Experience Seamless Full Ebook Downloads for Every Genre at textbookfull.

com

Charge-Sharing SAR ADCs for Low-Voltage Low-Power

Applications 1st Edition Taimur Rabuske

https://textbookfull.com/product/charge-sharing-sar-adcs-
for-low-voltage-low-power-applications-1st-edition-taimur-
rabuske/

OR CLICK BUTTON

DOWNLOAD NOW

Explore and download more ebook at https://textbookfull.com

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

Low-Power Millimeter Wave Transmitters for High Data Rate

Applications Khaled Khalaf

https://textbookfull.com/product/low-power-millimeter-wave-
transmitters-for-high-data-rate-applications-khaled-khalaf/

textboxfull.com

System on Chip Interfaces for Low Power Design 1st Edition

Mishra

https://textbookfull.com/product/system-on-chip-interfaces-for-low-
power-design-1st-edition-mishra/

textboxfull.com

Low Power Wireless Receivers for IoT Applications with

Multi-band Calibration Algorithms 1st Edition Michael W.
Rawlins
https://textbookfull.com/product/low-power-wireless-receivers-for-iot-
applications-with-multi-band-calibration-algorithms-1st-edition-
michael-w-rawlins/
textboxfull.com

High-Density Integrated Electrocortical Neural Interfaces:

Low-Noise Low-Power System-on-Chip Design Methodology 1st
Edition Sohmyung Ha
https://textbookfull.com/product/high-density-integrated-
electrocortical-neural-interfaces-low-noise-low-power-system-on-chip-
design-methodology-1st-edition-sohmyung-ha/
textboxfull.com
Low Power CMOS Digital Pixel Imagers for High Speed
Uncooled PbSe IR Applications 1st Edition Josep Maria
Margarit (Auth.)
https://textbookfull.com/product/low-power-cmos-digital-pixel-imagers-
for-high-speed-uncooled-pbse-ir-applications-1st-edition-josep-maria-
margarit-auth/
textboxfull.com

Multiphase Reactor Engineering for Clean and Low Carbon

Energy Applications 1st Edition Yi Cheng

https://textbookfull.com/product/multiphase-reactor-engineering-for-
clean-and-low-carbon-energy-applications-1st-edition-yi-cheng/

textboxfull.com

Low Energy Electrons Fundamentals and Applications 1st

Edition Oddur Ingólfsson (Editor)

https://textbookfull.com/product/low-energy-electrons-fundamentals-
and-applications-1st-edition-oddur-ingolfsson-editor/

textboxfull.com

Renal diet cookbook the low sodium low potassium healthy

kidney cookbook 1st Edition Press

https://textbookfull.com/product/renal-diet-cookbook-the-low-sodium-
low-potassium-healthy-kidney-cookbook-1st-edition-press/

textboxfull.com

IoT and Low-Power Wireless : Circuits, Architectures, and

Techniques First Edition Siu

https://textbookfull.com/product/iot-and-low-power-wireless-circuits-
architectures-and-techniques-first-edition-siu/

textboxfull.com
ACSP · Analog Circuits And Signal Processing

Taimur Rabuske
Jorge Fernandes

Charge-Sharing
SAR ADCs for
Low-Voltage
Low-Power
Applications
Analog Circuits and Signal Processing

Series editors
Mohammed Ismail, Dublin, USA
Mohamad Sawan, Montreal, Canada

More information about this series at http://www.springer.com/series/7381

Taimur Rabuske • Jorge Fernandes

Charge-Sharing SAR ADCs

for Low-Voltage Low-Power
Applications

123
Taimur Rabuske Jorge Fernandes
INESC-ID INESC-ID
Instituto Superior Técnico Instituto Superior Técnico
Universidade de Lisboa Universidade de Lisboa
Lisboa, Portugal Lisboa, Portugal

ISSN 1872-082X ISSN 2197-1854 (electronic)

Analog Circuits and Signal Processing
ISBN 978-3-319-39623-1 ISBN 978-3-319-39624-8 (eBook)
DOI 10.1007/978-3-319-39624-8

Library of Congress Control Number: 2016942569

© Springer International Publishing Switzerland 2017

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, express or implied, with respect to the material contained herein or for any
errors or omissions that may have been made.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG Switzerland
To our families
Preface

The integrated successive-approximation-register (SAR) analog-to-digital con-

verter (ADC) is known to present a remarkable energy efficiency. Additionally,
the SAR ADC is a very scaling-friendly architecture, due to its highly digital and
switching-intensive nature and its ability to effortlessly accommodate rail-to-rail
signals without resorting to precision amplifiers. These characteristics have been
boosting the popularity of SAR ADCs as process scaling reduces the transistor’s
intrinsic gain and supply voltages. The most basic form of an SAR ADC requires
a voltage comparator, a digital controller, a track-and-hold (TH) circuitry, and a
digital-to-analog converter (DAC).
In most of the reported designs, the DAC of an SAR ADC is capacitive and relies
on the charge-redistribution (CR) principle. Alternatively, an SAR ADC that relies
on the charge-sharing (CS) principle has been proposed more recently. The main
advantages of the CS-ADC are that it is immune to inaccuracies in the reference
buffer during the conversion and requires less demanding buffers for the input signal
and the reference voltage. Additionally, the CS-ADC allows the use of nonlinear
capacitors. On the other hand, the CS-ADC is less tolerant to noise and to the
comparator offset, which limits the competitiveness of the architecture. Finally, the
literature lacks a comprehensive analysis of the limiting factors of CS-ADCs.
The aim of this book is to fill the gap in knowledge that exists for CS-ADCs
by providing an in-depth analysis of the architecture and quantifying its limiting
factors. Finally, the insights gained through this analysis are applied to devise tech-
niques that mitigate the critical drawbacks of the architecture and allow the design
of energy-efficient SAR ADCs for low-voltage applications. These techniques are
validated through the design of two SAR ADCs that operate at supply voltages
down to 0.35 and 0.4 V. Experimental results show that the two reported ADCs
present the best energy efficiency among ADCs designed in technologies down to
0.13 µm. Additionally, one of the presented prototypes is the first demonstration
of SAR ADC that uses a DAC with very-nonlinear capacitors, benefitting from
the improved capacitance density and matching characteristics of metal-oxide-
semiconductor (MOS) capacitors.

vii
viii Preface

Much of the material presented in this monograph originates from the work
carried out by the first author for his Ph.D., at Universidade de Lisboa, Portugal.
This work includes original research results that have been presented at international
conferences (ESSCIRC, ISCAS, and ICECS, among others), published in Springer
Analog Integrated Circuits and Signal Processing, IEEE Transactions on Very Large
Scale Integration (VLSI) Systems and IEEE Journal of Solid-State Circuits.
The research reported in this book received contributions from many generous
and gifted people. The authors would like to thank all colleagues and friends at
INESC-ID Lisboa, Portugal, and UFSM, Brazil.
Finally, this work was financially supported by:
• FCT, Fundação para a Ciência e a Tecnologia (Portugal), under projects
UID/CEC/50021/2013 and DISRUPTIVE (EXCL/EEI-ELC/0261/2012)
• CNPq, Conselho Nacional de Desenvolvimento Científico e Tecnológico
(Brazil), under the Ph.D. grant 201887/2011-8

Lisboa, Portugal Taimur Rabuske

Lisboa, Portugal Jorge Fernandes
April 2016
Contents

1 Introduction .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 1
1.1 Background.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 1
1.2 Research Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 6
1.3 Book Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 6
1.4 Original Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 7
1.5 Final Remarks .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 7
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 8
2 ADCs for Low-Voltage Low-Power Applications . . . .. . . . . . . . . . . . . . . . . . . . 11
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 11
2.2 Sampling and Quantization .. . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 14
2.3 Search Methods for Nyquist ADCs . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 15
2.3.1 Direct Search (Flash ADCs). . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 17
2.3.2 Binary Search (SAR ADCs). . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 18
2.3.3 Pipelined Binary Search (Pipeline ADCs) .. . . . . . . . . . . . . . . . . . . . 19
2.3.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 21
2.4 The SAR ADC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 21
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 23
3 Review of SAR ADC Switching Schemes . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 25
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 25
3.2 Operation Modes of Charge-Based SAR ADCs . . .. . . . . . . . . . . . . . . . . . . . 26
3.2.1 The Charge Redistribution Principle .. . . . . .. . . . . . . . . . . . . . . . . . . . 26
3.2.2 The Charge Sharing Principle .. . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 27
3.3 Charge Redistribution Switching Schemes . . . . . . . .. . . . . . . . . . . . . . . . . . . . 28
3.3.1 Conventional Switching . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 28
3.3.2 Monotonic or “Set-and-Down” Switching . . . . . . . . . . . . . . . . . . . . 30
3.3.3 VCM -Based Capacitor Switching .. . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 32
3.3.4 Tri-Level Capacitor Switching . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 34
3.4 Charge Sharing Switching Scheme . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 36
3.5 Comparison of Reviewed Switching Schemes . . . .. . . . . . . . . . . . . . . . . . . . 42
3.6 State of the Art in CS-ADCs . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 46

ix
x Contents

3.6.1 Craninckx and van der Plas [8] . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 46

3.6.2 Giannini et al. [10] . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 46
3.6.3 Tsai et al. [11] .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 47
3.6.4 Malki et al. [9, 12] . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 48
3.6.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 49
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 66
4 Effects of Nonidealities on the Performance of CS-ADCs . . . . . . . . . . . . . . 69
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 69
4.2 Trajectory of the DAC Voltage in a CS-ADC. . . . . .. . . . . . . . . . . . . . . . . . . . 70
4.3 Analog-to-Digital Conversion Gain .. . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 73
4.4 Errors Caused by Mismatch . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 75
4.4.1 Mismatch Between TH Capacitors . . . . . . . .. . . . . . . . . . . . . . . . . . . . 75
4.4.2 Mismatch Between TH and Total DAC Capacitance . . . . . . . . . 80
4.4.3 Mismatch Between DAC Capacitors . . . . . .. . . . . . . . . . . . . . . . . . . . 82
4.5 Errors Caused by Parasitics . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 82
4.6 Errors Caused by Comparator Offset . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 86
4.6.1 Differential Nonlinearity . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 88
4.6.2 Integral Nonlinearity.. . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 90
4.6.3 ENOB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 92
4.6.4 Model Verification . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 97
4.6.5 Discussion .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 98
4.7 Errors Caused by Noise . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 98
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 102
5 Noise-Aware Synthesis and Optimization of Voltage
Comparators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 103
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 103
5.2 Review on Comparator Noise Calculation .. . . . . . . .. . . . . . . . . . . . . . . . . . . . 104
5.3 Comparator Noise Measurement . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 105
5.4 Multi-Objective Optimization Framework .. . . . . . . .. . . . . . . . . . . . . . . . . . . . 107
5.5 Design Example .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 107
5.6 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 108
5.7 Discussion .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 111
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 113
6 An 8-Bit 0.35-V CS-ADC with Comparator Offset
Auto-Zero and Voltage Boosting . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 115
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 115
6.2 CS-ADC with Background Comparator Offset Auto-Zeroing .. . . . . . . 116
6.2.1 Operation Principle of the Proposed CS-ADC . . . . . . . . . . . . . . . . 117
6.2.2 Background Comparator Offset Self-Zeroing . . . . . . . . . . . . . . . . . 118
6.3 SAR ADC Implementation .. . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 118
6.3.1 Self-Calibrated Comparator . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 118
6.3.2 Local Voltage Boosting .. . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 124
Contents xi

6.3.3 Full-Custom Self-Timed SAR Controller .. . . . . . . . . . . . . . . . . . . . 125

6.3.4 TH and DAC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 126
6.4 Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 127
6.5 Discussion .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 130
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 131
7 A 9-Bit 0.6-V CS-ADC with a MOSCAP-DAC . . . . . . .. . . . . . . . . . . . . . . . . . . . 133
7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 133
7.2 CS-ADC with a MOSCAP-Based DAC . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 134
7.3 SAR ADC Implementation .. . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 140
7.3.1 Comparator.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 140
7.3.2 Configurable TH with Boost-and-Bootstrap Switches .. . . . . . . 141
7.3.3 Voltage Boosters . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 142
7.3.4 DAC Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 143
7.4 Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 145
7.5 Discussion .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 150
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 153
8 Conclusions and Future Work . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 155
8.1 Final Remarks .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 155
8.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 156
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 158

Acronyms . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 159

Symbols . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 161

Index . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 163
Chapter 1
Introduction

Contents
1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Research Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.3 Book Outline. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.4 Original Contributions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.5 Final Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

1.1 Background

Before the Industrial Revolution, delivering a message to another person depended

on rudimentary communication systems, such as horseback post riders and pigeon
post. In that era, information was propagated in a speed that hardly exceeded a
few kilometers per hour. The outcomes of these lagging methods are perplexing.
For instance, the Battle of New Orleans, which was fought in January of 1815 in
Louisiana, USA, took place after the Treaty of Ghent was signed between the United
Kingdom and the USA 2 weeks before. The treaty was supposed to end the war and
hostilities between the parties, but since the message did not arrive from Europe
soon enough, more than 400 soldiers lost their life in that battle [1]. Only in 1844
these figures started to change significantly, after the invention of the telegraph and
the consequent dissemination of this technology towards different parts of the world.
The Mino–Owari earthquake, occurred in Japan in 1891, was reported to London on
only 1 day of delay, despite the almost 6000 km separating the countries. Later, the
introduction of the vacuum tube, in 1907, the transistor, in 1947, and the integrated
circuit, in 1959, enabled further advancements in electronics, that enhanced the
telecommunication systems at a similar rate. In the current days, electronics is an
omnipresent element of culture, and information crosses the globe in a fraction of a
second.

© Springer International Publishing Switzerland 2017 1

T. Rabuske, J. Fernandes, Charge-Sharing SAR ADCs for Low-Voltage
Low-Power Applications, Analog Circuits and Signal Processing,
DOI 10.1007/978-3-319-39624-8_1
2 1 Introduction

In the past five decades, scientists and engineers worked towards increasing the
level of integration of transistors in integrated circuits and reducing their fabrication
costs. In 2009, it was estimated that around one billion transistors were made for
every person in the world every year, and that rate was growing fast [2]. Back in
1969, the first integrated circuit with silicon gate from Intel, the 1101 256-bit static
random access memory (RAM) chip, comprised approximately 2000 transistors.
To put into perspective, a modern graphics processing unit (GPU), such as the
Nvidia’s GK110, uses 7.1 billion transistors in a 28-nm process. The advancements
in integration brought together improvements in the performance and cost of the
integrated circuits. A modern central processing unit (CPU) chip fabricated in
a 22-nm process runs with a clock over 4000 times faster, and each transistor
consumes about 5000 times less energy when compared to the first microprocessor,
the Intel 4004, introduced in 1971. Simultaneously, the price per transistor has
dropped by a factor of about 50,000 [3].
The unprecedented evolution of speed, power efficiency, and level of integration
in modern complementary metal-oxide-semiconductor (CMOS) technologies moti-
vates the replacement of conventional analog-domain signal processing and storage
by digital alternatives. On the other hand, the world is still analog, where the signals
appear in continuous time and with continuous amplitude. Consequently, there is a
need for interfacing circuits to bridge the two signal domains. The role is fulfilled
by the digital-to-analog converter (DAC) and the analog-to-digital converter (ADC).
The performance of these circuits is measured concerning the triad: power, speed,
and resolution. The requirements of resolution and speed for an ADC or DAC are
dictated by the system specifications and the characteristics of the signal to be
processed. Commonly, power consumption should be simultaneously minimized,
specially in battery-powered applications.
The search for improvements in the energy efficiency of an ADC is a persistent
engineering problem and is a critical issue for many of the modern technology
trends, such as sensor networks, wearable computing, and biomedical electronics.
Additionally, many of these applications consume energy that is collected from
the environment, e.g. through solar cells or radio-frequency (RF) harvesters. These
energy scavengers can provide only a limited amount of power to the circuits and,
in most of the cases, a limited voltage as well. A measure commonly employed to
quantify the energetic efficiency of an ADC is the figure-of-merit (FOM), which is
defined in (1.1), given that P is the total power drained by the ADC, effective number
of bits (ENOB) is a quantity that is related to the accuracy of the conversion, and fS
is the sampling frequency.

P
FOM D : (1.1)
2ENOB fS

Many topologies of ADCs exist, covering different applications and require-

ments. The graph in Fig. 1.1 summarizes the ADCs published from 2000 to 2015
in International Solid-State Circuits Conference (ISSCC) and Symposium on VLSI
Circuits (VLSIC), in terms of effective resolution, speed, FOM, and architecture [4].
1.1 Background 3

Fig. 1.1 Summary of ADCs published in ISSCC and VLSIC from 2000 to 2015 [4]

The plot reveals a trade-off involving power, speed, and energetic efficiency. The
fastest ADCs have lower resolutions and worse energy-efficiency; the ADCs with
higher resolutions are slower and less energy-efficient; finally, the most energy-
efficient ADCs are far from the leading-edge of speed or resolution.
The trade-off may be reasoned as follows. As the ADCs approach the leading-
edge of conversion speed, the architecture to be employed requires a higher level
of parallelism. It is more difficult to increase the resolution of highly parallel archi-
tectures, as the circuit complexity grows exponentially with resolution. Moreover,
highly parallel architectures are naturally less energy efficient, because a larger
number of components are activated simultaneously, draining more power for the
same resolution. On the other extreme, the ADCs designed towards the leading-
edge regarding resolution present lower sampling rate as they take advantage of
oversampling to minimize the quantization noise. High-resolution ADCs are, in
most of the cases, noise-limited, requiring large internal capacitances to assist in
noise filtering, ultimately increasing the power consumption. Another important
issue is that process scaling does not directly improve the noise characteristics of
integrated circuits, and therefore, the benefit of scaling in noise-limited topologies
is minor. As the supply voltage shrinks, it limits the input swing of the ADCs,
consequently reducing the available dynamic range while making the ADCs less
noise-tolerant. Hence, the resolution of modern integrated ADCs operating in the
4 1 Introduction

Fig. 1.2 SAR ADCs (including time-interleaved SAR and pipelined-SAR) versus total of ADCs
published in ISSCC and VLSIC per year since 1997 [4]

Nyquist range hardly exceeds 14 bits [5]. Fortuitously, the applications on both high-
ends, high-resolution or high-speed, concern performance much more than energy
efficiency and may tolerate an increased power consumption.
In the middle of the range in the plot of Fig. 1.1, we find the zone of moderate
resolution (8–12 bits) and moderate speed (tens of kSps to tens of MSps) ADCs.
According to the figure, these are the ADCs that provide the best energy efficiency.
In the past decade, the successive approximation register (SAR) ADC became the
dominant topology for this set of specifications while drawing a lot of attention
from the scientific community. In Fig. 1.2 it is shown the number of SAR ADCs
contrasted to the total number of ADCs published in ISSCC and VLSIC from 1997
to 2015 [4].
Before 2006, SAR ADCs appeared only occasionally in these conferences
and most of the specialized publications. Starting in that same year, the number
of published SAR ADCs grew consistently. In 2015, one-half of all the ADCs
published in ISSCC and VLSIC were SAR or used SAR ADCs as building blocks
(for example, in pipelined or time-interleaved architectures). Most of the success of
the SAR ADC comes from its scaling-friendly nature, that enables the architecture
to present outstanding energy efficiency when implemented in modern technologies.
One of the aspects that makes the SAR ADC a befitting topology for deep sub-
micron processes is that it does not resort to precision amplifiers, which are known
to perform dreadfully in these technologies due to low intrinsic transistor gain.
Moreover, since the SAR is based on switched capacitors, it benefits directly from a
faster transistor speed, as it leads to also faster-switching speed.
The majority of the SAR ADCs rely on a DAC that uses the charge-redistribution
(CR) principle [6], a very mature technique that has been successively employed
1.1 Background 5

for the past four decades. Enhanced switching schemes based on the CR operating
principle have been devised in the recent years, leading to further improvements in
the energy efficiency of the architecture. A different approach for implementing the
DAC of a SAR ADC was proposed in 2007 [7], and is known as the charge-sharing
(CS)-ADC. The CS-based topology shares all the merits of SAR ADCs, such as
requiring only a comparator as active circuit and operating following a highly digital
procedure. Similarly, the CS-ADC also shares most of the advantages of the CR
architecture, as it requires very similar circuit blocks. In some aspects, however,
the architectures diverge, and the CS-ADC presents advantages and drawbacks
when compared to the CR-ADC. For example, the CS-ADC presents an entirely
passive operation after the precharge phase, obviating or at least greatly relaxing
the reference buffer. The possibility of relaxing the buffer specifications makes the
CS-ADC a good choice of topology in applications that operate under low-voltage
supplies, where the design of the voltage buffers is burdened by the limited voltage
headroom. On the other hand, the CS-ADC is not as tolerant to the comparator
offset as the CR-ADC, requiring some sort of offset calibration in the comparator to
achieve high linearity.
Nevertheless, the CS-ADC never achieved the same popularity of the CR-based
architecture and had been employed in a very limited number of designs. Figure 1.3
shows the total number of CR- and CS-ADC designs presented in ISSCC and
VLSIC, from 2007 (the year when the topology was introduced) to 2015 [4]. In
these conferences, only four publications present CS-ADCs [7–10]. In addition,
to the author’s knowledge, there are only two other references of silicon-verified
CS-ADCs published in other conferences and journals [11] and [12] excluding the
outcomes of this book. Still, [12] reports the same design as [10].
Given the limited literature, and recognizing that all the works except for [12]
were published in conferences with limited room for theory, the amount of available
tutorial material is scarce. While most of the characteristics of the more mature

Fig. 1.3 CS and CR SAR ADC publications per year [4]

6 1 Introduction

CR-ADC have been extensively modeled and verified, including the power con-
sumption [6, 13–19], linearity [20], noise [21], and statistical properties of resolution
and yield [22], the same analyses are not available to the CS counterpart. Still,
despite the reduced number of publications on CS-ADCs, there is no indication
in literature that either of the architectures is universally superior. The CR-ADC
is much more often used, probably because of the popularity and maturity of the
architecture.

1.2 Research Goals

This book aims to fill the gap in knowledge that exists for CS-ADCs, by providing
an in-depth analysis of the architecture and modeling of its limiting factors. The
achieved understanding is exploited to improve the performance and to mitigate
some critical pitfalls of the architecture. While pushing the limits of the topology,
the performance of the CS-ADC is brought closer to that of state-of-the-art designs.
Finally, the insights gained through the research are applied to the design of energy-
efficient SAR ADCs for ultra-low voltage applications.

1.3 Book Outline

This book is organized into eight chapters. Following this introduction, Chap. 2
presents the engineering problem and shows the motivation for the design of low-
voltage low-power (LVLP) ADCs. Fundamental aspects of data conversion such
as sampling and quantization are also revisited, and the three most widely used
search methods in Nyquist-rate ADCs are reviewed. This brief review is carried
out to support the choice of the SAR topology for LVLP applications. In Chap. 3,
the most relevant switching schemes employed in SAR ADCs are reviewed and
compared. A summary shows that the CS scheme shows attractive features for
LVLP applications. In Chap. 4, the most important sources of error in CS-ADC are
analyzed and quantified. Since the performance of the voltage comparator employed
in the ADC directly and strongly affects the overall ADC performance, Chap. 5
introduces a noise-aware technique for synthesis and optimization of efficient
comparators. Chapter 6 introduces the local voltage boosting and the comparator
offset background auto-zeroing techniques. These techniques are experimentally
validated with the design of an 8-bit CS-ADC. Chapter 7 introduces a DAC
topology based on the use of a metal-oxide-semiconductor capacitor (MOSCAP)
as unitary capacitance cell and investigates its advantages in CS-ADCs over the
conventional DAC based on linear capacitors. A 9-bit SAR ADC is designed using
this DAC topology and the experimental results validate the technique. Finally,
Chap. 8 summarizes and concludes this book.
1.5 Final Remarks 7

1.4 Original Contributions

The main contributions of this book are as follows:

• Chapter 4 carries out an in-depth analysis of the limiting factors to the perfor-
mance of the CS-ADC. The impact of several nonidealities is analyzed, including
mismatch, parasitics, comparator offset, and noise.
• Chapter 5 introduces a noise-aware computational framework for synthesis
and optimization of dynamic comparators. The framework employs genetic
algorithms and simulation techniques that are traditionally used for RF circuits,
such as periodic steady-state and cyclostationary noise analysis [23];
• Chapter 6 proposes two techniques that enable implementations of CS-ADC to
present low power consumption and operate with low-voltage supplies: offset
auto-zeroing for dynamic comparators and local voltage boosting to decrease
the switches resistance. The mentioned techniques are experimentally validated
through the implementation and measurements of an 8-bit ADC prototype [24].
• Additionally, a topology of a custom-designed digital successive approximation
controller which relies on dynamic logic is proposed and validated through
simulations and experimentally. This controller is employed in the fabricated
ADC prototypes presented in this book.
• Finally, a topology of DAC for CS-ADCs that uses only MOSCAPs instead
of linear capacitors is proposed and validated. MOSCAPs provide the best
capacitance density among all the possible implementations of capacitors in
integrated circuits. Additionally, it is demonstrated in Chap. 7 that the use
of MOSCAPs greatly mitigates the drawbacks of the conventional CS-ADC,
namely susceptibility to comparator offset and decreased tolerance to noise. This
technique is validated with the design and experimental measurements (including
characterization in a temperature chamber) of a 9-bit CS-ADC prototype [25, 26].
• The two presented ADCs are, to the authors’ knowledge, the two ADCs with the
best FOM for technologies down to 0.13 µm, and the two CS-based ADCs with
the best FOM reported so far.

1.5 Final Remarks

During the research that led up to this book, a few other techniques were proposed.
However, since we were not able to experimentally validate those ideas, they were
not included in this book. In [27], a CS-ADC that utilizes an integrated step-
down voltage converter to generate the precharge voltage for the DAC is reported.
The integrated voltage conversion provides significant energy savings. In [28],
a 12-bit CS-ADC that employs background self-calibration of the DAC mismatches
is reported. The mismatches are calibrated using dynamic body-biasing. Thus, the
bodies of the MOSCAPs employed in the DAC are properly biased to compensate
for mismatches. The techniques proposed in these two publications were validated
through circuit-level simulations.
8 1 Introduction

References

1. G. Clark, A Farewell to Alms: A Brief Economic History of the World. The Princeton
Economic History of the Western World (Princeton University Press, Princeton, 2007).
ISBN:9781400827817 (cit. on p. 1)
2. S. Franssila, Introduction to Microfabrication, 2nd edn. (Wiley, New York, 2010). ISBN:
9781119991892. doi:10.1002/9781119990413 (cit. on p. 2)
3. Intel Corporation, Fun facts. October 2014 (Online). Available: http://www.intel.com/content/
dam/www/public/us/en/documents/corporate-information/history-moores-law-fun-facts-
factsheet.pdf (cit. on p. 2)
4. B. Murmann, ADC performance survey 1997–2015. (Online). Available: http://web.stanford.
edu/~murmann/adcsurvey.html (visited on 02/2016) (cit. on pp. 2, 3, 4, and 5)
5. W. Liu, Low-power high-performance SAR ADC design with digital calibration techniques.
Ph.D. thesis, University of Illinois at Urbana-Champaign, 2010 (cit. on p. 4)
6. J. McCreary, P. Gray, All-MOS charge redistribution analog-to-digital conversion techniques.
I. IEEE J. Solid State Circuits 10(6), 371–379 (1975). doi:10.1109/JSSC.1975.1050629
(cit. on pp. 4 and 6)
7. J. Craninckx, G. van der Plas, A 65fJ/conversion-step 0-to-50MS/s 0- to-0.7mW 9b charge-
sharing SAR ADC in 90nm digital CMOS, in IEEE International Solid-State Circuits
Conference (ISSCC) (IEEE, New York, 2007), pp. 246–600. doi:10.1109/ISSCC.2007.373386
(cit. on p. 5)
8. V. Giannini, P. Nuzzo, V. Chironi, A. Baschirotto, G. Van der Plas, J. Craninckx, An
820w 9b 40MS/s noise-tolerant dynamic-SAR ADC in 90 nm digital CMOS, in IEEE
International Solid-State Circuits Conference (ISSCC) (IEEE, New York, 2008), pp. 238–610.
doi:10.1109/ISSCC.2008.4523145
9. J.-H. Tsai, Y.-J. Chen, M.-H. Shen, P.-C. Huang, 1-V, 8b, 40MS/s, 113W charge-recycling
SAR ADC with a 14W asynchronous controller, in IEEE Symposium on VLSI Circuits
(2011), pp. 264–265
10. B. Malki, T. Yamamoto, B. Verbruggen, P. Wambacq, J. Craninckx, A 70dB DR 10b
0-to-80MS/s current-integrating SAR ADC with adaptive dynamic range, in IEEE Inter-
national Solid-State Circuits Conference (ISSCC) (IEEE, New York, 2012), pp. 470–472.
doi:10.1109/ISSCC.2012.6177095 (cit. on p. 5)
11. Y.-J. Chen, J.-H. Tsai, M.-H. Shen, P.-C. Huang, A 1-V 8-bit 100kS/s- to-4MS/s asynchronous
SAR ADC with 46fJ/conv.-step, in International Symposium on VLSI Design, Automation and
Test (IEEE, New York, 2011), pp. 1–4. doi:10.1109/VDAT.2011.5783590 (cit. on p. 5)
12. B. Malki, T. Yamamoto, B. Verbruggen, P. Wambacq, J. Craninckx, A 70 dB DR 10b 0-to-80
MS/s current-integrating SAR ADC with adaptive dynamic range. IEEE J. Solid State Circuits
49(5), 1173–1183 (2014). doi:10.1109/JSSC.2014.2309086 (cit. on p. 5)
13. V. Hariprasath, J. Guerber, S.-H. Lee, U.-K. Moon, Merged capacitor switching based
SAR ADC with highest switching energy-efficiency. Electron. Lett. 46(9), 620 (2010).
doi:10.1049/el.2010.0706 (cit. on p. 6)
14. Y. Zhu, C.-H. Chan, U.-F. Chio, S.-W. Sin, S.-P. U, R.P. Martins, F. Maloberti, A 10-bit 100-
MS/s reference-free SAR ADC in 90 nm CMOS. IEEE J. Solid State Circuits 45(6), 1111–
1121 (2010). doi:10.1109/JSSC.2010.2048498
15. C.-C. Liu, S.-J. Chang, G.-Y. Huang, Y.-Z. Lin, A 10-bit 50-MS/s SAR ADC with a
monotonic capacitor switching procedure. IEEE J. Solid State Circuits 45(4), 731–740 (2010).
doi:10.1109/JSSC.2010.2042254
16. C. Yuan, Y. Lam, Low-energy and area-efficient tri-level switching scheme for SAR ADC.
Electron. Lett. 48(9), 482 (2012). doi:10.1049/el.2011.4001
17. L. Sun, K.-P. Pun, W.-T. Ng, Capacitive digital-to-analogue converters with least significant
bit down in differential successive approximation register ADCs. J. Eng. 1–4 (2014).
doi:10.1049/joe.2013.0219
References 9

18. X. Song, Y. Xiao, Z. Zhu, Vcm-based monotonic capacitor switching scheme for SAR ADC.
Electron. Lett. 49(5), 327–329 (2013). doi:10.1049/el.2012.3332
19. E. Rahimi, M. Yavari, Energy-efficient high-accuracy switching method for SAR ADCs.
Electron. Lett. 50(7), 499–501 (2014). doi:10.1049/el.2013.3451 (cit. on p. 6)
20. M. Saberi, R. Lotfi, K. Mafinezhad, W.A. Serdijn, Analysis of power consumption and linear-
ity in capacitive digital-to-analog converters used in successive approximation ADCs. IEEE
Trans. Circuits Syst. I Regul. Pap. 58(8), 1736–1748 (2011). doi:10.1109/TCSI.2011.2107214
(cit. on p. 6)
21. W.P. Zhang, X. Tong, Noise modeling and analysis of SAR ADCs. IEEE Trans. Very Large
Scale Integr. Syst. 23(12), 2922–2930 (2015). doi:10.1109/TVLSI.2014.2379613 (cit. on p. 6)
22. J.A. Fredenburg, M.P. Flynn, Statistical analysis of ENOB and yield in binary weighted ADCs
and DACS with random element mismatch. IEEE Trans. Circuits Syst. I Regul. Pap. 59(7),
1396–1408 (2012). doi:10.1109/TCSI.2011.2177006 (cit. on p. 6)
23. T. Rabuske, J. Fernandes, Noise-aware simulation-based sizing and optimization
of clocked comparators. Analog Integr. Circ. Sig. Process 81(3), 723–728 (2014).
doi:10.1007/s10470-014-0428-4 (cit. on p. 7)
24. T. Rabuske, F. Rabuske, J. Fernandes, C. Rodrigues, An 8-bit 0.35- V 5.04-fJ/ conversion-step
SAR ADC with background self-calibration of comparator offset. IEEE Trans. Very Large
Scale Integr. Syst. (VLSI) 23(7), 1301–1307 (2015). doi:10.1109/TVLSI.2014.2337236
(cit. on p. 7)
25. T. Rabuske, J. Fernandes, A 9-b 0.4-V charge-mode SAR ADC with 1.6-V input swing
and a MOSCAP-only DAC, in Proceedings of European Solid-State Circuits Confer-
ence (IEEE, New York, 2015), pp. 311–314. ISBN:978-1-4673-7470-5. doi:10.1109/ESS-
CIRC.2015.7313889 (cit. on p. 7)
26. T. Rabuske, J. Fernandes, A SAR ADC With a MOSCAP-DAC, IEEE J. Solid-State Circuits,
51(6), 1410–1422 (2016). doi:10.1109/JSSC.2016.2548486 (cit. on p. 7)
27. T. Rabuske, J. Fernandes, F. Rabuske, C. Rodrigues, M.B. dos Santos, A self-calibrated 10-
bit 1 MSps SAR ADC with reduced-voltage charge- sharing DAC, in IEEE International
Symposium on Circuits and Systems (ISCAS) (IEEE, New York, 2013), pp. 2452–2455.
doi:10.1109/ISCAS.2013.6572375 (cit. on p. 7)
28. T. Rabuske, J. Fernandes, A 12-bit SAR ADC with background self-calibration based on a
MOSCAP-DAC with dynamic body-biasing, in IEEE International Symposium on Circuits
and Systems (ISCAS) (IEEE, New York, 2016), pp. 1482–1485 (cit. on p. 7)
Chapter 2
ADCs for Low-Voltage Low-Power Applications

Contents
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.2 Sampling and Quantization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.3 Search Methods for Nyquist ADCs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.3.1 Direct Search (Flash ADCs). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.3.2 Binary Search (SAR ADCs). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.3.3 Pipelined Binary Search (Pipeline ADCs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.3.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.4 The SAR ADC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

2.1 Introduction

There is an assortment of emerging applications for which energy consumption is a

key metric. Implantable biomedical systems, for instance, are of keen interest in a
variety of medical areas, as they can be utilized to treat disorders such as epileptic
seizures and Parkinson’s disease [1]. These units also led to promising results in
rehabilitative prosthetic devices to help disabilities such as deafness and blindness
[2]. Another example of highly energy-constrained application is a wireless sensor
network (WSN), that relies on the inter-cooperation of many nodes, distributed
throughout an area of interest, to sense environmental parameters. This approach
to computational networking has been shown to enable a viable solution for health,
structural, industrial, military, and habitat monitoring [3–7].
Minimizing the power consumption of all the circuits used in these applications
is critical, because it is usually impractical to supply by wires the energy that these
devices need to operate. Also, the size of the units is frequently prone to serious
limitations, to allow them to be conveniently placed or implanted, consequently
restricting the volume of the batteries. As the energy storage in a battery is

© Springer International Publishing Switzerland 2017 11

T. Rabuske, J. Fernandes, Charge-Sharing SAR ADCs for Low-Voltage
Low-Power Applications, Analog Circuits and Signal Processing,
DOI 10.1007/978-3-319-39624-8_2
12 2 ADCs for Low-Voltage Low-Power Applications

Fig. 2.1 Block diagram of a

sensor node in a WSN

proportional to its volume, the available energy is very limited. Another aggravating
factor is that, in many applications, it is inconvenient to replace the battery of the
devices: in the case of medical devices, a battery replacement may require a surgical
procedure; in the case of WSNs, the nodes may be in harsh environments or be
mobile.1
For WSNs and many biomedical devices, each unit is a piece of hardware that
performs sensing, computation, and communication. They consist of sensors, a
data acquisition block, a micro-controller, radio communication circuitry, a power
management unit (PMU), and power sources. A block diagram of a WSN node
is shown in Fig. 2.1. The sensor converts an environmental parameter such as
temperature, humidity, or pH to an electrical signal delivered as voltage or current.
Then, the data acquisition block performs preprocessing and amplification on the
signal, which is finally converted to the digital form by an ADC. This conditioned
signal is processed and stored in the micro-controller. The controller also provides
some level of intelligence to the sensor node, such as time scheduling and data
compressing. Finally, the RF block allows the node to communicate with the
neighboring nodes and the base-station, depending on the network topology. The
energy sunk by these modules is provided by the energy sources and is managed by
the PMU.
Fortunately, energy harvesting technology has emerged as a promising solution
to enable self-sustainable devices. The ultimate goal in this context is that the energy
consumption is sufficiently low that can be totally harvested from the environment,
presenting a theoretical unlimited lifetime. Still, a rechargeable battery or a super-
capacitor is employed to bear the power peaks and to provide a backup source when
the environment is unable to suffice the energy required by the unit. Table 2.1
shows power densities of four forms of energy suitable to be harvested in an
approximate scale [9, 10]. The outdoor solar energy presents the highest density and
RF presents the lowest. While the values provided in Table 2.1 are valid for specific
test conditions, the environmental conditions may be subject to large variations and

1
Take, for example, the ZebraNet [8], which is a habitat monitoring system that requires zebras
to wear global positioning system (GPS) collars. The specimens would have to be re-captured for
every battery replacement.
2.1 Introduction 13

Table 2.1 Power densities for different forms of power harvesting

[9, 10]
Source Condition Power density
Solar Direct sunlight 7.5 mW/cm2
Indoor 100 µW/cm2
Thermal T D 5 ı C 60 µW/cm2
T D 10 ı C 135 µW/cm2
Vibrational 1 m/s2 (vibrations in the Hz-range) 4 µW/cm3
RF Separated 2 m from the RF transmitter 14 mW/cm2
Unless near a RF transmitter <1 µW/cm2

greatly affect the values of available power. For example, the hours of direct sunlight
may be severely reduced during the winter, and a rainy day may leave the solar
cell with no direct sunlight at all. These factors have to be accounted for while
dimensioning the power circuit of the node, and remind us that it is beneficial to
minimize the power consumption of the circuit blocks employed in the system. In
addition, most of the forms of energy harvesting generate a very limited voltage.
A typical single solar cell can output 400–500 mV, depending on lighting conditions,
while thermoelectric generators are able to generate 150–250 mV [11]. Therefore,
it is necessary that the PMU performs voltage conversion and regulation, to provide
a convenient and reliable voltage for the remaining of the circuit blocks in the
sensor node.
Some topologies of voltage-converters are most efficient when the values of the
input and output voltage are close. In that case, it is of great interest that the voltage
supply required by the circuits in the sensor node is also low, to be close to the low
output voltage of the energy harvesting sources. Still, even if the voltage-conversion
efficiency does not benefit from input/output voltages proximity, it is beneficial from
the energy standpoint to supply the circuits with low voltage. The main outcomes
of lowering the supply voltage in most of the circuits are reductions in the operation
speed and the power consumption. The former is not as critical in LVLP applications
as it is in other scenarios. This category of systems generally requires low operation
speeds, because the signals to be sensed and processed vary slowly, reaching time
constants of seconds or minutes in biomedical and environment monitoring systems
[12]. Regarding power consumption, in switching-intensive circuits (as it is the case
for CMOS digital gates and switched-capacitor circuits) the energy is proportional
to the capacitance being switched and to the square of the supply voltage, as in

EDD / Cswitching VDD 2 : (2.1)

It is also pointed out in [13] that the leakage current is strongly correlated and
decreases exponentially with the supply voltage. In other words, the usage of ultra-
14 2 ADCs for Low-Voltage Low-Power Applications

low voltage supplies helps in reducing both the static and dynamic components
of power consumption. Also, it becomes evident that, as long as the timing
requirements are fulfilled, the LVLP applications may exploit the usage of very-low
voltage supplies in order to increase the battery lifetime or to relax the harvesting
specifications. It is also advantageous that all the circuits in the system operate at the
same supply voltage, so that there is no need for the power unit to generate multiple
voltages, minimizing losses.
This scenario, summed up with the projection of unceasing decrease in the supply
voltage of leading edge fabrication technologies [14], has motivated extensive
research in LVLP designs and techniques. Many implementations of different
classes of circuits have been reported to work with extremely low supply voltages,
including digital processors [15] and memories [16], analog amplifiers [17] and
filters [18], RF front-ends [19, 20], and ADCs [21–25].
ADCs are mandatory blocks in WSN and in many biomedical implantable/-
portable systems and other applications that require LVLP operation. In the case
of WSN, the requirements for the ADCs are in most cases low sampling rate
(kSps range), moderate resolution (8–12 bits), and minimum power consumption.
This book is focused on the design of ADCs in this class. In the remaining of this
chapter, basic concepts in data conversion are revisited. The three most commonly
used search methods for ADCs operating in the Nyquist frequency range are also
reviewed, to determine which one is most befitting for the aimed specifications.

2.2 Sampling and Quantization

ADCs convert signals from the analog domain (continuous time and continuous
amplitude) into the digital domain (discrete time and discrete amplitude). Thus,
the operation of the ADC may be split up into two different processes: sampling
(time discretization) and quantization (amplitude discretization).
In the context of ADCs, sampling is a simple process that is frequently
implemented by a track-and-hold (TH) circuit. The TH provides two operation
modes that are selected according to the polarity of a clock signal , as depicted
in Fig. 2.2. While in the track mode, the TH output follows (tracks) the input signal;
when the circuit is switched to the hold mode, the output is kept steady (held), so that
the amplitude can be quantized independently of changes in the input. The period
of the clock signal that controls the TH dictates the sampling period, and its inverse
is called sampling rate or sampling frequency.
Quantization, on the other hand, is a more complicated task that usually involves
more complex circuitry. The quantizer is responsible for searching the output code
that better represents the analog sample. However, since the number of possibilities
of output codes is finite, the circuit introduces quantization error, that is the
difference between the decision level and the analog sample. In an ideal quantizer,
the distance between decision levels corresponds to 1 least-significant bit (LSB),
and the magnitude of the quantization error is always lower than or equal to half
2.3 Search Methods for Nyquist ADCs 15

a b

Fig. 2.2 Simple TH circuit: (a) implementation and (b) corresponding waveforms

LSB. Most of the practical quantizers provide a binary digital output. Thus, an ideal
B-bit quantizer divides the full input range into 2B decision levels. The sampling and
quantization processes are illustrated in Fig. 2.3, for ideal quantizers of 2, 3, and 4
bits, respectively. Note that the quantization error is always bounded to  12 LSB
and 12 LSB. However, since one LSB corresponds to a smaller range of the input
signal as the resolution increases, the quantization error is inversely proportional to
the resolution.
In sampling theory, the Nyquist frequency is defined as half of the sampling
frequency (fS =2). According to the relationship between the Nyquist frequency and
the maximum bandwidth of the input signal that the device can process, the ADCs
can be laid down in two major categories. The Nyquist-rate ADCs (frequently
abbreviated as Nyquist ADCs) are able to process signals with a bandwidth of
half the sampling frequency. The oversampled ADCs, on the other hand, can only
process signals with a bandwidth significantly lower than the Nyquist frequency as
they sacrifice input bandwidth to improve resolution and reduce noise. As a result,
the implementations of quantizers are different for Nyquist and oversampled ADCs.
In the next section, the main methods that are used in the quantizers of Nyquist
ADCs to search the correct decision level are reviewed. The oversampled ADCs are
out of the scope of this book, and will not be further investigated.

2.3 Search Methods for Nyquist ADCs

Most of the implementations of Nyquist ADCs rely on one out of the three following
methods to search for the correct output level: direct search, binary search, and
pipelined binary search.
16 2 ADCs for Low-Voltage Low-Power Applications

Fig. 2.3 Sampling and quantization processes for 3, 4, and 5-bit quantizers
2.3 Search Methods for Nyquist ADCs 17

a b

Fig. 2.4 Direct search method: (a) illustrative waveforms and (b) example of Flash ADC
implementation

2.3.1 Direct Search (Flash ADCs)

In the direct search, the proper output code is found through the simultaneous
comparison with all the decision levels covered by the ADC. Then, the comparison
results are mapped into a binary representation of the sampled input. If the input falls
between the i-th and the (iC1)-th decision level, the output is i in the digital form. If
the input lies below or above all the decision levels, the ADC output returns 00: : :0
or 11: : :1, respectively. The operation is exemplified in Fig. 2.4a. In the example,
the input voltage range of 2 V (1 to C1 V) is mapped into 16 output possibilities,
representing a 4-bit quantizer. Since the input voltage of 0.65 V is greater than the
decision level of the code 1101 (0.625 V) but smaller than the decision level of the
code 1110 (0.75 V), it is converted to the binary code 1101.
The flash ADC falls into the category of direct search ADC. A typical imple-
mentation of the flash ADC is shown in Fig. 2.4b. The TH circuit samples the
input and holds it steady during the quantization. The resistive ladder generates
the decision levels in the voltage domain, which are fed to the comparators and
act as reference voltages. Ideally, with a negligible offset voltage, the comparator
outputs form a thermometer code that indicates where the input voltage sits. This
code is transformed into a binary word by the decoder. In practice, the decoder also
includes bubble-removal logic to deal with the comparator offsets.
Since the entire conversion occurs within a single cycle, this search method
is very time-efficient and is frequently employed in high-frequency applications.
On the other hand, the high level of parallelism requires that many components
18 2 ADCs for Low-Voltage Low-Power Applications

are activated simultaneously, and that raises the power consumption significantly.
Moreover, the hardware complexity grows exponentially ( 2B ) with the number of
bits B, becoming impractical for applications that require higher resolutions.

2.3.2 Binary Search (SAR ADCs)

The binary search method uses multiple cycles to find the correct output, trading
off operation frequency for a reduction in hardware complexity. One bit is resolved
in each cycle, and this allows for the search range to be halved for the following
cycle. Figure 2.5a depicts the process for a 4-bit quantizer, using an input signal
of 0.65 V. The search range for the first cycle is the full ADC input range (1 to
C1 V), and 0 V is used as a reference to evaluate if the input sits in the upper or
lower half. Since 0.65 V is larger than 0 V, the search is bounded to the upper half in
the second cycle. Now, 0.5 V is used as a reference to decide if the input sits in the
third or fourth quarter of the full range. The method proceeds similarly for the two
following cycles, and the ADC outputs the code 1101.
The SAR ADC falls into the category of binary search ADC. A typical
implementation of the SAR ADC is shown in Fig. 2.5b. The TH circuit samples the
input and holds steady during the quantization. The SAR controller starts feeding the
DAC with the digital code 100: : :0, that sets the DAC output to half of the reference
voltage. According to the comparison results, the SAR controller adjusts the digital
word that feeds the DAC, bringing the difference between Vhold and VDAC towards
zero.

Fig. 2.5 Binary search method: (a) illustrative waveforms and (b) example implementation of the
SAR ADC
Random documents with unrelated
content Scribd suggests to you:
child, a bit of the index held between the antagonised thumb and
medius is shown. The same sign expresses both parents, with
additional explanations. To say, for instance, my mother, you would
first pantomime “I,” or, which is the same thing, my, then woman, and
finally, the symbol of parentage. My grandmother would be conveyed
in the same way, adding to the end, clasped hands, closed eyes, and
like an old woman’s bent back. The sign for brother and sister is
perhaps the prettiest; the two first finger-tips are put into the mouth,
denoting that they fed from the same breast. For the wife—squaw is
now becoming a word of reproach amongst the Indians—the dexter
forefinger is passed between the extended thumb and index of the
left.
Of course there is a sign for every weapon. The knife—scalp or
other—is shown by cutting the sinister palm with the dexter ferient
downward and towards oneself: if the cuts be made upward with the
palm downwards, meat is understood. The tomahawk, hatchet, or
axe, is denoted by chopping the left hand with the right; the sword by
the motion of drawing it: the bow by the movement of bending it, and
a spear or lance by an imitation of darting it. For the gun the dexter
thumb or fingers are flashed or scattered, i.e. thrown outwards and
upwards, to denote fire. The same movement made lower down
expresses a pistol. The arrow is expressed by knocking it upon an
imaginary bow, and by snapping with the index and medius. The
shield is shown by pointing with the index over the left shoulder
where it is slung ready to be brought over the breast when required.
The pantomime, as may be seen, is capable of expressing
detailed narratives. For instance, supposing an Indian would tell the
following tale:—“Early this morning I mounted my horse, rode off at a
gallop, traversed a ravine, then over a mountain to a plain where
there was no water, sighted bisons, followed them, killed three of
them, skinned them, packed the flesh upon my pony, remounted,
and returned home,”—he would symbolize it thus:
Touches nose—“I.”
Opens out the palms of his hand—“this morning.”
Points to east—“early.”
 Places two dexter forefingers astraddle over sinister index
—“mounted my horse.”
Moves both hands upwards and rocking-horse fashion towards
the left—“galloped.”
Passes the dexter hand right through thumb and forefinger of the
sinister, which are widely extended—“traversed a ravine.”
Closes the finger-tips high over the head and waves both palms
outwards—“over a mountain to a plain.”
Scoops up with the hand imaginary water into the mouth, and
waves the hand from the face to denote no—“where there was no
water.”
Touches eye—“sighted.”
Raises the forefingers crooked inwards on both sides of the head
—“bison.”
Smites the sinister palm downwards with the dexter first—“killed.”
Shows three fingers—“three of them.”
Scrapes the left palm with the edge of the right hand—“skinned
them.”
Places the dexter on the sinister palm and then the dexter palm
on the sinister dorsum—“packed the flesh upon my pony.”
Straddles the two forefingers on the index of the left
—“remounted.”
Finally, beckons towards self—“returned home.”
“While on the subject of savage modes of correspondence, it may
not be out of place to quote an amusing incident furnished by the
Western African traveller Hutchinson. There was, it seems, a
newspaper established in the region in question for the benefit of the
civilized inhabitants, and an old native lady having a grievance,
“writes to the editor.” Let us give her epistle, and afterwards Mr.
Hutchinson’s explanation of it:
 “To Daddy Nah, Tampin Office.
“Ha Daddy,—Do yah nah beg you tell dem people for me make
dem Sally own pussin know—Do yah. Berrah well. Ah lib nah
Pademba Road—one buoy lib dah ober side lakah dem two docta lib
overside you Tampin office. Berrah well. Dah buoy head big too much
—he say nah Militie Ban—he got one long long ting—so so brass
someting lib da dah go flip flap dem call am key. Berry well. Had dah
buoy kin blow she—ah na marnin, oh na sun time, oh na evenin, oh
nah middle night oh—all same—no make pussin sleep. Not ebry bit
dat more lib dah One Boney buoy lib overside nah he like blow bugle.
When dem two woh woh buoy blow dem ting de nize too much to
much. When white man blow dat ting and pussin sleep he kin tap wah
make dem buoy carn do so. Dem buoy kin blow ebry day, eben
Sunday dem kin blow. When ah yerry dem blow Sunday ah wish dah
bugle kin blow dem head bone inside. Do nah beg you yah tell all dem
people bout dah ting, wah dem to buoy dah blow. Tell am Amstrang
Boboh hab feber bad. Tell am Titty carn sleep nah night. Dah nize go
kill me two picken oh. Plabba done—Good by, Daddy.
“Crashey Jane.”
“For the information of those not accustomed to the Anglo-African
style of writing or speaking, I deem a commentary necessary in order
to make this epistle intelligible. The whole gist of Crashey Jane’s
complaint is against two black boys who are torturing her morning,
noon, and night—Sunday as well as every day in the week—by
blowing into some ‘long, long brass ting,’ as well as a bugle. Though
there might appear to some unbelievers a doubt as to the possibility
of the boys furnishing wind for such a lengthened performance, still
the complaint is not more extravagant than those made by many
scribbling grievance-mongers amongst ourselves about the organ
nuisance.
“The appellative Daddy is used by the Africans as expressive of
their respect as well as confidence. ‘To Daddy in the stamping (alias
printing) office,’ which is the literal rendering of the foregoing
address, contains a much more respectful appeal than ‘To the Editor’
would convey, and the words ‘Berrah well’ at the end of the first
sentence are ludicrously expressive of the writer’s having opened
the subject of complaint to her own satisfaction and of being
prepared to go on with what follows without any dread of failure.
 “The epithet ‘woh-woh’ applied to the censured boys means to
entitle them very bad; and I understand this term, which is general
over the coast, is derived from the belief that those persons to whom
it is applied have a capacity to bring double woe on all who have
dealings with them. ‘Amstrang Boboh,’ who has the fever bad, is
Robert Armstrong, the stipendiary magistrate of Sierra Leone, and
the inversion of his name in this manner is as expressive of negro
classicality as was the title of Jupiter Tonans to the dwellers on
Mount Olympus.”
It is probable that to his passion for “picture making” Mr. Catlin is
indebted for his great success among North-American children of the
wilderness. A glance through the two big volumes published by that
gentleman shows at once that he could have little time either for
eating, drinking, or sleeping; his pencil was all in all to him. No one
would suppose it by the specimens Mr. Catlin has presented to the
public, but we have his word for it, that some of the likenesses he
painted of the chiefs were marvels of perfection—so much so,
indeed, that he was almost tomahawked as a witch in consequence.
He says:
“I had trouble brewing from another source; one of the medicines
commenced howling and haranguing around my domicile amongst
the throng that was outside, proclaiming that all who were inside and
being painted were fools and would soon die, and very naturally
affecting thereby my popularity. I, however, sent for him, and called
him in the next morning when I was alone, having only the interpreter
with me, telling him that I had had my eye upon him for several days
and had been so well pleased with his looks that I had taken great
pains to find out his history, which had been explained by all as one
of a most extraordinary kind, and his character and standing in his
tribe as worthy of my particular notice; and that I had several days
since resolved, that as soon as I had practised my hand long enough
upon the others to get the stiffness out of it (after paddling my canoe
so far as I had) and make it to work easily and succesfully, I would
begin on his portrait, which I was then prepared to commence on
that day, and that I felt as if I could do him justice. He shook me by
the hand, giving me the Doctor’s grip, and beckoned me to sit down,
which I did, and we smoked a pipe together. After this was over he
told me that he had no inimical feelings towards me, although he had
been telling the chiefs that they were all fools and all would die who
had their portraits painted; that although he had set the old women
and children all crying, and even made some of the young warriors
tremble, yet he had no unfriendly feelings towards me, nor any fear
or dread of my art. ‘I know you are a good man (said he), I know you
will do no harm to any one; your medicine is great, and you are a
great medicine-man. I would like to see myself very well, and so
would all of the chiefs; but they have all been many days in this
medicine-house, and they all know me well, and they have not asked
me to come in and be made alive with paints. My friend, I am glad
that my people have told you who I am; my heart is glad; I will go to
my wigwam and eat, and in a little while I will come and you may go
to work.’ Another pipe was lit and smoked, and he got up and went
off. I prepared my canvass and palette, and whistled away the time
until twelve o’clock, before he made his appearance, having
employed the whole forepart of the day at his toilette, arranging his
dress and ornamenting his body for his picture.
“At that hour then, bedaubed and streaked with paints of various
colours, with bear’s-grease and charcoal, with medicine-pipes in his
hands, and foxes’ tails attached to his heels, entered Mah-to-he-bah
(the old bear) with a train of his profession, who seated themselves
around him, and also a number of boys whom it was requested
should remain with him, and whom I supposed it possible might have
been his pupils whom he was instructing in the mysteries of his art.
He took his position in the middle of the room, waving his evil
calumets in each hand and singing the medicine song which he
sings over his dying patient, looking me full in the face until I
completed his picture at full length. His vanity has been completely
gratified in the operation; he lies for hours together day after day in
my room in front of his picture gazing intently upon it, lights my pipe
for me while I am painting, shakes hands with me a dozen times
each day, and talks of me and enlarges upon my medicine virtues
and my talents wherever he goes, so that this new difficulty is now
removed, and instead of preaching against me he is one of my
strongest and most enthusiastic friends and aids in the country.
 “Perhaps nothing ever more completely astonished these people
than the operations of my brush. The art of portrait painting was a
subject entirely new to them and of course unthought of, and my
appearance here has commenced a new era in the arcana of
medicine or mystery. Soon after arriving here I commenced and
finished the portraits of the two principal chiefs. This was done
without having awakened the curiosity of the villagers, as they had
heard nothing of what was going on, and even the chiefs themselves
seemed to be ignorant of my designs until the pictures were
completed. No one else was admitted into my lodge during the
operation, and when finished it was exceedingly amusing to see
them mutually recognizing each other’s likeness and assuring each
other of the striking resemblance which they bore to the originals.
Both of these pressed their hand over their mouths awhile in dead
silence (a custom amongst most tribes when anything surprises
them very much); looking attentively upon the portraits and myself
and upon the palette and colours with which these unaccountable
effects had been produced.
“Then they walked up to me in the most gentle manner, taking me
in turn by the hand with a firm grip, and, with head and eyes inclined
downwards, in a tone of a little above a whisper, pronounced the
words te-ho-pe-nee Wash-ee, and walked off.
“Readers, at that moment I was christened with a new and a great
name, one by which I am now familiarly hailed and talked of in this
village, and no doubt will be as long as traditions last in this strange
community.
“That moment conferred an honour on me which you, as yet, do
not understand. I took the degree (not of Doctor of Law, nor Bachelor
of Arts) of Master of Arts—of mysteries, of magic, and of hocus
pocus. I was recognized in that short sentence as a great medicine
white man, and since that time have been regularly installed
medicine, or mystery,—which is the most honourable degree that
could be conferred upon me here, and I now hold a place amongst
the most eminent and envied personages, the doctors and conjurati
of this titled community.
 “Te-ho-pe-nee Wash-ee—pronounced ‘tup’penny’—is the name I
now go by, and it will prove to me no doubt of more value than gold,
for I have been called upon and feasted by the doctors, who are all
mystery-men, and it has been an easy and successful passport
already to many strange and mysterious places, and has put me in
possession of a vast deal of curious and interesting information
which I am sure I never should have otherwise learned. I am daily
growing in the estimation of the medicine-men and the chiefs, and by
assuming all the gravity and circumspection due from so high a
dignity (and even considerably more), and endeavouring to perform
now and then some art or trick that is unfathomable, I am in hopes of
supporting my standing until the great annual ceremony
commences, on which occasion I may possibly be allowed a seat in
the medicine lodge by the doctors, who are the sole conductors of
this great source and fountain of all priestcraft and conjuration in this
country. After I had finished the portraits of the two chiefs and they
had returned to their wigwams and deliberately seated themselves
by their respective firesides and silently smoked a pipe or two
(according to an universal custom), they gradually began to tell what
had taken place; and at length crowds of gaping listeners, with
mouths wide open, thronged their lodges, and a throng of women
and girls were about my house, and through every crack and crevice
I could see their glistening eyes which were piercing my hut in a
hundred places, from a natural and restless propensity—a curiosity
to see what was going on within. An hour or more passed in this way
and the soft and silken throng continually increased until some
hundreds of them were clung and piled about my wigwam like a
swarm of bees hanging on the front and sides of their hive. During
this time not a man made his appearance about the premises; after
awhile, however, they could be seen folded in their robes gradually
sidling up towards the lodge with a silly look upon their faces, which
confessed at once that curiosity was leading them reluctantly where
their pride checked and forbade them to go. The rush soon after
became general, and the chiefs and medicine-men took possession
of my room, placing soldiers (braves, with spears in their hands) at
the door, admitting no one but such as were allowed by the chiefs to
come in. The likenesses were instantly recognized, and many of the
gaping multitude commenced yelping; some were stamping off in the
jarring dance, others were singing, and others again were crying;
hundreds covered their mouth with their hands and were mute;
others, indignant, drove their spears frightfully into the ground, and
some threw a reddened arrow at the sun and went home to their
wigwams.
“The pictures seen, the next curiosity was to see the man who
made them, and I was called forth. Readers, if you have any
imagination, save me the trouble of painting this scene. I stepped
forth and was instantly hemmed in in the throng. Women were
gazing, and warriors and braves were offering me their hands, whilst
little boys and girls by dozens were struggling through the crowd to
touch me with the ends of their fingers, and while I was engaged
from the waist upwards in fending off the throng and shaking hands
my legs were assailed (not unlike the nibbling of little fish when I
have been standing in deep water) by children who were creeping
between the legs of the bystanders for the curiosity or honour of
touching me with the end of their finger. The eager curiosity and
expression of astonishment with which they gazed upon me plainly
showed that they looked upon me as some strange and
unaccountable being. They pronounced me the greatest medicine-
man in the world, for they said I had made a living being; they said
they could see their chief alive in two places—those that I had made
were a little alive; they could see their eyes move, could see them
smile and laugh; they could certainly speak if they should try, and
they must therefore have some life in them.
“The squaws generally agreed that they had discovered life
enough in them to render my medicine too great for the Mandans,
saying that such an operation could not be performed without taking
away from the original something of his existence, which I put in the
picture, and they could see it move, see it stir.
“This curtailing of the natural existence for the purpose of instilling
life into the secondary one they decided to be an useless and
destructive operation, and one which was calculated to do great
mischief in their happy community, and they commenced a mournful
and doleful chant against me, crying and weeping bitterly through the
village, proclaiming me a most dangerous man, one who could make
living persons by looking at them, and at the same time could, as a
matter of course, destroy life in the same way, if I chose; that my
medicine was dangerous to their lives and that I must leave the
village immediately; that bad luck would happen to those whom I
painted, and that when they died they would never sleep quiet in
their graves.
“In this way the women and some old quack medicine-men
together had succeeded in raising an opposition against me, and the
reasons they assigned were so plausible and so exactly suited for
their superstitious feelings, that they completely succeeded in
exciting fears and a general panic in the minds of a number of chiefs
who had agreed to sit for their portraits, and my operations were of
course for several days completely at a stand. A grave council was
held on the subject from day to day, and there seemed great
difficulty in deciding what was to be done with me and the dangerous
art which I was practising and which had far exceeded their original
expectations. I finally got admitted to their sacred conclave and
assured them that I was but a man like themselves, that my art had
no medicine or mystery about it, but could be learned by any of
them, if they would practice it as long as I had; that my intentions
towards them were of the most friendly kind, and that in the country
where I lived brave men never allowed their squaws to frighten them
with their foolish whims and stories. They all immediately arose,
shook me by the hand, and dressed themselves for their pictures.
After this there was no further difficulty about sitting, all were ready
to be painted; the squaws were silent, and my painting-room was a
continual resort for the chiefs and braves and medicine-men, where
they waited with impatience for the completion of each one’s picture,
that they could decide as to the likeness as it came from under the
brush, that they could laugh and yell and sing a new song, and
smoke a fresh pipe to the health and success of him who had just
been safely delivered from the hands and the mystic operation of the
white medicine.”
The Mandans celebrate the anniversary of the feast of the deluge
with great pomp. During the first four days of this religious ceremony
they perform the buffalo dances four times the first day, eight the
second, twelve the third, and sixteen the fourth day, around the great
canoe placed in the centre of the village. This canoe represents the
ark which saved the human race from the flood, and the total-
number of the dances executed is forty, in commemoration of the
forty nights during which the rain did not cease to fall upon the earth.
The dancers chosen for this occasion are eight in number and
divided into four pairs corresponding to the four cardinal points. They
are naked and painted various colours; round their ankles they wear
tufts of buffalo hair; a skin of the same animal with the head and
horns is thrown over their shoulders; the head serves as a mask to
the dancers. In one of their hands they hold a racket, in the other a
lance, or rather a long inoffensive stick. On their shoulders is bound
a bundle of branches. In dancing they stoop down towards the
ground and imitate the movements and the bellowing of buffaloes.
Alternating with these pairs is a single dancer, also naked and
painted, and wearing no other garments than a beautiful girdle and a
head-dress of eagles’ feathers mingled with the fur of the ermine.
These four dancers also carry each a racket and a stick in their
hands; in dancing they turn their backs to the great canoe. Two of
them are painted black with white spots all over their bodies to
represent the sky and stars. The two others are painted red to
represent the day, with white marks to signify the spirits chased
away by the first rays of the sun. None but these twelve individuals
dance in this ceremony of solemnity. During the dance the master of
the ceremonies stands by the great canoe and smokes in honour of
each of the cardinal points. Four old men also approach the great
canoe, and during the whole dance, which continues a quarter of an
hour, the actors sing and make all the noise possible with their
instruments, but always preserving the measure.
Besides the dancers and musicians there are other actors who
represent symbolical characters and have a peculiar dress during
this festival. Near the great canoe are two men dressed like bears
who growl continually and try to interrupt the actors. In order to
appease them women continually bring them plates of food, which
two other Indians disguised as eagles often seize and carry off into
the prairie. The bears are then chased by troops of children, naked
and painted like fawns and representing antelopes, which eagerly
devour the food that is served. This is an allegory, signifying that in
the end Providence always causes the innocent to triumph over the
wicked.
All at once on the fourth day the women begin to weep and
lament, the children cry out, the dogs bark, the men are
overwhelmed with profound despair. This is the cause: A naked man
painted of a brilliant black like the plumage of a raven and marked
with white lines, having a bear’s tusk painted at each side of his
mouth, and holding a long wand in his hand, appears on the prairie
running in a zigzag direction, but still advancing rapidly towards the
village and uttering the most terrific cries. Arriving at the place where
the dance is performing he strikes right and left at men, women, and
children, and dogs, who fly in all directions to avoid the blows of this
singular being, who is a symbol of the evil spirit.
The master of the ceremonies on perceiving the disorder quits his
post near the great canoe and goes toward the enemy with his
medicine-pipe, and the evil spirit, charmed by the magic calumet,
becomes as gentle as a child and as ashamed as a fox caught
stealing a fowl. At this sudden change the terror of the crowd
changes to laughter, and the women cease to tremble at the evil
spirit and take to pelting him with mud; he is overtaken and deprived
of his wand and is glad to take to his heels and escape from the
village as quickly as he can.
It is to be hoped that the North-American Indian when
communicating with Kitchi-Manitou does not forget to pray to be
cured of his intolerable vice of covetousness. He can let nothing odd
or valuable pass him without yearning for it, or so says every
traveller whose lot it has been to sojourn among Red men. So says
Mr. Murray, and quotes a rather ludicrous case in support of the
assertion:
“While I was sitting near my packs of goods, like an Israelite in
Monmouth Street, an elderly chief approached and signified his wish
to trade. Our squaws placed some meat before him, after which I
gave him the pipe, and in the meantime had desired my servant to
search my saddle bags, and to add to the heap of saleable articles
everything of every kind beyond what was absolutely necessary for
my covering on my return. A spare shirt, a handkerchief, and a
waistcoat were thus drafted, and among other things was a kind of
elastic flannel waistcoat made for wearing next to the skin and to be
drawn over the head as it was without buttons or any opening in
front. It was too small for me and altogether so tight and
uncomfortable, although elastic, that I determined to part with it.
 The Covetous Pawnee.
“To this last article my new customer took a great fancy and he
made me describe to him the method of putting it on and the warmth
and comfort of it when on. Be it remembered that he was a very
large corpulent man, probably weighing sixteen stone. I knew him to
be very good-natured, as I had hunted once with his son and on
returning to the lodge the father had feasted me, chatted by signs,
and taught me some of the most extraordinary Indian methods of
communication. He said he should like to try on the jacket, and as he
threw the buffalo robe off his huge shoulders I could scarcely keep
my gravity when I compared their dimensions with the garment into
which we were about to attempt their introduction. At last by dint of
great industry and care, we contrived to get him into it. In the body it
was a foot too short, and fitted him so close that every thread was
stretched to the uttermost; the sleeves reached a very little way
above his elbow. However, he looked upon his arms and person with
great complacency and elicited many smiles from the squaws at the
drollery of his attire; but as the weather was very hot he soon began
to find himself too warm and confined, and he wished to take it off
again. He moved his arms, he pulled his sleeves, he twisted and
turned himself in every direction, but in vain. The old man exerted
himself till the drops of perspiration fell from his forehead, but had I
not been there he must either have made some person cut it up or
have sat in it till this minute.
“For some time I enjoyed this scene with malicious and demure
gravity, and then I showed him that he must try and pull it off over his
head. A lad who stood by then drew it till it enveloped his nose, eyes,
mouth, and ears; his arms were raised above his head, and for some
minutes he remained in that melancholy plight, blinded, choked, and
smothered, with his hands rendered useless for the time. He rolled
about, sneezing, sputtering, and struggling, until all around him were
convulsed with laughter and our squaws shrieked in their
ungovernable mirth in a manner that I had never before witnessed.
At length I slit a piece of the edge and released the old fellow from
his straight-waistcoat confinement; he turned it round often in his
hands and made a kind of comic-grave address to it, of which I could
only gather a few words: I believe the import of them was that it
would be ‘a good creature’ in the ice-month of the village. I was so
pleased with his good humour that I gave it to him to warm his
squaw in the ‘ice-month.’”
As this will probably be the last occasion of discussing in this
volume the physical and moral characteristics of the North American
Indian, it may not be out of place here to give a brief descriptive
sketch of the chief tribes with an account of their strength and power
in bygone times and their present condition. The names of Murray,
Dominech, Catlin, etc., afford sufficient guarantee of the accuracy of
the information here supplied.
The Ojibbeway nation occupies a large amount of territory, partly
within the United States, and partly within British America. They are
the largest community of savages in North America: the entire
population, in 1842, amounted to thirty thousand. That part of the
tribe occupying territory within the United States inhabit all the
northern part of Michigan, the whole northern portion of Wisconsin
Territory, all the south shore of Lake Superior, for eight hundred
miles, the upper part of the Mississippi, and Sandy, Leech, and Red
Lakes. Those of the nation living within the British dominions occupy
all Western Canada, the north of Lake Huron, the north of Lake
Superior, the north of Lake Winnibeg, and the north of Red River
Lake, about one hundred miles. The whole extent of territory
occupied by this single nation, extends one thousand nine hundred
miles east and west, and from two to three hundred miles north and
south. There are about five thousand in British America, and twenty-
five thousand in the United States. Of their past history nothing is
known, except what may be gathered from their traditions. All the
chiefs and elder men of the tribe agree that they originally migrated
from the west. A great number of their traditions are doubtless
unworthy of credence, but a few that relate to the foundation of the
world, the subsequent disobedience of the people,—which, the
Ojibbeways say, was brought about by climbing of a vine that
connected the world of spirits with the human race, which was strictly
forbidden the mortals below, and how they were punished by the
introduction of disease and death, which before they knew not;—all
this and much more of the same nature, is a subject of more than
ordinary interest to the contemplative mind.
Their first intercourse with Europeans was in 1609, when they, as
well as many of the other tribes belonging to the Algonquin stock,
met Champlain, the adventurous French trader. They were described
by him as the most polished in manners of the northern tribes; but
depended for subsistence entirely on the chase, disdaining
altogether the more effeminate occupation of the cultivation of the
soil. From that time they eagerly sought and very soon obtained the
friendship of the French. The more so that their ancient and
inveterate foes, the Iroquois, were extremely jealous of the intrusive
white men. With the help of the French they gained many bloody and
decisive battles over the Iroquois, and considerably extended their
territories. The history of the nation from this time is not very
interesting. From the ravages of war and disease the tribe, as may
be perceived from a comparison with many others, has escaped with
more than ordinary success; partly owing to the simplicity and
general intelligence of the tribe in guarding against these evils.
Their religion is very simple, the fundamental points of which are
nearly the same as all the North American Indians. They believe in
one Ruler or Great Spirit—He-sha-mon-e-doo, “Benevolent Spirit,” or
He-ehe-mon-edoo, ”“Great Spirit.” This spirit is over the universe at
the same time, but under different names, as the “God of man,” the
“God of fish,” and many others. It is supposed by many travellers
that sun-worship was a part of their mythology, from the extreme
respect which they were observed to pay to that luminary. But we
find the reason of this supposed homage is, that the Indian regards
the sun as the wigwam of the Great Spirit, and is naturally an object
of great veneration. In this particular, perhaps, they are not greater
idolaters than civilized people, who have every advantage that art
and nature can bestow. The Indian, because the sun doesn’t shine
to-day, won’t transfer his adoration to the moon to-morrow; and in
this respect at least is superior to many a wise and educated “pale
face.”
In addition to the good spirit they have a bad spirit, whom,
however, they believe to be inferior to the good spirit. He is
supposed to have the power of inflicting all manner of evils, and,
moreover, to take a delight in doing so. This spirit was sent to them
as a punishment for their original disobediences. They have, besides
these, spirits innumerable. In their idea every little flower of the field,
every beast of the land, and every fish in the water, possesses one.
Pawnees.—This tribe, which is scattered between Kansas and
Nebraska, was at one time very numerous and powerful, but at the
present time numbers no more than about ten thousand. They have
an established reputation for daring, cunning, and dishonesty. In the
year 1832 small-pox made its appearance among the Pawnees, and
in the course of a few months destroyed fully half their numbers.
They shave the head, all but the scalp lock. They cultivate a little
Indian corn, but are passionately fond of hunting and adventure. The
use of the Indian corn is confined to the women and old men. The
warriors feed on the game they kill on the great prairies, or on
animals they steal from those who cross their territory. The Pawnees
are divided into four bands, with each a chief. Above these four
chiefs is a single one, whom the whole nation obey. This tribe has
four villages, situated near the Nebraska. It is allied with the
neighbouring tribe of the Omahas and Ottoes. It was till recently the
custom of these people to torture their prisoners, but it is now
discontinued, owing to the fact of a squaw of the hostile tribe being
snatched from the stake by a white man. The circumstance was
regarded as a direct interposition of the Great Spirit, and as an
expression of his will that torture should he discontinued. They do
not appear to possess any historical traditions, but on certain other
subjects preserve some curious legends. The “sign” of the Pawnees
is the two forefingers held at the sides of the head in imitation of a
wolf’s ears.
The Delawares.—This ancient people, once the most renowned
and powerful among American Indians, has of late years so dwindled
that were the entire nation to be gathered, it would scarcely count
one thousand souls. They are now settled in the Valley of the
Canadian river, and their pursuits are almost strictly agricultural.
According to their traditions, several centuries ago they inhabited the
western part of the American continent, but afterwards emigrated in
a body to the banks of the Mississippi, where they met the Iroquois,
who, like themselves, had abandoned the far west and settled near
the same river. In a short time, however, the new comers and the
previous holders of the land, the Allegavis, ceased to be on friendly
terms, and the combined Delawares and Iroquois declared war
against them to settle the question. The combined forces were
victorious, and divided the land of the Allegavis between them. After
living peaceably for two hundred years, another migration was
resolved upon, and, according to some accounts, the whole of both
nations, and according to others, but part of them, settled on the
shores of the four great rivers, the Delaware, the Hudson, the
Susquehanna, and the Potomac. Up to this time the Delawares
remained, as they had ever been, superior to the Iroquois, and by-
and-by the latter grew jealous of their powerful neighbours, and by
way of thinning their numbers sought to breed a deadly feud
between the Delawares and certain other near-living tribes, amongst
which were the warlike Cherokees. This was an easy matter. The
arms of every tribe are more or less peculiar and may be safely
sworn to by any other. Stealing a Delaware axe, an Iroquois lay wait
for a Cherokee, and having brained him with the weapon laid it by
the side of the scalpless body. The bait took, and speedily the
Delawares and the Cherokees were plunged into deadly strife.
 An Iroquois Warrior.
The Iroquois, however, were not destined to escape scot free for
their diabolical trick. The Delawares discovered it, and swore in
council to exterminate their malicious neighbours. But the latter were
much too wise to attempt a single-handed struggle with their justly
incensed foes, so soliciting the attention of the other tribes they set
out their grievances in so artful a manner that the others resolved to
help them, and there was straightway formed against the
unoffending Delawares a confederation called the Six Nations.
“This,” says the Abbé Dominech, “was about the end of the fifteenth
and beginning of the sixteenth century, and from this period dates
the commencement of the most bloody battles the New World has
witnessed. The Delawares were generally victorious. It was during
this war that the French landed in Canada, and the Iroquois not
wishing them to settle in the country took arms against them; but
finding themselves thus placed between two fires, and despairing of
subduing the Delawares by force of arms, they had recourse to a
stratagem in order to make peace with the latter, and induce them to
join the war against the French. Their plan was to destroy the
Delawares’ fame for military bravery, and to make them (to use an
Indian expression) into old women. To make the plan of the Iroquois
understood, we must mention that most of the wars between these
tribes are brought to an end only by the intervention of the women.
They adjure the warriors by all they hold dear to take pity on their
poor wives and on the children who weep for their fathers, to lay
aside their arms and to smoke the calumet of peace with their
enemies. These discourses rarely fail in their effect and the women
place themselves in an advantageous position as peace-makers.
The Iroquois persuaded the Delawares that it would be no disgrace
to become “women,” but that on the contrary, it would be an honour
to a nation so powerful, and which could not be suspected of
deficiency in courage or strength, to be the means of bringing about
a general peace and of preserving the Indian race from further
extermination. These representations determined the Delawares to
become “women” by asking for peace. So they came to be
contemptuously known by other tribes as “Iroquois Squaws,” and
losing heart, from that time grew more few.
Shawnees.—The ancient “hunting grounds” of this important tribe
were Pennsylvania and New Jersey; but they are now found in the
Valley of the Canadian. “Some authors are of opinion,” says the
author of “The Deserts of North America,” “that these Indians come
from Eastern Florida, because there is in that country a river called
Su-wa-nee, whence the word Shawanas, which is also used to
design the Shawnees, might be derived. It is certain, however, that
they were known on the coast of the Atlantic, near Delaware and
Chesapeak, subsequent to the historical era: that is to say, after the
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

textbookfull.com