370

RID-AI774 CONFORMAL MICROSTRIP SLOT ANTENNA AND ANTENNA ARRAYCU) 1/1

MOORE SCHOOL OF ELECTRICAL ENGINEERING PHILADELPHIA PA
IVALLEY M KISLIUK ET AL SEP 86 RADC-TR-86-Bl14
USI FE F1688--0iFG95 N
 11U. L228 25
L36

4CROCOPY RESOLU1101 TESt HARI

0ANAl I A , qbN
AN
 RADC-TR-86-1 14
Interim Report
Septemberi9e6

CONFORMAL MICROSTRIP SLOT ANTENNA

0 AND ANTENNA ARRAY

University of Pennsylvania

Moshe Kisliluk, Bernard Steinberg and William Whistler

APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED DTIC

NV2 51986

W-- ROME AIR DEVELOPMENT CENTER

Air Force Systems Command
Griffiss Air Force Base, NY 13441-5700

S.' ]1
 This report has been reviewed by the RADC Public Affairs Office (PA) and
is releasable to the National Technical Information Service (NTIS). At NTIS it
will be releasable to the general public, including foreign nations.

RADC-TR-86-114 has been reviewed and is approved for publication.

APPROVED: -L&

ZACHARY WHITE
Project Engineer

APPROVED: /'4
6 IL LIz-t C-C -----.
ALLAN C. SCHELL
Chief, Electromagnetic Sciences Directorate

FOR THE COIOIANDER:544

JOHN A. RITZ
Plans and Programs Directorate

If your address has changed or if you wish to be removed from the RADC mailing
list, or if the addressee is no longer employed by your organization, please
notify RADC (EEAA) Hanscom AFB MA 01731-5000. This will assist us in
maintaining a current mailing list.

Do not return copies of this report unless contractual obligations or notices

on a specific document requires that it be returned.
 SIECURITY CLASSIFICArION OF THIS PAGE & "1ViA -il

REPORT DOCUMENTATION PAGE

la. REPORT SECURITY CLASSIFiCATiON lb. RESTRICTIVE MARKINGS
UNCLASSIFIED N/A
2a SECURITY CLASSIFICATION AUTHORITY 3 DISTRIBUTION/ AVAILABILITY OF REPORT
N/A Approved for public release;
2b DECLASSIFICA fION /DOWNGRADING SCHEDULE distribution unlimited
N/A
4. PERFORMING ORGANIZATION REPORT NUMBER(S) 5 MONITORING ORGANIZATION REPORT NUMBER(S)
N/A RADC-TR-86-114

6. NAME OF PERFORMING ORGANIZATION 6b OFFICE SYMBOL 7a NAME OF MONITORING ORGANIZATION

(If applicable)
-University of Pennsylvania Rome Air Development Center (EEAA)
6c. ADDRESS (Ciy, State, and ZIP Code) 7b ADDRESS (City, State, and ZIP Code)
Valley Forge Research Center
Moore School of Electrical Engineering
Philadelphia PA 19104
8a, NAME OF FUNDING/SPONSORING 8b OFFICE SYMBOL 9 PROCUREMENT INSTRUMENT IDENTIFICATION NUMBER
ORGANIZATION (if applicable)

Rome Air Development Center EEAA F19628-84-K-0021

Bc. ADDRESS(City, State, and ZIP Code) 10 SOURCE OF FUNDING NUMBERS
Hanscom AFB MA 01731-5000 PROGRAM
ELEMENT NO PROJECT
NO TASK
NO WORK UNIT NO
A.CCESSION

I 61102F 2305 J3 47
11 TITLE (Include Security Classification)

CONFORMAL MICROSTRIP SLOT ANTENNA AND ANTENNA ARRAY

12 PERSONAL AUTHOR(S)
Moshe Kisliuk, Bernard Steinberg, William Whistler
13a. TYPE OF REPORT 13b TIME COVERED 14 DATE OF REPORT (Year, Month, Day) S PAGE COUNT
Interim I FROM Mar 85 TO Feb 86 September 1986 52
16 SUPPLEMENTARY NOTATION
N/A

17. COSATI CODES 18 SUBJECT TERMS (Continue on reverse if necessary and identify by block number)
FIELD GROUP SUB-GROUP Airborne
Airborne Arrays
09 03 Conformal Antennae
Microslot Antennae AWACS
Radio Camera Sensors
09 Conformal Arrays Beamforming
9. ABSTRACT (Continue on reverse if necessary and identify by block number)

This study explores the wideband potential of a microstrip-fed slot radiator as an element of
a planar or conformal antenna array. The microstrip slot is seen to have 40% input impedance
bandwidth under 1.5 VSWR and a demonstrated 10% gain bandwidth to 3db.
*. The report describes an analytic slot model and experimental tests. A conclusion of this
*" report is that useful high efficiency radiators (greater than 80%) can be made by using the
microstrip slot in N-element serial arrays.

20. DISTRIBUTION /AVAILABILITv OF ABSTRACT 21 ABSTRACT SECURITY CLASSIFICATION

OUNCLASSIFIED/UNLIMITED I2i SAME AS RPT [ODTIC USERS UNCTATFTEID
'22a. NAME OF IJESPONSIBLE INDIVIDUAL 22b TELEPHONE (Include Area Code) 22c OFFICE SYMBOL
" Zadhar, White (617) 861-2055 RADC (EEAA)
DD FORM 1473 84 MAR 83 APR edition may be used until exhausted SECURITY CLASSIFICATION OF THIS PAGE
All other editions are obsolete U C A S F D

".~
%V~~ ~ ~ ~~~UCAS
. _' * FT .---
ED.1 %
 TABLE OF CONTENTS

Preface
Summary
Related Publications
and Presentations .................... 1
1. High Efficiency Microslot Study...2.
1.1. Determination of
Matching Technique .............. 2
1.2. Effect of Feeding System
on Antenna Patterns ............. 4
1.3. Effects of Slot Area
on Antenna Gain................6
1.4. Comparison of Shunt Microstrip
Slot and Microstrip Patch
Radiators ....................... 9
2. Series Slot Investigation ........ 10
2.1. Match Improvement Techniques...11
2.2. Dielectric Layer Truncation
Effects ........................ 13
2.3. Series Slot Pattern ............ 16
2.4. Gain and Efficiency ............ 21
3. Radiation of Microstrip
Meander-Line Slot Antenna ........ 22
3.1. Fields in a Spherical
Coordinate System .............. 23
3.2. Radiation from the Slot ........ 26
3.3. Radiation from Spillover
Currents ....................... 29
3.4. Radiation from the Hybrid
Surface Wave at the
Truncations of the
Dielectric ..................... 32
3.5. Tapered Meander Line Slot
Array Investigations ........... 35
4. Conclusions ...................... 38
References ..................... ....40
Accession For
XTIS GRA&I
DTIC TAB
Unannouncod Q
i+ Justific'.tlo3

Distribution/
Availability Codes

_'
"'I/
I'vail Ind/ar
'Dist SPe'cial
 RELATED PUBLICATIONS AND PRESENTATIONS

1. M. Kisliuk, "The Conformal Microstrip Slot Antenna,"

1983 IEEE Int. Symp. Antennas and Prop., Vol. 1,
Houston, TX, pp. 166-169.

2. M. Kisliuk, "The Voltage Distribution on Waveguide Fed

Slot and Slit Antennas," IEEE Benjamin Franklin
Symposium on Advances in Antennas and Microwave
Technology Digest, Philadelphia, PA, 1983, pp. 13-15.

3. M. Kisliuk, "Lossy Transmission Line Model of

Microstrip Slot Antenna," VFRC Quarterly Progress Report
No. 43, April-September 1983, pp. 76-81.
4. M. Kisliuk, "Microstrip-fed Radiating Slots and Slot
Arrays," VFRC Quarterly Progress Report No. 45, April-
September 1984, pp. 46-48.

5. W. Whistler, "Microstrip Slot Antenna-Experimental

Investigation, VFRC Quarterly Progress Report No. 45,
April-September 1984.

6. M. Kisliuk, B.D. Steinberg and W. Whistler, "Conformal

Microstrip Slot Antenna and Antenna Array," First
Interim R & D Report, December 1985.

7. M. Kisliuk, A. Axelrod and Y. Servant, "Experimental

Results on an H-Guide Slot Antenna with an 80% (and
higher) Efficiency in a 21% Frequency Band," submitted
in December 1985 for publication to Microwaves and RF.

-V-

J
 LIST OF FIGURES

1.1 3-D Sketch of Micrcslot Shunt Radiator ............... 3

1.2 Microslot Shunt Radiator: Pattern with Coaxial
Edge-Board Feed ...................................... 5
1.3 Microslot Shunt Radiator: Pattern with Coaxial
Rear-Feed through Ground Plane ....................... 5
1.4 Photo of vee Slot, .5 mm and .2 mm Microslot
Radiators used in Antenna Gain Studies ............... 6
1.5 Pattern of .2 x 3.2 mm Microslot Antenna
Measured at 9.154 GHz ................................ 7
1.6 Pattern of .5 x 10.9 mm Microslot Antenna
Measured at 8.535 GHz ................................ 8
1.7 Pattern of a .5 mm Vee Slot Antenna shown in
Figure 4 at 9.607 GHz .............................. 9
2.1 Front and Rear View of .2 x 12.6 mm Serial
Microstrip Slot Test Board .......................... 12
2.2 Comparison of Single X-Band Slot VSWR Measure-
ments at Tel-Aviv University and at VFRC ............ 12
2.3 Antenna Pattern of .5 x 12.6 mm Microslot on 15.3 cm
Dielectric Substrate ................................ i3
2.4 Antenna Pattern: Comparison of Patterns
of .5 x 12.6 mm Slot Radiator with (Solid Line)
and without 38 cm x 20 cm Ground Plane .............. 15
2.5 VSWR pf .5 x 12.6 mm Slot Radiator used in
Pattern Tests ....................................... 17
2.6 Antenna Pattern: .5 x 12.6 mm Slot at 8.0 GHz ....... 18
2.7 Antenna Pattern: .5 x 12.6 mm Slot at 9.0 GHz ....... 13
2.8 Antenna Pattern: .5 x 12.6 mm Slot at 10.5 GHz ...... 19
2.9 Antenna Pattern: .5 mm Vee Slot Antenna at 9.0 GHz..19
2.10 Gain and Beam Tilt vs. Frequency Graphs
for .5 x 12.6 mm and .2 x 14 mm Microstrip Slots
and .5 mn Vee Slot Radiators ....................... 20
3.1 The Microstrip Meander-Line Slot Antenna ............ 23
3.2 Spherical Coordinates in the Far-field Zone ......... 24
3.3 Simplified Equivalent Circuit of a Microstrip
Meander-Line Slot Antenna ........................... 28
3.4 13-Slot Array. The Width of Slots is Tapered ........ 36
3.5 E-Plane Radiation Patterns of the Slot Array
of Figure 3.4 ....................................... 37
3.6 10-Slot Array. The Width of the Microstrip Lines
Feeding the Slot is Tapered ..........................
3.7 E-Plane Radiation Patterns of the Slot Array
of Figure 3.6..................................... 38

-vii-
PREFACE

The Valley Forge Research Center initiated this research to formulate

theory and to develop microstrip antennae having radiating nonresonant

slots in the upper conductor. Preliminary calculations indicated that such

antennae can operate in a wider frequency band than the conventional patch

microstrip antennae. Our original work to obtain high angular resolution

microwave imaging led to the design of the Radio Camera. Implementation of

the Radio Camera led to techniques for beamforming an array which is com-

posed of randomly spaced elements, each element moving in a random fashicn.

The results of our experiments led us to believe that we could design a

conformal microstrip slot antenna. The overall goal of this research ef-

fort is to develop an antenna element that may be flush-mounted in the

outer skin of the aircraft and placed in an arbitrary location. These ele-

ments would be connected to a solid state RF located on the interior

surface of the airframe by an electrical feedthrough.

SUMMARY

In the First Interim Report, a series of analytic developments were

completed concerning the input impedance, pattern efficiency and losses of

the microstrip slot antenna. Experimental verification was attempted at

both Tel-Aviv University and VFRC with mixed results. One of the main im-

plications for further research was the need to investigate improvements in

antenna pattern measurement techniques. A section of this report is

devoted to this topic.

The design of a high efficiency microstrip slot shunt radiator com-

parablp to the microstrip patch antenna is a second area of investigation

reported here. Capacity coupling to the microstrip slot as a rescnator

produced antenna gain and efficiencies that approaches, but did not exceed,

the directivity of the patch radiator.

-1-
 Other conclusions of the shunt radiator study were that increasing

slot radiator area improved antenna gain and efficiency. In addition, the

effects of metal plate edges and dielectric truncations were more easily

studied on the higher efficiency shunt microstrip slot radiator.

Analytic investigations of microstrip slot characteristics continued

in this report period. Emphasis was placed on hybrid surface waves, a non-

radiating mode and radiation from a truncated dielectric edge.

1. HIGH EFFICIENCY MICROSLOT STUDY

It is evident that a high efficiency slot would be more universally

useful than the low efficiency slot, since the latter can only achieve use-

ful efficiencies (> 80%) by the formation of a series slot array. An

antenna element with broad beam pattern in two orthogonal planes is more

desirable for wide angle two dimensional arrays, so a study of methods of

terminating a line with a microstrip slot was initiated.

A method of matching by capacity coupling has been found and is

described. An antenna study has been conducted which shows the microstrip

slot radiator exhibiting an increase in gain as the slot area is increased.

Currents in the metal surrounding the slot excite nonradiating hybrid modes

in the dielectric, which radiate at the edges of the dielectric. When com-

pared to the microstrip patch radiator, the microslot radiator's pattern is

broader and gain is lower. Its bandwidth in this mode appears lower than

the patch radiator's for the same dielectric thickness.

1.1. DETERMINATION OF MATCHING TECHNIQUE

To begin the study, experiments were made to find a suitable matching

network for a radiator terminating a 50 line microstrip line. A matching

stub following the micrcslct was investigated and rejected, after trials of

a series of stub lengths of various impedances produced no indication of

-2-
 matching potential. Attempts at matching by varying the input line im-

pedance to present different impedance levels to the microslot also

produced only high input reflection conditions.

A successful matching system was finally obtained by capacity coupling

to the transverse microstrip line containing the slot, treating this last

as a resonant circuit. This system could always be made to match by ad-

justing capacity by varying coupling slot width. The resulting

configuration is shown in Figure 1.1.

F d

FED
" POINT

COPPER
DIELECTRIC
COPPER
METAL
CENTRAL CONDUCTOR
INSULATION
CONNECTOR BODY
CROSS-SECTION THROUGH FEED POINT

FIGURE 1.1 . 3-D SKETCH OF MICROSLOT SHUNT RADIATOR

-3-

Z z ~z
 A useful measurement parameter is the cavity Q of this resonant

circuit. For any shunt resonant circuit, cavity resonance is indicated by

zero reactance, or a normalized impedance of 1 + jO. The reciprocal of the

frequency difference between the + 45 degree phase conditions or between

normalized impedances of 1 + Ji is defined as the cavity Q. VSWR at the

level 1 + j1 impedance levels is 2.6:1 : this level is similar to the 2:1

and 3:1 bandwidths common in discussing microstrip patches arid other

radiators, and is convenient for comparing this radiator with others.

Resonant frequency of the shunt microslot radiator is directly related

dtc6he total path length of the 100 ohm metal lint- surrounding the

microstrip slot using the effective dielectric constant calculated for the

100 ohm line. Typical cavity Q's for the radiators measured here varied be-

tween .6% and .76% for a .031" (.787 mm) thick substrate. Resonant

frequencies varied from 8 GHz to 9.6 GHz for various configurations; for a

,' representative frequency of 9 GHz, this thickness is .024 A.

1.2 EFFECT OF FEEDING SYSTEM ON ANTENNA PATTERNS

The First Interim Report noted in its conclusions the need for improv-

Ing the feed system to reduce feed effects on pattern. The pattern in

Figure 1.2 is typical of the edge feeding system. The beam peak of the

shunt-fed microslot radiator shows a beam tilt of 48 degrees, indicating

Jthe influence of direct radiation from the coaxial feed at board edge.

%The pattern improvement obtained by using the rear feed with entry

point through the ground plane is clearly shown in Figure 3. This pattern

shows a more uniform pattern with the beam tilt reduced to 15 degrees. The

beam tilt remaining is believed to be caused by radiation from the coaxial

feedline or the .33 Ao long 50 ohm microstrip feedline.

-14-
 ' PATTERN No.15 DATE: 2/20/86
/ ] HIGH EFFICIENCY CONFORMAL ELEMENT
1 CAPACITY-COUPLED SLOT, 0.5x 10.9mm
-- 2-SCREW CONNECTOR IN 1/8" PLATE
-I - 0.085" DIELECTRIC DIAMETER

7
-7.
-
7
. ..
---
I
- --*--'-'
"- ' '-

- -- - - T -+ -,
.. .. .. -- -. - - . - -- . .1_ . . 4

A- -1-

90 -'- I
60 30 I 0, . -30
.. -60 -90 .

FIGURE 1.2 MICROSLOT SHUNT RADIATOR: PATTERN WITH COAXIAL

EDGE-BOARD FEED,

-PATTERN
J- - - ,7 ___ - __4. d _Jki L No- DATE 2/21/86
... _ 1_-_

_T-4 REAR FEED

• ,.,.
i. .,,.
.. .. _ i ." .
90
_. t,• .' . ', .. ,., ...
60
'.-.
30
'' i''."r.. '
0 -Ii -30
, " . ".. ,.-.
-.
-60
. ".. .,,"- ..
+-zv
-90
.. "-•.",. . . "' . ".". " . .. ,

FIGURE 1.3. MICROSLOT SHUNT RADIATOR: PATTERN WITH COAXIAL REAR-FEED

THROUGH GROUND PLANE.

-5-
 The rear feed approach han-B been adopted in all succeeding testing be-

cause of its clear reduction of interference from direct radiation with the

microslot radiator patterns.

1.3 EFFECTS OF SLOT AREA ON ANTENNA GAIN

The microslot as a shunt radiator acts like a slot antenna in a ground

plane with respect to antenna gain, having a low radiation efficiency be-

cause of low physical area. To investigate the effect of radiator area on

gain, the three substrates shown in Figure 1.4 were fabricated.

'4 I

SMETm 2 3 4 15

FIGURE 1.4. PHOTO OF VEE SLOT, .5 MM AND .2 M MICROSLOT RADIATORS

USED IN ANTENNA GAIN STUDIES.

The .2 mm microslot's antenna pattern is shown in Figure 1.5. The ap-

pearance of shoulders on the main lobe pattern is believed due to radiation

-6-
 from the dielectric truncations at + 1.78 cm from the center of the micro-

slot(the substrate edges). They are most prominent in this low-gain

microslot because they are similar in magnitude to the direct microslot

radiation. The extraneous radiation at + 90 degrees and at 180 degree

angles cannot be caused by the microslot because its pattern factor ap-

proaches zero for end fire direction. End fire radiation thus must be

caused by radiation of hybrid mode energy from the truncated dielectric

edges of the substrate.

210 180 150 120 90 60 30 0 -30 -60

FIGURE 1.5. PATTER4 OF .2 X 13.2 MM MiCROSLOT ANTENNA MEASURED AT

9.154 GHz. MEASURED GAIN IS - 1.5 dB BELOW THAT OF A
DIPOLE RADIATOR (-1.5 DBD).

7
 The .5 mm shunt microslot antenna pattern of Figure 1.6 shows con-

siderable reduction of shoulders from dielectric edge radiation, probably

because of the 2.3 dB increase in gain for this radiator.

x
... - j- - -- | -. - ...- -- "' : ....

--- I
St
--
l-I ...... - I....T
_ _-

180 150 120 90 60 30 0 -30 -60

FIGURE 1.6. PATTERN OF .5 X 10.9 MM MICROSLOT ANTENNA MEASURED AT

-55GHz. MEASURED ANTENNA GAIN IS + .8 d-d.
The gain increase with the wider slot prompted the design of the "'vee-

slot" radiator which was hand fabricated because of time pressures. The

resulting pattern shown in Figure 1.7 exhibits a 4.2 dB gain increase over

the .5 mm microslot pattern. This is a significant gain imprcvenent and is

the shunt microstrip slot radiator form most closely competitive with the

microstrip patch radiator.

8
 II I I

- :& 4.-_1

180 150 120 90 60 30 0 -30 -60 -90

FIGURE 1.7. PATTERN OF .5 MM VEE SLOT ANTENNA SHOWN IN FIGURE ~4AT

9.607 GHz. MEASURED ANTENNA GAIN IS 5 dB.

1 .4 COMPARISON OF SHUNT MICROSTRIP SLOT AND MICROSTRIP PATCH RADIATORS

A comparison of the microstrip vee-slot and patch radiators is given

in Table 1.

TABLE 1. COMPARISON OF PATCH AND VEE SLOT RADIATORS

Vee-Slot Microslot Microstrip Patch*

Gain +5 dBd measured +6.7 dBd Calculated

Bandwidth .64% measured 2.4% Calculated

2:IVSWR for t/A.o .035 for t/A o - .035

*Source: Reference [113

9

..... , . -- i iill~~l,~ill~., ~ii.ilIlL clIi~l

;l . Ji< il 'ivi"
'; " '- i .",", i1:
..
i2r>.; .$ k2,l
,':.i .I::4K2
',
 Most significant is that the vee-slot antenna measured gain approaches

that of the calculated microstrip patch gain.

Bandwidth for VSWR = 2:1 indicates superior performance for the patch

antenna. These are directions indicated for improving vee slot bandwidth,

the principal one being to use a lower impedance line to surround the

microstrip slot. However, it appears impossible for the vee slot bandwidth

to approach the patch bandwidth.

The capacity coupling method used in feeding the vee slot is difficult

to control tolerances on if edge coupling of microstrip lines is the

capacity mechanism. Any other capacity mechanism, such as soldered-on chip

capacitors, nullifies the microstrip advantage of being a low cost fabrica-

tion process.

To summarize, the vee slot antenna appears to be the microstrip slot

implementation that is the closest competitor for the microstrip patch

antenna. Its lower bandwidth and tolerance-sensitive coupling mechanism

make it inferior to the patch for similar applications in its present form.

2. SERIES SLOT INVESTIGATION

The series microstrip slot as originally conceived by Prof. Kisliuk

continues to be the main thrust of this contract's effort. Efforts were

made in this report period to improve pattern measurement techniques and to

broaden our understanding of the microstrip slot antenna.

A pattern study of a series slot radiator is detailed, with patterns

taken in a test fixture that minimizes the pattern interference of the

dielectric truncation radiation.

10
2.1 MATCH IMPROVEMENT TECHNIQUES

An impedance study of the rear-mounted connector was made using two

rear-mounted connectors moanted on a 13.72 cm (5.4") 50 ohm line. The con-

nector matching time cycle was shortened by the discovery that a tapered

resistive card laid on top of the microstrip offered more than 10 dB inser-

tion loss and thus isolated reflections from the distant connector from

those of the input port connector by attenuation. This system allowed

studying connector impedance by modifying only the input connector

dimensions.

A connector having a .085" teflon diameter was chosen because the

.093" wide 50 ohm line will block direct radiation from the feedline. The

criginal .010" diameter center conductor introduced a fixed series

inductance: this was minimized by increasing the center conductor diameter

to .027".

The final connector configuration is shown in Figure 2.1. The result-

ing single connector VSWR was under 1.1 VSWR from 8.0 to 11.0 GHz. This

design is duplicated in all test fixtures used in this report period.

11
 v[T'ir 12 111 13 III1 14

FIGURE 2.1 FRONT AND REAR VIEW OF .2 X 12.6 MI-M

SERIAL MICROSTRIP SLOT
TEST BOARD

Li -1~~..~T~ Aviv da 8/84.

-J -17X
-4 14 -+, _

VP;P

p I
44T 177
J. .5 9.5 9.05x.
IFIREQiENCY IN H

FIGURE 2.2. COMPARISON OF SINGLE X-I3AND SLOT V W MEA S'JR MLNTS AT TEL-
AVIV UNIVERSITY AND AT 'JFRC. DIWLIONl0PS (in mm) d = 12.1,
w=. 145, w1 = .90, S5 .1, h = 79, 2

-12-
 Mitering slot ends and microstrip line ends at each connector made

small reductions in VSWR and these changes were incorporated. The result-

ing impedance performance of a .2 mm x 12.1 mm slot on a 5.4" run is shown

in Figure 2.2; the VFRC test data compares favorably with the Tel-Aviv

University data also illustrated.

2.2 DIELECTRIC LAYER TRUNCATION EFFECTS

Dielectric-coated conductors, as noted in [7] and discussed in Section

3, can support a non-radiating hybrid mode which radiates at a dielectric

truncation. Figure 2.3 shows an example of this effect: the radiation pat-

tern of a single .2 mm slot in the center of a 15.25 cm board having both

metal ground plane and dielectric sheet ending at + 7.63 cm from the slot.

The multilobed pattern results from the reradiation at the board ends which

are 4.32 wavelengths apart at 8.5 GHz, the test frequency.

-- T--PATTERN No.1 DATE: 2/24/86

__ CONFORMAL ELEMENT
I0.2 x 10.92 mm SLOT
E PLANE f = 8500 MHz
6" BOARD, REAR FEED
'7 -
-- -- ---

------

-60 S-30-90

.4i
FIGURE 2.3 ANTENNA PATTERN OF .5 X 12.6 MM MICROSLOT ON 15.3 CM
DIELECTRIC SUBSTRATE. cr - 2.2, CONNECTORS BEHIND GROUND

PLANE.
90.. 60 30 0 -3 -6V9
-13-
 In Figure 2.3, there are two predominant cyclic pattern systems, one

having a 22 degree period and the other a 14 degree period. The separation

of pattern nulls should be related to sin- I/L, where L is the separation

of radiators in the two-radiator system.

It is interesting to try the spillover radiation hypothesis of

Section 3 on this system, considering L as 2.16 X, half the board length,

and assuming the 50 ohm line velocity of 1.37 times free space velocity.

Then

AO - 1.37 (2.16A) sin 22 deg. = 399 deg.

Since AO is by definition 360 deg., a velocity error of (1 - 399/360)
- 10.9% is indicated by the computation. To force the 360 deg. result, a
reduction of /e from 1.37 to 1.236 is required. Thus, a spillover current-
e

generated hybrid mode with a velocity 10.9% lower than the 50 ohm

microstrip velocity is indicated.

Testing then continued on the 35.6 x 50.8 mm board with the .5 mm x

12.6 mm shunt slot sketched in Figure 1.1. The slot radiation pattern shown

dotted in Figure 2.4 has a theoretical slot's beamwidth of 55 deg., but

shows deviations at the -15 dB level, indicating the possible presence of

radiation from dielectric truncations.

-14-
 - PATTERN No. 39

-4r ~I- CONFORMAL ELEMEINT

VA- CAPACITY COUPLED HI-P SHUNT M.ICROSLOT

-SLOT CENTERED IN 36cm G.P.

_.___. DIELECTRIC TRUNCATED AT - 1.78 cr.

L -t--.:--,J i -- -, +'

120 90 60 30 0 -30 -60 -90 -120 -150

FIGURE 2.4. ANTENNA PATTERN: COMPARISON OF PATTERNS OF .5 X 12.6 MM

SLOT RADIATOR WITH (SOLID LINE) AND WITHOUT 38 CM X 20 CM
GROUND PLANE

The solid trace is the same radiator set in a 38 cm long by 20 m wide

ground plane: the metal edges are now spaced 38 cm while the dielectric

sheet width remains at 3.56 cm. The metal sheet edge effects are clearly

indicated by a fine ripple relating to the 10.8 wavelength separation of

the metal edges. The overall pattern distortion remains present in the

radiation pattern, however, indicating the probability of dielectric

truncation radiation effects.

Two possible sources of excitation of hybrid modes are apparent:

-the spillover currents carried on the top of the 50 ohm feedlines as

discussed in Section 3.

-radiation from the microstrip open circuit at the junction of the

microstrip-to-coaxial line adapter.

-15-
 As described above, it is possible that spillover currents on the 6.85

cm feedlines are significant radiation source in the 15.25 cm board pattern

shown in Figure 2.2. On the short runs of the 3.56 cm board patterns shown

in Figure 2.3, however, it is possible that the microstrip open circuit of

the microstrip-coax adapter is the major contribution to hybrid mode

excitation/dielectric truncation radiation. This possibility is also dis-

cussed in connection with the serial slot patterns of the next section.

2.3 SERIES SLOT PATTERN

One goal of this study has been to determine the microslot antenna's

useful bandwidth as a radiator. The final unit tested shown in Figure 2.1

was freeest from extraneous radiation effects and was measured over the 8.0

tc 11.0 freuency band to permit that measurement. Its VSWR is given in

Figure 2.5.

The vee-slot antenna was also tested because it had achieved higher

gain than the slot in the shunt radiator version. It had lower measured

gain than the .5 mm slot radiator, which confirmed the validity of Dr.

Kisliuk's analysis of the microstrip slot as a coupled line system, a fea-

ture not present in the vee-slot design. Its pattern at 9 GHz is shown in

Figure 2.9.
Antenna patterns were taken from 8 to 11 GHz; typical patterns for the
.5 x 12.6 mm slot are shown in Figs. 2.6 - 2.8 and summarized in the graphs

cf Figure 2.10. At 8 GHz, the pattern beamwidth is slightly greater than

theoretical (58 deg. vs. 56 deg. theoretical), has good pattern symmetry to

a -15 dB level, and shows little beam tilt. The 9 GHz pattern shows a
shoulder developing near + 90 deg., a lobe developing at -90 deg. and a

-16-
 change of sign of the beam tilt. In the 10.5 GHz pattern, the energy at -

90 deg. has reached -7.5 dB with a definite null appearing at +23 deg.

right.

The pattern beam tilt appears related to the substrate board width:at

the zero beam tilt frequency of 8.4 GHz seen in Figure 2.10b, the substrate

is exactly 1.0 wavelengths wide. The pattern null, however, appears re-

lated to radiation from the microstrip open circuit at the input

microstrip/coax adapter. Consider that the substrate edge and the micro-

slotact as a two radiator set: at 10.5 GHz, computation of path lengths

shows that these two sources are 180 deg. out of phase when viewed at -23

deg., the angle of the pattern null. The fact that this effect occurs only

on the input connector side of the pattern indicates that only input con-

nector radiation is involved in distorting the pattern.

1.6

1.4
0.5mm x 12.6mm SLOT

1.2

I.O
8.0 8.5 9.0 9.5 10.0 10.5 11.0
FREQUENCY (GHz)

FIGURE 2.5. VSWR OF .5 X 12.6 IV SLOT RADIATOR USED 1N PATTERN TESTS.

-17-

Is ' ' " " . '- - " 'e , . , '' ' ' ., - ' " '-' " " . " - '" " , . . '' ' " ' ' -
 PATTERN N .t,

II0.5 x 12.6 mm SLOT

1. .

I i II:

120 90 60 30 0 -30 -60 -90

FIGURE 2.6. ANTENNA PATTERN: .5 x 12.6 mm SLOT AT 8.0 GHz

I PATTERN No. 62
0.2 x 12.6 mr, SLOT
I-f - 9.06 GHz
'I _

17
/_

I- I'

2.4.. I Ii /

150 120 90 60 30 0 -30 -60 -90 -120

FIGURE 2.7. ANTENNA PATTERN: 5 X 12.6 MM SLOT AT 9.0 GHz

-18-
 I
.... PATTERN No.65

--- - 0.5 x 12. 6mm SLOT

4I~IV 'If =10.5 GHz

J J - V-

150 120 90 60 30 0 -30 -60 -90 -120

FIGURE 2.8. ANTENNA PATTERN: .5 X 12.6 MM SLOT AT 10.5 GHz.

1
.. PATTRN No. .7
""VE .i "ESL OT f:9.0o H

-- -- r
I - -- .... .... -..... -- --- . ...
-

-I I

120 90 60 30 0 -30 -60 -90 -120 -150

FIGURE 2.9. ANTENNA PATTERN; .5 MM VEE SLOT ANTENNA AT 9.0 GHz.

-19-
 .9

.63

O.0 - .-- x :5051O1 -.

32
S 12.6mam SLUT
x.5mm --

20 ~
-59

z -VE SLOT
13 '

-10.

I5

o0 9.0 10.0 11.0

FREQUEN(CY ((;Hz)

-- - - -

x 0.2mm x 14mm SLOT

* --" 0.5mm x 12.6mm SLOT

*-'- -/

' 0 ,0 10.0 11.0

r'i FRlF'() II'NCY ((',1tz )

Fi V 2.10. GAIN AND BEAM TILT VS FREQUENCY GRAPHS FOR .5 X 12.6 MM AND
.2 X 14 ,Df MICROSTRIP SIOT AND .5 MM VEE SLOT RADIATORS.
-20-

"." .%*. -- . .- ~.
. . . . - . . . . - - .. - .... -. .• - . ... " ."-. " . ." ' ,. , . "-'. .,-,- '.
2.4 GAIN AND EFFICIENCY

The antenna gains measured in this study are referred to the gain of a

half-wave dipole for testing csnvenience. To relate to the average power,

radiated a fictitious isotropic radiator is defined which has uniform

radiation over a spherical shell surrounding the antenna. The dipole gain

relative to an isotrope is given in Reference 13 and other places as 1.64

or + 2.15 dB. One should also note that all measured gains include
2
dielectric and I R losses in the antenna. Directivity also relates to main
2
beam peak power relative to an isotropic source, but excludes any I R or

dielectric losses. Directivity, therefore, can be defined analytically

without involving the specific antenna material or other considerations.

Efficiency must be defined relative to the theoretical gain or direc-

tivity of the antenna type chosen, e.g., the slot radiating into a half

space defined by the infinite ground plane containing the slot. The
theoretical gain into a half space is + 3 dB above a dipole or + 5.15 dB

above an isotrope. An efficiency scale calculated on this basis is shown

in the right-hand ordinate of Figure 2.10's gain slot.

The patterns in the 8.0 - 9.0 GHz region are relatively undistorted by

the reradiatlon from the dielectric truncation as evidenced both by pattern

symmetry and absence of endfire lobe at -90 deg., and the efficiency shown

in this region (20-25%) relates closely to the values measured and reported

in the First Interim Report.

The values of gain above 9 GHz are suspect because of distortion from

truncation radiation. This increased radiation, it is noted, correlates

-21-

.. *~,~
*N
9
with the increased input reflection coefficient: it is expected that im-

proved connector match would eliminate the end-fire radiation and improve

the gain correlation with the lower frequency readings.

3. RADIATION OF MICROSTRIP MEANDER-LINE SLOT ANTENNA

The microstrip meander line (Franklin) antenna was introduced in 1953

by Fubini [1]. Later investigations of the Franklin microstrip antenna

(and array) were reported by Solbach [2] and Kisliuk [3]. Explicit expres-

sions for input impedance and radiation efficiency of a single microstrip

meander-line slot antenna were derived by Kisliuk [4]. Radiation patterns

of single slot antennas and antenna arrays were reported in [2], [4] and

[5]. It was observed in [4] that the electromagnetic radiation of

micrcstrip meander-line slot antennas is generated by the voltage distribu-

tion on the slot (magnetic current), and by the fractional currents flowing

on the upper surface of the straight microstrips [6] feeding the meander

line (see Figure 3.1). Hybrid surface waves excited on both sides of the

feeding microstrips [7] generate radiation (space) waves at the input and

output truncations of the microstrip dielectric board (z = z ) and z - z2

radiation generated by surface waves at the truncation of the dielectric

btoard has been reported in [8].

-22-

=. -~ ' £
 i2

/ GROUND
DIELECTRIC 'PLANE
BOARD h

FIGURE 3.1. THE MICROSTRIP MEANDER-LINE SLOT ANTENNA

This section presents a detailed derivation of the radiation-zone

electric fields generated by the slot, by the spillover currents, and by

the hybrid surface waves at the truncations of the dielectric board on

which the slot antenna has been etched.

3.1 FIELDS IN A SPHERICAL COORDINATE SYSTEM

The computation of the field components in the radiation (far-field)

zone requires the introduction of spherical coordinates with the polar axis

in the x direction, as shown in Figure 3.2; the Cartesian coordirates in

Figure 3.2 are the same as in Figure 3.1. The transition frcm Cartesian

unit vectors to the spherical ones are given by the following formulas:

-23-

=
y, :."
% , "',4 ¢. ." ,r'" . - .. - , "v., ,.. -, . - . .. . . . -
 x = rcosO - Osine, (3.1a)

y = rsinecoso + ecosecos - 4sin4, (3.1b)

z = rsin6sin¢ + gcosOsino + $cos¢, (3.1c)

,'5

FIGURE 3.2. SPHERICAL COORDINATES IN THE FAR-FIELD ZON

In the region above the microstrip where y > 0, the range of the polar

angle e is from 0 to r, and the range of the azimuthal angle * is from -1T/2

to n/2. The vector potential functions generated by electric and magnetic

currents are [9]:

A= 4 e exp(-jkR) dxodyodZ o , (3.2)

Aexp(-jkR) dx dy dz °
V
(3.3)
-24-
where Je and j are the vector densities of the electric and magnetic cur-

rents, respectively,

R = [(x-x )2 + (Yy )2 + (Z-Z )2] (3.4)

and k = o. x, y, z are the coordinates of the observation

(field) point, x0 , YO Zo are the coordinates of the source point.

When the distance r from the observation point to the coordinate

origin (Figure 3.1) is very large compared to the dimensions of the radiat-

ing body, the following standard approximations are used: R - r - x0 cosO -

YOsinOcos - z sinesino, and the integrals (2) and (3) are simplified.

i exp(-jkr) -
A r J eexp(joi)dx dy dz o ,
V

A, 0 ( jexp(-jkr)
) dd
A = 4 r J mexp( dx0dy0dz0 ,
V (3.6)

where 81x° + z° + 3 o
(3.7a)

kcosO, (3.7b)

82 = ksinsinn, (3.7c)

83 = ksin~coso (3.7d)

The angular components of the potential functions (5) and (6) are ob-

tained from (3.1)

A0 =
0-Ax sine + AyzcosOcos4 + A cos~sin4, (3.8a)
A= -Ay sinO + A cos€. (3.8b)
z

-25-
 The angular (transverse) components of the electric field in the far-

field (radiation) zone are obtained from (8) utilizing the basic equation

for the vector potential functions:

E = L V x A - jwA -
0
where 0 is the scalar potential function [9]. In the far-field zone (VO)tr

k 0, and _ _ A A
V x A 3-r jk(OA* OA)

Thus we obtain explicit expressions for the transverse components of the

electric field in the radiation zone:

E jk(- A8 -- A), (3.9a)

00

E - '
Jk(- A
Eo A) (3.9b)

3.2 RADIATION FROM THE SLOT

The voltage across the slot is

V =2V,
slot o (3.10)
where V is the voltage of the odd mode in the meander line enveloping the

slot [4].

It has been shown in [4] that if the width of the meander line is less

than the thickness of the dielectric board (Figure 3.1), the influence of

the right-angle bends at the ends of the slot, as well as the influence of
the T-junctions at the input and output of the slot radiator, can be

neglected. Simplified expressions for such a meander-line radiating slot

V.

_.'e derived in Reference 12. As shown in Reference 12, equation (10),

abovt, yields

sinh[y(d/2-11)]
slot s sinh(y °d/2) (3.11)
where

V =2V. + p
s inc 1 + S/tanh(Y d/2) o (1
(3.12)

is the voltage at the center of the slot, Vin c is the voltage wave inci-

dent upon the input (Figure 3.3) of the meander line, and p is the

reflection coefficient from the input of the meander line radiating slot

(Figure 3.3). Explicit expressions for S and p from Reference 12 are:

cosho in-Z c
v=V S i
e o cosh8 in+z c

2Y ZL+tanhO
2Y ZLtanhO +1

The equivalent magnetic current density of the slot is

- x 2 Vslot 6(y)
In S
(3.13)

where s is the width of the slot, and the factor 2 accounts for the in-

fluence of the conductive plane surrounding the slot. The substitution of

(3.13) into (3.6) yields

A = -= _
-f C
A 0 V D p,(-Ikr)
s r (3.14)

where
sf2[ 0/2
[ s i nh (y ( d / 2-
lx l

D exp(j 2 z)dz 0 exp(j ix)dx.

s sinh(yod/2) (3.15)
-s12 -d/2

-27-
 +Vslot

N I(Z) I(Z) .

m N

zC
zC
ctLOAD

+ +
0 V V LOAD

PV inc PLV t

FIGURE 3.3. SIMPLIFIED EQUIVALENT CIRCUIT OF A MICROSTRIP MEANDER-LINE

SLOT ANTENNA
a. With full meander line
b. With folded meander line

The integrals in (15) are readily obtained:

2Y [cosh(y d/2) - cos(8 1d/2)]

D 0 0s nc s ,
2~'2), (3.16)

where
sine (x) = sin(x3
x (3.17)

The angular ccmponents of the vector potential function (14) are found from
(8):

A o V s Dsin , exp(-Jkr
I0=2' r

(3.18a)
A 0. (3.18b)

-28-
 -~ ~~~~~~~~~~ '~ ~~ lr 1 r Jfrr y M ~ ~~fr ---------

Equations (3.9) and (3.18) yield the angular (transverse) components

of the electric field generated by the slot in the far field zone:

E
slot =
k x(-k
j -2Tr VsDsin6 e r (3.19a)
Eslot =0.

(3.19b)

3.3 RADIATION FROM SPILLOVER CURRENTS

The fractional currents that flow on top of the microstrips [6] are

called spillover currents. The back currents flow on the ground plane in

the region jxj > w/2 (Figure 3.1) and can be replaced by an image strip,

carrying the same spillover current in the opposite direction, and placed

at a distance 2h from the real strip; it is assumed that the space between

the real strip and the image strip (0 < y < 2h) is filled with the

dielectric material of the dielectric board.

If the distribution of the spillover currents across the top of the

microstrip (x-direction, Figure 3.1) is assumed uniform, the current den-

sity vector of the spillover currents is given by

^ I (z)
6 (y) - 6(y+2h)], (3.20)
T z s
e w
where
I (z) = qI(z) ( 1
S W - q (3.21)
is the spillover current, 1(z) is the total current of the microstrip line,

and q is a constant [6].

The current distribution on the straight microstrip lines feeding 'he

meander-line slot radiator is found from the equivalent circuit shown in

Figure 3.3.

-29-
 V.
Inc [exp(-jz) - pexpQjz)] for z < z < 0

Vvt
- lexp(-jz) - pLexp(jz)j for 0 < z < (3.22)
c

where Vinc is the voltage wave incident on the reference plane z - -0, p is

the reflection coefficient at z - -0, Vt is the transmitted voltage wave at

the output of the meander line slot radiator (z - +0), and PL is the

reflection coefficient of the "load" at the output terminals of the slot

radiators (Figure 3.3); Z c is the characteristic impedance of the feeding

microstrip lines.

The substitution of (3.20), (3.21) and (3.22) into (3.5) yields

o Vinc exp(-jkr)
2 z c r (3.23)
C

wr~re

C = jRsin(, 4 b)sinc( 1W/2)exp(-Ji3h),

(3.24)

4 r ' 3'

(3.25)

-3C-
.5., NsL,~!
. '9
 R =R1 + R2 + +
R3 R4'

(3.26a)
l-exp[jz 1 (32 _()
R=
I (2-)
jB '

1-exp[jz (U 2 +)] (3.26b)

R2 _pP J(B2 + ) '

VL exp(jz 2((a -)3-1 (3.26c)

R3 Vinc(l+pL) j (2-5)

VL exp~jz 2 ((32 +B) - i

R4 - PL V. (+P) 2(U2+) (3.26d)
inc L 2

V L is the output voltage of the meander-line slot radiator (see Figure

3.3).

The angular components of the potential function and of the radiation

fields generated by the spillover currents are

A o exp(-jkr) (3.27a)
A0 = qVinc 2TrZ r
c

1=0 CCOs exp(-Jkr) (3.27b)

qVinc 27tZ r
~C

I c
]E~urr = CCOSk exp(-jkr)
=-Jnqkr (3.29)
rr
E cu jV k Ccosgsin ,ep(jr
eic 2Zcr (3.29)

For a standard 500 microstrip line (Zc =500' the impedance ratio in

(3.28) and (3.29) is

n 120= 1.2 (3.30)

21TZ T007

-31-
 . RALIAFlICN FR&M THE HYBRID SURFACE WAVE AT THE TRUNCATIONS OF THE
DIELECTRIC

T,&e microstrip quasi-TEM mode in a miorostrip line generates a hybrid

surface wave [7] that propagates along the microstrip with the phase

velocity of the microstrip mode. It has been shown in [8] that the trunca-

ticns co the dielectric are sources of radiation generated by a surface

wave. The electric and magnetic fields of the hybrid surface wave that are

gential to the truncation planes z = z and z = z 2 (Figure 3.1) are

E = xE x + yEyy (3.31a)
33a

H = xH x + yH . (3.31b)
(33b

, tht- input truncation plane z = z1 the outward unit normal is n I -Z i

'..rce the equivalent current densities are

Jel = (XH y-yH Xl)6(z-z ) , (3.32)

A A

Jml = (-xE X6(z-z1) (333)

tihe cztput truncation plane z = z the outward unit normal is n.. = Z,

., equivalent current densities are

Je2 (-xHy2+YHx2) z-z2 (3.3)

= (Ey 2 -Ex 2 )(z-z 2 )" (3.35)

It h.:s been shown in [7] that the field components of a hybrid surface wave

Sxi f3 for y < h

E (3.36a)
x Sf for y > h
xe 4

SEyif I for y < h

E ="- .36b)
E f for y > h
V ye 2

-32-
 Hi fl/h for y < h
Xi 1 .
11xe f" for y > h

1 yif 3 /r for y < h

Hi 3
Y .Hyef /h for y > h (3.36d)
4

Explicit expressions for the computation of E xi, etc. are given in

Reference 12. The functions f in (3.36) are:

fl sVcos( y)exp(-caV i jaz), (3.37a)

f2 = sVexp(-a v- u - jz), (3-

x y' (3.37c)-

f3 = sVsgn(x)sin(Uy)exp(- xV j~z), (3.37c)

f 4x
= sVsgn(x)exp(- xv-C y u Jz) (3d
(3.37d )"

(3.38)
U - n,

x-w/2 for x > w/2

v (3.39)
-x-W/2 for x < -w/2 ,,

1 for x > w/2

sgn (x) = fr /
-- for x < -w/2

-33-
 .s.otage of tho. micr:ostrip quasi-TEM mode, arid s is the excitation

:,on_:t, wnich fcr a Duroid 5880 (E = 2.2) 50Ql line is 0.01.

substitution of (3.34)-(3.40) into the integrands of (3.5) and

is

K" 'n 4# Q5 (xQI+YQ2 ) exp(-.jkr) (3.M1

A= ~ L c - Q6(xQ3+YQ4
A-" V ^Q exp(-jkr)
c c 4o 63 4 r (3.42)

-(i P4+y P 6)P2' Q2 (llxiP 3+HxeP5)P1'

Q3= (E iP e5PIQ4 = -(EP 4+E xeP 6)P2

'3 yiP3+yeP5) 1' x

Q5= Q7-Q8
' Q6= Q7+Q8'

..t irage c-,poernts of the electric and magnetic currents in the

-. 7_die to the conductive ground plane (which is assumed infinite)

u ritiruations
ri'al of the functions (3.37a) and (3.37c); the

comrponents have an even image, while the "sine" components have an

.ie.e. Thus, the hybrid wav,; radiation is generated ! stly by the EY

U omponents of tr. hyt'id wave.

7-e transverse cc=mprients of the radiation field generated by the

t surface wave due to the truncation of tht. dielectric ire:

- vbr k exp(-jkr) (33

.c 6 (Q4cosecos _Q3 sin) + Q2 Q5 sinq] r )
-3c-
Ehybr .k . i-kr

E =rjVi s - [Q (Qlsine-Q cosecos ) + QQsinn exp(-kr)

inc 4rr S1 2 46 r (3.44)

The total radiation field is the sum of the fields derived above.

I
slot curr+Ehybr
E E (3.45)

_curr Ehybr (3.46)

The radial component of the pointing vector is

It= (rE 12 + IE0 i2 )/(240T), (3.47)

and the total power radiated by the microstrip slot antenna which is
Tr/2 7T
Prad = i dO IdOr 2 sine (3.48)
-,/2 0

can be found by numerical calculations, which will be reported later.

3.5 TAPERED MEANDER LINE SLOT ARRAY INVESTIGATIONS

Preliminary investigations of tapered meander-line slot arrays are

reported in [10].

An array of 13 slots is shown in Figure 3.4; the width of slots is

tapered leading to a taper in the amplitude of the voltage at the center of

the slots. The widths of the slots are:

= 40
Sl = s13 = 0.374 mm s2 s12 = 0.282 mn

= slO = 0.141 mm
s3= S = 0.189 mm s4

s6 = s = 0.104 mm
s 5 = S 9 = 0.116 mm
s = 0.100 im

-35-
 44D 12 13-
--
2.4 00.0.2.4
12.10. --,,- 45
12.1 1

FIGURE 3.4. 13-SLOT ARRAY. THE WIDTH OF THE SLOTS IS TAPERED.

The radiation patterns of that array (Figure 3.4) are shown on Figure

.thesidelobe level is about -15 dB.

-36-
 0o 0
* dB
10.5 GNz 10.6 6HE
-10. -10

-20 -20

-30. -30.I

-40 -40
0 50 10 ISO -40 50 100 15'0
#(dog) #dg

FIGURE 3.5. E-PLANE RADIATION PATTERNS OF THE SLOT ARRAY SHOWN IN FIG. 3.24.

Another array of 10 slots, where the width w of the straight

rnicrostrip lines feeding the slots is tapered, is shown in Figure 3.6. The

df-sign of this array is based on the relationship [2]

R (Q) 45

p. ~ ~"2~ W w01

FIGURE 3.6. 10-SLOT ARRAY. THE WIDTH OF MICROSTRIP LINES

FEEDING THE SLCTS IS TAPERED.
-37-
 The radiation patterns of the 10-slot array in Figure 3.6 art shown in

Figure 3.4; the sidelobe level is lower than -15 dB. The measured radia-

tion efficiency of these arrays is above 75%.

dB do
9.3 tOGH
9.4: I
-10, -1O
1 -----
-20 -20
I

-30 III -30

-40; -40
0 (del) t (dog)

FIGURE 3.7. E-PLANE RADIATION PATTERNS OF THE SLOT ARRAY SHOWN IN FIG. 3.6.

CONCLUSIONS

The experimental and analytical studies described in this report

detail information orn two forms of microstrip slot radiator, the shunt form

terminating a feedline and the serial form.

The shunt micrcslct radiator has a dipole pattern and exceeds 60%

ctfficiency. It has a bandwidth determined by the stub line impedance and

the substrate thickness and material. Its bandwidth in the present form is

.6% and requires improvement before being a useful competitor for the patch

antenna. The present microslot design compares unfavorably with the patch

radiator in bandwidth.

The serial form of microslot radiator when measured with improved test

methods exhibited dipole-like patterns when not distorted by radiation frzrm

-"electrie truncations, and exhibited 20 to 25% efficiency. The pattern

t_ ndwidth exceeding 1 GHz was shown by patterns. Nonradiating hybrid modes

Pan be generated by the microstrip launcher open circuit stub line as well

-38-
as by spillover currents on conductors. These modes transform to radiating

modes at dielectric edge truncations. This radiation's magnitude is

strongly correlated with input impedance match.

-9

-39-
 REFERENCES
[I] E.G. Fubini, "Stripline Radiators," IRE Trans. Microwave Theory
Techn., Vol. MTT-3, pp. 149-156, 1955.
[2] K. Solbach, "Microstrip-Franklin Antenna," IEEE Trans. Antennas
Propagat., Vol. AP-30, pp. 733-735, 1982.
[3] M. Kisliuk, "The Conformal Microstrip Slot Antenna," 1983 Int. Symp.
Antennas Propagat. Dig., Vol 1, pp. 166-169, Houston, Texas.
[4] M. Kisliuk, "Microstrip-fed Radiating Slots and Slot Arrays," VFRC
Quarterly Progress Report No. 45, pp. 46-58, Apr.-Sept. 1984.
[5] W. Whistler, "Microstrip Slot Antenna - Experimental Investigation,"
VFRC Quarterly Progress Report No. 45, pp. 58-65, Apr.-Sept. 1984.
[6] L. Lewin and T. Reuhle, "A Note on the Complex Pointing Vector and on
the Fractional Current on the Upper Surface of a Microstrip Line,"
IEEE Trans. Antennas Propagat., Vol. AP-29, pp. 144-147, 1981.
[7] M. Kisliuk and A. Axelrod, "Hybrid Surface Waves in Microstrip Lines,"
VFRC Quarterly Progress Report No. 47, pp. 69-75. Apr.-June 1985.
[8] S.B.A. Fonseca and A.J. Giarola, "Microstrip Disk Antennas, Part 2:
The Problem of Surface Wave Radiation by Dielectric Truncation," IEEE
Trans. Antennas Propagat., Vol. AP-32, pp. 568-573, 1984.
[9] C.A. Balanis, Antenna Theory: Analysis and Design, Harper & Row, Inc.,
New York, 1982.
[10] G. Elazar, "Microstrip Slot Arrays," M.Sc. Thesis (in Hebrew), Tel-
Aviv University, 1985. Adviser: Prof. M. Kisliuk.
[11] Bailey and Deshpande, Integral Equation Formulation of Microstrip
Antennas, IEEE Trans. Antennas & Propagat., Vol. AP-30, No. 4, July
1982, pp. 651-656.
[12] M. Kisliuk and W. Whistler, Radiation of Microstrip Meander-Line Slot
Antenna, VFRC Quarterly Progress Report No. 48, pp. 44-47, July-
September 1986.
[13] C.A. Balanis, Antenna Theory Analysis and Design, Harper & Row, 1982,
pp. 124-126.

-40-

- -* -'
WP -
ow

IV '