IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 39, NO.

3, MARCH 1991

332

Design of Multioctave Spiral-Mode

Microstrip Antennas
Johnson J. H. Wang, Senior Member, IEEE, and Victor K. Tripp, Senior Member, IEEE
Abstract-Existing broad-band planar spiral antennas have a cavity
loaded with absorbing material. AS a result, they are bulky and lossy.
Existing microstrip antennas are narrow-banded aess than 10%).In this
paper, the design of spiral-mode microstrip antennas with a bandwidth
of 6:l is demonstrated. This new antenna has a small dissipative loss due
to a resistive loading outside its active region.

I. INTRODUCTION
HE planar spiral-mode antennas, such as the equiangular
frequency bandwidths. Unfortunately, they radiate to both sides
of the spiral plane, whereas applications generally require a
unidirectional pattern. To overcome this difficultv over a large
frequency bandwidth, the usual approach is to-place a lossy
cavity on one side of the spiral or sinuous structure to absorb all
the undesired radiation in that direction. This cavity-backed
planar spiral was perfected more than a decade ago, achieving
typically a 9: 1 frequency bandwidth of, say, 2- 18 GHz.
The lossy cavity of the spiral and sinuous antennas has two
undesirable effects: 1) at least half of the radiated power is lost
in the dissipative cavity, 2) the cavity is deeper than the radius of
the spiral and therefore not suitable for low-profile surface
mounting.
It has been recognized, however, that the planar spiral can be
designed with the backing of a lossless cavity [3] or a conducting
plane [4]to achieve a bandwidth of 40 or 20%, respectively.
When low profile and conformability are of primary concern,
the microstrip antenna [5]is generally the antenna of choice and
has been widely used. However, existing microstrip antennas are
limited to a narrow bandwidth of less than 10%. Wood [6]
initially investigated this possibility of broad-banding by using a
single microstrip line wound as an Archimedian spiral. Wood
concluded that the achievement of a wide band analogous to the
conventional spiral was not feasible because the radiation patterns tend to exhibit a large axial ratio.
Being unaware of Woods earlier work, the senior author had
also studied the broad-band potential of the spiral microstrip
antenna [7]. He conducted impedance measurements on equiangular spirals etched on substrates of the same material but with
different thicknesses. The spiral was excited in its first mode
with a broad-band coaxial balun [8]. As expected, impedance
matching deteriorated as the substrates became thinner and the
ground plane was placed closer to the spiral. Yet even when the
substrate was only 1/16-in thick, a return loss below 10 dB (or a
voltage standing-wave ratio (VSWR) lower than 1.92) was observed over the wide frequency range of 6-16 GHz.
Although the impedance data were impressive, he soon learned
Manuscript received November 27, 1989; revised September 24, 1990.
The authors are with the Georgia Tech Research Institute, Georgia
Institute of Technolow, Atlanta. GA 30332.
IEEE Log Numbe;bOll1803.
~~

the disappointing findings of Wood regarding radiation patterns

and halted the design effort. The present renewed study was
et al. 141.
stimulated by the recent work of Nabo
that the ground
A theoretical investigation led to the
plane is compatible with the spiral modes, and we believe that
the poor radiation patterns are due to a small amount of residual
power after the electric current on the spiral has passed through
the first-mode active region (which is on a ring about one
wavelength in circumference). Thus, if we can remove the
residual power from radiation, we should be able to obtain good

11. BASICDESIGNPRINCIPLES
FOR THE SPIRAL-MODE
MICROSTRIPANTENNA
The basic planar spiral antenna, which consists of a planar
sheet of an infinitely large spiral or sinuous structure, radiates on
both sides of the spiral in a symmetric manner. When radiating
in the n = 1 mode, most of the radiation occurs near a circular
ring around the center of the spiral whose circumference is about
one wavelength. As a result, one can truncate the spiral outside
this radiation ring without too much effect on the radiation
pattern.
In practical applications, there is virtually no need for a spiral
radiating on both sides of the spiral plane. In a cavity-backed
planar spiral, one side of the radiated power is absorbed by a
lossy cavity, otherwise the undesired radiation on the cavity side
may interfere with the desired radiation on the other side.
When a ground plane is placed behind a planar spiral antenna,
its interference with the radiation pattern has long been recognized, even though little has been documented. We have examined these effects theoretically by considering the existence and
compatibility of the radiation modes of an infinite, planar spiral
structure backed by a conducting plate as shown in Fig. 1. This
analysis shows that as far as the spiral radiation is concerned, a
ground plane appears quite tolerable. The analysis also shows
that the effects of dielectric backing are detrimental, and therefore the dielectric constant of the substrate should be kept low.
The method of analysis follows that of Cheo et al. [9].
A simple technique to remove the residual power is to place a
ring of absorbing material at the truncated edge of the spiral
outside the radiation zone. This scheme allows the absorption of
the residual power which would radiate in an n = - 1 mode
(opposite sense CP) and n = + 2 modes, causing drastic deterioration of radiation patterns, especially at off-boresight angles.
We present below the experimental results of our design effort
on the spiral-mode microstrip antenna.

111. EXPERIMENTAL
RESULTS
The experimental models are illustrated in Fig. 2. As we have
said, the power not radiated can Only disrupt the radiation
pattern because it will radiate in in = - 1, +_2modes or as
diffracted fields at the point of spiral truncation. Thus a simple

0018-926X/91/0300-0332$01.00 0 1991 IEEE

11

333

WANG AND TRIPP: SPIRAL-MODE MICROSTRIP ANTENNAS

10
Frequency(GHr)

20

Fig. 3. Measured VSWR of an Archimedian spiral microstrip antenna of

Fig. 2.
10 0 ,

__5

z=o

5 0

z = -d

ground plane

-5.0

Fig. 1. An infinite planar spiral over a ground plane.

-10 0

18 -in diameter

-15 0
-20

Y
I

-25 0

-30 -90
Or 0

-60 0

30 0

-30 0

Angle

60 0

90 0

(Deg)

Fig. 4. Measured pattern for an Archimedian spiral microstrip antenna of

Fig. 2 at 2 GHz.

I3-in diameter

*be
/

spiral

+ - in wide ring
of absorber

/ / / / / /

coax balun

//
x

ground plane

Fig. 2. A 3-in diameter spiral microstrip antenna on an 18-in diameter

ground plane.

solution is to let the residual power be dissipated in a resistive

load in the truncation region. We have tried several dissipative
loads, including a resistive paint on the spiral arm at the region
of truncation and a lumped resistive load. The loading technique that has effectively dissipated the residual power over a
wide bandwidth is shown in Fig. 2. A 112411 wide ring of
microwave absorbing material, thick enough to fit the gap
between the spiral and the ground plane, is placed half inside the
truncated spiral and half outside, as shown in Fig. 2. The
antenna is fed by a broad-band tapered coaxial balun based on
the design of Duncan and Minerva [8].
A number of experimental models were fabricated and tested.
They include both equiangular and Archimedian spirals, and

with various gap widths d between 0.06 and 0.30 in. They all
exhibit good impedance and pattern performances over a 6:l
bandwidth, typically over 2-12 GHz. The main effect of a
reduced gap width d is a decrease in gain and efficiency at lower
operating frequencies.
The measured VSWR of an Archimedian spiral microstrip
antenna depicted in Fig. 2 with d = 0.25 in is shown in Fig. 3.
As can be seen, the VSWR is mostly below 1.5 between 2 and
18 GHz.
Good rotating-linear radiation patterns have been obtained
experimentally in the 2-12 GHz range as exemplified by Figs.
4-7 at 2, 4, 8, and 12 GHz for an Archimedian spiral shown in
Fig. 2 with d = 0.25 in. The pattern cuts are over the angle 8 in
Fig. 2 on a selected plane, say, the x = 0 plane, with respect to
a rotating-linear-polarization horn. The antenna gain is with
respect to a linear antenna (dBiL). If polarizations are matched,
and if the axial ratio is low, the gain in dBi is generally 3 dB
higher.
Although these patterns exhibit a larger axial ratio than the
fine-tuned 2- 18 GHz cavity-loaded spirals commercially available at this time, they are better than the patterns of unloaded
conical and planar spirals in the literature. Some of the pattern
ripples and beam asymmetries are due to the imbalance of the
balun feed and the edge of the finite ground plane, and can be
considerably improved.
A number of models, with the spiral geometry being equiangular or Archimedian, but of the same 3-in diameter, were
constructed and measured. The gap between the spiral and the
ground plane, d in Figs. 1 and 2, is varied between 0.06 and 0.3
in. All the impedance and pattern performances are about as

IEEE TRANSACTIONSON ANTENNAS AND PROPAGATION, VOL. 39, NO. 3, MARCH 1991

334

5 0
0

U
-

-5 0

-10

-15 0

C
C

m
Y

-20

4
C

-25 0

1 , )

- 3 0 0 ~ " " " " " " " " " " " " " " " " " "
0
-90 0
-60 0
-30 0

60 0

90 0

(Oeg)

Angle

Fig. 5.

30 0

Measured pattern for an Archimedian spiral microstrip antenna of

Fig. 2 at 4 GHz.
in o

~ \ A A .

5.0
I

0
-5 0

n.,

2 1
-90 0

-60

-30

Angle

30 0

60 0

90 0

r"""'

M~~

&dmmdan (d-0.lOhch)

- --

(Deg)

Fig. 6. Measured pattern for an Archimedian spiral microstrip antenna of

Fig. 2 at 8 GHz.

-2

6
7
8
9
Frequency (GHz)

10

11

12

Fig. 8. Antenna gain.

good as those in Figs. 3-1. However, when d is decreased to

below 0.15 in, the antenna gain at lower frequencies decreases,
even though the impedance and pattern performance do not
deteriorate appreciably.
We have also measured the gain of these antennas in an effort
to determine their efficiency and dissipative loss due to edge
loading. The gain and efficiency of several antenna models are
displayed in Figs. 8 and 9 in comparison with a cavity-loaded
planar spiral of about the same aperture size (2.5-in diameter).
Since gain measurements may have errors up to 1 dB, and since
the directivity of the antenna is computed by integrating the
power with some approximations, the efficiency data thus derived are not very accurate.
It is difficult to measure antenna efficiency. Our approach is to
perform a full three-dimensional spatial integration of the measured gain pattern to determine the directivity and the gain of the
antenna with d = 0.3 in. Interpolation of discrete data points is
undertaken to obtain digitized data for full spatial coverage. The
resulting efficiency, based on two independent sets of measured
data, is found to be within the upper and lower bounds shown in
Fig. 9. As can be seen, there is a definite improvement of
efficiency as compared with the cavity-loaded spiral in the range
of 1-3 dB. The accuracy of the efficiency can be improved by
measuring a more detailed spatial coverage of radiation patterns.

IV. CONCLUSION
We have demonstrated that spiral-mode microstrip antennas
can achieve a bandwidth as wide as 6:l for patterns, and even
wider for VSWR. As a specific measure of bandwidth at microwave frequencies, the spiral-mode microstrip antenna gain is

loo

20 -

MiCrOStrip

Upper Bound

---- ------- ---

Lower Bound

Cavity-LoadedSpiral

higher than that of the conventional loaded-cavity spiral over a

5:1 band if the spacing d is 0.25 in or larger over the 2- 12 GHz
range. Cursory efficiency calculations show consistent improvement over a loaded-cavity spiral. Finally, a study of ground
plane spacing showed that the antenna gain remains above the
conventional spiral gain as the spacing d is reduced until it is as
small as 0.1 in.

ACKNOWLEDGMENT
The authors would like to acknowledge the valuable assistance
of Mr. E. E. Weaver in the experimental phase of this work and

WANG AND TFUFP: SPIRAL-MODE MICROSTRIP ANTENNAS

Mr. Robert L. Davis and Dr. Charles J. Drane for their encouragement and support of this research.

335

B. R. S . Cheo, V. H. Rumsey, and W. J. Welch, A solution to

frequency-independent antenna problem, IRE Trans. Antennas
Propagat., vol. AP-9, pp. 521-534, Nov. 1961.

REFERENCES
V. H. Rumsey; Frequency Independent Antennas. New York:

Academic, 1966.
R. H. DuHamel, Dual polarized sinuous antennas, European
Patent Application 019 8578, filed Feb. 19, 1986 (U.S.Patent
7 03 042, Feb. 19, 1985).
R. C. Hansen, Ed., Microwave Scanning Antennas, Vol. 2 ,
Array Theory and Practice. New York: Academic, 1966, pp.
116-127.
H. Nakano, K. Nagami, S . Arai, H. M i d , and J. Yamauchi,
A spiral antenna backed by a conducting plane reflector, ZEEE
Trans. Antennas Propagat., vol. AP-34, pp. 791-796, June
1986.
R. E. Munson, Conformal microstrip antennas and microstrip
phase array, IEEE Trans. Antennas Propagat., vol. AP-22,
pp. 74-78, Jan. 1974.
C. Wood, Curved microstrip lines as compact wideband circularly polarized antennas, Inst. Elec. Eng. Microwaves, Optics
and Acoustics, vol. 3, pp. 5-13, Jan. 1979.
J. J. H. Wang, A study of the spiraphase and anisotropic
substrates in microstrip antennas, Rome Air Development Center, Griffiss AFB, NY, Interim Rep. RADC-TR-85-146, Aug.
1985.
J. W. Duncan and V. P. Minerva, 100:l bandwidth balun
transformer, Proc. ZEEE, vol. 48, p. 156-164, Feb. 1960.

Johnson J. H. Wang (M68-SM79) was born

in Hunan, China, on October 24, 1938. He
received the B.S.E.E. degree from National
Taiwan University and the Ph.D. degree from
The Ohio State University, Columbus.
From 1968 to 1975, he conducted research in
several industrial firms. Since 1975 he has been
with Georgia Institute of Technology, Atlanta.
His research activities include antennas, phased
arrays, computer solution of electromagnetic
problems, microwave imaging, bioelectromagnetics, radiation and scattering measurements, and recently in the theory
of chaos and superconductor.
Dr. Wang is a member of Sigma Xi and the Electromagnetics
Academy. He is the author of a book, Generalized Moment Methods
in Electromagnetics-Formulation and Computer Solution of Zntegral Equations, (Wiley, 1991).

Victor K. Tnpp (M73-SM90), for a photograph and biography please

see page 1414 of the November 1989 issue of this TRANSACTIONS.