

Page 47: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

ADC Parameters

• Sampling Frequency

• Resolution

• SFDR

• Input Bandwidth

• Latency

Page 48: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

System Aspects…

• Signal Bandwidth

• System SNR requirements

• Signal characteristics– Dynamic Range– Interferers present– Single/Multi Carrier

Page 49: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

Typical Real world signals

ADC requirements derived from Strengths of the Signal

and interferers.

-114dBm Noise

-104dBm

-14dBm

-40dBm

Blocker

Weak Signal

Page 50: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

Sampling Frequency

Page 51: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

Filters: DSL

30KHz 130KHz 1.1MHz

Upstream Downstream

FDM filters: Filters to filter-off the local-echo in a Frequency-Division-Multiplexed system

Page 52: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

Filtering: Wireless signals

Filter to remove adjacent channel and other interferers.

-114dBm Noise

-104dBm

-40dBm

Weak Signal

Page 53: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

Sampling Freq - Filtering

• Ideal anti-aliasing filter placed before an ADC – infinite attenuation for frequencies above

the cutoff frequency.

– sampling at fs=2fmax with no aliasing

• Very High order filters

– Typically > 8th order filters !!

Page 54: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

High order filter

• High sensitivity, tougher to implement

• Process variation: requires calibration of filter

• High Power and Silicon area

• Non-linear phase response

• Effective solution is digital filters for anti-aliasing.

Page 55: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

Receive channel

• Oversample the ADC– Simpler Lower order Analog filter– Anti aliasing done by the digital filter

• Filter –ADC complexity trade off

• ADC up in the signal chain– Signal processing in Digital domain

ADC

Digital filter Analog Filter ReceiverA/D Converter

Page 56: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

Resolution

Page 57: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

Resolution Specs

• SNR requirement

• Dynamic range of the input signal

• Signal characteristics:– Peak to Average (crest factor)

• Interferers present

Page 58: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

ADC Requirement for DSL

SNR for QAM Modulation

(3dB/bit; 15 bits)

45dB

Crest Factor for clipping < 10-8 15dB

Dynamic Range lost due to Echo 0dB

Margin from ADC 10dB

70dB

System optimisation

Page 59: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

Filter - ADC trade off

-114dBm Noise

-40dBm

Weak Signal

-90dBm

• Interferer higher power than the signal.

• For the signal to use the dynamic range of the ADC– Sufficient attenuation of the interferer ( < signal power)– AGC to gain the signal post filtering

• High order filter

-114dBm Noise

-40dBm

-90dBm

Page 60: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

Filter - ADC trade off

-114dBm Noise

-40dBm

Weak Signal

-90dBm

• Trade off between filter order and dynamic range lost

in the ADC

-114dBm Noise

-40dBm

Weak Signal

-90dBm

Page 61: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

SFDR

Page 62: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

SFDR

• Caused by Harmonics and Intermodulation

products

Page 63: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

SFDR

• Important parameter when digitizing

broadband multicarrier signals

• Desired signal then is obtained by using a

narrowband digital bandpass filter

• The SNR is improved by this digital-filtering

process – Advantage of oversampling

Page 64: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

SFDR Performance of the ADC

• a spurious component may fall within the bandwidth of the digital filter

• SFDR: noise added in-band– does not improve with digital-filtering.

• SFDR spec > than the SNR after the digital filtering

• Example: 15b 65MSPS ADC with – SNR = 73dB– SFDR = 85dB

Page 65: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

Frequency

Power

Spectrum

Missing Tone Test

180KHz 1104KHz

SFDR in a multi-tone system (DSL)

Page 66: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

Input Bandwidth

• Critical parameter for systems that use

undersampling.

• Direct sampling of the IF signal in a wireless

system– Eg: 14b 125MSPS ADC sampling a

200MHz IF

• Jitter performance of the sampling clock for

this high input frequency..

Page 67: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

Latency

• systems that use the ADC in control loops

• In communications– ADC used for AGC, power control

Page 68: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

ADC: Bits vs. Speed Chart

5 10 15 20

1MHz

10MHz

100MHz

500MHz

1GHz

AD

C S

pee

d

Resolution (Bits)

Flash/Folding

Pipeline

Sigma-Delta

Basestations, Graphics

WLAN, DSL

Bluetooth

UWB, Disk Drives, Instrumentation

Page 69: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

Pipelined A/D Converters

Page 70: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

PIPELINE ADC PRINCIPLE - RECAP

Page 71: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

Advances in ADC Speed

ADC Speed for 14-bit Resolution

100

150

200

2002 2003 2005

Year

Sp

eed

(M

SP

S)

• ADC speed increase driven by desire to bring ADC closer to antenna

Page 72: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

Advances in Power Dissipation

ADC Power for 14-bit 105MSPS ADC

0

500

1000

1500

2000

2002 2004 2005

Year

Po

wer

(m

W)

• ADC power reduction driven by portability, SOC integration

• Near 2X reduction every 18 months

Page 73: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

Continuous-time A/D Converters

Page 74: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

C-T ADC: BASICS

• Architecture trades off speed for resolution; Ideal for applications where signal bandwidth is relatively small (example: Bluetooth, DSL, DVB-H)

• Signal sampled by the internal ADC free anti-alias filtering by the loop filter

• Important limitations are SNR degradation due to clock jitter (DAC) and stability concerns in higher-order loops due to excess delay (loop filter + internal quantizer)

ADCLoop Filter

H(s)

DAC

gm

vin bits

Page 75: System Aspects of ADC Design SHANTHI PAVAN Assistant Professor Department of Electrical Engineering Indian Institute of Technology, Madras 05/18/01 V4.3

Advances in Power Dissipation

ADC Power for 12-bit ADC

0

50

100

150

2003 2005

Year

Po

wer

(m

W)

BW=40MHz; 74dB SNR

BW=10MHz; 67dB SNR

• Significant advances in the last two years; 4X improvement in speed, 5X reduction in power dissipation