Meta-Smith Charts and

Their Potential Applications

Copyright © 2010 by Morgan & Claypool

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in
any form or by any means—electronic, mechanical, photocopy, recording, or any other except for brief quotations in
printed reviews, without the prior permission of the publisher.

Meta-Smith Charts and Their Potential Applications

Danai Torrungrueng
www.morganclaypool.com

ISBN: 9781608455805 paperback

ISBN: 9781608455812 ebook

DOI 10.2200/S00276ED1V01Y201009ANT010

A Publication in the Morgan & Claypool Publishers series

SYNTHESIS LECTURES ON ANTENNAS

Lecture #10
Series Editor: Constantine A. Balanis, Arizona State University
Series ISSN
Synthesis Lectures on Antennas
Print 1932-6076 Electronic 1932-6084
 Synthesis Lectures on Antennas
Editor
Constantine A. Balanis, Arizona State University
Synthesis Lectures on Antennas will publish 50- to 100-page publications on topics that include both
classic and advanced antenna configurations. Each lecture covers, for that topic, the fundamental
principles in a unified manner, develops underlying concepts needed for sequential material, and
progresses to the more advanced designs. State-of-the-art advances made in antennas are also included.
Computer software, when appropriate and available, is included for computation, visualization and
design. The authors selected to write the lectures are leading experts on the subject who have extensive
background in the theory, design and measurement of antenna characteristics.
The series is designed to meet the demands of 21st century technology and its advancements on
antenna analysis, design and measurements for engineers, scientists, technologists and engineering
managers in the fields of wireless communication, radiation, propagation, communication, navigation,
radar, RF systems, remote sensing, and radio astronomy who require a better understanding of the
underlying concepts, designs, advancements and applications of antennas.

Meta-Smith Charts and Their Potential Applications

Danai Torrungrueng
2010

Generalized Transmission Line Method to Study the Far-zone Radiation of Antennas under
a Multilayer Structure
Xuan Hui Wu, Ahmed A. Kishk, and Allen W. Glisson
2008

Narrowband Direction of Arrival Estimation for Antenna Arrays

Jeffrey Foutz, Andreas Spanias, and Mahesh K. Banavar
2008

Multiantenna Systems for MIMO Communications

Franco De Flaviis, Lluis Jofre, Jordi Romeu, and Alfred Grau
2008

Reconfigurable Antennas
Jennifer T. Bernhard
2007

Electronically Scanned Arrays

Robert J. Mailloux
2007
iv

Introduction to Smart Antennas

Constantine A. Balanis and Panayiotis I. Ioannides
2007

Antennas with Non-Foster Matching Networks

James T. Aberle and Robert Loepsinger-Romak
2007

Implanted Antennas in Medical Wireless Communications

Yahya Rahmat-Samii and Jaehoon Kim
2006
Meta-Smith Charts and
Their Potential Applications

Danai Torrungrueng
Asian University, Thailand

SYNTHESIS LECTURES ON ANTENNAS #10

M
&C Morgan & cLaypool publishers
ABSTRACT
This book presents the developments and potential applications of Meta-Smith charts, which can be
applied to practical and useful transmission line problems (e.g., metamaterial transmission lines and
nonreciprocal transmission lines). These problems are beyond the capability of the standard Smith
chart to be applied effectively. As any RF engineer is aware, a key property of the Smith chart is the
insight it provides, even in very complex design processes. Like the Smith chart, Meta-Smith charts
provide a useful way of visualizing transmission line phenomenon. They provide useful physical
insight, and they can also assist in solving related problems effectively. This book can be used as a
companion guide in studying “Microwave Engineering” for senior undergraduate students as well as
for graduate students. It is also recommended for researchers in the RF community, especially those
working with periodic transmission line structures and metamaterial transmission lines. Problems
are also provided at the end of each chapter for readers to gain a better understanding of material
presented in this book.

KEYWORDS
Smith Chart, Meta-Smith Charts, transmission line, metamaterial transmission line,
nonreciprocal transmission line, periodic transmission line structure, exponentially ta-
pered nonuniform transmission line, conjugately characteristic-impedance transmission
line, bi-characteristic-impedance transmission line, negative characteristic resistance,
nonnegative characteristic resistance, RF circuit
 This book is dedicated to
imagination and creativity.

The constant support of my family,

in particular my mom and dad,
is deeply appreciated.
 ix

Contents
Foreword I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

Foreword II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii

List of Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxi

Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxiii

1 Essential Transmission Line Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Transmission Line Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Terminated Uniform Transmission Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3.1 Terminated Lossy Uniform Transmission Lines . . . . . . . . . . . . . . . . . . . . . . . 6
1.3.2 Terminated Lossless Uniform Transmission Lines . . . . . . . . . . . . . . . . . . . . . 8
1.4 Terminated Nonuniform Transmission Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.4.1 Terminated Exponentially Tapered Lossless Nonuniform Transmission
Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.6 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

2 Theory of Conjugately Characteristic-Impedance Transmission Lines

(CCITLs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.2 Nonreciprocal Lossless Uniform Transmission Lines . . . . . . . . . . . . . . . . . . . . . . . . 18
2.3 Reciprocal Lossless Uniform Transmission Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.4 Exponentially Tapered Lossless Nonuniform Transmission Lines . . . . . . . . . . . . . 20
2.5 Finite Reciprocal Lossless Periodic Transmission Line Structures . . . . . . . . . . . . . 21
2.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
x
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.7 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

3 Theory of Bi-Characteristic-Impedance Transmission Lines (BCITLs) . . . . . . . . 31

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.2 Nonreciprocal Lossy Uniform Transmission Lines . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.3 Finite Reciprocal Lossy Periodic Transmission Line Structures . . . . . . . . . . . . . . . 33
3.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.5 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

4 Meta-Smith Charts for CCITLs and BCITLs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
4.2 The Meta-Smith Chart for CCITLS with Nonnegative Characteristic
Resistances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.2.1 The Z Meta-Smith Chart for CCITLs with NNCRs . . . . . . . . . . . . . . . . . 42
4.2.2 The Y Meta-Smith Chart for CCITLs with NNCRs . . . . . . . . . . . . . . . . . 44
4.2.3 Numerical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.3 The Meta-Smith Chart for CCITLS with Negative Characteristic Resistances . 48
4.3.1 The Z and Y Meta-Smith Charts for CCITLs with NCRs . . . . . . . . . . . . 49
4.3.2 Numerical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.4 The Meta-Smith Chart for BCITLS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.4.1 The Z Meta-Smith Chart for BCITLs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
4.4.2 The Y Meta-Smith Chart for BCITLs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.4.3 Numerical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.6 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

5 Applications of Meta-Smith Charts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.2 Nonreciprocal Stub Tuners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.2.1 Nonreciprocal Single-Stub Series Tuners . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
5.2.2 Nonreciprocal Double-Stub Shunt Tuners . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
5.3 Terminated Exponentially Tapered Lossless Nonuniform Transmission Lines . . . 73
5.3.1 Procedures of Using the Meta-Smith Chart . . . . . . . . . . . . . . . . . . . . . . . . . 73
5.3.2 Numerical Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
 xi
5.4 Terminated Finite Reciprocal Lossless Periodic Transmission Line Structures
with Nonnegative Characteristic Resistances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
5.5 CCITLS with Negative Characteristic Resistances . . . . . . . . . . . . . . . . . . . . . . . . . 77
5.6 Reciprocal Lossy BCITLS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
5.8 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

A The Transmission (ABCD) Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84

B The Required Condition for Nonreciprocal CCITLs . . . . . . . . . . . . . . . . . . . . . . . . 85

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

C Derivation of Z0+ = (Z0− )∗ for CCITLs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

D Derivation of the Magnitude of the Voltage Reflection Coefficient at the Load

for Terminated CCITLs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

E Power Consideration for Terminated Finite Reciprocal Lossless Periodic

Transmission Line Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92

Author’s Biography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
 Foreword I
Metamaterial microwave circuits and related printed structures have gained significant interest in
recent years. Their wave slow-down properties have led to smaller matching circuits, microwave
filters, feed networks and antenna arrays. Nonreciprocal propagation is another characteristic in
some of these structures. With the growing interest in using metamaterials, Smith charts that can
be used for designing related RF components are badly needed.
This book is a well-written text with useful problems at the end of each chapter to serve
as a guide for students and instructors in using the proposed meta-charts (appropriately meaning
the next-chart or after-chart upon translating the Greek word meta). The book presents meta-
charts for exponentially tapered (varying characteristic impedance as a function of position along
the transmission line), nonreciprocal lossless, and nonreciprocal lossy transmission lines, among
others. Several applications are given, including the standard single and double stub tuners.
This is a well-written and easy-to-follow book. With the increasing use of metamaterials
in design, it can be an important companion to any microwave engineering text and to users of
commercial and research computational tools.

John L. Volakis
R.&L. Chope Chair Professor
The Ohio State University
 Foreword II
Dramatic enhancements in technologies for manipulating materials and their properties are spurring
innovations in many engineering disciplines, including those involving microwave and electromag-
netic systems. The creation of new antennas, enhanced transmit and/or receive circuit devices such as
switches, filters, mixers, and amplifiers, and the integration of these systems into smaller and smaller
packages are all subjects of extensive current research. Achieving operation over wider bandwidths,
electronically adaptable (“software defined”) performance, and platform conformal integration, all
at reduced cost and manufacturing complexity, are among the potential gains enabled by material
innovations. I believe that the future will see our present time as the beginning of a revolution in
RF technologies and the capabilities they provide due to the new materials that are currently being
investigated.
While system components will be the drivers of this future, innovations in engineering design
procedures will also be necessary in order to understand and to utilize the new “tools” in the engineer’s
toolbox that will be available. The Smith chart is one of the most fundamental tools in RF design,
and accompanies the basic transmission line structures that provide connections between RF devices.
As any RF engineer is aware, a key property of the Smith chart is the insight it provides in even
complex design processes. Enhancement of Smith charts into “Meta-Smith charts” to enable their
continued use even with transmission lines that are nonreciprocal or contain metamaterials keeps this
insight available for designs using future materials. Prof.Torrungrueng makes a valuable contribution
with this volume whose benefits will be realized at present, but even more so as the metamaterial
revolution continues.

Joel T. Johnson
Professor, Electrical and Computer Engineering
The Ohio State University
 Preface
The Smith chart was developed in early 1937 by Phillip H. Smith. In early 1939, an article on the
Smith chart was first published in the Electronics magazine. Radio-frequency (RF) engineers have
found that the Smith chart is much more than merely a graphical tool; not only does it allow the user
to gain more physical insight when visualizing transmission line (TL) phenomenon, but also assists
in solving associated problems effectively. Although powerful computers are used as the dominant
design tool nowadays, the Smith chart still remains widely in use. It has come to provide the basis
for both modern computer and measurement instrument displays.
I became impressed with the usefulness of the application of the Smith chart whilst studying
courses on antennas and frequency-selective surfaces with the late Prof. Dr. Ben A. Munk during my
graduate studies at The Ohio State University [1], [2]. Prof. Munk had shown that use of the Smith
chart can provide a physical understanding of many associated complex phenomena. In 2002, while
preparing my lecture notes on the transmission line theory from Chapter 2 in [3] (see Problem 2.29),
I asked myself the question: “Can a graphical tool, like the Smith chart, be applied for nonreciprocal
lossless transmission lines?” The author found that the answer to this question was “No”, and so I
started developing a new graphical tool for nonreciprocal lossless transmission lines. In 2004, the
first journal paper on the new graphical tool, called a generalized ZY Smith chart or “T-Chart”, was
published [4]. I and my team have continued developing the theory of T-Charts, including their
practical applications [4-10].
In the references [4-10], the authors employed the terminology “T-Charts” for graphical
tools associated with conjugately characteristic-impedance transmission lines (CCITLs) and bi-
characteristic-impedance transmission lines (BCITLs). These transmission lines are discussed in
detail later in this book. It should be pointed out that a nonreciprocal lossless TL in [4] is an
example of CCITLs. In this book, the new terminology “Meta-Smith charts” is used instead of “T-
Charts” to make readers, who are familiar with the Smith chart, see the originality and generality of
the proposed graphical tools: the Meta-Smith charts. Note that the Greek word meta means beyond.
The Meta-Smith charts can be applied to more practical and useful TL problems; i.e., CCITLs and
BCITLs, beyond the capability of the Smith chart. Another reason for using this new terminology
is that the Meta-Smith charts can be applied to analyze metamaterial TLs, where the Smith chart
cannot be employed effectively. It will be shown in the book that the Smith chart is a special case of
the Meta-Smith charts.
This book is organized into five chapters and five appendices. Chapter 1 provides the essential
transmission line theory. Both uniform and nonuniform transmission lines are discussed in this
chapter. The material covered in this chapter will serve as an essential grounding in understanding
subsequent material in the book. Chapters 2 and 3 give the theories of CCITLs and BCITLs
xviii PREFACE
respectively, including practical examples. These two chapters will serve as the necessary background
in developing the Meta-Smith charts for CCITLs and BCITLs in Chapter 4. Chapter 5 provides
several applications of the Meta-Smith charts for solving practical problems associated with CCITLs
and BCITLs. Finally, five appendices are also provided to supplement material in the chapters
discussed earlier. In this book, the Meta-Smith charts are applied to analyze and design passive
circuits only. However, they can also be applied for problems associated with active circuits. Due to
the fact that the Meta-Smith charts depend on parameters of CCITLs and BCITLs, computerized
Meta-Smith chart programs are indispensable, and they are developed to generate plots of the
Meta-Smith charts.
This book can be used as a companion guide in studying “Microwave Engineering” for senior
undergraduate students as well as for graduate students. It is also recommended for researchers in the
RF community, especially those working with periodic TL structures and metamaterial TLs. Prob-
lems are also provided at the end of each chapter for readers to gain a better understanding of material
presented in this book. For convenience in usage of Meta-Smith charts, a computerized Meta-Smith
chart program, developed in Java, can be accessed via http://www.meta-smithcharts.org/.
I hope that the Meta-Smith charts will be used extensively by students, professors and re-
searchers in the RF community in the near future to develop a physical understanding of many
sophisticated phenomena of complex transmission lines and related problems. Useful suggestions
and comments from readers are welcome, and they can be submitted by e-mail to the author via
[email protected] (cc: [email protected]).

DanaiTorrungrueng
Chon Buri, Thailand

BIBLIOGRAPHY
[1] B. A. Munk, Frequency Selective Surfaces, Theory and Design. New York: Wiley Interscience,
2000. xvii

[2] B. A. Munk, Finite Arrays and FSS. New York: Wiley Interscience, 2003. xvii

[3] D. M. Pozar, Microwave Engineering, 2nd ed. New Jersey: Wiley, 1998. xvii

[4] D. Torrungrueng and C. Thimaporn, “A generalized ZY Smith chart for solvingnon reciprocal
uniform transmission line problems,” Microwave and Optical Technology Letters, vol. 40, no. 1,
pp. 57–61, Jan. 2004. DOI: 10.1002/mop.11284 xvii

[5] D. Torrungrueng, P.Y. Chou, and M. Krairiksh, “An extended ZY T-chart for
conjugatelycharacteristic-impedance transmission lines with active characteristic impedances,”
Microwave and Optical Technology Letters, vol. 49, no. 8, pp. 1961–1964, Aug. 2007.
DOI: 10.1002/mop.22626
 BIBLIOGRAPHY xix
[6] D. Torrungrueng, P.Y. Chou, and M. Krairiksh, “A graphical tool for analysis and designof bi-
characteristic-impedance transmission lines,” Microwave and Optical Technology Letters, vol. 49,
no. 10, pp. 2368–2372, Oct. 2007. DOI: 10.1002/mop.22801

[7] D. Torrungrueng, P.Y. Chou, and M. Krairiksh, “Erratum: A graphical tool for analysis and
design of bi-characteristic-impedance transmission lines,” Microwave and Optical Technology
Letters, vol. 51, no. 4, pp. 1154, Apr. 2009. DOI: 10.1002/mop.24260

[8] D. Torrungrueng and C. Thimaporn, “Applications of the ZY T-chart for nonreciprocal stub
tuners,” Microwave and Optical Technology Letters, vol. 45, no. 3, pp. 259–262, May 2005.
DOI: 10.1002/mop.20789

[9] D. TorrungruengandC. Thimaporn, “Application of the T-chart for solving exponentially ta-
pered lossless nonuniform transmission line problems,” Microwave and Optical Technology Let-
ters, vol. 45, no. 5, pp. 402–406, Jun. 2005. DOI: 10.1002/mop.20836

[10] D.Torrungrueng, C.Thimaporn, and N. Chamnandechakun,“An application of the T-chart for

solving problems associated with terminated finite lossless periodic structures,” Microwave and
Optical Technology Letters, vol. 47, no. 6, pp. 594–597, Dec. 2005. DOI: 10.1002/mop.21239
 List of Abbreviations

BCITL Bi-Characteristic-Impedance Transmission Line

CCITL Conjugately Characteristic-Impedance Transmission Line

ETLNUTL Exponentially Tapered Lossless Nonuniform Transmission Line

NCR Negative Characteristic Resistance

NNCR Nonnegative Characteristic Resistance

NRI Negative Refractive Index

NRLSUTL Nonreciprocal Lossy Uniform Transmission Line

NRLUTL Nonreciprocal Lossless Uniform Transmission Line

OC Open-Circuited

RF Radio-Frequency

RHS Right-Hand Side

RLSUTL Reciprocal Lossy Uniform Transmission Line

RLUTL Reciprocal Lossless Uniform Transmission Line

SC Short-Circuited

TL Transmission Line
 Acknowledgments
I would like to thank my students who contributed to the development of the Meta-Smith charts;
namely: Ms. Chananya Thimaporn, Mr. Po-Yen Chou, Ms. Kanokwan Vudhivorn, Ms. Natcha
Chamnandechakun, Mr. Theeraputh Mekathikom, Mr. Alongkorn Darawankul, Mr. Attakorn
Wongwattanarat, and Mr. Sanchai Eardprab, including my professional colleagues: Dr. Suthasi-
nee Lamultree, Assistant Prof. Dr. Rardchawadee Silapunt, Dr. Chatrpol Lertsirimit, Assistant
Prof. Dr. Chuwong Phongcharoenpanich, and Prof. Dr. Monai Krairiksh. In addition, I would like
to thank Prof. Dr. Ben A. Munk of The Ohio State University for his valuable discussions lead-
ing to the discovery of the Meta-Smith charts, including his sincere encouragement and valuable
suggestions for writing this book.
Furthermore, I would like to express my sincere gratitude to all reviewers for their useful
feedback to improve the quality of this book; namely: Dr. Panuwat Janpugdee, Prof. Dr. Monai
Krairiksh, Prof. Dr. Joel T. Johnson, Prof. Dr. John L. Volakis, Prof. Dr. Yahya Rahmat-Samii,
Associate Prof. Dr. Mitchai Chongcheawchamnan, and Associate Prof. Dr. Prayoot Akkaraekthalin.
Finally, I would like to thank Mr. Sanchai Eardprab for his creative artwork, including the book
format arrangement.

Danai Torrungrueng
September 2010
 1

CHAPTER 1

Essential Transmission Line

Theory
1.1 INTRODUCTION
The transmission line (TL) theory is a crucial part in the general background for microwave engineers.
Basically, a transmission line consists of two or more conductors, and it is employed to deliver the
power from a source to a load. Figure 1.1 illustrates a two-conductor TL of length  connecting
the generator, with the source voltage ES and the source impedance ZS , to the load impedance ZL .
Typical TLs include a two-wire line, a coaxial cable and a microstrip line.

Zs

Es Z0, γ ZL

z
–ℓ 0

Figure 1.1: A two-conductor TL connecting the generator to the load.

A transmission line is a distributed-parameter two-port network, where the variations of its

voltage v(z, t) and current i(z, t) can be observed over its length. Note that both v(z, t) and i(z, t)
are functions of both space z and time t. TL problems can be analyzed using an extension of circuit
theory or a specialization of Maxwell’s equations [1]. In this chapter, the former approach is employed
to study wave propagation on TLs due to its simpler mathematical formulation.
In this book, cosine-based phasors
√ with ej ωt time-harmonic convention are assumed in the
formulation, where j is defined as −1, ω = 2πf is the radian frequency, f is the frequency, and t is
the time.Thus, the conversion from phasor quantities to real time-varying quantities is accomplished
by multiplying the phasor by ej ωt and taking the real part. For example, the real time-varying voltage
v(z, t) and current i(z, t) can be determined from the corresponding phasor voltage V (z) and phasor
2 1. ESSENTIAL TRANSMISSION LINE THEORY
current I (z) as

v(z, t) = Re{V (z)ej ωt }, (1.1)

i(z, t) = Re{I (z)ej ωt }, (1.2)

where Re{} denotes the symbol of real part.

This chapter is organized as follows. Section 1.2 presents the TL equations derived from a
lumped-element equivalent circuit model of a TL. Uniform TLs for both lossy and lossless cases
terminated in a load impedance are discussed in Section 1.3. Section 1.4 presents terminated nonuni-
form TLs.Then, conclusions are provided in Section 1.5. Note that the material covered in this chap-
ter will serve as an essential grounding in understanding subsequent material in the book. However,
extensive treatment of the TL theory can be found in [1]–[3].

1.2 TRANSMISSION LINE EQUATIONS

Using an extension of circuit theory, a two-conductor TL of an infinitesimal length z can be
modeled as an L-type lumped-element equivalent circuit model as shown in Figure 1.2. The TL
parameters R, L, G and C are per unit length quantities, which are constants for uniform TLs. Note
that R and L represent series resistance per unit length (/m) and series inductance per unit length
(H/m), respectively, while G and C represent shunt conductance per unit length (S/m) and shunt
capacitance per unit length (F/m), respectively. Physically, R and G represent conductor loss and
dielectric loss in the material between the conductors, respectively. Thus, lossless TLs possess R= 0
/m and G= 0 S/m. It should be pointed out that other types of lumped-element equivalent circuit
models can be used as well in deriving TL equations; e.g., -type and T -type [4].

Figure 1.2: An L-type lumped-element equivalent circuit model of a two-conductor TL of an infinites-

imal length z.

Applying Kirchhoff ’s voltage law and Kirchhoff ’s current law to the equivalent circuit model of
Figure 1.2 and taking the limit as z → 0 yield the following first-order coupled partial differential
 1.2. TRANSMISSION LINE EQUATIONS 3
TL equations in the time-domain form:

∂v(z, t) ∂i(z, t)
= −Ri(z, t) − L , (1.3)
∂z ∂t
∂i(z, t) ∂v(z, t)
= −Gv(z, t) − C , (1.4)
∂z ∂t
respectively.
For the sinusoidal steady-state condition, the phasor voltage V (z) and phasor current I (z)
defined in (1.1) and (1.2) can be employed to simplify (1.3) and (1.4) considerably. Substitut-
ing (1.1) and (1.2) into (1.3) and (1.4) yields the following first-order coupled ordinary differential
TL equations in the frequency-domain form:

dV (z)
= −(R + j ωL)I (z), (1.5)
dz
dI (z)
= −(G + j ωC)V (z). (1.6)
dz
Solving (1.5) and (1.6) simultaneously yields the following identical wave equations for V (z) and
I (z):

d 2 V (z)
− γ 2 V (z) = 0, (1.7)
dz2
d 2 I (z)
− γ 2 I (z) = 0, (1.8)
dz2
where γ is the complex frequency-dependent propagation constant of the TL defined as

γ ≡ α + jβ = (R + j ωL)(G + j ωC). (1.9)

Note that α and β are the attenuation constant and propagation constant of the TL, respectively.
Solving (1.7) and (1.8) yields the following traveling wave solutions for V (z) and I (z) :

V (z) = V0+ e−γ z + V0− eγ z , (1.10)

I (z) = I0+ e−γ z + I0− eγ z . (1.11)

In (1.10) and (1.11), the terms e−γ z and eγ z represent wave propagation in the +z and −z directions,
respectively, for ej ωt time-harmonic convention. Note that V0+ and V0− (I0+ and I0− ) are defined
as the amplitudes of incident and reflected voltage (current) waves referenced at z = 0 (at the load)
respectively, (see Figure 1.1). Substituting (1.10) into (1.5), I (z) can be written in terms of V0+ and
V0− as
1  + −γ z 
I (z) = V0 e − V0− eγ z , (1.12)
Z0
4 1. ESSENTIAL TRANSMISSION LINE THEORY
where Z0 is the complex frequency-dependent characteristic impedance of the TL defined as

V0+ V0− R + j ωL
Z0 ≡ + = − − = . (1.13)
I0 I0 G + j ωC

It should be pointed out that the relationships between V (z) and I (z) are given as in (1.5) and (1.6),
and in general
V (z)
I (z)  = , (1.14)
Z0
unless V0− = 0, (i.e., no reflected wave). Once V (z) and I (z) are known, v(z, t) and i(z, t) can be
determined using (1.1) and (1.2), respectively.
For lossless TLs, TL parameters R and G representing loss are equal to zero. Substituting
R = G =0 into (1.9) and (1.13) yields

α = 0, (1.15)
√
β = ω LC, (1.16)

L
Z0 = . (1.17)
C
Note that the attenuation constant α is zero for the lossless case as expected, and the characteristic
impedance Z0 is purely real. Thus, (1.10) and (1.12) can be expressed for a lossless TL as

V (z) = V0+ e−jβz + V0− ejβz , (1.18)

1  + −jβz 
I (z) = V0 e − V0− ejβz . (1.19)
Z0
To determine the wavelength λ and the phase velocity vp of waves propagating along lossless TLs,
consider v(z, t) in the time domain. For convenience in manipulation, V0+ and V0− can be represented
in the polar form as
+
V0+ = V0+ ej φ0 , (1.20)
j φ0−
V0− = V0− e , (1.21)

where φ0± are the arguments of V0± . Substituting (1.18) into (1.1) and employing (1.20) and (1.21),
v(z, t) can be expressed as

v(z, t) = V0+ cos(ωt − βz + φ0+ ) + V0− cos(ωt + βz + φ0− ). (1.22)

As pointed out earlier, the first term represents a wave propagating in the +z direction, while the
second term represents a wave propagating in the −z direction. Without loss of generality, consider
only the first term in determining the wavelength λ and the phase velocity vp . By definition, the
 1.3. TERMINATED UNIFORM TRANSMISSION LINES 5
wavelength is defined as the distance between two successive reference point on the wave, at a fixed
instant of time, possessing the phase difference of 2π. Using this definition, λ can be expressed in
terms of β as
2π
λ= . (1.23)
β
For the phase velocity, it is defined as the velocity at which a fixed phase point on the wave travels.
Consider a fixed phase point φ c = constant; i.e.,

ωt − βz = φc . (1.24)

Using (1.24) and (1.23), the phase velocity vp can be determined as

dz ω
vp = = = f λ. (1.25)
dt β

For lossless TLs, substituting (1.16) into (1.25) yields

1
vp = √ . (1.26)
LC

1.3 TERMINATED UNIFORM TRANSMISSION LINES

In this section, effects of load termination on uniform TLs will be studied for both lossy and lossless
cases. Figure 1.3 illustrates a uniform TL terminated in an arbitrary load impedance ZL . Only passive
load impedances (Re{ZL } ≥ 0) are of interest in this section.

Figure 1.3: A uniform TL terminated in a load impedance ZL .

6 1. ESSENTIAL TRANSMISSION LINE THEORY
1.3.1 TERMINATED LOSSY UNIFORM TRANSMISSION LINES
Lossy TLs were studied intensively in [5]. For lossy uniform TLs, the complex propagation constant
γ and the characteristic impedance Z0 are complex quantities. Let us define the voltage reflection
coefficient  at the load as the ratio of the amplitude of the reflected voltage wave to that of incident
one; i.e.,
V−
 ≡ 0+ . (1.27)
V0
Using (1.27), (1.10), and (1.12) can be rewritten in terms of  as
 
V (z) = V0+ e−γ z + eγ z , (1.28)
V0+  
I (z) = e−γ z − eγ z , (1.29)
Z0
where  can be determined by considering the load condition; i.e.,

V (z = 0) 1+
ZL ≡ = Z0 . (1.30)
I (z = 0) 1−

Solving (1.30) for  yields

ZL − Z0
= . (1.31)
ZL + Z 0
Note that  = 0 when the load impedance ZL is equal to the characteristic impedance Z0 . Under
this condition, the TL matches to the load with no reflection of the incident wave as can be seen
from (1.28) and (1.29) with  = 0. In general, the voltage reflection coefficient (z= – d) at an
arbitrary location z = – d along the TL can be defined as the ratio of the reflected component to the
incident one, (see (1.28)); i.e.,

V0+ e−γ d
(z = −d) ≡ = e−2γ d . (1.32)
V0+ eγ d

Note that (z= – d) and  are closely related as given in (1.32), and (z= – d) is equal to zero when
 = 0 as expected.
The input impedance Zin as shown in Figure 1.3 can be determined from (1.28) and (1.29)
as
V (z = −) 1 + e−2γ 
Zin ≡ = Z0 . (1.33)
I (z = −) 1 − e−2γ 
Using (1.32), (1.33) can be rewritten compactly in terms of the voltage reflection coefficient in at
the input, where in = (z = −), as

1 + in
Zin = Z0 . (1.34)
1 − in
 1.3. TERMINATED UNIFORM TRANSMISSION LINES 7
For convenience in calculation, Zin in (1.33) can alternatively be expressed in terms of the load
impedance ZL , by using (1.31), as

ZL + Z0 tanh (γ )
Zin = Z0 . (1.35)
Z0 + ZL tanh (γ )

Comparing (1.30) and (1.34), it is obvious that they are in the identical form, which is useful for
constructing a graphical tool, called the Smith chart [3], for solving reciprocal uniform TL problems
(see Problem 1.3 at the end of this chapter). It should be pointed out that the Smith chart is more
than a graphical tool. It usually provides a useful way with more physical insight of visualizing TL
phenomenon and solving TL problems effectively [1]. Extensive discussion and useful applications
of the Smith chart can be found in detail in [3].
It is interesting to point out that the magnitude of the voltage reflection coefficient at the
load || can exceed unity for reciprocal lossy uniform TLs (RLSUTLs) terminated by a passive load
impedance ZL . This can be seen from (1.31) as follows. Let us define the normalized impedance z
as
Z
z≡ , (1.36)
Z0
where Z is the impedance of interest. Using (1.36),  in (1.31) can be rewritten compactly as:

zL − 1
= , (1.37)
zL + 1
where zL is the normalized load impedance. Since the characteristic impedance Z0 and the load
impedance ZL are complex quantities in general, let us conveniently define them in the polar form
as:

Z0 = |Z0 | e−j φ0 , (1.38)

ZL = |ZL | e j φL
, (1.39)

where |Z0 | and –φ 0 are the magnitude and the argument of Z0 respectively, and |ZL | and φ L
represent the magnitude and the argument of ZL , respectively. Since all parameters of TLs in (1.13)
are nonnegative, it is obvious that the angle φ 0 must lie in the following range:

−45◦ ≤ φ0 ≤ 45◦ . (1.40)

In addition, for a passive load, (1.39) implies that

−90◦ ≤ φL ≤ 90◦ . (1.41)

Let us define the normalized load impedance zL as:

zL = rL + j xL , (1.42)
8 1. ESSENTIAL TRANSMISSION LINE THEORY
where rL and xL are the real and imaginary parts of zL , respectively. From (1.36), it should be
pointed out that the arguments of the load impedance ZL and the normalized load impedance zL
are generally different since the characteristic impedance Z0 is a complex quantity. For example, when
ZL is real and Z0 is complex, the normalized load impedance zL is complex. Using (1.36), (1.38),
and (1.39), it can be shown readily that [6]
ZL
rL = cos(φL + φ0 ). (1.43)
Z0
Note that if cos(φL + φ0 ) < 0, (e.g., φL = 90◦ and φ0 = 30◦ ), rL will become a negative quantity.
Using (1.37) and (1.42), the magnitude of the voltage reflection coefficient at the load || can be
written as: 
(rL − 1)2 + xL2
|| = . (1.44)
(rL + 1)2 + xL2
From (1.44), it is obvious that when rL < 0, || can exceed unity since the additional term xL2 is the
same for both numerator and denominator.

1.3.2 TERMINATED LOSSLESS UNIFORM TRANSMISSION LINES

As pointed out earlier in Section 1.2, the attenuation constant α of lossless TLs is zero; i.e., the
complex propagation constant γ is purely imaginary, and its characteristic impedance Z0 is purely
real. Note that a lossless TL is a special case of a lossy TL when the loss is vanished. Following the
same analysis as in the lossy case, the voltage reflection coefficient  at the load for the lossless case
is given as in (1.31). In addition, the input impedance Zin for the lossless case is given as in (1.33),
where γ is equal to jβ. For convenience in calculation, Zin in (1.35) for the lossless case can be
expressed in terms of the load impedance ZL as

ZL + j Z0 tan (β)
Zin = Z0 . (1.45)
Z0 + j ZL tan (β)
For reciprocal lossless uniform TLs terminated by a passive load impedance ZL , the magnitude
of the voltage reflection coefficient at the load || cannot exceed unity (|| ≤ 1). This can be seen
by considering (1.44), and using the fact that rL defined in (1.43) is always greater than or equal to
zero (rL ≥ 0) for the lossless case (φ0 = 0◦ in (1.38) and φL is given in (1.41)). In the next section,
terminated nonuniform TLs are presented.

1.4 TERMINATED NONUNIFORM TRANSMISSION LINES

Nonuniform transmission lines (NUTLs) have several advantages over uniform ones, and they
have found various applications in microwave technology [7]–[13]; i.e., impedance transformers,
resonators, delay equalizers, directional couplers and filters. For applications on impedance trans-
formers, NUTLs are generally used for matching two unequal impedances over a wide band of
 1.4. TERMINATED NONUNIFORM TRANSMISSION LINES 9
frequencies. The reflection coefficient of voltage/current along NUTLs can be described by a non-
linear Ricatti-type differential equation, and its general solution does not exist analytically [13].
To illustrate analytical procedures for analyzing NUTLs, a simple and well-known NUTL
is considered in the following subsection; i.e., the exponentially tapered lossless nonuniform trans-
mission line (ETLNUTL). To be practically useful, ETLNUTLs are considered in the presence of
passive load terminations.

1.4.1 TERMINATED EXPONENTIALLY TAPERED LOSSLESS

NONUNIFORM TRANSMISSION LINES
One of the NUTLs, the ETLNUTL, has found various applications in microwave technology; e.g.,
a resonator, an antenna, and a matching device suitable for matching two unequal impedances over a
wideband of frequencies [14]–[17]. By definition, an ETLNUTL is a TL whose impedance per unit
length Z (z) and admittance per unit length Y (z) vary exponentially with distance down the line.
Note that its impedance and admittance per unit length can be expressed mathematically as [15]

Z (z) = j ωL0 eqz , (1.46)

Y (z) = j ωC0 e−qz , (1.47)

respectively. Note that L0 and C0 are the inductance and capacitance per unit length at the load (z
= 0) respectively, and q is the real taper factor, which can be positive, zero or negative. In Figure 1.4,
the ETLNUTL of length  is employed to match a passive real load impedance ZL to the lossless
feed line of characteristic impedance Z0 , where Z0 is real.

Figure 1.4: An ETLNUTL of length  is employed to match a passive real load impedance ZL to the
lossless feed line of characteristic impedance Z0 .
10 1. ESSENTIAL TRANSMISSION LINE THEORY
For an ETLNUTL, the characteristic impedance Zc (z) along the ETLNUTL is defined
as [15] 
Z (z)
Zc (z) = . (1.48)
Y (z)
Substituting (1.46) and (1.47) into (1.48), the characteristic impedance Zc (z) can be expressed as

L0 qz
Zc (z) = Zc (z = 0)e =
qz
e . (1.49)
C0

To match the load to the feed line, the characteristic impedances at the input (z = −) and at the
load (z = 0) are required as follows, (see Figure 1.4):

Zc (z = −) = e−q Zc (z = 0) = Z0 , (1.50)

Zc (z = 0) = ZL . (1.51)

Note that the phasor voltage V (z) and the phasor current I (z) distributed along the ETLNUTL can
be determined exactly by solving the following first-order coupled ordinary differential equations [14],
[15]:
dV (z)
= −Z (z)I (z), (1.52)
dz
dI (z)
= −Y (z)V (z). (1.53)
dz
Note that (1.52) and (1.53) can be rearranged into the set of the second-order uncoupled ordinary
differential equations as follows:

d 2 V (z) 1 dZ (z) dV (z)

2
− − [Z (z)Y (z)] V (z) = 0, (1.54)
dz Z (z) dz dz
d 2 I (z) 1 dY (z) dI (z)
2
− − [Z (z)Y (z)] I (z) = 0. (1.55)
dz Y (z) dz dz

Using (1.46) and (1.47), (1.54) and (1.55) can be simplified as

d 2 V (z) dV (z)
−q + ω2 L0 C0 V (z) = 0, (1.56)
dz2 dz
d 2 I (z) dI (z)
+ q + ω2 L0 C0 I (z) = 0, (1.57)
dz2 dz
which are the wave equations of V (z) and I (z) for ETLNUTLs, respectively. Unlike the wave
equations for uniform TLs, (see (1.7) and (1.8)), (1.56) and (1.57) are different. Solving (1.56)
and (1.57) analytically, it is found that the phasor voltage V (z) and the phasor current I (z) can be
 1.4. TERMINATED NONUNIFORM TRANSMISSION LINES 11
written compactly in terms of the superposition of two traveling waves propagating in the opposite
directions as follows:

V (z) = V1 e−γV ,1 z + V2 e−γV ,2 z , (1.58)

I (z) = I1 e−γI,1 z + I2 e−γI,2 z , (1.59)

where the subscripts 1 and 2 are associated with waves propagating in the +z and –z directions,
respectively, and
q
γV ,1 = − + jβ, (1.60)
2
q
γI,1 = + jβ. (1.61)
2
Note that γV ,2 and γI,2 in (1.58) and (1.59) are complex conjugate of γV ,1 and γI,1 , respectively, and
the propagation constant β is defined as
  q 2
β ≡ ω 2 L0 C0 − . (1.62)
2
For convenience in analysis, β is assumed to be real, and only nonnegative values of β are considered.
In (1.58) and (1.59), Vi and Ii are voltage and current amplitudes of propagating waves referenced at
the load respectively, where i = 1 and 2. In (1.58), it is observed that the magnitudes of the first and
second terms on the right-hand side (RHS) of the phasor-voltage equation increase as z increases.
On the contrary, the magnitudes of the first and second terms on the RHS of the phasor-current
equation, (see (1.59)) decrease as z increases. However, it can be shown rigorously that the time-
average real power is constant along the ETLNUTL for real q and nonnegative β; i.e., there is no
power loss when waves propagate along the ETLNUTL. This result agrees with the fact that the
ETLNUTL is lossless.
Next, substituting (1.58) and (1.59) into (1.52) and (1.53), one obtains the following rela-
tionships between Vi and Ii (i= 1 and 2):

V1
I1 = , (1.63)
Z0+
V2
I2 = − − , (1.64)
Z0

where Z0± are the effective characteristic impedances of ETLNUTLs defined as

γI,1
Z0+ ≡ , (1.65)
j ωC0
γI,2
Z0− ≡ − . (1.66)
j ωC0
12 1. ESSENTIAL TRANSMISSION LINE THEORY

For real q and nonnegative β, it can be shown rigorously that Z0+ and Z0− are related as follows:
 ∗
Z0+ = Z0− , (1.67)
where the superscript “*” denotes the complex conjugate symbol. Note that the total voltage V (z =
0) and current I (z = 0) at the load are related to the load impedance ZL as
V (z = 0)
ZL = . (1.68)
I (z = 0)
The voltage reflection coefficient at the load for ETLNUTLs is defined as:
V2
≡ . (1.69)
V1
Using (1.58), (1.59), (1.63), (1.64), (1.68), and (1.69),  can be expressed compactly in terms of Z0+ ,
Z0− , and ZL as
ZL Z0− − Z0+ Z0−
= . (1.70)
ZL Z0+ + Z0+ Z0−
Next, let us consider the input impedance Zin of ETLNUTLs as shown in Figure 1.4, which is
defined as
V (z = −)
Zin ≡ . (1.71)
I (z = −)
Substituting (1.58) and (1.59) into (1.71) and using (1.60), (1.61), (1.63), (1.64), and (1.69), Zin
can be expressed compactly as
 
1 + e −j 2β
Zin = Z0+ Z0− e−q . (1.72)
Z0− − Z0+ e−j 2β
To verify that the expression of Zin in (1.72) is valid, one can compute the input reflection coefficient
in from Zin in (1.72), where in is defined as, (see Figure 1.4)
Zin − Z0
in ≡ , (1.73)
Zin + Z0
and compare the above result to the analytical solution for in given in [13]. To achieve this, it is
more convenient to express Zin in the following form:
       
ZL Z0+ + Z0− + j ZL Z0+ − Z0− + 2Z0+ Z0− tan β −q
Zin =   e . (1.74)
Z0− + Z0+ + j Z0− − Z0+ + 2ZL tan β
Note that (1.74) is obtained by substituting  in (1.70) into (1.72) and rearranging terms.
Finally, substituting (1.74) into (1.73) and manipulating terms, in can be written compactly as
q sin β
in = √ . (1.75)
2β cos β + j 2ω L0 C0 sin β
It is found that the resulting input reflection coefficient in (1.75) is identical to the existing one
in [13]. Thus, it can be concluded that the derived input impedance Zin in (1.72) is valid indeed.
 1.5. CONCLUSIONS 13

1.5 CONCLUSIONS
In this chapter, the covered material provides a necessary grounding in understanding material in
subsequent chapters, including mathematical notations and conventions employed in the book. Both
uniform and nonuniform TLs are discussed in detail, and associated TL equations and parameters
are obtained. The formulas of two important parameters, the voltage reflection coefficient and the
input impedance, are derived for both uniform and nonuniform TLs. These parameters will be
reconsidered when constructing and applying associated Meta-Smith charts in Chapters 4 and 5.

BIBLIOGRAPHY
[1] D. M. Pozar, Microwave Engineering, 3rd ed. New Jersey: Wiley, 2005. 1, 2, 7

[2] U. S. Inan and A. S. Inan, Engineering Electromagnetics. California: Addison Wesley Longman,
1999.

[3] P. H. Smith, Electronic Applications of the Smith Chart. Georgia: Noble Publishing, 2000. 2, 7

[4] M.N.O. Sadiku, Elements of Electromagnetics, 4th ed. New York: Oxford University Press,
2007. 2

[5] F. E.Gardiol, Lossy Transmission Lines. Massachusetts: Artech House, 1987. 6

[6] D.Torrungrueng, A. Wongwattanarat, and M. Krairiksh, “Magnitude of the voltage reflection

coefficient of terminated reciprocal uniform lossy transmission lines,”Microwave and Optical
Technology Letters, vol. 49, no. 7, pp. 1516–1519, Jul. 2007. DOI: 10.1002/mop.22490 8

[7] R. W. Klopfenstein, “A transmission line taper of improved design,” Proc. IRE, vol. 44, pp. 31–
35, Jan. 1956. DOI: 10.1109/JRPROC.1956.274847 8

[8] D. Youla, “Analysis and synthesis of arbitrarily lossless nonuniform lines,” IEEE Trans. on
Circuit Theory, vol. 12, no. 1, pp. 145–145, Mar. 1965.

[9] E. N. Protonotarios and O. Wing,“Analysis and intrinsic properties of the general nonuniform
transmission line,” IEEE Trans. on Microwave Theory and Tech., vol. MTT-15, no. 3, pp. 142–
150, Mar. 1967. DOI: 10.1109/TMTT.1967.1126403

[10] A. Bergquist,“Wave propagation on nonuniform transmission lines,” IEEE Trans.

On Microwave Theory and Tech., vol. MTT-20, no. 8, pp. 557–558, Aug. 1972.
DOI: 10.1109/TMTT.1972.1127809

[11] S. Uysal, Nonuniform Line Microstrip Directional Couplers and Filters. Massachusetts:Artech
House, 1993.
14 1. ESSENTIAL TRANSMISSION LINE THEORY
[12] K. Lu, “An efficient method for analysis of arbitrary nonuniform transmission lines,”
IEEE Trans. on Microwave Theory and Tech., vol. MTT-45, no. 1, pp. 9–14, Jan. 1997.
DOI: 10.1109/22.552026

[13] R. E. Collin, Foundations for Microwave Engineering, 2nd ed. New York: McGraw-Hill, 1992.
8, 9, 12

[14] R. N.Ghose, “Exponential transmission lines as resonators and transformers,” IRE

Transactions on Microwave Theory and Tech., vol. 5, no. 3, pp. 213–217, Jul. 1957.
DOI: 10.1109/TMTT.1957.1125143 9, 10

[15] N. H. Younan, B. L. Cox, C. D. Taylor, andW.D. Prather, “An exponentially tapered trans-
mission line antenna,” IEEE Transactions on Electromagnetic Compatibility, vol. 36, no. 2,
pp. 141–144, May 1994. DOI: 10.1109/15.293276 9, 10

[16] S. P. Mathur and A. K. Sinha, “Design of microstrip exponentially tapered lines to match
helical antennas to standard coaxial transmission lines,” IEE Proceedings H on Microwaves,
Antennas and Propagation, vol.135, no. 4, pp. 272 – 274, Aug. 1988.

[17] D. Torrungrueng and C. Thimaporn, “Application of the T-chart for solving exponentially
tapered lossless nonuniform transmission line problems,” Microwave and Optical Technology
Letters, vol. 45, no. 5, pp. 402–406, Jun. 2005. DOI: 10.1002/mop.20836 9

1.6 PROBLEMS
1.1. Derive the partial differential TL equations of (1.3) and (1.4) by using Kirchhoff ’s voltage
law and Kirchhoff ’s current law to the equivalent circuit model of Figure 1.2 and taking the
limit as z → 0.

1.2. Show that the input impedance Zin of a terminated lossy uniform transmission lines as
shown in Figure 1.3 is given as in (1.35).

1.3. The Smith chart is a polar plot of the voltage reflection coefficient  at the load of terminated
uniform transmission lines. It can be employed to convert from voltage reflection coefficients
to normalized impedances (or admittances), and vice versa. Using (1.31) and (1.36), show
that the equations of resistance and reactance circles in the Smith chart are given as
2 2
rL 1
r − + i2 = ,
1 + rL 1 + rL
2 2
1 1
(r − 1)2 + i − = ,
xL xL
 1.6. PROBLEMS 15
respectively. Note that r and i are the real and imaginary parts of , and rL and xL are the
real and imaginary parts of the normalized load impedance zL defined in (1.42), respectively.
In addition, plot the families of resistance and reactance circles for several values of rL and
xL in the  plane, respectively. Do the resistance and reactance circles intersect at the right
angle? Please explain in detail.
Using the same concept as above, derive the equations of conductance and susceptance circles
in the Smith chart as well. In addition, plot the families of conductance and susceptance
circles for several values of gL and bL in the  plane, where gL and bL are the real and
imaginary parts of the normalized load admittance yL = 1/zL , respectively.
1.4. Using the fact that all parameters of TLs in (1.13) are nonnegative, show that the angle φ 0
of Z0 , (see (1.38)) must lie in the range defined in (1.40).
1.5. Using (1.52) and (1.53), derive (1.54) and (1.55) rigorously. In addition, show that the
phasor voltage V (z) and the phasor current I (z) given in (1.58) and (1.59) are the solutions
of (1.56) and (1.57), respectively.
1.6. Using (1.58) and (1.59), show that thetime-average real power P av (z) along the ETLNUTL,
defined as
1  
Pav (z) = Re V (z)I ∗ (z) ,
2
is constant for real q and nonnegative β, (see (1.60) to (1.62)). Does this result make physical
sense? Please explain. In addition, derive (1.67) for real q and nonnegative β.
1.7. Show that the voltage reflection coefficient  at the load and the input impedance Zin of
terminated ETLNUTLs are given in (1.70) and (1.72), respectively. Determine the condi-
tion when terminated ETLNUTLs have no reflected wave. What is the input impedance
Zin under this condition?
1.8. Rigorously derive (1.74) and (1.75) for terminated ETLNUTLs.
 17

CHAPTER 2

Theory of Conjugately
Characteristic-Impedance
Transmission Lines (CCITLs)
2.1 INTRODUCTION

A conjugately characteristic-impedance transmission line (CCITL) is a class of transmission lines

(TLs) used in microwave technology. By definition, CCITLs are lossless and possess the complex-
conjugate characteristic impedances, Z0+ and Z0− , of waves propagating in forward and reverse direc-
tions, respectively; i.e., Z0+ = (Z0− )∗ . In general, CCITLs can possess different propagation constants
β + and β − for propagation in the forward and reverse directions, respectively, as shown in Figure 2.1;
i.e., CCITLs can be nonreciprocal transmission lines as well. Note that β + and β − are real while Z0+
and Z0− are generally complex.
In Figure 2.1, the CCITL is terminated in a passive load impedance ZL (Re{ZL } ≥ 0), where
 is the voltage reflection coefficient at the load and Zin is the input impedance of the terminated
CCITL. Examples of CCITLs are nonreciprocal lossless uniform transmission lines (NRLUTLs),
reciprocal lossless uniform transmission lines (RLUTLs), exponentially tapered lossless nonuniform
transmission lines (ETLNUTLs) and finite reciprocal lossless periodic TL structures operated in
passbands. These TLs will be discussed in detail in Sections 2.2 to 2.5.
In practice, CCITLs can exhibit nonnegative characteristic resistances (NNCRs) and negative
characteristic resistances (NCRs); i.e., Re{Z0± } ≥ 0 and Re{Z0± } < 0, respectively. For example, finite
reciprocal lossless periodic TL structures can exhibit both NNCRs and NCRs in passbands, which
will be described in more detail in Sections 2.5, 5.4, and 5.5 [1], [2].
This chapter is organized as follows. Section 2.2 presents nonreciprocal lossless uniform
transmission lines as the first example of CCITLs. Reciprocal lossless uniform transmission lines as
a special case of CCITLs are discussed in Section 2.3, and exponentially tapered lossless nonuniform
transmission lines are presented in Section 2.4. Section 2.5 shows the theory of finite reciprocal
lossless periodic TL structures and illustrates that they can be analyzed using reciprocal CCITLs.
Finally, conclusions are provided in Section 2.6.
18 2. THEORY OF CCITLS

Figure 2.1: A nonreciprocal CCITL terminated in a passive load ZL .

2.2 NONRECIPROCAL LOSSLESS UNIFORM

TRANSMISSION LINES
Nonreciprocal TLs have found various applications in microwave technology; e.g., wave guiding
structures, microstrip and lens antennas, and reflectionless coating. Note that a nonreciprocal TL
has different TL parameters for waves propagating in opposite directions.The concept of nonreciprocal
TLs is particularly useful when analyzing electromagnetic wave propagation in layered structures of
bi-isotropic media [3]. The theory of nonreciprocal TLs was studied thoroughly in [3], [4].
Figure 2.1 illustrates a model of a nonreciprocal TL of length  terminated in a passive load
impedance ZL . Note that the nonreciprocal TL possesses the different propagation constants β + and
β − , with corresponding characteristic impedances, Z0+ and Z0− , for propagation in the forward and
reverse directions, respectively. In Appendix B, it is shown that β +  = β − is the required condition
for nonreciprocal lossless uniform TLs. On the nonreciprocal TL, the traveling wave equations for the
phasor voltage V (z) and the phasor current I (z) can be written as

+z −
V (z) = V0+ e−jβ + V0− ejβ z , (2.1)
V0+ −jβ + z V0− jβ − z
I (z) = e − e , (2.2)
Z0+ Z0−

+ −
where e−jβ z and ejβ z terms represent waves propagating in the +z and –z directions, respectively.
Note that V0+ and V0− are defined as the amplitudes of incident and reflected voltage waves referenced
at z = 0, respectively.
For a nonreciprocal lossless uniform transmission line, the time-average power must be constant
along the line. Using this condition, it can be shown rigorously in Appendix C that the characteristic
 2.3. RECIPROCAL LOSSLESS UNIFORM TLS 19

impedances Z0± must be complex conjugate of one another, which can be expressed as
 ∗
Z0+ = Z0− . (2.3)

Thus, NRLUTLs are nonreciprocal CCITLs. For convenience in manipulation later, let us define
Z0± in the polar form as
Z0± = Z0± e∓j φ , (2.4)

where |Z0± | and φ are the absolute value and the argument of Z0− , respectively.
Note that the total voltage and current at the load (V (0) and I (0)) are related by the load
impedance ZL as follows:
V (0)
ZL = . (2.5)
I (0)
The voltage reflection coefficient  at the load is defined as:

V0−
≡ . (2.6)
V0+

Using (2.1), (2.2), (2.5), and (2.6),  can be mathematically expressed in terms of ZL and Z0 as

ZL Z0− − Z0+ Z0−

= . (2.7)
ZL Z0+ + Z0+ Z0−

In Figure 2.1, the input impedance at the point z = − is given by

V (z = −) 1 + e−j 2β̃

Zin = = Z0+ Z0− − , (2.8)
I (z = −) Z − Z + e−j 2β̃
0 0

where (2.1), (2.2), and (2.6) are employed. Note that β̃ in (2.8) is defined as the arithmetic mean of
β ± , which can be expressed as
1 +  2π
β̃ ≡ β + β− ≡ , (2.9)
2 λ̃
where λ̃ is defined as the effective wavelength of waves propagating along NRLUTLs. In the next
section, reciprocal lossless uniform TLs are discussed as an example of CCITLs.

2.3 RECIPROCAL LOSSLESS UNIFORM TRANSMISSION

LINES
Reciprocal lossless uniform transmission lines are discussed earlier in Section 1.3.2, where their
complex propagation constant γ = jβ is purely imaginary, and their characteristic impedance Z0 is
20 2. THEORY OF CCITLS
purely real. The phasor voltage V (z) and phasor current I (z) of RLUTLs are given as, (see (1.18)
and (1.19))

V (z) = V0+ e−jβz + V0− ejβz , (2.10)

1  + −jβz 
I (z) = V0 e − V0− ejβz , (2.11)
Z0
where V0+ and V0− are defined as the amplitudes of incident and reflected voltage waves, respectively.
Comparing (2.10) and (2.11) to (2.1) and (2.2) for NRLUTLs, it is found that the RLUTL is a special
case of NRLUTLs when Z0+ = Z0− = Z0 and β + = β − = β. It should be pointed out that (2.7)
and (2.8) ( and Zin for NRLUTLs) are reduced to (1.31) and (1.33) ( and Zin for RLUTLs
with γ = jβ), respectively, when Z0+ = Z0− = Z0 and β + = β − = β. Note that Z0+ = Z0− = Z0
(which is purely real) implies that Z0+ = (Z0− )∗ . Thus, the RLUTL is a special case of CCITLs with
the identical and real Z0+ and Z0− ; i.e., the CCITL is a generalization of the RLUTL, which allow
us to handle broader class of useful TL problems. In the next section, exponentially tapered lossless
nonuniform TLs are revisited (discussed earlier in Section 1.4.1).

2.4 EXPONENTIALLY TAPERED LOSSLESS NONUNIFORM

TRANSMISSION LINES
In Section 1.4.1, it is found that the phasor voltage V (z) and the phasor current I (z) can be written
explicitly in terms of the superposition of two traveling waves propagating in the opposite directions
as follows (see (1.58) to (1.61), (1.63), and (1.64)):
  qz
V (z) = V1 e−jβz + V2 ejβz e 2 , (2.12)
 
V1 −jβz V2 qz
I (z) = +e − − ejβz e− 2 , (2.13)
Z0 Z0

where the subscripts 1 and 2 are associated with waves propagating in the +z and –z directions,
respectively, and (see (1.62), (1.65), and (1.67))
  q 2
β = ω 2 L0 C0 − , (2.14)
2
γI,1  ∗
Z0+ = = Z0− . (2.15)
j ωC0
Note that L0 and C0 are the inductance and capacitance per unit length at the load, respectively,
and γ I,1 is defined as in (1.61). For convenience in analysis, β is assumed to be real, and only
nonnegative values of β are considered. This can be achieved when the real taper factor q is chosen
√
as q ≤ 2ω L0 C0 , (see (2.14)). From (2.12), (2.13), and (2.15), it is found that ETLNUTLs are
√
reciprocal CCITLs with NNCRs for q ≤ 2ω L0 C0 . In the next section, finite reciprocal lossless
periodic TL structures are discussed as a useful and interesting example of CCITLs.
 2.5. FINITE RECIPROCAL LOSSLESS PERIODIC TL STRUCTURES 21

2.5 FINITE RECIPROCAL LOSSLESS PERIODIC

TRANSMISSION LINE STRUCTURES
Periodic structures of TLs have several applications in microwave technology; e.g., microwave filters,
slow-wave components, traveling-wave amplifier, phase shifters, and antennas. A periodic structure
usually refers to an infinite TL or waveguide periodically loaded with reactive elements [5]. Periodic
TL structures support slow-wave propagation, and they possess passband and stopband character-
istics similar to those of filters. However, periodic structures are always finite in length in practical
applications. Effects of finite size of periodic TL structures and load termination will be studied in
this section. It should be pointed out that periodically loaded TLs can also be designed to simul-
taneously exhibit effective negative refractive index (NRI) and negative group delay in the frequency
bands of interest by using a resonant circuit embedded within each loaded TL unit cell, which are
metamaterial TLs having useful engineering applications [6], [7].
In this section, only reciprocal lossless uniform TLs are considered as elements of each unit cell
of periodic TL structures. It is assumed that periodic TL structures are operated in passbands only,
where nonattenuating and propagating waves exist on periodic TL structures [5]. Each unit cell of
periodic TL structures can be symmetrical or unsymmetrical as well.
Figure 2.2 illustrates an example of an unsymmetrical unit cell of a lossless periodic TL struc-
ture.The unit cell consists of a reciprocal lossless uniform TL of length d loaded with lossless lumped
elements (C1 , L, C and L1 ) across the midpoint of the line. It is assumed that these lumped elements
have no physical length at the frequency band of interest. Note that this lossless unit cell is similar to
that in [6] for metamaterial TLs (see Figure 1 in [6]), which simultaneously exhibits effective NRI
and negative group delay. It should be pointed out that this periodic TL structure of the unit cell in
Figure 2.2 is dispersive. The characteristic impedance and the propagation constant of the unloaded
uniform TL are defined as Z1 and k, respectively, as shown in Figure 2.2.

Figure 2.2: An example of an unsymmetrical unit cell of a lossless periodic TL structure.

The traveling wave equations for the phasor voltage Vm and the phasor current Im in the
passsbands at the terminal of the mth unit cell of terminated reciprocal lossless periodic TL structures
22 2. THEORY OF CCITLS
as shown in Figure 2.3 (where m = 1, 2, …, M) can be written compactly as [5]

Vm = V0+ e−j mβd + V0− ej mβd , (2.16)

V0+ −j mβd V0− j mβd
Im = e + e , (2.17)
ZB+ ZB−

where V0+ and V0− are defined as the amplitudes of the incident and reflected voltage waves referenced
at the input of the periodic TL structure, respectively, and ZB± are referred to as the Bloch impedances.
Note that V0 and I0 in Figure 2.3 denote the total input voltage and the total input current of the

I0 I1 I2 IM

1
st
+ 2
nd
M th
+
V0 Unit V1
+
Unit
+
V2 ... Unit VM
+
ZL
− cell − − cell − cell −
Zin,M d
M Unit cells

Figure 2.3: A finite reciprocal lossless periodic TL structure of M unit cells terminated in a load
impedance ZL .

terminated finite reciprocal lossless periodic TL structure of M unit cells, respectively. V0 and I0 can
be obtained from (2.16) and (2.17) by letting m = 0, respectively. Using the ABCD matrix technique
(see Appendix A) and eigenanalysis [5], it can be shown rigorously that the Bloch impedances ZB±
can be expressed in terms of the total ABCD parameters as

−2B
ZB± =  , (2.18)
A − D ∓ (A + D)2 − 4

where ZB+ and ZB− are the Bloch impedances for forward and reverse traveling waves, respectively.
From Figure 2.2, the total ABCD matrix of the unsymmetrical unit cell can be calculated by using
the cascading property of the ABCD matrix [5].
In the passband operation, it is required that [5]

|A + D| ≤ 2. (2.19)

Thus, it is more convenient to express ZB± in (2.18) for the passband operation as

−2B
ZB± =  . (2.20)
A − D ∓ j 4 − (A + D)2
 2.5. FINITE RECIPROCAL LOSSLESS PERIODIC TL STRUCTURES 23
In (2.16) and (2.17), β is the effective propagation constant of the periodic TL structure operating
in passbands, and it can be determined from the following dispersion relation
A+D
cos βd = , (2.21)
2
which is a result of eigenanalysis of periodic TL structures [5]. Using (2.19), the magnitude of the
right-hand side of (2.21) is always less than or equal to 1.0 in passbands; i.e., β is always real in
passbands.
For a reciprocal lossless unit cell, it can be shown rigorously that its total ABCD matrix possesses
the following properties; i.e., A and D are purely real and B and C are purely imaginary. In addition,
for a symmetric unit cell, it can be shown rigorously that A and D are identical (A = D). Using (2.20)
and (2.21), the Bloch impedances ZB± and the effective propagation constant β of the periodic TL
structure with symmetric unit cells operating in passbands satisfy the following equations:
∓j B
ZB± = √ , (2.22)
1 − A2
cos βd = A, (2.23)

where A ≤ 1 (see (2.19)). It is observed that ZB± are purely real with different signs for reciprocal lossless
symmetric unit cells operating in passbands since B is purely imaginary and A ≤ 1.
In practice, periodic TL structures are finite in size, and they are usually connected to a load
impedance. Let us consider a finite reciprocal lossless periodic TL structure of M unit cells terminated
in a load impedance ZL as illustrated in Figure 2.3. The input impedance Zin,M at the input terminal
of the terminated reciprocal lossless periodic TL structure can be expressed compactly as (see the
concept in [8])
V0 1 + e−j 2Mβd
Zin,M ≡ = ZB+ ZB− − , (2.24)
I0 ZB + ZB+ e−j 2Mβd
where  is the voltage reflection coefficient at the load, which is mathematically given as

V0− ej Mβd ZL ZB− − ZB+ ZB−

≡ =− . (2.25)
V0+ e−j Mβd ZL ZB+ − ZB+ ZB−
Next, it will be shown that reciprocal CCITLs can be used to analyze finite reciprocal lossless periodic
TL structures.
Figure 2.4 illustrates a reciprocal CCITL, terminated in a passive load impedance ZL , pos-
sessing the unique propagation constant β(β + = β − = β in (2.1) and (2.2)) with conjugate charac-
teristic impedances, Z0+ and Z0− , for propagation in the forward and reverse directions, respectively.
Using (2.8) and (2.9) with β + = β − = β, the input impedance Zin of the terminated reciprocal
CCITL can be written as
1 + e−j 2β
Zin = Z0+ Z0− − , (2.26)
Z0 − Z0+ e−j 2β
24 2. THEORY OF CCITLS

Figure 2.4: A reciprocal CCITL terminated in a passive load impedance ZL .

where  is the voltage reflection coefficient at the load defined in (2.7). Comparing (2.24) to (2.26),
ZB± and Z0± are related as
ZB± = ±Z0± . (2.27)

It should be pointed out that, with the aid of (2.27), the expressions of  defined in (2.7) and (2.25)
are identical. Using (2.24) and (2.27), the input impedance Zin,M at the input terminal of the
terminated periodic TL structure shown in Figure 2.3 can be rewritten in terms of Z0± as

1 + e−j 2Mβd
Zin,M = Z0+ Z0− , (2.28)
Z0− − Z0+ e−j 2Mβd

where Z0± are defined as the effective characteristic impedances of the periodic TL structure.
From (2.28), it can be concluded that the equation of the input impedance of terminated finite
reciprocal lossless periodic TL structures is the same as that of terminated reciprocal CCITLs even
M = 1 (one unit cell). Thus, these periodic TL structures when considering at the input terminal of
each unit cell can be effectively analyzed using reciprocal CCITLs. Due to flexibility in fabricating
periodic TL structures and their usefulness in practice, there are several applications in microwave
technology as pointed out earlier. It should be pointed out that lossless periodic TL structures are
probably the most useful example of CCITLs in practice.
From (2.20), (2.21), and (2.27), the effective characteristic impedances Z0± and the effective
propagation constant β of the reciprocal lossless periodic TL structure are generally dependent on
various parameters of each unit cell of the reciprocal lossless periodic TL structure through the total
ABCD parameters of each unit cell in a complicated fashion. As pointed out earlier in Section 2.2,
the characteristic impedances Z0± of CCITLs must be complex conjugate of one another as shown
in (2.3). For convenience, Z0± are defined in a polar form as shown in (2.4). It is interesting
 ±  to
point out that reciprocal
  lossless periodic TL structures can exhibit both NNCRs (Re Z0 ≥ 0)
and NCRs (Re Z0± < 0) in passbands as shown in an example below. For periodic TL structures
 2.5. FINITE RECIPROCAL LOSSLESS PERIODIC TL STRUCTURES 25

with NNCRs, (2.4) implies that the argument φ of Z0− must lie in the following range:

−90◦ ≤ φ ≤ 90◦ . (2.29)

However, the argument φ for periodic TL structures with NCRs must lie in the following ranges:

−180◦ < φ < −90◦ ∪ 90◦ < φ ≤ 180◦ , (2.30)

where the symbol “∪” denotes the union in set theory. Note that, for periodic TL structures with
reciprocal lossless symmetric unit cells operating in passbands, Z0± are identical and purely real ; i.e., the
argument φ is equal to 0◦ for NNCRs or 180◦ for NCRs (see (2.22) and (2.27)).
From (2.28) and (2.7), it is obvious that the matching condition resulting in Zin,M = ZL
for terminated periodic TL structures can be obtained when ZL = Z0+ for the NNCR case and
ZL = −Z0− for the NCR case; i.e., the magnitude of the voltage reflection coefficient at the load,
−
||, is equal to zero and approaches infinity, respectively. It should −be pointed out that ZL = −Z0
for the NCR case can be realizable with passive loads since Re Z0 < 0. Appendix D shows the
derivation of || for both NNCR and NCR cases. It is interestingly found that || is always less
than or equal to unity for the NNCR case, and always greater than or equal to unity for the NCR
case. However, associated powers for both cases are still conserved for passive load terminations as
shown in detail in Appendix E.
An example of periodic TL structures exhibiting both NNCRs and NCRs is illustrated in
Figure 2.5, where the number of unit cells is equal to one (M = 1) for simplicity. The lossless
unsymmetrical unit cell is connected with a passive load and a generator with the source voltage Es and
the source impedance Zs .The unit cell is implemented by using capacitively loaded reciprocal lossless
uniform TLs with the propagation constant k and the characteristic impedance Z1 , where d/2 is the
length of each uniform TL as shown in Figure 2.5. In this example, d = 2 cm, Z1 = 150  , Cs = 5.6

Cs
Zs
Es
Z1 ,k CshZ1 ,k ZL

Zin d/2 d/2

Unit cell

Figure 2.5: A lossless unsymmetrical unit cell connected with a passive load and a generator.

pF, and Csh = 0.44 pF are employed. In this example, it is assumed that the phase velocity of wave
propagating along the uniform TLs is equal to 3×108 m/s. In addition, ZL = 200  is employed,
26 2. THEORY OF CCITLS
and the frequency of interest ranges from 0.5 to 10 GHz. The passbands can be observed when
| cosβd| ≤ 1 as shown in Figure 2.6. It is found that there are two passbands in this frequency
range; i.e., the 1st passband (0.67–4.15 GHz) and the 2nd passband (7.56– 9.73 GHz). Between the
1st and 2nd passbands, a stopband exists in the frequency range between 4.15 GHz and 7.56 GHz,
where | cosβd| ≥ 1 is observed.

1.6
1.4
1st passband 2nd passband
1.2
1.0
|cosβd|

0.8
0.6
0.4
0.2
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
f (GHz)

Figure 2.6: The plot of the magnitude of cos βd versus frequency.

Figures 2.7 and 2.8 plot the magnitudes and the arguments of Z0± versus frequency, respectively.
Note that φ + and φ − in Figure 2.8 are the arguments of Z0+ and Z0− respectively, and φ − is the
same as φ defined in (2.4). From Figures 2.7 and 2.8, it is observed that Z0+ and Z0− are complex
conjugate of each other in the passbands only as expected; i.e., |Z0+ | = |Z0− |. From Figure 2.8, it is
found that φ − is in the ranges given in (2.29) and (2.30) for the 1st and 2nd passbands, respectively;
i.e., the periodic TL structure (M ≥ 1) for the unit cell given in Figure 2.5 exhibits NNCRs in the
1st passband and NCRs in the 2nd passband. Away from the passbands, (i.e., stopbands), Z0± are
purely imaginary with different magnitudes. Figure 2.9 illustrates the plot of the magnitude of 
versus frequency. From the plot, it is found that || ≤ 1 and || ≥ 1, corresponding to NNCRs
and NCRs, are observed in the 1st and 2nd passbands, respectively. These results agree very well with
the derivation given in Appendix D.

2.6 CONCLUSIONS
In this chapter, the theory of CCITLs is discussed in detail. Practical examples of CCITLs are also
provided, and lossless periodic TL structures are probably the most useful example of CCITLs. It is
interestingly found that CCITLs can exhibit nonnegative characteristic resistances (Re{Z0± } ≥ 0)
and negative characteristic resistances (Re{Z0± } < 0). It is rigorously shown that the magnitude of
 2.6. CONCLUSIONS 27

Magnitude of Impedance (Ω) Ω)

10000
Z 0+
Z 0−
1000
1st passband
100
2nd passband

10

1
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
f (GHz)

Figure 2.7: The plot of the magnitudes of Z0+ and Z0− versus frequency.

Figure 2.8: The plot of φ + and φ − versus frequency.

the voltage reflection coefficient at the load is always less than or equal to unity for the NNCR
case, and always greater than or equal to unity for the NCR case. However, associated powers for
both cases are still conserved for passive load terminations. To gain more physical understanding
and simplify the analysis and design of CCITLs, graphical tools based on the Smith chart will be
developed in Chapter 4. In the next chapter, CCITLs will be generalized to include effects of loss
of associated TLs and/or lumped elements.
28 2. THEORY OF CCITLS

2.6
2.4
2.2
2.0 1st passband
1.8
1.6

|Γ|
1.4
1.2
1.0
0.8
0.6 2nd passband
0.4
0.2
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
f (GHz)

Figure 2.9: The plot of the magnitude of  versus frequency.

BIBLIOGRAPHY
[1] S. Lamultree and D. Torrungrueng, “On the characteristics of conjugately characteristic-
impedance transmission lines with active characteristic impedance,” in 2006 Asia-Pacific Mi-
crowave Conference Proceedings, Dec. 2006, vol. 1, pp. 225–228.
DOI: 10.1109/APMC.2006.4429411 17

[2] D. Torrungrueng, S. Lamultree, C. Phongcharoenpanich and M. Krairiksh, “In-depthanalysis

of reciprocal periodic structures of transmission lines,” IET Microwaves,Antennas & Propa-
gation, vol. 3, no. 4, pp. 591–600, Jun. 2009. DOI: 10.1049/iet-map.2008.0205 17

[3] I. V. Lindell, M. Valtonen, and A. H. Sihvola, “Theory of nonreciprocal and nonsymmetri-

cuniform transmission lines,” IEEE Transactions on Microwave Theory and Techniques,vol. 42,
no. 2, pp. 291–297, Feb. 1994. DOI: 10.1109/22.275260 18

[4] I. V. Lindell, A. H. Sihvola, S. A. Tretyakov, and A. J. Viitanen, Electromagnetic Wavesin Chiral

and Bi-Isotropic Media. Massachusetts: Artech House, 1994. 18

[5] D. M. Pozar, Microwave Engineering, 3rd ed. New Jersey: Wiley, 2005. 21, 22, 23, 29

[6] O. F. Siddiqui, M. Mojahedi, and G. V. Eleftheriades, “Periodically loaded transmissionline

with effective negative refractive index and negative group velocity,” IEEE Transactions on
Antennas and Propagation, vol. 51, no. 10, pp. 2619–2625, Oct. 2003.
DOI: 10.1109/TAP.2003.817556 21

[7] G. V. Eleftheriades and K. G. Balmain, Negative-Refraction Metamaterials. New Jersey: Wiley

Interscience, 2005. 21
 2.7. PROBLEMS 29
[8] N. Chamnandechakun and D. Torrungrueng, “An Analytical Study of Terminated FinitePe-
riodic Structures,” in Proceeding of the 27 th Electrical Engineering Conference, Nov. 2004, vol. 2,
pp. 369–372. 23

2.7 PROBLEMS
2.1. Show that the voltage reflection coefficient  at the load and the input impedance Zin of
NRLUTLs as shown in Figure 2.1 are given as in (2.7) and (2.8), respectively.
2.2. Show that the phasor voltage V (z) and the phasor current I (z) distributed along ETL-
NUTL scan be expressed explicitly in terms of the superposition of two traveling waves
propagating in the opposite directions as shown in (2.12) and (2.13), respectively.
2.3. Using the ABCD matrix technique and eigenanalysis [5], show that the formula of the
Bloch impedances ZB± in terms of the total ABCD parameters of the unit cell of reciprocal
lossless periodic TL structures is given as (2.18).
2.4. Determine the total ABCD parameters of the unsymmetrical unit cell in Figure 2.2 as a
function of the angular frequency ω. Find the determinant of the total ABCD matrix. Is the
determinant equal to unity independent of ω? Explain in detail.
2.5. Show that the total ABCD matrix of a reciprocal lossless unit cell possesses the following
properties: A and D are purely real and B and C are purely imaginary.
2.6. Show that the total ABCD matrix of a symmetric unit cell possesses the following property:
A and D are identical.
2.7. Derive (2.24) and (2.25) for the terminated reciprocal lossless periodic TL structure.
2.8. Verify that the matching condition for terminated periodic TL structures can be obtained
when ZL = Z0+ for the NNCR case and ZL = −Z0− for the NCR case.
2.9. Determine the total ABCD parameters of the lossless unsymmetrical unit cell in Figure 2.5
as a function of the angular frequency ω by using d = 2 cm, Z1 = 150  , Cs = 5.6 pF, and
Csh = 0.44 pF. It is assumed that the phase velocity of wave propagating along the uniform
TLs is equal to 3×108 m/s. In addition, plot both magnitude and phase of each ABCD
parameter from 0.5 to 10 GHz.
2.10. Verify the plots in Figures 2.6 to 2.9.
 31

CHAPTER 3

Theory of
Bi-Characteristic-Impedance
Transmission Lines (BCITLs)
3.1 INTRODUCTION
In Chapter 2, the theory of CCITLs is developed in detail. Due to the fact that CCITLs are lossless,
they cannot account for losses associated with lossy TLs and/or lossy lumped elements to obtain
higher accuracy in modeling practical TL problems.Thus, one needs to resort to a more general model
for these lossy elements, which is the model based on bi-characteristic-impedance transmission lines
(BCITLs). By definition, BCITLs are generally lossy transmission lines possessing two complex
characteristic impedances Z0+ and Z0− for forward and reverse waves respectively, where Z0± are
generally different and not complex conjugate of one another, as shown in Figure 3.1. In addition,
BCITLs generally possess different complex propagation constants γ + and γ − for forward and
reverse waves, respectively. In Figure 3.1, the BCITL of length  is terminated in a passive load
impedance ZL , where  is the voltage reflection coefficient at the load and Zin is the input impedance
of the terminated BCITL. Using the same concept as in Appendix B, it can be shown rigorously
that γ +  = γ − is the required condition for nonreciprocal BCITLs.

Figure 3.1: A BCITL terminated in a passive load impedance ZL .

32 3. THEORY OF BCITLS

It should be pointed out that the CCITL is a special case of the BCITL when Z0+ and
Z0− are complex conjugate of one another; i.e., when losses associated with the BCITL vanish,
the BCITL is essentially the CCITL.Examples of BCITLs are RLSUTLs, nonreciprocal lossy
uniform transmission lines (NRLSUTLs), finite reciprocal lossy periodic TL structures and active
loaded TLs [1]–[3]. Note that RLSUTLs are discussed in detail earlier in Section 1.3.1, where their
characteristic impedances Z0 are complex. Thus, the RLSUTL is a special case of BCITLs with the
identical Z0+ and Z0− ; i.e., Z0+ = Z0− = Z0 . For active loaded TLs, active devices (e.g., microwave
transistor amplifiers) are connected to transmission lines in cascade, which are not discussed in this
book. Details of active loaded TLs can be found in [3]. In this chapter, only NRLSUTLs and finite
reciprocal lossy periodic TL structures are discussed.
This chapter is organized as follows. Section 3.2 presents nonreciprocal lossy uniform trans-
mission lines. Finite reciprocal lossy periodic TL structures are discussed in Section 3.3. Finally,
conclusions are provided in Section 3.4.

3.2 NONRECIPROCAL LOSSY UNIFORM TRANSMISSION

LINES
In Figure 3.1, the traveling wave equations for the phasor voltage V (z) and the phasor current I (z)
along NRLSUTLs can be written as
+z −z
V (z) = V0+ e−γ + V0− eγ , (3.1)
V0+ −γ + z V0− γ − z
I (z) = e − e . (3.2)
Z0+ Z0−

Note that V0+ and V0− are the voltage amplitudes for incident and reflected waves at the load (z=
0), respectively. Z0+ and Z0− are generally different, and γ + generally differs from γ − as well. Thus,
NRLSUTLs are nonreciprocal BCITLs. It is observed that these voltage and current equations for
NRLSUTLs are similar to those for NRLUTLs (see (2.1) and (2.2)). Note that the exponential
factors in (3.1) and (3.2) include loss effects of transmission lines, which are more general than those
in (2.1) and (2.2).
For convenience in constructing a graphical tool for analysis and design of BCITLs in Chap-
ter 4, the characteristic impedances Z0± of NRLSUTLs are represented in the rectangular form
as

Z0+ = P + j Q, (3.3)
Z0− = U + j V , (3.4)

where P and Q are the real and imaginary parts of the complex characteristic impedance Z0+ respec-
tively, and U and V are the real and imaginary parts of the complex characteristic impedance Z0− ,
respectively. Applying the load condition and using (3.1) and (3.2), the voltage reflection coefficient
 3.3. FINITE RECIPROCAL LOSSY PERIODIC TL STRUCTURES 33
at the load  as shown in Figure 3.1 is found to be

V0− ZL Z0− − Z0+ Z0−

≡ = (3.5)
V0+ ZL Z0+ + Z0+ Z0−

In addition, the input impedance Zin can be computed from  as follows:

V (z = −) 1 + e−2γ̃ 
Zin ≡ = Z0+ Z0− − , (3.6)
I (z = −) Z0 − Z0+ e−2γ̃ 

where (3.1), (3.2), and (3.5) are employed. Note that γ̃ in (3.6) is the arithmetic mean of γ ± , which
can be expressed as
1 + 
γ̃ = γ + γ− . (3.7)
2
It is observed that the formulas of  for both NRLSUTLs and NRLUTLs are identical (see (3.5)
and (2.7)). However, the formula of Zin for NRLSUTLs is different from that of NRLUTLs in
that the former employs the arithmetic mean of γ ± while the latter uses the arithmetic mean of β ±
(See (2.9)) as shown in (3.6) and (2.8), respectively.

3.3 FINITE RECIPROCAL LOSSY PERIODIC

TRANSMISSION LINE STRUCTURES
In Section 2.5, finite reciprocal lossless periodic TL structures are discussed in detail. However,
periodic TL structures may contain lossy elements in practice; e.g., lossy TLs and lossy lumped
elements. Figure 3.2 illustrates an example of an unsymmetrical unit cell associated with reciprocal
lossy periodic TL structures. The unit cell consists of a length d of reciprocal lossless uniform
transmission line loaded with lossy lumped resistors (RS and RSH ) across the midpoint of the line.
It is assumed that these lumped resistors have no physical length at the frequency band of interest.The
characteristic impedance and the propagation constant of the unloaded reciprocal lossless uniform
transmission line are defined as Z1 and k, respectively.
The wave equations for the phasor voltage Vm and the phasor current Im at the terminal of
the mth unit cell as shown in Figure 3.3 (where m = 1, 2, …, M) can be written compactly as [4]

Vm = V0+ e−mγ d + V0− emγ d , (3.8)

V0+ −mγ d V0− mγ d
Im = e + e , (3.9)
ZB+ ZB−

where V0+ and V0− are defined as the amplitudes of the incident and reflected voltage waves referenced
at the input of the reciprocal lossy periodic TL structure, respectively. Note that V0 and I0 in
Figure 3.3 denote the total input voltage and the total input current of the terminated periodic TL
structure of M unit cells, respectively. V0 and I0 can be obtained from (3.8) and (3.9) by letting m
34 3. THEORY OF BCITLS

Figure 3.2: An example of an unsymmetrical unit cell associated with reciprocal lossy periodic TL struc-
tures.

I0 I1 I2 IM

1
st
+ + 2
nd
M th
V ...
+ + +
V0 Unit V1 Unit 2 Unit VM ZL
− cell − − cell − cell −
Zin,M d
M Unit cells

Figure 3.3: A finite reciprocal lossy periodic TL structure of M unit cells terminated in a load impedance
ZL .

= 0, respectively. In (3.9), ZB± are referred to as the Bloch impedances, which can be expressed in
terms of the total ABCD parameters of the unit cell as

−2B
ZB± =  , (3.10)
A − D ∓ j 4 − (A + D)2

which is identical to (2.20) for the lossless case. Note that ZB+ and ZB− are the Bloch impedances
for forward and reverse waves, respectively. In (3.8) and (3.9), γ = α + jβ is the effective complex
propagation constant of the reciprocal lossy periodic TL structure, where α and β are the effective
attenuation constant and effective propagation constant of the structure, respectively. Mathematically,
γ can be determined from the following equation [4]

A+D
cosh (γ d) = . (3.11)
2
 3.3. FINITE RECIPROCAL LOSSY PERIODIC TL STRUCTURES 35
Let us consider a finite reciprocal lossy periodic TL structure of M unit cells terminated in a
load impedance ZL as illustrated in Figure 3.3. It can be shown that the input impedance Zin,M at
the input terminal of the terminated structure can be expressed as (see (2.24))

V0 1 + e−2Mγ d
Zin,M ≡ = ZB+ ZB− − , (3.12)
I0 ZB + ZB+ e−2Mγ d
where  is the voltage reflection coefficient at the load, which is given by

ZL ZB− − ZB+ ZB−

=− . (3.13)
ZL ZB+ − ZB+ ZB−
Note that reciprocal BCITLs can be used to model finite reciprocal lossy periodic TL structures as
shown in Figure 3.1 with γ + = γ − = γ . Using (3.6) with γ + = γ − = γ , the input impedance Zin
of the terminated reciprocal BCITL can be written as

1 + e−2γ 
Zin = Z0+ Z0− , (3.14)
Z0− − Z0+ e−2γ 

where  is the voltage reflection coefficient at the load defined as in (3.5). Comparing (3.12)
to (3.14), ZB± and Z0± are related as
ZB± = ±Z0± . (3.15)
Using (3.12) and (3.15), the input impedance Zin,M at the input terminal of the terminated structure
in Figure 3.3 can be rewritten in terms of Z0± as

1 + e−2Mγ d
Zin,M = Z0+ Z0− , (3.16)
Z0− − Z0+ e−2Mγ d

where Z0± in (3.16) are defined as the effective characteristic impedances of the reciprocal lossy
periodic TL structure. From (3.16), it can be concluded that the equation of the input impedance of
terminated finite reciprocal lossy periodic TL structures is the same as that of terminated reciprocal
BCITLs. Thus, these lossy periodic TL structures when considering at the input terminal of each unit
cell can be effectively analyzed using reciprocal BCITLs. Like CCITLs, lossy periodic TL structures
are probably the most useful example of BCITLs in practice.
It is interesting to point out that reciprocal lossy periodic TL structures can exhibit both
NNCRs (Re{Z0± } ≥ 0) and NCRs (Re{Z0± } < 0) as shown in an example below. For reciprocal
lossy periodic TL structures with NNCRs, (3.3) and (3.4) implies that P ≥ 0 and U ≥ 0. For
those with NCRs, P < 0 and U < 0 are required. Like the CCITL case (see Section 2.5), it can
be shown using (3.5) and (3.16) that the matching condition resulting in Z in,M = ZL for reciprocal
lossy periodic TL structures can be obtained when ZL = Z0+ for the NNCR case and ZL = −Z0−
for the NCR case; i.e., the magnitude of the voltage reflection coefficient at the load, ||, is equal to
zero and approaches infinity, respectively.
36 3. THEORY OF BCITLS
An example of reciprocal lossy periodic TL structures of Figure 3.2 terminated in a passive
load impedance ZL is shown in Figure 3.4, where the number of unit cells is equal to one (M=1)
for simplicity. Note that Zin is the input impedance of the terminated reciprocal lossy periodic TL
structures. In this example, the parameters in Figure 3.4 are given as follows: d = 2 cm, Z1 = 50 ,
RS = 75 , RSH = 200  and ZL = 50  [5]. It is assumed that the phase velocity of wave propagating
along the reciprocal lossless uniform TL is equal to 3×108 m/s. The frequency of interest ranges from
1 GHz to 10 GHz.

Figure 3.4: An example of an unsymmetrical unit cell associated with reciprocal lossy periodic TL struc-
tures (M = 1) terminated in a passive load impedance ZL .

Using (3.10) and (3.15), the plots of the magnitudes and the arguments of Z0± versus frequency
are shown in Figures 3.3 and 3.6, respectively. Note that φ + and φ − in Figure 3.6 are the arguments
of Z0+ and Z0− , respectively. From Figure 3.6, the reciprocal lossy periodic TL structure (M≥1) for
the unit cell given in Figure 3.4 exhibits NNCRs (−90◦ ≤ φ ± ≤ 90◦ ) in the frequency range of 1.0
to 7.5 GHz and NCRs (−180◦ < φ ± < −90◦ ∪ 90◦ < φ ± ≤ 180◦ ) in the frequency range of 7.5
to 10 GHz. It is observed from Figures 3.3 and 3.6 that different magnitudes and phases of complex
characteristic impedances, Z0+ and Z0− , are obtained as expected.
Using (3.11) to determine the effective complex propagation constant γ = α + jβ, it is ob-
served that α is positive and negative for NNCRs and NCRs, respectively, as shown in Figure 3.7.
Using the method based on short-circuited and open-circuited terminations [5], [6], it can be shown
rigorously that α of BCITLs, implemented using finite reciprocal lossy periodic TL structures, is al-
ways nonnegative and nonpositive for NNCRs and NCRs, respectively. In addition, Figure 3.7 shows
that β is positive in the frequency range of interest for both NNCRs and NCRs. Furthermore,
Figure 3.8 illustrates the plot of the magnitude of  versus frequency. From the plot, it is observed
that || ≤ 1 and || ≥ 1 for NNCRs and NCRs in this example, respectively. However, it should
be pointed out that this observation for the magnitude of  is not valid for general BCITLs and
general passive load impedances.
 3.4. CONCLUSIONS 37

100
+

Magnitudes of Impedance (Ω)

90 Z0
−
80 Z0
70
60
50
40
30
20
10
NNCR NCR
0
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
Frequency (GHz)

Figure 3.5: The plot of the magnitudes of Z0+ and Z0− versus frequency.

3.4 CONCLUSIONS
In this chapter, the theory of BCITLs is discussed in detail. Practical examples of BCITLs are also
provided, and lossy periodic TL structures are probably the most useful example of BCITLs. Like
CCITLs, it is interestingly found that BCITLs can exhibit nonnegative characteristic resistances
(Re{Z0± } ≥ 0) and negative characteristic resistances (Re{Z0± } < 0). In addition, the effective at-
tenuation constant α of BCITLs, implemented using finite reciprocal lossy periodic TL structures,
is always nonnegative and nonpositive for NNCRs and NCRs, respectively. To gain more physical
understanding and simplify the analysis and design of BCITLs, graphical tools based on the Smith
chart will be developed in Chapter 4.

BIBLIOGRAPHY
[1] D. Torrungrueng, P.Y. Chou, and M. Krairiksh, “A graphical tool for analysis and designof bi-
characteristic-impedance transmission lines,” Microwave and OpticalTechnologyLetters, vol. 49,
no. 10, pp. 2368–2372, Oct. 2007. DOI: 10.1002/mop.22801 32
[2] D. Torrungrueng, P.Y. Chou, and M. Krairiksh, “Erratum:A graphical tool for analysis and
design of bi-characteristic-impedance transmission lines,” Microwave and OpticalTechnolo-
gyLetters, vol. 51, no. 4, pp. 1154, Apr. 2009. DOI: 10.1002/mop.22801
[3] C. Lertsirimitand D. Torrungrueng, “Analysis of active loaded transmission line usingan equiv-
alent BCITL model,” in 2007 Asia-Pacific Microwave Conference Proceedings, Dec. 2007, vol. 4,
pp. 2353–2356. DOI: 10.1109/APMC.2007.4554830 32
38 3. THEORY OF BCITLS

Figure 3.6: The plot of φ + and φ − versus frequency.

40 160

30 140

20 120
α α (neper/m)

β (β (radian/m)
10 100

)
0 80
NCR
-10 60

-20 40

-30 NNCR 20

-40 0
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0

Frequency (GHz)

Figure 3.7: The plot of α and β versus frequency.

[4] D. M. Pozar, Microwave Engineering, 3rd ed. New Jersey: Wiley, 2005. 33, 34

[5] S. Lamultree, D.Torrungrueng and C. Phongcharoenpanich, “Measurement of BCITLparam-

eters,” in Proceedings of the 2008 ECTI International Conference, May 2008, vol. 1, pp. 257–260.
DOI: 10.1109/ECTICON.2008.4600421 36
 3.5. PROBLEMS 39

3.0
NNCR
2.5

2.0
|Γ|

1.5

1.0
NCR
0.5

0.0
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
Frequency (GHz)

Figure 3.8: The plot of the magnitude of  versus frequency.

[6] D. Torrungrueng, S. Lamultree, C. Phongcharoenpanich and M. Krairiksh, “In-depth analysis

of reciprocal periodic structures of transmission lines,” IET Microwaves,Antennas & Propaga-
tion, vol. 3, no. 4, pp. 591–600, Jun. 2009. DOI: 10.1049/iet-map.2008.0205 36

3.5 PROBLEMS
3.1. Show that the input impedance Zin of NRLSUTLs as shown in Figure 3.1 is given as
in (3.6).
3.2. Determine the total ABCD parameters of the lossy unsymmetrical unit cell in Figure 3.4 as
a function of the angular frequency ω by using d = 2 cm, Z1 = 50 , RS = 75  and RSH =
200 . It is assumed that the phase velocity of wave propagating along the uniform TL is
equal to 3×108 m/s. In addition, plot both magnitude and phase of each ABCD parameter
from 1 to 10 GHz.
3.3. Verify the plots in Figures 3.5 to 3.8.
 41

CHAPTER 4

Meta-Smith Charts for CCITLs

and BCITLs
4.1 INTRODUCTION
Analytical approaches shown in Chapters 2 and 3 for solving problems associated with CCITLs
and BCITLs usually provide quite complex formulae. In order to assist in the analysis and design of
CCITLs and BCITLs, the need for graphical tools is obvious. Like the Smith chart, graphical tools
usually provide a useful way with more physical insight of visualizing associated TL phenomenon
and solving related problems effectively. In this chapter, the Smith chart is generalized to problems
associated with CCITLs and BCITLs, called Meta-Smith charts. Note that the Greek word meta
means beyond. Meta-Smith charts can be applied to more practical and useful TL problems; i.e.,
CCITLs and BCITLs, beyond the capability of the Smith chart. The Meta-Smith charts are con-
structed using the same approach as in constructing the Smith chart. In this chapter, the Meta-Smith
charts for CCITLs with NNCRs and NCRs are developed, including the Meta-Smith chart for
BCITLs. Only passive loads (Re{ZL } ≥ 0) are of interest, where ZL is the load impedance.
In the references [1]–[7], the authors employed the terminology “T-Charts” for graphical
tools associated with CCITLs and BCITLs. In this book, the terminology“Meta-Smith charts” is
used instead of “T-Charts” to make readers, who are familiar with the Smith chart, see the originality
and generality of the proposed graphical tools, Meta-Smith charts. In addition, it should be pointed
out that in the literature [2], [8] the terminologies “passive characteristic impedances” and “active
characteristic impedances” refer to NNCRs and NCRs, respectively. In this book, NNCRs and
NCRs are employed instead to avoid any misleading of those terminologies, especially for active
characteristic impedances.
For convenience in derivation later, let Z0± be the characteristic impedances of both CCITLs
and BCITLs. For CCITLs, Z0+ and Z0− are complex conjugate of one another. For BCITLs, Z0+
and Z0− are generally different and not complex conjugate of one another. However, it is found in
Chapters 2 and 3 that the voltage reflection coefficient at the load  of both CCITLs and BCITLs
is given compactly in terms of ZL and Z0± as (see (2.7) and (3.5))

ZL Z0− − Z0+ Z0−

= . (4.1)
ZL Z0+ + Z0+ Z0−
Like the Smith chart, each Meta-Smith chart is essentially a polar plot of the voltage reflection
coefficient at the load  of CCITLs or BCITLs. It can be employed to convert from voltage
42 4. META-SMITH CHARTS FOR CCITLS AND BCITLS
reflection coefficients to normalized impedances (or admittances), and vice versa. For convenience
in studying CCITLs and BCITLs, computerized Meta-Smith chart programs are developed, and
they are employed in generating Meta-Smith charts in this book.
This chapter is organized as follows. The Meta-Smith chart for CCITLs with NNCRs is
discussed in detail in Section 4.2, and Section 4.3 presents the Meta-Smith chart for CCITLs with
NCRs. The Meta-Smith chart for BCITLs with NNCRs and NCRs is illustrated in Section 4.4.
Finally, conclusions are provided in Section 4.5.

4.2 THE META-SMITH CHART FOR CCITLS WITH

NONNEGATIVE CHARACTERISTIC RESISTANCES
The Meta-Smith chart for CCITLs with NNCRs (Re{ZL } ≥ 0) [1], [9] is constructed by using
the same concept as in constructing the standard ZY Smith chart in [10] (see Problem 1.3 at the
end of Chapter 1 as well), and it is discussed in detail in this section. Note that this Meta-Smith
chart is a superposition of the Z Meta-Smith chart and the Y Meta-Smith chart. The Z Meta-Smith
chart consists of resistance and reactance circles in the  plane, while the Y one is a superposition
of conductance and susceptance circles in the  plane.
For convenience, let us define Z0± in the polar form as

Z0± = Z0± e∓j φ , (4.2)

where |Z0± | and φ are the absolute value and the argument of Z0− , respectively. As pointed out earlier
in Chapter 2, the argument φ for CCITLs with NNCRs must lie in the following range:

−90◦ ≤ φ ≤ 90◦ . (4.3)

Due to the fact that the magnitude of the voltage reflection coefficient at the load || is always less
than or equal to unity for terminated CCITLs with NNCRs, the region of interest is always inside
or on the unit circle in the  plane for passive load terminations (see Appendix D).

4.2.1 THE Z META-SMITH CHART FOR CCITLS WITH NNCRS

For CCITLs, the normalized impedance z is defined as [1]

Z
z≡ , (4.4)
Z̃0

where Z̃0 is the geometric mean of the characteristic impedances Z0± ; i.e.,

Z̃0 = Z0+ Z0− = Z0± . (4.5)
 4.2. THE META-SMITH CHART FOR CCITLS WITH NNCRS 43

From (4.5), it should be noted that Z̃0 is always real and positive for physically realizable CCITLs.
Using (4.1) and (4.4),  can be rewritten compactly as

zL z0− − 1
= , (4.6)
zL z0+ + 1

where zL is the normalized load impedance and z0± are the normalized characteristic impedances.
Using (4.2), (4.4), and (4.5), z0± are found to be

z0± = e∓j φ . (4.7)

Note that z0± in (4.7) depend only on the argument φ only. To derive equations for resistance and
reactance circles, it is better to express zL explicitly in terms of . Using (4.6), zL can be expressed
compactly as
1+
zL = − . (4.8)
z0 − z0+ 

For convenience in manipulation, let us express  and zL as  = r + j i and zL = rL + jx L , where

r and i are the real and imaginary parts of  respectively, and rL and xL are the real and imaginary
parts of zL , respectively. Substituting these expressions of  and zL into (4.8) and using (4.7) yield

1 + r + j i
R L + j xL = . (4.9)
ej φ − e−j φ (r + j i )

After some straightforward algebraic manipulation, rL and xL in (4.9) can be expressed in terms of
r , i , and φ as follows:

cos φ − r2 cos φ − i2 cos φ

rL = , (4.10)
1 − 2r cos 2φ − 4i cos φ sin φ + r2 + i2
− sin φ − 2r sin φ + 2i cos φ − r2 sin φ − i2 sin φ
xL = . (4.11)
1 − 2r cos 2φ − 4i cos φ sin φ + r2 + i2

Rearranging (4.10) and (4.11) appropriately, the following circle equations are obtained [1]:

2 2 2
rL cos 2φ rL sin 2φ cos φ
r − + i − = , (4.12)
rL + cos φ rL + cos φ rL + cos φ
2 2 2
(xL cos 2φ − sin φ) (xL sin 2φ + cos φ) cos φ
r − + i − = , (4.13)
xL + sin φ xL + sin φ xL + sin φ

for resistance and reactance circles, respectively.

44 4. META-SMITH CHARTS FOR CCITLS AND BCITLS
4.2.2 THE Y META-SMITH CHART FOR CCITLS WITH NNCRS
To construct the Y Meta-Smith Chart, the normalized admittance y must be considered instead of
the normalized impedance z, where y is defined as
1
y≡ . (4.14)
z
Let yL = gL + jb L be the normalized load admittance, where gL and bL are the normalized load
conductance and susceptance, respectively. Using the same technique as in constructing the Z Meta-
Smith Chart above, the equations for conductance and susceptance circles (see (4.15) and (4.16),
respectively) are obtained as follows [1]:
2 2
gL cos φ
r + + i2 = , (4.15)
gL + cos φ gL + cos φ
2 2
cos φ cos φ
(r + 1)2 + i + = . (4.16)
bL − sin φ bL − sin φ
Note that these four circle equations (see (4.12), (4.13), (4.15), and (4.16)) depend on the argument
φ of Z0± in a complicated fashion. It is interesting to point out that these circle equations reduce
to the corresponding circle equations for the standard ZY Smith chart when φ = 0◦ as expected
(see Problem 1.3 at the end of Chapter 1) i.e., the standard ZY Smith chart is a special case of
the Meta-Smith chart for CCITLs with NNCRs. After deriving these equations, the Meta-Smith
chart for CCITLs with NNCRs can be constructed readily. In the next subsection, numerical results
of the Meta-Smith chart will be illustrated.

4.2.3 NUMERICAL RESULTS

From Sections 4.2.1 and 4.2.2, it is found that the Meta-Smith chart for CCITLs with NNCRs
depends on the argument φ. To illustrate this dependence, let us consider the Meta-Smith chart
when φ = 30◦ . Figure 4.1 shows the plot of resistance circles in the  plane for φ = 30◦ . Note that
each circle corresponds to different values of normalized load resistances rL . The largest circle occurs
when rL = 0. As rL increases, the radius of resistance circle becomes smaller as shown in Figure 4.1.
The resistance circles between the rL = 0 circle and the rL = 1 circle correspond to the circles with
0 < rL < 1. In addition, the resistance circles inside the rL = 1 circle correspond to the circles with
rL > 1.
Define θ as the angle between the horizontal line and the line OP drawn from the origin O
in the  plane to the touching point P of all resistance circles as shown in Figure 4.1. Using (4.12), it
can be shown rigorously that the relationship between the argument φ and the angle θ is given by

θ = 2φ. (4.17)

In this case, θ is equal to 60◦ for φ = 30◦ as shown in Figure 4.1. It should be emphasized that these
resistance circles of the Meta-Smith chart cannot be obtained by simply rotating the corresponding
 4.2. THE META-SMITH CHART FOR CCITLS WITH NNCRS 45

Figure 4.1: Plot of resistance circles in the  plane for φ = 30◦ .

circles of the standard Smith chart by the angle θ in the counterclockwise direction. This is due to the
fact that the scale of each resistance circle depends on the argument φ as well.
Figure 4.2 illustrates the plot of reactance circles in the  plane for φ = 30◦ . From the figure,
the horizontal line and the line OP drawn from the origin O in the  plane to the touching point
P of reactance circles intersect at the angle θ defined in (4.17). Note that the point P in Figures 4.4
and 4.2 is the same point. Unlike the standard Smith chart, the reactance circle xL = 0 for the
Meta-Smith chart is a curve (i.e., a part of circle) instead of a horizontal line as in the standard
Smith chart [10]. Inside the unit circle in Figure 4.2, all reactance circles above the circle xL = 0
correspond to xL > 0, and those below the circle xL = 0 correspond to xL < 0.
The Y Meta-Smith chart consists of conductance and susceptance circles as shown in Fig-
ures 4.3 and 4.4 for φ = 30◦ , respectively. Interestingly, the touching point of conductance circles
in Figure 4.3 is at the same position as in the Y Smith Chart (i.e., at the point (–1,0) in the 
plane) [10]. It is found that this point will not change when the argument φ changes. Each circle
corresponds to different values of normalized load conductances gL . The largest circle, which is the
unit circle, occurs when gL =0. As gL increases, the radius of conductance circle becomes smaller
as shown in Figure 4.3. The conductance circles between the gL = 0 circle and the gL = 1 circle
correspond to the circles with 0 < gL < 1. In addition, the conductance circles inside the gL = 1
circle correspond to the circles with gL > 1.
The plot of susceptance circles in the  plane for φ = 30◦ is illustrated in Figure 4.4. Like
reactance circles in Figure 4.2, the susceptance circle bL = 0 for the Y Meta-Smith chart as shown in
46 4. META-SMITH CHARTS FOR CCITLS AND BCITLS

Figure 4.2: Plot of reactance circles in the  plane for φ = 30◦ .

Figure 4.4 is also a curve (i.e., a part of circle) instead of a straight line as in the case of the Y Smith
chart [10]. Inside the unit circle in Figure 4.4, all susceptance circles above and below the circle bL =
0 correspond to bL < 0 and bL > 0, respectively. As pointed out earlier, the Y Meta-Smith chart is
a superposition of conductance and susceptance circles in the  plane. Note that the Y Meta-Smith
chart looks similar to the Y Smith chart, but its admittance scale is totally different.
The ZY Meta-Smith chart for CCITLs with NNCRs can be obtained by superimposing these
four families of circles in Figures 4.1 to 4.4 in the  plane as shown in Figure 4.5 for φ = 30◦ as an
example. From the figure, it is observed that the circle xL = 0 and the circle bL = 0 are identical. This
can be verified by comparing (4.13) and (4.16) when xL = 0 and bL = 0, respectively. It is interesting
to note that the circle xL = 0 joins the touching point of resistance (or reactance) circles (i.e., the
open-circuited (OC) point) and the touching point of conductance (or susceptance) circles (i.e., the
short-circuited (SC) point) as shown in Figure 4.5. In addition, it should be pointed out that the
procedures of using the ZY Meta-Smith chart in determining the input impedance of terminated
CCITLs with NNCRs are similar to those for reciprocal lossless uniform transmission lines using
the Smith chart provided in [10], [11], and they will be discussed in detail in Chapter 5.
It should be pointed out that the ZY Meta-Smith chart for CCITLs with NNCRs strongly
depends on the definition of the normalized impedance defined in (4.4). Using different Z̃0 results
in different graphical representations of the ZY Meta-Smith chart for CCITLs with NNCRs [9].
For example, if the arithmetic mean of the characteristic impedances Z0± is employed as Z̃0 instead
 4.2. THE META-SMITH CHART FOR CCITLS WITH NNCRS 47

Figure 4.3: Plot of conductance circles in the  plane for φ = 30◦ .

of the geometric mean of Z0± (see (4.5)); i.e.,

1 + 
Z̃0 = Z0 + Z0− = Z0± cos φ, (4.18)
2
the new four circle equations of resistance, reactance, conductance and susceptance can be obtained
as follows [9]:
2 2 2
rL cos 2φ rL sin 2φ 1
r − + i − = , (4.19)
rL + 1 rL + 1 rL + 1
2 2 2
(xL cos 2φ − tan φ) (xL sin 2φ + 1) 1
r − + i − = , (4.20)
xL + tan φ xL + tan φ xL + tan φ
2 2
gL cos2 φ
r + + i2 = , (4.21)
gL + cos2 φ gL + cos2 φ
2 2
cos2 φ cos2 φ
(r + 1) + i +
2
= , (4.22)
bL − sin φ cos φ bL − sin φ cos φ

respectively. Comparing these four circle equations with those of using the geometric mean of Z0± as
Z̃0 (see (4.12), (4.13), (4.15), and (4.16)), it is found that the corresponding equations are different
indeed. However, different ZY Meta-Smith charts must yield the same results; e.g., calculated
48 4. META-SMITH CHARTS FOR CCITLS AND BCITLS

Figure 4.4: Plot of susceptance circles in the  plane for φ = 30◦ .

input impedances of terminated CCITLs with NNCRs. Thus, it can be concluded that the ZY
Meta-Smith chart for CCITLs with NNCRs is not unique. More details of nonuniqueness of ZY
Meta-Smith charts can be found in [9].

4.3 THE META-SMITH CHART FOR CCITLS WITH

NEGATIVE CHARACTERISTIC RESISTANCES
As discussed in Chapter 2, some CCITLs can exhibit negative characteristic resistances (Re{Z0± } <
0); e.g., finite reciprocal lossless periodic TL structures operated in passbands. For this case, the
magnitude of the associated voltage reflection coefficient is always greater than or equal to unity
for passive load terminations as shown in Section 2.5 (see Appendix D). Due to this reason, some
modifications to the ZY Meta-Smith chart for CCITLs with NNCRs developed in Section 4.2 are
necessary. The modified chart is called the ZY Meta-Smith chart for CCITLs with NCRs, where the
range of plotting is always outside or on the unit circle of the voltage reflection coefficient  plane
for passive load terminations [2].
As pointed out in Chapter 2, the argument φ of Z0− for CCITLs with NCRs must lie in the
following ranges:
−180◦ < φ < −90◦ ∪ 90◦ < φ ≤ 180◦ . (4.23)
 4.3. THE META-SMITH CHART FOR CCITLS WITH NCRS 49

Figure 4.5: Plot of the ZY Meta-Smith chart in the  plane for φ = 30◦ .

The ZY Meta-Smith chart for CCITLs with NCRs can be constructed by using the same concept as
in constructing the ZY Meta-Smith chart for CCITLs with NNCRs.The ZY Meta-Smith chart for
CCITLs with NCRs is also a superposition of the impedance (Z) and admittance (Y ) Meta-Smith
charts as discussed below.

4.3.1 THE Z AND Y META-SMITH CHARTS FOR CCITLS WITH NCRS

Due to the fact that the formulas of the voltage reflection coefficient at the load  of CCITLs with
NNCRs and NCRs are identical (see (4.1)), all circle equations discussed in Section 4.2 (see (4.12),
(4.13), (4.15), and (4.16)) remain the same for CCITLs with NCRs. However, these circle contours
for CCITLs with NCRs are only considered outside or on the unit circle with the valid ranges of the
argument φ defined in (4.23), for passive load terminations [2]. In the next subsection, numerical
results of the ZY Meta-Smith chart for CCITLs with NCRs are illustrated.

4.3.2 NUMERICAL RESULTS

From the previous subsection, it is found that the ZY Meta-Smith chart for CCITLs with NCRs
also depends on the argument φ. To illustrate this dependence, consider the case when φ = 150◦ ,
which is in the ranges defined in (4.23). Figure 4.6 illustrates the plot of resistance circles in the 
plane. Each circle corresponds to different values of normalized load resistances rL . The boundary
circle, separating the regions of passive (rL ≥ 0) and active (rL < 0) loads, occurs when the rL equals
50 4. META-SMITH CHARTS FOR CCITLS AND BCITLS
to zero, which is the unit circle. Note that the region of active loads is within the unit circle. As rL
increases, the radius of the resistance circle becomes larger until it reaches a “turning point,” where
from this point on, the radius of the circle decreases, as shown in Figure 4.6. From (4.12), this point
corresponds to
rL = − cos φ, (4.24)
which makes the denominators in all associated terms of (4.12) equal to zero.

Figure 4.6: Plot of resistance circles in the  plane for φ = 150◦ .

Define θ as the angle between the horizontal line and the line OP , drawn from the origin O
in the  plane to the touching point P of all resistance circles, in the counterclockwise direction as
shown in Figure 4.6. The relationship between the argument φ and the angle θ are given by

θ = 2φ, (4.25)

which θ is equal to 300◦ for φ = 150◦ . Note that the relationship between φ and θ in (4.25) is
identical to (4.17) for the Meta-Smith chart for CCITLs with NNCRs.
Figure 4.7 illustrates the plot of reactance circles in the  plane for φ = 150◦ . From this
figure, the horizontal line and the line OP , drawn from the origin O to the intersecting point P of
all reactance circles, forms the angle θ (in counterclockwise direction) as well. Note that the point
P in Figures 4.4 and 4.7 is the same point. In addition, it is observed that the reactance circle xL =
0 for the Z Meta-Smith chart for CCITLs with NCRs is a part of circle. Furthermore, all reactance
 4.3. THE META-SMITH CHART FOR CCITLS WITH NCRS 51
circles outside the circle xL = 0 correspond to xL < 0, and those inside the circle xL = 0 correspond
to xL > 0 for 90◦ < φ ≤ 180◦ , and vice versa for −180◦ < φ < −90◦ .

Figure 4.7: Plot of reactance circles in the  plane for φ = 150◦ .

The Y Meta-Smith chart for CCITLs with NCRs consists of conductance circles and sus-
ceptance circles as shown in Figure 4.8 for φ = 150◦ . It can be observed that these circles intersect
(or touch) at the same point on the chart, which is the point (–1, 0) in the  plane. It is interesting to
point out that this point remains the same when the argument φ changes. For conductance circles,
the boundary circle separating the regions of passive and active loads occurs when gL = 0, which is
the unit circle. As gL increases to the turning point which corresponds to
gL = − cos φ, (4.26)
the radius of the conductance circle becomes larger, while after that, the radius becomes smaller.
Unlike the reactance circles in Figure 4.7, the susceptance circles inside and outside the circle
bL = 0 corresponds to bL < 0 and bL > 0 for 90◦ < φ ≤ 180◦ respectively, and vice versa for
−180◦ < φ < −90◦ .
Finally, the ZY Meta-Smith chart for CCITLs with NCRs can be obtained by superimposing
these four families of circles in Figures 4.6 to 4.8 in the  plane, as shown in Figure 4.9 for φ = 150◦ .
From the figure, it is observed that the circle xL = 0 and the circle bL = 0 are the same. This can be
verified by comparing (4.13) and (4.16) where xL = 0 and bL = 0, respectively. It is interesting to
note that the circle xL = 0 joins the intersecting (or touching) point of the Z Meta-Smith chart for
CCITLs with NCRs (the point P in Figures 4.6 and 4.7) and the intersecting (or touching) point of
the Y one (the point (–1, 0)), as shown in Figure 4.9. Furthermore, the former and latter intersecting
52 4. META-SMITH CHARTS FOR CCITLS AND BCITLS

Figure 4.8: Plot of the Y Meta-Smith chart for CCITLs with NCRs in the  plane for φ = 150◦ .

(or touching) points are the OC and SC points, respectively. In addition, it should be pointed out
that the resistance and reactance circles intersect at the right angle due to the conformal mapping
property [12]. Similarly, the conductance and susceptance circles intersect at the right angle as well.
To illustrate one more example of the ZY Meta-Smith chart for CCITLs with NCRs, Figure 4.10
shows the ZY Meta-Smith chart in the  plane for φ = −150◦ . Note that the circles rL = 1 and xL
= 0 are reversed to the opposite side of the horizontal axis in the  plane compared to the previous
case (see Figure 4.9). For other values of φ defined in (4.23), all observations above for φ = ±150◦
can be applied as well.
It should be pointed out that the procedures of using the ZY Meta-Smith chart for CCITLs
with NCRs in determining the input impedance of terminated CCITLs are similar to those for
reciprocal lossless uniform transmission lines using the Smith chart, and they will be discussed in
detail in Chapter 5. However, the operating region of the ZY Meta-Smith chart for CCITLs with
NCRs is outside or on the unit circle in the  plane for passive load terminations, which is different
from that of the Smith chart.
For more convenience in usage, the ZY Meta-Smith chart for CCITLs with NCRs can be
modified appropriately, called a modified ZY Meta-Smith chart for CCITLs with NCRs (originally
called a modified extended ZY T-chart [13], [14]), by mapping the region outside the unit circle
of the former to the inside one of the latter using an appropriate bilinear transformation [12]. The
modified chart is plotted in the plane of the inverse of the voltage reflection coefficient instead of the
 4.4. THE META-SMITH CHART FOR BCITLS 53

Figure 4.9: Plot of the ZY Meta-Smith chart for CCITLs with NCRs in the  plane for φ = 150◦ .

 plane. It is found that the modified chart also depends on the argument φ defined in (4.23) in a
complicated fashion.

4.4 THE META-SMITH CHART FOR BCITLS

In this section, the Meta-Smith charts for CCITLs with NNCRs and NCRs are generalized to treat
BCITL problems, called the Meta-Smith chart for BCITLs [3], [4]. Due to the fact that Z0± of BCITLs
are generally not complex conjugate of one another, associated equations for the Meta-Smith chart
for BCITLs will be found to be more complicated compared to those for the Meta-Smith charts for
CCITLs with NNCRs and NCRs. As pointed out in Chapter 3, BCITLs can exhibit both NNCRs
and NCRs, which will be treated simultaneously.
The Meta-Smith chart for BCITLs can be constructed by using the same concept as in
constructing the Meta-Smith charts for CCITLs developed in Sections 4.2 and 4.3. Using (4.1), it
can be shown that the magnitude of the voltage reflection coefficient at the load || associated with
terminated BCITLs may be greater than unity for some passive load impedances ZL . This implies
that associated circles are not only plotted inside the unit circle but as well as outside of the unit
circle in the complex  plane. As pointed out in Section 1.3.1, reciprocal lossy uniform TLs also
exhibit this phenomenon; i.e., their magnitude of the voltage reflection coefficient at the load could
be greater than unity for some passive load terminations. As pointed out in Chapter 3 , a reciprocal
54 4. META-SMITH CHARTS FOR CCITLS AND BCITLS

Figure 4.10: Plot of the ZY Meta-Smith chart for CCITLs with NCRs in the  plane for φ = −150◦ .

lossy uniform TL is one type of BCITLs. Thus, it is not surprised to see the same phenomenon
discussed above for terminated BCITLs.
Note that the Meta-Smith chart for BCITLs is a superposition of the Z Meta-Smith chart and
the Y Meta-Smith chart. Next, the key steps to obtain associated circle equations of the Meta-Smith
chart for BCITLs are discussed below.

4.4.1 THE Z META-SMITH CHART FOR BCITLS

For BCITLs, the normalized impedance z of any impedance Z is defined as in the case of CCITLs
(see (4.4) and (4.5)); i.e.,
Z
z=  , (4.27)
Z0+ Z0−

where Z0± are defined in the rectangular form as in (3.3) and (3.4). Unlike CCITLs, the term

Z0+ Z0− in (4.27) is complex in general. Using (4.41) for the voltage reflection coefficient at the load
for BCITLs and (4.27), the normalized load impedance zL can be expressed compactly in terms of
the normalized characteristic impedances z0± and  as

1+
zL = , (4.28)
z0−− z0+ 
 4.4. THE META-SMITH CHART FOR BCITLS 55

where z0+ and z0− are defined as

Z+ P + jQ
z0+ ≡  0 = , (4.29)
+ −
Z0 Z0 Z0+ Z0−
Z− U + jV
z0− ≡  0 = . (4.30)
Z0+ Z0− Z0+ Z0−

For convenience in manipulation, let  = r + j i and zL = rL + jx L . Substituting  and zL

into (4.28) and considering only the real and imaginary parts of (4.28), with some straightforward
manipulations, the following resistance and reactance circle equations are obtained:
2 2 2 2
A B A B
r + + i + = + − C, (4.31)
2 2 2 2
2 2 2 2
D E D E
r + + i + = + − F, (4.32)
2 2 2 2
respectively. Note that A, B, and C in (4.31) are dependent on the normalized load resistance rL and
the normalized characteristic impedance parameters, p, q, u, and v, which are defined as follows:
[2rL (up + vq) + (u − p)]
A=−   , (4.33)
rL p 2 + q 2 + p
[2rL (uq − vp) − (v + q)]
B=   , (4.34)
rL p2 + q 2 + p
 
rL u2 + v 2 − u
C=   . (4.35)
rL p 2 + q 2 + p
Note that p and q are the real and imaginary parts of z0+ (see (4.29)) respectively, and u and v are
the real and imaginary parts of z0− (see (4.30)), respectively. In (4.32), D, E, and F are dependent
on the normalized load reactance xL and the parameters p, q, u, and v, which are defined as
[2xL (up + vq) + (q − v)]
D=−   , (4.36)
xL p 2 + q 2 − q
[2xL (uq − vp) − (p + u)]
E=   , (4.37)
xL p 2 + q 2 − q
 
xL u2 + v 2 + v
F =   . (4.38)
xL p 2 + q 2 − q
It can be shown that (4.31) and (4.32) can be reduced to the equations of the resistance and reactance
circles of the Z Meta-Smith charts for CCITLs (see (4.12) and (4.13)), respectively, when p = u
and q = –v, (i.e., Z0+ = (Z0− )∗ ). Thus, the Z Meta-Smith charts for CCITLs are special cases of the
Z Meta-Smith chart for BCITLs when Z0+ = (Z0− )∗ indeed.
56 4. META-SMITH CHARTS FOR CCITLS AND BCITLS
4.4.2 THE Y META-SMITH CHART FOR BCITLS
To construct the Y Meta-Smith chart for BCITLs, the normalized load admittance yL must be
considered instead, where yL is the reciprocal of zL . Let yL = gL + jbL be the normalized load
admittance as defined in Section 4.2.2. Following the same procedures as in constructing the Z Meta-
Smith chart for BCITLs, the equations for the conductance and susceptance circles are obtained,
respectively, as follows:
2 2 2 2
G H G H
r + + i + = + − I, (4.39)
2 2 2 2
2 2 2 2
J K J K
r + + i + = + − L. (4.40)
2 2 2 2

In (4.39) and (4.40), the parameters G, H , I , J , K, and L are defined below:

2gL + p − u
G= , (4.41)
gL + p
− (v + q)
H = , (4.42)
gL + p
gL − u
I= , (4.43)
gL + p
2bL + q − v
J = , (4.44)
bL + q
p+u
K= , (4.45)
bL + q
bL − v
L= . (4.46)
bL + q

It can be shown that (4.39) and (4.40) can be reduced to the conductance and susceptance circles of
the Y Meta-Smith charts for CCITLs (see (4.15) and (4.16)), respectively, when Z0+ = (Z0− )∗ as
well. Thus, the Y Meta-Smith charts for CCITLs are special cases of the Y Meta-Smith chart for
BCITLs when Z0+ = (Z0− )∗ indeed.
After deriving these four circle equations (see (4.31), (4.32), (4.39), and (4.40)), the Meta-
Smith chart for BCITLs with NNCRs and NCRs can be constructed readily. In the next subsection,
numerical results of the Meta-Smith chart for BCITLs will be illustrated.

4.4.3 NUMERICAL RESULTS

In this subsection, two examples of the Meta-Smith chart for BCITLs with NNCRs and NCRs are
provided. First, consider a BCITL with NNCRs possessing Z0+ = 149.3 – j 7.9  and Z0− = 146.9
– j 39.6 , where Z0+ and Z0− are different with (Re{Z0± } ≥ 0). This BCITL can be implemented
using a finite lossy periodic TL structure discussed in detail later in Section 5.6. Substituting the
 4.4. THE META-SMITH CHART FOR BCITLS 57

given Z0± into (4.31) and (4.32), the resistance and reactance circles can be plotted as shown in
Figures 4.11 and 4.12, respectively. In addition, Figure 4.13 is the Z Meta-Smith chart for BCITLs
with NNCRs, which is the superposition of Figures 4.11 and 4.12. For resistance circles, the circles
with rL > 0 are inside the circle rL = 0 while those with rL < 0 are outside the circle rL = 0. For
reactance circles, the contours above xL = 0 correspond to xL > 0 and those below xL = 0 correspond
to xL < 0. Note that the resistance and reactance circles intersect at the right angle due to the
conformal mapping property.

Figure 4.11: Plot of resistance circles for Z0+ = 149.3 − j 7.9  and Z0− = 146.9 − j 39.6 .

Using (4.39) and (4.40), the Y Meta-Smith chart for BCITLs with NNCRs, which is the
superposition of conductance and susceptance circles, can be constructed as illustrated in Figure 4.14.
For conductance circles, the circles with gL > 0 are inside the circle gL = 0 while those with gL < 0
are outside the circle gL = 0. For susceptance circles, the contours above bL = 0 correspond to bL <
0, and those below bL = 0 correspond to bL > 0. Note that the conductance and susceptance circles
intersect at the right angle as well.
Figure 4.15 shows the ZY Meta-Smith chart for BCITLs with NNCRs for Z0+ = 149.3 –
j 7.9  and Z0− = 146.9 – j 39.6  by superimposing the Z Meta-Smith chart in Figure 4.13 and
the Y Meta-Smith chart in Figure 4.14 together. From Figure 4.15, it is observed that the circle xL
= 0 and the circle bL = 0 are the same. This can be verified by comparing (4.32) and (4.40) when
xL = 0 and bL = 0, respectively. In addition, the circle xL = 0 also joins the OC point and the SC
point as shown in Figure 4.15.
58 4. META-SMITH CHARTS FOR CCITLS AND BCITLS

Figure 4.12: Plot of reactance circles for Z0+ = 149.3 − j 7.9  and Z0− = 146.9 − j 39.6 .

Next, consider a BCITL with NCRs possessing Z0+ = –35.825 – j 12.480  and Z0− = –24.785
– j 18.261 , where Z0+ and Z0− are different with Re{Z0± } < 0. This BCITL can be implemented
using a finite lossy periodic TL structure [15]. Substituting the given Z0± into (4.31) and (4.32), the
resistance and reactance circles can be plotted together in the  plane resulting in the Z Meta-Smith
chart for BCITLs with NCRs as shown in Figure 4.16. For resistance circles, the circles with rL <
0 are inside the circle rL = 0 while those with rL > 0 are outside the circle rL = 0. Note that this
characteristic of resistance circles is opposite to that of the Z Meta-Smith chart for BCITLs with
NNCRs as shown earlier in Figure 4.13. For reactance circles, the contours below xL = 0 correspond
to xL > 0 and those above xL = 0 correspond to xL < 0. It is observed that this characteristic of
reactance circles is opposite to that of the Z Meta-Smith chart for BCITLs with NNCRs as shown
in Figure 4.13. Note that the resistance and reactance circles in Figure 4.16 intersect at the right
angle.
Using (4.39) and (4.40), the Y Meta-Smith chart for BCITLs with NCRs, which is the
superposition of conductance and susceptance circles, can be constructed as shown in Figure 4.17.
For conductance circles, the circles with gL < 0 are inside the circle gL = 0 while those with gL > 0
are outside the circle gL = 0. For susceptance circles, the contours below bL = 0 correspond to bL <
0, and those above bL = 0 correspond to bL > 0. Note that these characteristics of conductance and
susceptance circles are opposite to those of the Y Meta-Smith chart for BCITLs with NNCRs as
shown earlier in Figure 4.14. In addition, it is observed that the conductance and susceptance circles
in Figure 4.17 intersect at the right angle.
 4.5. CONCLUSIONS 59

Figure 4.13: The Z Meta-Smith chart for BCITLs with NNCRs for Z0+ = 149.3 − j 7.9  and Z0− =
146.9 − j 39.6 .

Figure 4.18 shows the ZY Meta-Smith chart for BCITLs with NCRs for Z0+ = – 35.825 –
j 12.480  and Z0− = – 24.785 – j 18.261  by superimposing the Z Meta-Smith chart in Figure 4.16
and the Y Meta-Smith chart in Figure 4.17 together. From Figure 4.18, it is observed that the circle
xL = 0 and the circle bL = 0 are identical. In addition, the circle xL = 0 also joins the OC point
and the SC point as shown in Figure 4.18.

4.5 CONCLUSIONS
In this chapter, the Meta-Smith charts for CCITLs with NNCRs and NCRs are developed, in-
cluding the Meta-Smith chart for BCITLs. For the Meta-Smith chart for CCITLs with NNCRs,
the region of interest is inside or on the unit circle in the  plane for passive load terminations. For
the case of CCITLs with NCRs, the plotting range of the Meta-Smith chart is always outside or on
the unit circle in the  plane for passive loads. It is found that the Meta-Smith charts for CCITLs
with NNCRs and NCRs depend on the argument φ of Z0± in a complicated fashion. In addition,
the Smith chart is a special case for the Meta-Smith chart for CCITLs with NNCRs when φ = 0◦ .
Examples are also provided for these Meta-Smith charts. For the Meta-Smith chart for BCITLs, it
is found that the magnitude of the voltage reflection coefficient at the load || may be greater than
unity for some passive load impedances ZL . Thus, associated circles of the Meta-Smith chart for
BCITLs are not only plotted inside the unit circle but as well as outside of the unit circle in the 
plane. Two examples of the Meta-Smith chart for BCITLs with NNCRs and NCRs are provided.
60 4. META-SMITH CHARTS FOR CCITLS AND BCITLS

Figure 4.14: The Y Meta-Smith chart for BCITLs with NNCRs for Z0+ = 149.3 − j 7.9  and Z0− =
146.9 − j 39.6 .

These Meta-Smith charts are very useful in the analysis and design of CCITLs and BCITLs since
they can provide more physical insight of visualizing associated TL phenomenon and solving related
problems effectively. In the next chapter, several applications of the Meta-Smith charts for CCITLs
and BCITLs will be illustrated in detail.

BIBLIOGRAPHY
[1] D. Torrungrueng and C. Thimaporn, “A generalized ZY Smith chart for solvingnonreciprocal
uniform transmission line problems,” Microwave and Optical Technology Letters, vol. 40, no. 1,
pp. 57–61, Jan. 2004. DOI: 10.1002/mop.11284 41, 42, 43, 44

[2] D. Torrungrueng, P.Y. Chou, and M. Krairiksh, “An extended ZY T-chart for conju-
gately characteristic-impedance transmission lines with active characteristic impedances,”
Microwave and Optical Technology Letters, vol. 49, no. 8, pp. 1961–1964, Aug. 2007.
DOI: 10.1002/mop.22626 41, 48, 49

[3] D. Torrungrueng, P.Y. Chou, and M. Krairiksh, “A graphical tool for analysis and designof bi-
characteristic-impedance transmission lines,” Microwave and Optical Technology Letters, vol. 49,
no. 10, pp. 2368–2372, Oct. 2007. DOI: 10.1002/mop.22801 53
 BIBLIOGRAPHY 61

Figure 4.15: The ZY Meta-Smith chart for BCITLs with NNCRs for Z0+ = 149.3 − j 7.9  and
Z0− = 146.9 − j 39.6 .

[4] D. Torrungrueng, P.Y. Chou, andM. Krairiksh, “Erratum: A graphical tool for analysis and
design of bi-characteristic-impedance transmission lines,” Microwave and Optical Technology
Letters, vol. 51, no. 4, pp. 1154, Apr. 2009. DOI: 10.1002/mop.24260 53

[5] D. Torrungrueng and C. Thimaporn, “Applications of the ZY T-chart for nonreciprocal stub
tuners,” Microwave and Optical Technology Letters, vol. 45, no. 3, pp. 259–262, May 2005.
DOI: 10.1002/mop.20789

[6] D. Torrungrueng and C. Thimaporn, “Application of the T-chart for solving exponentially
tapered lossless nonuniform transmission line problems,” Microwave and Optical Technology
Letters, vol. 45, no. 5, pp. 402–406, Jun. 2005. DOI: 10.1002/mop.20836

[7] D. Torrungrueng,C. Thimaporn, and N. Chamnandechakun, “An application of the T-Chart

for solving problems associated with terminated finite lossless periodic structures,” Microwave
and Optical Technology Letters, vol. 47, no. 6, pp. 594–597, Dec. 2005. DOI: 10.1002/mop.21239
41

[8] S. Lamultree and D. Torrungrueng, “On the characteristics of conjugately characteristic-

impedance transmission lines with active characteristic impedance,” in 2006 Asia-Pacific Mi-
crowave Conference Proceedings, Dec. 2006, vol. 1, pp. 225–228.
DOI: 10.1109/APMC.2006.4429411 41
62 4. META-SMITH CHARTS FOR CCITLS AND BCITLS

Figure 4.16: The Z Meta-Smith chart for BCITLs with NCRs for Z0+ = −35.825 − j 12.480  and
Z0− = −24.785 − j 18.261 .

[9] K. Vudhivorn, D. Torrungrueng and C. Thimaporn, “Non-uniqueness of T-charts forsolving

CCITL problems with passive characteristic impedances,” Progress inElectromagnetics Research,
vol. 7, pp. 193–205, 2009.
DOI: 10.2528/PIERM09052803 42, 46, 47, 48
[10] P. H. Smith, Electronic Applications of the Smith Chart. Georgia: Noble Publishing, 2000. 42,
45, 46
[11] D. M. Pozar, Microwave Engineering, 3rd ed. New Jersey: Wiley, 2005. 46
[12] E. Kreyszig, Advanced Engineering Mathematics, 8th ed. New Jersey: Wiley, 1998. 52
[13] K. Vudhivorn and D. Torrungrueng, “A modified extended ZY T-chart for
conjugatelycharacteristic-impedance transmission lines with active characteristic
impedances,”Microwave and Optical Technology Letters, vol. 51, no. 3, pp. 621–625, Mar. 2009.
DOI: 10.1002/mop.24124 52
[14] K. Vudhivorn, D. Torrungrueng and T. Angkaew, “Applications of the Modified Extended
ZY T-Chart for Problems Associated with CCITLs with Active CharacteristicImpedances,”
in Proceedings of the 2009 ECTI International Conference, May 2009, vol. 2, pp. 822–825.
DOI: 10.1109/ECTICON.2009.5137172 52
 4.6. PROBLEMS 63

Figure 4.17: The Y Meta-Smith chart for BCITLs with NCRs for Z0+ = −35.825 − j 12.480  and
Z0− = −24.785 − j 18.261 .

[15] S. Lamultree, D.Torrungrueng and C. Phongcharoenpanich, “Measurement of BCITLparam-

eters,” in Proceedings of the 2008 ECTI International Conference, May 2008, vol. 1, pp. 257–260.
DOI: 10.1109/ECTICON.2008.4600421 58

4.6 PROBLEMS
4.1. Using (4.9), show that the equations for resistance and reactance circles for CCITLs are
given as in (4.12) and (4.13), respectively.

4.2. Show that the equations for conductance and susceptance circles for CCITLs are given as
in (4.15) and (4.16), respectively.

4.3. Rigorously show that the relationship between the argument φ and the angle θ (see Fig-
ure 4.2 as an example) for CCITLs with NNCRs is given as in (4.17).

4.4. Plot the ZY Meta-Smith chart for CCITLs with NNCRs in the  plane for φ = 0◦ , ±30◦
and ±60◦ (see Figure 4.5 as an example).

4.5. Using Z̃0 defined in (4.18), show that the equations for resistance, reactance, conductance
and susceptance circles for CCITLs with NNCRs are given as in (4.19) to (4.22), respec-
tively.
64 4. META-SMITH CHARTS FOR CCITLS AND BCITLS

Figure 4.18: The ZY Meta-Smith chart for BCITLs with NCRs for Z0+ = −35.825 − j 12.480  and
Z0− = −24.785 − j 18.261 .

4.6. Using (4.19) to (4.22), plot the ZY Meta-Smith chart for CCITLs with NNCRs in the 
plane for φ = 0◦ , ±30◦ and ±60◦ . Compare these results with those of Problem 4.4.

4.7. Rigorously show that the relationship between the argument φ and the angle θ (see Fig-
ure 4.6 as an example) for CCITLs with NCRs is given as in (4.25).

4.8. Plot the ZY Meta-Smith chart for CCITLs with NCRs in the  plane for φ = 120◦ ,
±135◦ and ±150◦ (see Figure 4.9 as an example).

4.9. Using (4.28), show that the equations for resistance and reactance circles for BCITLs are
given as in (4.31) and (4.32), respectively.

4.10. Show that (4.31) and (4.32) can be reduced to the equations of the resistance and reactance
circles of the
∗ Z Meta-Smith charts for CCITLs (see (4.12) and (4.13)) respectively, when
Z0+ = Z0− .

4.11. Show that the equations for conductance and susceptance circles for BCITLs are given as
in (4.39) and (4.40), respectively.

4.12. Verify the plot of the ZY Meta-Smith chart for BCITLs with NNCRs in Figure 4.15 for
Z0+ = 149.3 − j 7.9  and Z0− = 146.9 − j 39.6 .
 4.6. PROBLEMS 65
4.13. Verify the plot of the ZY Meta-Smith chart for BCITLs with NCRs in Figure 4.18 for
Z0+ = −35.825 − j 12.480  and Z0− = −24.785 − j 18.261 .
 67

CHAPTER 5

Applications of Meta-Smith
Charts
5.1 INTRODUCTION
In Chapter 4, the Meta-Smith charts for CCITLs and BCITLs with NNCRs and NCRs are
developed. These charts usually provide more physical insight of visualizing phenomenon associated
CCITLs and BCITLs and solving related problems effectively as can be seen in several examples,
involving impedance matching and impedance calculation associated with CCITLs and BCITLs,
discussed in this chapter. It will be shown later in this chapter that graphical solutions based on
computerized Meta-Smith charts are fast, intuitive and accurate, compared to analytical solutions.
This chapter is organized as follows. Nonreciprocal stub tuners are designed using the Meta-
Smith chart for CCITLs with NNCRs for both single-stub and double-stub matching networks
in Section 5.2. Sections 5.3 and 5.4 illustrate the usage of the Meta-Smith chart for CCITLs with
NNCRs in solving problems associated with exponentially tapered lossless nonuniform transmission
lines and terminated finite reciprocal lossless periodic TL structures with NNCRs, respectively. A
CCITL with NCRs is analyzed using the Meta-Smith chart for CCITLs with NCRs in Section 5.5.
Section 5.6 presents the usage of the Meta-Smith chart for BCITLs with NNCRs in solving a
problem associated with reciprocal lossy BCITLs with NNCRs. Finally, conclusions are provided
in Section 5.7.

5.2 NONRECIPROCAL STUB TUNERS

Impedance matching is a part of design process for a microwave system to achieve the maximum
power delivered to the load. Stub-tuning is one of conventional matching techniques, which is
usually implemented using single-stub and double-stub tuners. A stub is an open- or short-circuited
TL connected either in parallel or in series with the transmission feed line at a certain distance from
the load [1]. In this section, the Meta-Smith chart for CCITLs with NNCRs is applied for solving
problems associated with nonreciprocal stub tuners composed of nonreciprocal lossless uniform
transmission lines, where these TLs are CCITLs with NNCRs. Both nonreciprocal single-stub and
double-stub matching networks are studied graphically using the Meta-Smith chart for CCITLs
with NNCRs. The stubs can be connected in series or in parallel with the main nonreciprocal lossless
feed line. It will show that basic procedures of using the Meta-Smith chart in solving problems of
nonreciprocal stub tuners are similar to those of applying the Smith chart for reciprocal ones discussed
68 5. APPLICATIONS OF META-SMITH CHARTS
in [1]. It should be pointed out that nonreciprocal stub tuners can be designed using analytical and
graphical solutions as well [2]– [3]. Although analytical solutions are more accurate and useful for
computer analysis and design, they are complicated [2]. On the other hand, graphical solutions based
on the Meta-Smith chart are found to be fast, intuitive, and sufficiently accurate in practice [3], which
will be illustrated subsequently.
In this section, it is assumed for simplicity that all TLs of interest are nonreciprocal and lossless,
possessing the same characteristic impedances Z0± = Z0± e∓j φ with NNCRs (Re{Z0± } ≥ 0), and
the same propagation constants β ± . First, graphical solutions based on the Meta-Smith chart for
CCITLs with NNCRs for nonreciprocal single-stub series tuners are discussed in Section 5.2.1.
Next, the Meta-Smith chart for CCITLs with NNCRs is applied for nonreciprocal double-stub
shunt tuners as illustrated in Section 5.2.2. It should be pointed out that the Meta-Smith chart can
be applied for nonreciprocal stub tuners when each nonreciprocal TL of interest is different as well.

5.2.1 NONRECIPROCAL SINGLE-STUB SERIES TUNERS

In the single-stub series tuning, the design parameters are the distance d from the load to the stub
and the length  of the OC or SC stub as shown in Figure 5.1. The single-stub series tuning circuit
can be designed using the Z Meta-Smith chart for CCITLs with NNCRs. In order to match a load

Feed Line d
Z 0+ , Z 0–
Zd Z 0+ , Z 0–
– – ZL
β +, β Z st β +, β

Z 0+ , Z 0– l
–
β +, β

Open or
Shorted stub

Figure 5.1: A nonreciprocal single-stub series tuner.

with the input nonreciprocal feed line, it is required to make the voltage reflection coefficient looking
into the matching network equal to zero. Using (2.7) for the desired voltage reflection coefficient,
the above requirement can be achieved by choosing the distance d and the length  such that
Z0+ = Zd + Zst , (5.1)
where Zd is the input impedance looking into the nonreciprocal TL terminated in a passive load
ZL , and Zst is the input impedance looking into the nonreciprocal stub as shown in Figure 5.1.
 5.2. NONRECIPROCAL STUB TUNERS 69
Using (2.7) and (2.8), it can be shown that Zst is purely imaginary due to OC or SC termination of
the stub. The normalized version of (5.1) can be written as

z0+ = zd + zst , (5.2)

where z0+ , zd , and zst are the normalized impedances (see (4.4) and (4.5)) of Z0+ , Zd , and Zst ,
respectively. The Meta-Smith chart solutions can be illustrated best by an example as shown below.
Consider an example of a single-stub series tuning network implemented using an OC stub
when ZL = 50 , |Z0+ | = 50  and φ = 45◦ . Let us define β̃ be the arithmetic mean of β ± ; i.e.,

1 +  2π
β̃ ≡ β + β− ≡ , (5.3)
2 λ̃
where λ̃ is the effective wavelength of waves propagating along NRLUTLs. From the given data,
the normalized load impedance zL is equal to 1. For this case, the Z Meta-Smith chart for CCITLs
with NNCRs is employed with φ = 45◦ as shown in Figure 5.2.

Figure 5.2: The Z Meta-Smith chart solutions for a single-stub series tuner with φ = 45◦ .

The first step is to draw a circle, centered at the point  = 0, passing through the point zL on
the Z Meta-Smith chart for CCITLs with NNCRs. It should be pointed out that the point  = 0
corresponds to the point z0+ , which can be seen from (4.6) and (4.7). Note that the circle intersects
the resistance circle, passing through the point  = 0, in the  plane at two intersection points
denoted as zd,1 and zd,2 as shown in Figure 5.2. Let us consider the first solution corresponding
70 5. APPLICATIONS OF META-SMITH CHARTS
to zd,1 first, where zd,1 can be read from the computerized Z Meta-Smith chart as zd,1 = 0.708 –
j 0.063. From Figure 5.2, the shortest distance d1 from the load to the nonreciprocal stub can be
determined as d1 = [0.5 − (0.063 − 0.034)] λ̃ = 0.471 λ̃. In order to determine the required stub
length , one equates the imaginary part of (5.2); i.e.,

xst,1 = I m{z0+ } − xd,1 , (5.4)

where xst,1 and xd,1 are the normalized reactances of zst,1 and zd,1 respectively, and zst,1 is the
normalized impedance zst of the first solution. Note that the symbol Im{} in (5.4) denotes the
imaginary part of its argument. From (5.4) with zd,1 = 0.708 – j 0.063, the first solution requires a
nonreciprocal stub with the normalized reactance of xst,1 = – 0.644. The length 1 of an OC stub
that provides xst,1 can be found using the computerized Z Meta-Smith chart by starting at the OC
point, and moving along the outer edge of the Z Meta-Smith chart, corresponding to the resistance
circle of rL = 0 (see (4.12)), toward the generator (in the clockwise direction) to the point xst,1 as
shown in Figure 5.2. This provides a stub length of 1 = 0.264 λ̃. Similarly, the distance d2 from the
load and the required OC stub length 2 of the second solution are found to be: d2 = 0.153 λ̃ and
2 = 0.423 λ̃. It is interesting to point out that these results are sufficiently accurate compared with
analytical solutions [2]. This completes the nonreciprocal single-stub series tuner design. Note that
the nonreciprocal single-stub shunt tuner can be employed as well by connecting the nonreciprocal
stub in parallel instead. Its procedures of using the Meta-Smith chart are similar to those of the
series one. In Section 5.2.2, the design of nonreciprocal double-stub shunt tuners is discussed in
detail.

5.2.2 NONRECIPROCAL DOUBLE-STUB SHUNT TUNERS

In nonreciprocal double-stub shunt tuning, the lengths of the two nonreciprocal OC or SC stubs
are required, where the distance d between them is fixed as shown in Figure 5.3. Note that the load
admittance YL is located at the position of the first nonreciprocal stub. In Figure 5.3, B1 and B2
are the input susceptances of the first and second nonreciprocal stubs, respectively. The Meta-Smith
chart solutions are illustrated by the following example.
Consider a nonreciprocal double-stub shunt tuning network implemented using SC nonrecip-
rocal stubs when YL = 10 + j 50 S, |Y0± |= 50 S, φ = 45◦ , and d = λ̃/8, where characteristic admittances
of all nonreciprocal transmission lines with NNCRs are defined as Y0± = 1/Z0± = |Y0± |e±j φ with
|Y0± | = 1/|Z0± |. From the given data, the normalized load admittance yL = YL /|Y0± | is equal to 0.2
+ j . As in the case of the nonreciprocal single-stub tuners, two common solutions are possible. After
the normalized load admittance yL is plotted on the Y Meta-Smith chart with φ = 45◦ as shown in
Figure 5.4, the conductance circle passing through the point  = 0 is rotated by moving every point
on this circle λ̃/8 toward the load (in the counterclockwise direction). Then, the normalized input
susceptance b1 of the first solution of the first nonreciprocal stub can be obtained by starting at the
point yL and moving along the conductance circle, passing through yL , to the intersection point y1
between the conductance circle and the rotated circle as shown in Figure 5.4. Using the comput-
 5.2. NONRECIPROCAL STUB TUNERS 71

d
– –
Y0+, Y0 jB2 Y0+, Y0 jB1 YL
β ,β
+ –
β ,β
+ –

0 –

0 –
Y

Y
0 ,+

0 ,+
–

–
β

β
Y

Y
,

,
+

+
Open l2 Open l1
β

β
or or
shorted shorted
stub stub

Figure 5.3: A nonreciprocal double-stub shunt tuner.

erized Y Meta-Smith chart, it is found that b1 = 0.908. Now one transforms y1 through the λ̃/8
section of nonreciprocal transmission line by rotating along a constant radius circle (passing through
y1 ) λ̃/8 toward the generator (in the clockwise direction). This brings the first solution y1 to the point
y2 = 0.706 – j 1.733. Then, the normalized input susceptance b2 of the second nonreciprocal stub
of the first solution is obtained by starting at the point y2 and moving along the conductance circle
(passing through the point  = 0) to the point y0+ as shown in Figure 5.4. It should be pointed out
that the point  = 0 corresponds to the point y0+ . Using the computerized Y Meta-Smith chart, it
is found that b2 = 2.441.
Once b1 and b2 are known, the lengths of the first and second SC nonreciprocal stubs for
the first solution are found from the computerized Y Meta-Smith chart to be 1 = 0.294 λ̃ and
2 = 0.438 λ̃, respectively. Similarly, the second solution can be determined as 1 = 0.116 λ̃ and
2 = 0.0615 λ̃. This completes the nonreciprocal double-stub shunt tuning design. Note that the
nonreciprocal double-stub series tuning circuit can be used as well by connecting the nonreciprocal
stubs in series instead.
For a double-stub tuner, there is a specific region, where the load cannot be matched, called a
forbidden region [1]. For convenience, only shunt stubs are considered. From the above discussion, it is
obvious that if the normalized load admittance yL were inside the shaded region, called the forbidden
region, as shown in Figure 5.4, no value of shunt double-stub susceptances could bring the point yL
to intersect the rotated conductance circle. It should be pointed out that the forbidden region can be
identified obviously on the Meta-Smith chart. This is an advantage of using the Meta-Smith chart
in this problem. From analytical solutions of the nonreciprocal double-stub shunt tuning, it can be
shown rigorously that the equation of the conductance circle governing the forbidden region can be
72 5. APPLICATIONS OF META-SMITH CHARTS

Figure 5.4: A Y Meta-Smith chart diagram for the operation of a nonreciprocal double-stub shunt tuner
with φ = 45◦ .

expressed compactly as
 2
2
1 sin2 β̃d
r + + i2 = . (5.5)
1 + sin2 β̃d 1 + sin2 β̃d

Comparing (5.5) to (4.15), it is found that the conductance circle equation in (5.5) corresponds to
the normalized conductance of g = gmax , where
cos φ
gmax = . (5.6)
sin2 β̃d
From (5.5), it is interesting to point out that the forbidden region for the nonreciprocal double-stub
shunt tuners depends only on the phase√β̃d. In this example, the normalized load conductance gL is
equal to 0.2, which is less than gmax = 2 computed via (5.6). Thus, the nonreciprocal double-stub
shunt tuners can be applied to match the given load to the nonreciprocal feed line as shown earlier.
In conclusion, nonreciprocal stub tuners can be designed using analytical and Meta-Smith
chart solutions. Although analytical solutions are more accurate and useful for computer analysis,
they are quite complicated and difficult to obtain for nonreciprocal stub tuners. In contrast, solutions
based on the computerized Meta-Smith chart are fast, intuitive and accurate. It is found that basic
procedures and concepts for designing nonreciprocal stub tuners are similar to those of reciprocal
ones. In the next section, the Meta-Smith chart for CCITLs with NNCRs is applied to solve
problems associated with ETLNUTLs.
 5.3. TERMINATED EXPONENTIALLY TAPERED LOSSLESS NONUNIFORM TLS 73

5.3 TERMINATED EXPONENTIALLY TAPERED LOSSLESS

NONUNIFORM TRANSMISSION LINES
Due to complicated equations associated with terminated ETLNUTLs as shown in Figure 1.4,
a graphical solution assisting in analysis and design is desirable. In the literature [4], a general-
ized Smith chart, which is one of the graphical solutions, was developed and employed to study
ETLNUTLs. However, this chart is not convenient in solving ETLNUTL problems due to its
complicated graphical representation.
Due to the fact that practical ETLNUTLs are reciprocal CCITLs with NNCRs when TL
parameters and frequency are chosen appropriately (see Section 2.4), the Meta-Smith chart for
CCITLs with NNCRs can be employed to solve ETLNUTL problems effectively. It should be
pointed out that this Meta-Smith chart is totally different from another graphical tool proposed
in [4], and it is more convenient and intuitive in solving these problems. Note that this Meta-Smith
chart depends on the argument φ of the effective characteristic impedances Z0± of ETLNUTLs
defined earlier in (1.65) and (1.66), where Z0± are dependent on parameters of ETLNUTLs and
frequency. It will be illustrated that the procedures of using this Meta-Smith chart are similar to those
of using the standard ZY Smith chart in solving problems associated with RLUTLs. As pointed out
earlier in Section 4.2, this Meta-Smith chart can be reduced to the standard ZY Smith chart when
φ = 0◦ ; i.e., ETLNUTLs become RLUTLs in this case.
To apply the Meta-Smith chart for CCITLs with NNCRs, the input impedance equation of
terminated ETLNUTLs given in (1.72) must be rearranged into the following convenient form [5]:
 
1 + e −j 2β
Z̃in ≡ Zin eq = Z0+ Z0− , (5.7)
Z0− − Z0+ e−j 2β
where Z̃in is the modified input impedance of the ETLNUTL terminated in a passive real load
impedance ZL . Notice that the exponential term e−q in (1.72) is moved to the left-hand side as
shown in (5.7). It is found that (5.7) is in the same form as the equation of the input impedance of
CCITLs terminated in a load impedance ZL (see (2.26)). Thus, the Meta-Smith chart for CCITLs
with NNCRs can be applied to solve ETLNUTL problems indeed by working with the modified
input impedance Z̃in rather than the input impedance Zin .

5.3.1 PROCEDURES OF USING THE META-SMITH CHART

To apply the Meta-Smith chart for CCITLs with NNCRs, the normalized impedance z of ETL-
NUTLs is employed, and it is defined as in (4.4) and (4.5). Thus, the normalized modified input
impedance z̃in of Z̃in defined in (5.7) can be expressed as
1 + e−j 2β
z̃in = , (5.8)
z0− − z0+ e−j 2β
where z0± are the normalized effective characteristic impedances defined in (4.7). Once z̃in is known,
the modified input impedance Z̃in can be obtained by using (4.4) and (4.5), and the input impedance
74 5. APPLICATIONS OF META-SMITH CHARTS
Zin can be readily determined from (5.7) as

Zin = Z̃in e−q . (5.9)

The procedures of determining z̃in by using the Meta-Smith chart for CCITLs with NNCRs are
described below.
Consider the normalized load impedance zL defined as in (4.8). One can see that (5.8) for
z̃in is in the same form as (4.8), except for the extra phase factor e−j 2βl multiplying  for both
numerator and denominator in (5.8). For convenience in discussion, define the voltage reflection
coefficient  at the load in the polar form as ||ej ψ , where || and ψ are the absolute value and
the argument of , respectively. The first step in applying the Meta-Smith chart for CCITLs with
NNCRs is to plot the reflection coefficient at the load ||ej ψ on the chart, which corresponds to the
normalized load impedance point zL . Then, the normalized modified input impedance z̃in can be
determined by simply rotating the zL point clockwise (toward the generator) an amount of 2β (i.e.,
subtracting 2β from ψ) around the origin of the Meta-Smith chart. One can see that the above
procedures of applying the Meta-Smith chart for CCITLs with NNCRs in solving ETLNUTL
problems are similar to those of using the standard Smith chart in solving problems associated with
RLUTLs. These procedures of using the Meta-Smith chart for CCITLs with NNCRs can be seen
more clearly by considering a numerical example given in the next subsection.

5.3.2 NUMERICAL EXAMPLE

In this subsection, an example will be provided to show that solutions obtained from the Meta-
Smith chart for CCITLs with NNCRs are valid by comparing with results obtained from associated
formulas. Consider the case when C0 = 30 pF/m, q = 2 Np/m, f = 1 GHz, L0 = 1.688 nH/m,
β = 45◦ and ZL = 7.5  [5]. Using the given data and (1.62), (1.65), and (1.66), β and Z0± can be
◦
computed as β = 1 rad/m and Z0± = 7.5 e∓j 45 . Note that |Z0± | and φ are equal to 7.5  and
45◦ in this case, respectively. Using (4.4) and (4.5), the normalized load impedance zL is determined
as zL = rL + jx L = 1. In this case, rL = 1 and xL = 0 correspond to a real load, which can be plotted
as the zL point on the Meta-Smith chart for CCITLs with NNCRs with φ = 45◦ as shown in
Figure 5.5. Next, a circle passing through the zL point is drawn on the Meta-Smith chart. Then,
the normalized modified input impedance z̃in can be found by rotating the zL point clockwise along
the circle by an amount of 2β = 90◦ . Note that the z̃in point can be read from the computerized
Meta-Smith chart for CCITLs with NNCRs as 0.98 – j 1.41. Once z̃in is known, the modified input
impedance Z̃in can be calculated as Z̃in = z̃in Z0± = 7.35 − j 10.57 , and the input impedance
Zin of the terminated ETLNUTL can be determined by using (5.9) as Zin = 1.527 – j 2.197 .
Alternatively, the input impedance Zin can be determined analytically by using (1.72) as Zin = 1.559
– j 2.204 . Comparing these results, it is found that both are in very good agreement. Thus, this
example illustrates that the computerized Meta-Smith chart for CCITLs with NNCRs provides
accurate results indeed, and it can be employed readily and intuitively in analysis and design of
ETLNUTLs.
 5.4. TERMINATED FINITE RECIPROCAL LOSSLESS PERIODIC TL STRUCTURES 75

Figure 5.5: Illustration of the procedures of applying the Meta-Smith chart to solve an ETLNUTL
problem with φ = 45◦ .

5.4 TERMINATED FINITE RECIPROCAL LOSSLESS

PERIODIC TRANSMISSION LINE STRUCTURES WITH
NONNEGATIVE CHARACTERISTIC RESISTANCES
As discussed earlier in Section 2.5, periodic TL structures have several useful applications in mi-
crowave technology. In addition, lossless periodic TL structures are probably the most useful example
of CCITLs in practice. Due to complicated equations involved with terminated finite reciprocal loss-
less periodic TL structures, a graphical solution assisting in analysis and design of these structures
is desirable. In this section, the Meta-Smith chart for CCITLs with NNCRs is applied to solve
problems associated with these structures operating in passbands readily and effectively, especially
for unsymmetrical unit cells associated with periodic TL structures. The procedures of applying the
Meta-Smith chart for these periodic TL structures are provided, including an example. Only the
periodic TL structures with NNCRs are considered in this section.
Consider the terminated finite reciprocal lossless periodic TL structure of M unit cells in
Figure 2.3. Using (2.28), (4.4), and (4.5), the normalized input impedance zin,M at the input terminal
of the terminated periodic TL structure can be expressed as

1 + e−j 2Mβd
zin,M = , (5.10)
z0− − z0+ e−j 2Mβd

where z0± are the normalized effective characteristic impedances of Z0± . The normalized load
impedance zL can be obtained from (5.10) by setting d = 0 resulting in (4.8). Note that (4.8) is
76 5. APPLICATIONS OF META-SMITH CHARTS

in the same form as (5.10), except for the extra phase factor e−j 2Mβd multiplying the  term. For
convenience in discussion, let us define the voltage reflection coefficient  at the load as ||ej ψ .
Thus, if one has plotted the reflection coefficient ||ej ψ at the load, corresponding to the normal-
ized load impedance point zL on the Meta-Smith chart, the normalized input impedance zin,M can
be determined by simply rotating the zL point clockwise an amount of 2Mβd (subtracting 2Mβd
from ψ) around the origin of the Meta-Smith chart for CCITLs with NNCRs. Now, it is ready to
illustrate the use of the Meta-Smith chart through a specific example.
Let us consider the unit cell as illustrated in Figure 2.2 with the following parameters: d =
2 cm, k = 120.323 rad/m, Z1 = 13.928 , C1 = 1 pF, L = 2.7 nH, C = 5 pF, L1 = 11 nH and f =
3.458 GHz [6]. In this example, three unit cells (M = 3) terminated in ZL =1.153 + j 1.153  are
employed for illustration. Using the cascading property of the ABCD matrix (see Appendix A), the
total ABCD matrix for this case is found to be
A B −0.749 j 1.857
= . (5.11)
C D j 0.349 − 0.469
◦
Using (2.20), (2.21), and (2.27), Z0± and β are found to be Z0± = 2.306 e±j 10  and β = 37.09
rad/m. Note that the argument φ of Z0− in this case is equal to −10◦ , and the finite periodic TL
structure is operated in a passband since β is real. In addition, the effective propagation constant β of
the periodic TL structure is significantly different from the propagation constant k of the unloaded
transmission line in Figure 2.2.
Using (4.4) and (4.5), the normalized load impedance zL for this example is found to be
zL = rL + jx L = 0.5 + j 0.5. In this case, it is found that rL = 0.5 and xL = 0.5, which can be plotted
as the zL point on the Meta-Smith chart with φ = −10◦ as shown in Figure 5.6. Next, a circle
passing through the zL point is drawn on the Meta-Smith chart. On the chart, the normalized
input impedance zin,M can be found by rotating the zL point clockwise an amount of 2Mβd = 255◦
around the origin of the chart (moving toward the generator) as shown in Figure 5.6. Thus, the zin,M
point can be read as 0.54 – j 0.24, and the input impedance Zin,M can be calculated as

Zin,M = zin,M |Z0± | = 1.25 − j 0.56 . (5.12)

Note that the analytical result obtained from (2.24) is equal to Zin,M = 1.2501 – j 0.5604 , which is
in very good agreement with the one obtained from the computerized Meta-Smith chart in (5.12).
Therefore, the computerized Meta-Smith chart for CCITLs with NNCRs provides accurate results,
and it can be used in analysis and design of terminated finite reciprocal lossless periodic TL structures
indeed.
It should be pointed out that the Smith chart can be used to determine the input impedance
Zin,M of the terminated finite periodic TL structure in this example as well. However, using the
Smith chart for calculation is not convenient compared to using the Meta-Smith chart for CCITLs
with NNCRs, especially when each unit cell is complex and the number of unit cells is large. In
addition, the Smith chart does not provide physical insight in solving this problem compared to the
 5.5. CCITLS WITH NEGATIVE CHARACTERISTIC RESISTANCES 77
Meta-Smith chart since the information on periodicity of the TL structure is not employed in the
Smith chart calculation at all. For this particular example, it is found that both the Meta-Smith chart
for CCITLs with NNCRs and the Smith chart yield almost identical results as expected. In the next
section, the Meta-Smith chart for CCITLs with NCRs is employed to solve problems associated
with reciprocal CCITLs with NCRs.

Figure 5.6: The Meta-Smith chart solution for the terminated finite periodic TL structure with NNCRs
when φ = −10◦ .

5.5 CCITLS WITH NEGATIVE CHARACTERISTIC

RESISTANCES
Consider a reciprocal CCITL with NCRs of length λ/8 with φ = −150◦ , terminated in a nor-
malized load impedance of zL = 0.8 + j 1.2 as shown in Figure 2.4, where λ is the wavelength of
waves propagating along the CCITL. Note that this reciprocal CCITL can be implemented using
appropriate finite reciprocal lossless periodic TL structures [7]. Using (2.26), (4.4), and (4.5), the
normalized input impedance of the terminated CCITL zin can be expressed as

1 + e−j 2β
zin = , (5.13)
z0−− z0+ e−j 2β

where z0± = e∓j φ are the normalized characteristic impedances of the CCITL. The procedures of
using the Z Meta-Smith chart for a CCITL with NCRs for impedance calculation are similar to
those of using the Z Meta-Smith chart for CCITLs with NNCRs given earlier in Section 5.4 due
to the fact that the formulas for computing the normalized input impedance for both cases are in
the same form.
78 5. APPLICATIONS OF META-SMITH CHARTS
Figure 5.7 illustrates the impedance calculation on the Z Meta-Smith chart for the CCITL
with NCRs of the above example. Note that the operating region is outside or on the unit circle for
passive load terminations as shown in Figure 5.7, and the magnitude of the reflection coefficient
at the load || in this case is found to be 2.57 (see (4.6)), which is greater than unity as expected.
Starting at zL = 0.8 + j 1.2, move along the circle passing through zL in the clockwise direction
(moving toward the generator) for λ/8(2βl = 90◦ ) finally arrive at zin = 0.4 + j 0.6, which is very
close to the result obtained from the formula in (5.13); i.e., zin = 0.388 +j 0.606.

x =0
L
3

x = - 0.5
2 L
r =2
L

1 r =1
L
i

0
Γ

z
-1 r =0
L
L
x = 0.5
L
-2

-3 z r = 0.5
in L
-3 -2 -1 0 1 2 3
Γ
r

Figure 5.7: Impedance calculation on the Z Meta-Smith chart for the CCITL with NCRs of length
λ/8 terminated in zL = 0.8 + j 1.2 when φ = −150◦ .

5.6 RECIPROCAL LOSSY BCITLS

In this section, a finite reciprocal lossy periodic TL structure with NNCRs is employed as an example
of reciprocal lossy BCITLs with NNCRs. Consider a reciprocal lossy unit cell of a finite lossy periodic
TL structure terminated in a passive load impedance ZL as illustrated in Figure 3.4 with d = 2 cm,
the characteristic impedance Z1 = 150 , the wavelength along the reciprocal lossless uniform TL
λ = 7.5 cm, a series resistance RS = 75 , a shunt resistance RSH = 200  and ZL = 50  [8],
[9]. Only one unit cell (M = 1) is considered in this example. Note that the propagation constant k
of the reciprocal lossless uniform TL is equal to 2 π/λ. In addition, the length () of the equivalent
BCITL is equal to d in this case. The total ABCD matrix of the unit cell for this case is readily found
 5.7. CONCLUSIONS 79
as
A B 0.063 + j 0.622 −28.550 + j 177.149
= . (5.14)
C D j 0.008 −0.312 + j 0.622
Once the total ABCD matrix is known, the characteristic impedances Z0± and the complex propaga-
tion constant γ of the equivalent BCITL can be determined by using associated formulas given in
Section 3.3. Using (3.10), (3.11), and (3.15), it is numerically found that Z0+ = 149.3 – j 7.9 , Z0− =
146.9 – j 39.6  and γ = α + jβ = 29.5 + j 83.8 l/m. Note that Z0+ and Z0− are not complex conju-
gate of one another as expected since the network is lossy. Using (3.5), it is found that the magnitude
of the reflection coefficient at the load || is equal to 0.505.
Using (4.27), the normalized load impedance zL in this example is found to be zL = rL + jx L
= 0.330 – j 0.035. In this case, it is found that rL = 0.330 and xL = – 0.035, which can be plotted as
the zL point on the Z Meta-Smith chart for BCITLs with NNCRs as shown in Figure 5.8. Due to
the loss of the BCITL, an exponentially attenuating spiral starting from the zL point is drawn on
the Z Meta-Smith chart for BCITLs with NNCRs. On the chart, the normalized input impedance
zin can be found by rotating the zL point clockwise an amount of 2βl = 192◦ along the spiral path
due to the lossy nature of the BCITL (see (3.6) and (3.7) with γ + = γ − = γ for reciprocal lossy
BCITLs). Thus, the zin point can be read from the computerized Z Meta-Smith chart as 1.27 –
j 0.34, and the input impedance Zin can be computed from zin as

Zin = zin Z0+ Z0− = 195.93 − j 30.91 . (5.15)

Alternatively, using the standard cascading approach for the TL calculation in Figure 3.4, it is
found that Zin = 196.09 – j 31.35 , which is in very good agreement with that obtained from
the computerized Z Meta-Smith chart for BCITLs with NNCRs in (5.15). Therefore, the chart
can provide sufficiently accurate results indeed. Note that the usage of the Z Meta-Smith chart for
solving BCITLs with NNCRs as shown above is similar to that of the standard Smith chart for
solving RLSUTLs.

5.7 CONCLUSIONS
In this chapter, the Meta-Smith charts for CCITLs and BCITLs developed in Chapter 4 are applied
to solve TL problems involving impedance matching and impedance calculation. It is found that
graphical solutions based on the computerized Meta-Smith charts are fast, intuitive and accurate,
indeed. In addition, basic procedures of using the Meta-Smith charts for solving CCITL and BCITL
problems are similar to those of applying the standard Smith chart for solving problems associated
with reciprocal uniform TLs. Thus, the Meta-Smith charts are useful in analysis and design of
complicated CCITL and BCITL problems.
It should be pointed out that the Meta-Smith charts are applied to analyze and design pas-
sive circuits only in this book. However, they can be applied for problems associated with active
circuits as well; e.g., active loaded TLs [10]–[11] and microwave transistor amplifiers in the CCITL
80 5. APPLICATIONS OF META-SMITH CHARTS

Figure 5.8: Impedance calculation on the Z Meta-Smith chart for BCITLs with NNCRs for Z0+ = 149.3
– j 7.9 , Z0− = 146.9 – j 39.6  and γ = 29.5 + j 83.8 l/m.

system [12]–[14]. Further applications of the Meta-Smith charts will be illustrated in the future.
Finally, it is hoped that the Meta-Smith charts will be used extensively by students, professors and
researchers in the RF community in the near future to develop a physical understanding of many
sophisticated phenomena of complex TLs and related problems.

BIBLIOGRAPHY
[1] D. M. Pozar, Microwave Engineering, 3rd ed. New Jersey: Wiley, 2005. 67, 68, 71

[2] A. Darawankuland D. Torrungrueng, “A single-stub tuning for nonreciprocal uniformtrans-

mission lines,” in Proceedings of the First ECTI International Conference, May 2004, vol. 1,
pp. 83–86. 68, 70

[3] D. Torrungrueng and C. Thimaporn, “Applications of the ZY T-chart for nonreciprocal stub
tuners,” Microwave and Optical Technology Letters, vol. 45, no. 3, pp. 259–262, May 2005.
DOI: 10.1002/mop.20789 68

[4] F. Urbani, L. Vegni, and A. Toscano, “A generalized Smith chart for an exponential tapered
nonuniform transmission line,” Microwave and Optical Technology Letters, vol. 14, no. 1, pp. 36–
39, Jan. 1997.
DOI: 10.1002/(SICI)1098-2760(199701)14:1%3C36::AID-MOP11%3E3.0.CO;2-A 73
 BIBLIOGRAPHY 81
[5] D. Torrungrueng and C. Thimaporn, “Application of the T-chart for solving exponentially-
tapered lossless nonuniform transmission line problems,” Microwave and Optical Technology
Letters, vol. 45, no. 5, pp. 402–406, Jun. 2005. DOI: 10.1002/mop.20836 73, 74

[6] D. Torrungrueng, C. Thimaporn, and N. Chamnandechakun, “An application of the T-Chart

for solving problems associated with terminated finite lossless periodic structures,” Microwave
and Optical Technology Letters, vol. 47, no. 6, pp. 594–597, Dec. 2005. DOI: 10.1002/mop.21239
76

[7] S. Lamultree and D. Torrungrueng, “On the characteristics of conjugately characteristic-

impedance transmission lines with active characteristic impedance,” in 2006 Asia-Pacific Mi-
crowave Conference Proceedings, Dec. 2006, vol. 1, pp. 225–228.
DOI: 10.1109/APMC.2006.4429411 77

[8] D. Torrungrueng, P.Y. Chou, and M. Krairiksh, “A graphical tool for analysis and designof bi-
characteristic-impedance transmission lines,” Microwave and OpticalTechnologyLetters, vol. 49,
no. 10, pp. 2368–2372, Oct. 2007. DOI: 10.1002/mop.22801 78

[9] D. Torrungrueng, P.Y. Chou, and M. Krairiksh, “Erratum: A graphical tool for analysis and
design of bi-characteristic-impedance transmission lines,” Microwave and Optical Technology
Letters, vol. 51, no. 4, pp. 1154, Apr. 2009. 78

[10] C. Lertsirimit and D. Torrungrueng, “Analysis of active loaded transmission line using an
equivalent BCITL model,” in 2007 Asia-Pacific Microwave Conference Proceedings, Dec. 2007,
vol. 4, pp. 2353–2356. DOI: 10.1109/APMC.2007.4554830 79

[11] C. Lertsirimit and D. Torrungrueng, “Graphical tool in analysis of active loadedtransmission

lines,” in Proceedings of the 2009 ECTI International Conference, May 2009,vol. 2, pp. 826–829.
DOI: 10.1109/ECTICON.2009.5137173 79

[12] R. Silapunt and D.Torrungrueng,“A novel analysis of two-port networks in the system of conju-
gately characteristic-impedance transmission lines (CCITLs),” in 2005 Asia-Pacific Microwave
Conference Proceedings, Dec. 2005, vol. 3, pp. 1662–1665. DOI: 10.1109/APMC.2005.1606613
80

[13] R. Silapunt and D. Torrungrueng, “The VSWR considerations in potentially unstablemi-

crowave amplifiers design in the CCITL system,” in 2007 Asia-Pacific Microwave Conference
Proceedings, Dec. 2007, vol.2, pp. 811–814. DOI: 10.1109/APMC.2007.4554840

[14] R. Silapuntand D. Torrungrueng, “An alternative approach in deriving associated circleequa-

tions for microwaveamplifiers in the CCITL system,” in Proceedings of the 2008 ECTI Inter-
national Conference, May 2008, vol. 1, pp. 233–236. DOI: 10.1109/ECTICON.2008.4600415
80
82 5. APPLICATIONS OF META-SMITH CHARTS

5.8 PROBLEMS
5.1. Show that the input impedance of nonreciprocal lossless uniform OC or SC stubs is purely
imaginary.
5.2. Design nonreciprocal lossless single-stub shunt tuners implemented using an OC stub as
shown in Figure 5.1 (except that the stub is connected in shunt) when ZL = 50 , |Z0± |=
50 , and φ = 45◦ .
5.3. Design nonreciprocal lossless double-stub shunt tuners implemented using OC nonrecip-
rocal stubs as shown in Figure 5.3 when YL = 10 + j 50 S, |Y0± |= 50 S, and φ = 30◦ , and
d = 3 λ̃/8. Compute the normalized conductance gmax defined in (5.6) as well. Interpret
your result in terms of a forbidden region.
5.4. Verify that the equation of the conductance circle governing the forbidden region is given
as in (5.5).
5.5. Using the Meta-Smith chart for CCITLs with NNCRs, compute the input impedance
of the terminated ETLNUTLs with C0 = 30 pF/m, q = 2 Np/m, f = 5 GHz, L0 = 1.688
nH/m, β = 45◦ and ZL =7.5  (defined in Section 5.3.2). Compare this graphical solution
with the analytical one.
5.6. Verify the total ABCD matrix in (5.11).
5.7. In Section 5.4, the Meta-Smith chart for CCITLs with NNCRs is employed to determine
the input impedance Zin,M in (5.12) of the terminated finite reciprocal lossless periodic TL
structure of three unit cells in Figure 2.3, where each unit cell is given as in Figure 2.2. Using
the Smith chart, determine Zin,M for this case, and compare with the result in (5.12). In
addition, if the number of unit cells is increased to six, determine Zin,M using the Meta-
Smith chart for CCITLs with NNCRs and the Smith chart, and compare these results. Is
the Smith chart convenient to use in computing Zin,M for these cases? Explain in detail.
5.8. Using the Z Meta-Smith chart for CCITLs with NCRs, compute the normalized input
impedance of the reciprocal CCITL with NCRs of length λ/8 with φ = 150◦ , terminated
in a normalized load impedance of zL = 0.8 + j 1.2 as shown in Figure 2.4, where λ is the
wavelength of waves propagating along the CCITL.
5.9. In Section 5.6, the Z Meta-Smith chart for BCITLs with NNCRs is employed to determine
the input impedance Zin in (5.15) of the terminated reciprocal lossy unit cell in Figure 3.4.
Using the Smith chart, determine Zin for this case, and compare with the result in (5.15).
 83

APPENDIX A

The Transmission (ABCD)

Matrix
The transmission (ABCD) matrix is a useful representation to characterize a two-port network,
especially for two-port networks with cascade connection. Figure A.1 illustrates a two-port network
represented by the transmission (ABCD) matrix. The ABCD matrix is defined in terms of the total


Figure A.1: A two-port network represented by the transmission (ABCD) matrix.

voltages (V1 and V2 ) and total currents (I1 and I2 ) as [1]

V1 AB V2
= . (A.1)
I1 CD I2

Note that the ABCD matrix relates the voltage and current at port 1 of the network (V1 and I1 )
to those at port 2 (V2 and I2 ). From the defining relations of (A.1), each ABCD parameter can be
determined as follows:
V1
A= , (A.2)
V2 I2 =0
V1
B= , (A.3)
I2 V2 =0
I1
C= , (A.4)
V2 I2 =0
I1
D= . (A.5)
I2 V2 =0
84 A. THE TRANSMISSION (ABCD) MATRIX
Note that A and C are OC parameters; i.e., they are determined by open-circuiting at port 2 (I2 =
0). However, B and D are SC parameters; i.e., they are determined by short-circuiting at port 2 (V2
= 0). It should be pointed out that A and D are dimensionless, while the units of B and C are in 
and S, respectively. The ABCD parameters of some useful two-port circuits can be found in [1].
Figure A.2 shows a cascade connection of two two-port networks. The ABCD parameters
of the first two-port network are A1 , B1 , C1 , and D1 , while those of the second two-port network
are A2 , B2 , C2 , and D2 . Using (A.1), the relationships of the voltage and current at each port in

Figure A.2: A cascade connection of two two-port networks.

Figure A.2 can be written as

V1 A1 B1 V2
= , (A.6)
I1 C1 D1 I2
V2 A2 B2 V3
= . (A.7)
I2 C2 D2 I3

Substituting (A.7) into (A.6) yields

V1 A1 B1 A 2 B2 V3
= . (A.8)
I1 C1 D1 C2 D2 I3

This shows that the ABCD matrix of the cascade connection of two two-port networks is equal to
the multiplication of the ABCD matrices of each two-port network. This result can be generalized to
three or more two-port networks in a straightforward manner. In summary, the ABCD matrix of the
cascade connection of two or more two-port networks is equal to the multiplication of the ABCD
matrices of each two-port network, where the matrix-multiplication order must be identical to the
order in which the networks are arranged. Thus, the ABCD matrix of more complicated networks
consisting of cascades of simpler two-port networks can be readily found using the above property
of the ABCD matrix.

BIBLIOGRAPHY
[1] D. M. Pozar, Microwave Engineering, 3rd ed. New Jersey: Wiley, 2005. 83, 84
 85

APPENDIX B

The Required Condition for

Nonreciprocal CCITLs
In this appendix, it will be shown that β +  = β − is the required condition for nonreciprocal CC-
ITLs (e.g., nonreciprocal lossless uniform TLs), where β + and β − are the propagation constants
of waves propagating in forward and reverse directions along nonreciprocal CCITLs, respectively.
Using (2.1), (2.2) and the definitions of ABCD parameters given in Appendix A (see (A.2) to (A.5)),
the transmission (ABCD) matrix of a nonreciprocal CCITLs of length  as shown in Figure 2.1 can
be determined as
 + −
 + −
 
AB 1 Z0+ ejβ  +Z0− e−jβ  Z0+ Z0− ejβ  − e−jβ 
= + . (B.1)
CD Z0 + Z0− +
ejβ  − e−jβ 
− + −
Z − ejβ  +Z + e−jβ 
0 0

For nonreciprocal CCITLs, it is required that the determinant of the ABCD matrix in (B.1) must
not be equal to unity [1]. After some straightforward manipulations, it is found that

AB + −β − )
det = ej (β . (B.2)
CD

From (B.2), the determinant of the ABCD matrix is not equal to unity if and only if β + = β −
for an arbitrary length  of nonreciprocal CCITLs. Thus, β +  = β − is the required condition for
nonreciprocal CCITLs indeed.

BIBLIOGRAPHY
[1] D. M. Pozar, Microwave Engineering, 3rd ed. New Jersey: Wiley, 2005. 85
 87

APPENDIX C

Derivation of Z0+ = (Z0−)∗ for

CCITLs
In Section 2.2, the traveling wave equations for the phasor voltage V (z) and the phasor current I (z)
on CCITLs (e.g., nonreciprocal lossless uniform TLs) are given as in (2.1) and (2.2), respectively.
Using the fact that CCITLs are lossless, the time-average power flow Pav (z) along CCITLs at an
arbitrary point z must be constant. By definition, Pav (z) can be computed as
1  
Pav (z) = Re V (z)I ∗ (z) . (C.1)
2
Substituting (2.1) and (2.2) into (C.1) yields
  ∗  ∗ 2
V0+ V0− V0+ j (β + +β − )z V0+ V0− −j (β + +β − )z V0−
2
1
Pav (z) = Re  + ∗ +  + ∗ e −  − ∗ e −  − ∗ . (C.2)
2 Z0 Z0 Z0 Z0
Using the fact that Pav (z) is constant, the rate of change of Pav (z) with respect to z must be equal
to zero; i.e.,
d
(Pav (z)) = 0. (C.3)
dz
Substituting (C.2) into (C.3) yields
  ∗ 
+  − ∗
j V0− V0+  +  + − j V V   + −
Re  + ∗ β +β − ej (β +β )z + 0 − 0∗ β + + β − e−j (β +β )z = 0. (C.4)
Z0 Z0

Using the fact that β + and β − are real, (C.4) can be rewritten compactly as

Re {j T } = 0, (C.5)

where  ∗  ∗
V0− V0+ j (β + +β − )z V0+ V0− −j (β + +β − )z
T ≡  + ∗ e +  − ∗ e . (C.6)
Z0 Z0
To satisfy (C.5), it is required that the term T must be real. Note that (C.6) can be rewritten as
  ∗ ∗  ∗
V0+ V0− −j (β + +β − )z V0+ V0− −j (β + +β − )z
T = e +  − ∗ e . (C.7)
Z0+ Z0
88 C. DERIVATION OF Z0+ = (Z0− )∗ FOR CCITLS
To make the term T real, it is obvious from (C.7) that

Z0+ = (Z0− )∗ . (C.8)

Thus, CCITLs possess Z0+ = (Z0− )∗ indeed.

 89

APPENDIX D

Derivation of the Magnitude of

the Voltage Reflection
Coefficient at the Load for
Terminated CCITLs
This appendix shows the derivation of the magnitude of the voltage reflection coefficient at the load,
||, for terminated CCITLs, where  is defined as in (2.7). For convenience in derivation,  is
rewritten in terms of normalized impedances as
 +
− 2 zL − z 0
 = (z0 )  , (D.1)
zL + z0−

where zL is the normalized load impedance and z0± are the normalized characteristic impedances
(see (4.4), (4.5), and (4.7)). Using the fact that |z0− | = 1 and defining zL ≡ rL + j xL , || can be
expressed as
1/2
(rL − cos φ)2 + (xL + sin φ)2
|| = , (D.2)
(rL + cos φ)2 + (xL + sin φ)2
where rL ≥ 0 for passive loads and φ is the argument of Z0− . In (D.2), when rL is equal to zero, ||
is always equal to one. In addition, it is obvious that the denominator will be greater than or equal
to the numerator for the NNCR case since cos φ ≥ 0. Therefore, || is always less than or equal
to unity for this case. On the other hand, || is always greater than or equal to unity for the NCR
case due to the fact that the numerator is always greater than or equal to the denominator when
cos φ < 0.
Since || can be greater than or equal to unity for the NCR case, one might wonder if this
results in the violation of the conservation of power, especially when || approaches infinity (when
zL = −z0− ). For the NCR case, || can approach infinity with passive loads (Re{zL } ≥ 0) due to
the fact that (Re{z0− } < 0). Appendix E shows that the associated power is still conserved for this
case with passive load terminations.
 91

APPENDIX E

Power Consideration for

Terminated Finite Reciprocal
Lossless Periodic Transmission
Line Structures
Figure E.1 illustrates a finite reciprocal lossless periodic TL structure of M unit cells connected to a
passive load impedance ZL and a generator, where ES and ZS are the source voltage and the source
impedance, respectively. The total time-average power Pm distributed along the mth terminal of the

I0 I1 I2 IM
ZS + 1
st
+ + 2
nd
+ M th +
ES V0 Unit V1 Unit V2 Unit VM Z
L
− cell − − cell − cell −
Zin,M d
M Unit cells

Figure E.1: A finite reciprocal lossless periodic TL structure of M unit cells connected to a passive load
impedance ZL and a generator.

periodic TL structure can be computed as

1  
Pm = Re Vm Im∗ , (E.1)
2

where the phasor voltage Vm and the phasor current Im in the passsbands at the terminal of the mth
unit cell (where m = 1, 2, …, M) can be written as (see (2.16), (2.17), and (2.27))

Vm = V0+ e−j mβd + V0− ej mβd , (E.2)

V+ V−
Im = 0+ e−j mβd − 0− ej mβd , (E.3)
Z0 Z0
92 E. POWER CONSIDERATION FOR TERMINATED FINITE RECIPROCAL TLS

where Z0± are the effective characteristic impedances and β is the effective propagation constant of
the periodic TL structure. Substituting (E.2) and (E.3) into (E.1), Pm in passbands can be expressed
compactly as follows [1]:
 
|V0 |2 1 − ||2
Pm = cos φ, (E.4)
2 |Z0 | 1 + e−j 2Mβd 2

where the voltage reflection coefficient at the load  in (E.4) is defined as

V0− ej Mβd
≡ ≡ 0 ej 2Mβd . (E.5)
V0+ e−j Mβd

Note that 0 is the voltage reflection coefficient at the input, where 0 is defined as

0 ≡ V0− /V0+ , (E.6)

and the input voltage V0 can be written in terms of 0 as

V0 = V0+ (1 + 0 ) . (E.7)

Alternatively, V0 can be determined from the circuit in Figure E.1 as

Zin,M
V0 = Es , (E.8)
Zin,M + Zs

where Zin,M is the input impedance of the terminated periodic TL structure. It can be seen from (E.4)
that, for both NNCR (|| ≤ 1) and NCR cases (|| ≥ 1), Pm is still nonnegative since cos φ ≥ 0
for the NNCR case and cos φ < 0 for the NCR case.
From (E.4) and (E.8), if || approaches infinity (when zL = −z0− as discussed in Appendix D)
, Pm becomes [1]
2 2
1 Zin,M Es
Pm = − cos φ , (E.9)
2 |Z0 | Zin,M + Zs
which is nonnegative as well since cos φ < 0 for the NCR case. Thus, the power is still conserved
for the NCR case with passive load terminations even when || approaches infinity. For a fixed
passive load impedance and a fixed generator, Pm is constant along each terminal of the periodic TL
structure operating in passbands since the structure of interest is lossless.

BIBLIOGRAPHY
[1] S. Lamultree and D. Torrungrueng, “On the characteristics of conjugately characteristic-
impedance transmission lines with active characteristic impedance,” in 2006 Asia-Pacific Mi-
crowave Conference Proceedings, vol. 1, pp. 225–228, Dec. 2006. 92
 93

Author’s Biography

DANAI TORRUNGRUENG
Danai Torrungrueng received his B.Eng. degree in electrical engi-
neering from Chulalongkorn University, Bangkok, Thailand, in 1993.
He obtained his M.S. and Ph.D. degrees in electrical engineering from
The Ohio State University in 1996 and 2000, respectively. From 1995
to 2000, he was a Graduate Research Assistant (GRA) in the De-
partment of Electrical Engineering, ElectroScience Laboratory of The
Ohio State University. Prior to joining Asian University, he worked as
a senior engineer in the USA, involved in research and development of
the urban propagation modeling project. At present, he is an associate
professor in the Electrical and Electronic Engineering Department in
the Faculty of Engineering and Technology at Asian University, Thai-
land.
In 2000, he won an award in the National URSI Student Paper competition at the 2000
National Radio Science Meeting in Boulder, Colorado. During 2004 to 2009, he invented and
developed Meta-Smith charts (visit http://www.meta-smithcharts.org) for solving prob-
lems associated with conjugately characteristic-impedance transmission lines (CCITLs) and bi-
characteristic-impedance transmission lines (BCITLs) graphically. His research interests are in the
areas of fast computational electromagnetics, antenna design, RFID, rough surface scattering, prop-
agation modeling and electromagnetic wave theory. He is currently a senior member of the IEEE
and a member of the Electrical Engineering/Electronics, Computer, Telecommunications and In-
formation Technology Association of Thailand (ECTI).
 95

Index

ABCD matrix, 22, 23, 29, 76, 78, 79, 82–85 characteristic impedance, 4, 6–11, 17–19, 21,
arithmetic mean, 19, 33, 46, 69 23–25, 28, 31–33, 35, 36, 41–43, 46,
55, 60–62, 68, 73, 74, 76–79, 81, 89,
BCITL, 31, 32, 35–38, 41, 42, 53, 54, 56–65, 91, 92
67, 78–82 conformal mapping, 52, 57
ABCD matrix, 78 cosine-based phasor, 1
input impedance, 31, 35, 79, 82
nonreciprocal, 31, 32 double-stub shunt tuning, 70, 71
reciprocal, 31, 32, 35–37, 53, 54, 56, 67,
78, 79, 82 effective characteristic impedance, 11, 24, 35,
reflection coefficient, 31, 41, 53, 54, 59 73, 74, 76, 91
Y Meta-Smith chart for, 56–61, 63–65 effective negative refractive index, 28
Z Meta-Smith chart for, 56–59, 62, 79, effective propagation constant, 23, 24, 34, 76,
80, 82 91
bilinear transformation, 52 eigenanalysis, 22, 23, 29
equation, 1–3, 8, 10, 11, 13–15, 18, 21, 23, 24,
CCITL, 17–21, 23, 24, 26, 27, 31, 32, 35, 37, 32–35, 43, 44, 47, 49, 53–56, 63, 64,
41–44, 46–54, 56, 59, 60, 62–64, 71–73, 75, 81, 82, 87
67–69, 72–82, 85, 87–89 conductance circle, 71, 72, 82
ABCD matrix, 85 ETLNUTL, 10, 73
condition for nonreciprocal, 85 reactance circle, 14, 43, 55, 56, 63, 64
NCR, 17, 20, 41–44, 46–54, 59, 63, 64, Ricatti-type differential, 8
67–69, 72–78, 82 susceptance circle, 15, 44, 56, 63, 64
NNCR, 17, 20, 41–44, 46–50, 53, 59, 63, TL, 1–3, 8, 10, 13, 14, 18, 24, 32, 35, 44,
64, 67–69, 72–77, 82 47, 49, 53, 54, 56, 63, 64, 73, 75, 81,
normalized input impedance, 77, 82 87
reciprocal, 17–20, 23, 24, 32, 35, 46, 48, ETLNUTL, 9–12, 15, 17, 20, 29, 72–75, 82
52, 67, 68, 73, 75–77, 79, 82, 85, 87 admittance, 9
Y Meta-Smith chart for, 46, 48, 49, characteristic impedance, 9, 11, 73
51–54, 56, 63, 64 equation, 10, 73
Z Meta-Smith chart for, 50, 51, 56, 68, impedance per unit length , 9
69, 77, 78, 82 input impedance, 12, 15, 73, 74, 82
96 INDEX
load impedance, 9, 73 impedance, 6, 8, 10, 12–15, 17, 19, 22–24,
normalized impedance, 73 29, 31, 33, 35, 36, 39, 46–48, 52, 68,
reflection coefficient, 12 73–79, 82, 92
reflection coefficient, 6, 12, 13, 17, 31, 68,
finite reciprocal lossless periodic TL structure,
91
17, 20, 22–24, 33, 48, 67, 75–77, 82,
voltage, 6, 13, 17, 22, 31, 33, 68, 91, 92
91
intersecting point, 50
reflection coefficient, 48
finite reciprocal lossy periodic TL structure, 32, lossless periodic TL, 17, 20–24, 26, 29, 33, 48,
35–37, 78 67, 75–77, 82, 91
input impedance, 35 lossless uniform TL, 8, 19, 21, 25, 36, 78, 85, 87
forbidden region, 71, 72, 82 lossy periodic TL, 32–37, 57, 58, 78
geometric mean, 42, 47 lossy uniform TL, 6, 7, 53, 54
reflection coefficient, 53
impedance, 1, 2, 4–15, 17–25, 28, 29, 31–36,
39, 41–44, 46–49, 52–55, 60–62, Meta-Smith chart, 13, 41, 42, 44–54, 56–65,
67–70, 73–79, 81, 82, 89, 91, 92 67–80, 82
Bloch, 22, 23, 29, 34 BCITL, 41, 42, 53, 54, 56–65, 67, 79, 80,
characteristic, 4, 6–11, 17–19, 21, 23–25, 82
28, 31–33, 35, 36, 41–43, 46, 54, 55, CCITL, 41, 42, 44, 46–54, 56, 59, 60, 63,
60–62, 68, 73, 74, 76–79, 81, 89, 91, 64, 67–69, 72–79, 82
92 NCR, 41, 42, 44, 46–54, 56–65, 67–69,
ETLNUTL, 9, 11, 12, 15, 73, 74, 82 72–80, 82
input, 6, 8, 10, 12–15, 17, 19, 23, 24, 29, NNCR, 41, 42, 44, 46–50, 53, 56–61, 63,
31, 33, 35, 36, 39, 46–48, 52, 68, 64, 67–69, 72–77, 79, 80, 82
73–79, 82, 92 metamaterial TL, 21
load, 1, 2, 5–11, 14, 15, 17–19, 21–25, 29,
31–36, 41, 43, 53, 54, 60, 68, 69, 73, NCR, 17, 24–27, 29, 35–37, 41, 42, 48–54,
74, 76–79, 82, 89, 91, 92 56–60, 62–65, 67, 77, 78, 82, 89, 92
matching, 8, 9, 67, 79 CCITL with, 67, 77, 78, 82
normalized, 7, 8, 14, 42–44, 46, 54, 55, 69, Y Meta-Smith chart for , 48, 49, 51–54,
70, 73–79, 82, 89 58, 59, 63–65
normalized characteristic, 43, 54, 55, 77, Z Meta-Smith chart for, 50, 51, 58, 62,
89 77, 78, 82
normalized input, 75–79, 82 ZY Meta-Smith chart for, 48, 49, 51–54,
normalized load, 7, 8, 14, 43, 54, 69, 74, 59, 64, 65
76, 77, 79, 82, 89 negative group delay, 21
input, 6, 8, 10, 12–15, 17, 19, 22–24, 29, 31, 33, NNCR, 17, 20, 24–27, 29, 35–37, 41–44,
35, 36, 39, 46–48, 52, 68, 70, 71, 46–50, 53, 56–61, 63, 64, 67–70,
73–79, 82, 91, 92 72–80, 82, 89, 92
 INDEX 97
CCITL with, 67 reciprocal lossless periodic TL, 17, 20–24, 29,
Y Meta-Smith chart for , 46, 48, 49, 57, 33, 48, 67, 75–77, 82, 91
59–61, 63, 64 reciprocal lossy periodic TL, 32–37, 78
Z Meta-Smith chart for, 57–59, 68, 69, reflection coefficient, 6–8, 12–15, 17, 19,
77, 79, 80, 82 23–25, 27, 29, 31–33, 35, 41, 42, 48,
ZY Meta-Smith chart for, 46, 48, 49, 57, 49, 52–54, 59, 68, 74, 76, 78, 79, 89,
61, 63, 64 91
nonreciprocal double-stub shunt tuning, 70, 71 Ricatti-type differential equation, 8
nonreciprocal single-stub series tuner, 68, 70 RLSUTL, 7, 32, 79
nonreciprocal TL, 18, 68 RLUTL, 17, 20, 73, 74
characteristic impedance, 18
short-circuited, 36, 46, 67
input impedance, 68
single-stub series tuner, 68–70
normalized impedance, 7, 14, 42, 44, 46, 54,
Smith chart , 7, 14, 15, 27, 37, 41, 42, 44–46,
69, 70, 73, 89
59, 60, 67, 73, 74, 76, 77, 79, 80, 82
NRI, 21
stopband, 21, 26
NRLSUTL, 32, 33, 39
symmetric unit cell, 23, 29
NRLUTL, 17, 19, 20, 29, 32, 33, 69
T-Chart, 61, 62, 81
open-circuited , 36, 46 time-average power, 18, 87, 91
TL, 1–10, 13–15, 17–27, 29, 31–37, 39, 41, 48,
passband, 17, 21–26, 48, 75, 76, 91, 92 53, 54, 57, 58, 60, 67, 68, 73, 75–80,
periodic TL structure, 17, 20–26, 29, 32–37, 82, 85, 87, 91, 92
48, 57, 58, 67, 75–78, 82, 91, 92 characteristic impedance, 4, 6, 8, 18, 21,
phasor, 1–3, 10, 11, 15, 18, 20, 21, 29, 32, 33, 24, 25, 35, 68, 78
87, 91 equation, 1–3, 5, 10, 13, 14, 18, 23, 24, 35,
cosine-based, 1 68, 75
current, 1–3, 10, 11, 15, 18, 20, 21, 29, 32, parameter, 2, 4, 7, 13, 15, 18, 24, 29, 36,
33, 87, 91 39, 73
voltage, 1, 3, 10, 11, 15, 18, 20, 21, 29, 32, phase velocity, 4, 25, 29, 39
33, 87, 91 phasor current, 18, 87
propagation constant, 3, 6, 8, 11, 17–19, 21, phasor voltage, 18
23–25, 31, 33, 34, 36, 68, 76, 78, 79, propagation constant, 3, 6, 8, 18, 21,
85, 91 23–25, 34, 76, 78, 85, 91
theory, 1, 2, 17, 18
reciprocal BCITL, 35
touching point, 44–46, 50–52
input impedance, 35 transmission matrix, see ABCD matrix
reciprocal CCITL, 17, 20, 23, 24, 73, 77, 82 two-port network, 1, 81, 83, 84
input impedance, 23, 82
normalized input impedance, 82 unit cell, 21–26, 29, 33–36, 39, 75–79, 82, 91
98 INDEX
unsymmetrical unit cell, 21, 22, 25, 29, 33, 34, Z Meta-Smith chart for BCITL, 56–59, 62,
36, 39, 75 79, 80, 82
Z Meta-Smith chart for CCITL, 50, 51, 68,
69, 77, 82
voltage reflection coefficient, 6–8, 12–15, 17,
normalized load impedance, 69
19, 23–25, 27, 29, 31–33, 35, 41, 42,
48, 49, 52–54, 59, 68, 74, 76, 89, 91 reactance circle, 50
ZY Meta-Smith chart for BCITL, 57, 59, 61,
64, 65
Y Meta-Smith chart for BCITL, 56–60, 63 ZY Meta-Smith chart for CCITL, 46, 48, 49,
susceptance circle, 57–59 51–54, 63, 64
Y Meta-Smith chart for CCITL, 51 ZY Smith chart , 42, 44, 60, 73