close

Вход

Забыли?

вход по аккаунту

?

Near-field imaging of microwave circuits with electromagnetic probes

код для вставкиСкачать
Near-Field Imaging of Microwave Circuits with
Electromagnetic Probes
by
Ruifeng Zhai
A Thesis
Subm itted to the Faculty o f Graduate Studies
in Partial Fulfillm ent of the Requirem ents
for the D egree of
M aster o f Science
Departm ent of Electrical and Com puter Engineering
University o f M anitoba
W innipeg, M anitoba
Canada
© June, 2004
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
1*1
Library and
A rchives C a n a d a
Bibliotheque et
A rchives C a n a d a
Published H eritag e
Branch
Direction du
P atrim oine d e I'edition
395 Wellington Street
Ottawa ON K1A 0N 4
C anada
395, rue Wellington
Ottawa ON K1A 0N 4
C anada
Your file Votre reference
ISBN: 0-612-97248-8
Our file Notre reference
ISBN: 0-612-97248-8
The author has granted a non­
exclusive license allowing the
Library and Archives Canada to
reproduce, loan, distribute or sell
copies of this thesis in microform,
paper or electronic formats.
L'auteur a accorde une licence non
exclusive permettant a la
Bibliotheque et Archives Canada de
reproduire, preter, distribuer ou
vendre des copies de cette these sous
la forme de microfiche/film, de
reproduction sur papier ou sur format
electronique.
The author retains ownership of the
copyright in this thesis. Neither the
thesis nor substantial extracts from it
may be printed or otherwise
reproduced without the author's
permission.
L'auteur conserve la propriete du
droit d'auteur qui protege cette these.
Ni la these ni des extraits substantiels
de celle-ci ne doivent etre imprimes
ou aturement reproduits sans son
autorisation.
In compliance with the Canadian
Privacy Act some supporting
forms may have been removed
from this thesis.
Conformement a la loi canadienne
sur la protection de la vie privee,
quelques formulaires secondaires
ont ete enleves de cette these.
While these forms may be included
in the document page count,
their removal does not represent
any loss of content from the
thesis.
Bien que ces formulaires
aient inclus dans la pagination,
il n'y aura aucun contenu manquant.
Canada
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
THE UNIVERSITY OF MANITOBA
FACULTY OF GRADUATE STUDIES
*****
COPYRIGHT PERMISSION
Near-Field Imaging of Microwave Circuits w ith
Electromagnetic Probes
BY
Ruifeng Zhai
A Thesis/Practicum submitted to the Faculty of Graduate Studies of The University of
Manitoba in partial fulfillment of the requirement of the degree
Of
MASTER OF SCIENCE
Ruifeng Zhai © 2004
Permission has been granted to the Library of the University of Manitoba to lend or sell copies of
this thesis/practieum, to the National Library of Canada to microfilm this thesis and to lend or sell
copies of the film, and to University Microfilms Inc. to publish an abstract of this thesis/practicum.
This reproduction or copy of this thesis has been made available by authority of the copyright
owner solely for the purpose of private study and research, and may only be reproduced and copied
as permitted by copyright laws or with express written authorization from the copyright owner.
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Abstra ct
The focus of this thesis is to investigate the feasibility of utilization of non-contact
near-field probing techniques with m iniature electrom agnetic probes to evaluate the
perform ance o f M icrow ave Integrated Circuits (MICs). U nlike conventional microwave
m easurem ent methods, such as using a vector network analyzer to directly probe the pins
o f the circuit under test, non-contact near-field probing techniques are non-invasive and
capable o f m easuring at arbitrary spots within the circuit with potentially high temporal
and spatial resolution. T he thesis gives an overview on non-contact probing systems
currently developed for signal m easurem ent and w aveform extraction techniques and
near-field m apping and im aging techniques.
For the purpose of investigating the feasibility of non-contact near-field m easurement
techniques, an autom ated near-field probing platform w as developed. An analytic model
o f the probing system was derived and discussed. Both electric and magnetic probes
(monopole probe, coplanar w aveguide probe, and single/multi loop probe) were
designed,
fabricated
and
installed
onto
the
platform
for
various
applications.
A pproxim ate circuit models for these probes were created and analyzed. Various
m icrowave circuits, including coplanar waveguide transm ission line (CPW), single patch
antenna and m icrostrip antenna array, were tested by using the near-field probing system.
To verify the m easurem ent results of CPW structures using the electric near-field probing
system, the electric field variation over the surface o f a CPW circuit was num erically
evaluated by using a finite difference approach, with a SO R iterative matrix solver, to
solve the Laplace B oundary Value problem.
In order to im prove the spatial resolution to enable application to on-w afer
microwave circuit probing, a m icro-fabricated CPW probe, at a scale less than 100 pm ,
was evaluated using a m icrom achinm g process at the A lberta M icroelectronics Centre.
The fabricated probe was installed onto the autom ated near-field probing platform and
measurement of coplanar w aveguide transm ission lines was made.
i
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Single-loop and m ulti-loop magnetic near-field probing was investigated as an
alternative to the electric near-field probing technique as a method of production line
testing o f printed circuit boards. The magnetic near-field probes were used to determine
the detection capability of faults in multi-layer Printed Circuit boards. Both indirect and
direct excitation m easurem ents were conducted and evaluated. The signal frequency in
such applications is lim ited up to 300 M Hz range, and strong coupling between probes
was found to have considerable impact on the performance of this near-field
m easurem ent approach.
11
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Acknow ledgem ents
I w ould like to express my gratitude to m y supervisor, Dr. Greg Bridges, whose
expertise, understanding, and patience, added considerably to my graduate experience. I
appreciate his vast know ledge and skill in the areas o f electrom agnetics and microwave,
scanning probe technologies and m icroelectronics, and his assistance in writing reports. I
w ould like to thank the other members of the SPM Lab for the assistance they provided at
all levels of the research project.
I w ould also like to thank m y family for the support they provided me through my
M .Sc. program and in particular, I m ust acknow ledge my wife, Liting, without whose
love, encouragem ent and editing assistance, I w ould not have finished this thesis.
The financial support from the National Sciences and Engineering Research Council
o f Canada (NSERC) is also gratefully acknowledged.
iii
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
T a b le o f C o n t e n t s
ABSTRACT................................................................................
ACKNOWLEDGEMENTS
I
......
Ill
TABLE OF CONTENTS.............................................................
TABLE OF FIGURES
CHAPTER 1
1.1
l .2
IV
.........
INTRODUCTION
VI
.................................
1
M o t i v a t i o n .....................................................................................................................................................................1
T h e s i s O u t l i n e ..............................................................................................................................................................3
CHAPTER 2
REVIEW OF NON-CONTACT NEAR-FIELD PROBING TECHNIQUES... 4
2. l
S i g n a l M e a s u r e m e n t a n d W a v e f o r m E x t r a c t i o n T e c h n i q u e s ................................................ 4
E lectron Beam P ro b in g ........................................................................................................................................ 4
SF M based p ro b in g ................................................................................................................................................ 7
2 .1 .2 .1 S c a n n in g F o rc e P o te n tio m e try ......................................................................................................................7
2 .1 .2 .2 A C F ie ld M e a s u re m e n t w ith S F M B a se d P ro b in g S y s te m ............................................................. 10
2.1.3
N oninvasive Electrostatic F orce M icro sco p y ............................................................................................ 13
2.1.1
2.1.2
2.2
F i e l d M a p p i n g a n d I m a g i n g T e c h n i q u e s ................................................................................................. 18
2.2.1
Electro-O ptic P ro b in g ....................................................................................................................................... 18
2.2.2
M odulated Scattering T ech n iq u e .................................................................................................................... 26
2.2.3
P assive Electrom agnetic P ro b in g ...................................................................................................................2 9
2 .2 .3 .1
2 .2 .3 .2
D o u b le /S in g le L o o p M a g n e tic P r o b e ..................................................................................................... 29
P a ss iv e E le c tric -fie ld P r o b e ....................................................................................................................... 35
CHAPTER 3
AUTOMATED NEAR-FIELD SCANNING SYSTEM................
40
3.1
3 .2
3.3
F u n d a m e n t a l s o f t h e S c a n n in g S y s t e m ................................................................................................. 4 0
A p p r o x i m a t e C ir c u it M o d e l l i n g o f t h e M o n o p o l e P r o b e ........................................................ 4 3
S c a n n i n g S y s t e m H a r d w a r e D e s c r i p t i o n ............................................................................................. 4 6
3 .4
S o f t w a r e D e v e l o p m e n t w i t h L a b V I E W ™ ............................................................................................. 5 0
3.4.1
3.4.2
Introduction o f L a b V IE W .................................................................................................................................. 5 0
The L abV IE W ™ Test P rogram f o r C o m p u ter C ontrolled Scanning S y s te m .................................52
CHAPTER 4
NEAR-FIELD SCAN USING MONOPOLE AND COPLANAR
WAVEGUIDE PROBES .......................................................
4.1
M o n o p o l e P r o b e s t o M e a s u r e 5 0 Q C P W T - L i n e ...............................................................................55
4.1.1
C oplanar W aveguide T ransm ission L ine A n a ly s is .................................................................................. 56
4 .1 .1 .1
A n a ly sis o f T ra n sv e rse E le c tric F ie ld V a r ia tio n ................................................................................. 58
4 .1 .1 .2 A n a ly sis o f L o n g itu d in a l E le c tric F ie ld V a ria tio n ............................................................................. 63
4 .7 .2
5 0 Q C P W T- L ine M easurem ent with M o n o p o le P robes o f D ifferent Tip L en g th s ................. 66
4 .2
M o n o p o l e P r o b e t o M e a s u r e S i n g l e P a t c h M i c r o s t r ip A n t e n n a ....................................... 7 0
P roperties o f Single Patch M icrostrip A n te n n a ........................................................................................ 70
S ingle Patch M icrostrip A n tenna M e a su re m e n t ..................................................................................... 75
4 .3
M o n o p o l e P r o b e t o M e a s u r e M i c r o s t r i p A n t e n n a A r r a y ....................................................... 7 9
4.3.1
M icrostrip A ntenna A rra y T opologies a n d Signal Feeding N e tw o rk ............................................ 79
4 .4
C P W P r o b e t o M e a s u r e 50£2 C P W T - L i n e ..............................................................................................85
4 .2 .1
4.2.2
4.4.1
C P W T-line P robe C ircuit M o d e llin g ........................................................................................................... 86
iv
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
4 .4 .2
4 .4 .3
4.4.4
4 .5
5 0 Q C P W T- Line M easurem ent with D ifferent C P W P robe O rien ta tio n s .................................. 87
5 0 Q C P W T- Line M easurem ent with C P W P robe a t D ifferent H eights above D U T .............. 89
5 0 Q C P W T- Line M easurem ent w ith C P W P robe a n d M onopole P robe - A Com parison 92
C o n c l u s i o n ...................................................................................
CHAPTER 5
MICROFABRICATED CPW PROBES
5 .1
I n t r o d u c t i o n .............................................................................................
5 .2
B
a s ic
5.2.1
5 .2 .2
5 .2 .3
5.3
5.4
Procedures
...........95
. .. 9 5
M i c r o m a c h i n i n g ............................................................................................................ 9 6
M ic r o m a c h in e d
CPW P r o b e ....................................................................................................................................... 99
103
C o n c l u s i o n ...................................................................................
SIMPLE PCB MEASUREMENT WITH MAGNETIC LOOP PROBES .... 104
N e a r -fie ld
RF I n s p e c t i o n
o f
PCB s ...................................................................................................................... 105
M e a s u r e m e n t s o n I n d i r e c t e x c i t a t i o n .......................................................................................................... 107
6.2.1
6.2.2
M agnetic Coupling between L oops over a P C B G round P la n e ...................................................... 107
M agnetic Coupling betw een loops o ver a P C B Trace with Both E nds Shorted to G round 117
6.2.2.1
6.2.2.2
6 .3
M
6.3.1
6.3.2
6.4
......
E tch ............................................................................................................................................................................ 96
D eposition .......................................................................................
97
Silicon M em brane ................................................................................................................................................ 98
CHAPTER 6
6.1
6.2
of
...9 4
Frequency Sweeping to Measure the Coupling Characteristic........................................117
Scanning along the PCB Trace..................................................................................... . 119
ea su r em en t on
D
ir e c t e x c i t a t i o n
..............................................................................................................1 2 5
F requency Sweeping to M ea su re the Coupling C h a ra cteristic ....................................................... 125
Transverse Scanning across the P C B T race ............................................................................................127
C o n c l u s i o n ............................................................................................................................................................................. 131
CHAPTER 7
REFERENCES
CONCLUSIONS...............................................................
......
132
134
v
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
T a b l e o f F ig u r e s
F i g u r e 2 -1 E l e c t r o n b e a m p r o b i n g e q u i p m e n t [ 2 ] ................................................................................................. 5
F i g u r e 2 -2 S e c o n d a r y e l e c t r o n s ( a ) t r a p p e d b y + 5 V p o t e n t i a l ; ( b ) r e p e l l e d b y -5 V
POTENTIAL [ 5 ] .......................................................................................................................................................... 5
F i g u r e 2 -3 W a v e f o r m s a t t h e a d d r e s s b u f f e r o u t p u t o f a 4 K b i t R A M [4 ].........................................6
F i g u r e 2 -4 P r o b e m e a s u r e m e n t s y s t e m a n d p o s i t i o n i n g [ 7 ] ........................................................................... 8
F i g u r e 2 -5 P r o b e m e a s u r e m e n t s y s t e m a n d p o s i t i o n i n g [ 8 ] ........................................................................... 9
F ig u r e 2 -6 S c h e m a t i c l a y o u t o f c o m b i n e d S F M / c o a x i a l t i p w it h a c o p l a n a r w a v e g u id e
a l o n g t h e c a n t i l e v e r a n d t h e c l o s e - u p o f c o a x i a l t i p p r o b i n g e l e c t r ic f i e l d
l i n e s b e t w e e n t w o c o n d u c t o r s [ 1 0 ] ................................................................................................... 11
F ig u r e 2 -7 a ) s c h e m a t i c a r r a n g e m e n t o f s h i e l d S F M t i p / c a n t i l e v e r c o n n e c t e d t o a
c o a x ia l o u t c a b l e ; b ) E x c e r p t s f r o m a l i n e a r s c a n a l o n g t h e N L T L d i o d e s
s h o w i n g w a v e f o r m c o m p r e s s i o n a t a 2 G H z [ 1 2 ] ...................................................................... 12
F i g u r e 2 -8 a ) D ia g r a m o f t h e c o n s t r u c t e d p o t e n t i o m e t r i c s c a n n i n g - p r o b e m ic r o s c o p e [1 3 ]
b ) P u l s e s a m p l e d h e t e r o d y n e e l e t r o s t a t ic f o r c e m e a s u r e m e n t o f t h e l o c a l
d i g it a l p a t t e r n
[ 1 4 ] .......................................................................................................................................14
F i g u r e 2 -9 a ) I n t e r c o n n e c t t e s t s t r u c t u r e ; b ) M e a s u r e d p o t e n t i a l a c r o s s t w o o f t h e
i n t e r c o n n e c t l i n e s s h o w n i n f i g u r e a a l o n g p a t h ( a ); c ) M e a s u r e d p o t e n t i a l
a c r o s s t h r e e o f t h e in t e r c o n n e c t l i n e s s h o w n in f i g u r e a a l o n g p a t h ( b ) [1 3 ] 16
F ig u r e 2 -1 0 a ) E x a m p l e o f p e r io d i c d i g i t a l w a v e f o r m v c ( x , y , t ) t o b e m e a s u r e d ; b ) i d e a l
s a m p l in g p u l s e
Vs(T) = G s(T-Tn), w h i c h i s s h i f t e d a l o n g t h e d i g it a l p a t t e r n [1 4 ]
................................................................
17
F i g u r e 2-1 1 M e a s u r e m e n t o f 1 6- b i t p a t t e r n a t 1 2 5 M b / s a ) a p p l i e d p a t t e r n f r o m
g e n e r a t o r ; b ) m e a s u r e d w a v e f o r m u s i n g E F M [ 1 7 ] .............................................................. 18
F ig u r e 2 -1 2 C o m m o n f o r m s o f e l e c t r o - o p t ic p r o b i n g a ) d i r e c t b a c k s id e p r o b i n g g e o m e t r y
FOR A COPLANAR WAVEGUIDE; B) EXTERNAL ELECTRO-OPTIC PROBING [ 5 ] ...........................19
F i g u r e 2 -1 3 E x t e r n a l e l e c t r o - o p t ic p r o b e s t a t i o n [ 2 1 ] .................................................................................21
F ig u r e 2 -1 4 E l e c t r o - o p t i c m a p p in g o f t h e n o r m a l f i e l d c o m p o n e n t o f t h e C P W e v e n m o d e SOLID LINE: NORMALIZED AMPLITUDE, DASH LINE; PHASE [21 ] ................................................... 21
F i g u r e 2 -1 5 E l e c t r o - o p t i c m a p p in g o f t h e t a n g e n t i a l f i e l d c o m p o n e n t o f t h e C P W e v e n
F ig u r e
F ig u r e
F ig u r e
F ig u r e
2 -1 6
2 -1 7
2 -1 8
2 -1 9
MODE - SOLID LINE: NORMALIZED AMPLITUDE, DASH LINE: PHASE [ 2 1 ] .................................. 2 2
A n e x p e r i m e n t a l s e t u p f o r c o m b i n e d e l e c t r o t h e r m a l m e a s u r e m e n t s [ 2 4 ] ........2 4
P r o b e a n d p o w e r - m e t e r m e a s u r e m e n t o f t h e M M IC [ 2 4 ] .................................................. 25
P r o b e -o n l y m e a s u r e m e n t s o f t h e M M I C [ 2 4 ] .............................................................................2 5
T e m p e r a t u r e - c a l i b r a t e d e l e c t r i c - f i e l d d a t a [ 2 4 ] ............................................................... 25
F ig u r e 2 -2 0 T h e n e a r f i e l d s c a t t e r i n g p r o b e s ; T h e d i p o l e is u s e d f o r s c a t t e r i n g t h e
TANGENTIAL ELECTRIC FIELD AND THE MONOPOLE IS USED FOR SCATTERING THE NORMAL
ELECTRIC FIELD [ 2 5 ] ........................................................................................................................................... 2 6
F ig u r e 2 -2 1 T h e m i c r o w a v e c i r c u i t e l e c t r ic f i e l d i m a g i n g s y s t e m u s in g m o d u l a t e d
SCATTERING PROBE [ 2 6 ] ................................................................................................................................... 2 7
F i g u r e 2 -2 2 a ) I n - p h a s e s i g n a l o f n o r m a l e l e c t r i c f i e l d ; b ) Q u a d r a t u r e s i g n a l o f n o r m a l
ELECTRIC FIELD; C) NORMAL ELECTRIC FIELD; D) NORMAL ELECTRIC FIELD PHASE DELAY
[ 2 5 ] .............................................................................................................................................................................. 2 9
F ig u r e 2 -2 3 D o u b l e l o o p m a g n e t ic p r o b e [ 2 8 ] ........................................................................................................3 0
F i g u r e 2 -2 4 D o u b l e - l o o p m a g n e t ic p r o b e c o u p l i n g t o m a g n e t i c f i e l d s o f a ) m i c r o s t r i p ; b )
c o p l a n a r ; c ) i n t e r a c t s b e t w e e n t w o l o o p s [ 2 8 ] ........................................................................31
F i g u r e 2 -2 5 M e a s u r e m e n t c o n f i g u r a t i o n [ 2 8 ] ...................................................................................................... 31
vi
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
F ig u r e 2 - 2 6 C o m p a r i s o n o f t h e e x p e c t e d a n d m e a s u r e d s t a n d in g w a v e a l o n g a m ic r o s t r i p
HAVING AN S W R = 1 .3 6 - A) NOT IMPROVED; B) IMPROVED BY THE PROBE-REVERSAL
TECHNIQUE [ 2 8 1 ........................................................................
32
F ig u r e 2 - 2 7 T r a n s v e r s e p a t t e r n o f a n o p e n - c i r c u i t e d m ic r o s t r i p - a ) t a k e n a t a c u r r e n t
m a x i m u m ; b ) t a k e n a t a c u r r e n t m in i m u m [ 2 8 ] ..........................................................................33
F ig u r e 2 - 2 8 S i n g l e s q u a r e l o o p m a g n e t ic p r o b e - a ) P r o b e f o r H z m a g n e t ic m e a s u r e m e n t ;
b)
P r o b e f o r H x a n d H y m a g n e t i c m e a s u r e m e n t [ 3 0 ] .......................................................... 3 4
F ig u r e 2 - 2 9 C o m p a r i s o n o f m e a s u r e d a n d c a l c u l a t e d s q u a r e d - a m p l i t u d e o f H z - a ) a l o n g
THE MICROSTRIP; B) PERPENDICULAR TO THE MICROSTRIP AT CURRENT MAXIMUM [3 0 ]. 35
F ig u r e 2 - 3 0 E l e c t r i c f i e l d p r o b e s - a ) E P Z ; b ) E P X Y w it h t r a n s m i s s i o n l i n e ; c ) e x p e n d e d
E P X Y WITHOUT TRANSMISSION LINE [3 2 ] ............................................................................................. 36
F ig u r e 2 - 3 1 a ) T h e l a y o u t o f t h e b a n d p a s s f i l t e r , a n d b ) 2 -D r e p r e s e n t a t i o n o f t h e
m e a s u r e d s q u a r e d f i e l d s t r e n g t h |E x|2 in t h e p a s s b a n d a t 1 2 .8 G H z [ 3 2 ] ................37
F ig u r e 2 - 3 2 M e a s u r e d s c a t t e r i n g p a r a m e t e r s f o r t h e b a n d p a s s f i l t e r [ 3 2 ] .................................37
F ig u r e 2 -3 3 T h e P F a n d r e f l e c t io n c o e f f i c i e n t s S l l o f t h e c o a x ia l e l e c t r j c - f i e l d p r o b e
[ 3 3 ] .............................................................................................................................................................................38
F ig u r e 2 - 3 4 C o m p a r i s o n b e t w e e n t h e m e a s u r e d a n d c a l c u l a t e d n o r m a l e l e c t r ic f i e l d a t
A CROSS SECTION OF THE MICROSTRIP LINE AT 1 1 .8 G H Z . ( SUBSTRATE: A l 20 3, Er = 9 .8 ,
H = 6 3 5 MM, w = 7 0 0 MM) [ 3 3 ] ........................................................................................................
39
F i g u r e 2 -3 5 T h e in v e s t i g a t io n o f t h e i n f l u e n c e o f t h e p r o b e o n t h e m i c r o s t r i p l in e [3 3 ] ...3 9
F ig u r e 3 -1 D ia g r a m o f N e a r - f i e l d S c a n n i n g S y s t e m ......................................................................................4 0
F i g u r e 3 -2 A n a p p r o x i m a t e m o n o p o l e p r o b e m o d e l a n d t h e c o r r e s p o n d i n g e q u i v a l e n t
c i r c u i t m o d e l [ 5 ] ............................................................................................................................................. 4 3
F ig u r e 3 -3 S i m p l if ie d e q u i v a l e n t c i r c u i t f o r m o n o p o l e p r o b e a s s u m i n g ....................................... ..4 4
F ig u r e 3 -4 C a l c u l a t i o n o f t h e c r i t i c a l c o u p l i n g c a p a c it a n c e C cl (C , l ) [ 5 ] ................................... 4 5
F ig u r e 3 -5 M e a s u r e m e n t s y s t e m h a r d w a r e ...........................................................................................................47
F i g u r e 3 -6 P l o t t e r m o u n t e d w it h m o n o p o l e p r o b e f o r m m r e s o l u t i o n s c a n s .............................. 4 8
F ig u r e 3 -7 C l o s e u p o f D C m o t o r t e s t s t a t i o n u s e d t o s c a n jam c i r c u i t s .......................................... 4 9
F ig u r e 3 -8 S e t u p o f t h e S c a n n in g S y s t e m f o r f m l e v e l m e a s u r e m e n t ........................................... ...4 9
F i g u r e 3 -9 F r o n t p a n e l o f t h e L a b V IE W t e s t p r o g r a m f o r C o m p u t e r C o n t r o l l e d S c a n n in g
S y s t e m ..................................................................................................................................................................... 53
F i g u r e 4 -1 S t r u c t u r a l d i a g r a m o f a c o p l a n a r w a v e g u id e T - l i n e .........................................................55
F ig u r e 4 -2 E l e c t r i c f i e l d v a r i a t io n a b o v e C P W T - l i n e a n d m e a s u r e m e n t w i t h m o n o p o l e
p r o b e ......................................................................................................................................................................... 5 6
F i g u r e 4 -3 a ) S t r u c t u r a l d i a g r a m o f C P W T - l i n e ; b ) T y p e B C P W u n d e r t e s t , s = 1 0 0 f m
and w
= 2 5 0 f m ................................................................................................................................................... 5 8
F ig u r e 4 - 4 L a p l a c e B o u n d a r y V a l u e p r o b l e m s o l v i n g .................................................................................. 5 9
F i g u r e 4 -5 C o n t o u r p l o t o f e l e c t r i c p o t e n t i a l
a p p l ie d t o t h e
y) o f T y p e a C P W p r o v i d e d t h a t 5 V
C P W s i g n a l l i n e ............................................................................................................6 1
F i g u r e 4 -6 E Y( x , y ) a t d i f f e r e n t h e i g h t a b o v e T y p e B C P W c i r c u i t p l a n e ......................................... 61
F i g u r e 4 -7 L a p l a c e B V p r o b l e m s o l u t i o n f o r T y p e A C P W p r o v i d e d t h a t d i f f e r e n t
m o n o p o l e p r o b e s a t s a m e h e ig h t o f
1 0 0 FM o v e r t h e c i r c u i t p l a n e .......................... 62
F ig u r e 4 -8 L a p l a c e B V p r o b l e m s o l u t i o n f o r T y p e B C P W p r o v i d e d t h a t a C P W p r o b e w it h
TIP LENGTH o f 1 8 0 FM a t d i f f e r e n t h e i g h t s o v e r t h e c i r c u i t p l a n e ............................6 2
F ig u r e 4 -9 E q u i v a l e n t c i r c u i t o f a s e c t i o n o f t r a n s m i s s i o n l i n e ............................................................ 63
F ig u r e 4 -1 0 S c h e m a t i c m o d e l o f a g e n e r a l l y l o a d e d t r a n s m i s s i o n l i n e ......................................... 6 4
F i g u r e 4 -1 1 M o n o p o l e p r o b e .........................................................................................................................................
66
F ig u r e 4 -1 2 T e s t s t r u c t u r e t o m e a s u r e C P W T - l i n e s ................................................................................... ..6 7
F i g u r e 4 -1 3 M a g n i t u d e a n d p h a s e p l o t o f t r a n s v e r s e s c a n o v e r T y p e A C P W w it h t w o
m o n o p o l e p r o b e s h a v i n g d i f f e r e n t t i p l e n g t h s ..................................................................... ..6 9
F i g u r e 4 -1 4 M a g n i t u d e a n d p h a s e p l o t o f l o n g i t u d i n a l s c a n a l o n g t h e s i g n a l l i n e o f T y p e
A C P W WITH 0 .5 MM TIP LENGTH MONOPOLE P R O B E ........................................................................6 9
F i g u r e 4 -1 5 D i m e n s i o n s o f t h e o p e n - e n d e d s i n g l e p a t c h m i c r o s t r i p a n t e n n a d e s i g n
..7 1
vii
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
4 -1 6 S tructural
d i a g r a m o f a s i n g l e p a t c h m i c r o s t r i p a n t e n n a d e s i g n ...........................7 1
F ig u r e 4 - 1 7 E q u iv a l e n t
c i r c u i t o f m i c r o s t r i p l i n e f e e d s i n g l e p a t c h a n t e n n a ............................ 7 3
f ig u r e
F ig u r e 4 - 1 8
a)
S in g l e
F ig u r e 4 - 1 9 3 -D
F ig u r e 4 - 2 0 M
patch a n t e n n a ; b )
p a t c h a n t e n n a t e s t s t r u c t u r e ...........................7 6
d is p l a y o f t h e m a g n it u d e p a t t e r n o f t h e s in g l e p a t c h a n t e n n a
a g n it u d e p a t t e r n o f t h e
F ig u r e 4 - 2 1 P h a s e
F ig u r e 4 - 2 2 M
S in g l e
pattern of the
a g n it u d e
(u p p e r )
S in g l e
S in g l e
patch a n ten n a
patch a n t e n n a
a n d ph ase
(l o w e r )
plo t
a n t e n n a patch alo n g the central l in e
-
........................7 6
.......................................................................7 7
....................................................................................7 7
l o n g it u d in a l s c a n n in g o v e r t h e
.................................................................................................7 8
F ig u r e 4 - 2 3 T he
layo ut o f type c anten n a a r r a y
F ig u r e 4 - 2 4 T h e
l a y o u t o f t y p e e a n t e n n a a r r a y ...................................................................................................... 81
..................................................................................................... 81
F ig u r e 4 - 2 5 N
e a r -f ie l d im a g e o f t y p e
C
a n t e n n a a r r a y d r iv e n b y
1 2 G F lz R F
s i g n a l .............. 8 2
F ig u r e 4 - 2 6 N
e a r -f ie l d im a g e o f t y p e
E
a n t e n n a a r r a y d r iv e n b y
12 G H z R F
s ig n a l
.............. 8 3
F ig u r e 4 - 2 7 N
e a r - f ie l d im a g e o f t y p e
E
a n t e n n a a r r a y d r iv e n b y
13 G H z R F
s ig n a l
.............. 8 4
F ig u r e 4 - 2 8 B
o r e s ig h t r a d ia t io n a n d in p u t im p e d a n c e c h a r a c t e r is t ic s o f
12 G H z 4 x 4
MICROSTRIP ARRAYS [ 4 3 ] ..........................................................................................................................................8 5
F ig u r e 4 -2 9
) C PW Pro be
a
F ig u r e 4 -3 0 A
b ) sn a psh o t of
CPW
probe a n d th e
D U T .......................................................... 8 5
n a p p r o x im a t e c o p l a n a r w a v e g u id e t r a n s m is s io n l in e p r o b e m o d e l a n d t h e
CORRESPONDING EQUIVALENT CIRCUIT M ODEL........................................................................................... 8 6
F ig u r e 4 -3 1
a
) T he C PW
p r o b e s u r f a c e n o r m a l is p a r a l l e l t o s i g n a l l i n e a n d b )
The CPW
PROBE SURFACE NORMAL IS ORTHOGONAL TO SIGNAL L IN E ................................................................ 8 7
F ig u r e 4 -3 2 C P W
probe t r a n sv er se sc a n over the
Type A CPW
w it h p r o b e s u r f a c e n o r m a l
PARALLEL TO THE C P W T-LINE UNDER TEST................................................................................................. 8 8
F ig u r e 4 -3 3 C P W
pr o be t r a n sv e r se sc a n o ver the
Type A C PW
w it h p r o b e s u r f a c e n o r m a l
ORTHOGONAL TO THE C P W T-LINE UNDER TEST.........................................................................................8 8
F ig u r e 4 - 3 4 C P W
pr o be t r a n sv e r se sc a n r esu lt o n
CPW
t r a n s m i s s i o n l in e , h e ig h t o f
0 .5 w
..................................................................................................................................................................................................... 8 9
F ig u r e 4 - 3 5 C P W
pro be t r a n sv e r se sc a n resu lt o n
CPW
t r a n s m is s io n l in e , h e ig h t o f w
F ig u r e 4 -3 6 C P W
pr o be t r a n sv e r se sc a n resu lt on
CPW
t r a n s m is s io n l in e , h e ig h t o f
F ig u r e 4 -3 7 C P W
pro be t r a n sv e r se sc a n r esu lt o n
CPW
t r a n s m is s i o n l in e , h e ig h t o f
F ig u r e 4 -3 8 M
e a su r e m e n t o f m a g n it u d e ,
h e i g h t o f Q .5 w , w ,
F ig u r e 4 - 3 9 C P W
F ig u r e 4 -4 0 M
pro be tr a n sv e r se sc a n o ver the
4 w . 91
at
T y p e B C P W T - L i n e .....................................9 2
pro be t r a n sv e r se sc a n r esu lt o n th e
T y p e B C P W T -L i n
e
.......................9 3
T y p e A C P W T - L i n e .....................................9 3
ono po le pr o b e t r a n sv e r se sc a n r esu lt o n th e
F i g u r e 5 -1 B o r o n
DUT
.....................................................................................................................................91
pro be t r a n sv e r se sc a n resu lt on th e
o no po le pr o b e tr a n sv e r se sc a n r esu lt o n th e
F ig u r e 4 -4 1 C P W
F ig u r e 4 -4 2 M
2w , 4 w
CPW
... 9 0
2 w .90
T y p e A C P W T -L i n e ...................... 9 4
m e m b r a n e in t h e f a b r ic a t io n o f s il ic o n in k je t n o z z l e
-
a ) a n d b ) sh o w
t h e e r r o r s in f i n a l n o z z l e s i z e w h i c h o c c u r s w h e n t h e w a f e r t h i c k n e s s v a r i e s
c)
s h o w s a b o r o n m e m b r a n e s t r u c t u r e t o m in im iz e t h e e f f e c t s o f t h e w a f e r
t h i c k n e s s ............................................................................................................................................................................. 9 8
F ig u r e 5 -2 D
im e n s io n a n d t h e p ic t u r e o f t h e m ic r o m a c h in e d
C oplanar W
a v e g u id e
(C P W )
p r o b e ........................................................................................................................................................................................9 9
F i g u r e 5 -3 T h e
layout of
F ig u r e 5 -4 T h e
s il ic o n w a f e r o f
F ig u r e 5 -5 C P W
F i g u r e 6 -1 T e s t
C o -P l a n a r W
CW P
a v e g u id e
(C P W ) P r o b e [ 6 2 ] ........................................................ 1 0 0
d u r in g t h e m ic r o m a c h in in g p r o c e s s in g [ 6 2 ]
102
p r o b e m a d e w i t h t h e m i c r o m a c h i n e d c a n t i l e v e r ........................................................... 1 0 3
se t u p fo r
PCB
F i g u r e 6 -2 T r a d i t i o n a l P C B
f a i l u r e d e t e c t i o n w i t h m a g n e t i c p r o b e s ........................................... 1 0 5
f a u l t d e t e c t i o n w i t h b e d o f n a i l s ................................................................. 1 0 5
F i g u r e 6 -3 I l l u s t r a t i o n
o f i n d i r e c t e x c i t a t i o n ......................................................................................................... 1 0 6
F i g u r e 6 - 4 Il l u s t r a t i o n
o f d ir e c t e x c it a t io n
F i g u r e 6 - 5 In d i r e c t
F ig u r e 6 -6 T w o
.............................................................................
e x c it a t io n o v e r a g r o u n d p l a n e
106
........................................................................................... 1 0 7
l o o p p r o b e s o v e r a l a r g e g r o u n d p l a n e , r in g s u r f a c e n o r m a l s a r e
p a r a l l e l ............................................................................................................................................................................. 1 0 8
F ig u r e 6 -7 E q u iv a l e n t
F ig u r e 6 -8 T w o
c ir c u it o f t w o p r o b e s o v e r a g r o u n d p l a n e
....................................................... 1 0 9
p r o b e s o v e r a g r o u n d p l a n e a n d t h e i r i m a g e s .................................................................... 111
viii
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
F ig u r e 6 - 9 T h e
n u m e r ic a l s im u l a t io n v s . e x p e r im e n t a l d a t a f o r t w o l o o p s o v e r a g r o u n d
..................................................... 1 1 2
PLANE AN D RING SURFACE NORMALS ARE PARALLEL
F ig u r e 6 - 1 0 T w o
l o o p p r o b e s o v e r a l a r g e g r o u n d p l a n e , r in g s u r f a c e n o r m a l s a r e
ORTHOGONAL.................................................................................................................................................................. 1 1 3
F i g u r e 6 - 1 1 S 21
p l o t f o r t w o l o o p s o v e r a g r o u n d p l a n e a n d r in g s u r f a c e n o r m a l s a r e
o r t h o g o n a l .................................................................................................................................................................. 1 1 3
F ig u r e 6 - 1 2 P e r f o r m a n c e
c o m p a r is o n o f s in g l e -l o o p a n d f o u r -l o o p p r o b e s , r in g s u r f a c e
NORMALS ARE PARALLEL....................................................................................................................................... 1 1 4
F ig u r e 6 -1 3 P e r f o r m a n c e
c o m p a r is o n o f s in g l e -l o o p a n d f o u r - l o o p p r o b e s , r in g s u r f a c e
NORMALS ARE ORTHOGONAL.....................................................................
F ig u r e 6 - 1 4 T w o
F ig u r e 6 - 1 5 F r e q u e n c y
F ig u r e 6 - 1 6 T w o
116
l o o p p r o b e s o v e r t h e c i r c u i t t r a c e w i t h b o t h e n d s s h o r t e d ............................. 1 1 7
s w e e p i n g r e s u l t s b y t w o s i n g l e - l o o p p r o b e s .................................................. 1 1 8
loo ps o ver a
F ig u r e 6 - 1 7 C a l c u l a t e d
PCB
t r a c e ..............................................
m a g n e t ic c o u p l in g b e t w e e n t h e
119
PCB
tr a c e a n d the m o v a b l e lo o p
p r o b e ....................................................................................................................................................................................1 2 1
F ig u r e 6 -1 8 P e r f o r m a n c e
c o m p a r is o n o f s in g l e -l o o p a n d f o u r - l o o p p r o b e s s c a n n i n g a l o n g
THE TRACE, RING SURFACE NORMALS ARE PARALLEL.......................................................................... 1 2 2
FIGURE 6 - 1 9 PERFORMANCE COMPARISON OF SINGLE-LOOP AND FOUR-LOOP PROBES SCANNING ALONG
THE TRACE, RING SURFACE NORMALS ARE ORTHOGONAL.................................................................. 1 2 3
F ig u r e 6 - 2 0 S c a n n in g
F i g u r e 6 -2 1 D
a l o n g t h e p a t h a s i d e f r o m b u t p a r a l l e l t o t h e t r a c e ................................1 2 3
is t a n c e s c a n n i n g r e s u l t s b y t w o s in g l e -l o o p p r o b e s in t h e c a s e o f c ir c u it
TRACE WITH EACH END SHORT CIRCUITED.................................................................................................... 1 2 4
F ig u r e 6 - 2 2 M
e a s u r e m e n t o n d i r e c t e x c i t a t i o n ....................................................................................................... 12 5
F ig u r e 6 -2 3 F r e q u e n c y
s w e e p in g f o r a s in g l e -l o o p p r o b e f r o m
40 M H z
F ig u r e 6 - 2 4 F r e q u e n c y
s w e e p in g f o r a s in g l e -l o o p p r o b e f r o m
40M H z
F ig u r e 6 -2 5 T r a n s v e r s e
F ig u r e 6 -2 6 A
s c a n n in g a c r o s s t h e c ir c u it t r a c e
to
to
1 0 0 0 M H z .. ....... 1 2 6
2 0 0 M H z ..................1 2 7
......................................................................... 1 2 8
n a p p r o x im a t e m o d e l f o r m a g n e t ic f ie l d c a l c u l a t io n w it h o u t c o n s id e r a t io n
o f i m a g e c u r r e n t e f f e c t ....................................................................................................................................1 2 8
F ig u r e 6 -2 7 S im u l a t io n
loop p r o b e
F ig u r e 6 -2 8 M
r e s u l t o f t r a n s v e r s e s c a n o v e r t h e c ir c u it t r a c e w it h a s in g l e ­
...................................................................................................
130
e a s u r e m e n t r e s u l t o f t r a n s v e r s e s c a n o v e r t h e c ir c u it t r a c e w it h a s in g l e ­
loop p r o b e
...........................................................................................................
ix
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
130
C h a p te r 1
I n t r o d u c t io n
The research work com pleted in this thesis is to investigate the feasibility of
applications o f electrom agnetically coupled probes as non-invasive diagnostic tools for a
broad range o f m icrowave circuits. An automated m easurem ent system has been designed
and built capable of producing inform ation unavailable from conventional measurement
techniques.
1.1
Motivation
R apid advances in the microelectronics field have enabled the developm ent of faster,
denser and more com plicated m icrowave integrated circuits (M ICs). The ability to test
the perform ance of a particular technology underlies the progress in developing that
technology. It is therefore im portant to develop appropriate tools to accurately evaluate
the perform ance o f these circuits and devices. The increasing chip com plexity has made
circuit m odelling difficult and the prediction of circuit behaviour by com puter
sim ulations alone is inadequate, Accurate m easurem ents for device evaluation are
sequentially essential. On-chip w aveform m easurem ents are often required for com plete
characterization o f devices and for failure analysis purposes. M easurem ents within a
circuit are carried out using internal probing techniques with the device driven by
external sources. W hile the m iniaturization and the enhanced speed performance of
integrated circuits are constantly being im proved, the experim ental m easurem ent of these
circuits is becom ing increasingly difficult. In fact, the perform ance characteristics of
many devices, especially those based on GaAs technology, have surpassed the capability
o f conventional electronic m easurem ent techniques.
Present measurements require probing techniques capable o f m easuring Gigahertz
signals with sub-micron spatial resolution and milli-volt sensitivity. The com plexity of
1
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
integrated circuits presents a num ber of challenges during testing at the chip level. The
ideal probing techniques should enable measurements with high temporal and spatial
resolution, high sensitivity and low invasiveness. The m easurem ent m ethod should also
be sim ple to operate and enable accurate results. Currently available instruments do not
com pletely satisfy all these requirements. It is therefore im portant to develop alternative
probing techniques.
C onventional circuit m easurem ent techniques rely on some form of direct electrical
contact w ith the point in the circuit being monitored. For exam ple, digital integrated
circuits are injected with test patterns and the outputs are m onitored to determine the
overall functionality of the circuit. This approach, how ever, does not provide any
inform ation on the identification o f individual internal faults when one arises. Likewise,
conventional netw ork analyser m easurem ents on m icrow ave circuits im pose similar
lim itations. Typically individual com ponents are designed with a high degree of accuracy
and reliability, how ever the integration of several com ponents on a common substrate
often gives rise to uncertainties in circuit perform ance. Often the performance is
degraded by mutual coupling o f elements due to the excitation o f surface waves,
particularly at m illim eter wavelengths. N etw ork analyser m easurem ents at the device
input and output ports are incapable of isolating such problem s.
An ideal m easurem ent technique should be non-invasive and capable o f measuring
w aveforms at arbitrary points within the circuits. D epending on the application it may be
desirable to perform voltage and/or current m easurem ents, preferably with magnitude and
phase data. In addition, the m easurem ent system m ust possess sufficient spatial and
tem poral resolution. D eveloping a m easurem ent system w ith such characteristics is
challenging and critical to the future developm ent o f reliable high quality integrated
circuits.
For this purpose, the feasibility o f using eletrom agnetically coupled probes as noninvasive diagnostic tools has been investigated. Probes are in various structure and shape,
2
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
including m onopole, coplanar waveguide transm ission line and looped-wire. Probes are
also in various physical sizes, from millim eter level down to pm level. An autom ated
m easurem ent system has been also designed and constructed capable of yielding
inform ation acquired by these probes.
1.2 Thesis Outline
Chapter 1 is an introduction of the research w ork to bring out the thesis.
In chapter 2, a review of the non-contact probing m easurem ent techniques currently
researched and developed within industry and academic com munity is presented. The
relative advantages and disadvantages o f each technique are discussed and sample
measurem ents are provided.
A description of the hardware and softw are required for the autom ated data
acquisition system is given in chapter 3.
Chapter 4 discusses the properties o f m onopole and CPW probes and presents the
experimental results o f measurements on m icrow ave circuits, i.e. transm ission lines,
printed antennas.
Chapter 5 gives a description on procedure to fabricate pm level probes by
m icromachining technology through VLSI process.
Chapter 6 discusses the magnetic loop probe application of printed circuit board
failure detection.
Finally, chapter 7 sum m arizes the research work and presents the future work.
3
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
C h a p te r 2
R e v ie w o f N o n - c o n t a c t N e a r - f ie l d
P r o b in g T e c h n iq u e s
W ith th e advent of near-field microscopy, conversional radio frequency (RF) and farinfrared (FIR) microscopy have gained m ore attention in recent two decades because of
their m any applications including material characterization and integrated circuit testing
[1] . This chapter gives an overview of non-contact near-field probing techniques, which
will cover the signal m easurem ent and waveform extraction technologies and the field
m apping and im aging technologies. Topics on Electron-beam Probe station, Scanning
Force M icroscopy probing, Electrostatic Force M icroscopy w ill be discussed for the
signal m easurem ent and waveform extraction technologies. The discussion on the field
m apping
and im aging
technologies,
including
Electro-optic
probing, M odulated
Scattering probing and near-field Electro/M agnetic probing will also be presented.
2.1 Signal Measurement and Waveform Extraction
Techniques
2.1.1 E lectron Beam Probing
Electron-beam probing technique is fully developed and used in industry and research
institutions. It is based on a m odified scanning electron m icroscope as shown in Figure
2-1 [ 2 ] .
T he electron gun generates low energy prim ary electrons. Em itted prim ary electrons
are accelerated through a potential on the order o f 0.5 to 2.5 kV and focussed via a lens
focussing system to a submicron spot on the device under test. On im pact with the device
under test, low energy secondary electrons are random ly em itted as a result o f elastic
collision. It can be seen in Figure 2-2 [5] that probing a +5V transm ission line will em it
less secondary electrons than that o f a -5V line. This is due to low energy electrons being
4
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
attractive to the metal conductor by the positive potential of the +5Y line while the
negative potential o f the -5V line repels such electrons.
-Electron gun
r
&
• Electron lenses
Primary electron beam
+ IOIW
Secondary electrons
Lfamptp®
/— H ;
\
Devtca stimulus
1C under test
Figure 2-1 Electron beam probing equipment [2]
primary e '
primary ©■
*
+SV
-5V
Figure 2-2 Secondary electrons (a) trapped by +5V potential; (b) repelled
by -5V potential [5]
The secondary electrons are then collected and counted. The m ethod o f collecting the
secondary electrons involves using either a high potential electrostatic extraction field
generated at the surface o f the device under test or a magnetic extraction field. The
secondary electrons are fed to a scintillator which converts the electrons to photons.
These photons are received by a photom ultiplier which generates a specific num ber of
5
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
electrons for each incident photon. In general, the num ber of secondary electrons is
inversely proportional to the potential within a lim ited area of the point o f im pact of the
prim ary electron beam. W hen the output of the photom ultiplier is converted to an input of
a video display, areas with a positive potential appear dark indicating low secondary
electron yield, w hile negative areas appear bright showing high secondary electron yield.
It is called voltage-contrast imaging [2] [3] .
In Figure 2-3 the address buffer output of a 4K Bit RAM is m easured using the
electron beam probe. A com puter simulation for reference is also included. The com puter
simulation and the electron beam probe m easurem ent com pare favourably with a delay
tim e of 2.2 ns, w hich is much im proved due to a tiny loading capacitance on the order of
10~5 pF [4] . A variety of applications such as m easuring internal w aveform s on a IK
Bipolar PRO M , a 1 GHz Gunn diode, a 4K M OS RA M can be found in [6 ].
(a)
CPU -SIMULATION
1 4 - ,
/ ; £r>s
J
1
1
\
ELECTRON PROBE
Figure 2-3 Waveforms at the address buffer output of a 4 Kbit RAM [4]
Probing bare conductors is relatively straightforw ard but probing buried (passivated)
conductors presents additional com plications. W hile probing buried conductors, the
limitation is that only AC signals can be m easured. Because the electrical behaviour of
the buried conductor can be inferred from the surface m easurem ent when the passivation
6
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
layer acts as dielectric which capacitively couples the surface potential to the buried
conductor. For a DC signal, the coupling effect will not function.
Though electron-beam probing is attractive due to its high tem poral resolution (1 ps 1 ns) and submicron spatial resolution it is a complex and expensive technology with
inherent limitations. It is required to follow some time consum ing and expensive
fabrication guidelines [3] . Im provem ent to make the DC signal m easurem ent available
on the buried conductors is also required. C rosstalk from the electric fields o f nearby
transm ission lines is another factor that can affect the results by roughly 10% [4] . As the
prim ary electrons are penetrating into the sensitive area of the device under test,
especially for the M OS devices, the electrical characteristics o f the device will be
affected [4] [6] .
2.1.2 SFM b ase d probing
2.1.2.1 Scanning Force Potentiometry
Scanning force potentiom etry is perform ed by monitoring the electrostatic force on a
conducting probe beam due to capacitively induced local charge on the circuit under test.
The physical principle o f this technique is the local interaction o f an atom icly sharp tip
m ounted on one end of the cantilever and a device under test (DUT) due to the attractive
or repulsive forces. The electrical interaction between the tip and the DUT causes a
detectible bending of the cantilever.
An exam ple of the direct capacitively-coupled probe has been designed by G. E.
Bridges, T. S. Forzly and D. J. Thom son [7] [8] . The probes were fabricated from
com m ercially available Si 3 N 4 insulating cantilevers (100 pm long, 10 pm wide, 0.3 pm
thick, with a single inverted pyram idal tip at their end.) A gold conducting electrode was
deposited on the top side of cantilever beam to make the cantilever respond to the charge
density in the localized region o f circuit under test at the position o f the tip and also to
provide a signal line. A grounded conducting electrode was deposited on the bottom side
7
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
o f beam to provide effective shield. The m etalized probe beam and its pyrex support
block w ere then m ounted into a 5 0 O stripline feed structure as shown in Figure 2 -4 [7]
which was connected to a network analyser or spectrum analyser measurem ent system.
The stripline feed is a ground-signal-ground structure and is designed to m atch the
m easurem ent system to the probe which is a parallel plate transmission line structure.
M echanical and piezoelectric manipulators attached to the stripline feed were used for
lateral and vertical m ovem ent of the probe tip position. By monitoring the vertical
position o f the probe tip, topographical and electrical characterization can be performed.
An equivalent circuit model of this probe was also developed for exam ining the expected
probe resolution and detailed resolution analysis was given in [7] [8] .
0 3um
S
G
2 828 nm
P robe b e a m a n d
pvrex mount
y
positioner
DUT
coax
S pectrum
• an a ly s e r
Figure 2-4 Probe measurement system and positioning [7]
The experimental result is com pared to the simulation o f scanning a transm ission line
in transverse direction with tw o tip-sam ple distances. Figure 2-5 [8] shows both
experimental and num erical results for two scan paths. The first case, h i, is w here the
probe was scanned at a specific height above the transm ission line. The second case, h2,
involved scanning the probe tip directly across the transm ission line in contact with the
8
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
S i0 2 substrate. It was found that as the probe approaches either side of the printed line of
transm ission line the coupling capacitance and thereby the relative induced signal pow er
increases. Both numerical sim ulation and experimental m easurem ent yield a resolution of
15 p m for the system. The 15 pm spatial resolution is how ever still insufficient to
exam ine many of the desired features o f integrated circuits. Since the spatial resolution of
this system is related to the geom etry and the dimension o f the probe, to enhance the
perform ance of the test system, m icrom achined ultra-small tip is necessary while the
trade-off should be concerned between the high spatial resolution and the lower signal
pow er output due to the decreased coupling capacitance of the ultra-sm all m icrom achined
probe.
31
o
Numerical
( 1GHz )
X
t—
a)
3
a
a.
<D
>
is
<D
-25
-20
-15
-10
■5
0
5
10
15
20
25
Probe Position (pm)
Measured
( XGHz )
0.5 -
CL
5 0.4 -
0.3 —
r
M easurments taken at indicated points only
-25
-20
15
-10
■5
0
5
10
15
20
Probe Position (pm)
Figure 2-5 Probe measurement system and positioning [8]
9
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
25
C. Bohm et al. also em ployed the SFM -based test system for MMIC internal
electrical characterization [9] . The frequency sweeping m easurem ent was conducted up
to 40 GHz to a set of coplanar waveguide structures and a travelling wave am plifier at
selected internal test points on wafer. The measurement results were com pared with that
resulted by an external netw ork analyzer, and dem onstrated the feasibility to use SFM based test system M M IC internal function and failure analysis because of the high spatial
resolution (<500 nm ) and the achieved high bandwidth (40 GHz).
2.1.2.2 AC Field Measurement with SFM Based Probing System
The active A CSFM is a non-contact m easurem ent technique w ith high spatial and
temporal resolution for AC electrical field measurement. A C SFM relies on the voltage
difference V between the tip at the cantilever’s end and that on the test point of device
under test. The force F on the cantilever goes as [10] :
P __
gpA Q ^
VDUT)
(2-1)
where £0 is the perm ittivity of free space, A is the effective tip area, and z is the
tip-sample distance. If the frequency of the voltage signal on the tip is slightly offset from
that on the sample, an inner product results from the (Vtip
- V d u t)2
term which gives a
sum and difference frequency. The difference frequency is a mechanical motion of the
cantilever that can be detected using the tip-distance control feedback loop o f the SFM. If
the fundamental and several harm onics o f this difference-frequency waveform fall below
the first mechanical resonance o f the cantilever, it can be displayed on a standard
oscilloscope and interpreted as a low -frequency replica o f the high-frequency signal on
the sample. A. Leyk, et al. have used this technique to m easure 104 GHz signals on the
gate of a GaAs field-effect transistor [11] . W hile the established “cantilever-m ixing”
approach for ACSFM dem onstrated the ability to probe devices with the required spatial
and temporal resolutions, it has some significant weaknesses in which it needs to drive
10
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
the SFM cantilever with an AC signal, which alone can cause unwanted interference or
perturbation, and which also should be phase synchronous with the unknown signal on
the device. Furtherm ore, because it uses the motion o f the cantilever for mixing, in its
basic form it cannot sim ultaneously acquire topographical and field information; the tipsample distance feedback loop o f the SFM must be broken and the mixing product taken
from this point will be the output signal. M ore im portantly, it cannot regulate the tipsample distance w hile m easure the electrical field. Since com pensation for drift during
the m easurem ent is made more difficult, there is poorer correspondence between
topographic and field im ages, absolute voltage calibrations are difficult to achieve over
the typical time o f m easurem ent. Finally, its reliance on standard SFM tip geom etry
permits long-range C oulom b interactions between the charge on the tip and that on
conductors near to the one being measured, meaning that its “electrical resolution” - the
ability to distinguish a signal on one conductor from that on a nearby conductor - is
lower than its “topographical resolution” .
Figure 2-6 Schematic layout of combined SFM/coaxial tip with a
coplanar waveguide along the cantilever and the close-up of coaxial tip
probing electric field lines between two conductors [10]
To solve the problem s in the above active ACSFM system, D. W. van der W eide and
P. Neuzil designed a reactive nanoscilloscope that integrated a shield SFM /coaxial tip
with a coplanar transm ission line along the cantilever [10] . Figure 2-6 illustrates the
first-order principle o f the near-field probe. A ssum ing the conductor spacing is 1 pm and
11
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
the instantaneous voltage across the conductors is 1 V, a field o f 104 V/cm exists. If the
pitch o f the coaxial probe is o f the order o f 0.1 fxm, a voltage of 100 mV, w hich can be
easily detected, develops between the tip and shield.
To realize the satisfactory operation at the nanom eter scale, a m icromachined
silicon/polysilicon coaxial tip with outer shield has been designed (Figure 2-7 a)) [12] .
The shield diam eter is <100 nm while the tip radius is -1 0 nm. The total height o f the tip
is >5 jam. D . W. van der W eide used this kind of probe to m easure a 30 ps w aveforms on
the nonlinear transm ission line (NLTL) diodes with a - 2 nm tip-sam ple distance while
the N LTL was driven by a 2 GHz signal. T he probe was estim ated to have Rcantiiever = 10
Q, Ccantikver = 0.03 pF, and Lbondtwire = 0.5 nH, and the tip-sam ple capacitance Cup = 0.4
pF. The experim ental results are shown in Figure 2-7 b). Although the three diodes are
approxim ately equidistant from each other, the first one (diode A) had a distortion due to
the w aveguide bend ju st before it. The Figure 2-7 b) shows the expected spatial
nonlinearity along the N LTL with a progressively steeper falling edge m oving along the
NLTL. This experim ent was made by m oving the probe with a step of < 0.1 jtm along the
NLTL.
Cant ilever
D iode A
D iode B
D iode C
l~ ...... t---
a
0
50
i
100
150
Time (ps)
-» ‘
200
b
Figure 2-7 a) schematic arrangement of shield SFM tip/cantilever
connected to a coaxial out cable; b) Excerpts from a linear scan along the
NLTL diodes showing waveform compression at a 2 GHz [12]
12
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
250
C om pared to the structure in the active ACSFM system , the techniques in the
nanoscilloscope have several advantages. It no longer needs to drive the SFM cantilever
with an AC signal. The cantilever height control feed back loop can be now used to
maintain a constant distance from the sample as well as to com pensate for any DC offset
in the voltages between tip and sample. This technology can be calibrated. It can
sim ultaneously acquire topographical and field inform ation. D ue to the shield around the
central tip, lim iting long range Coulom b interactions between the charge on the tip and
that on conductors near to the one being measured, A C SFM ’s “electrical resolution” will
be sim ilar to its “topographical resolution” . W hile we expect a wavelength-independent
spatial resolution in the 10 nm regim e, which is the radius o f the probe tip, further
theoretical and experimental w ork is needed to address the effects o f finite tip
conductivity, i.e., does skin depth affect resolution? The probe invasiveness m ust also be
answered, though we note that by having control over the p robe’s geometry, its radiation
im pedance can be designed closer to that o f free space, m inim izing the invasiveness.
2.1.3 N oninvasive E lectrostatic Force M icroscopy
N oninvasive Electrostatic Force M icroscopy has been developed for both static
potential m easurem ents and high frequency digital pattern extractions at internal points of
integrated circuits [13] - [17] .
The block diagrams of test fixture are shown in Figure 2-8 [13] [14]
for
m easurem ents o f both localized potential and local digital pattern. A conducting wire
probe with a sharp tip at the end is held in close proxim ity to the test circuit surface on
which the potential Vouii^, y), or Vc(x, y, t) is to be extracted. The wire probe is
configured so that it acts as a m echanical cantilever that can be deflected under an
applied force. There is a small capacitor C(x, y, z) form ed and charged up between probe
and circuit under test when a signal Vp externally applied to the probe to create a potential
difference between them. The induced capacitive charge produces an attractive force
resulting a deflection Az of the probe cantilever.
13
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
lock-in amplifier
L O C K -IN
A M PLIFIER
fiber
DEFLECTION
SEN SO R
fiber
positioner
probe
probe
control + bias
signals
Vptt)
Vs(t)
x y z circuit positioner
fiber
probe beam
TEST VECTOR
GENERATOR
b
test circuit
surface
a
Figure 2-8 a) Diagram of the constructed potentiometric scanning-probe
microscope [13] b) Pulse sampled heterodyne eletrostatic force
measurement of the local digital pattern [14]
The force can be deduced from the relationship between force and stored energy on
the capacitor as [14]
F-
2 dz
C( x, yjZ)[vp( t ) ~ v c(x,y,t) + m
2
(2-2)
The term A<b has been added to represent DC offset effects such as those due to
surface charge and material work function differences. A lock-in am plifier is used so the
probe deflection at f r only is m onitored, which can be evaluated as
14
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Az|/r = - ^ A ( ^ ) [ ( F D[/r- A $ ) - F r F m.
(2-3)
The capacitance derivative dC/dz also has to be evaluated to determine the sensitivity
of the instrum ent. G. E. Bridges and D. J. Thom son have addressed a numerical model
[7] to give the analysis o f variations of capacitance vs. distances.
By using the nulling technique in the instrum ent exact values for the terms C ’(x, y, z),
Q and k are not required, thus elim inating the need for com plex calibration and accurate
probe positioning. These term s become im portant though when noise, sensitivity, and
resolution issues are considered. Further, since the m ethod is independent of C(x, y, z)
m easurem ent o f passivated structures is possible without any m odifications to the
technique. This, however, comes with the penalty of a reduction in achievable spatial
resolution due to increased circuit-probe spacing.
The core o f nulling technique is to apply a sinusoidal signal o f frequency as much
close as the low est mechanical resonant frequency (os o r /, in Figure 2-8. By adjusting the
controllable param eter Vadj, which is either A or VT , the cantilever or the probe deflection
at 0 )s is nulled, the potential level at the internal tested point of the integrated circuit can
be determ ined (with a DC offset) as Vadj = VDUT - Ad>
R. A. Said and G. E. Bridges have given out the results on resolution [1 3 ], and one of
the test structures is shown in Figure 2-9 a), which consists of several interconnect lines
of 3 pm , 5 pm , and 10 pm in width. Cases both with and without a passivation layer on
top of the interconnect lines were studied. F or result shown in Figure 2-9 b) the substrate
was grounded and both interconnect lines were fixed at a +5 V potential. The cover
passivation was not present, with the probe located approxim ately 1 pm above the circuit
surface and scanned transversally across tw o o f the interconnect lines. Based on this
result, and using the half voltage level as the reference, the resolution of the instrum ent
for a probe height of 1 pm is approxim ately 3 pm . M easurem ents were also perform ed
15
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
with the presence of a 1 jam thick single passivation layer on top of the interconnect lines
(Figure 2-9 c)). It could be seen that the difference between the peaks and between the
m easured potential and the actual applied potential o f =5 V dc is greater than the case
when no passivation layer is present. Also, the substrate region between the interconnect
lines produces a higher measured potential. These differences are due to the increase in
the nonlocalized coupling effects caused by the larger separation between the probe tip
and interconnect lin e s ..
— i
iii®
S
>
>
a
o
a.
&
1
<-0
oC
8-
13C
P«o&e Position (pm)
w ees pomhon pm)
b
Figure 2-9 a) Interconnect test structure; b) Measured potential across
of the interconnect lines s h o w n in figure a along path (a); c) Measured
potential across three of the interconnect lines shown in figure a along
path (b) [13]
tw o
By introducing a pulse sampled heterodyne technique into the above m entioned
system, G. E. Bridges and R. A. Said, etc. have been able to internally extract the high
frequency bit-by-bit digital patterns on w afer [14] . A s shown in Figure 2-8 b), a m ulti-
16
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
channel test-vector generator supplies an N -bit patterns as indicated in Figure 2-10 a).
This N -bit patterns repeat with a period T = N/fb = 1/fo, where fb is the circuit clock
frequency. M eanw hile, the generator also provides a high-frequency single-bit sampling
pulse Vs(t) = G s ( t -
Tn).
The sam pling bit, as indicated in Figure 2-10 b), has a pulse
width 8 < 1/fb and can be shifted to a position xn along the N -bit test pattern. In this
manner, th e instrum ent will be able to m easure the logic level of the nth bit o f the pattern,
i.e., Vr(x, y,
t = Tn).
vc(t)
1/fb
(a)
vs(t)
Gs(l“ Xn}
-8
(b)
Y-l— t
%
Figure 2-10 a) Example of periodic digital waveform vc (x, y, t) to be
measured; b) ideal sampling pulse vs(t) = Gglt-v), which is shifted along
the digital pattern [14]
F o r very high speed applications, the nonideality of the sam pling w aveform would be
a constrain due to skew, finite rise and fall tim es, and possibly ringing. It requires that the
digital signals be reasonably flat so that the unknown signal voltage levels can be
determ ined when adjusting the param eter A to null the probe deflection. An 8-bit 1 M bit/s
digital pattern has been extracted w ith this technique from a CM OS test structure
consisting of four 10 pm wide interconnects [14] . D. Noruttun and G. Bridges have
achieved to be able to measure a 16-bit 125 M bit/s pattern stream with an im proved test
fixture [17] . These results are shown in Figure 2-11.
17
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
»uuT juuir
16
r--------------------- --------------- r-—T-•1
32
48
64
80
96
-
............
-
I
O
(------ " ------------------------- --------------- J
112
128
T im e (ns)
Figure 2-11 Measurement of 16- bit pattern at 125 Mb/s a) applied pattern
from generator; b) measured waveform using EFM [17].
2.2 Field Mapping and Imaging Techniques
2.2.1 Electro-O ptic Probing
Electro-optic probing is a technology developed based on the well know n electro­
optic effect. W hen an electrical field is applied across an electro-optic m edium the
electron distribution is perturbed so that the polarizability and hence the refractive index
of the m edium changes anisotroptically. As a result, the polarization o f the light passing
through the m edium is changed. This change which is linearly related to the electrical
field is know n as the Pockells effect and can be m easured optically. D ifferent types of
electro-optic probing system s have been developed in order to obtain m ore detailed
inform ation about the electrical field distribution inside the complex M onolithic
M icrowave Integrated Circuits (M M ICs). Internal (direct) and external electro-optic
probing systems help im proving electric m odels for sem iconductor devices and support
the developm ent of new integrated circuits.
In direct/internal electro-optic probing system , the device under test is fabricated on a
substrate that exhibits the electro-optic effect, such as GaAs. A pulsed and polarized laser
beam enters the M M IC from the backside o f the substrate. The electrical field in the
substrate changes the optical properties o f G aA s in such a m anner that the orthogonal
components o f a laser beam polarized in the y-z plane and travelling in x-direction get a
18
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
phase retardation of AF(y, z), which is solely proportional to the electrical field
com ponent Ex(x, y, z) being normal to the device surface (Figure 2-12 a)). The phase
retardation can be transform ed into an intensity variation AI(y, z) after the laser beam
passing a Pockells-cell arrangem ent which is relatively a little com plicated [18] [19] .
Usually the intensity variation AI(y, z) can be detected by photodiodes.
LongituOirtai
jyooe beam
Transverse
”oraM Ceam
Bectfo-OQftc
crystal
Figure 2-12 Common forms of electro-optic probing a) direct backside
probing geometry for a coplanar waveguide; b) external electro-optic
probing [5]
In contrast to the polarization analysis, there have been some reports about the FabryPerot enhanced internal probing [20] . The laser beam is focused on the M M IC from the
backside of the substrate inside which the Fabry-Perot resonance is present. This
approach requires an active wavelength control o f the laser beam to assure the FabryPerot resonance to happen. To detect the resonance, a pair o f focusing lens will be put in
front o f the reflected electro-optic signals from the M M IC. W hether the resonance is
observable or not, it depends on the choice of the num erical aperture o f the focusing lens
which is limited by the m axim um spatial resolution to be obtained.
For external electro-optic sam pling (Figure 2-12 b)), the device under test need not be
fabricated on an electro-optic substrate. Instead an electro-optic crystal is fused to the
sampling probe head, hence allow ing the electric fringing fields to penetrate the sam pling
crystal. The probe beam does not penetrate the device under test instead it is reflected off
19
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
a reflective coating on the bottom o f the crystal. The electrical field from the M M IC
around the bottom of the crystal will affect the polarization o f the laser beam. O perating
in either longitudinal or transverse probing m ode is dependent on the type of the crystal
utilized. An exam ple m easurem ent set up used shown in Figure 2-13 [21] . The optical
beam from a phase-stabilized Ti:Sapphire laser (pulse length 50 fs) is focused inside the
probe crystal. The reflected beam is analysed with respect to the change of the
polarization state. Due to the phase-locked-loop-electronics of the laser system, it is
possible to synthesize continuous wave (cw) signals from a m icrowave synthesizer to the
laser pulse train so that measurements in am plitude and phase can be performed. The
probes are fabricated from Bism uth Silicate (BSO) and Lithium Tantalate (L iT a03),
which
allow the determ ination of the norm al and tangential field com ponents,
respectively. The crystals have a footprint o f 90 x 70 p m for the BSO and 84 x 84 pm for
the L iT a0 3 . High spatial resolution is obtained by focusing the laser beam to a small spot
at the bottom of the probe. The distance between probe and D UT can be adjusted to be 57 pm . The m inim um detectable voltage is m easured to be around 1 mv and the sensitivity
is 40 pV /V /7z . The exam ined distribution netw ork circuit is a coplanar design fabricated
on high-resistivity Si with the designed working frequency o f 5 GHz and 15 GHz. Figure
2-14 and Figure 2-15 [21] dem onstrate typical electro-optic mapping of the electric
normal and tangential field com ponents for a CPW even m ode working at 5 GHz and
15GHz respectively in the transverse direction.
20
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
10 MHz Ref.
Ti:sapphire
synthesizer
Ref.
lock-inamplifier
50 f e '
polarizer
111
polarizer
microscopeobjective
photodiode
1/4
BSO or
LiTaO,
spectrumanalyser
DUT +
translation stage
micro wavesynthesizer
Figure 2-13 External electro-optic probe station [21]
100
50
an
v
*o
|
I*
0
0 .4
o 0.2
-50
-
Z
-100
0.0
-200
-100
100
0
Position (pm)
200
Figure 2-14 Electro-optic mapping of the normal field component of the
CPW even mode - solid line: normalized amplitude, dash line: phase [21]
21
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
150
100
0.6
0.4
-50
0,2
-100
-150
0.0
-200
100
-100
200
-200
Position (}«n)
Figure 2-15 Electro-optic mapping of the tangential field component of
the CPW even mode - solid line: normalized amplitude, dash line: phase
[21]
The electro-optic probing is a com plex
and expensive technology. But for
applications requiring high bandw idth and high spatial resolution, electro-optic probing is
a promising candidate. This is because it provides THz bandwidth and a spatial resolution
the size of the laser beam diam eter or even less. One issue which requires considerable
attention is that o f system calibration, such as fo r the internal Fabry-Perot electro-optic
probing system, the determ ination of the wavelength o f the laser beam to ensure the
Fabry-Perot resonance in each position of the M M IC due to the inconsistent thickness of
the substrate. A nother inconvenience is that fo r internal electro-optic probing system, its
application is lim ited only for testing D U T fabricated on special substrate that exhibits
electro-optic effect, such as GaAs. Finally, another area o f uncertainty is the possible
capacitive loading o f the electro-optic crystals such as L iT a0 3 in the external electro­
optic probing system. The perturbation o f the electric field above the D U T m ay affect the
normal operation of the DUT. In spite o f these difficulties, electro-optic sam pling shows
great promise as a high speed and resolution m easurem ent technique for a broad range of
devices.
22
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
M ore applications of the electrooptic probe are dem onstrated to diagnose the
m icrostrip patch antenna [22] and horn antenna for active am plifier array [23] . In
addition to the two-dim ensional m ovem ent o f the DUT, several optical com ponents have
freedom o f m ovem ent in the vertical direction, so that one m ay achieve a threedim ensional field-m apping capability. The m inim um detectable pow er em anating from
the patch is measured to be about 45 dBm, w here it corresponds to a field strength at the
probe o f approxim ately 30 V/m, and the sensitivity o f the m easurem ent system is 40 mV/
Hz. The results have been used to evaluate the performance o f the arrays and identify
im portant design issues. The high resolution and accuracy o f the electrooptic fieldm apping technique has proven to be advantageous in the design cycle o f am plifier arrays,
and has diagnosed problems such as nonuniform bias, nonuniform feed am plitude, and
m alfunctioning m onolithic microwave integrated circuits (M MICs).
A m ajor factor that has yet to be considered when applying electrooptic field mapping
techniques to the characterization o f active m icrowave circuits is the temperature
dependence o f the probe itself. The electrooptic coefficients that govern the response of
the probe to RF fields are known to vary with tem perature [24] . The flow of optical
pow er P through the sem iconductor is found to obey the tem perature dependence, 1/P °c
1 + kT, k is a constant related to the dim ension o f the sem i-conductor in the direction of
propagation. A system was developed [24] to sim ultaneously m easure the electric and
thermal fields utilizing an electrooptic sem iconductor probe (Figure 2-16). The Pockels
effect is em ployed within a gallium -arsenide probe to m easure electric fields, and the
effect o f photon absorption due to bandtail states in the sem iconductor is used to
determine temperature. By separating the electric field and tem perature signals in
frequency, both may be acquired with a single probe.
23
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Xa
Refefsmc®
Analyzer
EOCrysM JJ4
Potariabon
rTOO*
Diodft
#1
Gafts
w. b«9S
x Refactor
Pratoe
r*
|
Pass
Fife©?
Isolator
V80MH2
iThisrmal data)
m »t
tOMHte
RF DUT
X
rs»te
- '( B s c iric M
V
Pulsed
Laser
MMIC
A A Mcrosftip
™K»
’z m o c s s x
Low
Pass
Spectrum Lccfe-m
Analyzer Amp
d a te )
T
Electrical
Switch
3 MKz |
E x te rn a l!
R e f.
j
_ 'x r ± x j — '
x ;
Figure 2-16 An experimental setup for combined electrothermal
measurements [24]
Figure 2-17 clearly show s that there is a substantial difference between the behavior
o f the measured electric-field data obtained from the probe and the m easured power from
the independent pow er m eter. The explanation for the discrepancy is shown in Figure
2-18 as the absorption data from the probe is seen to decrease with time along with the
electric-field data. The change in the absorption signal is consistent with the expected
increase in temperature in the vicinity of the biased M M IC due to the dissipation o f heat.
Calibration of the tem perature effects o f the probe on the electric- field m easurem ents is
possible since the absorption signal is linearly proportional to the electric-field signal.
The temperature calibrated electric-field data is shown in Figure 2-19.
24
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
g 9.0x10 '
4 0.20
.£ 8.0x10
a£ 7.0x10
■fc 6. 0x 10
{Q
j 0.15 |
^O-O-
<5 5.0x10
0.10
2
j£ 4.0x10
tii
■o 3.0x10
- o — E le c tro -T h e rm # ! P r o b e
2 2.0x10
- » — P o w e r M e te r
fi 1.0x10
3
0
1
2
3
4
■0.05
SL
5
6
7
8
......
9
« nn
— ----
10
•2
11
T im e [min]
Figure 2-17 Probe and power-meter measurement of the MMIC [24]
7.0x 10
>
6.5x 10
£
6.0x 10
d
5.5x 10
a
5 .0 X 1 0
:
0.025
0.024 S
0.023:
- o — E le c t r ic - F ie i d D a t a
4.5x 10
-» _ »
- A b s o r p ti o n D a t a
g
c
Z>
4.0x 10' :
0 1 2 3 4 5 6 7 8 9
° 'x i .
V
0 .0 2 2 '
10 11
T i m e [m in ]
Figure 2-18 Probe-only measurements of the MMIC [24]
9.0X10
r ,-,-
,.r .
.
T, ,
020 2.
6. 0x 1O '
7.0x 1O '
0.10
4.0x 10'
3.0x 1O ' -
Q
2.0x 10' 1. 0x 10'
O
0.15
6.0x 1f f ’ 5.0x 1o '
XI
<t»
D
a.
§
TJ
O
€
%
s
fD
?
— o— C a lib r a te d P r o b e D a ta
.05 K<£tt
Z>
—-•*— P o w e r M e te r
••rrrx-p-r.-rr.-.-rp-rT-p'Tv
-i", -r"( v - ...................ir»'*-f.r r wr-rr-r-.-rw
0 1 2 3 4 5 6 7 8 9
<
o .o o * - *
10 11
T im e [m in ]
Figure 2-19 Temperature-calibrated electric-field data [24]
25
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
2.2.2 M odulated S cattering Technique
The m odulated scattering technique is a low cost technology of electromagnetic field
m apping o f m icrowave circuits. As a part of the basic operation of the m odulated
scattering system , a small dipole scatterer with a diode m ounted at the centre is placed in
the near filed of the circuit of interest. By m odulating the bias o f the diode at a frequency
much low er than the RF, a w eak m odulated scattered R F signal is returned to the
transm itter. The strength and phase o f the scattered signal are related to the square o f the
norm alized electric field distribution intercepted at the position o f the scattering probe.
B o n d in g
Pads
Chrome ThmFilm Resistors
Low Resistance
Schouky Diodes
Monopole
Scatterer
Figure 2 -2 0 The near field scattering probes; The dipole is used for
scattering the tangential electric fie ld and the monopole is used for
scattering the normal electric field [25]
Figure 2-6 [25] shows the design o f the dipole and m onopole scatterers. The probes
are fabricated on 125 pm thick low loss quartz. C hrom e resistive bias lines are used to
absorb any R F fields that may propagate along the bias lines and also serve as current
limiters to protect the diodes. These lines have a typical resistance of 100 Q/m m .
Comm ercially available low cost diodes are then silver epoxied between the dipole arms.
For monopole scatterer, tw o diodes are placed betw een the CPW ground planes and the
centre conductor. The diode has a 3 Q series resistance, 0.25 pF junction capacitance and
a 240 GHz cut-off frequency. The probes are fabricated using standard photolithographic
26
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
techniques w ith 1.0 jxm thick gold. The m onopole and dipole arms are both 100 (im and
are expected to have 40 dB dynamic ranges.
The general set up of the modulated scattering system [26] is shown in Figure 2-21.
In the R F section o f the near-field m odulated scattering experiment, part o f the R F signal
is sent as the LO to a w ideband quadrature m ixer as a reference. The other part is sent to
the input port o f the M M IC via a circulator w hich isolate the interference between the
reflected scattered R F signal and the input R F signal. U sually the power scattered to the
input/output port o f the M M IC by the probe is sm all, hom odyne m ixing is used to detect
the weakly m odulated signal. The in-phase and quadrature IF voltages from the m ixer are
detected by a lock-in amplifier. Through the use o f the absorptive R F switch, both the
scattered reflected and scattered transm itted w aves can be measured.
UNMODULATED
RF SIGNAL
r* MODULATED RF SIGNAL
WILKINSON
POWER DIVIDER
LOW FREQUENCY/fA
FUNCTION
CIRCULATOR
RF SOURCE
i
RESOURCE
REFERENCE
PLANE
^ UJ
DUT
REFERENCE
PLANE
QUADRATURE
MIXER
ABSORPTIVE
RF SWITCH
Figure 2-21 The m ic ro w a v e circuit electric fie ld imaging system using
m o d u la te d scattering probe [26]
The experim ental m easurem ents are verified by using a tw o inch long straight section
of 50 Q m icrostrip on RT/D uroid 5880 (er = 2.2, h = 0.015") at 9 GHz [25] . A monopole
probe is scanned over the microstrip line with 50 Q SM A term ination in the m atch mode.
27
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
The raw data collected from the in-phase and quadrature signal are shown in Figure 2-22
a) and b). The m axim a and minima along a transm ission line for the in-phase and
quadrature signal are separated by a distance A.eff/4 (A,eff = 26,000 pm ). The normal
electric field intensity and phase delay m easurem ents are display in Figure 2-22 c) and d),
respectively. The intensities are nearly constant along contours parallel to the microstrip.
An intensity ripple o f 0.1 dB is present which is due to the non-ideal match o f the
term ination or non-ideal connections m ade with the SM A connectors. The phase cycles
by
2 tc
every 13,000 pm which is A.eff/2 as can be seen in Figure 2-22 d). T he slight angle
in the phase is possible due to a variation in the height o f the probe across the microstrip.
Some m easurem ents were also conducted to a three stage coupled-line bandpass filter at
10 GHz [26] [27] and a distributed am plifier M M IC at 13.5 GHz [26] by the m odulated
scattering probing.
As a low cost and flexible electrom agnetic field mapping system, m odulated
scattering system is capable o f m apping the norm al and tangential electric field intensity
and phase delay above a M M IC with any substrate in the frequency range of 0.5 to 18
GHz. The spatial resolution is limited by the dim ension of the m onopole/dipole probe. It
is believed that the resolution can be pushed to 25 pm with sm aller and thinner probes
before the system becomes noise limit. F urther m ore by using a system o f circulator and
absorptive R F switches the bandwidth and the signal to noise ratio can be greatly
improved.
28
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
I=In Phase Signal
- m — 1 'T 'i
0
1
2
3
/
Q=Quadraiure Signal
Raw Data
i \— i— i— r
4
5
6
7
8
9
Position (/500 microns)
Position (/500 microns)
t v rvT
- 1 0 9 - 8 -*7-6 - 5 - 4 - 3 - 2 - 1 0 1 2 3 4 5 6 7
8
9 10
Linear Magnitude - Arbitrary Units
20 Logf iHTQTj
Electric Field Intensity (dB)
c.)
1 2
3
4
3
6
7
d)
Tan-'fO/T)
Electric Field Electrical Phase Delay
1
8
Position (/500 microns)
Normal Electric Field Intensity (dB)
-20 - t l
-16 -14
12
.10
2 3 4 3 6 7 8 9
Position 1/500 microns)
Phase (Degrees)
TT
-ISO -120 - »
-60 -30
o
30 6(1 90
120 ISO
Figure 2-22 a) In-phase signal of normal electric field; b) Quadrature
signal of normal electric field; c) Normal electric field; d) Normal electric
field phase delay [25]
2.2.3 P assiv e E lectrom agnetic Probing
2.2.3.1 Double/Single Loop Magnetic Probe
The design of the m agnetic field current probes started using conducting loops. It was
assum ed that the loop is im m ersed into a plane linearly polarized electrom agnetic w ave
in free space and is placed in the x-y, x-z, or y-z plane, perpendicular to the magnetic field
Hz-, Hy-, Hx- com ponents, which are considered to be constant over the area of the loop.
29
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
The system resolution is limited by the dim ensions o f the loop probe which should be
much sm aller than the wavelength.
A double-loop magnetic probe has been developed by S. S. Osofsky [28] with the size
o f 15 mm x 7 mm for each loop and physical form o f magnetic quadrupole (Figure 2-23).
Its field configuration matches the fields o f both m icrostrip and CPW , as shown in Figure
2-24. It was a large-scale model which can be easily constructed and tested in the 0.1-0.3
GHz range. The ultim ate goal, however, is construction of probes useful at 20 GHz or
higher by means of microfabrication technology. The advantage of this double-loop
configuration is:
•
For both microstrip and CPW , the w aveguides’ magnetic field goes up
through one loop and down through another, their contributions add.
•
Signals induced in the tw o loops by a nearly uniform field com ing from a
distant source, such as another w aveguide in the microcircuit, tend to
cancel.
The m ajor concern o f this m ethod how ever is the electric asymm etry of the probe
central conductor. That means the probe also works as an electric field probe in the
perpendicular direction. This asymm etry disturbs the magnetic m easurem ents, especially
in the position where the magnetic field is m inim um and the electrical field is maximum .
T o S p e c tr u m A n a ly z e r
Loops
Figure 2-23 Double loop magnetic probe [28]
30
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
L oops
<b)
(c>
Figure 2-24 Double-loop magnetic probe coupling to magnetic fie ld s of a)
microstrip; b) coplanar; c) interacts between two loops [28]
Variable
Phase Shifter
Combiner
Probe
Circuit liixSct Tcsi
Figure 2-25 Measurement configuration [28]
A typical m easurem ent configuration is shown in Figure 2-25. The current induced in
the probe passes through the CPW transm ission line to a spectrum analyser. B y this
means, measurements o f current am plitude can be made. To make phase m easurem ents, a
reference signal of the same frequency and adjustable phase is added to the probe signal.
The reference signal’s phase is varied until the am plitude seen by the spectrum analyser
31
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
is m axim um . The phase at the point of measurem ent is then com pared with that at some
other point of the circuit chosen as the reference. Thus the inherent phase delays o f the
probe and its transm ission line are of no concern.
60 00 .
55 (X)
• g 45 ( XU
40 00 h
35 00 r
-6QM
-55.00
-5000
■4500
-4000
-3500
-3000
Position Along Microstrip [cm]
(a)
70 00
65.00
E x p e c te d
60 00
Averaged
£ , 5 5 00
2f> 50 00
40 00
35 00
-60 00
-5500
-5000
-30 00
Position Along Microstrip [cm]
(b)
Figure 2-26 Comparison of the expected and measured standing wave
along a microstrip having an SWR = 1.36 - a) not improved; b) improved
by the probe-reversal technique [28]
The m easurem ent results of the probe signal along the m icrostrip are shown in Figure
2-26 a) and b). The current m aximum and m inim um are cycled along the m icrostrip. The
transverse patterns of an open-circuit m icrostrip taken at the current m axim um and
minimum points are shown in Figure 2-27 a) and b). These m easurem ents were made at
280 M Hz and the m icrostrip had a center conductor width of 5mm and a dielectric with er
32
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
= 12 and thickness o f 6.35 mm. The transverse pattern at a current minimum is quite
different from that at the current maximum. This observation reveals that there is an
unwanted contribution to the observed signal arising from electrostatic pickup. This
contribution is usually small, but at a point on a transm ission line corresponding to a
current m inim um o f the probe output, the magnetic field is reduced and the electric field
is increased, this causes the effect o f the electrostatic to appear. However, the error from
this source can be greatly rem oved through the expedient o f m aking two m easurem ents at
the point o f interest, between which the position o f the probe is rotated around a vertical
axis through 180°. By averaging these two m easurem ents, the error caused by parasitic
electric field coupling is greatly reduced (Figure 2-26 b)).
•65oo|~
aTE3
I
JS
A.
00 0
500
-500
tamo. A™, Mmmp tanl
a
F ig u r e
0.00
p ,,^
5(5
0
MioMtripIml
b
2-27 T r a n s v e r s e pattern of an open-circuited microstrip - a) taken
at a current maximum; b) taken at a current minimum [28]
Due to the electric asymm etry problem in the double-loop magnetic probe,
Grzybowski and Bansal [29] designed a half-loop m agnetic field probe in which a 1 mil
gold bond wire was connected to a Cascade M icrotech W P H -102-250 w afer probe test
head. Because the fabrication procedure is very com plicated and additionally the wafer
probe test head is expensive, it seem s not to be suitable for practical applications. In the
33
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
w ork m ade by Y. Gao and I. W olff [30] a single loop square m agnetic field probe was
designed (Figure 2-28). The square loop of the probe is 710 pm by 710 pm in size and its
conductor width is 55 pm. It is an inexpensive one which also solves the electric
asym m etry problem. But for both m icrostrip and CPW, the w aveguides’ magnetic field
goes up and down through the same loop and their contributions cancelled. Furthermore,
signals induced by a nearly uniform field com ing from a distant source can’t be avoided.
(a)
Probe
(b)
Figure 2-28 Single square loop magnetic probe - a) Probe for Hz magnetic
measurement; b) Probe for H* and Hy magnetic measurement [30]
The com parison o f m easured and calculated Hz- am plitude along and across the
shorted m icrostrip is shown in Figure 2-29. The m easurem ent w as made at 20 GHz. We
can see that the m easured and calculated signals agreed well. B ut there is a notch in the
amplitude curve o f Hz- perpendicular to the shorted m icrostripline at the center point.
The currents are cancelled while the m agnetic field goes up and down through the same
loop.
34
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
-20
-20
*23
-30
I
s
<N
-45
-50
-50
0
5000
2000
10000
X (pm)
mo
sooo
7000
8000
Y{ h*b>
Figure 2 -2 9 Comparison of measured and calculated squared-amplitude
of Hz - a) along the microstrip; b) perpendicular to the microstrip at
current maximum [30]
M agnetic field probe is a relatively simple device fo r circuit testing. The fabrication
o f reliable small scale probe for operation at high frequency is a m ajor issue. One o f the
difficulties is in establishing good electrical contact between the loops and CPW feed
which are right angled to each other. Another issue o f uncertainty is the electrostatic
coupling. If both o f these problem s solved, the magnetic probe could be a viable
diagnostic tool.
2.2.3.2 Passive Electric-fieid Probe
As discussed in the above section, the passive m agnetic-field probes appear to be
incompatible to strong radiating circuits, e.g., antennas and circuits on substrates with
small dielectric constant substrates. This made people com e back to the eclectic-field
probes. The early passive coaxial electric-field probe w as designed in 1980 by J. S.
D ahele [31] which used the central conductor of a coaxial cable to capacitively couple to
a microsrtip under test. W ith the developm ent of the m icrofabrication technology, the
probe can be made smaller. A m iniature probe designed by Y. G ao and I. W olff [32]
consists of a m iniature coaxial line with 508 pm outer diam eter and an inner conductor
diam eter of 112 pm. The inner conductor extends 300 p m beyond the outer conducting
shield, as shown in Figure 2-30 a). Coaxial electric-field probe is usually used to m easure
35
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
the electric field in the z-direction perpendicular to the substrate and is nam ed EPZ, while
the dipole electric field probe EPX Y shown in Figure 2-30 b) is designed for tangential
electric field m easurem ent. The dipole electric field probe EPX Y consists o f an electric
dipole and a coplanar transm ission line with 5042 impedance.
Transmission line
b
c
Figure 2-30 Electric field probes - a) EPZ; b) EPXY with transmission
line; c) expended EPXY without transmission line [32]
Some m easurem ent results using these electric field probes w ere discussed [32] . The
measurement results o f a bandpass filter with probe EPX Y is shown in Figure 2-31. It is
found when the probe enters the filter, the coupling effects between probe and the filter
change the reflection o f the filter (Figure 2-32). The S n changes in the passband are
about 7.0 dB. In order to im prove the accuracy o f the m easurem ent, a sm aller probe is
needed or set the probe in a higher position, but both of them will reduce the sensitivity
o f the probe and further m ore the second solution will also reduces the resolution o f the
measurement.
36
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
J-
i= = X
0
4000
SOOO
12000
16000
20000
X (jug)
Figure 2-31 a) The layout of the bandpass filter, and b) 2-D representation
of the measured squared field strength |EX| 2 in the passband at 12.8GHz
[32]
-10
24)
O
B
5a -30
e.
OD
e
-70
0
4
8
12
16
20
Frequency (GHz)
Figure 2-32 Measured scattering parameters for the bandpass filter [32]
A nother factor that m ust be taken into account is the probes’ influence to the D U T
when the probes are positioned near the D UT since the m easurem ent principle o f the
electric and magnetic near-field probes is to couple the electrom agnetic near field of
D UT to the probes. Further research w ork was done on EPZ probe (Figure 2-30 a)) to
reveal a new calibration technique [33] for application o f near-field probing. By using the
finite-difference tim e-dom ain (FDTD) m ethod, the probe shown in Figure 2-30 a) was
theoretically analyzed to determ ine its most sensitive probe segment. Taking the
37
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
am plitude of the normal electric field at this segment as a known field, the probe is
calibrated by defining a perform ance factor (PF), which is the ratio of the known field
am plitude to the probe signal amplitude. PF, which is derived as a function of frequency,
is shown in Figure 2-33. U sing the fitting curve of the PF as a calibration curve, the
m easured and calculated normal electric-field distribution over a cross section of a
m icrostrip line at 11.8 GHz is shown in Figure 2-34. Com paring the calculated results
with m easured results, the agreem ent is good.
W ith a calibrated probe, Y. Gao, etc estim ated the coupling effect between the probes
and the D UT, a microstrip line. The change of the scattering param eters S21 of the DUT
w as m easured with an N W A during the m ovem ent of the field probe perpendicular to the
line direction at a height of 50 pm above the microstrip line. It can be seen on Figure
2-35 that the influence of the probe on S21 is very small, about 0.05 dB. If the field probe
is placed higher than 50 pm above the D U T, the influence o f the probe is sm aller and can
be neglected.
140
120
100
PF
SI 1
E
sa3
T
fit.
Oh
-20
-40
0
2
4
6
8
10
12
14
16
18
20
Frequency (GHz)
Figure 2-33 The PF and reflection coefficients S ll of the coaxial electricfield probe [33]
38
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Measured
■•
Calculated
0 -10
0
2000
4000
6000
10000
8000
Y (pm )
Figure 2 -3 4 Comparison between the measured and calculated normal
electric field at a cross section of the microstrip line at 11.8 GHz. (
Substrate: A1 2 O3 , £r= 9.8 , h = 6 3 5 pm, w = 700 p ro ) [33]
f - 0 .4 7 5 ^
E-0.485 ^2!
-0.495
•0.505
16000
12000
Figure 2-35 The investigation of the influence of the probe on the
microstrip line [33]
39
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Chapter 3
A u t o m a t e d N e a r - f ie l d S c a n n in g
S ystem
An autom ated m easurem ent system is m andatory for investigating the feasibility of
various electrom agnetic near-field probing techniques. The hardw are and software o f the
system m ust be flexible in order to facilitate testing on w ide range o f applications. Since
experim ents may require jim or m m scale resolution depending on circuit dim ensions,
there has to be tw o sets o f hardw are apparatus. The corresponding probe design and its
characteristic dim ensions m ust com plim ent the application. The software required must
be user friendly and capable o f driving com m only used m easurem ent instrum ents, i.e.
N etw ork Analyzer, Spectrum A nalyzer and Oscilloscope. This chapter will introduce the
fundam entals o f near-field scanning system and describe the hardware and software
required for the system.
3.1
Fundamentals of the Scanning System
Network Analyzer
probe
received
p o rt2
transm itted
portl
V s tim u lu s
OUT
}i
_ {
S y ste m m e a s u re d b y V N A
!
Figure 3-1 Diagram of Near-field Scanning System
40
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Figure 3-1 shows the diagram o f a near-field scanning system. To analyse the
functionality and perform ance o f a device under test (DUT), it is required to find the
transfer function Tcircuit of DUT. As shown in Figure 3-1, given the circuit input signal
Vstimulus, the Tdrcuit can be calculated by sampling voltage signal V(x, y) at test point (x, y).
V ( x , y ) = Tnraiil( x ,y ) - V stimulus
(3-1)
The traditional contacting measurement m ethods have the probes directly touch on
the surface o f D U T to get the voltage signal V(x, y) at each test point. The near-field
probing concept is not intending to do a contacting m easurem ent on the DUT. In contrast,
by using a N etw ork A nalyzer to measure the S 21 param eter o f the system under test, as
shown in Figure 3-1, it derives the transfer function o f D U T from probing the
electrom agnetic field distribution above the surface o f DUT. Equation (3-2) shows how
the S 21 param eter is measured.
f
\V rfP / V received
S 2J {m ea su red ) = - V™ U = 201og,0 1
^
j tra n sm itted | z v transm itted
^ rtransm itted
/o
where Vtrammit,ed is the transm itted signal on port 1 o f N etw ork Analyzer, and
Vreceived is the signal received by Netw ork A nalyzer on port 2. Equation (3-3) shows the
relationship between Vlranmtitted and the input signal to D U T Vstimuius, whereas Equation (34) shows the relationship between VreceiVed and the induced signal picked up by probe Vinci.
^ s tim u lu s
( ^ )
^ tr a n s m itte d
(* ^
''''
(3-4)
^received
By substituting (3-3) and (3-4) into (3-2), S 21 can be form ulated as
41
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
S 2t (m easured ) =
V.received
l»
= r„cablel
Tcablel
■
ind
transmitted
(3-5)
»j
stimulus
As it w ill be discussed in the following section,
the induced voltage signal by
probe, can be form ulated as
V fcd
= T probe
(3-6)
-V(X,y)
Substitute Equation (3-6) into (3-5),
S 2 1 (m easured ) - 7 ^
•7 ^ , • r prote
E (x ,y )
(3-7)
V.stimulus
Given Equation (3-1),
Tcircuit (x
^ ? /v xW
S 21 (m ea su red ) = Tcnble2 - T ^ • Tprobe ■
V.stimulus
stimulus
(3-8)
^ c a b le l
^ c a b le l
tp r o b e
^c irc u it ( ^ > 3 0
The transfer function of D U T is then obtained from Equation (3-8),
S 21(m easured )
Tarru,A^y) = T
1 cable2
.T
1 crtWel
(3-9)
.T
1 proto?
In Equation (3-9), S?i, Tcabiei and
can be obtained by measurement. It is
expected to create a circuit model for the probe being used for near-field scanning. If the
transfer function Tpmbc is derivable from modelling, the idea described above is then
feasible. The following section will briefly discuss the m onopole probe modelling.
42
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
3.2 Approximate Circuit Modelling of the Monopole
Probe
In Equation (3-9), the transfer function of probe Tprot,e has to be found so that TCirmit
can be then calculated. Although an exact theoretical model is unavailable P.G. Frayne
[37] [38] has developed a prelim inary model to investigate the interaction between a
m onopole probe and an isolated conducting strip of finite w idth and infinite length.
50 Q
ind ----[
50Q
ind
—
lb
V(x,y)
Figure 3-2 An approximate monopole probe model and the corresponding
equivalent circuit model [5]
Figure 3-2 shows the model o f m onopole probe and an approxim ate equivalent circuit
on the right hand side [5] . The model assumes that the probe load impedance is much
greater than local circuit im pedance. Cci is the capacitance form ed between conductor of
probe and transm ission line circuit, CH is the capacitance form ed between probe
conductor and shield, C,/ is the capacitance form ed betw een shield of probe and
43
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
transm ission line circuit, Ccg is the capacitance form ed between probe conductor and
ground plane, Csg is the capacitance form ed between probe shield and ground plane.
Given that probe shield is at ground, Csg is large com pared to Csi, a sim plified
equivalent circuit can then be drawn as Figure 3-3.
Cd
-O
V(x,y)
Vt„d
r
C,
5 0 £2
U
Figure 3 -3 Simplified equivalent circuit for monopole probe a s s u m in g
1
Z //
jco(Csc + C„ )
VJ=V(*,y)~
z 0 //
= V(x, y)-
1
1
•+ jo )(C sc + Ccf, ) jcoCc!
______________Z 0__________
1 + Z 0j(o (C s< + CCK) + jcoCrlZ 0
for frequency w here l/jco (C sr+Clg) »
Vw = V ( x , y ) -
(3-10)
Z q, or ja>(Csc+Ccg) Z q «
j(oC t.[Z 0
(3-11)
1 + jcoCrlZ 0
Provided that jco Cc{Zo «
1
1
(3-12)
v ,nd = j a C , Z 0V (x, y)
44
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
From Equation (3-6) and (3-11),
Tprobe
can then be express as
Tpmbe = jcoC ,,Z0
(3-13)
By using a m ethod o f mom ent (M OM ) solution, the coupling capacitance Cci can be
calculated. W ith the potential specified as 1 Volt, T. Forzley gave a plot of Cci for various
heights h o f the probe above the transm ission line and incremental displacem ents relative
to the center o f the transm ission line [5] . As shown in Figure 3-4, for example, Cci is on
the order o f 5 fF for a 1/16" OD and tip length o f 0.5 m m probe at a height o f 0.5 mm
above the transm ission line.
7-i
—
—
—
—
r
e,»1.0, h-0.5 mm
e,»2.2, h«0.5 mm
£,*2.2, h»1.0mm
£,-2.2, h»1.5 mm
5-
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
Probe displacem ent relative to center of line (mm)
Figure 3-4 Calculation of the critical coupling capacitance Ccl (CH) [5]
Some fundam ental properties of the m onopole probe have been discussed [37] [38]
[5] . Firstly, the spatial resolution of the probe is not purely a factor of the coax outer
diameter. Instead, the operating frequency and standing wave ratio on the structure must
also be considered.
45
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Secondly, the sensitivity of the probe is directly proportional to the tip length t. As t
increases so does the num ber of electric field lines term inating on the center conductor,
hence a g reater potential difference betw een the conductors is induced. On the other
hand, the resolution as a function of tip length is difficult to calculate and cannot be
readily quantified. The most tactical approach to quantify the resolution w ould be to
incorporate both an experimental and theoretical analysis on probes of various sizes and
tip lengths.
Finally, the net probe response is due to both capacitive and m agnetic coupling. The
m agnetic field coupling is due to the transverse m agnetic field com ponent which is
proportional to the axial current density. Ideally, the m onopole probe should respond only
to electric field coupling, hence it is the natural dual o f the loop probe. In contrast, while
the m onopole probe m easurem ent error arises from magnetic field coupling, the error in
the loop probe m easurem ent is due to electric field coupling.
The usable bandwidth o f the m onopole probe is a function o f its dim ensions. Both the
tip length and outer diam eter must be a small fraction of the w avelength in the medium
being m easured. This restriction is necessary to avoid any spurious radiation by the
probe. The m easurem ent error incurred as a result o f undesirable m agnetic field coupling
is another limitation.
3.3 Scanning System Hardware Description
T he m easurem ent system shown in Figure 3-5 consists o f an H P8753E Network
A nalyser (for up to 6 GHz m easurem ent) or a W iltron 360 N etw ork A nalyzer (for up to
40 GHz measurement), a host PC, and a tw o-axis positioning stage. T he com puter acts as
the central controller which m anipulates the positioning state by RS-232 serial link,
conducts the Network A nalyzer operation and subsequent data acquisition via General
Purpose Interface Bus (GPIB or also com m only known as the IEEE- 488 standard).
46
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
HP8753E
Network Analyzer
Probe return
Test signal
GPIB
DUT
Two-axis Positioning Stage
Computer
Workstation
Serial port
Figure 3-5 Measurement system hardware
Tw o different positioning
stages were developed for targeting on
different
applications w here the D U T areas are varying and different resolutions are demanded.
F or applications requiring mm or sub-m m sam pling intervals, an H P7220C plotter has
been proven to be reliable and versatile. The scan area o f the plotter is lim ited by the
design o f the probe support. Currently scans on the order o f 15 cm by 28.5 cm with a
m inim um step resolution of 25 pm . For probing on tiny scale circuit, such as M M IC onw afer testing, where pm to sub-pm resolution is required, a D C m otor driven table was
alternatively used. Tw o lim itations of using D C m otors are backlash and speed. If the
backlash of each m otor is known it can be com pensated for in software. The minimum
step size of the m otors being currently used is 0.1 pm for each axis capable of 25 m m of
displacement.
47
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Figure 3-6 Plotter mounted with monopole probe for mm resolution scans
Figure 3-6 shows the test platform for m m level measurem ent of m icrow ave devices.
The probe is m ounted on the plotter which is controlled by a computer. The scanning
resolution o f the plotter is of 25 pm.
Figure 3-7 shows the apparatus required for p m resolution scans. The monopole
probe is m ounted on a plastic beam that is fixed on the DC m otor driven table. W hen
using the m onopole probes whose dim ensions inhibit direct connection to a SM A
connector some ingenuity is required. A m iniature coax (D0 = 0.32 m m and Z>, = 0.05
mm) is soldered directly to a large coax (D (>- 2.2 m m and D, = 0.51 mm ) which in turn
directly mates with a SM A connector. F or slightly large coax it is possible to use an
intermediary
m icrostrip
transm ission
line
to
construct
a
coax-m icrostrip-SM A
configuration. To investigate the effects o f the discontinuities introduced to the
transmission line, tim e domain m easurem ents can be performed.
48
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Figure 3-7 Close up of DC motor test station used to scan pm circuits
Figure 3-8 Setup of the Scanning System for pm level measurement
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Figure 3-8 shows the whole setup for the com puter controlled scanning system. The
DC m otors is connected to the com puter via a 9-pin serial port. The Netw ork Analyser is
com m unicated with the com puter through GPIB using an IEEE-488 cable. The test signal
from the N etw ork A nalyser is the input to the device under test, the feed back signal from
the probe is returned to the N etw ork Analyser. The com puter controls the movement of
the DC m otors and the operation of the Netw ork Analyzer, collects, processes and
displays the m easurem ent data by the softw are program m ed using L ab VIEW .
3.4 Software Development with LabVIEW™
3.4.1 Introduction of LabVIEW
The introduction of LabV IEW pioneered a new instrum entation approach called
virtual instrum entation. Virtual instrum entation em powers users to build their own
instrum entation systems with standard com puters and cost-effective hardware. These
softw are-centered system s leverage off the com putational, display and connectivity
capabilities of popular com puters to give users the pow er and flexibility to build each of
your instrum entation functions.
LabVIEW is a revolutionary graphical program m ing developm ent environm ent based
on the G program m ing language for data acquisition and control, data analysis, and data
presentation. LabVIEW features with the flexibility of a powerful program m ing language
without the associated difficulty and com plexity because its graphical programm ing
methodology is inherently intuitive to scientists and engineers.
LabVIEW is a m ulti-platform , available for W indows 9x/N T/2000/, Mac OS, Sun,
HP-UX, and Concurrent Pow erM A X . D ata can be acquired from thousands devices,
including GPIB, VXI, PXI, serial devices, PLCs, and plug-in data acquisition (DAQ)
boards. By using the LabV IEW inter-application com m unication tools such as ActiveX,
dynamic data exchange (DDE), and structured query language (SQL) database links,
50
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
remote data sources can be accessed via Internet. LabV IEW program can be called from
other environm ents. It can also call existing code in the form of a dynamic linked library
(DLL) on w indow s or a shared library on other O perating Systems.
Users can easily find and download almost all the drivers of m easurem ent equipment,
such as oscilloscope, logic analyzer, netw ork analyzer, spectrum analyzer, etc. in their
labs from either equipm ent makers or National Instrum ents, even from some other users
who have developed the drivers for the instrum ents. It em powers users a faster way to
program instrum entation, data acquisition and control system, to build their own
solutions fo r scientific and engineering systems. It gives the flexibility and performance
of a pow erful program m ing language w ithout the associated difficulty and complexity.
This will significantly reduce the time to develop a m easurem ent system, and make users
focused on their m easurem ent itself instead o f system development.
G program m ing language is the heart o f the LabVIEW . Like C or BA SIC, G is a
general purpose program m ing language with extensive libraries o f functions for any
program m ing task. G also includes conventional program m ing debugging tool, user can
set breakpoints, anim ate the execution to see how data passes through the program. G
programs are called virtual instrum ents (V is) because their appearance and operation can
imitate actual instrum ents. A VI consists o f an interactive user interface, a dataflow
diagram that severs as the source code, and icon connections that set up the VI so that it
can be called from higher level Vis.
The interactive user interface of a VI is called the front panel, because it simulates the
panel of a physical instrum ent. The front panel can contain knobs, pushbuttons, graphs,
and other controls and indicators. Users enter data using a m ouse and keyboard, and view
the results on the com puter screen.
51
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
The V I receives instructions from a block diagram, which user constructs in G. The
block diagram is a pictorial solution to a program m ing problem. T he block diagram is
also the source code o f the Vis.
V is are hierarchical and modular. U ser can use them as either top-level program s or
subprograms. The icon and connectors o f a VI work like a graphical param eter list so that
data can be passed between Vis.
W ith these features, G makes the best use of the conception o f the modular
program m ing. Each subVI can be perform ed and debugged by itself, this makes
program m ing much easier.
3.4.2 The LabVIEW™ T est Program for C om puter Controlled
S can n in g System
A series o f LabV IEW program s have been developed for controlling probe
m anoeuvring and data acquisition. T he program s are the third generation of the
automated scanning system software which is im proved over the second generation [5] .
It provides the way to easily integrate all hardware controlling involved in the
m easurem ent into a single graphic user interface, i.e. netw ork analyzer, DC motor,
plotter, spectrum analyzer, etc. The 2nd generation software EM D A P (ElectroM agnetic
Data A cquisition Program ) was developed based on the Turbo Pascal and provided the
control o f the scanning system by the com m and line interface on D O S Operating System.
The current software is designed with a graphical user-interface w hich m akes full use of
the control com ponents in the LabV IEW control palette (Figure 3-9).
Through GPIB (IEEE 488) interface, com m unication between the com puter and the
N etw ork A nalyser is established. All the features o f the N etw ork A nalyser can be
remotely accessed by the LabV IEW program in the Netw ork A nalyser Setup control box.
The test frequency can be specified as the centre frequency and the scanning span and the
data format can be log/linear m agnitude, phase, smith chart, polar and standing wave
52
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
ratio. U ser can also select the m easurem ent o f S-param eters, R-param eters or A/B
param eters. V ia the CO M 1/CO M 2 serial port, the D C M otor or plotter can be controlled
for pm or m m resolution scanning, respectively. In the DC M otor/plotter Control box,
user can define start point, end point and scanning step.
rlQi *>
.tn .v i *
F i»
EHl
O p a a te
Tocls
a -c w s e
ggndaw
'MS Mwior Control
Kelp
touu
s e r ia lp o s - i
^
n S .sggpy
' | -96,00i t .1
106.00
“
Network Analyzer S e tu p '■ 9-«S£+*
m easu rem en t
(0 : S I 1)
'
9 9*E+a
9.W5E+S
i.floce+9
ijmie+9
l.ooiEto
Tran*?.; F o r w a r d S 2 1
f o r m a t ( 0 : to g m a g )
III o q m a g n i t u d e '
fo r m a l. ( 0 : to g r n a g ) i o p y
center
span
I
mm-
o.ooJOO.O0
200.00-■
9.935E t -6 9. 99QF+S
9 999F-4-S
Data File Management
d e f a u l t p a t h ( - s a d d - is c ro tim )
i .oeh
200,00
l ,« M E
-»9
1.0G1E +9
01F i 9
i .O
FOOTE-I
WHKHm
We nnm e ■ . ■
■ '
Figure 3-9 Frontpanel of the LabVIEW test program for Computer
Controlled Scanning System
A real-time display o f the spectrum magnitude and phase along with the automatic
data filing is also provided in the LabV IEW program. A sum m ary and com parison o f the
capabilities between the tw o generations of softw are is provided in Table 3-1 Sum m ary
and comparison o f the tw o generation o f software.
53
R ep ro d u ced with p erm ission o f the copyright ow ner. Further reproduction prohibited w ithout perm ission.
Table 3-1 Summary and comparison of the two generation of software
Communication
Scans
Real time
display
Resolution
Program
G PIB & COM
1D&
2D
Magnitude
& phase
mm & pm
Command
line
Graphical
user
interface
EMDAP
Yes
Yes
No
Yes
Yes
No
LabVIEW :
1 DumScan.vi
2DumScan.vi
2Dmm Scan.vi
Yes
Yes
Yes
Yes
No
Yes
Control
54
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
C h a p te r 4
N e a r - f ie l d S c a n U s in g M o n o p o l e
C o p l a n a r W a v e g u id e P r o b e s
and
To further expand on previous m easurem ents [5] , more experiments w ere conducted
to illustrate the value o f the m onopole coaxial probes as m icrow ave circuit diagnostic
tools. Tests on applications ranging from coplanar w aveguide transm ission lines, single
patch m icrostrip antenna to m icrostrip antenna arrays have been done. As an alternative,
the C oplanar W aver Guide (CPW ) probes are also fabricated and tested. Some results
com paring m onopole probe with C PW probe is presented. D epending on the application
the sam pling interval m ay be in the m m to pm range. Typically, 50 to 100 field data
samples p er observation point are taken to m inim ize the noise effect to the m easurem ent
results.
4.1
Monopole Probes to Measure 5 0 0 CPW T- Line
Ground Lines
Signal L ine
Figure 4-1 Structural diagram of a coplanar waveguide T-line
55
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
The use o f coplanar waveguides (CPW ) in m icrowave integrated circuits is very
popular since it adds flexibility and im proves the perform ance o f some circuit functions
at high frequencies. The CPW architecture allows for the mounting of lum ped
com ponents in a shunt or series configuration, while elim inating the need for drilling
holes or slots through the substrate. Figure 4-1 shows the structure of a coplanar
w aveguide transm ission line circuit, w here w is the width o f the signal line, t is the
thickness o f the signal line, h is the substrate thickness. er is the relative dielectric
constant o f the substrate, and een is the effective dielectric constant o f the substrate.
4.1.1 C oplanar W aveguide T ransm ission Line A nalysis
y
*
._z_:
substrate
Figure 4-2 Electric field variation above CPW T-line and measurement
with monopole probe
Figure 4-2 is a diagram that illustrates the electric field variation over the surface of
CPW T-line. It also shows the m onopole probe displacem ent at different spot when it is
56
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
scanning transverse (x direction, position 1, 2, and 3) and longitudinal scanning (z
direction, position 1 and 2 ’) over the CPW transm ission line.
W hen it is in TEM m ode the distribution of electric field over the surface of CPW
transm ission line can be sum m arized as
E ( x , y , z ) ~ E T( x , y ) e ±n
(4-1)
w here E j only has x and y com ponents and the z-variation is of exponential form.
So E t should have same pattern at any given z expect for a phase and m agnitude change.
The electric field distribution over CPW T-line then can be discussed separately.
Tw o different CPW T-lines were scanned with m onopole and CPW probes. Their
geom etries and param eters are shown in Figure 4-3 and Table 4-1 respectively. The two
transm ission lines will be referred as Type A and Type B transm ission lines in the
following measurement.
Table 4-1 Parameters of two tested CPW transmission lines (Figure 4-1)
Type A
Type B
68.9*
2.54
49.2
(pm)
330
100
w (pm)
h (pm)
740
580
17
250
1.87
4.97
Z0 (Q)
£r
s
t
(pm)
£ e ff
9.8
280
5
*It is noted that Z 0 o f type A C PW calculated by P C A A D w as not close to 50 Q as expected. T his is
probably due to the change to the substrate property in the etching process. T he T ype A CPW is not a
com m ercial but a house-m ade product.
57
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Type A. s = 0.3 mm w = 0.7 mm
Type B: s = 0.1 mm w = 0.25 mm
(a)
(b)
Figure 4-3 a) Structural diagram of CPW T-line; b) Type B CPW under
test, s = 100 pm and w = 250 pm
4.1.1.1 Analysis of Transverse Electric Field Variation
An im portant property of TEM w ave on transm ission line is that the electric field can
be uniquely related to the voltage on the transm ission line. If the electric field E T(x , y )
due to the transm ission line is know n, an approxim ation to the induced voltage on the
probe is
= jE rU ,y)dl
Lfjp is very small
*
L ,iP E T y ( X t i p ^ u p )
(4-2)
path
w here L,ip is the probe tip length, E j y is the y com ponent o f electric field at the tip
location (xtip, ytip).
As shown in Figure 4-2, when the m onopole probe is right over the signal line of
CPW transm ission line at position 1, the electric field E T(x, y )o n ly has y com ponent,
the coupling voltage between the probe tip and the CPW circuit Vpc will then be
maxim um in magnitude. W ith the probe m oving to the right, E r (x, y )w ill have both x
and y com ponents till the probe m oves to position 2, where the probe is right above the
58
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
gap betw een signal line and ground line of CPW , E T (x, y) has now only x com ponent
and Vpc w ill then be zero. Again, E T (x, y) at position 3 only has y component with an
inversed phase, so Vpc will be seen at its second peak in magnitude.
To solve for E T( x , y ) , we solve the Laplace Boundary V alue problem for 4>(x, y),
and then use Ofx, y) to determ ine E T( x , y ) since E T (x, y) = -V<b(x, y ) . In order to
estim ate Vind for a particular CPW geometry, under electrostatic conditions we can solve
the Laplace Equation for that geom etry to yield E T (x, y) and then
(At, Ay)
o0
v.c
4--0
<e=v,'
* V<P1= Q
-►x
\i,b ! i i
Figure 4-4 Laplace Boundary Value problem solving
W e solve the Laplace BV problem num erically, as shown in Figure 4-4, using a finite
difference MOM approach with an SOR Iterative M atrix Solver [40] . We set
•
Ax = 25 fim, Ay = 25 pm and N x = 101 and N y = 121 to solve Laplace BV
problem for Type A CPW , and
59
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
•
Ax = 10 pm , Ay = 10 p m and N x = 101 and Ny = 111 to solve Laplace BV
problem for Type B CPW , and
Figure 4-5 is the plot o f electric potential <3>(x, y) on x-y plane (refer to Figure 4-2 and
Figure 4-4) by solving the Laplace BV problem . Given the contour plot o f <E>(x, y), the
electric field E T (x, y) will then be drawn since it is normal to d>(x, y).
Figure 4-6 is the plot that shows the Laplace BV problem solution for y com ponent of
the electric field Ey(x, y) at different height above the Type B CPW transm ission line
circuit plane.
Figure 4-7 and Figure 4-8 dem onstrate the solutions for Type A and Type B CPW Tline laplace problem using the SOR iteration algorithm , and they will be com pared with
the experim ental results on Type A and Type B CPW s later in section 4.1 and 4.4. Figure
4-7 shows the results on Type A CPW with different probe tip length at a certain height
over the circuit. Figure 4-8 shows the results on Type B CPW with probe at different
heights over the circuit. Since a CPW probe will be used to m easure the Type B CPW
circuit in the later experiment, we choose the height at 0.5w, w and 2w, where w = 120
pm is the signal line width of the CPW probe.
To confirm that the convergence o f FD m ethod is not changing with the grid size
AxAy, the calculation was repeated by using different values of Ax and Ay, nam ely Ax =
10 um and Ay = 10 um to calculate for the Type A CPW and Ax - 5 um and Ay = 5 urn to
calculate for the Type B CPW . The same results w ere obtained for each case.
60
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
200
400
600
800
1000
x (mm)
Figure 4-5 Contour plot of electric potential <J>(jt, y) of T^pe A CPW
provided that 5V applied to the CPW signal line
Ey(x, y) at different height above circuit plane
1600
.izzteghfelQOurr.
1400
hejght-200urr
- height=300urr
“-~-'-'heighf=400uhr'
— height=500urr
1200
1000
■g 800
fe> 600
400
200
-200
-400
0
Figure 4-6 Ey(x,
200
y)
400
600
displacement (um)
800
1000
at different height above Type B CPW circuit plane
61
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
0
500
1000
1500
displacement (um)
2000
2500
Figure 4-7 Laplace BV problem solution for Type A CPW provided that
different monopole probes at same height of 100 pm over the circuit plane
—- height - 60 um
height = 120 urr
height = 2 4 0 urr
-10
CD
-a
-20
-30
-40
-50
-60
0
200
600
400
displacement (um)
800
1000
Figure 4-8 Laplace BV problem solution for Type B CPW provided that a
CPW probe with tip length of 180 pm at different heights over the circuit
plane
62
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
4.1.1.2 Analysis of Longitudinal Electric Field Variation
To investigate the longitudinal variation of the electric field, we can analyze the
structures using circuit theory concepts, provided we break the problem into small parts
so that the circuit elem ent dim ensions will be much sm aller than a wavelength. T o do
this, we describe the transm ission-line by a series resistance per unit length R, series
inductance per unit length L, shunt conductance per unit length G, and shunt capacitance
per unit length C. A small section of transm ission-line with length dz thus has the
following equivalent circuit:
l (z)
L dz
o
R dz
l(z + dz)
-trmn— \AAA
+
V(z)
•o
o
dz
Figure 4-9 Equivalent circuit of a section of transmission line
A nalysis o f this circuit for tim e-harm onic signals gives the wave equations
(4-3)
dz
where y is the complex propagation constant given by
(4-4)
y - a + j P = , j( R + ja)L)(G + jcoC)
Given a lossless transm ission line, a = 0, and
63
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
y =
(4-5)
The solutions to the wave equations are superpositions o f forward and reverse waves,
V ( z ) = v;e-^
+v;e^
Just like with plane waves,
I ( z ) = I*e~yz + i y z
we
(4-6)
define a characteristic im pedance
as the ratio of
voltage to current (for positive travelling w aves)
(4-7)
O
O
and the wave velocity (or phase velocity) and wavelength are given by
n
(0
/a
\= -p
n\
( 4 ~8 )
Zs
Figure 4-10 S c h e m a tic
m odel
of a generally loaded transmission line
For the transm ission-line circuit above (Figure 4-10), we define a voltage reflection
coefficient (at the load Zi) as the ratio o f reflected voltage to incident voltage, which can
in general be com plex:
64
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Some special cases are:
Short: Z t = 0 —> F l = -1
Open: Z/, =
Match: ZL ~ Z o —> F l = 0
—» F L = +1
U sing the load reflection coefficient gives the input im pedance Z in for the lossless
case
Z L + j Z » t a n ffl
° Z 0 + j Z L tm/3l
The term (3/ is called the electrical length o f the transm ission-line. Some special cases
for Z in are considered with different load Z L\
2,71
Short line:
ZL = 0
Z m = j Z , tan $ = j Z 0 ta n — I
^8
Open circuit:
ZL -°°
Z in = - ;Z cot
^
2 71
= - j Z n c o t— I
M
(4-11)
(4-12)
.
Given Equation (4-6) and (4-9), V(z) can be produced,
V ( z ) = V0+[l + r L e 2ri] e - y:
(4-13)
where Vo+ is determ ined by the source V, (Figure 4-10) and its im pedance. For a
lossless transm ission line,
V(z) = V0+[l + T L e J2Pz]e- JPz
(4-14)
65
R eprod u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited without perm ission.
W hat are interested in is V(z) for different load conditions.
F or a m atched load, Z l = Z q and F l = 0, there is a linear phase variation with position
(4-15)
V ( z ) = VQ*e~if,z
For an OC (open circuit) load, Z L = °° and F L = +1,
V ( z ) = V0+[l + e j2fiz]e~jfil
(4-16)
= 2V0 cos(— z)
H ence, there is a standing w ave with m inim a at Ag/2 spacing. A long the transm ission
line, as shown in Figure 4-10, V(z) has a m inim um in magnitude at each test point w here
z = n k g/4, n = -1, -3, -5, ..., w here
=c f
■
4.1.2 50 O CPW T- Line M easurem ent with M onopole P ro b es of
Different Tip L engths
dD
►][C o a x ia l P ro b e
t= 0 .5 m m /2 m m
d o = 0 .3 2 m m
d i= 0 .0 5 m m
Figure 4-11 Monopole probe
66
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
C oaxial probe is the simplest and easiest m anufactured diagnostic tool in near field
scanning system . The purpose of this test is to com pare how difference of tip length t will
take effect to the probe performances. The tip length o f m onopole probe is the length by
which the inner conductor is made longer than the outer conductor (copper shield) at the
probing end.
Figure 4-11 is the m onopole probe used for scanning. The diam eter of outer shield is
320 pm , and the diam eter of inner conductor is 50 pm . The tests were conducted with
tw o identical coaxial probes with different tip lengths (t) o f 0.5 mm and 2 m m
respectively. The D C m otor scanning station (Figure 3-7) was used for testing.
Tw o 50 Q open circuit CPW T-Lines (Type A and Type B) were m easured at 5 GHz.
T heir param eters are listed in Table 4-1.
Network
Analyzer
486 PC
probe
OC
Figure 4 -1 2 Test structure to measure CPW T -Iin es
The test set up is shown in Figure 4-12. H P 8753A network analyzer (up to 6 GHz)
was used to excite the circuit under test and to pick up the signal to finally get the S 21
forw ard transm ission param eters. Tw o D C -controlled m otors com posed o f the scanning
stage that made the m ovable probe doing 1D/2D scanning over the circuit area. C om puter
workstation was controlling the activities o f the network analyzer (through GPIB bus)
67
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
and the scanning stage (through RS-232 port), and processing data from the netw ork
analyzer.
The transverse scan, for which the probe was moving along the path 2.5 cm away
form the edge of open circuit end o f the C PW T-Line, was run over the open-circuited
Type A C PW (w = 740 pm and s = 330 p m ) with both monopole probes (different tip
lengths) at frequency of 5 GHz. The probes were over the CPW circuit plane about 0.1
mm high. Figure 4-13 shows the scanning results with both m onopole probes. The 0.5
m m tip length m onopole probe found m inim um induced signal at displacem ent around
820 pm . T he Laplace BV problem solution for the same Type A CPW shows (Figure
4-7) that the m inim a occurs at displacem ent around 610 pm. The deviance o f num erical
and experim ental results is about 34%. The 2-m m tip length monopole probe, on the
other hand, was not able to reflect the proper electric field transverse variation as
expected from the Laplace BV problem solution in Figure 4-7.
The longitudinal scan was also applied to an open-circuited Type A C PW T-Line by
the 0.5 m m tip length monopole probe, the probe was over the CPW at about 0.1 m m
height, and the signal frequency was 5 GHz. The test results are shown in Figure 4-14.
From the standing wave pattern on the m agnitude plot, half wavelength o f guided wave
Xg/2, was m easured 21.5 mm. U sing Ag — c j f ^ £ eff , the
obtained by experim ent was
1.94. Noted that the seff calculated by PC A A D was 1.87, the deviance of simulation and
measurem ent was 3.7%.
68
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
-20
t= 2mm
t = 0.5mm
S ’ -40
~a
CD
-80
-100
■4
-3
-2
0
1
1
2
3
4
distance (mm)
200
t = 2mm
t = 0.5mm
oT 100
2ai
m
a
o
S -ioo
-200
-4
-3
-2
0
■1
1
2
3
4
distance (mm)
Figure 4-13 Magnitude and phase plot of transverse scan over Type A
CPW with two monopole probes having different tip lengths
-30
-40
-50
c -60
O
CCl
E -70
-80
0
5
10
15
20
25
30
35
40
25
30
35
40
distance (mm)
100
TS
-100
-150
0
5
10
15
20
distance (mm)
Figure 4-14 Magnitude and phase p lo t of longitudinal scan along the
signal line of Type A CPW with 0.5 mm tip length monopole probe
69
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
4.2 Monopole Probe to Measure Single Patch
Microstrip Antenna
Since th e mid 1970’s the rectangular patch has been one of the most w idely used
antennas in com m unication developm ents [41] . A single patch antenna is the
fundam ental elem ent in the design of an ideal antenna array. To determine the pattern of
amplitude and phase of a single patch antenna, a test system was set up as shown in
Figure 4-18 b) which is almost the same as the configuration used for CPW T-Lines
m easurem ents except for the DUT here is a single patch antenna.
To investigate an antenna com ponent, the m ost im portant measurements are:
•
V SW R for indication o f m ism atch
•
m ode excitation for elem ent cross-polarization
•
surface wave excitation for estim ation of the elem ent mutual coupling
level
4.2.1 P ro p erties of Single P atch M icrostrip A ntenna
Generally a single patch m icrostrip antenna is com posite o f three segm ent of
microstrip transm ission lines with specific dim ensions (width W and length L) working at
the expected operating frequency f r. Figure 4-15 shows the dim ensions o f a single patch
microstrip antenna working at 3.455 GHz. The three segments are signal feed line,
quarter wavelength transform er and rectangular patch built on the substrate with £r = 3.2 .
The single patch m icrostrip antenna is designed to be operating at its resonating mode,
for which the frequency o f signal fed to the patch is close to the resonant frequency f r.
In following analysis, we are sim ply considering rectangular patch in the first mode
that is the first order resonance frequency occurs on the patch.
70
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Rectangular
Signal feed line
L~14.7mm
L=34.0mm
substrate
h=1.56mm
L—
25.3mm
/
/
Quarter wavelength
transformer
Figure 4-15 Dimensions of the open-ended single patch microstrip
antenna design
Quarter wavelength
transformer
Rectangular
patch
Signal feed line
Vs
substrate
Figure 4-16 Structural diagram of a single patch microstrip antenna
design
71
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
A typical single patch antenna design is to com bine the m icrostrip transm ission line,
which is feeding the signal to antenna with the m icrostrip patch antenna on the same
substrate (Figure 4-16). To m atch the im pedance of the m icrostip signal feed line to the
microstrip patch, a quarter-wavelength transform er has to be inserted between the signal
feeding line and the patch. The quarter wavelength transform er technique is generally
used to m atch purely resistive loads [4 2 ].
It is also illustrated in Figure 4-16 that the electric fields, along the z direction,
undergo fringing at the edges of the patch since the open-ended patch is finite along both
length and width (x axis and z axis). This fringing effect has to be taken into account by
adding equivalent lengths AL at both ends.
(£eff + 0 .3 )(“ + 0.264)
— = 0 .4 1 2 ..................... h.................
h
(4-17)
(£eff -0 .2 5 8 )(—+ 0.8)
C. A. Balanis [41] gives out equations to calculate the param eters for rectangular
patch design at resonant frequency. Some values are calculated now for the designed
single patch antenna shown in Figure 4-15 and Figure 4-16.
The effective dielectric constant of the patch
£r +1
£. - I f .
1+
10h^\
-
2
3.03
(4-18)
v
Given the resonant frequency f = 3.455 G Hz,, the equivalent rectangular patch
length can be calculated by (4-19), where c is the free space speed o f light.
72
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Le = L + 2AL = ----- C
- r = = 0.0249 m
^ f r i £eff
(4-19)
In a typical design, the signal feed line is a 50 Q im pedance transm ission line. Given
the substrate thickness h and dielectric constant £r, the transm ission line width W and
effective dielectric constant zeg can be then calculated. W ith the patch width W is much
greater than the substrate thickness h (W/h »
1), the antenna will behave more as
transverse electrom agnetic wave (TEM ) transm ission line, so the entire line feed single
patch antenna can be equivalent to serialized transm ission lines with different line
im pedance and £
as shown in Figure 4-17.
feed line
A/4 transformer
patch
Vs
-,m -fe e d lin e
m -patch
Figure 4 -1 7 Equivalent circuit of microstrip line fe e d single patch antenna
For the design shown in Figure 4-15 and Figure 4-16, the radiation conduction at the
open end of rectangular patch is calculated as [41]
1 [ 60(k0W )
Gmd =
3 ^ 1 1 + (k0W )
2 'N
[l + 0.32 + 0.68cos(0.77k0Le)]
(4-20)
= 0 .0 3 8 8 0 ''
73
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
w here
k 0 = - ^ £ = = 41.55
C^ £eff
120tt
rj = —? = ~ 216.45
i e .»
From equivalent circuit Figure 4-17, the input im pedance o f rectangular patch can be
approxim ated as
Z „ -„ , = —
2 G -ad
(4-21)
= 1 2 .8 8 0
The characteristic im pedance o f the Xg/4 transform er can be calculated as
1 2 0 tz
J0U /4)
F
W-
t effai4) ^ £ + 1.393+ 0 .6 6 7 In
4 ^ -+ 1.444
h
(4-22)
- 75.590,
Z'1w / 4 - ( Z o u / 4 )) 2 / Z in- ^
= 4 4 3 .5 6 0
(4-23)
It is noted in (4-23) that the transform er appears not correctly designed since the lookin im pedance from the transform er input port is not 50 ohms as expected. There will be
m ism atching between the 50 ohm signal feed line and rest o f the antenna circuit.
For a lossless transm ission line shown in Figure 4-10
(4-24)
V{z ) = V- [ i + r L e m i z }e
74
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
where V)+ is determined by the source Vs and its impedance.
F or an OC load,
=
and T l = +1,
V (z) = V0+[l + e J2fi*]e~Jfil
(4-25)
+
2n
— 2V0 cos(— z)
H ence, there is a standing wave with a m inim um at XJA from the end of the Xs/2
patch.
4.2.2 Single Patch M icrostrip A ntenna M easurem ent
To m easure the single patch m icrostrip antenna shown in Figure 4-18 a), we excited
the antenna patch with a 3.544 G Hz R F signal from the netw ork analyzer. The coaxial
probe is right above the antenna at the height of about 0.2 mm. (The coaxial probe was
touch down an isolation of plastic film lying on the surface of the antenna patch, the
thickness o f the film was about 0.2 mm). The pattern shape in 3-D display for the
m agnitude measurem ent is shown in Figure 4-19. The m agnitude and phase pattern of the
single patch antenna are shown in Figure 4-20 and Figure 4-21, respectively.
Figure 4-22 is the m agnitude and phase plots along the central line in y direction. It is
noted that the m agnitude plot o f the 50 ohm signal feed line section is not flat as expected
due to mism atching of im pedance. Given the Zin.patch o f 12.88 ohms in (4-21), the A.g/4
transform er had to be designed such that the Zo(Ag/4 ) equals to 25.38 ohms in stead of
75.59 ohms (4-22) to make Z in.xg/4 close to 50 ohms.
75
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
substrate
Wiltron Network <— »->r| 486 PC
Analyzer
antenna
patch
25.10mm
scanning
area
signal feed line
b) test set up for single
patch antenna pattern
measurement
a) dimension of the
single patch antenna
Figure 4-18 a) Single patch antenna; b) Single patch antenna test
structure
-70
80
40
40
30
20
20
0 0
Figure 4-19 3-D display of the magnitude pattern of the single patch
antenna
76
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
x (m m i
Figure 4-20 Magnitude pattern of the Single patch antenna
150
W * i'z
100
-5 0
100
150
Figure 4-21
P hase
pattern of the S in g le patch antenna
77
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
A/4
p a tc h
tra n sfo rm e r
-3 0
-3 5
-4 0
5 -4 5
o>
-5 0
-5 5
-6 0
40
z(m m )
60
40
z(m m }
60
200
150
100
50
cn
T3
■§. -5 0
-100
-1 5 0
-200
20
Figure 4-22 Magnitude (upper) and phase (lower) plot - longitudinal
scanning over the antenna patch along the central line
78
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
4.3 Monopole Probe to Measure Microstrip Antenna
Array
4.3.1 M icrostrip A ntenna Array T opologies and Signal Feeding
Network
As an application of near-field probing technology to m easure the microwave circuits,
the m onopole probe scanning system was used for investigation on 4x4 antenna array
topologies at 12 GHz. A survey o f 4x4 array topologies at 12 GHz was conducted by D.
Gray, a Ph.D. student o f the D epartm ent of Electrical and C om puter Engineering at the
U niversity of M anitoba. Seven 4x4 M icrostrip antenna arrays fed with different singal
feeding netw orks were fabricated using 0.794 m m thick,
¥2
ounce copper clad Arlon
D iclad substrate (er = 2.5) [43] . Use o f a constant substrate and design frequency allows
for direct comparison of the arrays behaviour as a function o f topology alone. The
designed resonant frequency o f the antenna arrays are both between 11 GHz - 13 GHz.
Tw o types of antenna array, type C and type E [43] w ere chosen for the tests using
m onopole probe scanning system. The dim ensions o f type C and type E antenna array are
shown in Figure 4-23 and Figure 4-24, respectively.
The purpose of the m easurem ent is to experim entally investigate the performance of
m icrostrip feeding netw ork and radiation pattern o f each antenna patch. It was expected
to see same radiation pattern for both m agnitude and phase on all patches and
fundamental m ode as the only dom inant mode and no other higher order modes. It was
also investigated if the feeding netw ork is imbalance.
Figure 4-25 shows the m easurem ent results on the type C antenna array with 12 GHz
R F feed signal from the netw ork analyzer. The coaxial probe was at about 0.2 mm height
above the antenna. For patch 1 and 2, it was observed that they are fed with the same
79
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
am plitude and phase but with slight excitation of 3rd order mode. Patch 3 and 4 have
sim ilar behaviour show that higher order modes had predom inant effects.
The feeding netw ork show small VSW R in the section m arked in Figure 4-25, the
linear phase shift indicates impedance matching.
The sam e m easurem ent was applied to the antenna array type E fed with 12 GHz RF
signals, and the results are shown in Figure 4-26. The m odes generated on four patches
are m ore predom inantly the fundamental modes. The feeding network also show small
V SW R in the section m arked in Figure 4-26.
Type E has better perform ance than type C as shown in Figure 4-28, which agrees
well with the theoretical m odelling [43] .
80
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
f
■
t
rr*
| f
t
*
]
E
'
rr'
“
-scanning
area
f t
I
m
1[
f
t
f t
* 1
i
i
r j j j ]
Figure 4-23 The layout of ty p e c antenna array
scanning
area
::::
Z
.P i
n U
F
!
Figure 4-24 The layout of type e antenna array
81
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Mjnte Nf>.»
.y
Hs?:p
F eed line m atches
a)
m easured m agnitude pattern o f type c antenna array
driv en by 12GHz R F signal
Il*|UI4*\P .?
Fib
Hkip
b) m easured p hase pattern o f type c antenna array
driv en by 12G H z R F signal
Figure 4-25 Near-field image o f type C antenna array driven by 12 GHz
RF signal
82
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Feed line matches
». .
%1
i-.rie
100
a)
!?!>
m easured m agnitude pattern o f type e antenna array
driven by 12GHz R F signal
iigi.sf*\u *
if«
b)
t?fj
m easured phase pattern o f type e antenna array driven
by 12GHz R F signal
Figure 4-26 Near-field image of type E antenna array driven by 12 GHz
RF signal
83
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
5
10
15
20
25
30
x (mm)
a)
m easured m agnitude pattern o f type e antenna array
driven by 13G H z R F signal
5
10
15
20
25
UT
x (mm)
b)
m easured p hase pattern o f type e antenna array
d riven by 13G H z R F signal
Figure 4-27 Near-field image of type E antenna array driven by 13 GHz
RF signal
84
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
0
.......... r
— w
-5
............ t
!
?
.-77--.*"•
~~ —
f-10
T
1
1
— Type C
— Type D
TypeE
- Type 1
X :- j /, '
■\ / /
■V\ Cr'/
- 2 0 - ..................... .................................................................................. ............... '
i1I l'
i-'
t---------i1------ ■
--li—
t —— i-1-------- 1---------1
1————
_25L—......i---------t1—-----1
------ 1-—
11
1 1 .2
1 1 .4
1 1 .6
1 1 .8
12
1 2 .2
1 2 .4
1 2 .6
1 2 .8
13
Frequency (GHz)
Figure 4-28 Boresight radiation and input impedance characteristics of 12
G H z 4x4 microstrip arrays [43]
4.4 CPW Probe to Measure 500 CPW T- Line
jj s = 30pm
t = 180pm
w=
1 2 0 pm
a) C P W probe
b)
C PW probe to m easure the type B C P W
Figure 4-29 a) CPW Probe b) snapshot of CPW probe and the DUT
85
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
To im prove the spatial resolution of an electric near-field probe the m icrom achining
process has been em ployed to fabricate pm -scale probes. It is, however, unfeasible to
build probes in a form of coaxial cable by micromachining technique. The planar
structure o f coplanar waveguide has provided much flexibility in design, reduced weight
and volum e, and cost of production. It is for this reason that a CPW probe was fabricated
and tested as a scale model of the m icrom achined coplanar waveguide probe. The test
configuration and dim ension of the CPW probe is shown in Figure 4-29.
4.4.1 CPW T-line Probe C ircuit Modelling
50 Q.
ind
50Q
V(x,y)
Figure 4-30 An approximate coplanar waveguide transmission line probe
model and the corresponding equivalent circuit model
Figure 4-30 shows the model of coplanar waveguide transm ission line (CPW ) probe
and its approxim ate equivalent circuit. The m odel is quite similar to a monopole probe
model shown in Figure 3-2, so the m odelling and analysis will be also similar. The results
drawn from the analysis o f m onopole probe in section 3.2 is also applicable to CPW
86
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
probe m odelling. The model also assumes that the probe load im pedance is much greater
than local circuit impedance. Cci is the capacitance form ed between conductor o f probe
and transm ission line circuit, Csc is the capacitance form ed between probe conductor and
shield, Csi is the capacitance form ed between shield of probe and transm ission line
circuit, Ccg is the capacitance form ed between probe conductor and ground plane, Csg is
the capacitance formed between probe shield and ground plane.
4.4.2 50 Q CPW T- Line M easurem ent with Different CPW Probe
O rientations
Normal of CPW probe surface
Normal o f CPW probe surface
a)
b)
Figure 4-31 a) The CPW probe surface normal is parallel to signal line
and b) The CPW probe surface normal is orthogonal to signal line
This m easurement is perform ed to find the effect to the electric coupling level
between CPW probe and circuit under test when the probe is set to different orientations
over the circuit plane. The normal of C PW probe surface, as shown in Figure 4-31 a) and
b), is either parallel to or orthogonal to the signal line of CPW T-line. A Type A CPW Tline was measured and the results are shown in Figure 4-32 and Figure 4-33. From the
analysis to CPW in section 4.1.1.1, it is expected that the orientation of the CPW probe
not to make much difference on m easurem ent results, since the induced voltage on the
probe is mostly depending on probe height and tip length. Figure 4-32 and Figure 4-33
experim entally state the sam e conclusion.
87
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
-30
— -40
CD
3,
(D -50
-o
E -60
O
)
to
E -70
-80 >-2.5
-2
-1.5
•1
-0.5
0
0.5
distance (mm)
1
1.5
2
2.5
■2
-1.5
■1
-0.5
0
0.5
distance (mm)
1
1.5
2
2.5
g> -50
TJ.
i-100
-200 L-
-2.5
Figure 4-32 CPW probe transverse scan over the Type A CPW with probe
surface normal parallel to the CPW T-line under test
-40
cn -50
~u
-60
-70
■2
-1.5
■1
-0.5
0
0.5
distance (mm)
1
1.5
2
2.5
-2
-1.5
■1
-0.5
0
0.5
distance (mm)
1
1.5
2
2.5
50
S? -50
2 -100
^ -150
Figure 4-33 CPW probe transverse scan over the Type A CPW w it h probe
surface normal orthogonal to the CPW T-line under test
88
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
4.4.3 50 O CPW T- Line M easurem ent with CPW P robe at
Different H eights above DUT
The heights at which the probe is positioned over the CPW T-line have effects to the
m easurem ent performance. The tests discussed in this section w ere perform ed in order to
investigate these effects. The circuit was running at frequency o f 5 GHz, and the probe
was set at the height o f 0.5w, w, 2w and 4w , w here w is the width o f the signal line of
CPW T-line, to scan transversely over the DUT. A Type B CPW T-line (w = 0.2 m m and
s = 0.1 m m ) was m easured for this purpose. The m easurem ent results for the probe set at
different heights are presented in Figure 4-34, Figure 4-35, Figure 4-36 and Figure 4-37,
respectively. Figure 4-38 com pares the m agnitude o f m easurem ents and shows that
height o f 2w has the best performance.
-40
~ -50
m
Ta>J -60
I -70
E -80
-90 L
-
1.2
•1
-0.8
-0.6
I
i
-0.4
-0.2
0
distance (mm)
0.2
0.4
0.6
m
\tM ^
i
-1.2
-1
-0.8
i
-0.6
0.8
''A ll
r
-0.4
-0.2
0
distance (mm)
0.2
0.4
0.6
0.8
Figure 4-34 CPW probe transverse scan result on CPW transmission line,
height of 0.5 w
89
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
-50
S -60
o
•a
D) -70
-80 L-
1. 2
-1.2
■1
-0.8
I
)
-1
-0.8
-0.6
-0.4
-0.2
0
distance (mm)
j
-0.6
I
Lt i
-0.4
-0.2
0
distance (mm)
0.2
0.4
0.6
0.8
0.6
0.8
................
0.2
0.4
Figure 4-35 CPW probe transverse scan result on CPW transmission line,
height of w
-50
m -60
"O
-70
-80
-90 L—
-
1.2
1
-0.8
-0.6
0
-0.4
-0.2
distance (mm)
0.2
0.4
0.6
0.8
■1
-0.8
-0.6
-0.4
-0.2
0
distance (mm)
0.2
0.4
0.6
0.8
200
fj -100
Figure 4-36 CPW probe transverse scan result on CPW transmission line,
height of 2w
90
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
-50
CO
TS
■60
....
/~
/V
r,s,1w/ ...1
X !iffr.
..*
-80
-1.5
-0.5
0
distance (mm)
0.5
I
200
H vv*
■®
100
£
o>
m
S
m
vs
0
r
Vi
•S -100
-200
-1.5
-0.5
0
distance (mm)
H
0.5
Figure 4-37 CPW probe transverse scan result on CPW transmission line,
height of 4w
h = 0.5w
h=w
n = 4w
-1.5
-0.5
0
distance (mm)
Figure 4-38 Measurement of magnitude, CPW probe transverse scan over
the DUT at height of 0.5w, w, 2w, 4w
91
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
4.4.4 50 0 CPW T- Line M easurem ent with CPW Probe and
M onopole P robe - A C om parison
The transverse scan is applied to both Type A and B CPW T-Lines with CPW probe,
and the results are shown in Figure 4-39 and Figure 4-41. The probe is right above the
CPW w ith height between 25 pm to 50 pm.
T he sam e m easurem ent was repeated using the m onopole probe with the tip length t =
0.5 m m . T he results are shown in Figure 4-40 and Figure 4-42. T he m onopole probe is
also at a height between 25 pm to 50 pm over the DUT. These results can be referred to,
for evaluating the perform ance of the CPW probe.
-40
-50
-60
•£ -70
O
cC)
£ -80
0.2
0
distance (mm)
0.2
0.4
0.6
0.2
0
distance (mm)
0.2
0.4
0.6
-
0.6
-0.4
-
-
0.6
-0.4
-
150
„ 100
Figure 4-39 CPW probe transverse scan result on the Type B CPW T-Line
92
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
-40
-70 >-
0.8
-
-0.4
0.6
-
0
0.2
0.2
0.4
0.6
0.8
0.2
0.4
0.6
0.8
distance (mm)
100
O
a))
T3
-100
-
-0.4
0.6
-
0
0.2
distance (mm)
Figure 4-40 Monopole probe transverse scan result on the Type B CPW TLine
S'
-50
-60
-70
■2
-1.5
•1
-0.5
0
0.5
distance (mm)
1
1.5
2
2.5
■2
-1.5
■1
-0.5
0
0.5
distance (mm)
1
1.5
2
2.5
? -50
33
£ -100
-150
Figure 4-41 CPW probe transverse scan result on the Type A CPW T-Line
93
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
-20
m
-40
1
-60
o>
-80
-
100 *
-
-2.5
■2
-1.5
-0.5
0
0.5
distance (mm)
■1
.... ----!r— “i
1
----
1.5
2
2.5
--- -n
t
-2.5
-2
-1.5
-1
-0.5
0
0.5
distance (mm)
1
1.5
2
2.5
Figure 4-42 Monopole probe transverse scan r e s u lt on the Type A C P W TLine
4.5 Conclusion
V arious m icrowave devices were tested by m onopole and CPW probes with the
autom ated near-field scanning system described in C hapter 3. The performance o f both
monopole and CPW probes were investigated by m easuring tw o different CPW
transm ission lines. Both transverse and longitudinal scan were conducted and results
were com pared with the numerical analysis of transverse and longitudinal variation of
electric field in section 4.1.1. The probes were also used to m easure the single patch
antenna of 3.45 GHz resonance frequency and tw o types o f m icrostrip antenna arrays.
94
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
C h a p te r 5
M ic r o f a b r ic a t e d C P W P r o b e s
W ith constantly increasing frequencies in com m unications and integrated circuits,
there is a great dem and for low-cost, m iniature m icrowave components. The planar
transm ission lines, such as m icrostrip, stripline, and coplanar waveguide have provided
much flexibility in design, reduced weight and volum e, and cost o f production. Planar
com ponent size has reduced so much that the fabrication by conventional m achining
techniques has becom e too costly and difficult. It is for this reason that the recent
advanced m icrom achining techniques
have
found numerous
applications
in
the
m icrowave field.
5.1 Introduction
In conjunction with its conventional role as an electronic material, silicon can also be
exploited as a high-precision, high-strength and high-reliability mechanical m aterial by
taking advantage o f an already advanced m icro-fabrication technology. This is especially
applicable w herever m iniaturized mechanical devices and com ponents m ust be integrated
or interfaced w ith electronics. Texas Instrum ents has been m arketing a therm al point
head [46] in several com puter terminal and plotter products since early 1977 in which the
active printing elem ent abrasively contacting the paper is a silicon integrated circuit chip.
The crucial detector com ponent o f a high-bandw idth frequency synthesizer sold by
Hewlett-Packard [46] was made as a silicon chip in 1980 from which cantilever beams
have been etched to provide therm ally isolated regions for the diode detectors. H ighprecision alignm ent and coupling assem blies for fiber-optic com munication system s are
produced by W estern Electric [46] from anisotropically etched silicon chips sim ply
because it is the only technique capable of the high accuracies required. Som e recent
works carried out by G. M. Rebeiz [47] in the Radiation Laboratory in the U niversity of
M ichigan shows that m icrom aching techniques can be used to alter the electrom agnetic
95
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
characteristic o f the underlying substrate so as to result in better perform ance than
standard planar designs. They designed some high Q resonators, low -cost filters, and high
efficiency m icrom achined antennas. The sim ilar w ork is also done by V. M ilanovic [48]
on coplanar waveguides. It is verified that the absence of the lossy silicon substrate after
top side etching results in significantly im proved insertion-loss characteristics, dispersion
characteristics and phase velocity.
5.2 Basic Procedures of Micromachining
The m ostly used m icrom echanical processing techniques are etching, epitaxial
processes/deposition, therm om igration, therm al bonding, and m em brane techniques. The
etching, deposition and m em brane techniques w ere used in our application which will be
discussed in detail in section 5.4.
5.2.1 Etch
The com m only used etching system s are chem ical, plasma, sputtering and reactiveion etching (R1E). Chemical etching is suitable for large and deep area removal while the
other three etching techniques are usually used for surface/mask removal.
The chem ical etchants for silicon are num erous and they can be isotropic or
anisotropic, dopant dependent or not, and have varying degrees of selectivity to the
m asking materials. W hat we used in the m icrofabrication of CPW probe is KOH [46] ,
the anisotropic chemical etching w hich exhibits much higher etch rate ratios than any
other chem ical etching. A disadvantage o f K O H is that S i0 2 can ’t be used as a m ask in
many applications. Si3N 4 is the preferred m asking material for KOH.
In plasm a, sputtering and reactive-ion etching (RIE), the high-accelerated and heated
ions are applied to the surface o f silicon to rem ove the m aterials such as S i0 2 and Si3N 4.
96
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
5.2.2 D eposition
W hile etching is for the m aterial rem oval, the material addition in the form of film
deposition, metal plating is very im portant structural tool. D eposited thin films have
obvious applications in mask, passivation, w ear resistance, corrosion protection, fatigue
strength enhancem ent. The m ost common used deposition m ethods are chemical vapour
deposition (CAD), e-beam and sputtering deposition.
CA D is a widely used m ethod for depositing thin film s o f a large variety of m aterials
such as SiC>2 and Si 3 N 4 . The C PW probe discussed in section 5.3 was fabricated with this
technology. In a typical CAD process, reactant gases (often diluted in a carrier gas) at
room tem perature enter the reaction cham ber. The gas m ixture is heated as it approaches
the deposition surface, heated radiatively or placed upon a heated substrate. D epending
on the process and operating conditions, the reactant gases m ay undergo hom ogeneous
chem ical reactions in the vapour phase before striking the surface. N ear the surface
therm al, mom entum , and chem ical concentration boundary layers form as the gas stream
heats, slows down due to viscous drag, and the chem ical com position changes.
Heterogeneous reactions of the source gases or reactive interm ediate species (formed
from homogeneous pyrolysis) occur at the deposition surface forming the deposited
material.
The electron beam evaporator deposits thin metal film s by m eans of electron beam
heating and evaporation inside a high vacuum cham ber. This system is prim arily used for
the deposition of m icroelectronic contact or interconnects. To avoid contam inating the
metallization processes, the types o f source materials and substrates allowed into the
system is limited. Evaporation m aterials allowed are At (pure), A l-l% S i, Pt (user m ust
supply own Pt material). A llow ed Substrates are Inorganic substrates only - Si, glass,
Silicon-on-Sapphire,
com pound
sem iconductor.
Sputtering
deposition
use
the
cryopum ped cham ber for RF/D C sputtering to form a metal plane such as Al, Cr and gold
on the silicon surface.
97
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
5.2.3 Silicon M em brane
W hile etching and deposition are two essential procedures in micromachining, some
other procedures such as therm om igration, thermal bonding, especially the dopantdependent etching for the realization of thin silicon m em branes are all employed in many
applications. W e used the silicon membrane in our CPW manufacturing.
Silicon m em brane is im portant in controlling the etching processing. The etching
processing will stop while arriving at a silicon layer which is highly doped with boron. It
is simply because the etching rate inside the boron medium is really slow com pared to the
other normal silicon material.
One application of the boron-doped m em brane is illustrated in the following
application in the ink je t nozzle design. W ithout the m em brane, the hole size of the in k je t
nozzle will be affected by the thickness o f the wafer layer (Figure 5-1 a) and b)). The size
of the hole (1) equals L-(2t/tan0), w here the L is the m ask dim ension and t is the thickness
o f the w afer layer and 0 = 57.74°. If a boron doped silicon m em brane is produced and
suspended across the bottom o f the pit with an orifice in the center corresponding to the
location previously left undoped (Figure 5-1 c), the hole size can be precisely controlled.
(b)
(a)
(c )
Figure 5-1 Boron membrane in the fabrication of silicon ink jet nozzle - a)
and b) show the errors in final nozzle size w h ic h occurs when the wafer
thickness varies c) shows a boron membrane structure to minimize the
e ffe c ts of the wafer thickness.
98
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
5.3 Micromachined CPW Probe
Before we start to describe the processing procedures of the m anufacture of our CPW
probe, it is helpful to understand its architecture and dim ension.
Its dimension and a
picture of m icrom achined CPW probe are shown in Figure 5-2.
40 30 47 30 40
T h e d im e n sio n o f th e p ro b e tip
( u n it is p m )
A
Figure 5-2 Dimension and the picture of the micromachined Coplanar
Waveguide (CPW) probe
99
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
The probe was m anufactured on a 400 pm Si wafer, which is partially etched
previously until a 30 p m boron doped m em brane layer is reached. The extra layers of 1.2
pm Si02 and 0.1 pm Si 3 N 4 are deposited on the 400 p m Si w afer surface, which is the
platform o f the probe tip. The conductive material for the CPW three transmission line is
1.13 pm Au and 500 A Cr. (Figure 5-3 a) and b)). This was designed and fabricated at the
Alberta M icroelectronics Centre by Dr. C. Shafai in 1996 [6 2 ].
1.13 pm Au
„ 500 a
0.1 pm Si3N4
Cr
\
I
.„
~ 40 pm
'
■f ~ 30 um boron doped membrane
y
'
'
400 pm
Si wafer
partial backside etched
(a) the side view of the layout of the CPW probe
Au /Cr
30 pm Si
membrane
underneath
(b) top view of the layout of the CPW probe
Figure 5-3 The layout of Co-Planar W a v e g u id e
(C P W )
Probe [62]
100
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
The fabrication procedure [62] started with Si w afer <100> with 400 pm in thickness.
•
Dope the Si w afer with boron dopant on the surface with 30 pm thickness
to form a boron membrane.
•
D eposit 1.2 pm S i0 2 @ 300 °C on the surface o f Si wafer by Chemical
V apour D eposit (CVD), annual @ 1100 °C for 20 minutes, Figure 5-4 a).
•
D eposit 0.1 pm Si3N4 @ 300 °C on the surface of S i0 2 layer by CVD,
Figure 5-4 a).
•
D eposit 0.1 pm Si3N 4 @ 300 °C at the back o f the Si wafer by CVD.
Then pattern the backside Si3N4 layer with REE plasm a etch. This
patterned layer will be used as m ask for the next KOH etch procedure,
Figure 5-4 a).
•
K O H etch @ 80 °C for 6 hours to remove the Si until arriving the boron
m em brane layer to form Si diaphragm, Figure 5-4 b).
•
Pattern and etch top Si3N4 and S i0 2 dielectrics by REE to make Si
exposed, Figure 5-4 c).
•
D eposit 1.13 pm Au, 500A C r and 0.25 pm gold by sputtering deposition
and then patter and etch metals to form probe tip, Figure 5-4 d).
•
To rem ove the exposed Si, etch the Si diaphragm from back by RIE,
Figure 5-4 e).
101
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
( a ) the 400 (am Si w afer doped with boron
membrane and deposited with 1.2 (am S i0 2
and 0.J |am Si3N4mask
(b ) the Si wafer after KO H etch to form Si diaphragm.
(llllull I
(c ) the top view with exposed Si
jw tt.... hswwn;
exposed Si
(d) depositing the Cr and Au and gold
and then etch the metal layer to form the probe tip
(e) the final probe tip is form ed by etching the Si
diaphragm from back in RTE to remove the exposed Si
Figure 5-4 The silicon wafer of CWP during the micro machining
processing[62]
102
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
5.4 Conclusion
As show n in Figure 5-5, a m icrom achined cantilever was m ounted to a CPW circuit.
This created a near-field probe that was used to measure a Type B C PW transm ission line
(Figure 4-3). Unfortunately, the signal coupled to the probe was at noise level. The
structure o f the m icrom achined cantilever is the m ajor reason to cause the problem. Since
the signal line and the ground lines had the same length (Figure 5-2), there was no probe
tip constructed. The probe has to be redesigned and fabricated in the future work. O ther
than that, the gap between the probe tip and the CPW circuit plane has also to be
precisely controlled, and the vibration o f such a tiny cantilever has to be reduced as much
as possible.
i tlittiil
Figure 5 -5
C P W p ro b e
made
w ith
the micromachined cantilever
103
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
C h a p te r 6
S im p l e PCB M e a s u r e m e n t
M a g n e t ic L o o p P r o b e s
w ith
As an alternative to the electric field probing techniques, the magnetic near field
variation has been studied in recent years, and techniques have been subsequently
developed to m easure the magnetic near field produced by a radiating structure [28] [29]
[30] . O ne o f the developments is proposed for exam ining Printed Circuit Board (PCB)
through non-contact means. G oulette etc. developed a system fo r testing electrom agnetic
em ission and diagnosing faults in PBCs [55] [56] , the system com prises a planar sensor
array o f probes consisting of tw o serially connected wire loops aligned perpendicular to
one another. The sensor array is positioned close to the energized board under test such
that the m agnetic field produced by the currents within board conductors will be picked
up by the sensor elem ent located in the proximity. Similarly, Soiferman [57] utilized an
array o f planar printed spiral loop antennas as the non-contact sensors for mapping the
near-field patterns of bare PCB s. They both, for PCB failure detection, compare the near­
field im age pattern o f board under test to a reference pattern obtained from known nonfaulty boards.
This chapter is based on a set o f experim ent reports that w as part o f the investigation
to determ ine detection capabilities o f faults in Multi Layers B oards [58] [59] . The probe
design is such that the magnetic field generated by sources (such as neighbouring lines)
other than the printed line under test does not couple to the probe. Alternatively, printed
lines directly under the probe couple efficiently with the probe, inducing currents which
add constructively. The test set up is shown in Figure 6-1 w hich includes:
•
Single-loop/four-loop probes made from a rigid coaxial cable, the inner
conductive wire is coiled to form the ring(s). T he ring diam eter dr is about
10 mm and the inner wire diam eter dw is about 0.5 mm.
104
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
•
W iltron network analyzer, which was used to energize and to collect the
m agnetically coupled signal to get the S21 forw ard transm ission
parameters.
•
H P plotter, a scanning stage to drive the m ovable probe conducting 1D/2D
scanning over the circuitry area.
•
PC to control both network analyzer and plotter, and reading data from the
netw ork analyzer.
W iltron N etw ork 4 ------ ►
A nalyzer
fixed probe
d w = 0.5 mm
d r — 10 m m
435
pc
•+
m ovable probe
PCB board
Figure 6-1 Test setup for PCB failure detection with magnetic probes
6.1 Near-field RF Inspection of PCBs
E xcitation
Source
Signal
i Vic< n >r
r
PCB trace
Figure 6-2 Traditional PCB fault detection with bed of nails
105
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
A traditional solution of PCB fault detection is a contact m easurem ent as shown in
Figure 6-2. It requires a bed of nails for excitation and signal detection, which turns out
placem ent o f too many switches and yield o f low resolution.
W iltron N etw ork
A nalyzer
486 PC
fixed probe !'
m o v a b l e probe
C opper layer
Figure 6-3 Illustration of indirect excitation
W iltron N etw ork
A nalyzer
486 PC
m ovable probe
,p
xi
- .
r
|
i Si.unninc
h
C opper layer
|
slitCe
/
Figure 6-4 Illustration of direct excitation
Figure 6-3 shows the indirect excitation solution. Indirect excitation m easurem ent
induces currents on conducting paths on PCB by one o f the loop pair, and generates
m agnetic near-field pattern by the other loop. Indirect sources d o n ’t need to touch
106
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
circuits. Indirect excitation has some draw backs which are w eak induced single and
interfere due to direct source-loop coupling.
In contrast, direct excitation can induce currents on conducting paths on a printed
circuit board as shown in Figure 6-4. The advantage o f direct excitation is strong
coupling signal. The drawbacks are that it needs to contact PCB and needs bed of nails
for exciting.
6.2 Measurements on Indirect excitation
6.2.1 M agnetic Coupling betw een L oops over a PCB Ground
Plane
7-
Figure 6-5 Indirect excitation over a g r o u n d plane
The purpose o f the m easurem ents conducted in this section is to investigate the
effects of ground plane to indirect excitation. As shown in Figure 6-5, the ground plane
107
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
tends to repel the magnetic field away with the induced eddy-currents. H owever, the
m agnetic field can still go through the ground plane if the ground plane thickness is less
than a few skin depths [60] . For copper the skin depth can be calculated by
S = -yjl/ (Oficr , where
go
= 2nf, |i ~ p 0 = 4 n x l0 7 W b/m, and o ~ 5 .8 x l0 7 S/m. The
thickness o f copper sheet on the PCB tested is about 17 pm for a
¥2
oz. cladding. W hen
excited at 100 M Hz, the skin depth o f the copper sheet is about 6.6 pm , so the effect o f
ground plane can be m odelled by perfect im age current as shown in Figure 6-5.
Considering the scenario shown in Figure 6-6, tw o loop probes are placed over a PCB
ground plane. One probe is sitting relatively at xo, the other is movable and is m oved
from xi to
The port 1 o f netw ork analyzer feeds the fixed probe for indirect excitation
and the port 2 is connected to the movable probe for collecting m agnetically induced
signal. The ring surface norm als are parallel. Provided that the left edge of PCB is
m arked as x = 0, the fixed probe is located at x0 = 12 m m , the movable probe is scanning
from xi = 25 m m to X2 = 125 mm. Two loops are about h = 8 mm above the plane.
The equivalent circuit for Figure 6-6 is shown in Figure 6-7.
W iltron N etw ork
A nalyzer
4 8 6 PC
fixed probe
m o v a b le p ro b e
■sivrnai
=i=-
norrujfc'
G ro u n d p lan e
Figure 6-6 Two loop probes over a large ground plane, ring surface
normals are parallel
108
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
V,
50Q
5Q£2 5 0 Q
I,
I—^
+
V,
3 M 12
p M21
* L ii
V,
L 22 '
50Q
I
Figure 6-7 Equivalent circuit of two probes over a ground plane
The relations between Vj, V2, h , h can be expressed by the network theory as
following,
~M n ~
’A"
_A_
(6- 1)
t-J
1_
> r _ d ' hi
dt - M n
the mutual inductance between loops is [61]
M 21
]2 — -— ^ B io o p iir ^ d s '
A s2
n, - n 2 d V
(6 - 2)
f i 0AiA2
3
( « r « 2)
V“ -l ” 2
where i? is the distance between tw o loops, A/ and A? are the areas o f tw o loops,
and
, n 2 are the norm als to loop 1 and loop 2, respectively.
Given the same geometry, the self inductance of each loop is [61]
109
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
where r is the ring radius of loops and d is the wire diameter.
From the equivalent circuit in Figure 6-7, Vj and V2 can be calculated.
y =
j(° Ln
Vs + ------—
jO)Mn I 0
1 50 + j(t)Ln s 5 0 + jO)Lu
2 -
(6-4)
V2 = -5 0 1 2 = gA 50 y jcoM 2lI x
5 0 + jcoL 22
(6-5)
under unilateral approxim ation, which assumes loop 2 does not give any loopback
coupling to the source circuit, lo o p l, w e will have the approximation
/,
1
V
^
50 + jcoLv
(6-6)
substitute (6-6) into (6-5),
V,
V
jcoM 21-------- ^-----50 + j(oL22
50 + j(oLu
50
(6-7)
Hence, S 21 = V? / Vs can then be calculated with
5,, ~ —j c o M
-1
—--------------------------------------50 + jo)Ln 50 + j(t)Ln
The above discussion does not take into account the effects o f self-inductance and
mutual inductance o f the images. As shown in Figure 6-8, when considering the effects of
I 10
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
(6-8)
im ages, the self-inductance and m utual inductance equations (6-2) and (6-3) must be
recalculated.
2h
Figure 6-8 Two probes over a ground plane and their images
/6>AA
m :;™*
(6-9)
Atz^ R 1 +(2h)2)3
im age
r im
Lu
M,
11
Mo A A
(6 - 10)
47t{2hf
MoAA
4;r
0 V « 2)
(6 - 11)
(^R2 +(2h)2)3
( 6 - 12)
= ^rfln
—
- 2 . 4 5 )An(2h)
-^ A
0
d
Figure 6-9 shows both theoretical and experimental results o f S 21 - The calculations o f
either with the consideration o f im age effects (Equations (6-11) and (6-12)) or without
111
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
the consideration o f current image effects (Equations (6-2) and (6-3)) are plotted.
Regardless o f the m agnitude offset, the experimental result agrees well with the
numerical result. Figure 6-9 also shows that S 21 is at the noise level when the distance
between tw o loops is greater than 50 mm, and there is m inor magnetic coupling between
them. W hen the distance is less than 50 mm, the inter-probe coupling (interfere) is
dominant.
-20
—
—
Cal. w / image
Cal. w/o image
Measured
-60
-80
-100
IJ_
-120
-140
20
40
60
80
100
120
Distance between two loops (mm)
Figure 6-9 The numerical simulation vs. experimental data for two loops
over a g r o u n d plane and ring surface normals are parallel
A measurement was also conducted for the scenario w here ring surface norm als are
orthogonal, as shown in Figure 6-10. Sim ilarly, the fixed probe was located at Xo = 12
mm, the movable probe was scanning from Xi = 27 m m to x2 = 127 mm , tw o loops were
at 8 mm above the circuit plane. The test result is shown in Figure 6-11. As expected
from Equation (6-2), there was no magnetic coupling betw een two loops when their n ng
surface normals are perpendicular to each other, and S 21 plot is at the noise level.
112
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
W iltron N etw ork
A nalyzer
486 PC
fixed pro he
■m>\ able probe
•=^
G round plane
Figure 6-10 Two loop probes over a large ground plane, ring surface
normals are orthogonal
-65
-70
-75
-80 -85
v
-90
-95
u_
-105
-110
40
20
60
80
100
120
D istance b e tw e e n two loops (m m )
Figure 6-11
S 2i
plot f o r
over a ground plane and ring surface
normals are orthogonal
tw o lo o p s
Trying to enhance the perform ance of the m easuring system , a m ulti-loop (four-loop)
probe was also made and tested to be used as the m ovable probe. M easurem ents were
113
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
repeated w ith the four-loop probe for both configurations where normal vectors to ring
surfaces are parallel and orthogonal. Unless mentioned, the setups are the same as shown
in Figure 6-6 and Figure 6-10. The single-loop probe was fixed at xo = 0 mm, and the
movable four-loop probe was scanning from xi = 15 m m to X2 = 115 mm. Two probes
were at 8 m m above the ground plane. Figure 6-12 and Figure 6-13 show the single loop
probe vs. the four loop probe performance.
-35
-40
single-loop probe
four-loop probe
cq -45
-50
E -55
-80
-85
20
40
60
80
100
120
Distance between two loops (mm)
Figure 6-12 Performance comparison of single-loop and four-loop probes,
ring surface normals are parallel
It is noticed in Figure 6-12 that the inter-probe coupling level between two single­
loop probes does not change significantly by using a single-loop or a four-loop probe.
Provided that the self-inductance o f a single loop, i.e. loop 2 is L2 2 and the mutual
inductance between tw o single loop, i.e. loop 2 and loop 1 is M 2i, when loop 2 is
replaced with an N toroid loop, i.e. N = 4, the self-inductance o f the 4-loop probe will be
N 2L 22 , assuming that the loop 2 is an infinite thin w ire loop. The mutual inductance
114
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
between loop 2 and loop 1 is now N M 21 . Equation (6-8) can be rewritten for the four loop
configuration.
S*\onp « - j o M « r — — ^ - 7 ,------------ l—
21
1
5 0 + j(o L ^°p 5 0 + jo)Ln
(6-13)
L 2 2 turns out to be approxim ately 0.047 p H by Equation (6-3) given the ring
param eters shown in Figure 6-1. The inductance [jcoZ^I o f a single loop is roughly 30 Q
at frequency o f 100 M Hz. The S2i for both configurations can be calculated by Equation
(6-8) and (6-13).
50
1*^211
1
|5 0 + jco L 221 |50+ jtuL,,!
= \a)M 2l\ - - r J 2 = - ]-------1----V 502 + 302 |5 0 + j'tyL,,)
= 0.86|tyM 2 1 1
(6-14)
1
|50 + ja)Ln
1 1
1
50
S “oop\= \a)M *oop'
1 |50 + j a ) L ^ op\ |50 + jw L x\
= 4| a)M2
;
1
50
|5 0 + jl6 (o L n \ |5 0 + j(oLu \
(6-15)
200
1
= \o M ,, I
------------ .
.
V 502 + 4 8 0 2 |50 + j(t)Ln \
= 0 .4 lb M . . 1• ---------------r
1
2,1 \5 0 + ja )L u \
In actual loop probe design, the wire cannot be infinitely
thin and hence the self
inductance L 2 2 will be between N 2L22 and N L22 - The |s 2i4l0Op|will then be closer to |S2i|.
This gives an approxim ate explanation to m easurem ent result in Figure 6-12.
115
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Note th a t from Equation (6-15), S 21 is very sensitive to L 22 value accuracy, in another
words, is sensitive to the value of (0 L 22 vs. Zo. If
00
L 22 «
Zo, then the inter-loop coupling
will increase 4 tim es (~ +12 dB), while the coupling will decrease 4 times (~ -12 dB) if
(0 L 2 2 »
Zq.
Figure 6-13 shows the result when normal vectors to ring surfaces are orthogonal. As
expected from Equation (6-2) there is no coupling between two loops and S 21 plot is at
the noise level for this test configuration.
_ -a -
-€S
four—bop probe
single-loop profc
—
70
-9 5
-100
-1 0 5
-110
0
£0
eo
40
eo
100
D tetaroca ( m m )
Figure 6-13 Performance comparison of single-loop and four-loop probes,
ring surface normals are orthogonal
116
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
120
6.2.2 M agnetic Coupling betw een loops over a PCB Trace with
B oth E nds S horted to G round
This section will investigate the feasibility o f using tw o loop probes to detect PCB
trace failure. Instead of a large ground plane, a circuit board was used. This circuit board
consists o f a dielectric layer with dielectric constant er = 3.3, a copper layer on one side
as ground plane, and a 10 cm long by 1 m m w ide circuit trace on the other side. Both
ends of the trace were shorted to ground.
W iltron N etw ork
A nalyzer
*■ 4 8 6 PC
fixed probe
/
'v *
>.
rjliavable probe
A t
II
:------“— "
xw
1~ i|, i j stage
TT-
C o p p er layer
t
^
Figure 6-14 Two loop probes over the circuit trace with both ends shorted
6.2.2.1 Frequency Sweeping to Measure the Coupling Characteristic
As shown in Figure 6-14, tw o probes are at 8 mm above the circuit board. The fixed
probe is at xo, and the movable probe is at xi or x2. Port 1 of netw ork analyzer feeds the
fixed probe for excitation, port 2 w as connected to the m ovable one to collect the
inductive coupling signal.
The sweeping frequency is from 40 M H z to 1000 M Hz. Tw o measurements were
taken for the m ovable probe at tw o spots xi and x2 (Figure 6-14), xi is 7.5 cm away from
Xo along the trace, while x2 is 2 cm aw ay from Xi along the ring surface normal.
117
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
As show n in Figure 6-15, plot A is the m easurem ent o f m ovable probe being at xi and
plot B is fo r the m easurem ent at x2. The peaks around -50 dB are the trace resonant
frequencies. The first resonant frequency is at about 300 M Hz. W hen / < 300 M Hz, the
w avelength is less than 1 m, there will be no resonance, so the trace can then be treated as
lum ped coil (loop) inductive element. If / > 300 M Hz it will see line resonance effects.
Hence, the operation should be in quasi-static region, for which frequency is less than
250 M H z to have the inter-loop coupling m easured independent o f trace resonance.
START!
STOP:
STEP:
0 , 0 4 0 0 S M3!
1 . 0 0 DO QHZ
0 .0 1 3 0 GHZ
B A T S ST A R T !
GATS STOP:
GATE;
HXNpow:
ERROR CORFfc NONE
A V E R A fllN B :
* PT3
I F BNDWOTS-C REDUCED
8 2 1 FORWARD TRAMBMIS31 ON
oR0F+*-40 .OOOdB
1 0 . 0 0 0 0 B /D IV
SHARKER 3
o .a s e o SHE
- 4 6 . 2 8 3 cfS
MARKER TO MAX
MARKER TO MIN
1 0 .1 0 0 0 OHS
1
- e e 1.803 6B
F ig u r e 6-15 F r e q u e n c y s w e e p in g r e s u lts by tw o single-loop probes
118
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
6.2.2.2 Scanning along the PCB Trace
W hen tw o conductive wire loops are right over a trace, as shown in Figure 6-16, the
relations betw een Vj, V2, V, and /;, I2, It can be w ritten as,
~LU
Mn
Mu'
_ d M 21
L 22
M 2t
12
Ln
> r
~dt
y*.
Mn
m
(6-16)
/,
_
50Q 5QQ
L
-2dx= 0
Figure 6-16 Two loops over a PCB trace
For case that excited frequency is well below the resonance frequency,/ < 300 M Hz,
the PCB trace can be treated as a lum ped param eter device by ignoring the distributed
nature o f the transm ission line and thus can be m odelled as a single loop. The inductance
of the loop is L t and loop area is 2d-L.
Since the trace is shorted to ground, Vt = 0, hence
0 = |- [ M „ A + M ,]
at
=>
1 ,= -^ —
L„
119
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
given that Mbt s M ti,
I
v ,lo:aTp 2 =— - —j ® M t2<t
(6- 18)
-
Ls,u
50 + jcoL, 22 v % '2
V ^ = joM % _
Vs
L„
(6- i9)
50--------------- 1------50 + jcoL22 50 + j(oLu
w here L„ is the inductance per unit length o f m icrostrip t-line, and
, _ MqA2
= Mo ^ 2 ff —_ £2d
z _ _ —dx
M
11
..2
4^ JJa/(Vjc
2+/z2)3.
i f f
.
r 2 s3
4tt
2d
(6-21)
h 2J x 2 + h 2
From Equation (6-20), as the substrate thickness d is in order o f 1 mm, much less than
the distance h (about 8 m m ) betw een the loop and the trace, and also much less than the
PCB trace length L, the value o f M a can be approxim ately estim ated to be a very small
value. This will be dem onstrated in the experim ents that follow.
Figure 6-17 shows the num erical result of magnetic coupling between the PCB trace
and the loop probe by Equation (6-20). Provided that L t, is in order of 100 nH for the
PCB trace under test, the m agnetic coupling is very small at -76 dB level.
120
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
-78
2,
£ -80
-82
"E -84
-88
-90
40
60
Distance (mm)
80
100
120
Figure 6-17 Calculated magnetic coupling between the PCB trace and the
movable loop probe
Sim ilar to the experim ental investigation done in 6.2.1, both single loop and four-loop
movable probe were used for m easurem ent o f the PCB trace. The fixed probe was at xo =
1.0 cm, the m ovable probe was scanning from xi =2.6 cm to x2 = 9.5 cm. These tw o
loops were 0.8 cm high right above the circuit trace. Figure 6-18 shows the m easurem ent
results when tw o ring surface norm als were parallel. There is a 3-4 dB enhancem ent
when using a four-loop probe in place of a single-loop probe. The S2i plot is at noise
level when the distance between two probes is greater than 50 mm, and there is m inor
magnetic coupling between them. W hile the distance is less than 50 mm, the inter-probe
coupling (interfere) is dominant.
121
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
four-bop probe
single-loop probe
-5 0
E
-7 0
—9 0
-100
20
70
30
00
Figure 6-18 Performance comparison of single-loop and four-loop probes
scanning along the trace, ring surface normals are parallel
M easurem ent was also conducted for scenario that ring surface norm als were
orthogonal. Figure 6-19 shows there was m inor coupling between loops and between
loops and trace.
122
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
—65
-eo
4
fo u r -b o p probe
sin g le -lo o p probe;
1
r 1 .... r"....... - i
If
-65
K rt
-70
co
P.
i
©
&
4
j
1 M1
A
1
2 -so
v.
>t 'ifA „ / k1i
i
A
^
" !
*
i
-9 6
........
i
&1 \
i
A
i
•
:
f|f
1?
-90
100
120
\
[
1 ................. i ..................
-100
-106
4
........................ ...................
CM
<A
(t
*2**
-75
E
o
20
eo
40
80
Dfeta res £rnrn)
Figure 6-19 Performance comparison of single-loop and four-loop probes
scanning along the trace, ring surface normals are orthogonal
W iltron N etw o rk
A nalyzer
4
>
4 8 6 PC
I't
i»
i»
fixed probe
ijLil>\.ihle p in h c
x0'
.x iijy
p?
11■>
^Scan-mie
stage
?«5cT!l ‘
C o p p er layer
/
Figure 6-20 Scanning along the path aside from but parallel to the trace
1.23
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
To investigate the coupling level of scanning right over the trace from that of
scanning along a path aside from but parallel to the trace, a measurement also was
conducted as shown in Figure 6-20. In the m easurem ent two loops right over the circuit
trace, the fixed probe was located at xo = 3.6 cm , the m ovable probe was scanning from
xi = 5.3 cm to x2 = 12.3 cm. These two loops w ere directly over the trace at height of h =
0.8 cm.
In the test that tw o loops were both 3.5 cm aw ay from the trace, the fixed probe was
located at xo’ = 5.3 cm , the m ovable probe was scanning from x f = 7.0 cm to x2’ = 14.0
cm. Figure 6-21 shows the scanning results o f the S2i forward transm ission ratio versus
the distance between the fixed loop and the m ovable loop.
-4 0
right over the trace
far aw ay from the tra c;
C
.Q
.1
E
IS
JS -7 0
■0
§
£
10
20
30
40
50
60
70
80
Dfetanca (mm)
Figure 6-21 D is ta n c e scanning results b y two single-loop probes in the
case of circuit trace with each end short circuited
124
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
90
6.3 Measurement on Direct excitation
This section will discuss some m easurem ent results on direct excitation. As shown in
Figure 6-22, only one single-loop probe was used in the measurement. The circuit board
is the sam e as that used in 6.2. The p o rtl o f netw ork analyzer was directly connected to
one of the trace ends, and the port2 was connected to the movable probe to collect the
coupling signal from the trace. The other end o f the circuit was short circuited. Two
m easurem ents were conducted for this configuration:
•
Frequency sweeping to m easure the coupling characteristic versus the
stim ulus frequency.
•
Transverse scanning over the circuit trace.
W iltron N etw ork
A nalyzer
486 PC
m n \a M c pivho
slas>
C opper layer
Figure 6-22 Measurement on direct excitation
6.3.1 F requency Sw eeping to M easure th e Coupling
C haracteristic
The frequency sweeping m easurem ent w as taken at X] and X2 (Figure 6-22). Xj is 2.5
cm away from the signal feeding end, w hile X2 is 7.5 cm away from the end. The loop
was 0.7 cm high over the circuit trace. The sw eeping frequency is from 40 M Hz to 1000
125
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
M Hz. As show n in Figure 6-23, Plot A is the result at xi and Plot B is the result at x 2- It is
expected to see the similar results explained in 6.2.2.1.
BTA RT:
stop:
s te p :
0 . 0 4 0 0 GKfc
i.O D D O o m
Q .O iB O GHZ
GATE START:
bate sto p:
GATE:
w in d o w :
ERROR CORRj
A v e n A a lN o :
WONE
IF
REDUCED
BMOWDTHt
i
pts
CH 1
R E P . PLANE
aaa
forw ard tr a n sm
0 .0 0 0 0 am
ro s io n
1 0 .O O O O S /D IV
P R E F - 0 .OOOdB
^MARKER 2
0 .4 0 0 0 QHz
- B O . 4 3 4 dB
MARKER TO MAK
MARKER TO MIN
t
-8 0 .0 7 9 4)0
----- 1-----
Figure 6-23 Frequency sweeping for a single-loop probe from 40 MHz to
1000MHz
Figure 6-24 is a zoom ed-in plot for 40 M H z < / < 200 M Hz.
126
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
START!
STO P!
STEP:
0 .0 4 0 0
O .S O O O
0 .0 0 2 4
BHss
8H Z
GH Z
B A TS START!
SA TE STOW
GATE:
E R R O R CORR; NONE
AVSBAfflXMSI; i R T S
I F B M O H PT H : R E D U C E D
window:
OH 4
REF.
s s a.
PLA N E
0 . COO© w»m
S S I FORWARD TR&N8MIBBH3M
» R E F -~ ’5 0 . 0 0 0 d Q
S .O O O d B / D X V
>MAHK6R 2
0 . 1000
0 H j£
- 0 0 . STi dS
MARKER TO MAX
M ARKER T O M IN
Q.CMOO GHZ
- 8 6 . BV6 dB
“T
Figure 6-24 Frequency sweeping for a single-loop probe from 40MHz to
200MHz
6.3.2 T ran sv erse S canning a c ro s s th e PCB Trace
Transverse scanning over circuit trace with a single loop probe for direct excitation
was conducted as shown in Figure 6-25. The intersection xj o f the scanning path and
circuit trace was 2.5 cm away from the signal feeding end. The scanning range was 10
cm, 5 cm aside on each side o f the trace.
127
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
W iltron N etw ork
A nalyzer
4 86 PC
m ovable probe
C opper layer
F ig u r e
6-25 Transverse scanning across the circuit trace
Figure 6-26 An approximate model for magnetic field calculation without
consideration of image current e ffe c t
An approxim ate magnetic field distribution over the PCB surface is shown in Figure
6-26 when looking at the cross section of the PCB trace. W ith the consideration of image
128
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
current effect, the closed magnetic field curls in the very near field region can be
considered approxim ately as a group of circles having a com mon tangential line at point
(y = 0, z = 0). Provided that the electric current is along x axis, the y com ponent By of the
magnetic field B at the central point of the loop can be form ulated when the probe is at y.
B
-£ o l_ co sa
7
(6-22)
2 nr
where I is the current in circuit trace, r ’ is the distance between the central point o f the
loop and circuit trace, and a is the angle form ed by the vector B and y axis.
For a closed m agnetic field curl, the radius r is
r = h + A z - *j y2 + (A z)2
(6-23)
Az can then be expressed as
v2- h 2
Az = ^ — —
2h
(6-24)
Substitute Az in (6-23) with (6-24),
t
a
y 2+h2
r = h + Az = —-------2h
(6-25)
Hence,
B y -
;:■■■■cos a =
p =
....
2ft-yjy2 + h 2
2f t -Jy2 + h 2 y~ + h
129
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
(6-26)
-30
-40
-80
-90
-40
-30
-20
20
-10
Distance (mm)
Figure 6-27 Simulation result of transverse scan over the circuit trace with
a single-loop probe
_ .£ £ ,! _____________I_____________ 1____________ 1_____________I_____________i____________ I_____________ 1_____________ I------------------- 1-------------------
-5 0
-4 0
-3 0
-2 0
-1 0
0
10
20
30
40
60
□ s e ts n e e (m m )
Figure 6-28 Measurement result of transverse scan over the circuit trace
with a single-loop probe
130
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
Figure 6-27 is the m agnitude plot of By calculated by (6-26). The coupled current
signal to th e probe will be proportional to By. Figure 6-28 is the m easurem ent result. W ith
the assum ptions fo r the very near field region the m easurem ent result agrees w ith the
sim ulated result.
6.4 Conclusion
In this chapter, both indirect and direct excitation measurements with m agnetic
probes w ere investigated as an alternative to capacitively coupled probing technique.
Frequency sw eeping perform ed indicated that to elim inate trace resonance at high
frequency w here PCB traces cannot be treated as lum ped coil (loop) inductive elem ents,
the m easurem ent w ith m agnetic probes can be only done at frequency low er than 300
MHz.
The m easurem ent and simulation results also illustrated that in the indirect excitation
measurement, the inter-probe coupling (interfere) is dom inant when two probes were
close to each other. Increasing the loop coil num ber of probe did not enhance the
performance very m uch, for instance, using a four coil loop was giving a 3dB
performance enhancem ent.
131
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
C h a p te r 7
C o n c l u s io n s
T he feasibility o f the applications of electrom agnetically coupled probes as diagnostic
tools fo r a broad range o f m icrowave circuits has been investigated. Unlike a traditional
circuit m easurem ent technique, non-contact near-field probes are much less likely to
disturb the norm al operation of the device under test at Gigahertz frequency.
An autom ated near-field probing system has been designed and built capable of
producing inform ation unavailable from conventional m easurem ent techniques. A system
transfer model was created to prove that the transfer function o f device under test can be
obtained provided that the transfer function of near-field probe is derivable from
modelling. B y using a m ethod o f m om ent (M OM ) solution, T. Forzley has given a
approxim ate model for various heights o f the probe above the transm ission line and
incremental displacem ents relative to the center o f the transm ission line [5] . The system
hardware and softw are were built in a m odule-oriented m anner such that the system was
flexible and scalable to facilitate testing on w ide range of m icrowave devices. The system
is capable o f one-dim ension and tw o-dim ension scanning for mm and p m resolution
applications. The software developed based on Lab View
is user friendly and capable of
com m unicating to various m easurem ent instrum ents, i.e. Network Analyzer, Spectrum
Analyzer and Oscilloscope.
V arious microwave devices have been tested by monopole and CPW probes with the
automated near-field scanning system. The perform ance of both m onopole and CPW
probes were investigated by m easuring C PW
transm ission
lines. T o verify the
measurement results, CPW transm ission line analysis was perform ed. The transverse
electric field variation o f CPW transm ission line was m odeled by solving the Laplace
boundary value (BV) problem using a finite difference M OM approach with a SOR
Iterative Matrix Solver [40] . The longitudinal electric field variation m odelling was
132
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
perform ed by traditional transm ission line theory. Both transverse and longitudinal scan
over the C PW transm ission lines w ere conducted and the results were com pared with the
num erical models. The test results recom m end that the CPW probe and monopole probe
have a com parable m easurem ent perform ance, thus m icron-m eter electric near-field
probes can be fabricated in a form o f planar structure through the micromachining
process to improve the spatial resolution. In addition, the probes were also used to
m easure the GHz single patch antennas and m icrostrip antenna arrays. The test results
illustrate further proof to the feasibility o f electric near-field probe application.
As an alternative to the electric near-field probing technique, the m agnetic near-field
probing has been also investigated to determ ine detection capabilities of faults in Multi
Layers Boards [58] [59] . The signal frequency in such applications is limited up to 300
M Hz range, and strong coupling betw een m agnetic probes has been found to have
considerable im pact on the perform ance o f this near-field m easurem ent approach
A m icrom achined electric near-field probe designed and fabricated at Alberta
M icroelectronics Centre has been evaluated at scale less than 100 um measurement. The
electric coupling between the probe and the transm ission line under test was not strong
enough and data acquired was at noise level. This was m ainly due to the structure o f the
microm achined cantilever. There w as no probe tip constructed since the signal line and
the ground lines had same length. The probe has to be redesigned and fabricated in the
future work. O ther than that, the gap between the cantilever and the CPW circuit plane
has also to be precisely controlled, and the vibration o f such a tiny cantilever has to be
reduced as much as possible.
133
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
R eferen ces
[1] B. T. Rosner, D. W. van der W eide, “H igh-frequency near-field m icroscopy” ,
R eview of Scientific Instrum ents, Vol. 73, No. 7, July 2002, pp.2505-2525
[2] N. Richardson, “E-beam probing for VLSI circuit debug” , VLSI System design,
V ol. 8, Aug, 1987, pp.24
[3]
W . T. Lee, “Engineering a device for electron-beam probing”, IEEE design and
test o f Computers, June 1989, pp.36-48
[4]
H. Todokoro, S. Fukuhara, and T. Komoda, “Electron-beam LSI tester,” Japan
annual reviews in electronic, com puters and telecom m unications, Vol. 13, edited
by J. Nishizawa, N orth-H oli and, 1984, pp.373-382
[5] T. S. Forzley, “N on-Invasive D iagnostics o f M icrow ave Integrated Circuits Using
a Scanning N ear-Field Probe”, M .Sc. thesis, D epartm ent of Electrical and
C om puter Engineering, U niversity of M anitoba, 1992
[6]
K.U ra and H. Fujioka, “Function testing o f sem iconductor devices with
stroboscopic scanning electron m icroscope” , Japan annual reviews in electronics,
com puters and telecom m unications, edited by J. N ishizaw a, North-Holland, 1982,
pp.308-327
[7]
G. E. Bridges and D. J. Thom son, “H igh-frequency circuit characterization using
the AFM as a reactive near-field probe” , U ltram icroscopy 42-44, 1992, pp.321328
[8]
G. E. Bridges, T. S. Forzley, and D. J. Thom son, “N ovel near-field probe for onw afer intergrated circuit m easurem ents” , M icroeletronics Journal, Vol. 23, No. 5,
1992, pp.363-369
[9]
C. Bohm, C. Roths, E. Kubalek, “Contactless electrical characterization of
M M ICs by device internal electrical sam pling dcanning-force-m icroscopy”, IEEE
M TT-S Digest 1994, pp. 1605-1608
' [10] D. W. van der W eide and P. Neuzil, “The nanoscilloscope: Com bined topography
and AC field probing with a m icrom achined tip” , J. Vac. Sci. technol. B 14(6),
Nov/Dec 1996, p p .4 144-4147
134
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
[11] A. Leyk, C. Bohm , D. W. van der W eide, and E. Kubalek, “ 104 GHz signals
m easured by high frequency scanning force m icroscope test system ,” Electronics
Letters, vol. 31, 1995, pp. 1046-1047
[12] D. W . van der W eide, “Localized picosecond resolution with a near-field microw ave/scanning-force m icroscope”, Appl. Phys. Lett., Vol. 70, No. 6, 10 February
1997, pp.677-679
[13] R. A. Said, G. E. Gridges and D. J. Thom pson, “Noninvasive scanned probe
potentiom etry for integrated circuit diagnostics,” IEEE Transactions on
Instrum entation and M easurem ent, Vol. 43, No. 3, June 1994, pp.469-474
[14] G. E. Bridges, R. A. Said, M. M ittal and D. J. Thom pson, “H igh-frquency pattern
extraction in digital integrated circuits using scanning electrostatic force m icros­
copy,” J. Vac. Sci. Technol. B 13(3), M ay/June, 1995, pp.1375-1379
[15] G. E. Bridges, R. A. Said, M. M ittal, and D. J. Thom pson, “Sampled w aveform
m easurem ent in integrated circuits using heterodyne electrostatic force m icro­
scopy,” Rev. Sci. In strum. 65 (11), Nov. 1994, pp.3378-3381
[16] R. A. Said, G. E. G ridges and D. J. Thom pson, “High frequency potential probe
using electrostatic force m icroscopy,” J. Vac. Sci. Technol. A 12(4), Jul/A ug,
1994, pp.2591-2,594
[17] D. N oruttun, G. Bridges, and etc., “Failure analysis of high speed
m icroelectronics using non-contact probing,” 1997 M icronet Annual W orkshop,
Ottawa, 1997, p p .53-54
[18] W. M ertin, A. Leyk, and etc., ‘T w o-dim ensional m apping o f am plitude and phase
of microwave fields inside a M M IC using the direct electro-optic sam pling tech­
nique” , IEEE M TT-S D igest 1994, pp.1597-1600
[19] G. David, W .Schroeder, and etc., “2D electro-optic probing com bined with field
theory based m ultim ode wave am plitude extraction: a mew approach to on-w afer
m easurem ent” , IEEE M TT-S D igest 1995, pp.1049-1042
[20] P. O. M uller, D. Eras me, and B. Huyart, “W avelength oprim ized electro-optic
scanning of M M IC s with fabry-perot enhancem ent”, IEEE M TT-S D igest 1997,
pp.1329-1332
[21] K. Yang, D. D avid, S. Robertson, J. F. W hitaker and L. P. B. K atehi., “H ighresolution electro-optic mapping o f near-field distributions in integrated
mocrowave circuits”, IEEE M icrow ave Theory and Techniques D igest, 1998,
pp.949-952
135
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
[22] K. Y ang, G. David, et al, “Electrooptic mapping and finite-elem ent m odeling of
the near-field pattern o f a m icrostrip patch antenna”, IE E E Transactions on
M icrow ave Theory and Techniques, Vol. 48, No. 2, February 2000, pp.288-294
[23] K. Y ang, T. M arshall, et al, “A ctive-amplifier-array diagnostics using highresolution electrooptic field m apping” , IEEE Transactions on M icrowave Theory
and Techniques, Vol. 49, No. 5, M ay 2001, pp.849-857
[24] R. Reano, K. Yang, et al, “Sim ultaneous measurements o f electric and thermal
field utilizing an electrooptic sem iconductor probe” , IE E E Transactions on
M icrow ave Theory and Techniques, Vol. 49, No. 12, D ecem ber 2001, pp.25232531
[25] T. P. B udka and G. M. Rebeiz, “A m icrowave circuit electric field im ager” , IEEE
M TT-S D igest 1995, pp. 1139-1142
[26] T. P. Budka, E. M. Tentzeris and etc., “N ear-field m apping above a coupled-line
filter and a M M IC” , M icrow ave Journal, March 1998, pp.94-106
[27] T. P. Budka, E. M. Tentzeris, and etc., “An experim ental and theoretical com pari­
son o f the electric fields above a coupled line bandpass filter” , IEEE M TT-S
D igest 1995, pp. 1487-1490
[28] S. S. Osofsky and S. E. Schw arz, “D esign and perform ance o f a non-contacting
probe for m easurem ents on high-frequency planar circuits” , IEEE Transacrions on
M icrow ave Theory and Techniques, Vol. 40, No. 8, A ugust 1992, pp. 1701-1708
[29] R.Grzybowski and R.Bansal, “M agnetic field probe for m easuring surface current
distribution on millimetre wave m icrostrip antennas,” Electronics Lett., vol. 27,
Jan. 1991, p p .7 1-73
[30] Y. G ao and I. W olff, “A new m iniature magnetic field probe for m easuring threedim ensional fields in planar high-frequency circuits”, IEEE Transacrions on
M icrow ave Theory and Techniques, Vol. 44, No. 6, June 1996, pp.911-918
[31] J. S. D ahele and A. L. Cullen, “Electric probe m easurem ents on m icrostrip” , IEEE
Transacrions on M icrow ave Theory and Techniques, Vol. M TT-28, No. 7, July
1980, pp.752-755
[32] Y. G ao and I. Wolff, “M iniature electric near-field probes for m easuring 3-D
fields in planar m icrowave circuits” , IEEE Transacrions on M icrow ave Theory
and Techniques, Vol. 46, No. 7, July 1998, pp.907-913
136
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
[33] Y. G ao and I. W olff, “Calibration of Electric Coaxial N ear-Field Probes and
A pplications” , IEEE Transacrions on M icrowave Theory and Techniques, Vol.
46, No. 11, N ovem ber 1998, pp. 1694-1703
[34] K. karrai and R. D. Grober, “Piezoelectric tip-sam ple distance control for near
fie ld optical m icroscopes”, Appl. Phys. Lett., Vol. 66, No. 14, 3 April 1995,
pp. 1882- 1884
[35] T. W ei and X. D. Xiang, “Scanning tip m icrowave near-field m icroscope” , Appl.
P hys. Lett., Vol. 68, No. 24, 10 June 1996, pp.3506-3508
[36] S. H ubbell and D. J. A ngelakos, “A technique for m easuring the effictive
dielectric constant of a m icrostrip line” , IEEE Transacrions on M icrow ave Theory
and Techniques, Vol. M TT-31, No. 8, A ugust 1983, pp.687-688
[37] P. G . Frayne and J W hitehurst, “M ode analysis in radiating and no traiating planar
open structures using a scanning m icrowave probe” , 5th IC A P Conf., York, 1987,
p p .501-505
[38] P. G . Frayne, “H andbook of M icrostrip Antennas - M icrostrip field diagnostics” ,
[39] E. Levine, “H andbook o f M icrostrip Antennas - Special m easurem ent techniques
for printed antennas”,
[40] P. M . Knupp and S. Steinberg, “The fundam entals of grid generation”, CRC Press
LLC , 1993
[41] C. A. Balanis, “A ntenna Theory: A nalysis and D esign,” Second edition. John
W iley & Sons, 1997
[42] J. Zurcher and F. E. G ardiol, “B roadband Patch A ntennas,” Atrech House, 1995
[43] D. Gray, “O versized elem ents for hig efficiency EH F boresight array” , Ph.D.
candidacy report, D epartm ent of Electrical and Com puter Engineering, University
o f M anitoba, 2000
[44] A. Sebak, “Simulation o f EM em issions from printed board traces”, A N T EM ’98Symponium on A ntenna Technology and A pplied Electrom aginetics, pp.585-588
[45] F. Lan, etc., “A new m ethod for tapered open-ended coaxial line probe in
determining relative com plex perm ittivity at radio and micrewave frequencies” ,
A N TEM ’98-Sym ponium on A ntenna Technology and A pplied Electrom aginetics,
pp.497-500
137
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
[46] K. E . Petersen, “Sillicon as a mechanical m aterial,” Proc. IEEE, Vol. 70, No. 5,
M ay , 1982, pp.420-457
[47] G abriel M. Rebeiz, “M icrom achining for microwave and m illim eter-wave
applications” , A N TEM ’98-Symponium on A ntenna Technology and A pplied
E lectrom aginetics, pp. 11-14
[48] V eljko M ilanovic, ...etc. “M icrom achined microwave transm ission lines in
C M O S technology” , IEEE Transacrions on M icrow ave Theory and Techniques,
V ol. 45, No. 5, M ay 1997, pp.630-634
[49] K yung S. Oh, etc., “Computation o f the equivalent capacitance o f a via in a m ulti­
layered board using the closed-form G reen ’s function” , IEEE Transacrions on
M icrow ave Theory and Techniques, Vol. 44, No. 2, February 1996, pp.347-349
[50] B. J. Feenstra, etc., “Near-field scanning m icrew ave microscopy: m easuring local
m icrow ave properties and electric field distributions”, IEEE M icrowave Theory
and Techniques Digest, 1998, pp.965-968
[51] A. F. Lann, M. Golosovsky, and etc., “M apping the thickness o f conducting
layers by a m m -wave near-field m ocroscope”, IEEE M TT-S D igest 1998,
pp. 1337-1340
[52] G. D avid, J. F. W hitaker, and etc., “D C-to-m m -w ave absolute potential m easure­
m ents insde digital micrewave ICs using a m icrom achined photoconductive sam ­
pling probe” , IEEE M TT-S Digest 1998, pp.1333-1336
[53] M. G olosovsky and D. Davidov, “M M -w ave near-field scanning resistuvity
m ocroscope” , IEEE M TT-S D igest 1997, pp. 1333-1334
[54] T. Pfeifer, H. M. Heiliger, and etc, “H igh-frequency on-w afer testing with freely
prositionable silicon-on-sapphire photoconductive probes”, IEEE M TT-S D igest
1995, p p .1053-1056
[55] R. R. Goulette, S. K. Xavier, and R. L. G reenfield, “M ethod and apparatus for
m onitoring electrom agnetic emission level” , U s Patent No. 4,829,238, M ay 9,
1989
[56] R. R. G oulette, S. K. Xavier, and R. L. G reenfield, “M ethod and apparatus for
monitoring electrom agnetic em ission level” , Us Patent No. 5,006,788, A pril 9,
1991
[57] J. Soiferman, “Contactless test m ethod for testing printed circuit boards” , US
Patent No. 5,218,294, June 8, 1993
138
R ep ro d u ced with p erm ission o f th e copyright ow ner. Further reproduction prohibited w ithout perm ission.
[58] R. Zhai, G. Bridges, “Experim ent R eport - investigation on detection capabilities
o f fults in M ulti layers B oards” , the Dept, of Electrical and Com puter
Engineering, the University of M anitoba, November, 1998
[59] C irlog Inc. “M LB testing Capabilities” , M ay, 1997
[60] H. C. Han, “A report on the evaluation of ATTI contactless PCB test system” 349
D urham Computation Center, Iow a State University, Sept. 7 1993
[61] M . N. O. Sadiku, “Elem ents o f electrom agnetics”, Oxford, August, 2000, ISBN:
019513477X
[62] C. Shafai, G. Bridges, com m unications by em ails on M EM S cantilever fabrication
at A lberta M icroelectronics Centre, Aug. 1996
139
R ep ro d u ced with p erm ission o f th e copyright owner. Further reproduction prohibited without perm ission.
Документ
Категория
Без категории
Просмотров
0
Размер файла
11 673 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа