,1.) . Law.
1;». Hit)... \Wfiv
w. ;. 94%..

I
r... M. 1.3,"?!
{1.6.1 Marti:
. III, ? HVJW‘II‘J . ‘ ”Ind
‘ - M Q»; v .‘vr‘v I‘.‘ .2 .I
l J. 41411:! «1.. 5.
"1&le

 

 

 

    

  
    

    

      

   

   

 

:3 v.” -31.: us...-

Erkgfikgvfirfix. . .3 Save sa. .§-u.."...wx. Wu?u...wmamfi_ 23% ,

THESIS

63(3C3C)

llllllllllllllllllllllllllll'lllllllIll/lllllllll

293 01820 9068

This is to certify that the

thesis entitled

ANALYSIS OF DAIMLER-CHRYSLER/EUROPEAN RADIATED
EMISSIONS TESTING SETUP FOR MODULES

presented by
Christopher Mark Coleman
has been accepted towards fulfillment

of the requirements for

Master's degfiwin Electrical Eng

 

 

W\W

MajU r ofessor

Date 7'6 '39

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

LIBRARY

Michigan State

University

 

 

PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECAUJED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1m Wm“

ANALYSIS OF DAIMLER-CHRYSLER / EUROPEAN
RADIATED EMISSIONS TESTING SETUP FOR
MODULES

By

Christopher Mark Coleman

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE
Department of Electrical Engineering

1999

 

El:

(@3121!

lfifr‘lltl

I A - a ‘
KILE fit"
I

l2 fhr

i- -
mafia“

ABSTRACT
ANALYSIS OF DAIMLER-CHRYSLER / EUROPEAN RADIATED
EMISSIONS TESTING SETUP FOR MODULES
By

Christopher Mark Coleman

Electromagnetic compatibility testing standards and regulations have become
commonplace in industry. In order to ensure that one electronic device does not
interfere with the functional behavior of another, governments and other agencies
impose regulations on the radiated electromagnetic emissions of electronic devices.
In the automotive industry, the Comité International Spécial Des Perturbations Ra-
dioélectriques (CISPR) organization regulates the allowable radiated emissions from
an electronic module to be used in automobiles sold in Europe. A specific testing
setup is required by CISPR to perform the regulatory measurements. This thesis
presents a numerical model of the testing setup that can be used to predict the
effects of altering the testing setup on the measurements of a module’s radiated emis-
sions. The numerical model is compared to an experimental setup constructed at
MSU which also models the CISPR measurement setup. This experimental setup
will help verify the numerical model’s ability to predict the effects of varying certain

physical parameters of the testing setup.

 

 

COpyright by
Christopher Mark Coleman

1999

For my Parents

iv

I ll
and su
Trier:

.
enzrifgs

535%:

l W

3" r:~
lr‘ ‘ 3v
A‘n}‘._iu“

ACKNOWLEDGMENTS

I wish to thank fellow graduate student and friend Mike Havrilla for his help
and support given during my research. Without him, I would not have been able to
perform the measurements and calibration techniques essential to this research. His
enthusiasm and unending quest for knowledge are also a constant reminder of my
reasons for staying in graduate school.

I wish also to thank Dr. E.J. Rothwell for his friendship and help as a thesis
advisor. Dr. D.P. Nyquist also was a great source of ideas and answers throughout
my research. I also thank Dr. Leo Kempel for being a part of my thesis committee.

I want to thank Ahmet Kizilay for his help with using MEX and his thesis style
files. With his help, countless hours of thesis formatting have been saved.

I am grateful to the National Science Foundation. Their funding made work on
this thesis possible.

A special thanks needs to be given to Dave Schilling, Greg Brooks, Ponniah Illi-
varison, Andrew Xiong, Tom Livernois, Poul Anderson, Norman Welsh, and the other
employees at Daimler-Chrysler who dedicated their time to help me with this project.

Most of all, I would like to thank my fiancée Johnna for her support throughout
this thesis. She has helped me from crossing the edge of insanity many times in the

recent past. Johnna will be happier that this thesis is finally complete than I am.

llQ

L‘s;

V IL I A a l 1 ‘
me. E H a
.L VIA N x P ..

yrl\ hark.

HI
fur:
H
pl

if.
\
\

TABLE OF CONTENTS

LIST OF TABLES ................................. ix

LIST OF FIGURES ................................ x
CHAPTER 1

Introduction .................................... 1

1.1 Overview ................................ 1

1.2 NSF GOALI Project ......................... 3

1.3 MSU / Daimler-Chrysler Interaction ................ 3
CHAPTER 2

Project Description ................................ 5

2.1 Overview ................................ 5

2.2 CISPR EMC Requirements ..................... 5

2.3 Daimler-Chrysler’s Testing Setup at Auburn Hills ......... 7

2.4 Goal of Thesis Research ....................... 9
CHAPTER 3

Numerical Modeling ................................ 11

3.1 Overview ................................ 11

3.2 Numerical Electromagnetics Code (NEC) .............. 11

3.3 Geometry ............................... 13

3.4 Procedure ............................... 16

3.4.1 Z to S-Parameter Conversion ................ 19

3.5 Typical Numerical Results ...................... 22
CHAPTER 4

Experimental Work ................................ 35

4.1 Overview ................................ 35

4.2 Description of Experimental Setup ................. 35

4.2.1 Method 1: Transformer Included in the S-Parameter Mea-

surements ........................... 39

4.2.2 Method 2: Transformer Calibrated out of the Measurements 41

4.2.3 Method 3: Measurement with No Transformer Balun . . . 41

4.3 Comparison of Experimental Setup to CISPR 25 Requirements and
Numerical Model ........................... 43

vi

(HA?

Result

"Y'

' ' l
kn‘flip

. i
can

lDD"!
'11 I C.

NEC C

4.4 Typical Experimental Results .................... 46

CHAPTER 5
Results and Analysis ................................ 56
5.1 Overview ................................ 56
5.2 Comparison of Measurement Methods ................ 56
5.2.1 Method 1: Transformer Included in the S-Parameter Mea-
surements ........................... 57
5.2.2 Method 2: Transformer Calibrated out of the Measurements 67
5.2.3 Method 3: Measurement with No Transformer Balun . . . 77
5.3 Comparison of Measured Results to Numerical Results ...... 87
5.3.1 Results for a Wire Harness Location of d = 5 cm ..... 87
5.3.2 Results for a Wire Harness Location of d = 15 cm ..... 97
5.3.3 Results for a Wire Harness Location of d = 20 cm ..... 106
5.3.4 Numerical Results for All Wire Positions .......... 115
5.4 Conclusions .............................. 122
CHAPTER 6
Conclusion ..................................... 125
APPENDIX A
NEC Grid Generator Program .......................... 128
A.1 Overview ................................ 128
A2 Explanation of Program ....................... 128
A.3 Source Code for cmstl ........................ 131
APPENDIX B
Additional Fortran Programs ........................... 147
B.1 Overview ................................ 147
8.2 Z-Parameter Calculation Program .................. 147
BB Source Code for zparmake ...................... 148
BA Z-Parameter to S-Parameter Conversion and Data Processing Pro-
gram .................................. 152
B5 Source Code for sparconv ...................... 154
B.6 Phase Correction Program ...................... 162
8.7 Source Code for sphase ....................... 163

vii

APPENDIX C

NEC Input Code .................................. 165
C.1 NEC Input Code ........................... 165
BIBLIOGRAPHY ................................. 174

viii

LIST OF TABLES

Table 2.1 Lengths labeled in Figure 2.2 .....................

ix

‘«
1,2,“;
h,
h. 3
mm; .

Figure 2.1
Figure 2.2
Figure 3.1
Figure 3.2
Figure 3.3

Figure 3.4

Figure 3.5
Figure 3.6
Figure 3.7
Figure 3.8
Figure 3.9

Figure 3.10
Figure 3.11

Figure 3.12

Figure 3.13
Figure 3.14

Figure 4.1
Figure 4.2

Figure 4.3

Figure 4.4

Figure 4.5
Figure 4.6
Figure 4.7
Figure 4.8
Figure 4.9

LIST OF FIGURES

Diagram of the CISPR 25 radiated emissions module testing setup. 6

Diagram of Testing Setup at CTC. ................. 9
Diagram of system modeled in N EC. ................ 13
Diagram of table sections. ...................... 14
Terminal plane definitions and locations of VT, and In for a linear

two port network ............................ 15
Terminal plane definitions and locations of an and bn for a linear

two port network ............................ 19
Magnitude of 5'11 (d = 10 cm) found using NEC ........... 23
Phase of 511 (d = 10 cm) found using NEC. ............ 24
Magnitude of $21 (or 512) (d = 10 cm) found using NEC ...... 25
Phase of $21 (or .312) (d = 10 cm) found using NEC ......... 26
Magnitude of 822 (d = 10 cm) found using NEC ........... 27
Phase of 522 (d = 10 cm) found using NEC. ............ 28

Magnitude of 5'11 for an isolated table-wire harness antenna and

the table-antenna system. ...................... 29
Phase of 811 for an isolated table-wire harness antenna and the
table-antenna system. ........................ 30

Magnitude of 522 for an isolated dipole and the table-antenna system. 31

Phase of 822 for an isolated dipole and the table-antenna system. 32
Diagram of experimental setup at Michigan State University. . . . 34
Diagram of cable and adaptor arrangement for measurement

method 1. ............................... 38
Diagram of cable and adaptor arrangement for measurement

method 2. ............................... 40
Diagram of cable and adaptor arrangement for measurement

method 3. ............................... 41
Measured magnitude of 511 (d = 10 cm). .............. 45
Measured phase of 811 (d = 10 cm) .................. 46
Measured magnitude of Sn (in dB) ((1 = 10 cm). ......... 47
Measured phase of 321 (d = 10 cm) .................. 48

Measured magnitude of 512 (in dB) (d = 10 cm). ......... 49

Figure 4.10
Figure 4.11
Figure 4.12
Figure 5.1

Figure 5.2

Figure 5.3

Figure 5.4

Figure 5.5

Figure 5.6

Figure 5.7

Figure 5.8

Figure 5.9

Figure 5.10
Figure 5.11
Figure 5.12
Figure 5.13
Figure 5.14
Figure 5.15
Figure 5.16
Figure 5.17
Figure 5.18
Figure 5.19
Figure 5.20
Figure 5.21
Figure 5.22
Figure 5.23
Figure 5.24
Figure 5.25
Figure 5.26
Figure 5.27
Figure 5.28
Figure 5.29

Measured phase of Sn (d = 10 cm) .................. 50
Measured magnitude of 522 (d = 10 cm). .............. 51
Measured phase of 822 (d = 10 cm) .................. 52
Magnitude of 511 (d = 10 cm) using method 1 ............ 56
Phase of 511 (d = 10 cm) using method 1. ............. 57
Magnitude of 521 (in dB) (d = 10 cm) using method 1. ...... 58
Phase of 521 (d = 10 cm) using method 1. ............. 59
Magnitude of 512 (in dB) (d = 10 cm) using method 1. ...... 60
Phase of 512 (d = 10 cm) using method 1. ............. 61
Magnitude of 522 (d = 10 cm) using method 1 ............ 62
Phase of 322 (d = 10 cm) using method 1. ............. 63
Magnitude of SM (d = 10 cm) using method 2 ............ 66
Phase of 311 (d = 10 cm) using method 2. ............. 67
Magnitude of 521 (in dB) (d = 10 cm) using method 2. ...... 68
Phase of 521 (d = 10 cm) using method 2. ............. 69
Magnitude of $12 (in dB) (d = 10 cm) using method 2. ...... 70
Phase of 512 (d = 10 cm) using method 2. ............. 71
Magnitude of 522 (d = 10 cm) using method 2 ............ 72
Phase of 322 (d = 10 cm) using method 2. ............. 73
Magnitude of SM (d = 10 cm) using method 3 ............ 76
Phase of Sn ((1 = 10 cm) using method 3. ............. 77
Magnitude of 521 (in dB) ((1 = 10 cm) using method 3. ...... 78
Phase of 321 (d = 10 cm) using method 3. ............. 79
Magnitude of 512 (in dB) (d = 10 cm) using method 3. ...... 80
Phase of 312 (d = 10 cm) using method 3. ............. 81
Magnitude of 322 (d = 10 cm) using method 3 ............ 82
Phase of 822 (d = 10 cm) using method 3. ............. 83
Magnitude of Sn (d = 5 cm) using method 2. ........... 85
Phase of 311 (d = 5 cm) using method 2. .............. 86
Magnitude of 521 (in dB) (d = 5 cm) using method 2 ........ 87
Phase of 521 (d = 5 cm) using method 2. .............. 88
Magnitude of Sn (in dB) (d = 5 cm) using method 2 ........ 89

xi

Figure 5.30
Figure 5.31
Figure 5.32
Figure 5.33
Figure 5.34
Figure 5.35
Figure 5.36
Figure 5.37
Figure 5.38
Figure 5.39
Figure 5.40
Figure 5.41
Figure 5.42
Figure 5.43
Figure 5.44
Figure 5.45
Figure 5.46
Figure 5.47
Figure 5.48
Figure 5.49
Figure 5.50
Figure 5.51
Figure 5.52
Figure 5.53
Figure 5.54

Phase of 512 (d = 5 cm) using method 2. .............. 90
Magnitude of 322 (d = 5 cm) using method 2. ........... 91
Phase of 822 (d = 5 cm) using method 2. .............. 92
Magnitude of 511 (d = 15 cm) using method 2 ............ 94
Phase of Sn (d = 15 cm) using method 2. ............. 95
Magnitude of 521 (in dB) (d = 15 cm) using method 2. ...... 96
Phase of 821 (d = 15 cm) using method 2. ............. 97
Magnitude of 512 (in dB) (d = 15 cm) using method 2. ...... 98
Phase of 512 (d = 15 cm) using method 2. ............. 99
Magnitude of 522 (d = 15 cm) using method 2 ............ 100
Phase of 522 (d = 15 cm) using method 2. ............. 101
Magnitude of 811 (d = 20 cm) using method 2 ............ 103
Phase of Sn ((1 = 20 cm) using method 2. ............. 104
Magnitude of $21 (in dB) (d = 20 cm) using method 2. ...... 105
Phase of Sn (d = 20 cm) using method 2. ............. 106
Magnitude of 512 (in dB) ((1 = 20 cm) using method 2. ...... 107
Phase of 812 (d = 20 cm) using method 2. ............. 108
Magnitude of 522 (d = 20 cm) using method 2 ............ 109
Phase of 522 (d = 20 cm) using method 2. ............. 110
Numerical magnitudes of 511 for all wire harness positions ..... 112
Numerical phases of Sn for all wire harness positions ........ 113
Numerical magnitudes of 521 for all wire harness positions ..... 114
Numerical phases of Sm for all wire harness positions ........ 115
Numerical magnitudes of 522 for all wire harness positions ..... 116
Numerical phases of $22 for all wire harness positions ........ 117

xii

1.1

the

Li\_ .

4

rnep‘

5.11 . \ l
a.

utter

CHAPTER 1

INTRODUCTION

1.1 Overview

Electromagnetic compatibility (EMC) has become an important area of electrical en-
gineering in industry. Testing standards and regulations imposed by governments and
other agencies have forced companies to pay close attention to the electromagnetic
properties of their products. One of the regulated properties is the radiated electro-
magnetic energy from an electronic device. Most products sold in the United States
must comply with radiated emissions standards set by the Federal Communications
Commission (FCC) [1]. Automobiles manufactured in the United States are one
group of products that are not regulated by the FCC because of an exemption given
by the FCC. However, automobiles manufactured in the United States are subject to
regulations from the Comité International Spécial Des Perturbations Radioélectriques
(CISPR) organization if they are to be sold in Europe.

In order to comply with the CISPR regulations, the Daimler-Chrysler Corporation
has dedicated a semi-anechoic chamber to employ the testing method outlined in the
CISPR 25 document describing radiated electromagnetic emissions [2]. The testing
setup measures the radiated emissions from an electronic module such as a radio,
anti-lock brake system, etc. to be used in an automobile. One of the concerns of
Daimler-Chrysler is the possible effect of currents on a wire harness used to control
the electronic modules during the test on the measured radiated emissions. This thesis
investigates the effect of possible currents on the harness by developing a numerical
model of the testing setup and comparing the model to an experimental arrangement
built at Michigan State University (MSU) similar to the one specified by the CISPR

25 document.

Uflvlji
Chap”

.ML

5M1 ]~-

I

ID m E: d

 

This thesis is organized as follows. First, a brief explanation of the NSF GOALI
project which funded the research for this thesis will be given in Section 1.2 of this
chapter. Following that will be a discussion of the MSU / Chrysler interaction formed
through this project.

The remainder of this thesis is divided into several chapters. Chapter 2 will give
a description of the research project performed for this thesis. This will include
an overview of the CISPR organization’s EMC requirements found in Section 2.2, a
description of Chrysler’s CISPR testing facility found in Section 2.3, and an explana-
tion of the research important for Chrysler and others attempting to meet the CISPR
requirements. The goal of this research project will then be discussed in 2.4.

Chapter 3 is devoted to the numerical modeling of the CISPR testing arrangement.
Section 3.2 will explain how the Numerical Electromagnetics Code (NEC) will be used
to model the CISPR setup. Section 3.3 will then explain the exact geometry which
was modeled with NBC. Typical numerical results are then given in Section 3.5.

The experimental portion of this research is dealt with in Chapter 4. A description
of the experimental setup constructed at MSU is found in Section 4.2. A comparison
of the experimental setup to both the numerical model and the CISPR regulation
testing setup is given in Section 4.3. Typical experimental results are then given in
Section 4.4.

Next, Chapter 5 will give a presentation of all the numerical and experimental
results, as well as a comparison of the results. Section 5.2 will compare the results
using the three different measurement methods used for the experimental work done
for this thesis. Section 5.3 will then compare the remaining measured results with
the numerical results. Conclusions drawn from the numerical and measured data will
then be presented in Section 5.4.

After the conclusion of the thesis in Chapter 6, the Fortran based NEC grid

generating program written for this thesis is included and explained in Appendix A.

The it
found;
GOA}
C-‘LKISE
intera-
assists
turn in

E9311. i

l I
h . _ g '

- l
J} but“

1.3 .\T

A: 3154?;
3E Inter:
Pl ..i
all“. “’9‘?
. I"

, UHIIE
Tl- .
“J" phil

”5 made

i. fl“ I

 

Three other Fortran programs were written for this research and are discussed in
Appendix B. The NEC input code generated from the grid generating program can

be found in Appendix C.

1.2 NSF GOALI Project

The research completed for this thesis project was funded by the National Science
Foundation (NSF). The NSF Grant Opportunities for Academic Liaison with Industry
(GOALI) program helped fund the development of an electromagnetic compatibility
course to be taught at Michigan State University. The program was set up to foster an
interaction between industry and academia. The funding enabled graduate research
assistants from MSU to help design the EMC course with the advice and expertise
from industry by the Daimler-Chrysler Corporation. As well as the course develop-
ment, the funding allowed some of the graduate students to perform research guided

by both Daimler-Chrysler’s industry expertise and MSU’s academic experience.

1.3 MSU / Daimler-Chrysler Interaction

As discussed in the previous section, funding through the NSF GOALI project fostered
an interaction between MSU’s Department of Electrical Engineering and the Daimler-
Chrysler Corporation. Through this interaction, industry experts from the Electrical
/ Electronic Systems Compatibility (EESC) group at Daimler-Chrysler identified a
research project in the area of electromagnetic compatibility suitable for a thesis.
The professors at MSU then guided the research for this thesis to ensure that it fit
the academic mold required for a Master of Science thesis. Thus this thesis was
influenced both by industry and academia in the hope that it will be both practical

and scholarly.

32.3.2“
devices
5.1... .
out: df.’
'44:
Li‘: I, 1'
Clo
ll) (18%!

35930195

 

CHAPTER 2

PROJECT DESCRIPTION

2.1 Overview

In the automobile industry, electronic modules such as car radios, anti-lock brake
controllers, cruise controllers, trip computers, etc., are tested to determine if they
will behave properly in their electromagnetic environment. This is important for the
safety of the automobile and its passengers as well as the reputation of the company
and the satisfaction of its customers. Electronic modules must not radiate electro—
magnetic energy above acceptable levels or they might interfere with other electronic
devices. Thus, organizations such as CISPR decide what levels of radiation are accept-
able and then regulate the allowable radiated emissions from products using specific
testing methods. Companies such as Daimler-Chrysler must then test their products
to determine if they meet the requirements set by the organizations or government
agencies. As briefly discussed in Chapter 1, this thesis investigates CISPR’s regula-
tory testing method itself by developing a numerical model of the testing setup to
compare to an experimental setup constructed at MSU.

This chapter will discuss the goal of the research performed for this thesis. First an
introduction to the CISPR organization and its requirements will be given in Section
2.2. Then an explanation of the testing setup available for radiated emissions testing
at Daimler-Chrysler will be discussed in 2.3. Finally, the goal of this thesis will be

discussed in Section 2.4.

2.2 CISPR EMC Requirements

Most automobile modules such as radios, instrument clusters, anti-lock brake sys-
tems, etc., are connected to a wire harness that contains power lines and bus or

communication lines during normal operation. To operate these modules in the same

functional environment as they are operated in an automobile, these lines must be
connected to the module during a test. Electromagnetic energy coupled to these wires
has the capability of radiating. The shape and length of the wire harness will directly
influence the efficiency of this radiation. The radiated electromagnetic fields from
the wire harness might be received by the measurement antenna and interpreted as
measured radiated emissions from the module. This would be somewhat misleading,
since the origin of the emissions would then lie in the coupling of the energy to the
wire harness. Thus the situation is really a conducted emissions problem.

Because the lines leading to the module are necessary for the module’s operation,
they cannot be removed for a radiated emissions test. Therefore the CISPR 25 doc-
ument imposes requirements on the location of the module and the wire harness to
standardize the measurement setup. A diagram of the CISPR 25 radiated emissions
measurement setup is shown in Figure 2.1. The DUT referenced in the figure is the
Device Under Test. A module is placed in a specified location on a conducting table.
The metallic test bench is located .9 m above the ground plane. The wire harness
that controls the module is required to be located 10 cm back from the edge of the

table, 5 cm above the table, and to be 1.5 m in length.

 

 

 

Receiver _] Bulkhead Connector

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

AI»
Antenna 100 cm
above ground plane
l Wire harness 5 cm
100 cm above table
1 ...... iv ______ A rtificial
..... 43.10-39)..-“ / //Networks
T DUT <—150 cm
Metallic Test / Microwave
Bench ‘—-—— 250 cm > Absorber
I"
Shielded / Power Supply
Enclosure Filter

Figure 2.1. Diagram of the CISPR 25 radiated emissions module testing setup.

This measurement setup is required for any laboratory desiring to test electronic
modules for compliance to CISPR 25 standards. The next section will discuss a

measurement setup found in industry used to test radiated emissions from modules.

2.3 Daimler-Chrysler’s Testing Setup at Auburn Hills

At the Daimler-Chrysler Technology Center (CTC) in Auburn Hills, Michigan, a mea-
surement facility was constructed to test the radiated emissions from electronic mod-
ules. The measurement room is a semianechoic chamber located inside a shielded en-

closure to prevent external electromagnetic interference. This measurement room was

 

originally constructed to perform measurements self-imposed by the Daimler-Chrysler
Corporation, but it is now desired to use the room to perform module tests using the
measurement methods outlined in the CISPR 25 documentation. The automobile in-
dustry is usually exempted from government regulations concerning electromagnetic
compatibility because the automotive self-regulations are usually much stricter than
those imposed by governmental regulatory agencies. However, the Daimler-Chrysler
Corporation is interested in performing the CISPR 25 test as another self-regulatory
testing method.

A diagram of the measurement setup at CTC is shown in Figure 2.2. The con-
ducting table is located .9 m above the ground plane. Notice that the conducting
table where the module is placed does not have the exact shape or dimensions spec-
ified in the CISPR 25 documentation. Therefore, the Systems Compatibility group
at Daimler—Chrysler wishes to understand how small changes in the measurement
setup will affect the measured radiated emissions from an electronic module. If the
Daimler—Chrysler measurement group had a numerical model which could accurately
predict the changes in measured values due to changes in the table size or shape,
they could be used when interpreting the measured data. The dimensions labeled in

Figure 2.2 are listed in Table 2.1.

Table 2.1. Lengths labeled in Figure 2.2.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Physical
Label dimension
(meters)
11 7.55
lg 1.08
13 1.74
14 1.54
15 4.27
16 3.04
17 1.21
19 0.61
ha 2.35
[11 2.16
[12 1.20

 

 

 

 

2.4 Goal of Thesis Research

As discussed in this chapter, it would be valuable to industry to have a numerical
model of the CISPR 25 radiated emissions setup to see the effects on the measurements
by possible RF currents on the wire harness. This numerical model must be compared
to results found by experimental measurements to test its validity. The goals of the
research performed for this thesis are to develop a numerical model of the CISPR
25 radiated emissions testing setup which as closely as possible predicts experimental
measurements made using a similar setup, and then determine if the numerical model
can predict changes due to varying the physical parameters of the system.

A numerical model of the CISPR 25 testing setup was developed using Lawrence
Livermore National Laboratory’s Numerical Electromagnetics Code (NEC) at MSU.
To compliment the numerical work, an experimental arrangement was also con-
structed in the Electromagnetics Laboratory at MSU. Using a vector network an-

alyzer, measurements were performed to compare to the numerical results. Changes

 

[AAA/X

 

 

MN
u'\

 

.\

 

-----

 

table

.\
I!

 

 

 

 

 

 

------- §>€ microwave

antenna 100 cm absorber
above ground plane

 

 

 

 

 

'<———- door—r,

Figure 2.2. Diagram of Testing Setup at CTC.

in the distance of the wire harness to the table edge were made in both the numerical
model and the experimental measurements to verify if the numerical model could

accurately predict the testing setup.

l . .,
i. eipt
.i net“

i
tiff}: It]!

Thane,

siren
paramr

Sparta:

TL

in"? Hill
mirital
a filling
8.3933ch
Hittite}:
lll‘efm,

2’0! the .

“Tilers .

 

CHAPTER 3

NUMERICAL MODELING

3.1 Overview

This chapter discusses the numerical modeling performed for this thesis. The goal of
the numerical modeling done for this project is to find a property or quantity that
can be compared to experimental measurements. As will be discussed in Chapter
4, experimental measurements have been performed with a vector network analyzer.
A network analyzer measures the magnitude and phase properties of a two port
electronic system and computes and displays the S—parameters of the network [3]. A
scheme was developed to compute the Z-parameters of the theoretical table-antenna
system using NEC as described in Section 3.4 of this chapter. These theoretical Z-
parameters are then converted to S-parameters and will be compared to the measured

S—parameters of the system in Chapter 5.

3.2 Numerical Electromagnetics Code (NBC)

The numerical modeling performed for this project was done with the aid of the Nu-
merical Electromagnetics Code (NEC). NEC analyzes electromagnetic systems with
a numerical solution of integral equations established from the boundary conditions
associated with the electric and magnetic fields. NEC solves the integral equations nu-
merically by the method of moments. NEC has been under development at Lawrence
Livermore National Laboratory for over twenty years. Version 4.1 of NEC was used
for the numerical modeling done for this project. The two manuals produced by the
writers of NEC are the user’s manual [4] and the program theory description manual
[5].

N EC is able to calculate the currents on thin wire antenna structures due to various

source excitations. One of the reasons that NEC was chosen as the numerical tool for

10

natal '

ti‘rTura'

G’WJE

E
i" .
A Jqlraz‘,

to . r
’h V
.‘m,

this project is because it can, with sufficient accuracy, model a flat surface such as
the CISPR 25 testing table, thin wire antennae such as the dipole receiving antenna
and the wire harness antenna, and an infinite ground plane. An infinite ground plane
is modeled by modifying the Green’s function in the electric field integral equation.

NEC reads geometry, excitation, and frequency information from an input file
called nec . in. Except for the ground plane, the table and antenna system is modeled
as a system of wires, with each wire split into a certain number of wire segments. NEC
calculates the current on each of these segments. For accurate numerical results,
the segment length needs to be smaller than about .1/\ at the highest frequency
investigated, where /\ is the wavelength. The frequency band that will be investigated
for this project ranges from 50 MHz to 150 MHz. The geometries created for this
project use one segment per wire, so each wire must satisfy this length requirement.

There is no input command in NEC for a thin finite conducting plate such as a
metal table. Therefore, a wire mesh grid was created to model the table. To obtain
accurate numerical results, the mesh size or the length of each wire must be smaller
than about .1 A for all frequencies analyzed.

Another requirement for accurate numerical results is imposed on the wire radius.
The wire radius a is specified for each wire in NEC. Accurate results require that
27ra//\ << 1 for all frequencies studied. Also, the wire radius must be much smaller
than the length of the wire segment. To satisfy these criteria, a wire radius of 1
millimeter was chosen for all of the wires.

NEG-4 is provided in the form of Fortran source code by Lawrence Livermore
National Laboratory. This was very important for this project because the NEC
Fortran code had to be modified to create specialized current files to help in the
computation of an electrical system’s Z and S—parameters. The Digital Visual Fortran
Optimizing Compiler (Version 6.0, Update A) was used to compile all of the Fortran

programs used in this project, including NBC.

11

3.3 Geometry

A geometry was created using NEC to model the experimental setup. This was
accomplished by using the NEG-input code generator developed for this project. This
NEG—input code generator, “cmstl,” is described in detail in Appendix A. The N EC-
input code generator creates a file named “nec.in” that NEG-4 uses to run the
method-of-moments solution to the input file.

A diagram of the setup that was modeled in NEC is shown in Figure 3.1. Free
space is assumed for the surrounding media. A flat conducting surface 2.44 m long
and 1.22 m wide exists .9 m above an infinite ground plane. This table is broken
into six sections as shown in Figure 3.2. This is done so that two vertical wires that
touch the table can do so at the end of a segment of the table’s wires, regardless of
the grid spacing. This is important so that charge can flow freely from the vertical
wires on to the table. The two vertical wires are each .05 m long and connected by a
1.5 m long wire that is broken into many segments by the program cmstl. Each of
the six sections of the table are broken up into a grid of many small wire segments
so that the table appears to be a solid ground plane of finite width and length. The
wire radiator above the table is placed a distance d back from the table lip and the
ends of the wire are placed .5 m from both edges of the table so that the ends are
equidistant from the edge of the table, as shown in Figure 3.1.

A dipole receiving antenna is placed 1.0 m away from the lip of the table at a
height of 1.0 m above the ground plane. The length of the antenna was chosen such
that it will be resonant in the absence of the ground plane, table, and wire radiator
at the middle of the frequency band investigated. This frequency is 100 MHz. A half-
wave dipole (the antenna’s full length [being a half wavelength) will be approximately

resonant when

(3.1)

t0|>w

12

 

 

Top view of ‘f 1.5 m >'

 

 

 

 

 

 

 

 

 

 

setup: , \ , _ __
dipole antenna height
1 m above ground plane
1 in
table .9 m above
.. [
—.-— groun\p ane
l 3distance d
1.22 m wire harness 5 cm
above table
_V__.
:4 2.44 m =1:

Horizontal view

 

 

of table: wire harness location of
\ terminal voltage
l l/
I 5 cm
/
table \ ’ l
location of 50 _9 in
ground plane Ohm load 1
\

 

Figure 3.1. Diagram of system modeled in NBC.

13

 

 

wire harness

\

 

 

 

 

 

 

 

 

 

—‘F- Section |\ I_ distance d Section ISection
.... _J 5 | _5_ _
I l
| |
| |
1.22 in Section : Section : Section
1 2
| l
I: 1.50 m J
+47m4 }A7m+
JI_ .
I i
l: 2.44 m >‘i
i

Figure 3.2. Diagram of table sections.

14

The resonant length is then I = 1.5 m. Thus the antenna should have a half-length of
.75 In so that it will perform best in the middle of the frequency band. The program

cmstl generated the NEG-4 code for this dipole antenna.

3.4 Procedure

As discussed in Section 3.1, the theoretical Z—parameters of the table-dipole antenna
system are desired so that they can be converted to S-parameters and compared to
measured results. Since the most easily accessible quantity from NBC is currents on
wire segments, a method must be used to find the Z-parameters from the currents.

The method used is presented in this section.

 

 

 

Il I2
c —' _ o
+ Linear +
Vl Port 1 Two-Port Port 2 V2
. Network _
c O

 

 

 

 

 

Figure 3.3. Terminal plane definitions and locations of Vn and In for a linear two port
network.

A simple explanation of two-port networks and the Z and S-parameters is now
provided. Assume that two terminal planes, or ports, can be defined in a linear
electrical system. Figure 3.3 shows the location of the voltages and currents at the
terminal planes or ports of a two-port electrical network. If the network is linear, the

voltages can be written as a linear combination of the currents, or

[V] = [Z] [1] (3-2)

15

(litre

aid I

5?.va

where [Z] is the impedance,or Z-parameter matrix of the network [6]. For a two-port

network, the matrix description can be written as

V1 Zn 212 I1
= (3.3)
V2 Z21 Z22 12
Expressions for the voltages V1 and V2 are then found to be
M = 21111 +21212 (3.4)
V2 = 22111 + Z22 I2 - (35)

Since the electrical network describing the table-antenna system can be assumed to
be a reciprocal network, 212 = Z21 [6]. With this assumption, the voltage expressions

become

W = 21111+22112 (3.6)

V2 = 22111 + 222 12 . (3.7)

A method to find the Z -parameters is described as follows. Apply test voltages or
currents at specific ports under certain conditions (port 72. opened, shorted, or loaded)
and measure or calculate a resulting voltage or current due to the test excitation. The
Z parameters can then be found algebraically from these measurements. For the first

step, open-circuit Port 2 so that 12 = 0. Then

W = Z1111 . (3.8)

Next, apply a test voltage V1 at Port 1 and calculate or measure the resulting current

16

\|V[

ylul:

In, :
41.91”

4

11. Z11 IS then found as

Z11 = — . (3.9)

Similarly, if Port 1 is open-circuited (implying 11 = 0) and a test voltage V2 is applied

to Port 2 resulting in a measured or calculated current 12, then
222 = — . (3.10)
Finally, to obtain 221, short circuit Port 2 to force V2 2 0. Then
V2 = 22111+ 222 12 = 0. (3.11)

Apply a test voltage V1 to Port 1, and measure or calculate the resulting currents [1
and 12. Then
12

221 = —Z22 I— - (3.12)
1

Since 222 was found in a previous step, Z21 and therefore Z12 are found from the
above equation.

This method will now be applied to the NEC model. First, the two ports on the
table-antenna system need to be defined. Port 1 is defined to be the input terminals to
the wire radiator above the table. This port is taken to be the bottom wire segment in
the N EC model. Port 2 is taken to be the center wire segment on the dipole receiving
antenna. These port definitions are convenient because in NEC a test Voltage can be
applied over a wire segment, and the current on every wire segment is calculated.

To calculate the Z-parameters numerically, NEC must be run three separate
times under different port excitations and loads. The first run places a near infi-
nite impedance (100 M0) at Port 2 and applies a test voltage at Port 1. Along with
the normal NEC output, the file cur.p122.txt is created. This file contains the real

and imaginary parts of the currents for Port 1. The currents for Port 2 are unneces-

17

at i
cut
me.

IllE}

can

prim

sary because they are forced to zero due to the large impedance. Next, a second run
of NEC is performed with an large impedance at Port 1 and a test voltage applied
at Port 2. NEC produces the file cur_p223.txt that contains the real and imaginary
currents for port 2 during this step. During the final NEC run, Port 2 is shorted,
meaning there is no applied voltage or impedance loading across the Port 2 wire seg—
ment, and a test voltage is applied to Port 1. NEC produces two files for this step,
cur-p121.txt and cur_p2zl .txt, which contain the real and imaginary currents for
Port 1 and Port 2, respectively.

After the three NEC runs, the program zparmake reads in the port current files
(cur.p121.txt, cur_p221.txt, cur_plz2.txt and cur_p2z3.t:xt) and computes the
Z-parameters of the two-port network using the equations derived in this section. An
explanation of the Fortran code as well as the source file for the program zparmake
can be found in Appendix B. After the Z-parameters are calculated, the program
prints the real and imaginary parts of the Z-parameters to the files zl1_num.txt,

z21.num.txt and z22_num.txt.
3.4.1 Z to S-Parameter Conversion

Another program, sparconv, is used to convert the Z-parameter files into S-
parameters so that they can be more easily compared to measured results. 5'-
parameters relate the complex wave amplitudes that enter and exit an n—port linear
network. For the nth port, the complex wave amplitude entering the port is desig-
nated as an, and the complex wave amplitude exiting the port is designated as bu. A
diagram of a two port linear network defining the incoming and outgoing waves can
be found in Figure 3.4.

The scattering matrix, [S], of this network is defined by the relation

[bl = [S] [a] (3-13)

18

figure

HE‘ZWOI
Of

A full

ilrflil‘i‘ke;
found 1

C02Ver1

tawny].

Where

Stile 1d

 

 

 

 

at t 3 a2
Linear
Port 1 Two-Port port 2
Network
19,? . b2

 

 

 

 

Figure 3.4. Terminal plane definitions and locations of an and bn for a linear two port
network.

01‘

bl _ 511512 01 (314)

()2 S21 522 02

A full treatment relating Z—parameters to S—parameters, including the relationship
between the terminal voltages and currents to the complex wave amplitudes, can be
found in standard texts on microwave networks [6]. Of interest here are the formulas
converting between the two sets of parameters. If the Z-parameters of a two-port

network are known, the scattering matrix is found by

[S] = ([2] + [WV1 ([2] - [Ull (3-15)
where
1 O
[U] = (3.16)
0 1

is the identity matrix [6]. Writing expressions for each S-parameter gives

(Zn — Z0)(222 + Z0) — 212221

 

 

$11 2 AZ , (3.17)
_ 2212Zo
3,2 _ AZ , (3.18)

19

_ 2 Z21Z0

 

 

521 — AZ , (3.19)
and
(Zn + Zo)(Z22 — Zo) — 212221
.___ .2
522 M , (3 0)
where 20 is the characteristic impedance of the port and
AZ = (211+ Zo)(Z22 + Zo) —' Z12221 - (3-21)

It can be seen from equations 3.18 and 3.19 that since Z12 = 221, 512 = $21.

The Fortran program sparconv was used to read in the numerically-calculated
Z-parameter files (211_num.txt, z21_num.txt and z22.num.txt), convert the Z-
parameters to S-parameters, and then print out the numerically calculated S-
parameters to the files n311.txt, r1321 .txt, ns 12.txt, and ns22.txt. The program
sparconv was also used to format and manipulate the measured S-parameter data.
A description of the program sparconv, including the source code, can be found in
Appendix B.

Once this method of computing the S—parameters of the NEG-modeled system
was established, the S-parameters of the table-antenna system were generated for
several positions of the table’s wire radiator. The results are discussed in the next

section.

3.5 Typical Numerical Results

Using the method discussed in Section 3.4, the S-parameters of the table-antenna
system were generated numerically for several different cases. The S—parameters
were calculated for d = 5 cm, at = 10 cm, (1 = 15 cm, and d = 20 cm where d is
the distance along the table from the front lip of the table back to the elevated wire,

as shown in Figure 3.1. The S—parameters were determined for each case from 50

20

fl

it

16.1

MHz to 150 MHz in increments of 1 MHz for a total of 101 frequency points. As
discussed in Section 3.4, three NEC runs were needed for each wire position to be
able to determine the S-parameters of the table—antenna system.

After the S-parameters were calculated, they were imported and plotted using
the Axum commercial plotting software. Typical numerical results for the magnitude
and phase of the S-parameters are shown in Figure 3.5 through Figure 3.9 found in
the following pages. The distance d for these plots is 10 cm.

Resonance is defined as the phenomenon where the average electric and magnetic
stored energies of a system are equal [6]. This occurs for Port 1 when the input
impedance looking into Port 1 is purely real. Since the input impedance looking into

Port 1 when Port 2 is match terminated is

1+8“
1*511

 

Z.— = 20 (3.22)

where 20 is the characteristic impedance at Port 1, the input impedance is purely
real when the phase of $11 equals zero degrees.

The magnitude and phase plots for 511 are shown in Figure 3.5 and Figure 3.6,
respectively. These plots indicate that the numerical model predicts a resonance for
the table-wire harness antenna system around 91 MHz. The NEC model shows that
very little signal enters Port 1 outside of a band 40 to 50 MHz wide centered at
the resonant frequency. These magnitude and phase curves are plotted along with
the magnitude and phase of 511 for a table-wire harness antenna in the absence of
dipole antenna in Figure 3.11 and Figure 3.12, respectively. In the labeling for these
figures, the separation corresponds to the distance between the dipole antenna and
the conducting table. The 1 m separation is the normal separation investigated here,
and the infinite separation is the case investigated for the isolated dipole antenna.

Notice that both the magnitudes and phases of 5'11 are nearly identical for the isolated

21

MHz to 150 MHz in increments of 1 MHz for a total of 101 frequency points. As
discussed in Section 3.4, three NEC runs were needed for each wire position to be
able to determine the S-parameters of the table-antenna system.

After the S-parameters were calculated, they were imported and plotted using
the Axum commercial plotting software. Typical numerical results for the magnitude
and phase of the S-parameters are shown in Figure 3.5 through Figure 3.9 found in
the following pages. The distance d for these plots is 10 cm.

Resonance is defined as the phenomenon where the average electric and magnetic
stored energies of a system are equal [6]. This occurs for Port 1 when the input
impedance looking into Port 1 is purely real. Since the input impedance looking into

Port 1 when Port 2 is match terminated is

1+511
1’511

 

Z,” = Z0 (3.22)
where Z0 is the characteristic impedance at Port 1, the input impedance is purely
real when the phase of 511 equals zero degrees.

The magnitude and phase plots for Sn are shown in Figure 3.5 and Figure 3.6,
respectively. These plots indicate that the numerical model predicts a resonance for
the table-wire harness antenna system around 91 MHz. The NEC model shows that
very little signal enters Port 1 outside of a band 40 to 50 MHz wide centered at
the resonant frequency. These magnitude and phase curves are plotted along with
the magnitude and phase of 511 for a table-wire harness antenna in the absence of
dipole antenna in Figure 3.11 and Figure 3.12, respectively. In the labeling for these
figures, the separation corresponds to the distance between the dipole antenna and
the conducting table. The 1 m separation is the normal separation investigated here,
and the infinite separation is the case investigated for the isolated dipole antenna.

Notice that both the magnitudes and phases of S 11 are nearly identical for the isolated

21

Elli

5w

table-wire harness case and for the complete table-antenna system.

The magnitude and phase plots for 521 are shown in Figure 3.7 and Figure 3.6,
respectively. For this NEC model, 321 = 812, so plots for 512 are not included as
separate figures. Notice that the NEC model predicts a near linear phase over the
entire band studied. The model predicts a local maximum in the magnitude of 521
of around -90 dB at 96 MHz.

Figure 3.9 and Figure 3.10 show the magnitude and phase plots for 8'22 where Port
2 is the input to the dipole antenna. The NEC model predicts a resonance at 95 MHz.
These magnitude and phase curves are also plotted along with the magnitude and
phase curves of 822 for an isolated dipole antenna in Figure 3.13 and Figure 3.14. For
the isolated case, Port 2 is assumed to be the only port for convenience. Notice that
the numerical results for both the magnitude and phase of 5'22 of an isolated dipole
antenna closely match those of the dipole antenna in the complete table-antenna
system.

A complete discussion and analysis of the numerically generated S-parameters for
all of the wire positions can be found in Chapter 5. The next chapter will discuss
the experimental portion of the research completed for this thesis project. The 5-
parameters of a system similar to that modeled by NEC were measured at MSU’s
Electromagnetics Laboratory. These measured S—parameters will be compared to the

numerical results obtained using NEC in Chapter 5.

22

Magnitude of S11

1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1

 

l

' I fi—Ifi' T fiTfiI '

l

 

1 a L A 4 . l l A J . l . l . l A l A; J

50 60 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 3.5. Magnitude of S“ (d = 10 cm) found using NBC.

23

Phase of S11 in degrees

 

 

l A l l i l i l r l r l l l r L # l 1_ l

50 60 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 3.6. Phase of 511 (d = 10 cm) found using NBC.

24

Amie» tum. .0 0:3...CDCE

 

é

Magnitude of 821 (dB)
8

l

-120

 

 

-130 4 l r l i 1 . l 4 i r 1 1 1 a l . l . 1 . l
50 60 70 80 90 100 110 120 130 140 150

Frequency in MHz

Figure 3.7. Magnitude of 521 (or 512) (d = 10 cm) found using NBC.

25

Phase of $21 in degrees

100

-500

 

l A 1 m l i 1 i 1 i l A 1 l l r l . l . l

50 60 70 80 90 100 110 120 130 140 150

Frequency in MHz

Figure 3.8. Phase of 521 (or 512) (d = 10 cm) found using NBC.

26

Magnitude of $22

1.0

0.9

0.8

0.7

0.6

0.5

0.4

0.3

 

50 60 70 80 90

1 . l r 1 r 1 1 1 L 1 1 1 r 1 1 1 . l . J

100 110 120 130 140 150
Frequency in MHz

Figure 3.9. Magnitude of 822 (d = 10 cm) found using NBC.

27

Phase of $22 in degrees

30

20

10

 

L 1 i 1 l r l r 1 . I . l i l . l r l . l

 

50 60 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 3.10. Phase of 522 (d = 10 cm) found using NEC.

28

Magnitude of S11

1.0

0.9
0.8
0.7
0.6

0.5

0.4

0.3
0.2

0.1

 

 

 

 

 
 

Infinite Separation
——— 1 m Separation

 

 

 

 

1 r l l 1 l 1 A L r 1 1 1 1 J . 1 r 1

 

50 60 70 80 90 100 110 120 130 140 150

Frequency in MHz

Figure 3.11. Magnitude of 511 for an isolated table-wire harness antenna and the
table-antenna system.

29

 

 

50 1 Infinite Separation
40 r ——— 1 m Separation

30’
20?
10”

 

 

    

 

-10 f
-20 j
-30 ~
.40 ~

Phase of $11 in degrees

 

 

 

1 41 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

5O 60 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 3.12. Phase of 511 for an isolated table-wire harness antenna and the table-
antenna system.

30

‘III‘

. Jr

31

P.-.

.0 .0 :'"
m to o
l ' T ' l

p
\l
' I

   

.0
a)
' T

 

Magnitude of $22

Infinite Separation
——— 1 m Separation

 

 

.0
01
' r

 

.0
A
' i

p
(A)
r

 

1 41 1 1 1 1 1 1 1 1 1 1 1 L 1 1 . 1 1 1

50 6O 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 3.13. Magnitude of 522 for an isolated dipole and the table-antenna system.

31

30

20

10

Phase of $22 in degrees

 

   

 

Infinite Separation
——— 1 m Separation

   

 

 

 

1 41 l

140 150

. 1 1 1 1 1 l 1 1 1 1 1 1 1 1 1

50 60 7O 80 90 100 110 120 130
Frequency in MHz

Figure 3.14. Phase of 522 for an isolated dipole and the table—antenna system.

32

cue,

“1‘1””:
ii; p,
r3.

MIL!
he
. ‘ i

”-7
11.1""

CHAPTER 4

EXPERIMENTAL WORK

4.1 Overview

Experimental work was performed to compliment the numerical modeling portion
of this project. A setup similar to the one modeled in Chapter 3 was constructed
to verify the computer simulation. This chapter will explain the experimental work
performed for this project. Section 4.2 will discuss the apparatus and measurement
techniques. Next, Section 4.3 will give a description of how the experimental arrange-
ment compares to the testing requirements laid out in CISPR 25. Finally, Section
4.4 will give some typical experimental results. A complete presentation of the ex-
perimental results, along with a comparison to the numerical results, can be found in

Chapter 5.

4.2 Description of Experimental Setup

An experimental setup that closely meets the CISPR 25 requirements was constructed
at the Engineering Research Complex at Michigan State University. A diagram of
the setup can be found in Figure 4.1.

The Electromagnetics Laboratory at Michigan State University has access to a
large ground plane (20 feet by 24 feet) where many radar and target identification
experiments have been performed. This ground plane was chosen as the platform
to perform this experiment. One extraneous 4 foot by 8 foot section of the ground
plane was used as the conducting table top. The metal plate has a thickness of 1/8”.
The table dimensions are very close to those found in the CISPR 25 documentation
(2.44 In actual length versus 2.5 m specified length) [2]. The width of the table used
measured 1.22 m. This width is unspecified in the CISPR 25 documents.

Four wooden 4” x4” posts were cut and drilled to act as the legs for the table.

33

conducting table
2.44 m long and

 

   
    

1-22 m wide holes for wire
\ harness positions
\ wire harness 1.5 m of d = 5, 1o, 15,
long and and 20 cm

5 cm abovitable

 

distance to ground "a, ....................... -
plane = .9 m

 

 

to Port 1 of
Network Analyzer

 

location of 50 1" """""
Ohm load . antenna's teed

dipole antenna leads to Port 2 of
1'5 m long Network Analyzer

 

distance to ground
plane = 1.0 m

Figure 4.1. Diagram of experimental setup at Michigan State University.

34

The lengths of the posts were chosen such that the total height of the table measures
.9 m, as specified in the CISPR 25 documentation [2]. The table top and legs were
then assembled and the table was placed on top of the ground plane.

Next, a total of eight holes were drilled through the table as shown in Figure 4.1.
The two holes in each of the four sets were drilled 1.5 m apart and equidistant from
both the sides and the front of the table. The four sets were chosen to be spaced at
5 cm, 10 cm, 15 cm, and 20 cm back from the front edge of the table. This distance
is specified in the CISPR 25 document to be 10 cm. The other distances were chosen
to investigate the effect of distance on the measured S—parameters.

The holes were drilled so that a Type-N female—to-female microwave connector
can be threaded from under the table until the face of the connector is flush with the
table surface. This requires a 5/8”-24 tap, which calls for a pilot hole size ranging
from .580” to .590” in diameter. However, a drill bit with a diameter in this range
could not be found, so a 9/ 16” bit (.5625” diameter) was used. The holes were
then expanded by using a metal file until the tap snugly fit in each hole. The holes
were then hand threaded using the 5 / 8”-24 tap. Then two Type-N female-to-female
microwave connectors were threaded into the table from the underside.

Two Type-N male center conducting pins were then taken from two unused adap-
tors. These pins were soldered to the ends of a thin copper wire. This wire was
carefully measured to be 1.6 m long so that each pin could rest in the female center
conductor of the adaptors threaded into the table and so the wire could be strung 5
cm above the table across the entire 1.5 m distance between adaptors. The wire was
suspended above a conducting table by low dielectric constant Styrofoam supports.

A 50 9 load with a male Type-N connector was connected to one of the Type-N
female connectors on the underside of the table. A cable arrangement with a Type-
N male connector on one end was mated with the other Type-N connector on the

underside of the table. This cable would be used to provide energy to the table, wire

35

 

(PL:

on

[(1

(TL
1 ‘ti

[lit

ll

!
311C,

411‘

radiator and load arrangement.

A dipole antenna was then chosen to act as the receiving antenna for the system.
This type of antenna was selected for several reasons. A dipole antenna designed to
work around the 50 MHz to 150 MHz region is reasonable in physical size so that
it is not too difficult to construct and set up. A broadband antenna designed to
work at these low frequencies might be extremely large. A dipole antenna is very
inexpensive compared to some of the broadband antennae used in industry. With
limited funds, this is a major consideration. Most importantly, the dipole antenna
may be modeled in NEC simply and accurately, which is of primary importance for
this project. Some of the popular broadband antennae have complicated geometries
and feeding arrangements which are difficult to implement in a NEC input code
generator. As seen in Appendix A, the dipole antenna was simple to add to the grid-
generating program developed for this project. A common TV antenna was purchased
to be used as the dipole antenna with two adjustable length poles that can be set
at different angles. A 300 9 twin lead wire originally extended from the antenna’s
feed point. As will be discussed in the following subsections the twin lead is modified
depending on which measurement method is used.

One of the difficulties of these measurements was determining a suitable feeding
arrangement for the dipole receiving antenna. A well known problem exists when a
coaxial line is connected directly to a dipole antenna (inner conductor to one monopole
and outer conductor to the other monopole). The currents on the coaxial line have the
possibility of being unbalanced under this arrangement [7]. Therefore, three different
feeding and calibration methods were performed for one position of the wire radiator
(d = 10 cm). The method producing the best result for this distance was then chosen
and used to perform the measurements for the rest of the wire positions. The three

methods are described in the next subsections.

36

4.2.1 Method 1: Transformer Included in the S-Parameter Measure-

ments

Method 1 was the first antenna feed and calibration setup attempted during these
measurements. First, the dipole antenna’s 300 (2 twin lead was out about two cm
from the antenna’s feed point. A standard 300 Q to 75 Q VHF / UHF TV transformer
was then used to reduce the mismatch to the 50 9 cable used in the electromagnetics
laboratory and, more importantly, to act as a balun to balance the possible unbalanced
currents on the cable. As shown in Figure 4.2, a series of adaptors was then used
to connect the antenna to 50 Q cable. This cable routes the energy received by the
antenna back to the measurement equipment.

An HP 8510C Vector Network Analyzer (VNA) was used to make the measure-
ments for this experiment. The cable leading from the underside of the table was
connected to Test Port 1 of the VNA and the cable leading from the dipole antenna
was connected to Test Port 2. An HP 85054B Type-N calibration kit was used for
the calibration standards. The S—Parameters of the table-antenna network were then
measured from 50 MHz to 150 MHz using 801 frequency points. An IF averaging
factor of 128 was used, as well as a dwell time of 100 milliseconds using a stepped
frequency sweep. The results of these measurements for a wire distance of d = 10 cm
with a comparison to the numerical results are shown in Chapter 5 in Figure 5.1
through Figure 5.8. As will be discussed later, only the case where d = 10 cm is
included in this thesis for Method 1, just to be used as a comparison to Method 2,

the best measurement method.
4.2.2 Method 2: Transformer Calibrated out of the Measurements

Method 2 was the second setup used for this project’s measurements, and it proved to
be the most successful. Instead of allowing the transformer to be part of the measured

system, the calibration plane for this method was moved as close to the receiving

37

 

 

HP 8510C

Port 1

Network
Analyzer

Port 2

 

 

2.4 mm (1)

H

 

2.4 mm (m) to Type K (f) ~

 

3.5 mm (m) to 3.5 mm (m)—

lllllllll

 

 

SMA (f) to SMA (f)

 

 

SMA (m) /:

 

SMA (m)

 

 

SMA (1)10 Type-N (m)

 

Type-N (l) to Type-N (f) ~

llflll

 

 

Underside of Table /

ll

Network
Analyzer

Test Port
Cables

Cables

Calibration
Planes

L

 

—r

2.4 mm (m)

 

—2.4 mm (1) to Type K (t)

 

— 3.5 mm (m) to 3.5 mm (1)

 

—SMA (m) to SMA (f)

 

 

 

SMA (m)

SMA (m)

 

[Tlr\l—\__1llllllllllll

—SMA (f) to Type-N (m)

:— Type-N (t) to SMA (m)

 

—SMA (f) to BNC (r)

 

 

8th (m) to BNC (m)

 

Dklf‘

 

 

 

uuu (I) to Type-F (m)

 

Type-F (f) to
—Twin Lead 75 Ohm to
300 Ohm Transformer

 

Twin Lead to Dipole Antenna

 

 

 

Figure 4.2. Diagram of cable and adaptor arrangement for measurement method 1.

38

dipole as possible. The small amount of 300 (2 transmission line that extended from
the dipole antenna was completely stripped and a female BNC adaptor was soldered
directly to the dipole’s leads. A male BNC to female Type-N adaptor was then used
so that the Type-N calibration kit could be used to establish the calibration plane at
the female Type-N connector. The 300 9 line extending from the transformer was
stripped and a male Type-N connector was connected by again soldering a female
BNC adaptor to the leads and using adaptors to change to a male Type-N connector.
A diagram of the adaptor arrangement can be found in Figure 4.3.

The HP 8510C was again used to perform the S-parameter measurements. The
results obtained using this method more closely matched the numerical results than
results obtained using the other calibration methods. Typical results for this method
can be found in Section 4.4. Complete results for all wire distances can be found in

Chapter 5.

4.2.3 Method 3: Measurement with No Transformer Balun

Although measurements for all four wire distances were already performed using
Method 2 with acceptable results, one more measurement method was performed for
comparison. In the first two measurement methods, a transformer was used as a balun
to balance the potentially unbalanced currents on the coaxial cable feeding the dipole
antenna. As a comparison, the transformer and much of the adaptor assembly was
removed from the Method 2 system and a Type-N male connector was placed at the
end of the coaxial cable leading to the antenna. The system was calibrated to this male
connector, and the connector was mated to the female Type-N adaptor connected to
the dipole antenna. A diagram of the adaptor assembly for this measurement method
is shown in Figure 4.4.

Once again the HP 8510C network analyzer was used to perform the S-parameter
measurements for this measurement / calibration method. The results for the d =

10 cm case are shown in Figure 5.17 through Figure 5.24 found in Chapter 5. Notice

39

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

HP 85100
Network
Analyzer
Port 1 Port 2
l |
Network
Analyzer
Test Port
Cables
2.4 mm (f) -——-———j : 2.4 mm (m)
2.4 mm (m) to Type K 0) —El _ 2.4 mm (1) to Type K (t)
3.5 mm (m) to 3.5 mm (m)—E B— 3.5 mm (m) to 3.5 mm (0
SMA (r) to SMA (n—E E—SMA (m) to SMA (r)
SMA (m) ———————-l}'1 ——SMA (m)
Cables/ SMA (m)
El—SMA (r) to BNC (r)
SMA (m) m B—enc (m) to BNC (m)
SMA (ll 1° TYPe'N W —B B—BNC (r) to Type-F (m)
Type-N (t) to Type-N (t) —E] Type-F (r) to
""""" .—""-—/"'\ ——Twin Lead 75 Ohm to
Undersrde 0‘ Table 300 Ohm Transformer

Calibration (Output Soldered to BNC (1))
P'anes 3— BNC (m) to BNC (m)

\3— BNC (f) to Type-N (m)

3.2-TREE (BIO—Bi 5 (771)“

J BNC (f) Soldered to Dipole
Leads

 

 

 

 

Figure 4.3. Diagram of cable and adaptor arrangement for measurement method 2.

4O

 

 

 

 

 

 

 

 

 

 

 

2.4 mm (m)

 

 

 

 

 

2.4 mm (1) to Type K (i)

 

 

 

B— 3.5 mm (m) to 3.5 mm (f)

 

 

 

SMA (m) ——————'

HP 85100
Network
Analyzer
Port 1 Port 2
Network
Analyzer
Test Port
Cables
2.4 mm (f) -————3 5;.—
2.4 mm (m) to Type K (f) —: L.
3.5 mm (m) to 3.5 mm (m)_:
SMA (r) to SMA (r)——3

 

SMA (m)
SMA (f) to Type-N (m) .—

Type- -N (f) to Type- -N (f) —

B—SMA (m) to SMA (r)

 

Cables

SMA (m)

SMA (m)

 

 

 

CJ Calibration fB— Type- -N (f) to BNC (m)

 

Planes

El

Q BNC (f) Soldered to Dipole

 

H

Underside of Table /

Leads

Figure 4.4. Diagram of cable and adaptor arrangement for measurement method 3.

41

that the results do not match the numerical results as well as those obtained using

the second measurement / calibration method.

4.3 Comparison of Experimental Setup to CISPR 25 Requirements and
Numerical Model

As mentioned in Chapter 2, the goal of this project was to obtain an experimental
setup and a predictable numerical model that both resemble the testing setup outlined
in the CISPR 25 requirements [2]. In this section, the similarities and differences will
be discussed between the experimental setup, the numerical model, and the CISPR
25 requirements.

The table and wire harness radiator dimensions are very similar for the MSU
experimental setup, the NEC model, and the CISPR 25 requirements. The table
length for both the MSU experimental setup and the NEC model was 2.44 m. As
stated previously, this is close to the 2.5 m length specified in the CISPR document.
The radiating wire harness was centered left to right on the table at a distance (I
back from the front edge of the table and suspended 5 cm above the table in all
three models or setups. The distance d specified in the CISPR document is 10 cm
[2]. Measurements for the experimental setup at MSU as well as numerical modeling
was performed for d = 10 cm, as well as 5, 15, and 20 cm. Results from the other
distances were desired because they could possibly predict the effect of varying the
harness’ distance from the lip during a CISPR 25 test.

The distance between the receiving antenna and the table was specified to be
.9 m in the CISPR 25 document. Since the wire harness distance would be varied
from 5 cm to 20 cm in the numerical model and experimental setup at MSU, the .9
m distance between the antenna and the table lip was not used. Therefore, a 1 m
distance was chosen for these models as shown in Figure 3.1 and Figure 4.1.

The main difference between the CISPR 25 requirements and the experimental

42

setup at MSU is that while the CISPR 25 documentation requires the use of a semi-
anechoic chamber inside a shielded enclosure, the experimental setup at MSU was
performed openly above a ground plane in a large laboratory room. Reflections from
walls, the ceiling, metal cabinets, or other structures as well as external electro-
magnetic interference had the possibility of affecting the measurements performed
at MSU. Although the electromagnetics laboratory at MSU does have an anechoic
chamber, the microwave absorber on the walls is not very effective below 500 MHz,
which would not be very useful for these measurements. Also, the chamber is fully
anechoic, and therefore has no ground plane on the floor.

The numerical model follows the CISPR 25 guidelines more closely. No extraneous
objects or structures such as walls or ceilings are included in the numerical model,
which models the effect of the absorber well. No external electromagnetic interference
is present in the numerical model, which resembles the effect of the shielded enclosure
in the CISPR 25 requirements. The numerical model and the CISPR 25 requirements
therefore have similar differences to the experimental setup at MSU. One difference
between the numerical model and the CISPR 25 requirements is that an infinite
ground plane exists 90 cm below the table in the numerical model, and only a ground

plane as large as the semianechoic chamber is present in the CISPR 25 requirements.

4.4 Typical Experimental Results

Using the second measurement method of Section 4.2.2, S—parameter measurements
were performed for all four wire distances (d = 5, 10, 15, and 20 cm). Typical results
are found in this section in Figure 4.5 through Figure 4.12 using (1 = 10 cm. These
results will be discussed briefly in this section.

From the magnitude and phase plots of 5'11, Figure 4.5 and Figure 4.6, respectively,
it is seen that a resonance exists around 95 MHz. The measured data of the table /

wire harness antenna system is very similar to the numerical data calculated in NBC

43

and discussed in Chapter 3. A resonance effect exists at 91 MHz in the NEC model,
which is a close prediction to the measured 95 MHz value.

The magnitude and phase plots for 821 are shown in Figure 4.7 and Figure 4.8,
respectively. These measured plots appear much more jagged as a function of fre-
quency than their numerical counterparts found in Chapter 3. However, the general
envelope of the magnitude data is similar to the numerical model, and the phase
characteristics show a somewhat decreasing linear dependence. A better comparison
between the measured and numerical results would be to plot both sets of data on
the same figure. This is done for all wire distances in Chapter 5.

Comparing the magnitude and phase data for 521 (Figure 4.7 and Figure 4.8,
respectively) with the data for 312 (Figure 4.9 and Figure 4.10, respectively), it is seen
that the two sets of data are nearly identical. Therefore, the reciprocity assumptions
inherent in the NEC model are supported by these measured results.

Figure 4.11 and Figure 4.12 show the magnitude and phase characteristics of
the measured data for 322. Notice that although the data is less smooth than the
numerical results, the measured data shows a resonance at 95 MHz. This is the same
result found from the NEC model.

A complete analysis of the experimental results is presented in Chapter 5. The
measured results will be directly compared with those found from the NEC model

discussed in Chapter 3.

44

Magnitude of S11

1.0
0.9 _

0.8

0.7 y
0.6 _
0.5 _
0.4 g
0.3 g
0.2 .
0.1 .

0.0

 

 

g1+1 11 111 111111411 111

50 60 7O 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 4.5. Measured magnitude of 511 (d = 10 cm).

45

Phase of S11 in degrees

80

60

40

20

 

 

J_ 1 1 1 1 1 1 1 J 1 1 A 1 1 1 1 1 1 1 1 J

50 60 7O 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 4.6. Measured phase of 5'11 ((1 = 10 cm).

46

 

 

1 1 1 1 1 1 l 1 1 1 l 41 1 1 1 1 1 1 1 1 1 1—

50 60 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 4.7. Measured magnitude of 5'21 (in dB) (d = 10 cm).

47

Phase of 821 in degrees

8

Ii)
0
o

1

 

K

a
8
T T

 

r.
8

 

 

-600 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 M1 1
50 60 70 80 90 100 110 120 130 140 150

Frequency in MHz

Figure 4.8. Measured phase of 521 (d = 10 cm).

48

Magnitude of 812 (dB)

 

 

50 60 7O 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 4.9. Measured magnitude of $12 (in dB) (d = 10 cm).

49

_'1
8 a
' i T

§

 

Phase of S12 in degrees
8
o

 

1 1 1 1 1 11 1 A 1 1 1 1 1 1 1 1 1 1M1 1

50 60 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 4.10. Measured phase of 512 (d = 10 cm).

50

Magnitude of $22

1.2
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2

l

V T

I T ff”

 

50

1 1 1 4 1 1 1 1 1 1 1 1 l 1 1 1 1 _1 l 1 1

60 7O 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 4.11. Measured magnitude of 822 (d = 10 cm).

51

Phase of $22 in degrees

 

 

1 1 1 1 A 1

50 60 70 80 90 100 110 120 130 140 150
Frequency in MHz

14 J 1 1 1; L 1 1 1 1

 

Figure 4.12. Measured phase of 822 (d = 10 cm).

52

CHAPTER 5

RESULTS AND ANALYSIS

5.1 Overview

Chapters 3 and 4 discussed the numerical and experimental models of the CISPR 25
testing setup. Numerical and measured S—parameters were obtained for four different
positions of the wire harness on the table (d = 5 cm, 10 cm, 15 cm, and 20 cm). In
addition, three different measurement techniques were tried in order to decide which
method should be used.

This chapter will present the data collected from these experimental and numerical
models. The three different measurement techniques at one wire harness location
(d = 10 cm) will be compared in Section 5.2 to determine which measurement method
is to be used for the rest of the harness locations. The methods will be evaluated
by comparing each set of measured data to the numerical results obtained from the
NEC model. The measured S—parameters and the S-parameters found from the NEC
model for the remaining wire harness locations using the best measurement method

will then be compared in Section 5.3.

5.2 Comparison of Measurement Methods

As discussed in Section 4.2 of Chapter 4, three different measurement methods were
used with the wire harness located 10 cm from the table lip to determine the most
accurate method. The measured S—parameters for all three methods are presented in

this section and compared to the numerical results found by using NBC.

53

5.2.1 Method 1: Transformer Included in the S—Parameter Measure-

ments

Figure 5.1 through Figure 5.8 display both the measured and numerical S-parameters
for a wire harness distance of d = 10 cm. The measured results for these S—parameters
were found by using measurement Method 1. As discussed in Section 4.2.1, measure-
ment Method 1 used a transformer to balance the potentially unbalanced currents
on the cable feeding the dipole antenna, but the method included the transformer as
part of the measured network.

Figure 5.1 and Figure 5.2 are the magnitude and phase plots of Sn for both the
numerical results found using NEC and the measured results found using measure-
ment Method 1. The measured results seem to generally agree with the numerical
results over the entire frequency band. Both results show that a large portion of in-
cident energy is coupled into Port 1 only over a small frequency range where 511 dips
from its larger value to a minimum valley. However, the numerical results predict the
frequency of this minimum to occur at around 91 MHz, whereas the measured results
show the minimum to occur around 95 MHz. The spectra of the numerical results
appears to be shifted down in frequency by about 4 MHz compared to the measured
results across the entire frequency band. This is also true for the phase of Sn shown
in Figure 5.2. The phase of 5'11 indicates a resonance effect as it crosses zero degrees
at 91 and 95 MHz for the numerical and measured results, respectively.

The magnitude and phase plots of 521 are shown in Figure 5.3 and Figure 5.4,
respectively and the magnitude and phase plots of 312 are shown in Figure 5.3 and
Figure 5.4, respectively for both the numerical results and the measured results found
using measurement Method 1. Both the magnitude and the phase of 521 are nearly
identical to those for $12 for both the numerical and the measured results over the
entire frequency band. This is expected behavior for a reciprocal network or system.

Notice that the measured magnitudes are quickly varying as a function of frequency

54

over the frequency band, whereas the numerical magnitudes are slowly varying. How-
ever, the envelope of the measured magnitudes are close to the envelope of the nu-
merical magnitudes even though the measured results appear ragged. The measured
phases also exhibit a ragged behavior over the frequency band in comparison to the
numerical phases. Aside from a constant phase shift of about 150 degrees, the mea-
sured phases show roughly the same trend as the numerical phases. The constant
phase shift could possibly be present due to effects from including the transformer in
the measurements.

Figure 5.7 and Figure 5.8 show the magnitude and phase plots, respectively, of
$22 for both the numerical and measured data. The measured results shown here
using measurement Method 1 do not correspond well at all to the numerical results.
The phase of 522 from the numerical results crosses zero degrees at around 95 MHz
and the magnitude of the numerical results shows a minimum at the same frequency,
indicating a resonance at 95 MHz. The measured results show no such phenomenon
at this frequency. The general trend of the measured curves show no resemblance to
those of the numerical curves.

Inadequacies in the correlation between the measured results using measurement
Method 1 and the numerical results for 322 along with the general ragged behavior
of the measured $21 and 312 results prompted the search for a new measurement
method. Since the transformer balun was included in the measurements, perhaps cal-
ibrating the balun out of the measurements would improve the dipole antenna’s input
impedance and the ragged behavior shown in 321 and 512 would be deemphasized.
This was the reason for developing measurement Method 2, described in Section 4.2.2.

The measured results found using this method are described in the next subsection.

55

Magnitude of $11

1 .0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0

 

—_———

 

 

  

-— Measured
——— NEC

 

 

 

l

 

Yfi 'f v

 

Y I

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 l

50 60 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.1. Magnitude of SM (d = 10 cm) using method 1.

56

Phase of $11 in degrees

80

60

40

20

 

 

— Measured

 

 

 

 

1 1 1 1 1 l 1 1 1 1 1 1 1 1 11 1 1 1 1 1

50 60 70 80 90 100 110 120 130 140 150

Frequency in MHz

Figure 5.2. Phase of S“ (d = 10 cm) using method 1.

57

Magnitude of 821 (dB)

 

—-— Measured
-80 “ —-—— NEC

 

 

'8

-100 ”

 

I
.5
.5
O

I

 

 

 

 

 

-120 ”

   

-130 T

 

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 l 1 1 .

50 6O 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.3. Magnitude of 5'21 (in dB) (d = 10 cm) using method 1.

58

 

100

Phase of 821 in degrees

-500

-600

 

 

 

— Measured
——— NEC

 

 

 

 

 

l 1 1 1 1 1 1 1 1 1 1 1 1 11 1 1 1 1 1 1 1

 

50 60 7O 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.4. Phase of 521 (d = 10 cm) using method 1.

59

Magnitude of S12 (dB)

 

— Measured
~80 * ——— NEC

 

 

 

_'1
o
o

I

 

 

 

 

 

-120 T

 

   

-130 T

 

 

50 60 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.5. Magnitude of 512 (in dB) ((1 = 10 cm) using method 1.

60

Phase of S12 in degrees

 

100 ' —— Measured

 

l -——— NEC

 

 

I
O
I

a

ti:
o
o

"2%

a

-500 “

 

 

-600 L 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
50 60 70 80 90 100 110 120 130 140

Frequency in MHz

Figure 5.6. Phase of 512 (d = 10 cm) using method 1.

61

 

150

1

Magnitude of $22

1.0
0.9

0.8

0.7

0.6

0.5
0.4

0.3

 

 

\
\ ,,’
\ /”’
\\ ///
\ // — Measured
\ // ——— NEC

 

 

 

 

1 1 1 1 1 1 L 1 1 1 1 1 1 1 1 1 l 1 1 1 1

5O 60 7O 8O 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.7. Magnitude of 522 (d = 10 cm) using method 1.

62

100

Phase of $22 in degrees
r'o
8

 

 

— Measured
——— NEC

 

 

L 1 1 1 1 1 1 1 1 l 1 l 1 1 1 1 1 l 1 1

50 6O 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.8. Phase of 522 (d = 10 cm) using method 1.

63

5.2.2 Method 2: Transformer Calibrated out of the Measurements

Figure 5.9 through Figure 5.16 show the S-parameters of the table-antenna system
with a wire harness distance of d = 10 cm for both numerical results and measured re-
sults found using measurement Method 2. This measurement method was performed
by moving the calibration plane for Port 2 as close to the dipole antenna’s feed point
as possible in such a way to remove the impedance and network characteristics of the
transformer balun from the measurements while still utilizing the balancing charac-
teristics of the transformer. This method is described in detail in Section 4.2.2.

Figure 5.9 and Figure 5.10 show the magnitude and phase plots of Sn for the
numerical results and the measured results found using measurement Method 2. The
measured results for both the magnitude and phase again follow the general trend of
the numerical results but occurring about 4 MHz higher than the numerical results.
The measured results using Method 2 are nearly identical to those found using Method
1. This is because the feed and the calibration plane to Port 1 was not changed from
Method 1 to Method 2.

The magnitude and phase plots of 521 are shown in Figure 5.11 and Figure 5.12,
respectively, and the magnitude and phase plots of 512 are shown in Figure 5.13 and
Figure 5.14, respectively for both numerical and measured results. As in the results
shown using Method 1, these results found using Method 2 indicate that 521 = 512
over the entire frequency band, indicating a reciprocal system. The magnitudes found
using Method 2, however, follow the numerical results much more closely than those
found using Method 1. Although the results found using Method 2 are still quickly
varying over the frequency band, much of the rapidly varying deviations has been
reduced from the measured results found using Method 1. The measured phases
found using Method 2 also provide a better match to the numerical results than
those found using Method 1. A constant phase shift between the numerical and

measured data still exists across the frequency band, but it is much smaller than the

64

one found using Method 1. This implies that the phase shift due to the transformer
has been removed from the measured data.

Figure 5.15 and Figure 5.16 show the magnitude and phase curves of 5'22 for both
numerical and measured data. Notice for Method 2 that the measured magnitude
results are very close to those found numerically, unlike the results found using Method
1. Some quickly varying deviations still exist over the frequency band, but the general
trends of the measured results and the numerical results are similar. This is also true
for the phase curve, except that a constant phase shift does exist across the band.

The measured results found using Method 2 match the numerical results more
closely than the measured results found using Method 1. A third measurement
method was also performed using no balancing transformer. The results using this

measurement method are discussed in the next subsection.

65

Magnitude of S11

1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0

 

-————~

 

 

  

—— Measured
—-- NEC

 

 

 

 

 

1 1 l 1 1 1 l J 1 1 1 1 1 1 1 1 1 J 1 1

50 6O 7O 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.9. Magnitude of 511 (d = 10 cm) using method 2.

66

Phase of S11 in degrees

80

60

1

 

 

— Measured
——— NEC

  

 

 

 

 

 

 

 

1 1 1 1 1 1 1 1 1 1 1 1 l 1 1 1 J 1 I

90 100 110 120 130 140
Frequency in MHz

1 l

150

Figure 5.10. Phase of S” (d = 10 cm) using method 2.

67

Magnitude of 821 (dB)
at
O

 

50

Figure 5.11

 

 

— Measured

 

 

 

 

 

 

11 111 111 1 11 111 l 1 11 1

60 70 8O 90 100 110 120 130 140 150
Frequency in MHz

. Magnitude of 821 (in dB) (d = 10 cm) using method 2.

68

Phase of $21 in degrees

:1
o
o

a

-300

-400

-500

-600

1

 

 

—— Measured
--- NEC

 

 

 

 

 

 

 

11 1 1 1 1 1 1 111 1 1 1 1 1 1 1 1 111

50 60 70 80 90 100 110 120 130 140 150

Frequency in MHz

Figure 5.12. Phase of 321 (d = 10 cm) using method 2.

69

 

— Measured

 

 

 

 

 

11 1 1 1 1 1 1 1 1 1 11 1 1 1 1 1 1 1 1 1 1 l

50 60 7O 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.13. Magnitude of $12 (in dB) (d = 10 cm) using method 2.

70

1

 

S12 in degrees

Phase of
§

 

— Measured
100 * ‘TT NEC

 

 

 

2‘
O
o

r

it
8
' l

 

do
0
o

 

 

 

1 1 1 1 1 11 1 1 1 L1 1 1 1 1 1 1 1 111

50 60 7O 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.14. Phase of 312 (d = 10 cm) using method 2.

71

Magnitude of $22

1 .2
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2

 

— Measured
~ ——- NEC

 

 

 

I

 

v I w fl ‘1

 

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 l 1 4 1 J

50 60 70 80 90. 100 110 120 130 140 150
Frequency in MHz

Figure 5.15. Magnitude of 522 (d = 10 cm) using method 2.

72

Phase of $22 in degrees

 

 

— Measured
/—-\\ -—- NEC

 

 

 

 

 

1 4 1 1 1 l 1 A 1 1 1 1 L 1 1 1 l 1 1

50 60 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.16. Phase of 822 (d = 10 cm) using method 2.

73

(I)

fl

'1)
1“

that
115i!

{118

BPPPa
imbah
5nd 5;
Th-
“Dem.
the m“
Wit 35 I
diction
when I};

Currents;

5.2.3 Method 3: Measurement with No Transformer Balun

A third measurement method, Method 3, was discussed in Section 4.2.3. This mea-
surement method was performed by removing the transformer completely and per-
forming the experiment with the dipole leads connected directly to the inner and
outer conductors of the coaxial cable leading to the antenna. The S-parameters mea-
sured using Method 3 are plotted along with numerical results found using NEC for
a wire harness distance of d = 10 cm in Figure 5.17 through Figure 5.24.

Figure 5.17 and Figure 5.18 show the magnitude and phase plots, respectively,
of Sn for numerical results and results found using measurement Method 3. Notice
that the measured magnitude and phase for Method 3 are very similar to those found
using Methods 1 and 2. This is again because the measurement method did not affect
the calibration plane or feeding arrangement for Port 1.

The magnitude and phase curves of 821 are shown in Figure 5.19 and Figure 5.20,
respectively, and the the magnitude and phase curves of 512 are shown in Figure
5.21 and Figure 5.22, respectively, for numerical results and measured results using
Method 3. The measured magnitudes found using Method 3 are much farther from
the numerical results than the other two methods over the entire band. The phase also
appears to have more rapidly varying deviations. The resulting effect of the current
imbalance shows that Method 3 is inferior to the other methods for measuring 5'21
and $12.

The magnitude and phase plots of 822 are shown in Figure 5.23 and Figure 5.24,
respectively. Notice that the results found using Method 3 were better at following
the numerical predictions for the magnitude than those found using Method 1, but
not as close as those using Method 2. A large deviation between the numerical pre-
diction and the measured magnitude using Method 3 can be seen at around 140 MHz
when the measured magnitude dips considerably. This is probably due to imbalanced

currents on the coaxial cable leading to the dipole antenna. The phase curve is also

74

an in

from
resul'
nitud
prese
dista'

next

an improvement over the measured found using Method 1.

After comparing all three measurement methods with the numerical results found
from NBC, it is evident that Method 2 is the superior measurement method. The
results found using this method follow the numerical results the closest for the mag-
nitudes and phases of all of the S-parameters. Therefore, the measured results will be
presented along with corresponding numerical results for the remaining wire harness
distances studied only for measurement Method 2. These results are presented in the

next section.

75

"O ‘1‘ II in."
h 1 Cu at.‘

Magnitude of S11

1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0

 

—————‘

 

 

  

— Measured
——— NEC

 

 

 

 

 

1 1 1 1 1 1 1 l 1 1 1 1 1 1 1 1 1 1 1 1 1 J

50 60 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.17. Magnitude of Sn (d = 10 cm) using method 3.

76

«GECCE E :m ‘0 mmmzl

Phase of $11 in degrees

80

60

40

20

 

 

    

 

— Measured
——— NEC

 

 

 

1 1 1 1 1 1 A 1 1 1 1 1 1 1 1 1 1 I

50 60 7O 80 90 100 110 120 130 140
Frequency in MHz

Figure 5.18. Phase of Sn ((1 = 10 cm) using method 3.

77

 

150

 

53 .mm .0 033::me

Magnitude 01821 (dB)

:1 :1 :1 :1

#- on n: -*

c: c: c> c:
T

 

—— Measured
—-——- NEC

 

 

 

 

-1oo - ,1

 

 

 

 

-150 ‘

 

460 "

 

 

-170 1 1 1 1 1 '1 1 1 1 1 1 1 1 1 1 1 1 L 1 1 1 1 1
50 60 70 80 90 100 110 120 130 140 150

Frequency in MHz

Figure 5.19. Magnitude of 521 (in dB) (d = 10 cm) using method 3.

78

Phase of $21 in degrees

200

100

-400

-500

 

 

 

 

 

 

— Measured
——— NEC

 

 

 

 

\
\
50 60 70 80 90 100 110 120 130 140 150
Frequency in MHz
Figure 5.20. Phase of 521 (d = 10 cm) using method 3.

79

Magnitude of S12 (dB)

 

d:
c:
1

—— Measured
-—— NEC

 

 

 

 

 

 

 

 

 

_'1 _'1 _'1 :1 _'1 ,

a- co to .1 o to

O O O O O O
l

'Tr1

-150 t

 

 

-160 ’

 

l 1 1

-170 1 L 1 1 1 1 1 1 1 l 1 1 1 1 1 1 1 J. 1 1
50 60 70 80 90 100 110 120 130 140 150

Frequency in MHz

Figure 5.21. Magnitude of $12 (in dB) (d = 10 cm) using method 3.

80

 

— Measured
200 - --- NEC

 

 

100 ’

I
O

' I
/

 

-- -100 r . \

Phase of 812 n degrees
ii:
0
o

 

6:
o
o

1

 

 

'8
O

 

 

-500 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 J 1 l 1
50 60 70 80 90 100 110 120 130 140 150

Frequency in MHz

Figure 5.22. Phase of 312 (d = 10 cm) using method 3.

81

Magnitude of $22

1.0

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

 

 

 

 

 

 

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

50 60 70 80 90 100 110 120 130 140
Frequency in MHz

Figure 5.23. Magnitude of 522 (d = 10 cm) using method 3.

82

41

 

1

150

Phase of 822 in degrees

 

 

—— Measured
/‘\ ——— NEC

 

 

 

§~1

 

 

1 1 1 1 1 1 1 + 1 1 J 1 1 1 1 1 1 1 l 1 1

50 60 7O 80 90 100 110 120 130 140 150

Frequency in MHz

Figure 5.24. Phase of 522 (d = 10 cm) using method 3.

83

5.3 Comparison of Measured Results to Numerical Results

In the previous section, it was determined that Method 2 was the superior measure-
ment method, and that this method would be used to obtain the measured data for
the remaining wire harness locations studied. In this section, these measured results

found using Method 2 are presented along with corresponding numerical results.
5.3.1 Results for a Wire Harness Location of d = 5 cm

The wire harness location d was changed to 5 cm on the experimental arrangement
and the S-parameters of the system were measured using Method 2. New numerical
results were also found using N EC after modifying the wire harness location in the
numerical model. The magnitudes and phases of the measured and numerical S-
parameters can be found in Figure 5.25 through Figure 5.32.

Figure 5.25 and Figure 5.26 Show the magnitude and phase plots, respectively, of
Sn for both the numerical and the measured data. Notice that the the measured
resonant frequency remained around 95 MHz, although the resonant frequency found
numerically shifted to around 93 MHz. The resonant frequency calculated numerically
was previously 91 MHz for the d = 10 cm case. Other than this slight shift, the d = 5
cm results for Sn were very similar to those found for the d = 10 cm case.

The magnitude and phase curves of 521 are shown in Figure 5.27 and Figure 5.28,
respectively, and the the magnitude and phase curves of 512 are shown in Figure 5.29
and Figure 5.30, respectively, for both the numerical and measured data taken for
d = 5 cm. Notice that the magnitudes and phases of the measured data follow those
01' the numerical data similarly to the d = 10 cm case.

Figure 5.31 and Figure 5.32 show the numerical and measured data of the magni-
tude and phase of 322, respectively. The measured results follow the numerical results

Similarly to the d = 10 cm case for both the magnitude and phase of 822.

84

Magnitude 01811

1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0

 

’———_——~-

 

 

 

—— Measured
——- NEC

 

 

 

 

 

1 1 1 1 l 1 1 1 1 1 1 1 1 1 l 1 1 1 1 1 1

50 60 7O 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.25. Magnitude of Sn (d = 5 cm) using method 2.

85

Phase of $11 in degrees

80

60

40

r?

 

 

— Measured
——— NEC

 

 

 

 

 

 

1 1 1 1 1 1 1 1 1 1 L 1 1 1 1 1 l 1 1 1 l

50 60 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.26. Phase of 511 (d = 5 cm) using method 2.

86

-" .
o
o 8

Magnitude of 821 (dB)
3

-120

-130

~14O

 

 

— Measured
——- NEC

 

 

 

 

 

 

 

 

50 60 70 8O 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.27. Magnitude of 521 (in dB) (d = 5 cm) using method 2.

87

Phase of $21 in degrees

 

— Measured
100 1 -—-— NEC

 

 

 

 

 

 

 

 

-600 1 1 1 1 1 l 1 1 1 L 1 1 1 1 1 l1 1 11 1 1 1
50 60 7O 80 90 100 110 120 130 140 150

Frequency in MHz

Figure 5.28. Phase of 521 (d = 5 cm) using method 2.

88

Magnitude of S12 (dB)

 

—— Measured
——- NEC

 

 

 

ch
0
1

:1
o
O
' r
\

 

I
d
.5
O

. r

\

 

11

to

o
1
\

 

 

-130 ”

 

-14o -

 

1 1 1 1 1 1 1 1 L 1 1 1 1 1 1 1 1 1 1 1 1

 

50 60 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.29. Magnitude of $12 (in dB) (d = 5 cm) using method 2.

89

degrees

In

Phase of S12

é

 

* -— Measured
100 r "' NEC

 

 

 

_'1
o
o

,

I

-200

 

I
0)
o
O
I

 

T

 

-500

 

 

-600 1 11 1 1 1 1 1 1 1 1 1 1 1 1 L 1 1 1 1 1 1
50 60 7O 80 90 100 110 120 130 140 150

Frequency in MHz

Figure 5.30. Phase of 512 (d = 5 cm) using method 2.

90

Magnitude of 822

1 .2
1 .1
1 .0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2

I

Y T

 

 

——- Measured

 

 

 

 

1 1 1 1 1 1 1 1 1 1 1 1 1 1 l 1 1 1 1 1 1

50 60 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.31. Magnitude of 522 (d = 5 cm) using method 2.

91

Phase of $22 in degrees

 

 

—— Measured

 

 

 

-‘
~~
‘h-

 

 

1 1 1 1 1 l 1 1 1 1 1 1 1 1 1 1 1 I 1 1

50 60 7O 80 90 100 110 120 130 140 150

Frequency in MHz

Figure 5.32. Phase of 322 (d = 5 cm) using method 2.

92

5.3.2 Results for a Wire Harness Location of d = 15 cm

The wire harness location d was changed to 15 cm on the experimental setup and the
S-parameters of the system were again measured using Method 2. New numerical
results were also found using NEC after modifying the wire harness location in the
numerical model. The magnitudes and phases of the measured and numerical S-
parameters can be found in Figure 5.33 through Figure 5.40.

Figure 5.33 and Figure 5.34 show the magnitude and phase plots, respectively, of
Sn for both the numerical and the measured data. The resonant frequency of the
table wire harness antenna was predicted to be back down around 90 to 91 MHz
using the numerical results. The measured resonant frequency remained around 95
MHz. These magnitude and phase plots are also very similar to those found for the
case where d = 10 cm.

The magnitude and phase curves of 521 are shown in Figure 5.35 and Figure 5.36,
respectively, and the the magnitude and phase curves of 5'12 are shown in Figure
5.37 and Figure 5.38, respectively, for both the numerical and measured data taken
for d = 15 cm. Notice that the magnitudes and phases of the measured data follow
those of the numerical data similarly to the d = 10 cm case. The magnitudes are
just slightly lower in [amplitude because the transmitting and receiving antennae are
further away from each other than in the case where d = 10 cm.

Figure 5.39 and Figure 5.40 show the numerical and measured data of the magni-
tude and phase of 5'22, respectively. The measured results follow the numerical results

similarly to the d = 10 cm case for both the magnitude and phase of 5'22.

93

Magnitude of $11

1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0

 

————~~-

 

 

  

—— Measured
-——— NEC

 

 

 

 

 

1 1 1 1 1 1 1 1 1 1 1 1 l 1 l 11 1 1 1 1 I

50 60 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.33. Magnitude of 311 (d = 15 cm) using method 2.

94

Phase of $11 in degrees

80

60

4O

20

 

 

— Measured i

 

——- NEC

 

 

 

 

1 1 l 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

50 60 7O 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.34. Phase of S“ (d = 15 cm) using method 2.

95

Magnitude of 821 (dB)

 

, — Measured
_ ,1\ ——— NEC

 

 

 

o888
I
I
r
l
I
1
\
\

-120 F

 

 

 

.'.1.L1_'1_'1
0)th
oooo

 

-170 ’

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

 

 

50 60 7O 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.35. Magnitude of 521 (in dB) (d = 15 cm) using method 2.

96

200

Phase 01821 in degrees

 

 

— Measured
-—— NEC

 

 

 

 

 

1 J 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

50 60 7O 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.36. Phase of 821 (d = 15 cm) using method 2.

97

-11O ’
-120 ’

 

-130 ”

 

-14O '

Magnitude of 812 (dB)

-150 '

 

-160 ’

 

 

-17O “

1 1 1 1 1 1 1 1 1 1 1 1 1

50 60 7O 80 90 100 110
Frequency in MHz

 

 

— Measured
——— NEC

 

A 1L 1 l L 1'1

130 140 150

Figure 5.37. Magnitude of $12 (in dB) (d = 15 cm) using method 2.

98

 

Phase of $12 in degrees

 

200 P — Measured
i --- NEC

 

 

 

 

100 ”

I
o
' i

 

1'1 1'.) r'o :1

O O O O

O O O O
1

-500 ”

 

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

50 60 70 80 90 100 110 120 130 140 150

Frequency in MHz

Figure 5.38. Phase of 512 (d = 15 cm) using method 2.

99

Magnitude of 822

1.2
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2

 

 

-— Measuredj

 

 

 

1 1 J 1 1 1 1 1 _1 1 1 1 1 1 1 1 1 1 l 1 1

50 60 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.39. Magnitude of 522 (d = 15 cm) using method 2.

100

Phase 01822 in degrees

 

 

— Measured
/"‘\ _-'- NEC
\

 

 

 

-“
~~~
‘-

 

 

1 1 1 1 1 1 1 1 1 . 1 1 1 1 1 1 L 1 1 1 1

50 60 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.40. Phase of 522 (d = 15 cm) using method 2.

101

5.3.3 Results for a Wire Harness Location of d = 20 cm

The wire harness location (1 was changed to 20 cm on the experimental setup and the
S-parameters of the system were again measured using Method 2. Numerical results
were also found using N EC after modifying the wire harness location in the numerical
model. The magnitudes and phases of the measured and numerical S-parameters can
be found in Figure 5.41 through Figure 5.48.

Figure 5.41 and Figure 5.42 show the magnitude and phase plots, respectively, of
311 for both the numerical and the measured data. The resonant frequency of the
table wire harness antenna was found to be around 90 to 91 MHz from the numerical
results. The measured resonant frequency remained around 95 MHz. Again, these
magnitude and phase plots are very similar to those found for the case where d = 10
cm.

The magnitude and phase curves of 521 are shown in Figure 5.43 and Figure 5.44,
respectively, and the the magnitude and phase curves of 5'12 are shown in Figure 5.45
and Figure 5.46, respectively, for both the numerical and measured data taken for
d = 20 cm. Notice that the magnitudes and phases of the measured data follow those
of the numerical data similarly to the case where d = 10 cm, except that they are
slightly lower in amplitude due to the increased distances between the ports.

° Figure 5.47 and Figure 5.48 show the numerical and measured data of the magni-
tude and phase of 522, respectively. The measured results follow the numerical results

similarly to the d = 10 cm case for both the magnitude and phase of 522.

102

Magnitude of S11

1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0

T Y

 

1 1 1 1 J 1 1

50 60 70 80 90

 

———-~-

 

 

— Measured
——— NEC

 

 

 

l 1 1 1 1 1 1 1 1 1 l

100 110 120 130 140 150

Frequency in MHz

Figure 5.41. Magnitude of 511 (d = 20 cm) using method 2.

103

Phase of 811 in degrees

 

 

— Measured
—-- NEC

 

 

 

 

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 l 4 1

50 60 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.42. Phase of Sn ((1 = 20 cm) using method 2.

104

 

— Measured

 

 

 

I
_.L
.5
o
l

 

:1
or
o

r

 

Magnitude of 821 (dB)
:3
O

 

-170 "

 

 

 

-190 1 1 1 1 1 14 1 1 l 1 1 1 1 1 19 1 1 14L 1 1
50 60 70 80 90 100 110 120 130 140 150

Frequency in MHz

Figure 5.43. Magnitude of 521 (in dB) (d = 20 cm) using method 2.

105

Phase of $21 in degrees

-500

-600

-700

 

 

1 1 1 1 1 1 1 1

 

— Measured
——- NEC

 

 

 

 

 

 

1 1 1 1 1 1 1 1 1 1 1

 

50 60 70 80 90

100 110 120 130 140 150

Frequency in MHz

Figure 5.44. Phase of 521 (d =

106

20 cm) using method 2.

Magnitude 01812 (dB)

 

 

   

 

 

 

 

 

 

-70
i — Measured
-90 F ——- NEC
i
-110 ”
F
-130 “
i
-150 L
-170 r
i
-190 1 1 1 1 1 L 1 1 1 1 1 1 1 1 1 l 1 1 L 1

50 60 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.45. Magnitude of 512 (in dB) (d = 20 cm) using method 2.

107

Phase of $12 in degrees

100

-1 00
-200

-300

-500

-600

-700

 

 

—— Measured
-—— NEC

 

 

 

 

 

 

1 1 . 1 1 1 1 L 1 L 1 1 1 1 1 1 1 1 1 1

5O 60 70 80 90 100 110 120 130 140 150

Frequency in MHz

Figure 5.46. Phase of 812 (d = 20 cm) using method 2.

108

Magnitude of 822

1.2
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2

 

1 1 1 1 1 1 1 1 l 1 1 4% 1

50 60 70 80 90 100 110
Frequency in MHz

1

120

 

 

—— Measured
-—— NEC

 

 

 

1 1 1 m

130 140

Figure 5.47. Magnitude of 522 (d = 20 cm) using method 2.

109

1

150

L

Phase 01822 in degrees

 

 

— Measured
,—~\ ——— NEC
\

 

 

 

-‘
'IIIu.
H
“Vt—~-

 

 

1 1 A 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

50 60 70 80 90 100 110 120 130 140 150

Frequency in MHz

Figure 5.48. Phase of 522 (d : 20 cm) using method 2.

110

5.3.4 Numerical Results for All Wire Positions

Now that all of the numerical and measured results for all of the wire positions
have been presented, the numerical results for the four different wire positions will be
compared. Figure 5.49 and Figure 5.50 show the magnitudes and phases, respectively,
of 5'11 for all four wire harness positions. Notice that the resonant frequency of the
table-wire harness antenna decreases as the wire harness distance is increased. This is
possibly due to the interaction of the wire harness with the edge of the table changing
as the wire harness distance is varied.

Figure 5.51 and Figure 5.52 show the magnitudes and phases, respectively, of
521 for all four wire positions found using NEC. Notice that the magnitude of 521
decreases across the entire frequency band as the wire harness distance increases. This
is because the distance between the dipole antenna and the wire harness increases
as the wire harness distance (1 increases. However, the decrease in magnitude is not
uniform as a function of wire harness position. The decrease in magnitude between
the wire harness distances (1 = 10 cm and d = 15 cm is small in comparison to the
decrease in magnitude between the wire harness distances (1 = 15 cm and d = 20
cm. Therefore the relationship between the distance between the terminal ports and
the magnitudes of 521 are not simply related. The interaction of the wire harness
with the edge complicates this relationship. The SIOpes of the phases of 321 increase
slightly as the wire harness distance increases, indicating that the port separation is
increasing.

The magnitudes and phases of 522 found using NEC for all the wire harness
positions are shown in Figure 5.53 and Figure 5.54. It is evident that the wire harness
position does not dramatically affect either the magnitude or the phase of 1922. Both
the magnitude and phase curves for all the wire positions lie almost exactly on t0p

of each other.

111

 

 

Magnitude of S11 (dB)

 

 

 

 

 

? -1 ..........
0.9 f 4
0.8 —

0.7 f

0.6 r

0.5 f — d=05 cm
0.4 _ ......... d =10 cm
0.3 g ....... 333::
0.2 &

0.1 f

0.0 t

 

50 60 7O 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.49. Numerical magnitudes of Sn for all wire harness positions.

112

Phase of $11 in degrees

 

 

 

 

 

 

 

A L 1 1 1 1 1 1 1 1 1 1 1 1 m 1 1 1 1 1

50 60 7O 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.50. Numerical phases of S 11 for all wire harness positions.

113

Magnitude of 821 (dB)

-100 *

 

-110 F

_'1
N
o
f

 

 

 

 

-130 *

 

1 1 1 1 1 1 1 1 l

-140 1 1 1 1 1 1 1 11 1 1
50 60 7O 80 90 100 110 120 130 140 150

Frequency in MHz

Figure 5.51. Numerical magnitudes of $21 for all wire harness positions.

114

100

Phase of $21 in degrees
r'o
8

-400

-500

T

 

 

1

50

1

60

7O

1

1

80

1 1 1 1 l

90 100 110
Frequency in MHz

120

 

 

 

130

140

150

Figure 5.52. Numerical phases of 521 for all wire harness positions.

115

 

d
O
I

p
(O
I

p
m
I

.0
\l
I

 

p
O)
l

Magnitude of 822 (dB)
9
O1

 

 

 

.0
A
I

 

.0
ca

 

1 1 1 J 1 1 1 1 1 L L L 1 1 1 1 l 1 1

50 60 7O 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.53. Numerical magnitudes of 522 for all wire harness positions.

116

Phase of $22 in degrees

20*

10*

 

 

 

 

 

 

 

50 60 70 80 90 100 110 120 130 140 150
Frequency in MHz

Figure 5.54. Numerical phases of 5'22 for all wire harness positions.

117

5.4 Conclusions

In this chapter, the results of the numerical model of the CISPR testing setup using
NEC were compared with measured results. Three different methods were performed
to determine the best method, and then this method was applied to all of the wire
harness distances investigated. As shown in Section 5.2, it is evident that Method
2 is the best measurement method of the three investigated for this research. This
measurement method was therefore performed to find the S-parameters of the table-
antenna system for wire harness distances of d = 5, 10, 15, and 20 cm.

The measured S-parameters in general matched those found numerically from
N EC for all of the wire distances fairly well. For all of the wire harness distances,
the shape of the measured magnitudes and phases of 5'11 were close to those found
numerically, except that the numerically generated curves appeared slightly shifted
in frequency below the measured curves.

The measured magnitudes and phases of 521 and $12 to an acceptable degree
matched the numerical results found using NBC. The measured results had rapidly
varying fluctuations as a function of frequency over the entire frequency band, but
the general shape and amplitudes of the magnitudes were in fair agreement with the
numerical results.

Both the magnitudes and phases of 522 found from measurement followed the
general trend of the numerical results. Rapidly varying deviations in the magnitude
and phase also appeared in the measured results. A constant phase shift between the
measured and numerical curves also existed in the 322 phase plots.

Many factors could be the sources of the rapidly varying fluctuations found in
the 521, $12, and 522 measurements. The calibration routine and the S-parameter
network formulation in general rely on the assumption of linearity for validity. The

transformer balun could be a non-linear device that possibly cannot be calibrated

118

out of the measurements. Another possibility is that the transformer only balances
part of the common mode current on the cable leading to the dipole antenna, and the
remaining unbalanced current is interfering with the measurements. This is plausible
because the transformer-free measurements made using Method 3 exhibited more of
the deviations observed from the measurements using Method 2.

The most probable source for the variations comes from the laboratory itself.
The CISPR 25 specifications require the measurements to be performed in a shielded
semianechoic chamber, but the measurements performed at MSU were over a ground
plane open in a large laboratory. The laboratory contains several metallic cabinets,
metallic support beams above the ground plane, a metallic roof, and various other
conducting bodies. At the low frequency range investigated, 50 MHz to 150 MHz,
the free space wavelengths range from 6 meters down to 2 meters. These conducting
bodies are on the order of single wavelengths away from the table-antenna system.
The presence of these bodies could introduce constructive or destructive interference
depending on the frequency. Constructive and destructive interference could produce
the rapidly varying fluctuations. The numerical model did not include any of these
additional conducting structures, and this could be a prime source for the deviations
between the numerical and measured results.

Another possible source of error from the laboratory is external electromagnetic
interference from radio stations, computers, etc. that could be received by either
antenna and be interpreted as signals originating from the network analyzer. This
possibility of external electromagnetic interference was investigated by match termi-
nating one port and measuring the received signal for a through measurement. Thus
Port 1 was match terminated and 321 was measured, then Port 2 was match termi-
nated and 512 was measured. For both measurements, any external electromagnetic
interference present did not register above the noise floor of the HP 85100 network

analyzer. Thus the possibility of external electromagnetic interference influencing the

119

measurements was neglected.

As discussed in Chapter 2, the goal of this thesis research was to develop a numer-
ical model that could adequately predict the effects of currents existing on the wire
harness above a table on measurements performed for a CISPR 25 radiated emissions
test. The results discussed in this chapter show that the numerical model developed
using NEC is able to predict the general trends of the S-parameters for a physical
table / antenna system such as the CISPR 25 setup. As shown in Section 5.3.4, the
numerical model predicts that the magnitude of 521 of the table-antenna system will
decrease as the wire harness distance d increases. This agrees with physical intuition
because the separation between the ports increases as the wire harness distance d in-
creases. The slope of the phase of 521 increases as the wire harness distance increases,
also indicating that the distance between the ports increases.

The measurements performed at MSU have many rapidly varying fluctuations
compared to those predicted from the numerical model, but the general shapes of
the S-parameters as a function of frequency are very similar. Variations in the wire
harness location do not adversely affect the ability for the numerical model to predict
the measured S—parameters of the system. Thus the numerical model serves as an
adequate representation of the experimental setup at MSU and therefore should accu-
rately predict the effects of different parameters on the CISPR 25 radiated emissions

test.

120

CHAPTER 6
CONCLUSION

This thesis has presented a comparison between a numerical model and an experi-
mental arrangement that both model the radiated emissions testing setup for modules
as required by the CISPR 25 organization. The goals of this thesis, as discussed in
Chapter 2, are to develop a numerical model of the CISPR 25 testing setup that
predicts results obtained through experimental measurement of a similar setup, and
to verify that the numerical model can accurately predict the effect of varying the
wire harness distance on measurements. As discussed in Chapter 5, the numerical
model developed does adequately predict the S—parameters of the experimental table-
antenna arrangement constructed at MSU. Variations in the wire harness distance in
the numerical model also lead to predictable results.

A description of the numerical model developed for this project using the N umeri-
cal Electromagnetics Code was given in Chapter 3. The numerical model was used to
calculate the S-parameters of the table-antenna system. In Chapter 4, a description
of the experimental setup constructed at MSU was discussed, and the methods and
techniques used to measure the S—parameters of this system were explained.

Finally, Chapter 5 presented all of the numerical and measured results of the S-
parameters of the table-antenna system. The numerical results were compared with
the measured results for different measurement methods and various wire harness
locations. Conclusions about the success of the numerical model and the experimental
setup were drawn at the end of the chapter. As mentioned above, it was determined
that the numerical model was successful in predicting the S-parameters of the system.

Although the goal of this research was achieved, further research could produce
better results. N EC was used to numerically model the CISPR 25 testing setup for

the research presented here. NEC is a good simulation tool that can be used for a

121

wide variety of electromagnetic structures, but it may not have been the best tool
to use to model flat surfaces such as the metal table. A better numerical simulation
could be developed with a method of moments solution dedicated to this structure
instead of a multi-use program.

Another area of future research is further experimental verification of the nu-
merical model. The use of a shielded anechoic chamber would be beneficial if the
measurements were to be repeated. A better balun than the transformer used in the
experimental setup used here could possibly also improve the measurements. It would
still be helpful, however, to be able to de-embed the properties of the balun from the
measurements.

Another note is that many of the problems encountered while performing this
research seemed to stem from the frequency range investigated. The band of frequen-
cies from 50 MHz to 150 MHz is nearing the upper limit of frequencies in which NEC
can accurately model this system with the available computer power. However, the
frequency band is much lower than the frequencies typically studied at MSU. This
means that the measurement facilities at MSU are well equipped to investigate phe—
nomenon at frequencies greater than 1 GHz, but frequencies far below this are not
absorbed well by the microwave absorber available in MSU’s laboratories.

The area of electromagnetic compatibility is a fast growing and interesting field
in electrical engineering. As well as being an interest in academia, electromagnetic
compatibility is becoming a primary concern in industry as testing standards and
regulations become more prevalent. This thesis illustrates the need for more research

into testing standards, numerical simulations, and measurement techniques to be used

in this field.

122

APPENDICES

123

APPENDIX A
NEC GRID GENERATOR PROGRAM

A. 1 Overview

A NEC input file can potentially contain many hundreds of lines of code. Each
specification of a wire segment consists of a tag number and six wire coordinates
(three starting point coordinates and three endpoint coordinates). This information
not only has to be precisely calculated, it must be entered into the input file without
typographical errors. Using a separate computer program to calculate the required
position of each wire and then tabulate the results in the format that NEC recognizes
is much more efficient and reliable than attempting to generate a NEC input file by
hand. Thus, a NEC grid generating program was developed for this research project.
The program was written in Fortran 90 and compiled with the same Digital Visual
Fortran compiler that was used to compile the NEC-4 program used for numerical

modeling.

A.2 Explanation of Program

The program cmst1 is the grid computing code written to generate the N EC input file
for this project. The program first calls the ‘beginning’ subroutine. This subroutine
opens the output files that will be produced (nec . in, nec .par, and zparinfo .par),
initializes variables, and determines which Z—parameter will be tested for each NEC
run. The starting frequency, frequency range, and number of frequency points are
also determined. This frequency information is written to the file zparinfo.par,
which will be used by the program zparmake (explained Section B2 of Appendix B)
to aid in the computation of the Z -parameters of the table-antenna system.

Next, the subroutine ‘table’ is called. This subroutine calculates all of the coordi-

nates of the wire segments and stores them in the 6x5000 matrix named ‘coordinates’.

124

The coordinates are calculated by first breaking the table up into six sections (as dis-
cussed in Section 3.3), then separating the wires into :r-directed wires and y—directed
wires, and finally calculating the maximum spacing between wires for each section.
The subroutine then evaluates the positions of each wire. The coordinates of all the
wires comprising the dipole antenna were then computed in the ‘table’ subroutine.
At the end of the subroutine, the ‘coordinates’ matrix is used to print out all of
the wire coordinates to the nec.in file to be used by NEC. These coordinates are

given to NEC in the format
GW ITG NS XVI YWl 2W1 X142 W2 2W2 RAD

where GW is N EC’s straight wire specification command, ITG is the tag number for all
the segments of the wire, NS is the number of segments on the wire, X111, “11, and
2W1 are the 11:, y, and z coordinates of the first end of the wire, X112, W2, and 2142
are the 2:, y, and z coordinates of the second end of the wire, and RAD is the Wire’s
radius [4]. All coordinates and radii are assumed to be in meters.

Before the ‘table’ subroutine is ended, the nec .par file is created. This file contains
the information about which wires were x-directed, y-directed, and other possible
directions, as well as the wire tag numbers of the port 1 and port 2 that are used
for the network formulation of the table-antenna system discussed in Chapter 3. The
modified NEC-4 program will use these pieces of information to determine which
currents to write to different output files.

The next subroutine called is named ‘exite’. The purpose of this subroutine is
to end the wire geometry section, write the excitation and load commands, and add
a ground plane under the table to the nec.in file. A voltage source was chosen in
the Ex NEC structure excitation command. The voltage source was placed at either
the dipole’s feed point or the table’s driving point depending on which Z-parameter

calculation was chosen when the program was run. The NEC impedance loading

125

command LD was used to load the antenna at various positions, depending on which
Z-parameter was being calculated. The load was written to the nec.in file in a
resistance and reactance format. This corresponds to a load type ‘4’ in NEC [4].
Finally, the subroutine ‘finish’ is called to end the program cmstl. The NEC
command EN is written to the nec . in file, which identifies the end of the input for
the NEC program. The files are then closed, and the program is terminated. The

source code for the program cmstl is included in the next section.

126

A.3 Source Code for cmst1

The following is the Fortran source code for the program cmstl.
program cmstl

C************************************************

This program will generate a NEC4 input file for Chris Coleman’s
MS thesis project

The NEC input file generated is called ’nec.in’

latest revision: 5/20/99

0000000

c************************************************

call beginning
call table
call exite
call finish
end

subroutine beginning
c************************************************

c Subroutine ’beginning’ initializes the program. It

c specifies the frequencies, how the antenna terminals

c should be loaded, and what input and output files should
c be generated.

C************************************************

common/bbeginningl/minfreq,maxfreq,maxseg
common/bbeginningZ/spacing,wrad
common/btag/tag

common/gnd/gndflag

common/xqf/xqflag
common/frqstuff/freqflag,freqstep,freqnum
common/zpar_vars/zpar_flag

rea1*8 minfreq,maxfreq,freqstep,spacing,wrad

integer*4 tag,gndflag,xqflag,freqflag,freqnum,maxseg,
+ zpar_f1ag

127

00000

0000

’2par_flag’ will determine which voltages will be
applied and which segments will be shorted or Opened.
’2par_flag=1’ will be used to measure 221 and 212
’2par_flag=2’ will be used to measure 211
’zpar_flag=3’ will be used to measure 222

write(*,*) ’2par_f1ag=0 is a normal NEC run (300 Ohm load
+ at antenna termials)’
write(*,*) ’2par_flag=1 Applies V1, shorts port 2,
+ measures 221 and 212’
write(*,*) ’2par_f1ag=2 Applies V1, Opens port 2, and
+ measures 211’
write(*,*) ’2par_flag=3 Applies V2, opens port 1, and
+ measures 222’

write(*,*) ’ ’
write(*,*) ’Enter 2par_f1ag NOW!’
read(*,*) 2par_flag

’2par_f1ag’ is displayed on the screen so that the user
can check that the program does have the correct value.
Then an explanation of what type of operation will be made
for the particular choice of 2par_flag.
write(*.*) ’zpar_flag=’,2par_f1ag
if(2par_flag.eq.0) then
write(*,*) ’normal NEC run (300 Ohm load
+ at antenna termials)’
else if(2par_flag.eq.1) then
write(*,*) ’ Applies V1, measures 221 and 212’
else if(2par_flag eq.2) then
write(*,*) ’ Applies V1, opens port 2, and
+measures 211’
else if(2par_flag.eq.3) then
write(*,*) ’ Applies V2, opens port 1, and
+measures 222’
end if

Now the program will set values for geometries, frequencies,
and excitations

’gndflag=1’ puts a ground plane under the table.
This flag will be used by another subroutine

gndflag=1

The flag xqflag=1 will exite the table with a plane wave

128

IIDO
11()

xqflag=2 will exite the table with current from
the wire harness
xqflag=2

these next two variables specify the minimum and
maximum frequencies (in H2)

minfreq=50000000.

maxfreq=150000000.

’freqflag=1’ will choose frequency stepping, and
’freqflag.ne.1’ will not use frequency stepping
freqflag=1

The freqstep variable specifies the frequency stepping
increment (in MHz), and freqnum specifies the number
of points

freqnum=101

freqstep=(maxfreq-minfreq)/(freqnum-1)

’maxseg’ is the maximum number of segments
maxseg=3000

This ’spacing’ variable specifies the maximum allowable
grid spacing (lambda over 10). All grid spacing will be
equal or smaller than this

spacing=.1*(3.E+8 / maxfreq)

’wrad’ is the wire radius
wrad=.001

’tag’ is an important variable
used to keep track of the segment numbers
tag=1

Now ’nec.in’ will be opened. This is the input file
to NBC.

open(10,file=’nec.in’,status=’unknown’)

The following are just comments written for the NEC
input code

write(10,100)

write(10,110)

format("cm NEC input file for CISPR table")
format(’ce Written by Chris Coleman’)

’nec.par’ will be used to tell NEC4 information about

129

120

the wire tag numbers used for x and y directed segments
open(20,file=’nec.par’,status=’unknown’)

’2parinfo.par’ will be used by a z-parameter computation
program

Open(30,file=’2parinfo.par’,status=’unknown’)
write(30,120) minfreq, maxfreq, freqnum

write(30,*) ’minfreq, maxfreq, freqnum’
format(f13.0,f13.0,15)

return
and

C************************
C************************

subroutine table

C********
c subroutine table is the new table drawing routine
C********
rea1*8 1(50).
+ tabheight,wireh,
+ l_ant_ov2,h_r_ant,
+ minfreq,maxfreq,
+ spacing,wrad,
+ coordinates(6,5000),spac(50)
integer*4 tag,maxseg,nseg,ilist(50),wiretag(10).
+ xtmin,xtmax,ytmin,ytmax,2tmin,2tmax,otmin,otmax,
+ termina1_tag(50)
common/bbeginningl/minfreq,maxfreq,maxseg
common/bbeginning2/spacing,wrad
common/btag/tag
common/harness/wiretag
c ’nseg’ is the number of segments per wire
nseg=1
c ’wireh’ is the height of the wire above the table
wireh=.05
tabheight=.9
c ’1_ant_ov2’ is the length of the receiving antenna over 2,
c and ’h_r_ant’ is the height of the receiving antenna

1_ant_ov2=.75
h_r_ant=1

130

c These are the physical dimensions of the table in meters
1(1)=2.44
1(2)=1.22
1(7)=1.5
1(3)=(1(1)-1(7))/2
1(4)=1(1)-(l(3)+1(7))
1(5) is the distance the wire radiator lies back from the
table lip
1(5)=.05
1(6)=1(2)-1(5)
l(8)=l(1)/2
1(9)=1(2)+1
1(10)=h_r_ant
l(11)=l_ant_ov2

C******************************************
c*******This is the beginning of the x-directed wire section

xtmin=tag

c This section draws the

c x-directed back section 1 of the table

c ilist(1) is the number of points in the y-direction
ilist(1)=int(l(6)/spacing)+1

c ilist(2) is the number of points in the x-direction

ilist(2)=int(1(3)/spacing)+1

spac(1)=1(6)/ilist(1)

spac(2)=1(3)/ilist(2)

do j=1,ilist(1)

do i=1,ilist(2)
coordinates(i,tag)=spac(2)*(i-1)
coordinates(4,tag)=spac(2)*i
coordinates(2,tag)=spac(1)*(j-1)
coordinates(5,tag)=coordinates(2,tag)
coordinates(3,tag)=tabheight
coordinates(6,tag)=tabheight

tag=tag+1
end do
end do
c This section draws the

0

x-directed back section 2 of the table

131

ilist(l) is the number of points in the y-direction
ilist(3) is the number of points in the x-direction
ilist(3)=int(l(7)/spacing)+1
spac(3)=l(7)/ilist(3)
do j=1,ilist(1)
do i=1,ilist(3)
coordinates(1,tag)=1(3)+spac(3)*(i-1)
coordinates(4,tag)=l(3)+spac(3)*i
coordinates(2,tag)=spac(1)*(j-1)
coordinates(5,tag)=coordinates(2,tag)
coordinates(3,tag)=tabheight
coordinates(6,tag)=tabheight
tag=tag+1
end do
end do

This section draws the
x-directed back section 3 of the table

ilist(l) is the number of points in the y-direction

ilist(4) is the number of points in the x-direction

ilist(4)=int(l(4)/spacing)+1

spac(4)=l(4)/ilist(4)

do j=1,ilist(1)

do i=1,ilist(4)
coordinates(1,tag)=l(1)-1(4)+spac(4)*(i-l)
coordinates(4,tag)=1(1)-l(4)+spac(4)*i
coordinates(2,tag)=spac(1)*(j-1)
coordinates(5,tag)=coordinates(2,tag)
coordinates(3,tag)=tabheight
coordinates(6,tag)=tabheight
tag=tag+1

end do

end do

This section draws the
x-directed back section 4 of the table

ilist(5) is the number of points in the y-direction
ilist(5)=int(1(5)/spacing)+1

ilist(2) is the number of points in the x-direction
Spac(5)=1(5)/ilist(5)

132

do j=1,ilist(5)+1

do i=1,

end do
end do

ilist(2)
coordinates(1,tag)=spac(2)*(i-1)
coordinates(4,tag)=spac(2)*i
coordinates(2,tag)=l(6)+spac(5)*(j-l)
coordinates(5,tag)=coordinates(2,tag)
coordinates(3,tag)=tabheight
coordinates(6,tag)=tabheight
tag=tag+1

This section draws the
x-directed back section 5 of the table

ilist(S) is the number of points in the y-direction
ilist(3) is the number of points in the x-direction
do j=1,i1ist(5)+1

do i=1,

end do
end do

ilist(3)
coordinates(l,tag)=1(3)+spac(3)*(i-1)
coordinates(4,tag)=l(3)+spac(3)*i
coordinates(2,tag)=l(6)+spac(5)*(j-1)
coordinates(S,tag)=coordinates(2,tag)
coordinates(3,tag)=tabheight
coordinates(6,tag)=tabheight
tag=tag+1

This section draws the
x-directed back section 4 of the table

ilist(S) is the number of points in the y-direction
ilist(4) is the number of points in the x-direction
do j=1,i1ist(5)+1

do i=1,

ilist(4)
coordinates(l,tag)=1(1)-l(4)+spac(4)*(i-1)
coordinates(4,tag)=1(1)-1(4)+spac(4)*i
coordinates(2,tag)=1(6)+spac(5)*(j-1)
coordinates(5,tag)=coordinates(2,tag)
coordinates(3,tag)=tabheight
coordinates(6,tag)=tabheight

tag=tag+1

133

end do
end do

This is the end of the x-directed section
xtmax=tag

c*1H:***************************************
c*******This is the beginning of the y-directed wire section

00

0000

0000

ytmin=tag

This section draws the
y-directed back section 1 of the table

ilist(1)=int(1(6)/spacing)+1 from the x-directed section 1
spac(1)=l(6)/ilist(1) from the x-directed section 1
ilist(2)=int(l(3)/spacing)+1 from the x-directed section 1
spac(2)=1(3)/ilist(2) from the x-directed section 1
do i=1,ilist(2)
do j=1,ilist(1)
coordinates(l,tag)=spac(2)*(i-1)
coordinates(4,tag)=coordinates(1,tag)
coordinates(2,tag)=spac(1)*(j-1)
coordinates(5,tag)=spac(1)*j
coordinates(3,tag)=tabheight
coordinates(6,tag)=tabheight
tag=tag+1
end do
end do

This section draws the
y-directed back section 2 of the table

ilist(1)=int(l(6)/spacing)+1 from the x-directed section 1

spac(1)=l(6)/ilist(1) from the x-directed section 1

ilist(3)=int(l(7)/spacing)+1 from the x-directed section 2

spac(3)=l(7)/ilist(3) from the x-directed section 2

do i=1,ilist(3)

do j=1,ilist(1)
coordinates(l,tag)=l(3)+spac(3)*(i-1)
coordinates(4,tag)=coordinates(1,tag)
coordinates(2,tag)=spac(1)*(j-1)
coordinates(5,tag)=spac(1)*j

134

0

0000

0000

end do
end do

coordinates(3,tag)=tabheight
coordinates(6,tag)=tabheight
tag=tag+1

This section draws the
y-directed back section 3 of the table

ilist(1)=int(l(6)/spacing)+1 from the x-directed section 1
spac(1)=1(6)/ilist(1) from the x-directed section 1
ilist(4)=int(l(4)/spacing)+1 from the x-directed section 3
spac(4)=1(4)/ilist(4) from the x-directed section 3

do i=1,ilist(4)+1

do j=1,

and do
end do

ilist(l)
coordinates(l,tag)=l(1)-l(4)+spac(4)*(i-l)
coordinates(4,tag)=coordinates(1,tag)
coordinates(2,tag)=spac(1)*(j-l)
coordinates(5,tag)=spac(1)*j
coordinates(3,tag)=tabheight
coordinates(6,tag)=tabheight

tag=tag+1

This section draws the
y-directed front section 4 of the table

ilist(5)=int(l(5)/spacing)+1 from the x-directed section 4
spac(5)=l(5)/ilist(5) from the x-directed section 4
ilist(2)=int(l(2)/spacing)+1 from the x-directed section 1
spac(2)=1(4)/ilist(2) from the x-directed section 1

do i=1,ilist(2)

do j=1,

and do
end do

ilist(S)
coordinates(l,tag)=spac(2)*(i-1)
coordinates(4,tag)=coordinates(1,tag)
coordinates(2,tag)=l(6)+spac(5)*(j-l)
coordinates(5,tag)=1(6)+spac(5)*j
coordinates(3,tag)=tabheight
coordinates(6,tag)=tabheight
tag=tag+1

135

0000

0000

This section draws the
y-directed front section 5 of the table

ilist(5)=int(1(5)/spacing)+1 from the x-directed section 4
spac(5)=l(5)/ilist(5) from the x-directed section 4
i1ist(3)=int(1(3)/spacing)+1 from the x-directed section 3
spac(3)=l(3)/ilist(3) from the x-directed section 3
do i=1,i1ist(3)
do j=1,i1ist(5)
coordinates(1,tag)=l(3)+spac(3)*(i-1)
coordinates(4,tag)=coordinates(1,tag)
coordinates(2,tag)=1(6)+spac(5)*(j-l)
coordinates(5,tag)=l(6)+spac(5)*j
coordinates(3,tag)=tabheight
coordinates(6,tag)=tabheight
tag=tag+1
end do
end do

This section draws the
y-directed front section 6 of the table

ilist(5)=int(1(5)/spacing)+1 from the x-directed section 4
spac(5)=l(5)/ilist(5) from the x-directed section 4
ilist(4)=int(l(4)/spacing)+1 from the x-directed section 3
spac(4)=l(4)/ilist(4) from the x-directed section 3
do i=1,ilist(4)+1
do j=1,ilist(5)
coordinates(l,tag)=1(1)-1(4)+spac(4)*(i-1)
coordinates(4,tag)=coordinates(1,tag)
coordinates(2,tag)=l(6)+8pac(5)*(j-1)
coordinates(5,tag)=l(6)+spac(5)*j
coordinates(3,tag)=tabheight
coordinates(6,tag)=tabheight
tag=tag+1
end do
end do

This is the end of the y-directed section
ytmax=tag

136

c******************************************
c*******This is the beginning of the z-directed wire section
2tmin=tag

c This is the end of the 2-directed section

ztmax=tag

C******************************************
c*******This is the beginning of the other-directed wire section

otmin=tag
c This section draws the x-directed wire radiator
c This line sets the end wire of the harness to be exited

do i=1,ilist(3)*10
coordinates(l,tag)=1(3)+spac(3)*(i-1)/10
coordinates(4,tag)=1(3)+spac(3)*i/1O
coordinates(2,tag)=1(6)
coordinates(5,tag)=coordinates(2,tag)
coordinates(3,tag)=tabheight+wireh
coordinates(6,tag)=coordinates(3,tag)
tag=tag+1

end do

c This section draws the z-directed wire radiator
ilist(6)=int(10*wireh/spacing)+1
spac(6)=wireh/i1ist(6)
wiretag(1)=tag+ilist(6)/2
do i=1,ilist(6)
coordinates(1,tag)=l(3)
coordinates(4,tag)=coordinates(1,tag)
coordinates(2,tag)=1(6)
coordinates(5,tag)=coordinates(2,tag)
coordinates(3,tag)=tabheight+spac(6)*(i-1)
coordinates(6,tag)=tabheight+spac(6)*i
tag=tag+1

end do

c This section draws the 2-directed wire radiator
wiretag(2)=tag+ilist(6)/2
do i=1,ilist(6)
coordinates(1,tag)=1(1)-l(4)
coordinates(4,tag)=coordinates(1,tag)
coordinates(2,tag)=1(6)

137

coordinates(5,tag)=coordinates(2,tag)
coordinates(3,tag)=tabheight+spac(6)*(i-1)
coordinates(6,tag)=tabheight+spac(6)*i
tag=tag+1

end do

c*******This section draws the dipole antenna
ilist(7)=20
l(12)=l(11)/(2*ilist(7)+1)
spac(7)=(1(11)-l(12))/ilist(7)

c This next point defines the feed point
coordinates(l,tag)=1(8)-1(12)
coordinates(4,tag)=l(8)+1(12)
coordinates(2,tag)=1(9)
coordinates(S,tag)=coordinates(2,tag)
coordinates(3,tag)=l(10)
coordinates(6,tag)=l(10)
wiretag(3)=tag

tag=tag+1
c Now some of the important tag locations will be written
c to the screen

write(*,*) ’Port 1 (wiretag(1))=’, wiretag(l)
write(*,*) ’Location of 50 Ohm load (wiretag(2))=’, wiretag(2)
write(*,*) ’Port 2 (wiretag(3))=’, wiretag(3)

do i=1,ilist(7)
coordinates(l,tag)=l(8)-l(12)-(i-1)*spac(7)
coordinates(4,tag)=l(8)-1(12)-i*spac(7)
coordinates(2,tag)=1(9)
coordinates(5,tag)=coordinates(2,tag)
coordinates(3,tag)=l(10)
coordinates(6,tag)=coordinates(3,tag)
tag=tag+1

end do

do i=1,ilist(7)
coordinates(l,tag)=1(8)+l(12)+(i-1)*spac(7)
coordinates(4,tag)=l(8)+1(12)+i*spac(7)
coordinates(2,tag)=1(9)
coordinates(5,tag)=coordinates(2,tag)
coordinates(3,tag)=l(10)

138

coordinates(6,tag)=coordinates(3,tag)
tag=tag+1
end do

c this is the end of the other-directed section
otmax=tag

C***************************************
c*******This part writes out the wires from the coordinate matrix
c to the nec input file

do i=1,tag-1
write(10,100) ’gw’,i,nseg,coordinates(1,i),
+ coordinates(2,i),coordinates(3,i),coordinates(4,i),
+ coordinates(5,i),coordinates(6,i),wrad
end do

C***************************************
c*******This is the ’nec.par’ writing section

c write to the nec.par file
write(20,*) xtmin
write(20,*) xtmax
write(20,*) ytmin
write(20,*) ytmax
write(20,*) ztmin
write(20,*) ztmax
write(20,*) otmin
write(20,*) otmax
write(20,*) tag-1
write(20,*) wiretag(1)
write(20,*) wiretag(3)

100 format(a2,i6,i3,f10.4,F10.4,F10.4,F10.4,F10.4,F10.4,F10.4)
return
end

C************************

139

 

C************************

subroutine exite

c********

C
C

subroutine exite is a function used
to write the exitation card

c********

+

integer*4 gndflag,xqflag,freqflag,freqnum,maxseg,
wiretag(10),2par_flag

real*8 minfreq,maxfreq,freqstep,maxflOO,minflOO,
xcurp,cmom,

tabheight
common/bbeginning1/minfreq,maxfreq,maxseg
common/gnd/gndflag

common/xqf/xqflag
common/frqstuff/freqflag,freqstep,freqnum
common/harness/wiretag
common/2par_vars/zpar_flag

These nex frequencies are made so that they are in MHz.
maxf100=maxfreq/1000000.
minf100=minfreq/1000000.

tabheight is the table height
tabheight=.9

The geometry ends with the ’ge’ command
if(gndflag.eq.1) then
write(10,200) ’ge -1’
else
write(10,100) ’ge’
end if

if(xqflag.ne.0) then
if(freqflag.eq.1) then
write(10,105) ’fr’,’0’,freqnum,’0’,’O’,minf100,
freqstep/1000000.
else
write(10,110) ’fr’,’0’,’1’,’O’,’0’,maxf100
end if
end if

This section loads the

wire radiator at the end of the wire
write(10,140) ’1d’,’4’,wiretag(2),’1’,’1’,’50.’,’O’

140

 

These next lines load the

dipole receiving antenna at the center

The loading is dependent on

which z-parameter is being calculated

if(2par_flag.eq.0) then

write(10,140) ’ld’,’4’,wiretag(3),’1’,’1’,’300.’,’O’

end if '
if(2par_flag.eq.2) then

write(10,140) ’ld’,’4’,wiretag(3),’1’,’1’,’9999999999.’,’O’
end if

if(2par_flag.eq.3) then

write(10,140) ’ld’,’4’,wiretag(1),’1’,’1’,’9999999999.’,’O’
end if

0000

if(xqflag.eq.1) then
write(10,120) ’ex’,’1’,’1’,’1’,’0’,’0.’,’0.’,’0.’
end if

c************************************************
c This section will generate the voltage exitation on the
c wire harness (if xqflag=2)
if(xqflag.eq.2) then
if(2par_flag.eq.0) then
write(10,121) ’ex’,’0’,wiretag(1),’1’,’0’,’1.’
end if
if(2par_flag.eq.1) then
write(10,121) ’ex’,’0’,wiretag(1),’1’,’0’,’1.’
end if
if(2par_flag.eq.2) then
write(10,121) ’ex’,’0’,wiretag(1),’1’,’0’,’1.’
end if
if(2par_flag.eq.3) then
write(10,121) ’ex’,’0’,wiretag(3),’1’,’O’,’1.’
end if
end if

if(gndflag.eq.1) then

c This statement creates the ground plane under the table.
write(10,210) ’gn’,’1’
end if

141

100
105
110
120
121
125
130
140
200
210

if(xqflag.ne.0) then

NEC uses this statement as a signal to calculate the currents.
write(10,130) ’xq’

end if

format(a2)

format(a2,a6,i8,a8,a8,f8.2,f8.2)
format(a2,a6,a8,a8,a8,f8.2)
format(a2,a6,a8,a8,a8,a8,a8,a8)
format(a2,a6,i8,a8,a8,a8)
format(a2,a6,a6,a6,a6,f9.4,f9.4,f9.4,a6,a6,f9.4)
format(a2)

format(a2,a6,i8,a8,a8,a13,a8)

format(a8)

format(a2,a6)

return
and

C************************
C************************

subroutine finish

c********

C

subroutine finish completes the program

c********

100

write(10,100) ’en’
close(10)
close(20)
close(30)
format(a2)

return

end

142

APPENDIX B

ADDITIONAL FORTRAN PROGRAMS

B.1 Overview

As well as the grid-generating program cmstl discussed in Appendix A, additional
Fortran programs were written for use in this project. Three important programs,
zparmake, sparconv and sphase will be discussed in this appendix. The Fortran

source code for the three programs will also be included.

B.2 Z-Parameter Calculation Program

The program zparmake was written to take the input files cur.p121.txt,
cur_p221.txt, cur-p122.txt, and cur.p223.txt created from NBC during various
NEC runs, and use the data to compute the Z-parameters of the table-antenna sys-
tem. The equations governing the relations between the measured currents and the
Z -parameters were discussed in Chapter 3.

The subroutine ‘readfiles’ is called for the first step performed by zparmake. In this
subroutine all of the input and output files are opened and variables are initialized.
The input files are read and the currents are stored in complex arrays. These arrays
are passed to the next subroutine.

Next, the subroutine ‘zmake’ processes the raw currents stored in the arrays found
by the subroutine ‘readfiles’. The Z -parameters are computed in this subroutine using
the relations discussed in Section 3.4. These Z -parameters are then written to the files
211_num.txt, 221.num.txt and 222_num.txt. Finally, the subroutine ‘finish’ closes
all the files and finishes the program. The source code for the program zparmake is

included in the next section.

143

3.3 Source Code for zparmake

The following is the source code for the program zparmake.

program zparmake
c************************************************
This program will make z-parameters from measured currents
taken from NBC output

It is assumed that the applied voltage for either
port 1 or port 2 is 1 Volt

0000000

This program was last modified on 5/21/99
C************************************************

call readfiles
call zmake
call finish
end

subroutine readfiles
c************************************************

c This subroutine reads in the output files generated

c from the modified NEC program. The subroutine then

c puts them into matrices so that they can be manipulated
G into Z-parameters

C************************************************

common /matrix/ num_mat

common /freqs/ minfreq,maxfreq,nof

common /currents/ ncur_plzl,ncur_p122,ncur_p2zl,
+ ncur_p2z3

complex*16 num_mat(2,2,201),ncur_p121(201),
ncur-p122(201).

ncur_p123(201),

ncur_p2zl(201).

ncur_p223(201)

real*8 minfreq,maxfreq,trea1,timag

integer nof

++++

c Files 10,20,40,60 are the modified-NEC output files that
c contain information about the Z-parameters. File 70

144

(’Zparinfor.par’) contains frequency and number of data
point taken information.
open(10,file=’cur_plzl.txt’,status=’unknown’)
open(20,file=’cur_plz2.txt’,status=’unknown’)
Open(40,file=’cur_p2zl.txt’,status=’unknown’)
open(60,file=’cur_p223.txt’,status=’unknown’)
open(70,file=’zparinfo.par’,status=’unknown’)

Files 80-82 are the Z-parameter output files.
It is assumed that z21=212.
open(80,fi1e=’211_num.txt’,status=’unknown’)
open(81,file=’z21_num.txt’,status=’unknbwn’)
Open(82,file=’222_num.txt’,status=’unknown’)

Now the minimum and maximum frequencies, as well as the
number of frequency points will be read in from
’zparinfo.par’.

read(70,*) minfreq,maxfreq,nof

write(*,*) minfreq,maxfreq,nof

Now the NEC output files will be read
do i=1,nof

read(10,*) trea1,timag
ncur_p121(i)=dcmp1x(treal,timag)

end do

do i=1,nof

read(20,*) treal,timag
ncur_plz2(i)=dcmplx(treal,timag)
end do

do i=1,nof

read(40,*) treal,timag
ncur_p2zl(i)=dcmp1x(treal,timag)
end do

do i=1,nof
read(60,*) trea1,timag

ncur_p2z3(i)=dcmplx(treal,timag)
end do

return

145

end

subroutine zmake
c************************************************

c This subroutine takes the raw current from the
c NEC output and calculates the Z-parameters of the
c two port network.

C************************************************

common /matrix/ num_mat

common /freqs/ minfreq,maxfreq,nof

common /currents/ ncur_p121,ncur_plz2,ncur_p2zl,
+ ncur_p223

 

complex*16 num_mat(2,2,201),ncur_p121(201),
+ ncur_p122(201),ncur_p221(201),
+ ncur_p223(201),zi1(201),z21(201),
+ 222(201)
real*8 minfreq,maxfreq,Vapp,treal,timag
integer nof

 

c Vapp is the applied test voltage to the ports
Vapp=1

c Now the Z-parameters will be calculated.
do i=1,nof

211(i)=Vapp/ncur_plz2(i)
treal=drea1(211(i))
timag=dimag(211(i))
write(80,100) treal,timag
end do

do i=1,nof
z22(i)=Vapp/ncur_p223(i)
treal=dreal(z22(i))
timag=dimag(z22(i))
write(82,120) treal,timag
end do

 

do i=1,nof .
z21(i)=-z22(i)*ncur_p221(i)/ncur_p1zl(i)
treal=drea1(z21(i))
timag=dimag(z21(i))
write(81,110) treal,timag

146

end do

100 format(f9.4,’ ’,f9.4)
110 format(f10.5,’ ’,f10.5)
120 format(f10.4,’ ’,f10.4)
return
end

subroutine finish
c************************************************

c This subroutine will close all the files.
C************************************************

close(10)
close(20)
close(40)
close(60)
close(70)
close(80)
close(81)
close(82)

return
end

147

 

B.4 Z-Parameter to S-Parameter Conversion and Data Processing Pro-

gram

The program sparconv is another program used extensively for this project. After
measurements from the HP 8510C Network Analyzer were stored to disk and the
theoretical Z-parameters were calculated, the program sparconv was used to for-
mat all of the measured and numerical data so that it could be displayed using the
commercial plotting program Axum.

The program sparconv begins by calling the ‘readfiles’ subroutine. This subrou-
tine opens all of the input and output files that will be used by the program. The
input files that are opened are the Z—parameter output files created by the zparmake
program and an S-parameter measurement file created by the HP 8510C. The mea-
sured file created by the HP 85100 contains information about the measurement
performed (date, time, number of points, starting and stopping frequencies, format
of stored data, etc.) followed by the four S-parameters separated by text in real and
imaginary format. The subroutine then reads in the S—parameters and stores them
in a matrix.

Next, the subroutine ‘stoz’ is called. This subroutine calculates the Z-parameters
from the measured S-parameters and writes them to the file measzpar.txt. Since
S-parameters were chosen as a format to compare measured and theoretical data, the
measured Z-parameters are not included in this thesis.

The subroutine ‘ztos’ is called next. This subroutine reads in all of the theoretical
Z -parameters stored in the output files created by the program zparmake and converts
them to S-parameters. The theoretical S-parameters are then stored in a matrix to
be used in another subroutine.

The subroutine ‘data_output’ is then called to write the measured and theoretical

S-parameters to eight separate output files (four for the measured parameters and

148

four for the theoretical parameters) in magnitude and phase format. Before the
parameters are written to the output files, the phase is unwrapped so that it appears
more continuous. In fact, another short Fortran program was written to further
unwrap the phase of the S-parameters. This program, sphase, is discussed in the
Section B.6. After the ‘data_output’ subroutine has written the output files, all of
the input and output files are closed and the sparconv program ends. The source

code for this program is included in the next section.

149

 

B.5 Source Code for sparconv

The following is the source code for the program sparconv.

program sparconv
c************************************************

c this program will read in S-parameter measurements

c and Z-parameter numerical calculations, convert the numerical
c Z-parameters to S-parameters, and output the results into

c several files so that they can be easily plotted using Axum.
c This program was last modified on 5/19/99

C************************************************

call readfiles
call stoz

call ztos

call data_output
end

subroutine readfiles
c************************************************

c This subroutine will read in the measured S-parameters
c and put them into a 2 by 2 by nop matrix where nop is the
c number of points

C************************************************

common /matrix/ meas_matrix,nop
common /freqs/f_min,f_max
common /meas_s/ms

rea1*8 s_real,s_imag,f_min,f_max

integer readflag,nop,endflag

complex*16 meas_matrix(2,2,5000),ms(2,2,5000)
character btemp*80, filename*80

write(*,*) ’Enter measured S-parameter filename’
read(*,*) filename

c Now all of the files that are needed will be opened.
c File 10 is the measured S-parameter file taken from
c the HP 8510c Network analyzer.

open(10,file=filename,status=’unknown’)

c Files 11-13 are the NEC-calculated Z-parameters (221:212)
open(11,file=’211_num.txt’,status=’unknown’)

150

 

00000

00000

Open(12,file=’z21_num.txt’,status=’unknown’)
open(13,file=’222_num.txt’,status=’unknown’)

Files 20 and 30 are measured S and Z parameters put in
files so that they could be easily displayed in Matlab.
Open(20,file=’measspar.txt’,status=’unknown’)
open(30,fi1e=’measzpar.txt’,status=’unknown’)

Files 31-34 are the measured S-parameters in a format where
they can be plotted using Axum. The first column in each file
is the frequency (in MHz), the second column is the magnitude
of the S-parameter (S21 and S12 are in dB), and the third
column is the phase.

open(31,file=’sii.txt’,status=’unknown’)
open(32,file=’s21.txt’,status=’unknown’)
open(33,file=’312.txt’,status=’unknown’)
open(34,file=’s22.txt’,status=’unknown’)

Files 35-38 are the NEC-calculated S-parameters in a format
where they can be plotted using Axum. The first column in each
file is the frequency (in MHz), the second column is the
magnitude of the S-parameter (S21 and 812 are in dB), and the
third column is the phase.
Open(35,file=’ns11.txt’,status=’unknown’)
Open(36,fi1e=’ns21.txt’,status=’unknown’)
open(37,file=’nsl2.txt’,status=’unknown’)
open(38,file=’ns22.txt’,status=’unknown’)

readflag=0
This do loop looks in the measurement file for the letters
’SEG_LIST_BEGIN’ because frequency information will follow.
do while (readflag.eq.0)
read(10,*) btemp
if(btemp.eq.’SEG_LIST_BEGIN’) then
write(*,*) btemp
readflag=1
end if
end do

Now the minimum frequency, the maximum frequency, and the
number of points is read in. ’btemp’ is just a dummy
variable used to read in (but not use) a character string.
read(10,*) btemp,f_min,f_max,n0p

This next loop looks for the letters ’BEGIN’ so that it knows

151

 

100

when the S-parameter measurements start.
readflag=0
do while (readflag.eq.0)

read(10,*) btemp

if(btemp.eq.’BEGIN’) then

write(*.*) btemp

readflag=1

end if

end do

endflag=0
do j=1,2
do k=1,2
do i=1,nop
read(10,*) s_real,s_imag
meas_matrix(k,j,i) is the measured S_parameter matrix that
will be converted to a measured Z-parameter matrix for
use by other programs
meas_matrix(k,j,i)=dcmplx(s_real,s-imag)
’ms’ is the measured S-parameter matrix (S_kj matrix).
ms(k,j,i)=meas_matrix(k,j,i)
write(20,100) s-real,s_imag
end do
This extra read in of ’btemp’ is because there is character
information separating the S-parameter lists
read(10,*) btemp
if(btemp.ne.’END’) then
write(*,*) ’END not correctly reached’
end if
endflag=endf1ag+1
if(endflag.ne.4) then
read(10,*) btemp
end if
end do
end do

format(f14.10,’ ’,f14.10)
close(10)

close(20)

return

end

subroutine stoz

152

 

c************************************************
c this subroutine (pilfered from Rothwell’s 2port.for)

C converts s-parameters to z-parameters
c************************************************

common /matrix/ m,n0p

complex*16 m(2,2,5000),dummy(2,2,5000)
complex*16 den
integer nop

real*8 20
c 20 is the characteristic impedance of the two port network.
c It is assumed to be 50 Ohms for convenience.

20:50.

do i=1,4*nop
den=((1.-m(1,1,i))*(1.—m(2,2,i))-m(1,2,i)*m(2,1,i))/zO
dummy(1,1,i)=(1.+m(1,1,i))*(1.-m(2,2,i))+
+ m(1,2,i)*m(2,1,i)
dummy(1,2,i)=2.*m(1,2,i)
dummy(2,1,i)=2.*m(2,1,i)
dummy(2,2,i)=(1.-m(1,1,i))*(1.+m(2,2,i))+

+ m(1.2.i)*m(2,1.i)
c This next line checks the value of the denominator
c and then sets it to a small value if it is zero so that
c no divide by zero is performed.

if(cdabs(den).eq.0.) den=(1.e-20,0)

m(1,1,i)=dummy(1,1,i)/den

m(1,2,i)=dummy(1,2,i)/den

m(2,1,i)=dummy(2,1,i)/den

m(2,2,i)=dummy(2,2,i)/den
end do

return
end

subroutine ztos
c************************************************

c this subroutine takes the numerical z-parameters
c and converts them to s-parameters

153

 

C************************************************
common /freqs/f-min,f_max
common /matrix/m,nop
common /num_s/ns

real*8 f_min,f_max,z_rea1,z_imag

complex*16 m(2,2,5000),n(2,2,5000),ns(2,2,5000),
+ zO,delt_z

integer nop

c The characteristic impedance of the two port network is chosen
c to be 50 Ohms for convenience.
zO=dcmplx(50.,0.)

c This do loop will read in the numerical Z-parameters (’n’) and
c calculate the numerical S-parameters (’ns’)
do i=1,101

read(11,*) z_real,z_imag
n(1,1,i)=dcmplx(z_real,z_imag)
read(12,*) z_real,z_imag
n(2,1,i)=dcmplx(z_real,z_imag)
n(1,2,i)=n(2,1,i)

read(13,*) z_real,z_imag
n(2,2,i)=dcmplx(z_real,z_imag)

c now the numerical S-parameters (’ns’) will be calculated
de1t_z=(n(1,1,i)+zO)*(n(2,2,i)+zO)-n(1,2,i)*n(2,1,i)
ns(1,1,i)=((n(1,1,i)-zO)*(n(2,2,i)+zO)-n(1,2,i)*n(2,1,i))/

+ de1t_z
ns(2,1,i)=2*zO*n(2,1,i)/delt_z
ns(1,2,i)=ns(2,1,i)
ns(2,2,i)=((n(1,1,i)+zO)*(n(2,2,i)-zO)-n(1,2,i)*n(2,1,i))/
+ delt_z
end do

return
end

subroutine data_output
c************************************************

c this subroutine takes the z-parameters
c (converted from the measured s-parameters)
c and outputs them to the data file ’measzpar.txt’

C************************************************

154

 

0000

common /matrix/ meas_z,nop
common /num_s/ns

common /meas_s/ms

common /freqs/f_min,f-max

real*8 f_min,f_max,f,smag,sphase(5000),tmpphase,phasetol
complex*16 meas_z(2,2,5000),ns(2,2,5000),ms(2,2,5000)
integer nop

converting the frequency bounds so that they can be
read in MegaHertz

f_min=f_min/1.d6

f_max=f_max/1.d6

This do loop writes out the measured Z-parameters to the
file ’measzpar txt’
do j=1,2
do k=1,2
do i=1,nop
write(30,100) dreal(meas_z(k,j,i)).
dimag(meas_z(k,j,i))
end do
end do
end do

’phasetol’ and ’tmpphase’ are used in the phase unwrapping
section in the next do loop.

phasetol=250

tmpphase=0.

This next long do loop writes out the numerical and measured
S-parameters to the files 31-38. Column 1 of every file is
the frequency in MHz, column 2 is the magnitude (S21 and 812
are expressed in dB), and column 3 is the phase.
do j=1,2
do k=1,2
do i=1,nop
f=f_min+(i-1)*((f_max-f_min)/(n0p-1))
smag=abs(ms(k,j,i))

sphase(i)=(180/3.1415926)*atan2(dimag(ms(k,j,i)),
drea1(ms(k,j,i)))

155

This next ’if’ statement is so that the phase is ’unwrapped’
by testing whether the phase skips between -180 and 180 and
then forcing the phase to be continuous.
if(i.gt.1) then
tmpphase=sphase(i-1)
if(sphase(i)-tmpphase.gt.phasetol) then
sphase(i)=sphase(i)-360
else if(sphase(i)-tmpphase.1t.-phasetol) then
sphase(i)=sphase(i)+360
end if
end if

if(k.eq.1) then
if(j.eq.1) then
write(31,*) f,smag,sphase(i)
else .
write(33,*) f,20*log(smag),sphase(i) l
end if
else
if(j.eq.1) then
write(32,*) f,20*log(smag),sphase(i)

 

else
write(34,*) f,smag,sphase(i)
end if
end if
end do
do i=1,101

f=f_min+(i-1)*((f_max-f_min)/(101-1))

smag=abs(ns(k,j,i))

sphase(i)=(180/3.1415926)*atan2(dimag(ns(k,j,i)),
dreal(ns(k,j,i)))

This next ’if’ statement is so that the phase is ’unwrapped’
by testing whether the phase skips between -180 and 180 and
then forcing the phase to be continuous.
if(i.gt.1) then
tmpphase=sphase(i-1)
if(sphase(i)-tmpphase.gt.phasetol) then
sphase(i)=sphase(i)-360
write(*,*)
else if(sphase(i)-tmpphase.lt.-phasetol) then
sphase(i)=sphase(i)+360

156

100

writef*,*)
end if
end if

if(k.eq.1) then
if(j.eq.1) then
write(35,*) f,smag,sphase(i)
else
write(37,*) f,smag,sphase(i)

write(37,*) f,20*log(smag),sphase(i)

end if
else
if(j eq.1) then
write(36,*) f,smag,sphase(i)

write(36,*) f,20*log(smag),sphase(i)

else
write(38,*) f,smag,sphase(i)
end if
end if
write(35,*) f,smag,sphase(i)
end do
end do
end do

close(10)
close(11)
close(12)
close(13)
close(20)
close(30)
close(31)
close(32)
close(33)
close(34)
close(35)
close(36)
close(37)
close(38)

format(f14 8,’ ’,f14.8)

return
end

157

B.6 Phase Correction Program

The program sphase is a simple program written to help correct the phase of an S-
parameter. After the program sparconv has been run, sometimes large sections of the
phase of the S-parameters appear to be 360 degrees below the rest of the phase. The
program sphase corrects this problem. The band of frequencies that needs correction
is first determined visually. Then the beginning and ending frequencies of this band
are given as input to the program sphase during the program’s operation. Then
sphase then simply adds 360 degrees to the phase of the S-parameter in the band of
frequencies specified. The program then rewrites the entire S-parameter file with the

corrected phase. The source code for this program is included in the next section.

158

 

3.7 Source Code for sphase

The following is the source code for the program sphase.

program sphase
C************************************************

c This program will help unwrap the phase of

c S-parameters by adding 360 degrees to certain
c sections of the phase.

c This program was last modified on 5/20/99

C************************************************

real*8 f(5000),smag(5000),sp(5000),f_min,f_max, '
+ f_beg,f_end E

integer nop,i,ibeg,iend -

character filename*80

 

n0p=801

write(*,*) ’Enter measured S-parameter filename
+ to be unwrapped’

read(*,*) filename

open(10,file=filename,status=’unknown’)

write(*,*) ’Enter begin frequency in MHz’
read(*,*) f_beg

write(*,*) ’Enter end frequency in MHz’
read(*,*) f_end

c These do statements find the frequency at which the phase
c modification will begin, and then it will add 360 degrees
c to the phase until the end frequency is reached. Then
c the rest of the S-parameters will be read in so they can
c be rewritten in a later loop.

readflag=0

i=1

do while (readflag.eq.0)
read(10,*) f(i),smag(i),sp(i)
if(f(i).gt.f_beg) then
ibeg=i
readflag=1
end if

159

i=i+1
end do
write(*,*) ibeg,f(i-1)

readflag=0
do while (readflag.eq.0)
read(10,*) f(i),smag(i),sp(i)
Sp(i)=sp(i)+360
if(f(i).gt.f-end) then
iend=i
readflag=1
end if
i=i+1
end do

do i=iend+1,nop
read(10,*) f(i),smag(i),sp(i)
end do

close(lo, status=’delete’)
open(20,file=filename,status=’unknown’)

The S-parameters will now be re-written to the same file
with the modified phase.
do i=1,nop
write(20,*) f(i),smag(i),sp(i)
end do

close(20)
end

160

APPENDIX C

NEC INPUT CODE

C.1 NEC Input Code

As discussed in Appendix A, the program cmstl was used to generate an input file to
be used by the program N BC. This input file, nec . in, contains all of the information
pertaining to wire locations and radii, voltage excitations, loads, the ground plane,
and frequency data.

The following pages contain the N EC input code “nec . in”. This NEC input file
was generated to produce an Open circuit located at port 1, and a voltage source
applied at port 2. The output created from N EC using this input file will be used to
calculate the Z-parameters of the table / antenna system using the method discussed

in Chapter 3.

161

°2°3°2°2°2°2°3°2°2°3°3°2°2°2°2°3°3°2°2°3°3°2°2§°2°2°2°2§°2°2°2°2@fifififififififififififififi"2°20“:8S

Written by Chris Coleman
.0000
.1567
.3133
.0000
.1567
.3133
.0000
.1567
.3133
.0000
.1567
.3133
.0000
.1567
.3133
.0000
.1567
.3133
.4700
.6575
.8450
.0325
.2200
.4075
.5950
.7825
.4700
.6575
.8450
.0325
.2200
.4075
.5950
.7825
.4700
.6575
.8450
.0325
.2200
.4075
.5950
.7825
.4700
.6575
.8450
.0325
.2200
.4075
.5950
.7825
.4700
.6575

mflmmthD-b

(0

NEC input file for CISPR table

1

HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHp—bt—s»HHHHHHHHHHHHHHHH

OOHHHHHOOOHHHHHOOOHHHHHOOOHHHHHOOOCOOOOOOOOOOO000000

000000000OOOOOOOOOOOOOOOOOO0000000000000000000000COO

.0000
.0000
.0000
.1867
.1867
.1867
.3733
.3733
.3733
.5600
.5600
.5600
.7467
.7467
.7467
.9333
.9333
.9333
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.1867
.1867
.1867
.1867
.1867
.1867
.1867
.1867
.3733
.3733
.3733
.3733
.3733
.3733
.3733
.3733
.5600
.5600
.5600
.5600
.5600
.5600
.5600
.5600
.7467
.7467

OCOOCOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOCOOOOOOOOOOOOOOOO

.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000

162

OOHHHHHHOOHHHH-HHOOHHHHHHOOHHHHHHOOOooooooooooooooooo

.1567
.3133
.4700
.1567
.3133
.4700
.1567
.3133
.4700
.1567
.3133
.4700
.1567
.3133
.4700
.1567
.3133
.4700
.6575
.8450
.0325
.2200
.4075
.5950
.7825
.9700
.6575
.8450
.0325
.2200
.4075
.5950
.7825
.9700
.6575
.8450
.0325
.2200
.4075
.5950
.7825
.9700
.6575
.8450
.0325
.2200
.4075
.5950
.7825
.9700
.6575
.8450

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

.0000
.0000
.0000
.1867
.1867
.1867
.3733
.3733
.3733
.5600
.5600
.5600
.7467
.7467
.7467
.9333
.9333
.9333
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.1867
.1867
.1867
.1867
.1867
.1867
.1867
.1867
.3733
.3733
.3733
.3733
.3733
.3733
.3733
.3733
.5600
.5600
.5600
.5600
.5600
.5600
.5600
.5600
.7467
.7467

O00000000000OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000

OOOOOOOOOOOOOOOOCOCOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010

55222253033229.2222?”3822229229.?‘fi‘fififififififififi‘fi‘fifififi22222222522

53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106

HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH

i-‘HHPHOOCHHHHHoooooooooMMHMMHMMHMMHNNHMMHHHHHHOOOHHHHHO

.8450
.0325
.2200
.4075
.5950
.7825
.4700
.6575
.8450
.0325
.2200
.4075
.5950
.7825
.9700
.1267
.2833
.9700
.1267
.2833
.9700
.1267
.2833
.9700
.1267
.2833
.9700
.1267
.2833
.9700
.1267
.2833
.0000
.1567
.3133
.0000
.1567
.3133
.4700
.6575
.8450
.0325
.2200
.4075
.5950
.7825
.4700
.6575
.8450
.0325
.2200
.4075
.5950
.7825

HHHHHHHHHHHHHHHHHHHHHHOOOOOOOOOOOOOOOOOO00000000000000

.7467
.7467
.7467
.7467
.7467
.7467
.9333
.9333
.9333
.9333
.9333
.9333
.9333
.9333
.0000
.0000
.0000
.1867
.1867
.1867
.3733
.3733
.3733
.5600
.5600
.5600
.7467
.7467
.7467
.9333
.9333
.9333
.1200
.1200
.1200
.2200
.2200
.2200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200

COOOOOOOOOOOOOOOOOOOOOOOOOOOOO000000000000000000000000

.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000

163

HHHHHHCOHHHHHHOOOOOOOONNNNMMMNNMNNNMNNNNHHHHHHOCHHHHHH

.0325
.2200
.4075
.5950
.7825
.9700
.6575
.8450
.0325
.2200
.4075
.5950
.7825
.9700
.1267
.2833
.4400
.1267
.2833
.4400
.1267
.2833
.4400
.1267
.2833
.4400
.1267
.2833
.4400
.1267
.2833
.4400
.1567
.3133
.4700
.1567
.3133
.4700
.6575
.8450
.0325
.2200
.4075
.5950
.7825
.9700
.6575
.8450
.0325
.2200
.4075
.5950
.7825
.9700

HHHHHHHHHHHHHHHHHHHHHHOOOOOOOOOOOOOOOOOO00000000000000

.7467
.7467
.7467
.7467
.7467
.7467
.9333
.9333
.9333
.9333
.9333
.9333
.9333
.9333
.0000
.0000
.0000
.1867
.1867
.1867
.3733
.3733
.3733
.5600
.5600
.5600
.7467
.7467
.7467
.9333
.9333
.9333
.1200
.1200
.1200
.2200
.2200
.2200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200

COOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010

92222522225222132233°:"2220322332fi‘fifi‘fififi‘fififififififi‘fi‘fifififi‘fi582°22‘92”;

107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160

I-‘HHHHHHHHHHHHHHHHHHPHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH

HHHHHHHHHHPHOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOMMHMMH

.9700
.1267
.2833
.9700
.1267
.2833
.0000
.0000
.0000
.0000
.0000
.0000
.1567
.1567
.1567
.1567
.1567
.1567
.3133
.3133
.3133
.3133
.3133
.3133
.4700
.4700
.4700
.4700
.4700
.4700
.6575
.6575
.6575
.6575
.6575
.6575
.8450
.8450
.8450
.8450
.8450
.8450
.0325
.0325
.0325
.0325
.0325
.0325
.2200
.2200
.2200
.2200
.2200
.2200

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOHHHHHH

.1200
.1200
.1200
.2200
.2200
.2200
.0000
.1867
.3733
.5600
.7467
.9333
.0000
.1867
.3733
.5600
.7467
.9333
.0000
.1867
.3733
.5600
.7467
.9333
.0000
.1867
.3733
.5600
.7467
.9333
.0000
.1867
.3733
.5600
.7467
.9333
.0000
.1867
.3733
.5600
.7467
.9333
.0000
.1867
.3733
.5600
.7467
.9333
.0000
.1867
.3733
.5600
.7467
.9333

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO0000000000000

.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000

164

HHHHHHHHHHHHOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOMNNMMM

.1267
.2833
.4400
.1267
.2833
.4400
.0000
.0000
.0000
.0000
.0000
.0000
.1567
.1567
.1567
.1567
.1567
.1567
.3133
.3133
.3133
.3133
.3133
.3133
.4700
.4700
.4700
.4700
.4700
.4700
.6575
.6575
.6575
.6575
.6575
.6575
.8450
.8450
.8450
.8450
.8450
.8450
.0325
.0325
.0325
.0325
.0325
.0325
.2200
.2200
.2200
.2200
.2200
.2200

I-hOOOOOHOOOOOHOOOOOHOOOOOHOOOOOHOOOOOHOOOOOHOOOOOHHHHHH

.1200
.1200
.1200
.2200
.2200
.2200
.1867
.3733
.5600
.7467
.9333
.1200
.1867
.3733
.5600
.7467
.9333
.1200
.1867
.3733
.5600
.7467
.9333
.1200
.1867
.3733
.5600
.7467
.9333
.1200
.1867
.3733
.5600
.7467
.9333
.1200
.1867
.3733
.5600
.7467
.9333
.1200
.1867
.3733
.5600
.7467
.9333
.1200
.1867
.3733
.5600
.7467
.9333
.1200

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010

522522”225°22‘25”:59°22°2°2°2°2°2°2°2°3°2°2°2°2°2°2°§°2°2°2°2°2°2°288522552525553?

161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214

HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH

HHHHHHOOOOOONMNMMNMMMNMIOIOMKDMIOMHHHHMHHHHHHHHHHHHHHHHHHH

.4075
.4075
.4075
.4075
.4075
.4075
.5950
.5950
.5950
.5950
.5950
.5950
.7825
.7825
.7825
.7825
.7825
.7825
.9700
.9700
.9700
.9700
.9700
.9700
.1267
.1267
.1267
.1267
.1267
.1267
.2833
.2833
.2833
.2833
.2833
.2833
.4400
.4400
.4400
.4400
.4400
.4400
.0000
.1567
.3133
.4700
.6575
.8450
.0325
.2200
.4075
.5950
.7825
.9700

HHHHHHHHup»Hoooooooooooooooooooooooooooooooooooooooooo

.0000
.1867
.3733
.5600
.7467
.9333
.0000
.1867
.3733
.5600
.7467
.9333
.0000
.1867
.3733
.5600
.7467
.9333
.0000
.1867
.3733
.5600
.7467
.9333
.0000
.1867
.3733
.5600
.7467
.9333
.0000
.1867
.3733
.5600
.7467
.9333
.0000
.1867
.3733
.5600
.7467
.9333
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO00000

.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000

165

MHHHHHOOOOOOMNMMMMMNMNNMMMMNMMHHHHHHHHHHHHHHHHHHHHHHHH

.4075
.4075
.4075
.4075
.4075
.4075
.5950
.5950
.5950
.5950
.5950
.5950
.7825
.7825
.7825
.7825
.7825
.7825
.9700
.9700
.9700
.9700
.9700
.9700
.1267
.1267
.1267
.1267
.1267
.1267
.2833
.2833
.2833
.2833
.2833
.2833
.4400
.4400
.4400
.4400
.4400
.4400
.0000
.1567
.3133
.4700
.6575
.8450
.0325
.2200
.4075
.5950
.7825
.9700

HHHHHHHHHHHHHOOOOOHOOOOOHOOOOOHOOOOOHOOOOOHOOOOOHOOOOO

.1867
.3733
.5600
.7467
.9333
.1200
.1867
.3733
.5600
.7467
.9333
.1200
.1867
.3733
.5600
.7467
.9333
.1200
.1867
.3733
.5600
.7467
.9333
.1200
.1867
.3733
.5600
.7467
.9333
.1200
.1867
.3733
.5600
.7467
.9333
.1200
.1867
.3733
.5600
.7467
.9333
.1200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO0000000000000

.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000
.9000

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010

“22222259222":@222529222252855522255‘22°3°22°2°§°2§°2°2°2°2°2°2°2°2°2°2°2§

215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268

HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH

HHHHHHHHHHHHHHHHHHHh.Hh‘oooooooooooooooooooooooooooooMUN

.1267
.2833
.4400
.4700
.4888
.5075
.5263
.5450
.5638
.5825
.6013
.6200
.6388
.6575
.6763
.6950
.7138
.7325
.7513
.7700
.7888
.8075
.8263
.8450
.8638
.8825
.9013
.9200
.9388
.9575
.9763
.9950
.0138
.0325
.0513
.0700
.0888
.1075
.1263
.1450
.1638
.1825
.2013
.2200
.2388
.2575
.2763
.2950
.3138
.3325
.3513
.3700
.3888
.4075

HHHHI—‘HHHHHHHHF‘HHHHHHHHHHHHHHHHHHHHHHHPHHHHHHHHHHHHHHHH

.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO0000000000000OOOOOO

.9000
.9000
.9000
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500

166

HHHHHHHHHHHHMHHHHHHl-sHHHooooooooooooo000000000000OOOMMM

.1267
.2833
.4400
.4888
.5075
.5263
.5450
.5638
.5825
.6013
.6200
.6388
.6575
.6763
.6950
.7138
.7325
.7513
.7700
.7888
.8075
.8263
.8450
.8638
.8825
.9013
.9200
.9388
.9575
.9763
.9950
.0138
.0325
.0513
.0700
.0888
.1075
.1263
.1450
.1638
.1825
.2013
.2200
.2388
.2575
.2763
.2950
.3138
.3325
.3513
.3700
.3888
.4075
.4263

HHHHHHHHHHHHHHHHHHHHHHF‘HHHHFHHHHHHHHHHHHHHHHHHHHHHHHHH

.2200
.2200
.2200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

.9000
.9000
.9000
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500

OOOOOO00OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010

“322222222”322222222222222222222222222222222222222222222

269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322

HHHHHHHHHHHHHHHHHHHHHHHHt-LHHHHHHHHHHHHHHHHHHHHHHHHHHHHH

OOOOOOOOOOOOHHHHHHHHHHOOOHHHHHHHHHHHHHHHHHHHHHHHHHHHHH

.4263
.4450
.4638
.4825
.5013
.5200
.5388
.5575
.5763
.5950
.6138
.6325
.6513
.6700
.6888
.7075
.7263
.7450
.7638
.7825
.8013
.8200
.8388
.8575
.8763
.8950
.9138
.9325
.9513
.4700
.4700
.4700
.9700
.9700
.9700
.2017
.2017
.1651
.1285
.0920
.0554
.0188
.9822
.9456
.9090
.8724
.8359
.7993
.7627
.7261
.6895
.6529
.6163
.5798

MMNNNNMNMMMMMMMMNMIOHHHHHHHHHHHHHHHHHHHHHMHHHHHHHHHHHHH

.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200

HHHHHHHHHHHHHHHHHHHOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9000
.9167
.9333
.9000
.9167
.9333
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000

167

OOOOOOOOOOOOOHHHHHHH-HHOOOHHHHHHHHHHHHHHHHHHHHHHHHHHHHH

.4450
.4638
.4825
.5013
.5200
.5388
.5575
.5763
.5950
.6138
.6325
.6513
.6700
.6888
.7075
.7263
.7450
.7638
.7825
.8013
.8200
.8388
.8575
.8763
.8950
.9138
.9325
.9513
.9700
.4700
.4700
.4700
.9700
.9700
.9700
.2383
.1651
.1285
.0920
.0554
.0188
.9822
.9456
.9090
.8724
.8359
.7993
.7627
.7261
.6895
.6529
.6163
.5798
.5432

MMMMNMNMMMMMMMMMMMMHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHI-‘HHHH

.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.1200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200

HHHHHHHHHHHHHHHHHHH0000000000000000000000000OOOOOOOOCO

.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9500
.9167
.9333
.9500
.9167
.9333
.9500
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO000000000000OO

.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010

2222222222222222222222

323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344

HOthi—t

HHHHHHHHHHHHHHHHHHHHHH

HHHHHHHHHHHHHHHHHHHHOO

H
O
H

302
299
304

.5432
.5066
.2383
.2749
.3115
.3480
.3846
.4212
.4578
.4944
.5310
.5676
.6041
.6407
.6773
.7139
.7505
.7871
.8237
.8602
.8968
.9334

HHHO

MMMMMMMMMMMMMMMNMMMMNM

.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200

OHHO

1.0000 0.5066
1.0000 0.4700
1.0000 1.2749
1.0000 1.3115
1.0000 1.3480
1.0000 1.3846
1.0000 1.4212
1.0000 1.4578
1.0000 1.4944
1.0000 1.5310
1.0000 1.5676
1.0000 1.6041
1.0000 1.6407
1.0000 1.6773
1.0000 1.7139
1.0000 1.7505
1.0000 1.7871
1.0000 1.8237
1.0000 1.8602
1.0000 1.8968
1.0000 1.9334
1.0000 1.9700
50.00 1.00
50.
9999999999.

1.

168

O

MMMMMMMMMMMNMMMMMMMNMM

.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200
.2200

HHHHHHHHHHHHHHHHHHHHHH

.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000
.0000

0000000000000000000000

.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010
.0010

BIBLIOGRAPHY

169

BIBLIOGRAPHY

[1] C. R. Paul, Introduction to Electromagnetic Compatibility. New York: John Wiley
& Sons, Inc., 1992.

[2] International Electrotechnical Commission, Geneve, Suisse, CISPR 25: Limits
and Methods of Measurement of Radio Disturbance Characteristics for the Pro-
tection of Receivers Used on Board Vehicles, first ed., 1995.

[3] Hewlet-Packard Company, Santa Rosa, California, HP 8510C Network Analyzer
System Operating and Programming Manual, second ed., January 31 1994.

[4] Gerald J. Burke, Numerical Electromagnetics Code - NEC - 4 Method of Moments
Part 1: User’s Manual. Lawrence Livermore National Laboratory, January 1992.

[5] Gerald J. Burke, Numerical Electromagnetics Code - NEC - 4 Method of Moments
Part II: Program Description - Theory. Lawrence Livermore National Laboratory,
January 1992.

[6] D. M. Pozar, Microwave Engineering. Reading, Massachusetts: Addison-Wesley
Publishing Company, 1990.

[7] S. Ramo, J. R. Whinnery, and T. Van Duzer, Fields and Waves in Communication
Electronics. New York: John Wiley & Sons, Inc., Second ed., 1984.

170

"I[llllllllllllll'llllf