2011 m 25 shvaichenko

29
1 IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reserved Elena Shvaichenko METHODS OF DESIGNING ANALOG-TO-DIGITAL CONVERTERS Master’s Thesis Presentation Author: Olena Shvaichenko

Upload: alberto-dan

Post on 22-Jul-2016

18 views

Category:

Documents


1 download

DESCRIPTION

WQEDFFWE

TRANSCRIPT

Page 1: 2011 M 25 Shvaichenko

1

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

METHODS OF DESIGNING ANALOG-TO-DIGITAL

CONVERTERSMaster’s Thesis Presentation

Author: Olena Shvaichenko

Page 2: 2011 M 25 Shvaichenko

2

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

The purpose is to investigate the design of

14-bit resolution ADC with sampling

frequency 2 MHz in 0.13 um technology.

Purpose

Page 3: 2011 M 25 Shvaichenko

3

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Architecture Speed Conversiontime Resolution Area Power

consumption

Flash ADC High Constant Low (up till8 bits)

Increasesexponentiallywith resolution

Very high

PipelinedADC

Medium-high

Increases withresolution

Medium-high (up till 12-bits)

Increaseslinearly withresolution

Medium

Sigma-DeltaADC Medium Trade off with

resolutionHigh (up till24-bits)

Constant; no change with increase in resolution

Medium-low

SAR ADC Medium-low

Increases withresolution

High (up till18-bits)

Increaseslinearly withresolution

Medium-low

SpecResolution

14 bit

Frequency

2 MHz

Technology

SG13S

Power supply

1.2 V

ADC architectures

Comparison table

Page 4: 2011 M 25 Shvaichenko

4

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Advantages

•Low power consumption,

•Low circuit complexity

•Mostly digital circuitry.

Limitations

•Lower sampling rates

•Accuracy of the system

depends of the accuracy of the

DAC and the comparator.

Successive approximation ADC

Page 5: 2011 M 25 Shvaichenko

5

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Principle of operation

Successive approximation ADC

Page 6: 2011 M 25 Shvaichenko

6

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Mathematical model

Page 7: 2011 M 25 Shvaichenko

7

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Mathematical model

SAR

ADC

Input Output Ideal

DAC

Histogrammethod

FFT

• ramp >10 samples per

transition level

• sine-wavecoherent sampling

Static:Gain error, Offset, INL, DNL.

Dynamic:THD, SFDR, SINAD, ENOB.

Mismatches

Offset voltage

Settling time limitations

Imperfections

Parasitic capacitance

Thermal noise

Clockfeedthrough

Jitter

Page 8: 2011 M 25 Shvaichenko

8

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Modeling of Capacitor array mismatch effect

Common centroid structure

Page 9: 2011 M 25 Shvaichenko

9

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Simulation results of modeling Capacitor array mismatch effect

0 2000 4000 6000 8000 10000 12000 14000 16000 18000-0.5

0

0.5

Digital code

INL

(LS

B)

0 2000 4000 6000 8000 10000 12000 14000 16000 18000-0.4

-0.2

0

0.2

0.4

0.6

Digital code

DN

L (L

SB

)

ADC W1=2um, L1=2um, kp=10nmIdeal ADC

ADC W1=2um, L1=2um, kp=10nmIdeal ADC

0 5 10 15

x 10-14

75

80

85 X: 6.248e-015Y: 83.45

SINAD versus value of unit capacitor

Cu, (F)

SIN

AD

, (dB

)Mismatch Cu = 6.5 fF

Page 10: 2011 M 25 Shvaichenko

10

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Modeling of Settling time effect

Sampling frequency Switches sizing.

~

Sampling phase

0 0.5 1 1.5 2 2.5 3 3.555

60

65

70

75

80

85

90

X: 2.167Y: 83.43

SINAD versus time constant τ

τ, (ns)S

INA

D, (

dB)

Page 11: 2011 M 25 Shvaichenko

11

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Behavioural model

Differential triple reference charge-redistribution SAR ADC with monotonic switching procedure

Reduces Ctot by a factor of 4 compared to conventional

Page 12: 2011 M 25 Shvaichenko

12

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

SAR Logic

Page 13: 2011 M 25 Shvaichenko

13

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Behavioural model test bench

Page 14: 2011 M 25 Shvaichenko

14

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Simulation results of behavioralmodel

Page 15: 2011 M 25 Shvaichenko

15

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Simulation results of behavioralmodel

Page 16: 2011 M 25 Shvaichenko

16

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Transistor level model

Comparator architecture

Page 17: 2011 M 25 Shvaichenko

17

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Comparator schematic

Page 18: 2011 M 25 Shvaichenko

18

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Simulation results

Input and output waveforms

Page 19: 2011 M 25 Shvaichenko

19

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Conclusion

Analysis of ADC architectures was done.

SAR architecture was chosen as most appropriate architecture to meet

given specification.

Development of program for modeling successive-approximation analog-to-digital conversion in MATLAB was done.

Analysis of non–ideal effects in SAR ADC was done.

Modeling of mismatch, input referred dc offset and settling time effects was done.

Simulation of modeling mismatch effect shows that for SG013S IHP technology it is possible to get DNL ≤ 0.5 LSB, INL < 0.5 LSB, THD = -95.07 dB, SFDR = 85.75 dB, SINAD =84.84 dB, ENOB = 13.8 bit with probability 99.7%.

Simulation of modeling comparator offset voltage effect shows that for SG013S IHP technology with 3.2 mV input referred dc offset it is possible to get THD = -84.33 dB, SFDR = 76.10 dB, SINAD =74.95 dB, ENOB = 12.16 bit with probability 99.7%.

Page 20: 2011 M 25 Shvaichenko

20

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Conclusion

Simulation of modeling settling time effect shows that for SG013S IHP technology with 2.1 ns time constant it is possible to get SFDR = 83.97 dB, SINAD =83.7 dB, ENOB = 13.61 bit with probability 99.7%.

Verilog–A behavioral model of 14-bit differential charge–redistribution SAR ADC with monotonic switching procedure was developed.

Simulation of proposed high-speed comparator with resolution 40uV, clock frequency 100 MHz, supply voltage 1.2 V in SG013S IHP technology was done.

Analysis of more than 45 scientific sources up to 2011 year in the field of analog-to-digital conversion has been conducted.

Results of investigation have been published at the International conference on system analysis and information technologies.

The investigation was carried out for IHP - Innovations for High Performance Microelectronics company.

Page 21: 2011 M 25 Shvaichenko

21

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Thank You!!!

Page 22: 2011 M 25 Shvaichenko

22

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

General concept in designing ADC

Design flow for ADC• Mathematical model

• Algorithm is examined

• Functional description

• Behavioural model• Architecture is verified

• Behavioural description

of the blocks

• Transistor level model• Schematic is verified

• Transistor level of the blocks• Layout level

Abstract model in Matlab, C/C++

Verilog/VHDL model of the digital part;

Verilog-A/VHDL- AMS model for the analog

Synthesis to get gate level Verilog or VHDL

Schematic design of the analog

Mixed verification

Analog layout

Place&Rout of the digital

Post-layout simulation

Page 23: 2011 M 25 Shvaichenko

23

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Modeling of Capacitor array mismatch effect

Page 24: 2011 M 25 Shvaichenko

24

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Simulation results of modeling Settling time effect

0 0.5 1 1.5 2

x 106

-200

-150

-100

-50

0Power spectral density

Frequency (Hz)

Pow

er/fr

eque

ncy,

(dB

/Hz)

THD = -98.90 dB

TSD = -86.98 dB

SFDR = 83.97 dB

SINAD = 83.70 dB

ENOB = 13.61 bits

ADC with τ = 2.1ns

Page 25: 2011 M 25 Shvaichenko

25

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Waveforms of conventional and monotonic switching procedure

Page 26: 2011 M 25 Shvaichenko

26

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Waveforms of conventional and monotonic switching procedure

Waveform of monotonic switching procedure

Reduces Switching power in a factor of 5

Waveform of conventional switching procedure

Page 27: 2011 M 25 Shvaichenko

27

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Preamplifiers

Preamplifier with diode connected pMOS load and it’s gain

Preamplifier with resistor load and it’s gain

Modified Bult’s preamplifier and it’s gain

Song’s preamplifier and it’s gain

Page 28: 2011 M 25 Shvaichenko

28

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Transistor level model

Preamplifier

Amplifier architecture Gain, dB Bandwidth, Hz

Differential pair with resistive load 13.88 565.7

Differential pair with pmos diode

connected load

4.86 465.2

Modified Built’s preamplifier 20.13 201.3

Song’s preamplifier 20.42 178.1

Page 29: 2011 M 25 Shvaichenko

29

IHP Im Technologiepark 25 15236 Frankfurt (Oder) Germany www.ihp-microelectronics.com © 2011 - All rights reservedElena Shvaichenko

Transistor level model

Latch

Dynamic latchAdvantages

•zero dc current in reset mode,

•full logic levels after

regeneration,

• outputs are both reset to supply

voltage so they are well defined