Get the full ebook with Bonus Features for a Better Reading Experience on ebookgate.

com

Time domain computer analysis of nonlinear hybrid

systems 1st Edition Wenquan Sui

https://ebookgate.com/product/time-domain-computer-analysis-
of-nonlinear-hybrid-systems-1st-edition-wenquan-sui/

OR CLICK HERE

DOWLOAD NOW

Download more ebook instantly today at https://ebookgate.com

Instant digital products (PDF, ePub, MOBI) available
Download now and explore formats that suit you...

Analysis and Synthesis of Computer Systems 2nd Edition

Erol Gelenbe

https://ebookgate.com/product/analysis-and-synthesis-of-computer-
systems-2nd-edition-erol-gelenbe/

ebookgate.com

Real Time Systems Design and Analysis 3rd Edition Bruce

Fraser

https://ebookgate.com/product/real-time-systems-design-and-
analysis-3rd-edition-bruce-fraser/

ebookgate.com

Nonlinear Control Systems 1st Edition Zoran Vuki■

https://ebookgate.com/product/nonlinear-control-systems-1st-edition-
zoran-vukic/

ebookgate.com

Real Time Systems Scheduling Analysis and Verification 1st

Edition Albert M. K. Cheng

https://ebookgate.com/product/real-time-systems-scheduling-analysis-
and-verification-1st-edition-albert-m-k-cheng/

ebookgate.com
Nonlinear Dynamics of Production Systems 1st Edition
Günter Radons

https://ebookgate.com/product/nonlinear-dynamics-of-production-
systems-1st-edition-gunter-radons/

ebookgate.com

Control of Nonlinear Distributed Parameter Systems 1st

Edition Goong Chen

https://ebookgate.com/product/control-of-nonlinear-distributed-
parameter-systems-1st-edition-goong-chen/

ebookgate.com

Two Hybrid Systems Methods and Protocols 1st Edition

Stanley Fields

https://ebookgate.com/product/two-hybrid-systems-methods-and-
protocols-1st-edition-stanley-fields/

ebookgate.com

Domain Oriented Systems Development Practices and

Perspectives 1st Edition Kiyoshi Itoh

https://ebookgate.com/product/domain-oriented-systems-development-
practices-and-perspectives-1st-edition-kiyoshi-itoh/

ebookgate.com

The Patterns of Symbolic Communication 1st Edition Sui Yan

https://ebookgate.com/product/the-patterns-of-symbolic-
communication-1st-edition-sui-yan/

ebookgate.com
 Time-Domain
Computer Analysis
of Nonlinear
Hybrid Systems

Wenquan Sui
Conexant Systems
Chelmsford, Massachusetts

CRC PR E S S
Boca Raton London New York Washington, D.C.

©2002 CRC Press LLC

 Library of Congress Cataloging-in-Publication Data

Sui, Wenquan.
Time-domain computer analysis of nonlinear hybrid ststems / Wenquan Sui.
p. cm.
Includes bibliographical references and index.
ISBN 0-8493-1396-1 (alk. paper)
1. Electronic circuits—Computer simulation. 2. Electromagnetism—Computer
simulation. 3. System analysis—Data processing. 4. Time domain analysis. 5. Finite
differences. I. Title.

TK7867 .S95 2001

621.3815—dc21 2001043256

This book contains information obtained from authentic and highly regarded sources. Reprinted material
is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable
efforts have been made to publish reliable data and information, but the author and the publisher cannot
assume responsibility for the validity of all materials or for the consequences of their use.

Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic
or mechanical, including photocopying, microfilming, and recording, or by any information storage or
retrieval system, without prior permission in writing from the publisher.

The consent of CRC Press LLC does not extend to copying for general distribution, for promotion, for
creating new works, or for resale. Specific permission must be obtained in writing from CRC Press LLC
for such copying.

Direct all inquiries to CRC Press LLC, 2000 N.W. Corporate Blvd., Boca Raton, Florida 33431.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are
used only for identification and explanation, without intent to infringe.

Visit the CRC Press Web site at www.crcpress.com

© 2002 by CRC Press LLC

No claim to original U.S. Government works

International Standard Book Number 0-8493-1396-1
Library of Congress Card Number 2001043256
Printed in the United States of America 1 2 3 4 5 6 7 8 9 0
Printed on acid-free paper

©2002 CRC Press LLC

©2002 CRC Press LLC

 TABLE OF CONTENTS

Preface

The Author

Chapter 1 Introduction

1.1 Introduction
1.2 Electromagnetic Systems
1.3 Hybrid Electromagnetic Systems
1.4 Organization of the Book

Chapter 2 Electromagnetic Field Theory

2.1 Introduction
2.2 Electromagnetic Theory
2.2.1 Coulomb’s law
2.2.2 Gauss’s law
2.2.3 Faraday’s law
2.2.4 Ampere’s law
2.2.5 Continuity equation
2.2.6 Magnetic vector potential
2.2.7 Maxwell’s equations
2.2.8 Wave equations and field retardation
2.2.9 Time-harmonic field solution
2.2.10 Boundary conditions
2.3 Example of Solving Electromagnetic Field Distribution

Chapter 3 Circuit Equivalence and Transmission Line Theory

3.1 Circuit Theory as Field Approximation

3.1.1 Circuit basis under quasi-static approximation
3.1.2 Circuit equations for some lumped elements
3.1.3 Circuit model at different frequency ranges
3.1.4 Transient response of a lumped circuit
3.2 Transmission Line Theory
3.2.1 General transmission line solution
3.2.2 Lossless transmission line
3.2.3 Lumped-element equivalent model for a transmission line
3.3 Scattering Parameters of an n-port Network
3.3.1 Definition of S parameters
3.3.2 Definitions of other network parameters

©2002 CRC Press LLC

©2002 CRC Press LLC

Chapter 4 Finite-Difference Formulation

4.1 Introduction
4.2 Finite-Difference Method
4.2.1 Forward, backward and central differences
4.2.2 Finite-difference approximation in a nonuniform grid
4.3 System Solution and Stability Condition
4.3.1 Jacobian matrix and system solution
4.3.2 Application example
4.3.3 Stability condition

Chapter 5 Solving Electromagnetic Fields in the Time Domain-FDTD Method

5.1 Introduction
5.2 Finite-Difference Time-Domain Method
5.2.1 Maxwell's equations
5.2.2 Three-dimensional FDTD formulation
5.2.3 Two-dimensional FDTD formulation .
5.3 Issues of FDTD Numerical Implementation
5.3.1 Stability condition
5.3.2 Absorbing boundary conditions
5.3.3 Unconditionally stable FDTD algorithm
5.3.4 Numerical dispersion in FDTD
5.4 Examples of FDTD Application

Chapter 6 Circuit Formulation and Computer Simulation

6.1 Introduction
6.2 Constitutive Relation of Devices
6.3 Modified Nodal Formulation of Circuit Simulation
6.4 Transient Analysis of Linear Circuit
6.5 Nonlinear Device Models in Circuit Simulation
6.5.1 Diode model
6.5.2 Bipolar junction transistor model
6.5.3 MOS transistor model
6.6 Newton Method for Solving Systems with Nonlinear Devices
6.7 Timestep Control in Transient Simulation

Chapter 7 Formulation for Hybrid System Simulation in the Time Domain

7.1 Introduction
7.2 Maxwell’s Equations and Supplemental Current Equations
7.3 Hybrid Circuit Simulation with Lumped Elements
7.3.1 FDTD equations for RLC components
7.3.2 Examples of hybrid circuit simulation

©2002 CRC Press LLC

©2002 CRC Press LLC

 7.4 Electron Beam in FDTD Simulation
7.4.1 Interaction between electromagnetic field and an electron beam
7.4.2 FDTD algorithm for modeling an electron beam
7.4.3 Electron-beam modeling for a planar DC diode
7.4.4 Small-signal space-charge waves in FDTD

Chapter 8 Interfacing FDTD Field Solver with Lumped Systems

8.1 Introduction
8.2. Linking FDTD Method with a SPICE-like Circuit Simulator
8.2.1 Equivalent circuit model of a distributed system
8.2.2 Implementation of the circuit-field model for hybrid simulation
8.2.3 Example of the circuit-field model in FDTD
8.3 Modeling a Multiport S-Parameter Network in FDTD
8.3.1 Scattering parameters, port voltage, and port current
8.3.2 Modeling a S-parameter block in FDTD grid
8.4 Multiport Behavioral Model in FDTD
8.4.1 Behavioral model
8.4.2 Behavioral model block in an FDTD grid
8.5 Examples of General Hybrid System Cosimulation

Chapter 9 Simulation of Hybrid Electromagnetic Systems

9.1 Introduction
9.2 FDTD Characterization and De-embedding
9.3 Examples of Hybrid System Cosimulation
9.3.1 Commercial simulators
9.3.2 Application of the circuit-field model
9.3.3 Application of the multiport model
9.3.4 General hybrid system cosimulation
9.4 Analysis of Packaging Structure with On-chip Circuits
9.4.1 Analysis of packaging structures
9.4.2 Simulation of packaging structures with on-chip circuits

Chapter 10 Optical Device Simulation in FDTD

10.1 Introduction
10.2 Active Gain Media in VCSEL
10.3 FDTD Formulation for Systems with Nonlinear Gain Media
10.4 FDTD Analysis of VCSEL Structures
10.4.1. One-dimensional structures
10.4.2. Gain media in 2D structures
10.5 Cosimulation for VCSEL Source and other Circuits

©2002 CRC Press LLC

©2002 CRC Press LLC

Appendix I Vector Differential Operators and Vector Identities

I.1 Vector Differential Operators

I.2 Vector Identities

Appendix II Laplace Transformation

References

©2002 CRC Press LLC

©2002 CRC Press LLC

 PREFACE

History will view the last hundred years as an era of scientific discovery and
tremendous technological growth. What was once science fiction has become
reality. Our ancestors dreamed about going to the moon, exploring deep space,
and talking to people on the other side of the world — events we now consider
routine. Scientific research and practical application of knowledge about
electrical phenomena and others made these and other wonders possible.
Though no one knows who discovered electrical phenomena, they were known
to ancient peoples thousands of years ago. Yet only in the last hundred years
have we understood their mysteries and devised ways to apply them to benefit
people everywhere.
Electrical engineers of earlier generations were much less specialized than
engineers of today. The rapid development of the digital integrated circuit and
high-speed electronic systems led to specialization. Engineers who designed
low-frequency digital circuits considered signals in terms of one and zero, but
rarely thought the electrical circuits as distributed electromagnetic systems;
those who worked with microwaves were not interested in low-frequency
circuits. As the rapid development of telecommunication, wired and wireless,
and high-speed electronic systems, engineers start to realize that more need to be
considered in their designs, especially when dealing with some highly nonlinear
systems that include complicated digital circuits and high-frequency radio
systems, a natural outcome leading to time-domain computer analysis.
This book is a summary and review of research work on finite-difference
time-domain (FDTD) cosimulation for hybrid electromagnetic system that
started more than 10 years ago. The fundamental development of the FDTD
numerical technique, based on the time-domain solution of Maxwell’s
equations, was the original idea of Yee more than 30 years ago, and then the
first practical application of the method by Taflove along with contributions
from others. The numerical cosimulation methods described in this book and
used in various applications show great potential for many aspects of future
electrical system design and analysis. More work is still to be done to perfect the
method for practical daily applications, but a solid theoretical foundation and
required computing resources have been developed.
One motivation for writing this book was the need for a detailed discussion
of high-speed electronic system analyses. Another factor was the evolution of
the finite-difference time-domain (FDTD) method in simulating various types of
electrical systems. This book includes many of the recent developments in the
time-domain cosimulation of hybrid electromagnetic systems. While several
books and numerous papers cover different aspects of the subject, this book
concentrates on techniques for integrating time-domain field solutions, analog
circuit analysis, and a lumped-in-nature foreign system (like an n-port nonlinear
circuit) into a unified solution scheme.
The proliferation of wireless and optical communication devices and sub-
micro semiconductor technology triggered a need for circuit simulators able to

©2002 CRC Press LLC

©2002 CRC Press LLC

handle electrical systems at macro and micro levels. More detailed and extensive
studies are needed before such simulators would be ready for commercial use.
The discussions in this book should provide a foundation for work in that area.
Readers should have a solid background in electromagnetic and circuit theories
and numerical methods along with practical experience in applying the theories.
I hope this book will serve as a useful reference for researchers, practicing
engineers, and graduate-level students.
A number of people played important roles in the production of this book.
Drs. Douglas Christensen and Carl Durney kindled my interest in the finite-
difference time-domain (FDTD) method. Dr. Durney helped with the book
proposal submitted to my publisher. My friend and collaborator, Dr. Tong Li,
who co-authored Chapter 6, provided much helpful advice and many simulation
results for the book. Dr. Peter Li finished most of the packaging analyses
presented in this book for his graduate study. Lively discussions with Dr. Albert
Ruehli of IBM’s Yorktown Research Center yielded valuable information about
circuit and field cosimulation in general and his PEEC model in particular.
Part of this book was written at the homes of Dr. Fred Huang in East
Fishkill, NY, during my association with IBM, and Dr. Qing Guo in Marboror,
MA when I joined Conexant last year. I thank them for the hospitality and
interesting discussions we had. Dr. Phillip Frederick, a retired radiologist and
expert on radar and aircraft safety, encouraged me during the years to search for
unknowns and pursue new ideas.
This book would never have been finished without the patience of Nora
Konopka, my corresponding editor at CRC Press. She encouraged me to keep
writing and I appreciate her support.
I am greatly indebted to my family since my wife was under great pressure
while I concentrated on writing this book, mostly in the evenings and weekends
after work. In addition to managing our family during that period, she had to tell
our son why I spent so little time with him. I look forward to going fishing,
playing and watching soccer, and enjoying other activities with my family. I
plan to present a copy of this book to my mother when I visit China. I hope she
will understand that the great sacrifices she and my late father made were
appreciated.
It will be fascinating to see what electromagnetic marvels in research now
will enter commercial production in the near future. I want to share with the
readers the belief of a bright future, even though what what we have today is not
perfect. The ability of the Star Trek crew to zoom around space, cure human
maladies, and teleport to distant locations is not that far-fetched. Technological
progress, particularly in the last half of the 20th century, has far surpassed what
science fiction writers could have imagined earlier in the century. The message
is clear: keep exploring, thinking, and trying. Small setbacks are challenges to
overcome on the way to achievement.

Wenquan Sui
Chelmsford, MA

©2002 CRC Press LLC

©2002 CRC Press LLC

 THE AUTHOR

Wenquan Sui, Ph.D., was born in Qingdao, China. He spent his early age
during one of the most turbulent periods in China’s thousand year history.
Political and ideological differences led to extensive social unrest, which in turn
had a disastrous effect on education at all levels. Many young people lost their
opportunities to pursue higher education. His parents encouraged their children
to pursue academic excellence despite the uncertain political climate, and made
many sacrifices to achieve that goal.
Dr. Sui attended the world-renowned University of Science and Technology
of China (USTC) and earned undergraduate and graduate degrees in electrical
engineering. He then studied and worked in the research group of Drs. Douglas
Christensen and Carl Durney at the University of Utah in Salt Lake City, Utah.
His research there led to his interest in finite-difference time-domain (FDTD)
simulation and its applications in various areas.
Dr. Sui was a member of the technical staff at the Design Automation
Division of Bell Labs, Lucent Technologies, at Murray Hill, NJ, and was a
senior engineer/scientist of IBM’s Microelectronics Division at Fishkill, NY. He
has been a graduate advisor at New Jersey Institute of Technology during the
past years. He is now a principal engineer at the Boston Design Center, part of
the Wireless Communication Division of Conexant Systems, in Chelmsford,
MA, and continues to explore the many facets of numerical simulation in hybrid
electromagnetic systems. He is a senior member of the IEEE.

©2002 CRC Press LLC

©2002 CRC Press LLC

 CHAPTER 1

INTRODUCTION

1.1 Introduction

The focus of this book is about time-domain numerical solution of hybrid

electrical systems, in terms of electromagnetic fields, lumped current and
voltage, or their combinations. An electromagnetic system can be considered
either as distributed or lumped, sometimes as their combination. A distributed
system is a system that demonstrates field behaviors, like propagation and
interaction, and is described by field equations. A lumped circuit, on the other
hand, can be simplified as a system that distance need not to be included in
calculation of the field distribution. For a pure distributed system, field solution
by a field solving technique is desired; for a lumped-only system, there are many
mature methods to solve different states of the circuit. When an electronic
system has both a distributed component and lumped component, the system is
referred to as a hybrid system, and this will be the definition for hybrid system
in this book.
The finite-difference method is the foundation of many numerical
algorithms; its applications can be found in almost all the simulation techniques.
One specific application of the finite-difference technique is the finite-difference
time-domain (FDTD) method, where the finite-difference method is applied in
both time domain and spatial domain to approximate the partial derivatives in
Maxwell’s equations. During the past couple decades, FDTD method has been
utilized extensively for time-domain solutions to the Maxwell’s equations and
systems described by the equations. The method is explained in detail in various
chapters of this book, from its fundamentals to its extended applications. As one
of the groundbreaking developments almost ten years ago, FDTD method was
extended to be able to include lumped passive and active components. This new
approach opened the pathways for circuit and field cosimulation in a unified
environment. The combination of time-domain full-wave solution, through
FDTD method, and general lumped-system simulator would enable great
flexibility in complex system design and analysis.
The study of the FDTD method is a very active research area, and in
almost every issue of the related technical publications there are usually a few
papers about the FDTD method or its application in different disciplines. New
and exciting developments are reported as the technique progresses, and
therefore they further extend the applications of the method to new areas. There
are quite a few excellent reference books and reports on FDTD method and its
applications, and they provide very fine details about the theoretical basis and
applications of FDTD development. Therefore, the emphasis of this book is
steered toward the theory, algorithm development and application of
cosimulation for hybrid electromagnetic systems. New developments in hybrid
system cosimulation, some of which have not yet been published, are
incorporated throughout the chapters.

©2002 CRC Press LLC

©2002 CRC Press LLC

 1.2 Electromagnetic Systems

An electromagnetic system can be considered a system that involves active

or passive electromagnetic activities, regardless of the origin of the field.
Looking at the big picture, the universe is full of electromagnetic activities, from
the strongest pulsar signal, solar electromagnetic storm, the life-essential
magnetic field around the earth, to the hard-to-detect 3K microwave background
radiation in outer space. Around the earth atmosphere, electromagnetic waves
generated by radio and television broadcasting, satellite and wireless
communications, power lines, and even a remote controller for a toy car are
everywhere; no one can escape from them. Those electromagnetic waves are
traveling at the speed of light around and leaving the earth. If we count the first
man-made radio signal 100 years ago, the signal is about 1013 kilometers away
from the earth, or about 1/100 of the diameter of the galaxy in which our solar
system is residing. In a microscopic view, electrons and other charged particles
are spinning in their own orbits, colliding with different particles, sometimes
emitting light waves when their states change. In other words, the studies of
electromagnetic systems, from different viewpoints and different levels of
magnitude are fundamental to the understanding and progress of basic science,
the protection of the ecological system on earth, and the development of
technology we have been enjoying.
With the above somewhat too general definition of an electromagnetic
system, another interesting question would be could the human body, aside from
its complicated social-spiritual component, be treated as an electrical system?
The argument is strong. There are constant electrical activities inside the human
body, including those defined in terms as electrocardiogram (ECG),
electromyogram (EMG), electroencephalogram (EEG), electrooculogram, etc.,
not to mention all the other long medical terms for describing electrical
potentials recorded at various parts of the human body. The sources of these
electrophysiological activities are the cellular-level electrical potential changes,
and the normal ECG activity is synchronized by the pulsing sinus node inside
the right atrium of the heart. Every second or so, an electric pulse is generated
and propagated down the fibers concentrated around the heart ventricle muscle,
an efficient and accurate electrical system that is vital to human life. Electric
signals can be measured at almost all locations of the body, and that is how
those long medical words were coined. Even those considered high-level neuron
activities, like neural signal conduction, have their electrobiological basis.
Clinicians are relying on reading and explaining those sometimes very weak
signals to aid diagnosis. There are groups of researchers who are experts in
measurement, simulation, and analysis of these electrical activities, making up
their own approaches to probe the mechanism of the mysterious life phenomena.
Another important note is that human body tissues, as lossy dielectric media in
the eyes of physicists and electrical engineers, interact with incident
electromagnetic fields and this has led to a special study area, for example the
effects on the human body from power transmission lines and cellular phone

©2002 CRC Press LLC

©2002 CRC Press LLC

handsets. This has been one of the most active research areas for protective and
diagnostic purposes. Therefore, this seems somewhat off the topic at the
beginning, but it is not a totally isolated issue, and the original question is a
legalistic question and most people probably have to agree that the answer to the
question is positive.
Getting closer to what we are interested in, an electromagnetic system is a
functional electronic device that is designed to perform a certain specific task,
like signal transmitting and receiving, or the combination of those functions. The
vacuum tube and semiconductor transistor were invented in the early and middle
20th century and the integrated circuit (IC) started its crusade in the late 1960s.
Combined with technology such as fiber optics where laser signals are
transmitted, they revolutionized the electronics industry and everyday life of
human society. Pocket-sized devices can now perform functions that used to
require an apparatus hundreds of times larger and heavier. Further focusing on
the topic of this book, the design and analysis of such systems are the task of
electrical engineers, who portray and abstract an electronic device as a
schematic on paper, a portion of a doped semiconductor wafer, or a discrete
element on a circuit board.
Figure 1-1 gives an AM radio receiver that includes, typically, an antenna,
frequency mixing, tuning and filtering circuits, audio signal amplifier, and
speaker for audio output. This system is simple and most readers started to know
this circuit when they were working on their high-school science projects. A
high-frequency amplitude-modulated signal is detected and received by the
antenna, and then the circuit is tuned to mix one local oscillating frequency with
the receiving signal to select the desired signal channel. The demodulated low
frequency signal, audio signal in this case, is filtered and amplified for output.
What is more interesting about this diagram is that the receiving devices in
modern wireless communications, like in a CDMA cellular phone, works in a
similar fashion.

Antenna
Frequency Audio signal Output to
mixing amplification speaker

Local
oscillation
(tuning)

Figure 1-1. A block diagram for signal flow in an AM radio receiver.

©2002 CRC Press LLC

©2002 CRC Press LLC

 In a modern wireless communication system, a powerful digital signal
processing (DSP) core and memory block are integrated inside the apparatus and
provide powerful functionality and user-friendly interface. Similar to the AM
radio data flow, a high-frequency modulated radio frequency (RF) signal is fed
into a low-noise amplifier (LNA) before it is down-converted, through multiple
stages if necessary, by mixing with a local oscillating (LO) signal to get an
intermediate frequency (IF) signal. The output signal from the final down-
converting stage is an analog signal, and its frequency should be lower enough
for analog-to-digital (AD) conversion. Once in the digital world, DSP and other
programmed function blocks could process different types of data streams and
audio signals for information display and audio output. The transmitting diagram
is just the reversal of the receiving flow where a digital-to-analog (DA)
conversion is required, as both are shown in Figure 1-2. The number of stages
for frequency down converting depends on the frequency of the RF signal; it is
desirable to have fewer stages for circuit simplicity and performance. New
technology is converging in that direction, and signal direct conversion circuit
has been the trend in many wireless applications. As a mater of fact, if the future
AD/DA converter is fast enough, it is possible to sample the RF signal directly,
thereby totally eliminating the need for IF stages and associated LOs, and letting
the DSP core extract the information buried in the modulated signals.

Antenna LNA Frequency Buffer and A/D Digital signal

downconverting amplification converter processing
(DSP)

RF
IF1 IF2 IFn Analog Digital
······

LO1 LO2

Local Local
oscillation oscillation

Antenna LNA Frequency Buffer and D/A Digital signal

upconverting amplification converter processing
(DSP)

RF
IFn IFn-1 IF1 Analog Digital
······

LOn LOn-1

Local Local
oscillation oscillation

Figure 1-2. A block diagram for signal flow, receiving and transmitting, in a
modern wireless communication system.

©2002 CRC Press LLC

©2002 CRC Press LLC

 Any given electrical system is made of many components or subsystems.
The physical size of the electrical system varies dramatically, but the size of the
system is measured by its electrical size to the electrical engineers. Physical size
is compared with the corresponding wavelength of the frequency at which the
system is operated. If the physical size of a system is far less than the
wavelength, the system is called “electrically small”; similarly, when the
physical size is comparable or bigger than the wavelength, the system is
considered “electrically big.” This is an important concept in electrical system
analysis and this concept directly leads to the introduction of the concepts of
distributed and lumped electrical systems.
If anyone were able to open the plastic package of an integrated circuit chip
and look at it under a high-power microscope, then many the circuit components
and their connectivity would be seen. The topology of the circuit is more
complicated than is shown on the surface since all the chips today are using
multiple levels of metal for complex connectivity, ground and power grid. All
circuits start from their schematic designs, followed by verification of the design
with circuit performance simulation. Then the design is placed into a given
dimension with sometimes-artistic layout work before it is shipped to the silicon
foundry or circuit board manufacturer. Figure 1-3 shows a simplified flow chart
for a typical circuit design flow. After the initial concept design of the circuit,
most of the efforts of a design work are spent between the schematic input,
simulation, layout and its verification. Circuit simulation, either in the DC state,
time domain, or frequency domain, is the most important tool for verifying the
correctness of any design. Once the designed circuit is laid out in its physical
representation, the physical connection of the device is checked again with the
original design.

Function Schematic Performance

specification input simulation

Layout and
verification

Tapeout for
manufacture

Figure 1-3. A block diagram for integrated circuit design flow.

©2002 CRC Press LLC

©2002 CRC Press LLC

 Since the distributive nature of the circuit layout and the lumped nature of
the schematic reside in two different domains, the transition between those two
paradigms is realized by elaborate schemes for device modeling and parasitic
extraction. Electronic devices are primarily characterized by measurement and
simulation that lead to device models used in the schematic simulation. Device
models vary greatly depending on the operating frequency and the targeted
circuit application of the device. The most important factor, of course, is the
frequency at which the circuit is operating. At low frequency, some DC
measurements may well satisfy the need for the circuit design while at gigahertz
frequency range, a spiral inductor may have to call for full-wave field solution to
capture its complete equivalent model.
The density of modern integrated circuits is tremendous with multiple layers
of material and metal to connect components, provide power and ground
connections. Metal routings are much more complicated than a street map of
Manhattan and the parasitic effects of interconnects between components and
interactions between interconnects are yet another troubling concerns for circuit
designers. Various types of parasitic extraction techniques have been developed
and implemented to improve design efficiency and first-time success rate. The
extracted parasitic from devices and interconnects can be combined with the
original schematic for simulation with parasitic effects, mostly resistance and
capacitance. For higher-frequency circuits, effects of parasitic inductance due to
interconnects and devices have become increasingly important and some
commercial tools have emerged out of the development stage.
Figure 1-4 gives an example of modeling a p-type field-effect transistor
(FET) device, shown in Figure 1-4(a), with its simplified structure, an equivalent
circuit model used for digital application and another equivalent circuit for
small-signal analysis at high-frequency application. From the FET model in
Figure 1-4(b) and Figure 1-4(c), it is clear that the difference between the digital
model and the high-frequency one is very dramatic, and, in fact, as more
physical effects are included in the equivalent model, it is impossible to have the
whole model pictured on paper within a reasonable complexity. Instead, a long
list of parameters, some of them in hundreds, is given for the model. Therefore
each simple symbol, or component, on the schematic, actually contains a whole
set of device models, which is treated as a subcircuit, and it is the equivalent
circuit model that is used in the schematic-level simulation.
Gate
Source Drain
+
n n+

p-

substrate
Figure 1-4(a). A simplified structure view of a p-type field-effect transistor
device.

©2002 CRC Press LLC

©2002 CRC Press LLC

 Gate

Source Drain

Figure 1-4(b). A simple FET equivalent circuit model in digital circuit

simulation.

Gate

Source Drain

Bulk

Figure 1-4(c). A FET equivalent circuit model in high-frequency small-signal

simulation.

To some extent, the integrated circuit chip itself can be treated as a discrete
element, although it may contain hundreds of thousands integrated elements
inside. Inside a semiconductor chip, the behavior of each individual component
can be lumped or distributed depending on the operating frequency and other
factors. An IC chip on a circuit board can be treated as lumped if its
functionality is described by its external connections, while its internal details
are not of interest. There is no distinct boundary between distributed and lumped
systems; however, traditionally electronic circuits are separated as lumped and
distributed as if there were one. Engineers working on systems operating at

©2002 CRC Press LLC

©2002 CRC Press LLC

different frequency ranges have distinct ways of thinking, and they may consider
other electrical systems from totally different application areas. Traditional
microwave engineers consider the circuit from a field point of view, while a
digital or analog designer rarely knows or cares much about field theory. This
boundary is now starting to crumble from its base with the wireless
communication boom in recent years. Electronic circuits and systems operating
at RF frequency, which positions somewhere between the lumped and
distributed domains, have become high in demand. This led to a shortage of RF
engineers and therefore generated a great deal of interest in circuit design and
theory in the range of a few gigahertzes. In fact, the so-called digital RF circuit,
a digital circuit operating at RF range, has further fueled the transition from
conventional digital and analog circuit design to the more complex RF world.
Circuit theory and numerical analysis were matured for designing and
simulating such a circuit during the past decades. Huge amounts of resources
have been invested to perfect the art of circuit analysis and simulation. Various
techniques have been developed for designing sophisticated digital and analog
circuits. Computer technologies, benefited largely from the revolution of the
electronic industry, in turn stimulated and enhanced the evolution of electronic
design, therefore electronic design automation (EDA). Meanwhile, techniques
for solving field equations have been exclusively developed for high-frequency
applications, like those in radar and antenna system designs. Today, as the size
of microwave and semiconductor elements shrinks and their operating frequency
gets higher and higher, traditional circuit designers, who used to think of circuits
at lower frequency ranges, will find themselves surrounded by all the physical
effects not included in the lumped domain. This generates the need for a
simulation capability to combine the two different techniques for future system
design and analysis.
In the next few chapters, the fundamentals and relation between field theory
and circuit theory will be further explored in detail. In short, circuit theory
approximates field theory; it works for the majority of the low-frequency
circuits. The foundation of circuit theory, Kirchhoff’s current and voltage laws,
and many other well-known circuit equations can be derived from quasi-static
field theory, and they will be discussed in detail. The theory is presented with
focus on the circuit computer-aided analysis and design (CAD). In parallel, the
numerical solutions to full-wave field equations are addressed, with emphasis on
the time-domain approach, and its cosimulation with lumped systems is
introduced.

1.3 Hybrid Electromagnetic Systems

As defined earlier, a hybrid electromagnetic system is an electronic system

that contains both distributed subsystems and lumped subsystems. Examples of a
distributed system are microstrip lines and circuits built with them, waveguides,
and antennas; examples for lumped ones are resistors, capacitors, inductors,

©2002 CRC Press LLC

©2002 CRC Press LLC

diodes and transistors, and circuits made of those elements, such as filters and
power amplifiers. Distributed systems require a field solution in order to fully
characterize their electrical behaviors; a lumped system, on the other hand,
needs circuit theory to describe its performance. The fact that hybrid systems
exist poses the need of a generalized cosimulation method for analyzing such a
system.
Naturally, any device can be treated as a distributed device, but most of the
time this would prove to be too expensive for any reasonable circuit design. In
fact, many of the devices are modeled with the aid of a field solution, and then
the extracted device model parameters are used in the circuit simulation. The
best combination for hybrid analysis should be using circuit theory and field
theory together for each part of the corresponding subsystem. Usually the field
solution is more accurate but much more expensive and time-consuming
compared with lumped-circuit simulation. Therefore, the trade-off between
accuracy and computation cost in the circuit and field cosimulation determines
the division of the two analyses. It should be safe to conclude that an ideal
hybrid cosimulation model should be circuit based, with the ability of handling
distributed subcircuits with a full-wave solution. The exact threshold where a
field solution becomes an absolute necessity remains to be determined, but it is
on the horizon as the operating speed of the electronic system is ramping up
almost exponentially.
Historically, there used to be a distinct separation between distributed and
lumped electrical systems. Most of the time, electrical systems are treated as
lumped systems because the operating frequency of those systems is low enough
that lumped approximation causes little error. It used to be only those systems in
microwave range and above, mostly for high-power applications, required
analysis from the viewpoint of field theory. As will be discussed in later
chapters, this is no longer the case in many of the applications for modern high-
frequency communication.
A hybrid system is in general divided into two parts, the main system with
another alien system. In the case of hybrid electromagnetic system, they are the
distributed and the inserted alien subsystem, or vice versa, as illustrated in
Figure 1-5. A typical example of such a hybrid system is made of a multilayer
microstrip line system, a distributed component, and its impedance matching
circuit composed by discrete capacitors and inductors, some common lumped
elements. Depending on the emphasis of the study, the role of the main system
and alien system can be exchanged. For example, when a system analysis is
driven by the lumped circuit simulation, then the distributed subsystem can be
treated as the “inserted” alien part.
Each of the subsystems could include both linear and nonlinear elements.
Although there are some unsolved problems in nonlinear system analysis in
general, for the most part, there is not much difficulty in the analysis of a
nonlinear system numerically. Various algorithms have been developed for the
analyses of nonlinear circuits. As an example, the harmonic balance method is
widely used in high-frequency linear and nonlinear circuit analyses. Many of the

©2002 CRC Press LLC

©2002 CRC Press LLC

materials demonstrate nonlinear behaviors, especially at higher frequency
ranges, and this is another important consideration when solving the field
distribution within a structure including those materials.
In order to analyze a hybrid system, the most common and easiest approach
will be to approximate the distributed system with a lumped equivalent circuit,
and then the hybrid system is degraded into a complete lumped system.
Therefore the theory for lumped system can readily be applied to the analyses of
the hybrid systems. The transmission line matrix (TLM) method, partial
equivalent electric circuit (PEEC) method and other equivalent circuit model
approaches are some examples along this line. They have different numerical
properties, and, depending on the methodology itself, they are suitable for
various applications.
The numerical advantages of using the lumped approach are enormous.
First, the partial differential equations become ordinary differential equations in
most cases, and the distances between different parts of the system disappear,
sometimes replaced with a certain time delay. During the past decades, the art of
circuit simulation has been perfected, or closing to that status, with the
tremendous efforts of researchers and engineers. Different mathematical
techniques were developed for better efficiency and more capabilities. In fact,
there are separated tool sets for digital, analog and mixed-signal simulations.
Some digital and mixed-signal design tools have evolved into specialized
programming languages for specific design. SPICE and SPICE-like circuit
simulators are now industry standards in analog simulations and they are well
tested and readily available in variations.

Alien
System
Distributed System Distributed System

Figure 1-5. An example of hybrid electrical system, a system that includes both
a distributed system and an alien system (for example, a lumped circuit), is
illustrated.

©2002 CRC Press LLC

©2002 CRC Press LLC

 As computer technology and numerical techniques grew rapidly, field
solution techniques have been undergoing the same immense transition. Method
of moment, finite-element method, and others become standard tools for
electrical engineers, and even engineers in other disciplines. FDTD method,
being one of the most prominent time-domain full-wave field solvers, has grown
to its maturity during the past decade, and it is becoming more and more popular
in many routine designs.
The necessity of time-domain simulation relies on the fact that the
electronic circuit is operating at higher and higher frequencies and ever
increasing nonlinear effects have to be included in the system analysis.
Commercially available advanced semiconductor processes are supporting
devices working above frequencies of 100GHz for high-speed data transmission
and the trend is going still higher. In fact, researchers are working on circuit
with operating frequency at terahertz and above. As signal bit rate increases, it is
harder to analyze the circuit within a limited frequency bandwidth, regardless of
analog or digital signal.
Let us look at the frequency spectrum of a time-domain pulse series p(t), as
seen in Figure 1-6, through Fourier analysis. It can be shown that the time-
domain signal has a frequency spectrum of a sinc function, shown in Figure 1-7
(a) and (b) for magnitude and phase, respectively. The magnitude spectrum has a
main lobe within which most of the signal energy is concentrated, and the
spurious ringing in the spectrum is a clear indication of the dispersion nature of
the signal. The frequency f0 is sometimes defined as the bandwidth of the pulse
series, and it is inversely proportional to the pulse width (T). As the pulse gets
narrower in high-speed circuits, the spectrum bandwidth of the signal becomes
wider and wider. For example, when the pulse width T is 1ms, frequency f0
equals 1KHz and f0 becomes 1GHz when T is 1ns. As the pulse width is reduced
to one hundredth of a nano second, the bandwidth is up to 100GHz, a broadband
signal that would be really difficult to handle by any frequency solution
technique.
Although the above discussion is only to get a qualitative idea about the
circuit bit stream and required bandwidth, it clearly shows the demand for a
broadband approach for characterizing, modeling, and analyzing electronic
systems operating at high frequency. It may not be practical to utilize some of
the numerical methods described in this book for everyday large designs, yet it
would certainly become more valuable when other techniques fail. FDTD
method and its extension for hybrid systems would become more and more
applicable when the limitation of computing resources is decreasing
exponentially, sometimes in leapfrog fashion, and when broadband high-speed
circuit designs become routine.

©2002 CRC Press LLC

©2002 CRC Press LLC

 p(t)

-T/2 T/2 time

T0

Figure 1-6. A time-domain pulse series signal.

Magnitude
spectrum
AT/T0

-4f0 -3f0 -2f0 -f0 f0 2f0 3f0 4f0

frequency

Figure 1-7(a). Magnitude spectrum of the time-domain signal in Figure 1-6.

©2002 CRC Press LLC

©2002 CRC Press LLC

 Phase
spectrum

frequency

−π

Figure 1-7(b). Phase spectrum of the time-domain signal in Figure 1-6.

One of the main purposes of this book is to describe a general approach for
including lumped-in-nature alien subsystems into the time-domain full-wave
solution. The inclusion of a current, contributed from the inserted alien
subsystem in Maxwell’s equation, provides the mechanism for interaction
between the electromagnetic field and the lumped subsystem. Examples like
modeling a fluid-like electron beam and frequency-dependent gain media in a
laser cavity clearly demonstrate the broad range of the application. The circuit-
field model, a versatile interface between FDTD solver and a standard SPICE-
like analog circuit simulator, is presented for lumped circuit cosimulation.
Approaches to include the multiport model, described by its scattering
parameters and behavioral model in time-domain simulation, further extend the
capability of hybrid system cosimulation. Many of the materials, algorithms and
results presented in this book are being published for the first time and are state-
of-art in time-domain circuit and field cosimulation. They present a bright future
for applications in high-speed electronic system design and analysis. This book,
weighted heavily to field and circuit cosimulation, is constructed on that belief
and it is the author’s hope that its readers will find this book beneficial and
thought provoking.

1.4 Organization of the Book

During the past decades, research and application of the FDTD method have
been evolved into specialized areas and they are too wide to be covered in a few
hundred pages of context. Publications on different topics are numerous and are
available for reference. This book focuses on the simulation of electromagnetic
systems in the time domain, with special emphasis on solving hybrid systems

©2002 CRC Press LLC

©2002 CRC Press LLC

that have other subsystems with linear or nonlinear characteristics. Specifically,
the hybrid cosimulation model is described and applied to the simulation and
characterization of systems that include both distributed and lumped elements.
Throughout the first few chapters of this book, the basis of solving full-wave
field equations and lumped circuit is introduced and extended fully. This lays the
foundation of simulating hybrid systems using field solution, circuit equivalent
model or both, therefore hybrid system cosimulation. Different methods of
solving field equations are briefed, with most of the parts devoted to the finite-
different technique in the time-domain method. As background to the circuit and
field theory and simulation, a brief review of the electromagnetic theory, the
fundamentals of finite-difference technique and SPICE circuit simulator are
introduced before the introduction of cosimulation methods.
In Chapter 2, the theorems that make up the electromagnetic field theory are
introduced in a brief and historical view. Most of the basic field equations are
given, without much mathematical details, for easy reference. Fundamental
concepts in electromagnetic theory are discussed and illustrated to pave the easy
transition between field theory and circuit theory. Various examples for solving
field equations are given in different sections of the chapter. At the end of the
chapter, a couple of examples for the field solution are presented and the
solution process illustrated the application of boundary conditions and many
basic approaches in field theory. One of the examples is for cavity mode
analysis in frequency domain using method of separation variable, a bit contrary
to the time-domain theme of this book, but the field distribution of the two-
dimensional cavity structure is referred to in later chapters.
Since circuit theory is approximated from field theory, the transition from
field to circuit is derived and highlighted with some commonly seen electric
components at the beginning of Chapter 3, after the quasi-static field is
introduced. This chapter emphasizes and illustrates the link between field and
circuit theory, for example, lumped variables such as voltage and current are
related to field variables, electric field and magnetic field and the limitation of
those relations are discussed for preparation of future application. Other circuit
concepts such as Kirchhoff’s voltage and current laws are easily derived from
the field theory, as the current and voltage relation for some of the simple
lumped components are discussed in detail as well. Applications of those
equations are detailed in Chapter 6. The importance of system transient
behaviors is discussed and illustrated with a transient solution of a circuit, with
the help of Laplace transformation.
The transmission line model, which includes signal propagation delay and
characteristic impedance, is an important building block for circuits, and it
actually has lumped and some distributed properties. The transmission line
model is extracted from a one-dimensional wave equation, and it represents a
general group of structures that guide the propagation of signals. Many
important concepts, impedance and signal reflection, are introduced along with
the transmission line general solution. As an example of solving a transmission
line problem in the time domain, a transmission line system, with a lumped

©2002 CRC Press LLC

©2002 CRC Press LLC

voltage source excitation and lumped resistor load, is solved analytically, again
with the help of Laplace transformation. In the last section of the chapter,
scattering (S) parameters for n-port network system are introduced, based on the
results from the transmission line analysis in earlier sections. Other types of
network representations are briefly discussed as well.
In Chapter 4, finite difference method for approximating first-order and
second-order derivatives is first derived from Taylor’s series. Numerical
solution of ordinary and partial differential equations with the finite-difference
method is then discussed, along with an example of solving Poisson’s equation.
Solution of a linear matrix system, a product from approximating a linear system
with the finite-difference method, is discussed briefly. Stability condition, error
control, convergence and other important aspects of numerical solution are
briefly mentioned. An example of finding a numerical scheme with better
stability condition for a lumped circuit demonstrates the importance of selecting
the right algorithm in search of better accuracy and stability.
With the discussions of field theory and finite-difference method in
previous chapters, solution to Maxwell’s equations using FDTD method is
obtained by finite differencing of the partial derivatives in both time and spatial
domains. Chapter 5 introduces the Yee cell grid and includes the FDTD
expressions for uniform and nonuniform formulations in both homogenous and
inhomogeneous media. In addition to the full three-dimensional formulation,
two-dimensional FDTD equations are derived as well. Some of the important
aspects of the FDTD method, such as different types of absorbing boundary
condition, numerical stability and accuracy, unconditionally stable algorithm
and numerical dispersion, are discussed in some detail. General FDTD
formulation for inhomogeneous media in nonuniform grid is derived from the
integral form of Maxwell’s equations. The unconditionally stable FDTD
algorithm, a recent development, is described in the chapter, and it is expected to
play an important role in hybrid circuit simulation since it releases FDTD from
its rather restrictive timestep limitation. At the end of the chapter, a couple of
application examples of FDTD method are presented.
Chapter 6 gives readers a simplified introduction for a SPICE-like analog
circuit simulator; again, the content focuses on the time-domain transient
simulation and linear and nonlinear device modeling. General SPICE-like
analog circuit simulator can be built on the modified nodal analysis (MNA)
formulation, with Kirchhoff’s current and voltage laws as the basis of circuit
nodal analysis. Examples of the circuit solution from MNA formulation and
analog simulation are described. Modeling issues of some linear and nonlinear
devices, diode, FET and bipolar junction transistors are discussed with details.
In addition to the foundations of circuit simulator, such as timestep and accuracy
control, some related issues, such as language support and device model
interface, are discussed.
As the main theme of the book, in Chapter 7, general equations from
Maxwell’s equations for including various types of current-contributing
subsystems are discussed. This is the foundation for integrating an alien system

©2002 CRC Press LLC

©2002 CRC Press LLC

into the distributed system, as described by Maxwell’s equations. The additional
current stems from the “inserted” system residing inside or interacting with the
distributed circuit. The extended FDTD method has enhanced capability of
handling additional sub-modules, and these alien systems could be a standard
lumped circuit, a multiport scattering parameter block, a lumped behavioral
model or a system generating convection current due to an electron beam
flowing in vacuum. The general FDTD solution methodology for including alien
systems inside a distributed EM system is the basis of hybrid system
cosimulation. As examples of applying the theory described in the chapter,
individual lumped voltage source, individual RLC component and a quasi two-
dimensional traveling electron beam are included in FDTD scheme.
Based on the general relation described in previous chapters, Chapter 8
gives details about the interface between FDTD and a lumped circuit simulator
and general n-port system, described either by its frequency-domain scattering
(S) parameters or port behavioral model. A circuit-field cosimulation model for
distributed and lumped circuits, along with its implicit model implementation,
are presented. The circuit-field model acts like an interface between the
distributed and lumped circuits, with much increased numerical stability;
therefore, it bridges the data exchange and simulation control between the two
distinct electrical systems. Stability and accuracy of the extended FDTD
method, with the addition of the current contributed from the lumped system, is
illustrated by comparing the simulation solution with simulation results from
other known results.
A multiport model is described for integrating a general n-port network
system into FDTD simulation, based on its mathematical behaviors at each of
the input and output ports. The formulation is derived from the general solution
of transmission line theory, described in Chapter 3, and its convolutional
equation is suitable for both systems with frequency-domain descriptions, like S
parameters and those with time-domain descriptions.
The formulations presented in this chapter prepare for the complete
simulation of a hybrid system that includes circuit, lumped in nature, residing
inside a complex structure that requires three-dimensional full-wave solution,
such as the packaging analysis discussed in the next chapter. At the end of the
chapter a couple of examples are given to show the application of the
cosimulation models introduced in the chapter.
Chapter 9 demonstrates the applications of the circuit-field and multiport
cosimulation models in some high-frequency circuits, with attention to
broadband system characterization. Some commonly seen circuit structures are
simulated using the appropriate cosimulation model described in the last chapter.
The example circuits are relatively simple but representative, more detailed
studies are required to further complete the numerical cosimulation schemes.
Both two-dimensional and three-dimensional structures for the distributed
circuit are used for the calculations.
Packaging analysis is one of the most important FDTD applications since it
has many advantages over its frequency-domain field counterparts. Broadband

©2002 CRC Press LLC

©2002 CRC Press LLC

FDTD characterization is used for a three-layer flip-chip packaging structure
and the simulation results are compared closely with the measured data. As the
first attempt to simulate a whole chip structure that includes both the packaging
configuration and its on-chip circuit, the multiport model is applied to a flip-chip
structure where an amplifier, described by its S parameters, is sitting on top of
the structure.
One interesting topic in modern optical communication systems is the
design and analysis of a light source and its integration with the rest of the
electrical system. Communication systems operating at light-speed require better
simulation capabilities for analyzing and designing such a system, either at
system or subsystem level. Chapter 10 describes a generalized formulation for
extending the FDTD analysis technique in nonlinear optical systems. Here the
alien subsystem is a phenomenological nonlinear active medium inside a laser
microcavity. With the described current modeling approach, a vertical cavity
surface emitting laser (VCSEL) light source is simulated to study the detailed
field distribution inside and outside the cavity, especially for mode selection, a
major consideration in VCSEL design. The nonlinear gain medium inside the
VCSEL cell has a frequency-dependent gain distribution and gain saturation.
The inclusion of such nonlinear material in FDTD grid provides a realistic
simulation of the mode build-up process. At the conclusion of the chapter, a
system consisted of both a VCSEL light source and a simple transmission line
loaded with a lumped resistor is simulated to illustrate the potential of the
formulation in analyzing a complete optoelectronic circuit. The initial simulation
for combining an optical device with a distributed and lumped circuit is still
primitive, but it is definitely a good starting point for future research and
development.
For easy reference, many vector operators and identities are summarized in
Appendix I at the end of the book. Throughout this book, Laplace transform,
both forward and inverse, is used extensively for transient solutions or
discussion about frequency-domain behaviors of a system. Laplace transform is
briefly introduced, along with some of its properties and transform pairs, in
Appendix II. To give more details about the transformation between the time
domain and the Laplace domain, some commonly used techniques are illustrated
by examples at the end of the appendix.
At the end of book, some references on various application areas are listed
for further reading and study. Another useful on-line resource is a website at
www.fdtd.org, which provides a quite complete archive on FDTD applications.
To provide flexibility for selecting the appropriate references, some of the
chapters of this book can be skipped by those who are familiar with the
materials covered in them.
Hybrid electromagnetic system cosimulation is becoming more popular in
various application areas, and for the most part, it is still just a beginning of a
new adventure for next-generation simulation tools. This book attempts to
explore the potential applications of the extended field and circuit cosimulation
method in circuit design and analysis. The road that leads to the eventual

©2002 CRC Press LLC

©2002 CRC Press LLC

complete system simulation could be a long-time endeavor; the approaches
presented in this book may not be the best ones, but it would be the author’s
relief if readers found this book stimulating and helpful in searching the ultimate
goal of designing high-speed electronic circuitry. The concept of complete
system cosimulation will shine brighter as more and more design engineers
realize the need for it in the future and more simulation capabilities become
available.
This book was completed in a relatively short period of time, and it is the
author’s intention to include the latest research results in this book. It is
inevitable that something will have been overlooked and that there will be some
mistakes with the materials presented here. Revisiting the topics discussed in
some of the chapters in the future should find more mature and efficient
approaches for practical applications and it would make all the efforts
worthwhile if this book could serve as a stepstone toward that goal.

©2002 CRC Press LLC

©2002 CRC Press LLC

 CHAPTER 2

ELECTROMAGNETIC FIELD THEORY

2.1 Introduction

The discovery and understanding of electromagnetic phenomena in general

can be traced back thousands of years ago when ancient philosophers were
interested in the physical world around them. When amber was rubbed by fur, it
could attract small objects; in today’s theory, it had been electrically charged.
Lightning in the sky of course was the subject of heavenly worship, but complex
electromagnetic processes were involved in the natural phenomenon, which is
still being studied by scientists armed with the latest equipment. Those natural
phenomena led to the recognition of the existence of electric power, yet a truly
scientific explanation and quantization was completed a little more than a
century ago. Generation and utilization of such power for human life have fully
blossomed only in the last few decades. With the proliferation of modern
electronic devices, high-power electricity became a daily necessity, and a
pocket-sized gameboy, to a seven-year-old, is something that naturally exists,
like stars in the night sky.
It took generations of scientists in hundreds of years to complete the theory
of electromagnetics. Although not a focus of this book, a brief introduction of
electromagnetics would be helpful to start our journey, which will eventually
bring us to the state of art of computational electromagnetics. Predicting the
future is always dangerous, but there is an old Chinese saying “knowing the past
will help understanding the new.” Keeping that in mind, some historical review
is probably not too far out of the scope of this chapter. In fact, when those names
of the pioneers and their contributions are mentioned in this chapter, the theory
they developed, in the forms of theorem that many of them are familiar to high-
school seniors, will be introduced. Some of the equations are for static cases, but
they are used in many of the practical applications, for example the Poisson’s
equations for some of the two-dimensional or three-dimensional RC parasitic
extractions in integrated circuit design. At the end of the next section, the final
governing equations, which are well known today as Maxwell’s equations, are
introduced and discussed.
Besides the philosophical discussion in ancient Greece, theories about
electromagnetics started with experiments. At the early stage of the
electromagnetic theory development, experiment, observation and ingenuity
inspired most of the theories. Analytical ability was absolutely necessary for any
new deduction and breakthrough. Up until the middle of the 20th century when
the first computer was built, engineers relied on slide rulers, manual tables and
charts for their designs and calculations. Most of the commonly used equations
and formulae are tabulated for easy references. The complete time-domain

©2002 CRC Press LLC

©2002 CRC Press LLC

partial differential forms of Maxwell’s equations, the governing laws of
electromagnetic behaviors, were often time introduced, tabulated in textbooks,
but were almost never really solved, either analytically or numerically, for any
practical system. The most common solutions were computed in the frequency
domain, avoiding the mathematical difficulties from the time-domain partial
derivatives. From Maxwell’s equations, wave equations can be derived, and,
under certain circumstances, they are solved for field mode analyses within
many of the waveguide configurations. Even further, Poisson’s equation and
Laplace equation are two equations that can be considered as special cases of
electrodynamics. More electromagnetic problems with simpler structures can be
discussed and accurately computed.
Circuit theory is the quasi-static approximation of the field theory, namely
Maxwell’s equations and other equations. Circuit theory was developed
independently from electromagnetic theory since most of the earlier electrical
signal and systems studied were in DC or at the low-frequency range.
Understanding the relation between circuit and field theory would provide
deeper understanding of the electromagnetic system. Several equivalent circuit
models dedicated to link the circuit and field theory have been proposed and
designed for system analysis. These methods are normally for both time-domain
and frequency-domain analyses, but each has its own limitations as well.
The theory described in this chapter is a mixed bag of equations for both
general dynamic and static or quasi-static fields. A clear definition of a quasi-
static field is given in the next chapter when the circuit theory is deduced from
field theory. By no means is this chapter intended as an electromagnetic
textbook; there are many good books available for that purpose. With the limited
extent of the chapter, equations are explained for their applicable range, but
readers should be aware of limitations when applying each equation under
different circumstances. Most of the derivation and detailed discussion of those
formulations is omitted; interested readers can find many better references.
For reference purposes, many of the equations are given in both integral and
differential forms. Equations in either integral or differential forms are
equivalent in nature, but, under different circumstances, sometimes one is more
convenient to work with than the other. In later chapters of this book, equations
in both forms are used for derivations and discussions of their numerical
solution. Some of the vector operators and identities can be found in Appendix I
when necessary. Now let us start with the field equations along with their
historical calendar.

2.2 Electromagnetic Theory

2.2.1 Coulomb’s law

In the late 18th century, Coulomb’s law, also known as the law of force for
electric charge, was derived based on the results of Coulomb’s brilliant

©2002 CRC Press LLC

©2002 CRC Press LLC

measurement of force between two charged objects. A charge generates an
electric field in the surrounding space, and this field interacts with other charges
some distance away. The force, in the direction of spatial vector r, is
proportional to the product of the charges of both objects Qq and inversely
proportional to the square of the distance R. In mathematical form,

Qq
F= r (2.1)
4πε 0 R 2

where Q and q, with unit of coul, are two point charges separated by distance R,
in the direction of unit vector r. ε0 is the permittivity of free space

ε 0 = 8.854 × 10 −12 farad/m

in the MSK unit system.

Coulomb’s law is the foundation for understanding the basics of electrical
interaction. When a unit test charge is in the force field generated by a charge Q,
the electric field intensity E can be defined as

F Q
E= = r (2.2)
q 4πε 0 R 2

Electric field intensity E, with unit of volt/m, is a measurable vector field; it

is directly related to other field variables, such as scalar electric potential Φ.

E = −∇Φ (2.3)

and

dQ
Φ=∫ (2.4)
4πε 0 R

One important observation about Coulomb’s law is that by taking the curl
of (2.3) and using vector relation (see Appendix I), we can get

∇ × E = ∇ × (−∇Φ ) = 0 (2.5)

This equation states an important property about the quasi-static electrical

field, i.e. static electrical field is not circulative, contrary to the time-varying
field as shown in Maxwell’s equation seen later.

©2002 CRC Press LLC

©2002 CRC Press LLC

 Inside a static or quasi-static electric field E, when a unit charge is moved
from point A to point B, the amount of work associated with the movement is
defined as electric field potential between point A and point B.

B A
VAB = VB − VA = − ∫ E ⋅ dl = ∫ E ⋅ dl (2.6)
A B

The scalar electrical potential Φ in (2.3) and (2.4) is identical to the

electrical potential defined in (2.6). Equation (2.6) describes the relation
between electrical potential and electric field in static or quasi-static situation.
The definition of quasi-static approximation is introduced and discussed in detail
in the next chapter.
Figure 2-1 shows two points (A and B) inside an electric field generated by
a charged object, the potential difference VAB between the two points is
calculated by (2.6). The electric field produced by the charged object, rather
than a point source as in (2.2), is the general case in Gauss’s law discussed in the
next section. In fact, it can be proved that the path integral for VAB is
independent of the paths from A to B for the static case. Equation (2.6) will be
discussed further in later chapters and is often used in hybrid electrical circuit
simulations, where the electric field and voltage are related to link field and
circuit variables.

Charged
object
q VB

VA

Figure 2-1. Illustration of the electric field generated by a charged object (q) and
the electrical potential from point A to point B through different paths.

©2002 CRC Press LLC

©2002 CRC Press LLC

2.2.2 Gauss’s law

Gauss’s law relates electric flux and electric charge, the source of the flux,
and it was first discovered by Gauss through his experiment.

Q
∫ E ⋅ dS = ε
S 0
(2.7)

where Q is the total charge inside surface S. When there is a distribution of

charge density inside a volume enclosed by S, the total charge can be calculated
by volume integration. In differential form, for any point in space the flux is
measured by the divergence of the flux-density vector; therefore, (2.7) can be
written as

ρ
∇⋅ E = (2.8)
ε0

where ρ, of unit coul/m3, is the spatial distribution of the charge density.

Gauss’s law can be considered an extension of Coulomb’s law in a spatial
charge distribution over a volume. When electric field from a point charge is
integrated over a sphere surface, equation (2.7) reduces to (2.2) naturally. More
importantly, (2.7) states the fact that an electric field at one location can be
considered the supposition of the contribution from all the point sources, as
illustrated in Figure 2-2, and the principle of supposition is valid for all linear
systems.

z E
P(x,y,z)

y
x

e1

e2
e3 ei

Figure 2-2. Illustration of the supposition of field, where the total electric field E
is the sum of four field vectors generated by each charge in the system.

©2002 CRC Press LLC

©2002 CRC Press LLC

 Another note for Gauss’s law is that it relates the electric field and voltage
to its source, charge and charge distribution in any given volume. This relation
is critical for extracting capacitance of an electrical system, as will be seen in
later chapters.
Poisson’s equation (2.9) relates the static or quasi-static electrical potential
to the charge density at any point in space; it can be derived easily by combining
equations (2.3) and (2.8).

ρ
∇2Φ = − (2.9)
ε0

Poisson’s equation, (2.9), is valid for cases of static and quasi-static field
distribution, and it also holds for time-varying situations at each instance of
time. However, its simple form does not imply an easy solution for arbitrary
geometry. Only a limited number of relatively simple geometric structures and
boundary conditions can be solved analytically. For general problems, especially
in three-dimensional situations, a numerical scheme has to be employed to get
accurate electrical potential distribution. As an example in Chapter 4, a two-
dimensional region with two point charges as source is solved numerically, and
the spatial field distribution is plotted.
In a charge-free region, Poisson’s equation reduces to Laplace’s equation,

∇2Φ = 0 (2.10)

Like most of the differential equations, both Poisson’s equation and

Laplace’s equation are solved, in either two-dimensional or three-dimensional
space, with proper boundary conditions to ensure unique solutions. There are
many types of boundary conditions and their details are discussed later in this
chapter and in Chapter 4, where numerical solutions to ordinary and partial
differential equations are discussed. Specific boundary conditions for electric
and magnetic fields are listed in a later section.

2.2.3 Faraday’s law

In explaining his experiment observations, Faraday discovered that the

change of magnetic flux generated measurable electric current or voltage, thus
Faraday’s law. In modern mathematical expression, the physical observation can
be described by the following equation:

d
∫ E ⋅ dl = − µ
c
0
dt ∫S
H ⋅ dS (2.11)

©2002 CRC Press LLC

©2002 CRC Press LLC

where µ0 is defined as the permeability of the free space

µ 0 = 4π × 10 −7 henry/m

in the MSK unit system.

The equation in (2.11) states that in a closed surface, the integral of
electrical field (corresponding to voltage in a loop) is caused by the temporal
changing of the magnetic field. When integral surface S becomes smaller and
eventually shrinks to a point, Faraday’s law can be written in its differential
form as

∂H
∇ × E = −µ0 (2.12)
∂t

Equation (2.11) or (2.12) shows that the time-varying magnetic field,

represented by magnetic field intensity H of unit amp/m, causes the generation
of electric force, therefore the generation of current. When the magnetic field is
zero or remains constant, the right side of (2.12) becomes zero, indicating that
the electric field becomes static. In that case, equation (2.12) reduces to (2.5),
which is the expression for Coulomb’s law.
Equation (2.11) or (2.12) is the basis of modern electric generators, and,
more importantly, Faraday’s law was the fundamental contribution to the
development of Maxwell’s equations more than a century ago.

2.2.4 Ampere’s law

In parallel to Faraday’s discovery that electromagnetic force (EMF) could

be generated by a moving (changing) magnetic field, Ampere studied by
experiment the interaction between current-carrying wires, and he was the
pioneer of studying the origin of magnetic fields. Therefore, named in his honor,
Ampere’s law of force gives the force between two circuit-carrying loops:
v
µ 0 I1 I 2 dl2 × (dl1 × r )
4π C∫1 C∫2
F= (2.13)
R2

where current I1 and I2 have loops C1 and C2, respectively, and vector r is the
unit vector in R direction, as seen in Figure 2-3.
Ampere’s circuital law states that the magnetic field is related to a steady
current in a surface S closed by loop C by the following relation:

©2002 CRC Press LLC

©2002 CRC Press LLC

 I = ∫ H ⋅ dl = ∫ J ⋅ dS (2.14a)
c S

In its differential form, Ampere’s circuital law is written as

∇× H = J (2.14b)

Current density J, of unit amp/m2, is flowing in surface S, enclosed by

contour C. Current I, of unit amp, is the total steady current, which is the
integral of current density J that leaves the surface.
Equation (2.13) states the fact that a magnetic field is related to current; i.e.,
current generates a magnetic field around it. Actually, Ampere’s law can be
derived from the Biot-Savart law, which is stated as

1 J
H= ∫
4π V
∇ × ( ) dV '
R
(2.15)

Ampere’s law is very important in the sense that it gives the relationship
between the magnetic field and current, and this relation will be further
discussed in later chapters. As will be seen shortly, Ampere’s law forms the
basis for the second half of Maxwell’s equations by adding an additional
displacement current term.

S2

S1 I2

F21
I1 r dl2
dl1 R
F12
C2

C1

Figure 2-3. Interactions of two current loops, I1 and I 2, are labeled as F 12 and
F21, and they are opposite in direction. Steady current can also be calculated by
a loop integral as in (2.14a).

©2002 CRC Press LLC

©2002 CRC Press LLC

 When there are different forms of current that are contributing to the surface
integration of (2.13), current density J is considered the sum of all the currents
flowing through the surface. For example, when a material has a conductivity of
σ, the conduction current density, due to the electric field E, is calculated by

J = σE (2.16)

Conductivity varies greatly for different materials as some are listed in

Table 2-1. Metal has high conductivity, and a good insulator has very low
conductivity. Semiconductors lie between conductors and insulators and have
some special physical features that form the foundation of modern electronic
industry, i.e. semiconductor integrated circuit. A metal is considered an “ideal
metal” when its conductivity is assigned to be infinite in some of the theoretical
and numerical calculations.
When a charged particle, with charge density ρ, is moving with a velocity v,
the current density caused by the movement is

J = ρv (2.17)

This equation states that the current density is depending on the charge
density and the velocity of the charge movement. The direction of the current
density is either the same as the movement direction for positive charges (or
holes in a semiconductor) or the opposite direction for negative charges
(electrons), as illustrated in Figure 2-4.

ve J vh

Figure 2-4. Illustration of current flow due to the charged particles moving
under an electric field. Negative and positive charges are moving in the opposite
directions.

©2002 CRC Press LLC

©2002 CRC Press LLC

 Equation (2.17) is a general definition for current density of charge
movement, and it can be applied to a current calculation when charged particles
are moving in vacuums or inside a semiconductor block. In semiconductor
material, for example a lightly-doped p-type silicon, there are electrons and
holes co-existing, under a given electric field, current due to the movement of
electrons and holes can be expressed as

J = ρ eve + ρ hvh = (−eµe + eµ h ) E (2.18)

where e is the electron charge, and µe and µh are the mobility of the electron and
hole, respectively. As can be seen from the equation in (2.18), material with
higher mobility would generate bigger current density than the one with lower
mobility under the same electric field. When they collide, particles with positive
and negative charges can combine with each other and particle collision can
change the direction of their movement, therefore causing the changes in total
current.
Recall from equation (2.2) that the force on a charged particle is
proportional to the electric field. When a magnetic field is present and the
particle is moving with velocity of v, the magnetic field has force on the particle
as well.

F = q( E + v × B ) (2.19)

As seen from the above equation, the force from the magnetic field is
perpendicular to the moving direction of the particle. Equation (2.19) is called
the Lorentz force equation in the literature.

2.2.5 Continuity equation

The continuity equation is another fundamental law that must be satisfied in

any electromagnetic system. The equation describes the fact that the net change
of electrical charge inside a given volume is equal to the current flowing in or
out of the surface enclosing the volume. This is another way of stating the law of
charge conservation, meaning no charge can be generated or lost:

d
∫ J ⋅ dS = − dt ∫ ρdV
S V
(2.20)

where ρ is the spatial distribution of the charge density inside volume V.

In differential form, the continuity equation can be written as

©2002 CRC Press LLC

©2002 CRC Press LLC

 ∂ρ
∇⋅J = − (2.21)
∂t

In (2.20) or (2.21), spatial diversity of current density is related to the

changing rate of charge density in one equation. In other words, the equations
show that movement or change of charge is the source of current generation.
When the system is static or quasi-static, meaning there is no change of charge
in time, then (2.20) and (2.21) become

∫ J ⋅ dS = 0
S
(2.22a)

and

∇⋅J = 0 (2.22b)

Equation (2.22) is the current continuity equation in a static situation. As

will be seen in later sections, this corresponds to the current relation at each
node, with no charge storage, in circuit theory.

2.2.6 Magnetic vector potential

Mathematically, magnetic field vectors are defined to aid the solution of the
field distribution. Similar to the definition of electric potential Φ in (2.4),
magnetic vector potential A can be defined as

µ J(r' )
4π V∫ R
A= dV' (2.23)

Therefore, the magnetic field is related to A by the following equation:

1
H= ∇× A (2.24)
µ

Based on (2.24), the following equation can easily be derived (referring to

the vector identities in Appendix I), concerning the divergence of the magnetic
field.

∫ H ⋅ dS = 0
S
(2.25)

©2002 CRC Press LLC

©2002 CRC Press LLC

or in differential form

∇⋅ H = 0 (2.26)

Physically, (2.25) and (2.26) show no “magnetic charge” exists.

2.2.7 Maxwell’s equations

Based on generations of efforts in both theoretical and experimental

electromagnetic fields, most of which have been summarized by the equations in
this chapter, Maxwell synopsized all the available knowledge at that time and
with his genuine abstraction ability, he completed the theory that governs the
macroscopic behaviors of electromagnetic fields. Maxwell’s equations cover
electromagnetic fields from zero frequency (DC) to the visible and invisible
optical frequency range.
Maxwell’s equations are basically combinations of Faraday’s law and
Ampere’s law with the addition of a displacement current, a current generated
by the temporally changing electrical field. In integral form Maxwell’s
equations are written as

∂H
∫ E ⋅ dl = −µ ∫
c
0
S
∂t
⋅ dS (2.27a)

∂E
∫ H ⋅ dl = ∫ (J + ε
c S
0
∂t
) ⋅ dS (2.27b)

or in differential form

∂H
∇ × E = −µ0 (2.28a)
∂t

∂E
∇ × H = J + ε0 (2.28b)
∂t

where current J includes conduction current, given in (2.16), and contributions

from other possible current sources.
Using the vector identities, as given in Appendix I, differential equations in
(2.28) can be expressed in terms of each field component:

©2002 CRC Press LLC

©2002 CRC Press LLC

 ∂Ez ∂E y ∂H x
− = −µ (2.29a)
∂y ∂z ∂t

∂Ex ∂Ez ∂H y
− = −µ (2.29b)
∂z ∂x ∂t

∂E y ∂Ex ∂H z
− = −µ (2.29c)
∂x ∂y ∂t

and

∂H z ∂H y ∂E
− = Jx + ε x (2.30a)
∂y ∂z ∂t

∂H x ∂H z ∂E
− = Jy +ε y (2.30b)
∂z ∂x ∂t

∂H y ∂H x ∂E
− = Jz + ε z (2.30c)
∂x ∂y ∂t

The other two equations, (2.7) and (2.25) for the integral form and (2.8)
and (2.26) for the differential form, respectively, are treated as part of the
Maxwell equations in many textbooks. In fact, they are special cases of the
above equations and can be derived mathematically from the two fundamental
equations with some vector operations. For example, (2.5) is the static case of
(2.28a) and (2.14) is the same as (2.28b) without the displacement current.
Poisson’s equation can be derived from (2.28a) in addition to the Gauss’s law.
The continuity equation can be derived in a similar manner. These derivations
should make good exercises at leisure time, for both the physical understanding
of the equations and for practicing mathematical deductions, especially vector
operations.
The above Maxwell’s equations are written in the form for a field in free
space. When there are other types of materials, such as metal and dielectrics, in
a system, the permittivity (ε) and permeability (µ ) in the equations should be
adjusted accordingly.
For an electric field in a homogeneous material that has relative permittivity
εr, the ε0 term in the Maxwell’s equations will be replaced by ε, where

©2002 CRC Press LLC

©2002 CRC Press LLC

 ε = ε rε 0 (2.31)

In some literature, material loss is included as an imaginary part of

permittivity, while the relative permittivity is its real part. Permeability remains
as that in vacuum µ0 for most of the non-ferrite materials. In general cases, both
permittivity and permeability are tensors, which means they could vary in each
of the three dimensions. Making the situation even more complicated, these
material properties could be dispersive, meaning they are functions of
frequency.
At the optical frequency range, permittivity of a typical metal varies with
frequency dramatically. With some simplifications, the relative permittivity of a
metal is plotted against frequency normalized to the bounce-charge resonant
frequency in Figure 2-5; this characteristic is important when calculating the
time response of a system that contains this type of material. It should be
mentioned that at higher frequencies almost all the material properties,
conductivity and permittivity, exhibit a certain degree of dispersive behavior,
and this physical phenomenon should be included for analyzing high-speed
electronic systems. This type of nonlinear behavior makes the analysis of system
containing dispersive material much more complicated. As an example, a
nonlinear dispersive gain medium is considered and incorporated into a time-
domain simulation at optical frequency in Chapter 10. In addition, as illustrated
in the later chapter, numerical dispersion is compounding the complexity of the
analysis of such systems.

0.5
Relative permittivity (F/m)

-0.5

-1
0 1 2 3
Normalized Frequency

Figure 2-5. Permittivity of a typical metal as a function of a normalized

frequency. This figure shows the dispersive nature of materials.

©2002 CRC Press LLC

©2002 CRC Press LLC

 Before the advancement of computer and numerical techniques in the last
couple decades, Maxwell’s equations were not solved directly for most of the
field-solving efforts, mostly due to the mathematical and numerical difficulties.
Instead, other equivalent variations, as the wave equations and Green’s function
representation, are employed for getting a solution of field distribution. Most of
the solution methods are in frequency domain, and those topics are out of the
focus of this book. So, other than some examples in the later sections, we will
concentrate on the discussion in time-domain and numerical solution using
finite-difference technique, detailed in later chapters.

2.2.8 Wave equations and field retardation

Wave equations in vector forms are derived from Maxwell’s equations by

eliminating either the electric or magnetic field from the equation set, with the
help of the Poisson’s equation, the fact that there is no magnetic charge exists,
and vector identities, resulting in a second-order partial differential equation set
in the following:

∂2 ∂J ρ
(∇ 2 − µε )E = µ + ∇( ) (2.32a)
∂t 2
∂t ε

∂2
(∇ 2 − µε ) H = −∇ × J (2.32b)
∂t 2
Wave equations describe the propagation, including reflection and
transmission, of electromagnetic fields in a vacuum or dielectric media. The
right sides of (2.32) are the sources of the electromagnetic field, and in a source-
free region they all become zero which therefore simplifies the solution to wave
equations. Wave equations are solved analytically frequently for field analysis
using various techniques. At the end of this section, an example of a two-
dimensional cavity is analyzed for its spatial field mode using separation of
variable method.
The general equation for propagation speed of electromagnetic waves is
determined by the properties of the medium in which the wave is propagating:

1
c= (2.33)
µε

where ε and µ are the permittivity and permeability of the medium in which the
wave is propagating, respectively. For waves traveling in vacuums, both

©2002 CRC Press LLC

©2002 CRC Press LLC

permittivity and permeability are the values in vacuums (ε0 and µ0) and the
speed of light is designated as c0,

c0 = 2.998_108 meter/sec.

When an electromagnetic wave is propagating in more than one medium,

for example in the case of a microstrip line where a field could exist in both a
multilayer substrate and air, the wave propagation speed is not easy to calculate.
Most often the effective permittivity, ε eff, is estimated, and then propagation
speed is calculated by replacing ε in (2.33) with its effective value.

1
ceff = (2.34)
ε eff µ

The measurement of power propagation of an electromagnetic field is a

vector called the Poynting vector and defined as

p= E×H (2.35)

where power density p has unit of watts/m2. The definition of Poynting vector in
equation (2.35) clearly indicates that the direction of power propagation is
perpendicular to the directions of both the electric and magnetic field, as
illustrated in Figure 2-6.
An electrical signal at frequency f has a wavelength λ ; both quantities are
related to the propagation speed (c) of the wave

λω
c = λf = (2.36)
2π
where ω is the wave angular frequency, ω=2πf. This equation is applicable to
waves in media as well, where the speed and wavelength all depend on the
properties of the media, much like the dependency demonstrated in (2.33) and
(2.34).
In a source-free and homogenous region, wave equations can be written as

∂2
(∇ − 2 2 ) E = 0
2
(2.37a)
c ∂t

∂2
(∇ 2 − )H = 0 (2.37b)
c 2 ∂t 2

©2002 CRC Press LLC

©2002 CRC Press LLC

 E
p

Figure 2-6. Poynting vector shows the direction of power propagation as an

electromagnetic field travels in space.

Before the field retardation is introduced, a three-dimensional wave

equation is solved as an example, with the help of the Laplace transform (see
Appendix II).
Let’s consider an electric field E(x,y,z,t) that has an initial function form
F(t) at an initial position (x,y,z)=(0,0,0) and the wave goes only to the positive
direction. The wave equation for the electric field, according to (2.37a), is

∂2 ∂2 ∂2 ∂2
( 2 + 2 + 2 − 2 2 ) E(x, y , z,t) = 0 (2.38)
∂x ∂y ∂z c ∂t

where

E( 0 ,0,0, t) = F (t ) and x ≥ 0, y ≥ 0, z ≥ 0

are the two boundary conditions for E(x,y,z,t).

Transform equation (2.38) into Laplace domain, and it becomes

∂2 ∂2 ∂2 s2
( + + − ) E(x,y, z , s) = 0 (2.39)
∂x 2 ∂y 2 ∂z 2 c 2

Equation (2.39) contains spatial derivatives only, and the solution for
E(x,y,z,s) is straightforward if the separation of variable method is used.

©2002 CRC Press LLC

©2002 CRC Press LLC

 Assuming E(x,y,z,s) can be expressed as the multiplication of three
functions

E(x, y, z,s) = f ( x, s ) g ( y , s )h( z , s )

Equation (2.39) now can be written as

''
f x' ' g y hz' ' s 2
+ + = 2
f g h c

Since the sum of the above terms on the left is independent of variables x, y,
and z, each of the terms has to be independent of those variables as well.
Therefore

f x' ' g x'' hx''

= β x2 ( s ) = β y2 ( s ) = β z2 (s )
f g h

where the right side’s terms satisfy

s2
β x2 (s ) + β y2 ( s ) + β z2 ( s ) = β 2 ( s) = (2.40)
c2
It is easy to solve the three second-order differential equations above, and
the solution for the electric field could be written as

− ( β x ( s ) + β y ( s ) + β z ( s )) ( β x ( s ) + β y ( s ) + β z ( s ))
E(x, y, z,s) = p( s )e + q ( s )e (2.41)

where p(s) and q(s) are arbitrary functions to be determined with boundary
conditions.
Equation (2.40) represents the relation of propagation constants in each
direction in space; with the assumption of uniform propagation in a
homogeneous region, the propagation constant in each direction is equal to the
projection of the total propagation constant in the respective axis, as shown in
Figure 2-7.

x0 y0 z0
β x (s) = β ( s) β y ( s) = β ( s) β z ( s) = β ( s)
R R R

where distance

©2002 CRC Press LLC

©2002 CRC Press LLC

 R = x02 + y02 + z02

Equation (2.41) can then be written as

x02 y 02 z 02 s x 02 y 02 z 02 s R R
−( + + ) ( + + ) − s s
E(x, y , z,s) = p ( s )e R R R c
+ q ( s )e R R R c
= p (s )e c
+ q (s )e c

Using the Laplace transform pairs in Appendix II, the time-domain solution
can be transformed from the above expression as

R R
E(x,y, z , t) = p(t − ) + q (t + ) (2.42)
c c

where again p(⋅) and q(⋅) are arbitrary functions to be determined.

z0
R

β(s)

βz(s)
θ

y0

ϕ βx(s) y
x0
βy(s)

Figure 2-7. Illustration of relation between the total propagation constant β(s)
and its components in each direction.

©2002 CRC Press LLC

©2002 CRC Press LLC

 Arbitrary function p(⋅) is a wave propagating in the positive x-direction and
q(⋅) is the one in the opposite direction. Since function F(t) gives the boundary
distribution of E(x,y,z,t) at position (x,y,z)=(0,0,0), and since the solution is
confined in the positive space (x≥0, y≥0, z≥0), it is clear that function p(⋅) is in
fact the function F(t). Therefore, for the solution in the positive space quadrant,
the second term in (2.42) is dropped and the final electric field solution is

R
E(x,y, z , t) = F (t − ) = F (t − t0 )
c

The solution in the above equation represents a time waveform propagating

from one position to another in space, and it is illustrated in Figure 2-8, where
the waveform at a distant position (x0,y0,z0) is an exact copy of the function only
delayed by t0=R/c. This indicates the concept of retardation since a field has to
propagate to be seen from a distance, and the time delay is determined by the
speed of electromagnetic waves.
Equation (2.42) actually is the general solution to the three-dimensional
wave equations, and, in most cases, waves are traveling in both directions,
positive and negative directions, at the same time. As will be seen in the
discussion of the transmission line theory in the next chapter, the general
solution of a transmission line consists of two signals traveling in both positive
and negative directions as well. In fact, the transmission line equation can be
considered the special case of wave equations in a specific field-guiding
structure.

F(t)

t=t0 t

F(t)
(x0, y0, z0)
R
t=0 t z
y

(0,0,0) x

Figure 2-8. A time function started at position (0,0,0) is propagating in space,

according to the solution of wave equations in (2.42).

©2002 CRC Press LLC

©2002 CRC Press LLC

 Electromagnetic fields are retarded from a distance where they are
generated from sources, like a charge or current density distribution inside a
volume. The above three-dimensional example shows an existing waveform
travel from point to point in the spatial domain. Without derivation, retarded
vector and scalar potentials are given here for references; as stated earlier, both
electric and magnetic fields can be calculated from these potential functions:

µ0 J(r' ,t )
A(r,t - td ) =
4π ∫ R( r' ,r) dV'
V
(2.43)

1 ρ (r ' , t )
Φ (r , t − td ) = ∫
4πε 0 V R(r ' , r )
dV ' (2.44)

where td represents a retardation from the source position (vectors with prime) to
the observation point (vectors without prime), as illustrated in Figure 2-9.
Equations (2.43) and (2.44) describe the retardation relation between an
electromagnetic source and its generated field distribution at an observation
position at any given time. Once both vector potentials are known, they are in
general the solution of the distribution because both electrical and magnetic
fields can be derived from them mathematically. Since most of the solutions are
not easy to get, various approximations are applied to simplify the problem.

z A(r,t-td),Φ (r,t-td)

r
R=|r-r’|

V’
y
r’

Figure 2-9. Illustration of retarded potentials at a distance from the source

distribution.

©2002 CRC Press LLC

©2002 CRC Press LLC

 One of the most important factors in determining what type of
approximation to use is the distance R(r’,r) between the source and observation
point. The ruler used to measure the distance is its relative size to the
wavelength of the electromagnetic field, not the absolute dimension. For
example, in antenna design near-field and far-field regions are defined to
separate different field distributions and behaviors in each region, therefore
simplifying the analysis.
When the distance is three to five times bigger than the wavelength, the
field can be considered far field; otherwise, when it is less than one or two
wavelengths, it is included as near field. On the same note, a structure is called
“electrically small” if its physical size is much smaller than the signal
wavelength. Similarly, an “electrically large” structure has dimensions
comparable or larger than the wavelength. The relativity of large or small now
depends on both the physical structure and the frequency of the field. Therefore,
a physically very small circuit could be considered electrically large if its size is
comparable with the wavelength of the signal. In many systems there is a
frequency range of operation; the wavelength used for size comparison should
correspond to the frequency that is of most interest or in some cases a frequency
of an upper limit for better approximation. Most importantly, in the next chapter
circuit theory is introduced when the dimension of the circuit is much less than
the wavelength of the electromagnetic field, and quasi-static field approximation
is introduced.
When fields vary greatly inside an electrically large structure, field
distribution in both time and spatial domains is required to completely describe
the field behavior. Most times this complete solution, called full-wave solution,
is obtained by solving Maxwell’s equations or their equivalences, in
combination with specified boundary conditions.

2.2.9 Time-harmonic field solution

In many of the textbooks, a solution of field distribution starts from the

frequency-domain wave equations, in either scalar or vector form. When
converting an equation from time domain to frequency domain, the wave is
assumed to propagate in a single frequency; therefore, all the partial derivatives
with respect to time in the equations are replaced by jω , where again ω is the
angular frequency of the wave. When the electromagnetic wave is traveling in a
single frequency, called a time-harmonic field, all the field components can be
expressed in their complex forms:

E ( x, y, z , t ) = Re[ E ( x, y, z , t )e jωt ] (2.45a)

H ( x, y, z , t ) = Re[ H ( x, y, z , t )e jωt ] (2.45b)

©2002 CRC Press LLC

©2002 CRC Press LLC

where the time-domain expressions are the real parts of the complex functions.
Other field variables, current, charge, power, etc., could all be written in
similar expressions. Therefore, Maxwell’s equations and others can be
simplified by replacing the time derivative with a term of jω. After substitution,
the equation will have only a spatial derivative and for some geometries, an
explicit general solution can be relatively easily obtained with given boundary
conditions. A frequency-domain solution assumes the system is in a steady state
when all the transient processes died out.
Relations between electric and magnetic fields are simplified with a time-
harmonic expression. From Maxwell’s equations and appropriate vector
identities, the following equations can be derived:

E = η∇ × H (2.46a)

1v
H= n× E (2.46b)
η

where
µ0
η= ≈ 376.7343
ε0

is referred often as the wave impedance of free space.

One of the common applications of wave equations is for the mode analyses
of waveguides, in a rectangular, cylindrical, or spherical coordinate system
using method of variable separation, as shown in the example at the end of this
section. For example, in the frequency domain (2.37) can be written as

(∇ 2 + ω 2 µε ) E = (∇ 2 + β 2 ) E = 0 (2.47a)

(∇ 2 + ω 2 µε ) H = (∇ 2 + β 2 ) H = 0 (2.47b)

where β is defined as the phase propagation constant,

β 2 = ω 2 µε (2.48a)

or in the Cartesian coordinate system

β 2 = β x2 + β y2 + β z2 (2.48b)

©2002 CRC Press LLC

©2002 CRC Press LLC

Another Random Scribd Document
with Unrelated Content
 egli menando questo sconquasso. Imitato da Dante
Inf. 5. La bufèra infernal che mai non resta.
La qual ec. è perciò tale, da dirsi appunto una
continua quartana, ch'ammazza l'uomo. Contina
anche da se val febbre continua. Cadde malato in
Pisa, e d'una contina passò all'altra vita. M. Vill.
Stor.
Ch'ogni ec. così la stessa ventura provasse chiunque
sia nemico d'amore, talchè ne perisse.

Ferza, donna, 'n sul cul: le corna! svina,

Che ti verrà orlando il cappelletto,
E poi 'ngorbierò la pedoncina.

Ferza: sferza. Risponde egli: Una buona frusta ne sta

già dunque pendente sopra le natiche. Le corna!
canchero! Svina: svigna, fuggi a gambe; voce
contadinesca, come sfratta. Rid.
Verrà ec. altrimenti metterà in pompa di guarnizioni il
tuo bel cimiere; cioè più pompose renderà a' pubblici
sguardi le corna tue.
E poi ec. salvati ora, ch'avrem poi tempo pe' nostri
congressi. Ingorbiare è inzeppar nella gorbia, cioè in
quel calzuolo di ferro, in cui incastrasi la punta de'
bastoncini di maneggio. Pedoncina: pezzo degli
scacchi; scherzoso equivoco come quello, Sacch. 165.
Carmignano, vatti quella pedina?

Tu palpi il copertojo fuor del letto;

Ed inghiarando qua e là ti vai:
Che hai tu sott'i piè? diss'io: calcetto.
 Tu palpi ec. soggiunge ella: Tu con chi l'hai? In tanta
paura sembri tu uno sbalzato dal letto, che ne
tasteggia il copertojo, ma non ne gode.
E inghiarando ec. e fai nel tuo entusiasmo un tale
sbatter di piedi, che pare tu pesti ghiara e sassi.

Il gozzo volsi 'ndietro, e screpazzai:

Di mona Bruna crollava la testa:
Silenzio feci, e più non motteggiai.

Screpazzai: allora non mi potei più tenere; e scoppiai

a ridere, voltandomi colla faccia altrove per non riderle
in volto.

Di còmpito non sarie, disse 'n questa,

Se non m'avessi dato tal baciozzo:
E poi si volse intorno: chi mi pesta?

Còmpito: quella quantità di lavoro, che s'assegna

altrui in proporzione della sua attività. Non sarebbe
stata opera da par mio, se almeno ec. Compito è
anche un paese di Toscana, da cui forse in dettato
esser da compito per esser nelle sue cose completo
e perfetto.
Chi mi pesta: dicesi in prov. dentro è chi la pesta,
il cuore è il luogo delle mie pene. Qua dentro è chi la
pesta, Qui sono i miei dolori. Allegr. 43. Donde
viene un tal mio disturbo?

E mona Belcolore e Andreozzo

In guardaspensa entraron quinciritta,
 Mostrando 'l desioso e 'l berlingozzo.

Belcolore ec. io ed essa, quasi novelli Belcolore e

Andreozzo. Il Bocc. 82.
Guardaspensa: dispensa; qui per istanza
apparecchiata a segreti piaceri. Quinciritta: a diritto;
dal Lat. hinc recta.
Berlingozzo: pasta coll'uovo in forma di torta fatta a
spicchi; perciò intesa qui dal Salvini pel sesso
femminile, come pel virile il desioso.

Al levar delle tende parve afflitta;

E stette marcassata pur baciando:
To to gli dissi; ed ella: gitta gitta.

Al levar delle tende: al levar mano dall'opera; tolto

dal costume de' soldati, che levan le tende quando
sloggiano da una terra.
Marcassata: pensa il Ridolfi esser lo stesso che
stramazzata per la stanchezza: quasi, soggiunge il
Salvini, dal Lat. marcida.
To to: detto popolare con cui accompagnasi il porgersi
alcuna cosa; e val prendi.

Se pigne, non maligna tonfolando;

Nel culattaro letto parrà ch'abbia,
Per la gran giravolta scorteando.

Se pigne: sebben ardente si spinga oltre. Non

maligna: non vien però a farsi d'indole rea e nociva.
 Tonfolando: facendo il tonfo, come chi cade in un
pozzo.
Culattaro: scherzo come culiseo e culisburgo; in
cui parrà ch'abbia letto, cioè agiata capacità da
starvi comodamente.

Un bel fancel è Arno, e mena rabbia,

E comincia a svernare a rigoletto:
Col cul in man già si trovò in gabbia;

Fancel: sincope di fanticello, garzoncello. Minaccia

ella ora il rivale di questi suoi amori, presa la
similitudine dall'Arno, piacevole nella sua origine e poi
rovinoso.
Rigoletto: luogo appiè del monte di Falterona, ov'è la
sorgente dell'Arno. Rid. Meno sforzato è l'intendersi
ch'in fin del verno avanti di gonfiare per le sciolte nevi,
sembra esso un ruscello; dal Franc. rigole, canaletto
d'acqua.
Col cul in man: deluso nelle sue speranze; dal volgar
atto di reggersi il deretano nel tapinarsi. Già in
gabbia: altre volte finì a trovarsi per me nelle peste o
in una carcere.

Ma gli scappò, che non era soletto.

Così veggio vendetta de' crudeli,
Come si dice ch'è di Macometto.

Che non era ec. perchè avea de' compagni o degli

appoggi, perciò col loro ajuto gli venne fatto d'uscirne
coll'ossa sane.
 Macometto: si dice che questo celebre impostore
morisse d'un lento veleno, fattogli apprestare col
mezzo d'una sua femmina.

Capruggine, canestri, e cazzaveli,

Tartufi bergamaschi, e pece greca!
Mal cresce chi non pèggiora, direli.

Capruggine ec. l'ho per un di que' modi ammirativi,

come zoccoli! castagne! corbezzoli! la cui forza
dipende totalmente dal soggetto del discorso e
dall'uso. Quasi egli sul discorso di lei esclamasse:
Canchero! l'è piucchè mai furfante.
Mal cresce ec. intendesi d'uno, in cui la malizia sia
così passata in natura, che non possa far questa
progressi se non su' piedi di quella. Direli: gli direi; è
da affibbiargli quel detto ec.

Babbo, il farsetto va, disse 'l Suzzeca;

E misemi la pulce nell'orecchie,
Quando mi fè rimanere 'l manzeca.

Il farsetto va: perde il pelo, va a logorarsi. Trarre la

bambagia del farsetto vale snervare collo
smoderato uso de' piaceri. Sai che mi disse il furbo di
cotesto Suzzeca per impaurirmi, e così farmi
messere? Mi disse ch'il farsetto va.
Misemi ec. metter una pulce nell'orecchio ad alcuno è
maliziosamente suggerirgli cosa, che debba porlo in
apprensione.
Manzeca: cornuto; dall'Ebreo manser, adultero. Salv.
Forse sarà da radice meno straniera, cioè da manzo,
 bue.

Cascato egli è omai infra le vecchie;

E ben mi morse 'l granchio: a quella volta
Subito rinculai, dietro a parecchie

Cascato ec. è già invecchiato, non ne può più. Incalza

furbescamente il Suzzeca, e contesta che quindi
anche il suo farsetto già cascava a pezzi.
Mi morse ec. esser morso del granchio, è
comunemente usato per esser di mano avara. Qui
credo che parli d'una certa arida contrazion di potenza
rapporto a' piaceri: che perciò subito rinculò,
rispondendo con beffe al fischio (squilletta) di
parecchie, che l'invitavano.

Pisciar su la squilletta. Ed io: ascolta;

Allor mi misi la coda fra gambe;
E sperperato mi misson in volta.

Mi misi ec. lo stesso avvenne a me pure, ripiglia egli

per restituirgli la burla; e perciò feci a somiglianza de'
cani, i quali quand'han paura si metton la coda fralle
gambe, e fuggono.

Tra 'l piovano e la scotta fu' intrambe;

Perch'io ancora nespole non mondo,
Cucendo le gavarchie con le strambe.

Piovano: burlescamente per pioggia; onde andar col

piovano, andar quando piove. Scotta: siero.
 Intrambe: una cosa di mezzo. Rid. Pur troppo fui
anch'io in uno stato di consumazione, paragonabile a
quello che di te tu racconti; com'il siero più o men se la
batte coll'acqua piovana.
Nespole: non mondar nespole vale trovarsi a un
dipresso nelle circostanze, ch'un altro di se stesso
descrisse.
Gavarchie: ci convien confessar col Ridolfi di non
saperne indovinare il significato. Strambe: funi fatte
d'erba.

Povero 'n canna son col capo biondo;

Son più leggier d'un can di ventott'anni:
Or non avess'i' peggio all'altro mondo.

Povero ec. son sì macilento da sembrare un di quegli

squallidi pezzenti, che van mendicando per le terre con
canna in mano: eppur son giovine, col capo biondo.
Leggier ec. e son sì secco ed asciutto, che più non
posso le quoja come cane di 28. anni; il quale se si
desse, sarebbe affatto consumato dalla decrepitezza.
Salv.

A caricarli 'l basto tu t'ammanni:

Tu hai maggior ragion, ch'i quarteruoli,
E gongoli tu stesso de' tuoi danni.

A caricarli ec. risponde la donna: A quel ch'io veggio

tu (t'ammanni) ti vai bel bello preparando a farlo
rimaner un asino.
 Quarteruoli: monete d'ottone, che servono di
memoria in un conto. E ragion fate senza
quarteruoli. Sacch. Rim. Sai sì ben fare i tuoi conti,
che tu sei a te stesso un calcolo più sicuro de'
quarteruoli.
Gongoli ec. perciò quel che ti dovea esser di
rammarico e di danno, divien per te argomento di
felicità e di tripudio.

A Lunata impiccati i bugiarduoli:

E per lo fummaiuol tu te n'andrai,
Scoprendosi le torte romagnuoli.

A Lunata ec. rammentati però della fine che fanno a

Lunata i bugiardi. Lunata: terra del dominio Lucchese;
nè altra notizia abbiamo.
Andrai: ti ridurrai al niente, squaglierai com'il fumo
che sbocca da' fumajuoli di sopra a' tetti.
Romagnuoli: suol intendersi per astuti. Venendosi a
scoprire i tuoi furbeschi pasticci, e i tuoi raggiri.

Paura guarda vigna sempre mai,

Dice 'l proverbio, e non mala famiglia.
E zara vaglia, tosto confermai.

Paura ec. il timor d'un male che possa avvenire, ne

guarda dal male stesso: tu al contrario ti sei fidato
sopra mala famiglia, cioè su raggiri e furberie.
Zara: giuoco di dadi; e propriamente zare si dicon
que' punti, che per esser di caso assai raro non
 s'hanno in conto. Orsù, riprende egli, il mio caso è un
po' strano; ma ci scommetto ha da valer questa zara.

La botte piena, e la moglie ebra piglia:

E dormirà pe 'l sugo del sermento,
Mal non pensando, se fia da Corniglia.

La botte ec. cose incompassibili; cattivo

amministratore e prosperità d'affari. Tant'è la zara è
questa: sarà il bersaglio de' miei raggiri, e si crederà
d'andar molto bene.
Sugo del sermento: il vino. Lo farò dormir lunghi
sonni nell'oblio di sestesso, quasi immerso
nell'ubriachezza.
Corniglia: città di Brettagna; equivoco dal Franc.
Cornouaille. Nè penserà perciò se la sua donna
venga intanto a fargli scorno.

E non pensando male, sarà spento

Ogni peccato in lei; e 'n vita eterna
N'andrà diritta con poco tormento.

Non pensando ec. nè di ciò sospettando, la crederà

egli incapace di male; quasi estinto in lei fosse lo
stesso fomite del peccato, e quasi colomba da volar
dritta in paradiso.

Però usa chiarello la taverna:

Amore ha nome l'oste; un soldo rotto
Spendi, e non bere acqua di citerna.
 Chiarello: in gergo vale acqua. Perchè non sospetta di
sua donna, nè guardala con gelosia, perciò può star
bene nè saper come; aver cattiva merce e molto
spaccio.
Amore ec. amore è quel che la fa da oste e vi chiama
la folla de' concorrenti; un misero soldo che tu spenda,
là bei e ti diverti.

E del culo menate ben diciotto:

I' non ci metto se non culo e denti;
E 'l peto 'n cul mi ritorna di botto.

Ben diciotto: assai, quanto ti piace. Così tener

l'invito del diciotto per esser loquace quantomai. E
strombettate pur da dietro quanto v'aggrada per farvi
beffe di me. Il Ridolfi legge delle culmonate, altri
delle culmenate.
I' non ec. poco mi costa il rendervi la pariglia, e
beffarmi di vostra beffe; una stretta di denti,
un'allargata di fondo, ed è fatto.
E 'l peto ec. anzi me ne vedo bene. Riavere il peto
è un modo basso, che significa rinvigorirsi. Appoggiò
lietamente il corpo al desco. E come si suol dir,
riebbe il peto. Malm. 9. 6.

San Biagio è oggi, disson più di venti;

E minaccia madama di gran quello.
Tal cul ta' brache, e sarete contenti.

S. Biagio: festa in Firenze aspettata da' ragazzi. Sa a

quanti dì è S. Biagio dicesi d'un ragazzo da non
potersi burlare, ben sapendo quando gli tocca a far
 festa. Paoli Mod. Tosc. Deh non mi far dire; tu
intendi, ed io intendo; ed ognuno di noi sa a
quanti dì è S. Biagio. Macch. Cliz. 2. 3.
Gran quello: gran che; oppur molte quelle, cioè
molte smorfie. Quasi dicesser costoro: Tu non ci burli;
ben sappiamo il carattere di madama.
Tal cul ec. replica egli: Adattatevi al suo naturale;
conforme ha il piede, così le ponete la scarpa.

I' ho male campane, e non ti uccello,

Benchè sta notte sentii la tregenda:
E nel ventriglio ho l'asso, e nel cervello;

Ho male campane: sono un po duro d'orecchio.

Risponde fra tutti Buchino: Poco ci sento; cioè quel che
tu dici non mi suona.
Benchè ec. spiega com'abbia male campane, non
perchè non ci senta in realtà, avendo pur sentito
passar la tregenda; ma perchè non gli va a fagiuolo
quel ch'egli dice. Si finge esser la tregenda una
processione di notturni fantasmi, che vada attorno con
lumicini.
Nel ventriglio ec. di chi è molto dedito al giuoco
dicesi aver l'asso nel ventriglio. Lo spasso mi piace
sì, ma questa volta non ti sento.

Una meta di bue fu la merenda:

Va che ti buchi, mi disse Buchino:
Iddio per tutti, e ciascun per se spenda.
 Meta: coll'e stretto è una scaricata di ventre. Forse si
dovrà legger metà, essendo insolito ne' manoscritti
l'uso degli accenti. Salv. Par ch'alluda al costume de'
fanciulli Toscani, ch'apparecchiano la sera la merenda
alla befana (o tregenda) per esser ben da essa
trattati. Quasi dicesse; Buoni pasti abbiam già dati a
cotesta befana di donna; abbastanza ha mangiato alle
nostre spalle.
Ti buchi: è volgar detto che la befana buchi il corpo a'
bambini, e perciò le balie se ne servono accortamente
per ispauracchio de' medesimi. Salv. Se da cotesta
befana te la vuoi far ficcare, buon pro ti faccia; a me
non me la ficca certo.
Iddio ec. lascialo a Dio il pensar a tutti; tu pensa a te,
perchè a noi ci pensiam da noi stessi.

Chi prende moglie, e non ha del fiorino,

Non fia di meglio; e non fia canajuola,
Quando per lui non volgesse 'l mulino.

Chi prende ec. chi senza denari s'accasa, com'ha

fatto il marito di lei, ha da far così per istar bene;
lasciar a carico della moglie il far grasso quanto più
può.
Canajuola: specie d'uva che piace a' cani, i quali di lei
si sfamano quando loro manca il biscotto. Qui dice
ch'alla fame di lui non si troverà risorsa fuor della
moglie.

Il piè nell'O non gli hai, nè pur le suola;

Lascialo andar, ch'egli ha nome giuntone,
Perch'ha pregna la mamma e la figliuola.
 Nell'O: nel sedere. Avere un piè nel sedere ad
alcuno è esserselo guadagnato in modo di poter
disporre della sua volontà. Rid. Dopo averlo tanto
ingrassato che credi tu di cavarne? Sgarbi, e
nient'altro. Nè pur le suola: men ch'il piede.
Giuntone: nome fantastico d'un mal uomo, creduto
dal volgo corruttore della figlia e della madre; da
giuntare, ingannare. Quindi aver nome giuntone è
presso la plebe essere un indegno ingannatore. Rid.

Quel bene avrò a calen di mattone

Da quel che visse, solo da mia vaga,
Che per Enea la Reina Didone.

Quel bene ec. quel conforto che Didone ottenne da

Enea, io dalla mia vaga l'avrò solamente dopo calen
di mattone; anzi tanti anni dopo quanti son già gli
anni della vita di lei. Calen di mattone: calende d'un
mese che non vien mai, o verrà dopo il dì del giudizio.
Enea: è noto per l'epica di Virgilio il fatto della famosa
spelonca, in cui la sorte d'una procella recò Enea a
Didone.

E come 'l fregio sul palio s'adaga,

Ch'è pertugiato volte più di mille,
che l'un dall'altro niente si smaga:

Fregio: ornamento di guarnizione. S'adaga: s'adatta,

si soprappone; forse da adagiare.
Niente si smaga: non si stacca punto. Come sta il
fregio strettamente attaccato al panno, a cui è ben
cucito; così io ec.
Così fuss'io con le dolci mammille.
Ma credo ch'altri s'avrà tal diletto;
Ma drommi pace, se non mi ritrille.

Ritrille: quasi titille; cioè se nuovo solletico non

sorga in me a turbar la mia pace. Salv.

Ch'i' credo ben, che fusse maledetto

Il punto l'ora e 'l dì ch'i' nacqui al mondo;
Come chi è in pergamo interdetto.

In pergamo ec. come chi è scomunicato; secondo il

costume di pubblicarsi sul pergamo le scomuniche.

Cicutrennola stammi sempre a tondo,

Ed un putito, e quindici merdosi:
E 'l duol della marsupia sta facondo

Cicutrennola: è la cicutrenna uno stromento da

sonare simile al zufolo. Zufolo usasi per minchione. Sì
che tu resterai zufolo zufolo. Cecch. Il Salvini
arbitra d'intender cutrettola, uccello ch'ha parte
negl'incantesimi; e dicesi di donnicciuola dal menar la
coda.
Putito: puzzolente, che spira cattivo odore come
becco graveolente.
E 'l duol ec. i dolori della (marsupia) borsa, inferma
per esser vuota, son dolori facondi; cioè che rendono
al cercare eloquenti gli stessi chiozzati ec.

A' chiozzari, guascorti, ed acetosi.

 E vannomi le cose tutte bieche,
Per modo ch'io contento gl'invidiosi:

Chiozzati: pieni di chiozze, di gomme, e di piaghe,

com'i lebbrosi. Rid. Guascorti: guaschi in furbesco val
gentiluomini. Salv. Perciò sarà composto di guaschi e
di corti, gentiluomini di quattro soldi. Acetosi: acidi di
stomaco; che perciò spesso paton la fame.

Ed una Beca non mi fa due Beche.

Beca: accorciatura di Menica o di Domenica. Rid. Le

cose mi van sì male, com'a povero disgraziato
ch'avesse una donna nemmen buona a tanto di far due
femmine per servizio almen della casa.
 CAPITOLO DECIMO.

Preso 'l partito, è passato l'affanno?

Tutto cotesto è un manicar grosso:
Deh va che non ci nocci; sta con danno.

Manicar grosso: mangiar a grossi bocconi. È un

operare inconsiderato il far consistere tutto il difficile
nel primo passo di prender risoluzione; quasichè altro
non vi sia poi da temere.
Deh va ec. vatti con Dio, che non ci abbiano a
nuocere coteste tue massime. Sta con danno: abbili
per te que' malanni, di cui vai a caccia colla tua
inconsiderazione.

Gli occhi a' mochi non ebbi io nell'osso;

Ma col marretto mi colse un marritto,
Ch'i' traboccai alla bocca d'un fosso.

Mochi: picciolissima biada, ben guardata da' contadini

affinchè non sia mangiata in erba dagli animali
selvaggi, a cui piace moltissimo. Quindi aver l'occhio
a' mochi per badare con singolar diligenza a' proprj
interessi. Nell'osso: nella cassa dell'occhio. Rid.
Marretto: picciola marra. Marritto: colpo menato
colla dritta, e perciò più gagliardo. Io l'ho provato che
ti fa un operare inconsiderato: mi cascò quindi fra capo
e collo un colpo orribile.
E che ne sai? e che sonv'entro fitto?
Tanto ti sia rivolto che tu muoja,
Con algherìa mi disson con iscritto.

Che ne sai: rispondono alcuni in difesa di quella

massima: Preso 'l partito, è passato l'affanno; e
dicono: Qual maraviglia se talora ne segua male! Si
può forse penetrar l'interno d'ogni cosa, e tutte
prevederne le conseguenze?
Rivolto: non si può intender che del collo; detto da
quella sentenza giudiciaria: S'appenda finchè
muoja. Rid. Ti si possa stroncare il collo: puoi tu tutto
antivedere?
Algheria: con fasto, con voce altitonante. Con
iscritto: il Salvini legge conscritto, cioè senatore,
che dicesi di chi sta sul grave. Forse è gergo frall'uno e
l'altro.

Or s'i' avessi avuta l'epa croja,

Pur risi, come Dio vuol, a formaggi;
E spennacchiato rimasi con noja.

Epa croja: trippa dura com'un tamburo. Col pugno

gli percosse l'epa croja: Quella sonò come fosse
tamburo. Dant. Inf. 30.
A formaggi: a guisa de' formaggi; i quali ridono
quando si fendono e crepano. Perciò ridere a
formaggi sarà rider crepando per dispetto e per
rabbia. Rid. Come Dio vuol: non di cuore: ma a
strapazzo, e per non poter far altro.
Con noja: perchè mi scottava veramente il rimanervi
avvilito e confuso; come gallina spennata, che par che
 si vergogni di se medesima.

Non gir alla badia d'adalticaggi:

Ma feci un pa' di grotte con più doglie;
E di ciò fanno calli assai coraggi.

Adalticaggi: andar alla badia d'adalticaggi sembra

un dettato significante cader giù a piombo dall'alto.
Rid. Opportunamente gli soggiunsi: Non ti levare in
tanta ira, che non avessi a romperti il collo.
Adalticaggi: forse è il paese detto Altipassi da
Tolomeo da Lucca. Salv.
Ma feci ec. aggrottai però intanto un par di ciglia con
tanto d'occhi per interno livore.
Assai coraggi: più d'un cuore a siffatti incontri s'è
dovuto indurire. Sicchè amendue aggiam solo un
coraggio. Dant. Maj.

Già col tramaglio vi prese tre moglie:

Troppo mi se' riuscito del guscio,
Disse veggendo mutatomi scoglie.

Già ec. ah ah, riprese egli borbottando, a costui un

qualche gran fumo è salito alla testa. Credo di dover
così interpretar questo verso, avuto per un mero
bisticcio; riflettendo al proverb. Chi toglie una
moglie, merita una corona di pazienza; chi due,
una di pazzia; e simili. Tramaglio: ampia rete da
pescare.
Troppo ec. più bravo assai mi ti mostri di quel che ti
credevo; non ti tenea da tanto. Metafora tolta da'
pulcini.
 Mutatomi scoglie: quasi avessi cambiata scorza, e
fossi tutt'altro. Scoglia è la pelle ch'ogni anno muta la
serpe.

Ciascun ha l'impiccato suo all'uscio:

Così tre asso nel cul li traesse,
Perch'a mie spese rosecchio ed isguscio.

Ciascuno ec. e qual maraviglia? ognuno ha i suoi

difetti, e soffre ognun qualche eclissi. Quisque suos
patitur manes. Virg.
Così ec. sì tanto gli risposi; perchè popoi t'ho io in
quel servizio. Tre asso: quel che le persone modeste
dicon quattro. Rid. Quasi dicesse: Un corno che dietro
se gli ficchi: non campo (rosecchio ec.) già io alle
sue spalle, nè ho bisogno di lui.

Rezzajo rezzajo mostra che si stesse;

E l'ascoltava per ismemorato:
Col cucchiajo voto mostra, che 'l pascesse.

Rezzajo: quasi chi sta pigro e sonnolento al rezzo.

Qui rezzajo è di due sillabe secondo l'apostrofe
Fiorentina, che pronunzia rezza'.
Pascesse: per pascessi. Pascer col cucchiajo
vuoto dicesi di que' maestri, che fole porgono e non
dottrine. M'ascoltava sì svogliato, che ben mostrava
d'avermi per dettatore di sogni e di fole.

Tra que' che sanno, un sonno ebbi schiacciato;

E poi mi dette qualcosa col pane:
 Chi muta lato, disse, muta fato.

Schiacciato ec. schiacciar un sonno è dormir nella

grossa. Chi sembrai allor tra' sapienti? Sembrai un
uomo stupidito dal più grave sonno.
Mi dette ec. perciò quasi compassionando la mia
grossolana capacità, non col cucchiajo voto prese a
pascermi, ma con massime sostanziali.
Chi muta ec. e la massima fu che chi muta lato,
muta fato; massima più soda di quella di sopra, che
preso partito sia passato ogni affanno.

Poi disse: al badalucco fatti cane;

Allora i' mi ristrinsi nelle spalle:
Bocca pecciola fece 'n tre semmane.

Badalucco: trattenimento giocoso. Altro

ammaestramento mi diede, cioè ch'alle festevoli
radunanze m'accomodassi; imitando lo scherzevole
cagnolino, che con festa corrisponde alle feste che gli
si fanno. Rid.
Pecciola: aver bocca a peccioli è volgar detto
significante quel portar le labbra alzate, e più per
ischerno che per vezzo. Rid,

Ed il prete mangatto, e tre farfalle

Ad un bacin ben pien di giglio 'n giglio
Alla veletta stava per piglialle.

Mangatto: granfia di gatto, truffarello, Malgatto,

cioè astuto, leggerebbe il Ridolfi. Farfalle: persone di
Welcome to Our Bookstore - The Ultimate Destination for Book Lovers
Are you passionate about books and eager to explore new worlds of
knowledge? At our website, we offer a vast collection of books that
cater to every interest and age group. From classic literature to
specialized publications, self-help books, and children’s stories, we
have it all! Each book is a gateway to new adventures, helping you
expand your knowledge and nourish your soul
Experience Convenient and Enjoyable Book Shopping Our website is more
than just an online bookstore—it’s a bridge connecting readers to the
timeless values of culture and wisdom. With a sleek and user-friendly
interface and a smart search system, you can find your favorite books
quickly and easily. Enjoy special promotions, fast home delivery, and
a seamless shopping experience that saves you time and enhances your
love for reading.
Let us accompany you on the journey of exploring knowledge and
personal growth!

ebookgate.com