IEEE Standard: Test Procedure For Antennas

IEEE Standard
Test Procedure
for Antennas
Approved March 1, 1971
American National Standard Institute
Authorized licensed use limited to: MAPUA INSTITUTE OF TECHNOLOGY. Downloaded on October 05,2016 at 02:55:05 UTC from IEEE Xplore. Restrictions apply.
ANSI
C16.11-1971
IEEE
Std 149-1965
Reaffirmed 1971
IEEE Standard
Test Procedure
for Antennas
Approved March 1, 1971
American National Standard Institute

Corrected Edition
When IEEE Std 149-1965 was originally approved, it included Section 11 Definitions. That section was
superseded by IEEE Std 145-1969, IEEE Standard Definitions of Terms for Antennas. This edition has been
published without the superseded definitions, as reaffirmed in 1971 by the IEEE Standards Committee and approved
by the American National Standards Institute.
Copyright 1971 by
The Institute of Electrical and Electronics Engineers, Inc.

No part of this publication may be reproduced in any form,
in an electronic retrieval system or otherwise,
without the prior written permission of the publisher.
ACKNOWLEDGMENT
The Institute wishes to acknowledge its indebtedness to those who have so freely given of
their time and knowledge, and have conducted experimental work on which many of the IEEE
publications are based.
This publication was prepared by the Subcommittee on Methods of Testing Antennas of
the IEEE Antennas and Waveguides Committee.
Members of the Subcommittee on Methods of Testing Antennas were:
R. C. Hansen, Chairman
E. L. Bock
R. Krausz
R. W. Clapp
J. E. Holland
S. M. Kerber
T. Kinaga
D.
R.
K.
R.
E. Kreinheder
L. Mattingly
A. Norton
J. Stegen
L. C. Van Atta
Members of the Subcommittee on Definitions of Antenna Terms who contributed to this
document were:
P. H. Smith, Chairman
C. C. Allen
R. W. Klopfenstein
P. S. Carter
P. A. Loth
P. W. Hannan
E. N. Torgow
Members of the Antennas and Waveguides Committee who contributed to this document
over the past six years are as follows:
C. C. Allen
R. L. Mattingly
P. S. Carter
A. A. Oliner
H. V. Cottony
K. S. Packard
G. A. Deschamps
D. C. Ports
P. W. Hannan
S. W. Rubin
R. C. Hansen
W. Sichak
H. Jasik
C. J. Sletten
W. K. Kahn
R. W. Klopfenstein
D. J. Levine
P. A. Loth
P. H. Smith
W. V. Tilston
E. N. Torgow
M. S. Wheeler
The Chairmen during this period were G. A. Deschamps, A. A. Oliner, R. L. Mattingly, and
P. W. Hannan.
CONTENTS
1. I N T R O D U C T I O N
1.1 Foreword
1.2 Symbols and Units
1.3 Environmental Factors
1.4 Reciprocity
1.5 Polarization
1.6 Field Regions
1.7 Scale Models
5
5
5
5
6
6
8
9
2. MEASUREMENTS FOR DETERMINATION O F A M P L I T U D E

PATTERNS
2.1 General
2.2 On-Site Measurements
2.3 Off-Site Measurements
10
10
11
13
3. MEASUREMENT O F O T H E R PATTERN CHARACTERISTICS

3.1 Phase Measurements
3.2 Polarization Measurements
3.3 Special Measurements for Boresight and Angular Sensitivity
15
15
17
18
4. MEASUREMENT O F MAXIMUM POWER GAIN AND

DIRECTIVITY
4.1 General
4.2 Measurement of Maximum Power Gain
4.3 Determination of Directivity
19
19
20
21
5. D E T E R M I N A T I O N O F RADIATION EFFICIENCY
21
6. GROUND W A V E MEASUREMENTS
22
7. MEASUREMENTS O F IMPEDANCE
7.1 Input Impedance Measurements
24
24
7.2 Mutual Impedance Measurements
25
8. P O W E R HANDLING MEASUREMENTS
26
9. NOISE T E M P E R A T U R E MEASUREMENTS
27
10. D E T E R M I N A T I O N O F SCATTERING CROSS SECTION
28
Test Procedure for

ANTENNAS
1.
INTRODUCTION
1.1
Foreword
This Test
vious issue,
T e s t i n g " 48
ments of the
P r o c e d u r e , which supersedes the p r e " S t a n d a r d s on A n t e n n a s M e t h o d s of

I R E 2S2, is concerned with m e a s u r e properties which characterize a n t e n n a s .
electric field strength, volts per meter
=
=
=
=
=
frequency, cycles per second

maximum power gain of an antenna
directivity of an antenna
current, amperes
loss factor (ratio of power input to output)
z= power, watts
distance from antenna to observation point,

meters
= standing wave ratio (voltage or current)
= noise temperature, degrees Kelvin
= voltage, volts
= impedance, ohms
= co-latitude angle (see Figure 3)
= free-space wavelength, meters
voltage reflection coefficient
= azimuthal angle (see Figure 3)
= antenna rotation angle
Environmental Factors
One category of environmental factors may be defined

as directly affecting the material properties or structure of
an antenna, and thus indirectly affecting the electrical
characteristics. Mechanical loading of the antenna structure by wind and ice is a common, but nevertheless important, effect to be considered in the testing of many
antennas. Vibration and shock tests are often made to
assure that an antenna subject to severe accelerations will
maintain its structural integrity, as well as to determine
whether dynamic deformations are within allowable electrical limits. Antennas in exposed locations are often provided with lightning protection and anti-icing devices; the
effect of such devices on the electrical characteristic must
be evaluated.
Various natural or man-made environments may impose
special requirements: for example, shipborne antennas
may have to withstand water-wave impact and salt-water
corrosion. Antennas on hypersonic vehicles must withstand very high temperatures and pressures, and antennas
designed for satellite or space-probe application must
withstand intense ionizing radiation, hard vacuum, and extreme temperatures. Those ground-based antennas which
are intended to operate in the vicinity of a nuclear blast
should maintain their essential properties in the wake of
seismic waves, atmospheric shock waves, thermal radiation, ionizing radiation, blast-product erosion, and radioactive debris.
1.2 Symbols and Units

T h e basic symbols used in this T e s t P r o c e d u r e are
given in the following list. Rationalized M K S units
are employed, a n d are listed for each symbol.
/
Gmu
Gdmu
/
L
ratio of full-scale dimensions to scale-model

dimensions
An antenna can be considered adequately tested only

when the tests have recognized the environmental conditions of operation affecting antenna performance. Many
of these environmental factors are of a specialized nature,
and it is impractical to include the appropriate tests in
this Test Procedure. Instead, a few cases will be briefly
mentioned as examples of interest to the reader.
Throughout this document, certain terms are italicized where

they first appear in each section. Definitions of these terms are
given in IEEE Std 145-1969, IEEE Standard Definitions of
Terms for Antennas, and supersede those which appeared as
Section 11 of the 1965 edition of IEEE Std 149-1965.
half-power beamwidth, degrees

diameter or maximum dimension of an antenna aperture, meters
1.3
In Sections 2 t h r o u g h 10, those a n t e n n a characteristics which m a y require m e a s u r e m e n t are considered. T h e techniques associated with the m e a s u r e ment of each are discussed, but n o a t t e m p t is made
to furnish step-by-step procedural descriptions. References which are illustrative of m e a s u r e m e n t techniques are provided in which details m a y be found.
Since m e a s u r e m e n t techniques u n d e r g o continuing
refinement, the reader should be alert to references
on the subject of a n t e n n a m e a s u r e m e n t which a p pear after this T e s t P r o c e d u r e was p r e p a r e d .
=
=
S
T
V
Z
0
X
p
*
r
In Section 1.2, symbols and units to be used in this

T e s t P r o c e d u r e are defined. T h e effect of the environment on the a n t e n n a and its characteristics is
discussed in Section 1.3, and certain precautions to
be taken during the m e a s u r e m e n t s are suggested. In
Section 1.4, t h e usefulness of the reciprocity t h e o r e m
in a n t e n n a m e a s u r e m e n t s is cited, along with examples of situations in which this theorem does not
apply. T h e characteristics of electromagnetic wave
polarization are outlined in Section 1.5, with e m p h a sis on those aspects which are important in the
m e a s u r e m e n t of a n t e n n a s . Section 1.6 treats those
e l e c t r o m a g n e t i c field regions which are of particular
interest in a n t e n n a m e a s u r e m e n t s . In Section 1.7,
the features and limitations of scale models for antenna m e a s u r e m e n t s a r e discussed.
BW
D
Certain antenna applications necessitate unusual attention to tests involving quite ordinary aspects of the physical
Tests of radome effects are often made by observing the

change in antenna performance as the radome is revolved
around the antenna, or as the radome is alternately removed and replaced. Characteristics measured generally
include power gain, amplitude of radiation pattern, and
input impedance; other quantities sometimes investigated
are phase, polarization, boresight, power-handling capability, and noise temperature. In determining the effect of
a radome, it has usually been found that significant results
are obtained only when the radome is tested together with
exactly the same type of antenna with which it is to be
used.
When an antenna is covered by a radome, it is necessary, of course, to determine whether the radome will withstand the destructive factors of the physical environment
which otherwise would have been applied to the antenna
itself.
environment. For instance, rapid-scan antennas employing

ferrite components are especially sensitive to changes in
ambient temperature. Antennas in which intermodulation
distortion must be minimized may have difficulty with nonlinear corrosion films. In the case of precision tracking
antennas, a boresight error can be caused merely by deflections resulting from nonuniform heating of the antenna
by the sun. Large antennas rotated in the elevation plane
may also undergo significant deflections because of the
change in the effect of the gravitational force on the
antenna.
The other category of environmental factors may be
considered as directly affecting the electrical properties of
the antenna. In some cases this electrical environment is
permanent, and is included in the design as an integral
part of the antenna. Familiar examples of this situation
are the role of the earth in ground-based antennas, the
superstructure in shipborne antennas, and the vehicle in
antennas for airborne or space applications. Techniques
for testing in these circumstances have been developed and
are widely used; they are described in Sections 2.2 and 6
of this Test Procedure. Another example is the noise
environment of an antenna, including galactic radiation,
thermal atmospheric radiation, atmospheric and man-made
static, and thermal earth radiation; many of these factors
are recognized in Section 9, which describes tests of antenna noise temperature. A less familiar but more extreme
example of permanent electrical environment occurs with
low-frequency antennas designed to operate underground
or beneath the surface of the sea.
1.4 Reciprocity
The reciprocity principle3 is of fundamental importance
in the determination of many of the properties of antennas,
because it permits the determination of these properties
from measurements on the antennas in either the transmitting or the receiving condition. For example, the radiation pattern and gain characteristics of an antenna may be
determined from measurements of the antenna during reception, and this is often done for convenience.
In describing methods of making measurements, it is
generally desirable to limit the description to one condition of operation. It should be noted that reciprocity does
not imply that antenna current distributions are the same
for transmitting as for receiving. The reciprocity principle may not apply to antennas containing components
such as ferrite devices or semiconductor diodes, or to antennas in the presence of a medium such as an ionized gas
in a magnetic field.
In many cases the electrical environment of an antenna

is transient, and causes problems of a more elusive nature.
Moisture, when combined with impurities, can create conducting paths which are troublesome, especially in regions
of high electric field. Even pure water, if it exists in a
continuous film, may affect the performance of a microwave antenna by virtue of its high dielectric constant. Ice
or snow are also factors which can directly affect the
characteristics of an antenna. In the case of antennas for
missile application, ionized gases may exist in the exhaust
or may be generated by high vehicle velocity through the
atmosphere; these gases are also a part of the electrical
environment. Finally, certain antennas may be susceptible
to precipitation static, which is a series of noise pulses
created when charged particles such as raindrops discharge
on an antenna.
There is a particular type of antenna environment that
deserves special mention; namely, the radome. This is a
structure1 which often encloses an antenna to protect it
from the physical environment. However in so doing, it
introduces an electrical environment whose effects should
not be overlooked.
One typical class of radome has a pointed shape and
covers an antenna on the nose of an aircraft or missile.
Another common type 2 has the shape of a truncated
sphere; it is fixed to the ground and covers an antenna
which is sometimes quite large in size. Most radomes are
constructed essentially of dielectric materials, although
some contain a significant amount of metal for greater
strength.
1.5
Polarization
Polarization is a property of electromagnetic radiation

describing the spatial orientation of the field vectors; in
common practice only the electric field (E) vector is
specified. For the case of radiation by an antenna, the
electric field is generally sampled in small regions far from
the antenna, where the electromagnetic wave is nearly uniform and plane, and where the ^-vector lies in a plane
nearly perpendicular to the direction of wave propagation.
Coherent excitation of the antenna is always assumed,
yielding radiation which is said to be completely polarized.
The -vector, when observed at a fixed point in space,
can vary periodically in amplitude, direction, or both; the
manner in which it varies determines the type of polarization. When the tip of the -vector traces an ellipse in a
fixed plane perpendicular to the propagation direction, the
1 W. M. Cady, M. B. Karelitz, L. A. Turner "Radar Scanners
and Radomes", M.I.T. Radiation Laboratory Series, vol. 26,
McGraw-Hill, pp. 241-460; 1948.
2 W. Lavrench, "Electrical Performance of Rigid Ground Radomes", I R E Transactions on Antennas and Propagation, vol.
AP-8, No. 6, pp. 548-558; November 1960.
3 E. C. Jordan "Electromagnetic Waves and Radiating Systems",
Prentice-Hall, Chater 10; 1950.
wave is said to be "elliptically polarized". This is the most

general type of polarization, and is indicated in Figure
1 ( a ) . When the ellipse degenerates to a line, as indicated
in Figure 1 ( b ) , the polarization is called "linear polarization". When the ellipse becomes a circle, as indicated in
Figure 1 ( c ) , the term "circular polarization" is applied.
The direction of rotation of the ^-vector in a fixed
plane, observed looking in the direction of wave propaga-
tion, is the "handedness" of the elliptical or circular polarization.4- 5 Clockwise rotation is called right-handed and
counterclockwise is called left-handed. It is interesting
that the opposite sense of rotation is obtained if the fixed
plane is observed looking against the direction of propagation. The opposite sense is also obtained if a series of
planes are observed at one instant of time, and the resulting spiral (see Figure 1 ( a ) or ( c ) ) locus is used to define
DIRECTION
OF PROPAGATION
UNFIXED
PLANE
DIRECTION OF
ROTATION OF
E VECTOR IN
THIS PLANE
(a) Right-hand
elliptical
polarization
(b) Vertical
linear
polarization
( c ) Left-hand
circular
polarization
Figure 1
Electric field vectors of several types of polarization, s h o w n at various points
in space at one instant of time
the handedness. These two misinterpretations have been1
a source of some confusion.
Any given polarization can be resolved into two independent "orthogonal" components for convenience in representation or analysis. These are usually linear components>
in space quadrature, are sometimes oppositely-rotating circular components, and may also be elliptical components>
with a certain relationship. The relative magnitude and
phase of the orthogonal components must be known in
order to specify the polarization completely. It is interesting to note that two equal-magnitude linear components in
both space and time quadrature constitute circular polarization. The general case of elliptical polarization may also
be specified in terms of the "polarization ellipse" traced
out by the tip of the E-vector (see Figure 1 ( a ) ) . The

axial ratio of the ellipse, the orientation in space of the
major axis, and the direction of rotation comprise the complete description. This representation of polarization is
equivalent to and interchangeable with the resolution into
components described above; the appropriate transformations are described in the literature. 6
4 " I R E Standards on Wave Propagation: Definitions of Terms'*,
I E E E No. 211 (formerly 50 I R E 24 S I ) ; 1950.
5 J. D. Kraus, "Antennas", McGraw-Hill, pp. 470-471; 1950.
6 H. G. Booker, V. H. Rumsey, G. A. Deschamps, M. L. Kales,
J. I. Bohnert, "Techniques for Handling Elliptically Polarized
Waves with Special Reference to Antennas", Proc. IRE, vol.
39, pp. 535-552; May 1951.
lated from the angle between these two points at the center
of the sphere. This latter method is particularly useful for
the interpretation of complex situations.
The transfer of power between two antennas is dependent on the polarization characteristics of their radiated
waves. When the two antennas have the same polarizations in the appropriate directions, maximum power is
transferred. Any failure of correspondence in polarization
causes a loss in the antenna system; for the extreme case
of exact orthogonality between the two polarizations, no
power is transferred. A useful example of an intermediate
case involves a circularly-polarized and a linearly-polarized antenna; here the system loss is three decibels regardless of the relative orientations of the two antennas.
1.6 Field Regions

The distribution of field strength around an antenna is
a function of the distance from the antenna. In close proximity to the antenna, the field strength may include, in
addition to the radiating field, a significant reactive (nonradiating) field. The strength of the reactive-field components decays rapidly with the distance from the antenna
(inversely as distance raised to powers greater than unity).
That region of space immediately surrounding the antenna
in which the reactive components predominate, is known
as the reactive near-field region. The size of this region
varies for different antennas. For most antennas, however, the outer limit is of the order of a few wavelengths
or less. For the particular case of an electrically-small
dipole indicated in Figure 2 ( a ) , the reactive field predominates out to a distance of approximately \/2TT where the
radiating and reactive fields are equal.
When the polarization characteristics of the two antennas are known, the polarization loss may be computed.
A basic expression 7 for the ratio of the power actually
transferred to the power available for transfer with matched
polarizations i s :
^transferred _
^available
"
\En\En2 + c l c 2 p
( l ^ l l 2 + |cl|2) (|n2|2 + l ^ l 2 )
Beyond the reactive near-field region, the radiating field

predominates. The radiating region is divided into two
subregions: the radiating near-field region and the farfield region. The former region exists for most electrically-large antennas, but not for electrically-small antennas.
The latter region exists for all antennas.
where E is a complex quantity defining the amplitude and

phase of the electric field radiated by an antenna, the subscripts n and c refer to two orthogonal components of
polarization, and the subscripts 1 and 2 refer to the two
antennas involved in the power-transfer system. In defining these orthogonal components, one must be careful to
consider their right or left hand nature from the respective viewpoints of the two waves propagating in opposite
directions, while considering their space orientations from
a single viewpoint. For convenience, the first polarization
component may sometimes be regarded as the normal or
desired component having a reference phase of zero, and
the second as a cross-polarization or undesired component
having a phase angle relative to the normal component. It
may be seen from equation (1) that the polarization loss
in a system of two antennas is not, in general, simply the
sum of the individual losses into cross polarization; the
system loss may be either greater or less depending on the
phases of the two cross polarizations.
Various alternate methods for computing the
tion loss in the transfer of power between two
are available. Two methods employ equations
parameters of the polarization ellipse8* 9 or the
polarization components. 10 ' n
In the radiating near-field region, the relative angular

distribution of the field, (the usual radiation pattern) is
dependent on the distance from the antenna. The reason
for this behavior is two-fold: (1) the relative phase relationship of field contributions from different elements of
the antenna changes with distance, and (2) the relative
amplitudes of these field contributions also change with
distance. For an antenna focused at infinity, indicated in
Figure 2 ( b ) , the radiating near-field region is sometimes
referred to as the Fresnel region on the basis of analogy
to optical terminology.
In the far-field region, the relative angular distribution
of the field becomes essentially independent of distance.
Correspondingly, the amplitude of the field is given, in the
limit, by the reciprocal of the first power of distance. The
reason for this behavior is that the relative phase and
amplitude relationships between the field contributions
from different elements of the antenna approach a fixed
relationship. Although this situation is not attained precisely until the observation point is an infinite distance
from the antenna, the relative angular distribution of the
field at a comparatively short distance is often an adequate
approximation of the field distribution at infinity. Thus
the far-field region is considered to extend from this distance to infinity. For an antenna focused at infinity, the
far-field region is sometimes referred to as the Fraunhofer
region on the basis of analogy to optical terminology.
polarizaantennas
involving
circular-
Another method employs a graphical technique in which

the two polarizations are plotted on the surface of a Poincare polarization sphere, 1 2 , 1 3 and the loss is easily calcu-
7 H. Jasik, "Antenna Engineering Handbook", McGraw-Hill, p.

17-7; 1961.
8 W. Sichak and S. Milazzo, "Antennas for Circular Polarization", Proc. IRE, vol. 36, pp. 997-1001; April 1948.
9 "Reference Data for Radio Engineers", International Telephone
and Telegraph Corp., Fourth Edition, pp. 665-667; 1957.
10 L. Hatkin, "Elliptically Polarized Waves", Proc. IRE, vol. 38,
p. 1455; December 1950.
11 V. H. Rumsey, "Transmission Between Eliptically Polarized
Antennas", Proc. IRE, vol. 39, pp. 535-540; May 1951.
12 G. A. Deschamps, "Part IIGeometrical Representation of the

Polarization of a Plane Electromagnetic Wave", Proc. IRE, vol.
39, pp. 540-544; May 1951.
13 "Reference Data for Radio Engineers", International Telephone
and Telegraph Corp., Fourth Edition, pp. 667-670; 1957.
sometimes be used to infer the pattern in the far-field region (see Section 2.3).
For electrically-large antennas other than conventional
broadside-aperture types, there is no recognized criterion
defining the distance to the far-field inner boundary. However the 2D 2 /X criterion, when D is taken as the largest
linear dimension, will usually give a distance which is well
within the far-field region. Special precautions should be
observed when the environment of the antenna plays a
part in the formation of the radiation pattern. For such
situations, the distance to the inner boundary of the farfield region is determined by the dimension of the entire
radiating structure; this may involve a large metallic
supporting surface (for example, an aircraft fuselage) in
a mobile installation, or the ground terrain in a fixed
installation (see Section 2.2).
(a) Electrically small dipole
Certain measurement situations do not permit separation of the useful space into simple near- and far-field
regions. An important example of this class is that of a
vertically polarized antenna operating over a ground of
finite conductivity (see Section 6 ) .
RADIATING
NEAR-FIELD
REGION
1.7
Scale Models
Scale-modeling techniques are often used when the

measurement of the original antenna in its original environment is impractical. This situation frequently exists
for antennas located on large support structures, such as
ships, aircraft, and large man-made satellites, which influence the antenna properties. In such cases the original
radiating system frequently cannot be subjected to full
experimental control because of instability of the supporting vehicle or the surrounding medium, or because
some of the radiation directions of interest are inaccessible
to accurate measuring techniques of the usual types. In
very large radiation systems the accessibility problems
tend to become greater. In moving systems or changing
environments, the stability problem may compel an unreasonable amount of statistical measurement. Another
major situation for which scaling is often used is development work requiring successive modifications which are
costly due to the large size or, occasionally, the very small
size of the final antenna system. Thus the main motives
for scale modeling are to obtain greater experimental
control over the measurements, or to obtain an economic
advantage in the measurement program.
(b) Electrically large reflector

Figure 2
Field regions for t w o antenna types
For electrically-large antennas of the broadside-aperture
type such as that shown in Figure 2 ( b ) , a commonly used
criterion 14 to define the distance in free space to the boundary between the radiating near-field and the far-field regions is
K = 2D2
(2)
where D is the largest dimension of the aperture. The

diflFerence in path lengths between the center and edges of
the aperture to the point at the region boundary is X/16.
At this boundary, the antenna gain over most of the major
lobe of the radiation pattern differs from that at infinity by
a very small factor, the exact value depending on the shape
of the aperture and the illumination taper. However, for
directions corresponding to minor lobes and pattern minima, the gain at the region boundary may differ from that
at infinity by many decibels.
For most applications, it is the radiation pattern in the
far-field region that is important. It is, therefore, customary to make measurements in the far-field region. If
the distances involved are prohibitively great, measurements in what would normally be the near-field region can
Generally, the model is reduced in size from the fullscale antenna, but whether reduced or increased, the requirements 15 - 1 6 1 7 for exact simulation by the model are
as follows:
Linear dimensions of the model to be 1/n times those of
the full-scale antenna.
15 G. Sinclair. E. C. Jordan, E. W. Vaughan, "Measurement of
Aircraft Antenna Patterns Using Models", Proc. IRE, vol. 35,
pp. 1451-1462; December 1947.
16 J. A. Stratton, "Electromagnetic Theory". McGraw-Hill, pp.
488-490; 1941.
17 R. W. King, "Electromagnetic Engineering", McGraw-Hill,
vol. I, pp. 316-320; 1945.
14 S. Silver, "Microwave Antenna Theory and Design", M.I.T.

Radiation Laboratory Series, vol. 12, McGraw-Hill, Section
6.9; 1949.
Operating frequency and conductivity of the materials

used in the model to be n times those of the full-scale
antenna.
high-voltage breakdown (see Section 8) and the noisetemperature (see Section 9 ) , do not scale, because of the
frequency-dependent nature of the mechanisms involved.
Electric permittivity and magnetic permeability of

the materials used in the model to be the same as in
the full-scale antenna.
In the measurement of radiation patterns, it is likely that

cables to the scale model will perturb the patterns appreciably. T o overcome this the scale-model antenna may
transmit from a battery-operated transmitter, or, alternatively, the scale-model antenna may contain a receiver from
which the demodulated signal can be removed by means of
high-resistance wire leads which minimize the disturbance
of the radio frequency field. The support for the model
should receive particular attention if the pattern null
structure is to be accurately determined.
In the above, n is an arbitrary number (generally, but

not necessarily, greater than unity) which determines
the scale of the model. Also in the above, the electric
permittivity and magnetic permeability may be extended to
their complex form 18 to include electric and magnetic
type losses, respectively; this is a substitute for the ordinarily-used electrical conductivity and the little-used, hypothetical, magnetic conductivity 18 , respectively. More general forms of scaling, which permit additional parameters
to be changed, are occasionally desired; these are described in the literature. 19
In the case of an input-impedance measurement using

a scale model, it may sometimes be desired to determine
the effect of a full-scale matching section for the antenna.
T o obtain this information, the matching section may be
terminated in a network equivalent 20 to the impedance
measured on the scale model. Special measurements on
antennas which are electrically small, as is common at
low frequencies, can often be made by quasi-static methods
using electrostatic cages 21 . Here the charge induced by a
known field strength is measured, allowing an equivalent
area to be determined. For the case of scattering measurements of radar targets (see Section 10), model techniques
are quite common because of the considerable difficulty in
making measurements with the actual target.
In a practical model it is usually not feasible to satisfy

exactly the full set of requirements in the above list.
However, in antennas which are not highly resonant, the
error will usually be small if good conductors such as
copper or aluminum are used to simulate good conductors,
and if low-loss dielectric materials of identical electric
permittivity and magnetic permeability are used to simulate low-loss dielectrics. The principal problem is presented by poor conductors or lossy dielectrics; in such
cases it is not always possible to obtain materials which
satisfy the scale-modeling requirements.
In addition to the special probems of scale-model measurements referred to in this section, there are general procedures and precautions to be employed during any of the
measurements. These procedures are essentially the same
as those described in the following sections of this Test
Procedure for full-scale antennas.
In constructing a scale model, not only the antenna

needs to be simulated but also those portions of the surrounding structures and environment which have an
appreciable effect on the properties of the antenna. Good
judgment is required in determining how far to simulate
the surroundings. In many cases it is difficult to construct
simulated surroundings, due to their complexity and to
their electromagnetically accidental nature. For example,
when the antenna interacts with the earth, it has often
proven to be impractical to accurately scale the highly
variable and sometimes unknown properties of the soil.
2. M E A S U R E M E N T S F O R D E T E R M I N A T I O N
OF A M P L I T U D E P A T T E R N S
2.1
General
The radiation pattern of an antenna may represent any

of the several properties of electromagnetic wave radiation as a function of space coordinates. In Section 2,
only the pattern of radiation amplitude is considered; the
other properties, such as phase and polarization, will be
coyered in Section 3. Furthermore, Section 2 is concerned with only the relative amplitudes of the antenna
response; determination of the absolute values is treated
in Section 4 which discusses gain measurements.
While the radiation pattern is usually the antenna characteristic of most concern in model measurements, many
other antenna properties of interest can also be reliably
reproduced. If the exact scaling procedure described
above is followed throughout the antenna system, all
fields are reproduced exactly in shape, both externally and
within the feed line. Thus power gain, directivity, radiation efficiency, input impedance, mutual impedance, boresight error, and in general all properties dependent only
on field ratios, are preserved. If the modified scaling procedure is followed, efficiencies will not be reproduced but
the remaining properties will be reproduced accurately
enough for most purposes, providing that the antenna does
not have extremes of current concentration or mismatch.
Certain antenna characteristics, such as the power level for
A system of space coordinates often employed for the

two-dimensional pattern of an antenna is shown in Figure
3. In such a coordinate system, the radiation amplitude
is determined on the surface of a sphere whose center is
located at the antenna being tested. Since the distance
( R ) to the antenna is invariant, only the two angular
coordinates are variables in a given set of patterns. As
20 R. M. Fano, "Theoretical Limitations on the Broadband Matching of Arbitrary Impedances", Journal of the Franklin Institute, 249, pp. 57-154; 1950.
21 J. T. Bolljahn and R. F. Reese, "Electrically Small Antennas
and the Low-Frequency Aircraft Problem*', I R E Trans, on
Antennas and Propagation, vol. AP-1, pp. 46-54; October 1953.
18 N. Marcuvitz, "Waveguide Handbook", McGraw-Hill, M.I.T.

Radiation Laboratory Series, vol. 10, pp. 18-23; 1951.
19 G. Sinclair, "Theory of Models of Electromagnetic Systems",
Proc. IRE, vol. 36, pp. 1364-1370; November 1948.
10
discussed in Section 1.6, the angular coordinates are the

only significant ones in the far-field region of an antenna,
where radiation patterns are determined whenever possible.
contour represents a particular amplitude of antenna response, and a set of orthogonal coordinates on the chart
represents the two angular coordinates of the pattern.
This form of pattern permits a large volume of data to be
presented on a single sheet of paper.
Two angular coordinates which may be employed are

shown in Figure 3 ; these are the phi (<*>) coordinate and
the theta (0) coordinate. The direction of the polar axis
(0 = 0) of this spherical-coordinate system is chosen to
be appropriate to the particular geometry of the antenna
system being measured. For a fixed ground-based antenna the polar axis is usually vertical; in this case phi
is the azimuthal angle with respect to some convenient
reference direction, and theta, sometimes called the colatitude angle, is the complement of the elevation angle.
For a rotatable pencil-beam antenna, the coordinate system rotates with the antenna under test, and the polar
axis is sometimes specified in the direction of maximum
radiation so that symmetry can be exploited to the greatest advantage. More often, however, measurements on
such antennas are made with the polar axis horizontal
and orthogonal to the beam maximum; typically, rotation
of the antenna in azimuth would correspond to a variation
of theta while an elevation rotation would correspond to a
phi variation (Figure 6 illustrates such a measurement).
Measurements of amplitude patterns are generally made

for a particular polarization. This polarization, whether
linear, circular, or elliptical, (see Section 1.5) should be
that which the antenna is intended to radiate. When the
antenna response for cross polarisation is significant, the
amplitude patterns in this orthogonal polarization should
be measured as well. In the case of linear polarization,
it should be noted that the polarization direction is dependent on the angular coordinates of the pattern; for
example, the pair of orthogonal linear polarizations shown
in Figure 3 have electric field vectors E~ and E$ whose
directions depend on theta and phi.
The technique of amplitude-pattern measurements will
be discussed under two categories: on-site measurements
and off-site measurements. The on-site case (Section 2.2)
involves an antenna located on a site which may play a
significant part in pattern formation, although not in wave
propagation. T h e antennas falling in this category may
operate at any frequency, but are usually not found at the
extreme ends of the radio spectrum. The off-site case
(Section 2.3) covers those antennas whose patterns are
substantially independent of their surroundings. Such
antennas are more abundant at the microwave frequencies,
and are typified by the aperture-type highly-directive reflector antenna.
2.2 On-Site Measurements
The measurement of the complete two-dimensional
radiation patterns of a full-scale antenna located on its
ultimate site is a laborious and expensive task. Nevertheless, it is sometimes necessary to make such measurements. This situation occurs when the antenna radiation
is significantly affected by the site on which it is located,
or when construction of the antenna is practical only at
full scale on the site. Furthermore, on-site measurements
of an- antenna may have additional usefulness as a conclusive demonstration that the antenna is constructed
properly, is correctly excited, and interacts with its environment in the predicted manner.
Figure 3
A wide variety of techniques have been employed for

the on-site measurements of the amplitude patterns of an
antenna. The procedures outlined in the following discussion are commonly-used ones, and serve to illustrate
the problems involved. It should be emphasized, however, that each on-site antenna measurement has its special considerations requiring techniques appropriate to the
particular case.
System of spherical coordinate's

The amplitude patterns of an antenna may be plotted
in terms of voltage (electric field strength), although
power is sometimes preferred. An alternate form employs
decibels; this is particularly useful for presenting patterns
with low minor lobes. When decibels are used, or when
the major lobe is narrow, the patterns are generally
plotted on rectangular charts. When the major lobe is
broad, voltage or power is generally plotted on polar
charts. Most antennas require the measurement of at least
two patterns in orthogonal directions through the pattern
maximum; often many more are needed to sufficiently determine the antenna performance. Occasionally the amplitude patterns are presented as contour plots in which each
Figure 4 indicates the essential parts of the typical

measurement system. A distant source is carried by a
vehicle which is maneuvered through the space surrounding the antenna to provide a series of plane waves incident
from all the directions of interest. T h e direction to the
source is obtained from a tracking device, and this information serves to position the chart of a recording device.
The response of the antenna, which is operated in recep11
In the lower-frequency class ( H F and V H F ) , the

selection of a source antenna depends on polarization. For
horizontal polarization, a light sleeve-dipole antenna
trailed behind the aircraft has proven satisfactory. This
type can be readily fabricated from a length of standard
coaxial cable by removing the shielding braid for a distance of a quarter-wavelength, and it can be supported in
a horizontal position by a miniature parachute. For vertical polarization, useful results have been obtained with a
cage-type monopole where the framework of the aircraft
serves as the "ground." Since this antenna system is
essentially an electrically-short dipole, its radiation pattern
has a null in the vertical direction; if measurements are
made at appreciable elevation angles, the non-uniform
nature of the source antenna pattern should be recognized.
tion, controls the position of the pen of the recorder. This

data may then be processed to present the antenna patterns
in the desired form.
TO RECORDER
In the higher-frequency class (microwave region),

wing-tip antennas have been successfully utilized. This
placement minimizes excitation of the airplane structure
and provides a source radiation pattern which is as close
as possible to the pattern of the isolated source antenna.
TO RECORDER
As indicated in Figure 4, a tracking device is required

to establish the direction of the source. This process defines the direction of the aircraft with respect to the
tracker; the direction of the aircraft relative to the antenna under test must then be determined by correcting
the parallax error introduced by the known separation between the antenna under test and the tracker. In addition
to source direction, it is often necessary for the tracking
system to determine range to the source. This information
may be needed in the parallax correction, and, if the aircraft does not fly in a perfect circle about the antenna to
be measured, may also be needed for an inverse-distance
correction to the received signal level.
Figure 4
System for on-site measurements of
amplitude patterns
To carry the source, airborne vehicles, such as conventional airplanes, helicopters, blimps, and free and captive
balloons, have been employed on various occasions. As
discussed in Section 1.6, the source should be in the farfield region of the antenna system being measured. When
the far-field distance is greater than the maximum height
obtainable with airborne vehicles, man-made earth-orbiting satellites 22 have some usefulness. Even the sun 2 3 and
radio stars have been considered since they provide a
natural source. Of all these possible vehicles, light airplanes have proven most generally suitable.
Two types of tracking instruments are in common use:

optical trackers and radar trackers. An important distinction between the two is that the ordinary optical
tracker furnishes only the direction of the source, while
radar furnishes range as well. To determine range when
an optical system is employed, aircraft altitude may be
measured by an instrument in the aircraft; this information must then be transmitted to the test site so that
range can be calculated.
For measuring the on-site amplitude patterns of an
antenna, the system described in the preceding discussion
may be adequate. However, there are occasions when this
system is not sufficiently accurate, because the amplitude
or polarization of the wave radiated by the source toward
the antenna under test is too variable; this is particularly
likely at microwave frequencies. In these cases, a reference antenna must be introduced in the system, as indicated in Figure 4. This reference antenna should be
placed close enough to the antenna under test so that the
effect of variations in the strength of the wave from the
source can be substantially eliminated by normalizing the
signal received by the antenna under test with respect to
the signal received by the reference antenna. In addition
to the shape of the amplitude pattern, the reference antenna also may permit a measurement of power gain, as
discussed in Section 4.2. Finally, the reference antenna
can sometimes be designed to determine the polarization
If the attitude of the source antenna relative to the

antenna under test changes, a change in the received
signal is likely to occur. T o minimize this, the source antenna should be oriented so that its peak is in the direction
of the antenna being measured, and the useful portion of
the source pattern should be as uniform as possible. In
addition, the course flown by the aircraft should be chosen
to minimize changes in attitude; a favorable course for
this purpose lies along a circle centered on the antenna
being measured and contained in a plane perpendicular
to a vertical axis through the antenna. Since the radiation pattern of the source antenna is subject to modification by the aircraft on which it is carried, the design and
placement of the source antenna must include the effects
of this environment. This factor is greatly dependent
on the relative sizes of the operating wavelength and the
aircraft; the source antenna may be separated into two
classes according to this distinction.
22 H. Bruekman, "Antenna Pattern Measurement by Satellite",
IEEE Transactions on Antennas and Propagation", vol. AP-11,
pp. 143-148; March 1963.
23 J. T. Kennedy and J. W. Rosson, "The Use of Solar Radio
Emission for the Measurement of Radar Angle Errors", Bell
Sys. Tech. J., vol. 41, No. 6, pp. 1799-1812; November 1962.
12
of a set of source antennas, as discussed in Section 3.2;

this in turn may permit the polarization of the antenna
under test to be determined. It should be recognized, however, that unless the reference antenna system is carefully
designed, it may introduce errors as great as those it is
intended to cancel.
At microwave frequencies, it is desirable to make the
response of the reference antenna substantially independent of the ground and surrounding structures. This dictates the use of a reference antenna having a narrow
beam and low minor lobes, and may also require treatment of the ground and other structures near the reference
antenna with absorbing material or special fences to minimize reflections. The narrow beam in turn imposes a
pointing requirement on the reference antenna which is
usually satisfied by slaving the reference antenna to the
tracking device. At frequencies well below the microwave
range, the pattern of the reference antenna may be so
broad that ground reflections must be accepted as part of
the mechanism of pattern formation of the reference antenna, and the reference antenna may be practical only
when it is fixed in position rather than pointed at the
distance source. If the pattern of the reference antenna
system can be accurately determined, then it is still possible to employ its output as a means for correcting variations in the wave radiated by the source during measurement of the pattern of the antenna under test. However
at the lower frequencies such as H F , it is customary to
use the reference antenna only during the measurement of
power gain at one angle, as described at the end of Section
4.2. Regardless of the frequency being employed, the reference antenna system should be designed and located so
that its effect on the patterns of the antenna under test is
negligible.
The process of pattern measurement and recording may
involve either a point-by-point or a continuous method;
the latter is preferable. Commercially-available continuously-recording equipment can often be adapted to introduce automatically the various corrections which are
needed in both the signal and angle inputs. The original
data is recorded in a form dependent on the particular
measurement geometry; when the aircraft flies in circles
centered on the antenna under test, the patterns are plotted
in the form of antenna response vs. azimuth direction for
a series of elevation directions. The final antenna patterns
can be presented in different forms; the most comprehensive form is a contour plot, as mentioned in Section 2.1.
2.3 Off-Site Measurements
When the antenna pattern is essentially independent of
the enviroment, a simple measurement technique can be
used 24 ' 2 5 2 6 . This consists of operating the antenna under
test as a receiving antenna, rotating it, and measuring the
24 C. C. Cutler, A. P. King and W. E. Kock, "Microwave Antenna Measurements", Proc. IRE, vol. 35, pp. 1462-1471;
December 1947.
Radiation Laboratory Series, vol. 12, McGraw-Hill, Chaps.
15-16; 1949.
26 John D. Kraus, "Antennas", McGraw-Hill, Chap. 15; 1950.
signal it receives from a suitably-located fixed source

antenna. The method can be directly applied to most
microwave antennas and has been used with success at
any frequency which does not lead to impossible mechanical restrictions; it is also most suitable for scale-model
measurements.
This type of measurement tends to be divided into two
classes, one appropriate to antennas having electrically
large apertures and the other appropriate to antennas
having dimensions of the order of a wavelength or less.
For those antennas with dimensions much less than a
wavelength, the far-field region starts near X/2TT (see Section 1.6). This is such a short distance that the practical
consideration of producing a uniform field over the region
of rotary motion of the test antenna leads one well into the
desired far-field region for the placement of the source
antenna. Even for antennas having dimensions of the
order of a wavelength, the far-field region is usually conveniently close to the test antenna. In addition to convenience, it is desirable to have a short distance between a
small test antenna and the source antenna in order to
minimize the spurious power that may enter the test antenna by wide-angle scattering from obstructions in the
test range. Because antennas having dimensions of the
order of a wavelength or less have wide radiation patterns,
the rotating mechanism should provide 360 rotation of
the test antenna, and may have relatively low precision.
An example of a test range for such antennas is indicated
in Figure 5.
For antennas having electrically large apertures, the
2D 2 /X criterion for the far-field region (see Section 1.6)
becomes most important. The distance between the source
antenna and the antenna under test should equal or exceed
this value. If, as is usually the case, the maximum aperture dimension of the source antenna is equal to or less
than that of the test antenna, then the 2D 2 /X criterion also
fairly well ensures that the essential variation in amplitude
of the incident field across the aperture of the test antenna
is small enough to have a negligible effect. Thus, the
measured pattern taken at at least a 2D 2 /X separation differs from the far-field pattern by amounts usually within the
limits of instrumental error. If necessary, useful measurements can be. made with the separation between source
and receiving antennas as little as D 2 /X or even less.
However, the disparity 27 between the measured pattern
and the far-field pattern becomes more marked as the
separation decreases, and measurement results rapidly become less meaningful. With an electrically large aperture,
the radiation pattern is narrow and the antenna may be
physically large and heavy; the rotating mechanism
should therefore be a rugged platform of relatively high
angular precision. Provision should also be made for precise alignment of the antenna in the coordinate orthogonal
to the one being varied for the pattern measurement. An
example of a test range for antennas with electrically
large apertures is indicated in Figure 6.
27 E. V. Jull, "An Investigation of Near-Field Radiation Patterns
Measured with Large Antennas", IRE Trans, on Antennas and
Propagation, vol. AP-10, No. 4, pp. 363-369; July 1962.
13
ANTENNA
UNDER TEST
TRANSMITTER
TO RECORDER
Figure 5
Off-site pattern range for antennas having dimensions of a wavelength or less
SOURCE
ANTENNA
TRANSMITTER
TO RECORDI
ANTENNA
UNDER TEST
Figure 6
Off-site pattern range for antennas having electrically large apertures
14
With either large or small antennas, measurements of

patterns with low sidelobes require a test range which is
free of reflections with amplitudes greater than a small
fraction of the sidelobe level to be measured. Techniques
for minimizing undesired reflection effects include the
following:
(a) Mounting the source and test antenna on towers or
adjacent hills or buildings. This arrangement reduces the
ground reflections since the radiation directed downward
is in the sidelobe region of the test and source antenna
and this radiation propagates along a greater path length
between the two antennas than does the direct radiation.
(b) Employing a source antenna with maximum permissible directivity, consistent with maintaining a sufficiently uniform illumination across the aperture of the
receiving antenna.
(c) Designing the source antenna so that the ground
area causing reflections lies in the null region of the
source. This technique is sometimes frequency sensitive,
however, and is often incompatible with ( b ) .
(d) Utilizing diffraction fences or absorbing baffles to
either scatter or absorb the reflected radiation. A series
of such fences may be required to reduce the amplitude
of the interference pattern intercepted by the receiving
aperture to an acceptable value for the measurement precision required.
A less common alternative to minimizing ground reflections is to smooth the intervening ground between the
source and test antenna to produce specular reflections 28> 29 . The source and test antenna heights are then
adjusted in conjunction with the antenna spacing to yield
an interference pattern maximum in the vicinity of the
test antenna. If the test antenna is small relative to the
period of the interference pattern, it will intercept a quasiuniform plane wave. N o matter which of the above
methods is employed for reducing the harmful effects of
ground reflection, it should be considered standard procedure to probe the field across the aperture of the antenna under test with a small sampling antenna. By this
means the uniformity of the field can be determined directly, and the expected accuracy of the antenna pattern
measurement can be estimated.
a convergent spherical wavefront rather than a plane wave

across the antenna aperture. With phased-array antennas,
the phasing of the elements can in general be adjusted to
produce faithfully the far-field pattern distribution in the
near-field focal region. For the special case of arrays
requiring in-phase excitation, it may sometimes be convenient to physically adjust the radiating elements on a
concave spherical arc centered or focused at a usable
distance. 30 Even on a paraboloidal reflector, a moderate
motion of the primary feed away from the focus along
the reflector axis will produce a nearly perfect concave
spherical wavefront on the aperture. T o a good approximation the far-field radiation pattern characteristics
(major lobe and near-in minor lobes) of a large aperture
antenna as a function of angle remain constant as the
antenna is refocused at a finite distance provided that this
distance is very large compared with the antenna aperture.
In general, this equivalence does not apply to minor lobes
far from the major lobe.
Microwave patterns of antennas which are not physically large can also be measured in an anechoic chamber
or "microwave darkroom." 3 1 Such chambers are lined
with absorbing walls or baffles which greatly reduce the
reflected radiation, and may be designed so that reflected
rays can travel from one antenna to the other only by
multiple bounces. Anechoic chambers are generally valuable for design studies, but are usually not suitable for
accurate measurements of low minor lobes. The frequency
response of the absorbing material, reflection characteristics of walls, ceiling, and floor, and the influence on these
properties of humidity, temperature and traffic must be
considered prior to use.
3. M E A S U R E M E N T O F O T H E R
CHARACTERISTICS
3.1
PATTERN
Phase Measurement
The phase characteristic of an antenna radiation pattern

often provides important information for antenna design
and use. 32 - 3 3 In the far-field region of the antenna, the
variation of phase with pattern angle can be indicative of
antenna defocus; the variation in relative phase between
two patterns of a tracking antenna can also be a significant
quantity. In the near-field region or at the aperture of an
antenna, a knowledge of phase and amplitude as a function
of location around the antenna can permit accurate prediction of the far-field patterns. The phase center of an
antenna feed, which must be accurately located in a
For very large antennas having aperture dimensions

many times greater than the wavelength, the inner
boundary of the far-field region may extend too far to
allow a practical measurement to be made by the methods
so far described. However, the far-field pattern, which
exists at very large distances from large-aperture antennas, can sometimes be determined to a good approximation by refocusing the antenna to a focal region within
the radiating near-fie Id region, much closer to the antenna. As in analogous optical systems, focusing at a finite
distance rather than at infinity requires the generation of
30 R. W. Bickmore, "Fraunhofer Pattern Measurements in the

Fresnel Region", Canadian Journal of Physics, vol. 35, p.
1299; 1957.
31 A. J. Simmons and W. H. Emerson, "An Anechoic Chamber
Making use of a New Broadband Absorbing Material", IRE
Convention Record, Part I I , pp. 34-41; 1953.
32 C. C. Cutler, A. P . King, and W. E. Kock, "Microwave Antenna Measurements", Proc. I R E , vol. 35, pp. 1462-1471;
December 1947.
Radiation Laboratory Series, vol. 12, McGraw-Hill, pp. 564573; 1949.
28 W. A. Cumming, "Radiation Measurements at Radio FrequenciesA Survey of Current Techniques", Proc. IRE, vol. 47,
pp. 705-735 ; May 1959.
29 A. Cohen and A. W. Maltese, "The Lincoln Laboratory Antenna Test Range", Microwave Journal, vol. 4, No. 4, pp.
57-65; April 1961.
15
focusing antenna, may be determined by a measurement

of the phase of the feed radiation pattern as a function
of angle or displacement. Phase measurements may also
be useful in determining distortions caused by radomes,
lens steps, and the like.
reference and test signal paths, in view of the presence

of a movable joint or flexible cables in either or both lines,
and the avoidance of distortion of the radiated field by
reflections from the sampling probe, the reference antenna, or any support structures.
The measurement of far-field phase is usually made at a

constant distance (radius) from the antenna under test; it
may be thought of as being measured over a spherical
surface, where the antenna under test is located at the
center of the sphere. By analogy with the more common
amplitude radiation pattern, this form of measurement of
phase information is referred to as the phase pattern of
the antenna. The measurement of near-field phase requires
more attention to the definition and control of the measurement parameters because of the very rapid and often
unpredictable phase variations in this region. If the
location of interest is known, such as a planar surface
parallel to an antenna array, or a line parallel to a line
source, the phase sampling should be done at that location.
If constant-phase contours in a certain region are to be
determined, several "slices" can be taken at convenient increments of distance; in the case of measurements on
spherical surfaces, the location of the actual center of
rotation of the patterns should be known with respect to
the antenna itself. The contours of constant phase may be
determined by interpolation of the measured values of
phase. It may even at times be possible to measure a
constant-phase contour directly, by means of suitablydesigned apparatus.
ANTENNA
UNDER TEST
-PROBE
:x:
&
PHASE
MEAS.
CIRCUIT
FLEXIBLE
CABLE
(a) Near-field, or short distance, pattern phase

_ FIXED
ANTENNA
(b) Far-field pattern phase
PHASE
MEAS.
CIRCUIT
*T
TEST
ROTARY _ _ ^ * C 3
MOUNT ~~^
All of these measurements are made in a similar

fashion. Since phase is a relative quantity, a reference
signal must be provided at all times for comparison. For
measurements made at short distances, the antenna under
test may be used as the transmitting antenna and a simple
receiving antenna or probe used to sample the radiated
field, as shown in Figure 7 ( a ) . The reference signal is
coupled out from the transmitter line, and compared with
the received signal in a suitable bridge circuit. For measurements at distances too long to permit direct connection
of the reference and sampled signal, the arrangement of
Figure 7(b) may be used wherein the signal from a distant source is received simultaneously by the antenna under test and by a fixed reference antenna; the antenna under test is rotatable in the usual manner for measuring
radiation patterns. The distant source is also useful in the
arrangement of Figure 7 ( c ) for the measurement of relative phase between two ports of a multi-port antenna as
a function of angle.
ANTENNA
UNDER TEST
DISTANT
SOURCE
(c) Far-field relative phase between two patterns

Figure 7
Arrangements for measuring phase patterns
Figure 8 shows a very simple phase measuring circuit
which may be used in any of the arrangements described
above. The hybrid junction is a microwave four-port
component analogous to a bridge circuit. When equal
signals are fed into conjugate arms of the hybrid, the
signal emerging from each of the other ports is a function
of the relative phase between the two inputs. By properly
adjusting the phase of one of the input signals, a null can
be produced at one of the output ports.
VARIABLE
REFERENCE
(SEE
A number of techniques are described in the literature 3 4 ' 35 3 6 for performing the actual phase comparison
measurement; a typical method is described below. Among
the necessary precautions in phase-pattern measurements
are the preservation of constant phase lengths in the
DETECTOR
HYBRID
JUNCTION
SIGNAL
NOTE)
>
>
CALIBRATED
SHIFTER
34 H. Jasik, "Antenna Enginering Handbook", McGraw-Hill pp.

34-26 thru 34-28; 1961.
35 M. Wind, "Handbook of Electronic Measurements", Polytechnic Institute of Brooklyn, vol. II, Chapter 7, pp. 7-37, 7-38;
1956.
36 R. A. Sparks, "Microwave Phase Measurements", MicroWaves, pp. 14-21; Jan. 1963.
NOTE:
T E S T SIGNAL
TRANSMISSION PHASE OF ATTENUATOR

TO BE CONSTANT, OR CALIBRATED.
Figure 8
Phase measuring circuit
16
The test signal, which is assumed to be weaker than

the reference signal, is brought through a low-loss calibrated phase-shifter to one port of the hybrid junction.
The reference signal passes through a variable attenuator
having constant or calibrated phase shift, and enters the
conjugate port of the hybrid. (If the test signal is stronger than the reference signal, as is likely on the major lobe
of a highly directive antenna, then either the variable
attenuator or a fixed attenuator may be placed in the testsignal branch.) A detector and a termination are connected to the remaining ports of the junction. For accurate results, all reflections should be minimized. To
make a measurement, the phase shifter and attenuator are
varied until a null is observed at the detector; the phaseshifter setting, corrected for the attenuator phase if necessary, is noted. In this manner, a point-by-point phase pattern may be obtained as the antenna position is varied and
the circuit rebalanced.
Automatic recording of antenna phase patterns may be
accomplished in a number of ways by more elaborate
phase-measuring circuits. 37 Since the test signal will
usually vary widely in amplitude with respect to the
reference signal, either an amplitude-insensitive method of
phase sensing (such as a quadrature null detector), or
amplification and limiting preceding the phase discriminator must be used. Circuits having a quadrature-signal
null detector usually use it to servo-control an adjustable
phase shifter in the reference signal channel; the phaseshifter setting is the recorded indication of phase.38 In
such circuits, it is sometimes convenient to apply audio
modulation to only one of the comparison channels before
combination in the null detector. The "homodyne" technique39' 4 0 generates suppressed-carrier sideband signals
in one channel; the "subcarrier" technique41 employs
simple amplitude modulation. When the test and reference
signals are amplified and limited so that their amplitudes
remain relatively constant, conventional phase-discriminator circuits may be used to produce an output signal
proportional to the relative phase. 42 The principal problems are ambiguity and non-linearity in portions of the
output characteristic. Another means of deriving an output signal proportional to the relative phase of two limited
signals is the "coincident slicer" which reads the average
time during which both limited signals have the same
polarity.43 Although fairly insensitive to amplitude fluctuations, this circuit also is subject to problems of ambiguity and non-linearity.
3.2 Polarization Measurements

The radiation pattern of an antenna is not completely
determined unless, in addition to amplitude and phase,
polarization is also measured for all directions of interest.
The nominal polarization of the antenna is expected to
predominate near the peak of the major lobe, but accurate
determination of the cross-polarisation in this direction is
often essential. In other directions, the polarization usually
departs appreciably from the nominal value, and it is important in many cases to obtain comprehensive information
about this polarization variation.
In any specified direction, the polarization of an antenna
can be measured by sensing the relative amplitude and
phase of two orthogonal components (see Section 1.5) of
its radiation field. This is usually done in the far-field
region of the antenna, where the electric-field vector has
no significant radial component, so that it may be sampled
two-dimensionally. Two orthogonally-polarized but otherwise identical test antennas, such as crossed linear or
opposite-handed circular, may be used, and their corresponding responses compared in amplitude and phase to
determine 44 the unknown polarization. If it is not feasible
to make phase measurements, measurements of amplitude
response with four test antennas having certain known
polarizations can permit a complete determination45 of
polarization by a graphical computation.
An incomplete, but nevertheless informative, measurement of polarization is often made by observing the amplitude of response of a single rotating test dipole or other
linearly-polarized antenna. If the observed response is
plotted as a function of the rotation angle *r, a dumbbellshaped curve may be obtained, as indicated in Figure 9.
When sampling a linearly-polarized field in this manner,
the plot has two deep minima and follows a cosine variation
in polar coordinates. When sampling a circularly-polarized field, a circular plot is produced. For the general
case of an elliptically-polarized field, shown in Figure 9,
the axial ratio of the polarization ellipse is given by:
(3)
where Emax and Emin are the maximum and minimum

electric fields sampled by the rotating dipole. The orientation angle of the major axis of the polarization ellipse is
the angle of maximum response with the rotating dipole.
Thus the two main parameters of the polarization ellipse
are determined by this simple measurement. The handed^
ness of the elliptical polarization, however, is not determined unless the phase of the signal from the rotating dipole is also measured. However, if the antenna under test
is nominally circularly-polarized with a known handedness, it is usually safe to assume that the measured elliptical polarization has the same handedness.
37 W. A. dimming, "Radiation Measurements at Radio Frequencies, a Survey of Current Techniques", Proc. IRE, vol. 47,
pp. 705-735; May 1959.
38 W. F . Gabriel, "An Automatic Impedance Recorder for
X-Band", Proc. IRE, vol. 42, pp. 1410-1421; Sept. 1954.
39 F. L. Vernoh, Jr., "Application of the Microwave Homodyne",
Transactions PGAP, pp. 110-115; Dec. 1952.
40 S. D. Robertson, "A Method of Measuring Phase at Microwave
Frequencies", Bell System Tech. Jnl., pp. 99-103, January 1949.
41 G. E. Schafer, "A Modulated Subcarrier Technique of Measuring Microwave Phase Shifts", Transactions PGI, pp. 217-219,
September 1960.
42 R. M. Barrett and M. H. Barnes, "Automatic Antenna WaveFront Plotter", Electronics, pp. 120-125; January 1952.
43 Y. P. Yu, "Coincident Slicer Measures Phase Directly", Electronics, pp. 99-101; Sept. 12, 1958.
44 H. Jasik, "Antenna Etigineering Handbook", McGraw-Hill, pp.

17-1 to 17-8; 1961.
45 G. A. Deschamps, "Part IIGeometrical Representation of the
Polarization of a Plane Electromagnetic Wave", Proc. IRE, vol.
39, pp. 540-544; May 1951.
17
E max
MEASURED RESPONSE
(POLARIZATION PLOT)
electrical boresight with such antennas can be better than

about 1/100 the beamwidth of an individual lobe.
Conical scanning antennas use a single beam which
traverses a conical scan pattern which is usually circular;
Figure 10(a) represents a cross-section through the pattern in any plane. The response to an off-axis target
varies with beam position, so that an audio error signal
with a fundamental frequency equal to the scan rate is
generated. The strength of this audio signal, as a percentage modulation of the rf carrier, is proportional to the
target angular displacement from electrical boresight, and
the phase of the audio signal is an indication of the direction of the angular displacement.
LOBE 2,OR LOBE I

AT A LATER TIME
LOCUS OF TIP OF E VECTOR vs. TIME

(POLARIZATION ELLIPSE)
Figure 9
Relative amplitude of elliptically-polarized field
as detected by a rotating linearly-polarized antenna
(a) Lobe patterns, monopulse or conical scanning
3.3 Special Measurements for Boresight and

Angular Sensitivity
Directive antennas often require precise determination
of the direction of the beam or the tracking axis, based
on an electrical indication from the antenna system; such
a direction is called the electrical boresight. This direction is determined with respect to some reference direction, called the reference boresight, which may be a specified stationary direction, or may be obtained from the
antenna by a physical indication such as an optical or
mechanical axis, or by a prior electrical indication.
Measurements of boresight error are concerned with the
angular deviation of the electrical boresight of an antenna
from its reference boresight. Based on such measurement,
it may be necessary to adjust the antenna system to minimize boresight error, or to place the electrical boresight
in alignment or perpendicularity with mechanical axes or
other physical reference.
LOBE I + LOBE 2 (SUM)
LOBE I-LOBE 2
(DIFFERENCE)
DIFFERENCE- PATTERN
MINIMUM OR NULL
ANGLE
(b) Monopulse sum and difference patterns

Figure 10
Signals received by tracking antenna
versus angle of target
Monopulse tracking antennas, either amplitude comparison or phase comparison, commonly have separate
ports for three responses: the sum, the azimuth difference,
and the elevation difference. The sum (or range) pattern
has a single major lobe in the boresight direction. The
two orthogonal difference patterns (elevation and azimuth) each have a minimum response in the boresight
direction, and amplitude proportional to angle off axis
over a small central range. Figure 10(b) shows the sum
and a difference pattern in one plane; comparison with
( a ) indicates that they may be considered as the sum and
difference of the overlapping lobes. In general, an offaxis target generates a difference signal in each difference
channel; these signals, when referenced or normalized to
the sum signal, provide error signals whose amplitudes are
proportional to the target displacement from electrical
The beam direction of an antenna with a single major

lobe is usually determined by noting the direction of maximum response, or the direction halfway between equal
responses either side of the peak. The precision of such
determination is usually in the order of 1/10 the halfpower beamwidth.
Much greater precision of direction is demanded of
tracking antennas, where two or more overlapping beams
or lobes may be provided as indicated in Figure 10(a).
These lobes are compared to indicate a common direction
by equality of amplitude or phase. When the comparison
is done sequentially it is termed sequential lobing and
typically compares amplitude; two examples of this class
are conical scanning and lobe switching. When the comparison is done simultaneously it is called simultaneous
lobing or monopulse; in this case either amplitude or phase
may be compared 46 . The precision of determination of
46 D. R. Rhodes, "Introduction to Monopulse", McGraw-Hill;

1959.
18
be defined by equality of the lobe signals ("crossover" in

Figure 10(a) ) ; in a monopulse antenna, the differencesignal minimum (see Figure 10(b) ) would define the
"antenna" boresight. For this latter case, the additional
boresight errors contributed by interaction between the
antenna and the associated circuits can be determined from
additional measurements and a simple graphical computation. 47 As regards angle sensitivity, the information
usually desired is the slope of the lobe-signal patterns at
crossover, or the asymptotic slope of the difference-signal
patterns. During measurement, these slopes should be
referred to an absolute level of signal voltage, such as
that which would be produced by an isotropic radiator
or some other appropriate standard, 48 in order to permit
a significant evaluation of the antenna design and the
system signal-to-noise ratio.
boresight in the appropriate azimuth or elevation plane,

and whose phases indicate the sense of displacement.
The test setup for determining the electrical boresight
of a tracking antenna comprises the antenna under test,
the proper associated circuits for performing the appropriate signal processing to obtain an error signal, a distant
source, a precise means for continuously adjusting antenna
direction (or source location) and a highly accurate optical or mechanical indication of antenna direction (or
source location). The antenna (or source) is adjusted
for the condition of minimum output observed at the
appropriate error-signal port of the circuit. This antenna
direction (or source location) is then noted as the electrical boresight under the particular condition of frequency,
environment, or other variable undej test; the direction
may be compared with a reference direction to determine a
boresight error. If the antenna system output cannot be
noted continuously as a function of source angle, the minimum point may be interpolated. Since the absolute direction of the electrical boresight during test is often of less
interest than its variation with system parameters, the
chief requirement of the antenna direction-indicating
mechanism may be high precision, over a very small
angular range. A telescope rigidly fixed to the antenna
mount may be used to sight on a calibrated optical target
at the distant source, or a dial indicator on a long radius
arm may be used. A camera may be similarly employed
to measure dynamic errors between electrical and mechanical axes when tracking a moving target.
4. M E A S U R E M E N T O F M A X I M U M
GAIN A N D D I R E C T I V I T Y
4.1
POWER
General
The power gain of an antenna in a specified direction is

4TT times the ratio of the power radiated per unit solid
angle in that direction to the net power accepted by the
antenna from its generator. This term is an inherent property of the antenna, and does not involve system losses
arising from mismatch of impedance or of polarization.
T o determine power transfer in a complete system, the
antenna input impedance (see Section 7.1) and the antenna polarization (see Section 3.2) must be measured
and compared with the impedance and polarization of the
other appropriate components of the system.
Another measurement typical of tracking antennas is

that of the appropriate output error signal (percent modulation, or normalized difference signal) as a function of
angular displacement from electrical boresight. The test
setup described above may be used, with particular attention to any circuit properties affecting the output level,
such as automatic gain controls or limiters. The angular
sensitivity of the antenna is measured as the rate of change
of output error signal with angle in the vicinity of electrical boresight, and is related to the slopes of the patterns
indicated in Figure 10. For conical-scanning antennas,
this property is sometimes termed "modulation slope";
for monopulse antennas, it may be called "error slope."
For antennas in which the error-signal magnitude is used
to determine angular location of an off-axis target, the
linearity of error-signal output with angle is of interest,
as well as the range over which the linearity holds. Another property of interest is the region of angular pull-in,
within which the error-sensing circuits will generate an
error voltage acting to reduce the angular displacement
from electrical boresight.
The directive gain of an antenna in a specified direction

is Air times the ratio of the power radiated per unit solid
angle in that direction to the total power radiated by the
antenna. This term differs from power gain because it
does not include antenna dissipation losses. The ratio of
power gain to directive gain in the same direction is the
radiation efficiency (see Section 5) of the antenna.
In both of the above definitions of gain, the standard
for gain comparison is a hypothetical radiator called the
isotropic radiator, which radiates equally in all directions,
is lossless, and has a gain of unity.
The gain of an antenna is usually measured in the direction of its maximum value. Knowledge of the radiation
pattern amplitude (see Section 2) then permits determination of the gain in any other direction. In the direction of the maximum value, power gain is termed "maximum power gain" and directive gain is termed directivity.
These maximum values are often referred to merely as
the "antenna gain," but this usage is deprecated because
it is not definitive.
The measurements described above for boresight and

angular sensitivity involve the error signal, as derived
from the complete system including the circuits associated
with the antenna. While such measurements yield the
true performance of the system, some significant contributions of the antenna to system performance may not be
evident. For this reason, the signals obtained directly
from the antenna are sometimes used to determine an
electrical boresight and an angular sensitivity. In a
conical-scan antenna, such an "antenna" boresight would
47 H. W. Redlien, "The Monopulse Difference Chart", IEEE International Convention Record, Part I, Antennas and Propagation, pp. 129-131; 1963.
48 R. R. Kinsey, "Monopulse Difference Slope and Gain Standards", IRE Trans, on Antennas and Propagation, vol. AP-10,
pp. 343-344; May 1962.
19
4.2
Measurement of Maximum P o w e r Gain
fundamental method, and this can be done as follows when

two identical antennas are available.51 One antenna is
connected to a transmitter and the other to a receiver;
both antennas are impedance-matched to their respective
loads. The two antennas are placed a distance R apart
(where R is beyond the inner boundary of the far-field
region), and are aligned for maximum power transfer.
The maximum power gain is then:
(5)
In practice, the maximum power gain of an antenna is

almost always determined from a measurement of relative
gain involving a standard-gain antenna. The power received by the antenna under test is compared with the
power received by the standard-gain antenna when both
antennas are placed in a uniform plane-wave field and
oriented for maximum received power. The maximum
power gain of the antenna under test is determined from
the known maximum power gain of the standard-gain
antenna as follows:
(4)
Pr (test ant.)
Gmax p
/gt(j
a n t
Gmax (std. ant.) =
X~W~
where Pr and Pt are the received and transmitted power,

respectively. The ratio Pr/Pt must be measured; this
may be done by employing a calibrated receiver which is
transferred from the receiving antenna to the transmitter,
and rematched.
s X Gmax ( s t d . a n t . )
where Pr is proportional to the received power in each

case. It is common practice to express equation (4) in
decibels. Gmax in decibels then becomes the sum of two
terms, the first term being the relative antenna gain in
decibels and the second term being the gain of the standard antenna above isotropic in decibels. The sum thus
yields the gain of the test antenna above isotropic in
decibels.
Various sources of error must be considered when attempting to achieve accuracy in the measurement of maximum power gain. For example, at microwave frequencies
the antenna under test and the standard-gain antenna may
react differently to irregularities which are often present
in the field of even the most carefully designed test site.
Several techniques are available to minimize such errors;
for instance, if the standard antenna is small compared
with the antenna under test, it should be moved across the
aperture of the antenna under test, and the average power
recorded.
Two precautions must be observed in the above measurement. First, account must be*taken of the impedance
mismatch which may exist between each antenna and its
associated receiver. It is usually preferable to eliminate
such mismatches by means of tuning devices which are
adjusted to provide conjugate match in each case. Care
should be taken to account for any dissipation present in
the tuners. If the impedances cannot be matched, then
the mismatches should be separately determined from impedance measurements of each antenna (see Section 7.1)
and its associated receiver. The losses caused by such
mismatches may then be calculated (see Section 7.1), and
included in the computation of antenna gain.
At much lower frequencies, where the ground must be

regarded as an essential part of both the standard antenna
and the antenna under test, the measurement technique
depends on the polarization employed. For horizontal
polarization, if the height above ground of the antenna
under test is set for maximum power gain at a certain
elevation angle, the reference antenna may be a half-wave
dipole at the same height. In this case the gain of the
reference antenna in the direction of interest is taken as
8db, which represents the approximate sum of the 2.15
db free-space gain of a lossless half-wave dipole and the
6 db augmentation due to reflection from the assumed
perfectly-conducting earth. For an accurate measurement
of gain of a vertically-polarized low-frequency antenna,
perfect reflection from the ground cannot always be relied
on. If the ground were a perfect conductor, a reference
antenna comprising a quarter-wave monopole would have
a gain of about 5 db in the horizontal direction, representing the approximate sum of the 2.15 db free-space
gain of a lossless half-wave dipole and the 3 db augmentation due to imaging of the monopole by the assumed
perfectly-conducting earth; for directions above horizontal, this gain would be reduced according to the standard Zs-plane pattern of a half-wave dipole. 52 However
for an imperfectly-conducting ground, the gain at low
The second precaution to be observed involves the

polarizations of the antenna under test and the standardgain antenna. It is generally desirable to make these
polarizations identical in order to simplify the gain computation. Usually this situation is closely approximated
when the antennas are nominally linearly polarized and
are carefully aligned during the measurement. If the
polarizations of the two antennas are not the same, as in
the case of elliptically-polarized antennas having different
axial ratios, then it is necessary to account for the difference. In general, the polarization of both antennas should
be measured separately (see Section 3.2) and each compared with the measured polarization of the field produced by the source antenna. The losses caused by polarization mismatch may then be calculated (see Section 1.5),
and included in the computation of antenna gain.
Accurate knowledge of the maximum power gain of
the standard-gain antenna is essential to the measurement
just described. This gain can usually be calculated when
the standard antenna has negligible dissipation and a
simple configuration such as a half-wave dipole49 or a large
horn. 50 However it is desirable to measure its gain by a
49 J. D. Kraus, "Antennas", McGraw-Hill, p. 54; 1950.

50 H. Jasik, "Antenna Engineering Handbook", McGraw-Hill,
Chapter 10; 1961.
51 S. Silver, "Microwave Antenna Theory and Design", McGrawHill, M.I.T. Radiation Laboratory Series, vol. 12, Section
15.19; 1949.
20
elevation angles may be seriously reduced.52 Even with

the use of a large counterpoise, the pattern may deviate
appreciably from the ideal one because of diffraction and
reflection from the edge of the counterpoise.53
4.3 Determination of Directivity
When the complete radiation pattern of an antenna is
known or is measurable, it may be used to determine the
directivity of the antenna. The particular quality of the
pattern that is employed is the radiation intensity, which
is the power radiated from the antenna per unit solid
angle in a given direction. The directivity of an antenna
is the maximum radiation intensity divided by the average
radiation intensity. The latter quantity multipled by 4w
is the total power radiated. To compute the directivity
the following relation may be employed:
(6)
4?r X maximum radiation intensity
It is important that the radiation paterns which are

measured for determining directivity include the effect cf
cross polarisation. Radiation intensity involves not only
power radiated in the normal polarization, but crosspolarized power as well; the latter component is often
surprisingly large in directions away from the peak of the
pattern. To determine radiation intensity, two sets of
patterns may be measured using orthogonal polarizations,
such as 0 and <& linear, or right and left circular. Each
pair of patterns, plotted in terms of power, is added
together to yield the radiation intensity, which may then
be applied in formula (6) to obtain the directivity.
For estimation purposes, a rough value for directivity
is sometimes determined from a measurement of only the
major lobe, for only the normal polarization. When the
major lobe is reasonably narrow, the following relation
is employed:
^
,j^
Gd max =
BWi X BW 2
where BWi and BW2 are the half-power beamwidths in
/
degrees in the orthogonal principal planes through the
J
radiation intensity X sin 6 d0 d4>
major lobe, and k is a factor which may range54' 55 from
< P = 0 0=O
below 25,000 to about 40,000. The lower value of k is
where 6 and 4> are the spherical coordinates of the radiatypical for microwave reflector type antennas where spilltion pattern, as indicated in Figure 3.
over usually represents an appreciable fraction of power
Except for the simplest of antenna patterns, the integral
radiated into wide-angle minor lobes; the upper value
must be evaluated by graphical methods. There are two
can be approached in certain antennas having aperture
in general use: the "orange-slice" and the "conical-cut"
excitations with appreciable amplitude tapering and no
methods. In the orange-slice method a set of patterns is
spillover. It is clear that this simplified method should
obtained by measuring radiation intensity versus 0, for a
be used only with great care because unexpected antenna
number of discrete values of $. Each pattern must be
multiplied continuously by the sin $ weighting factor of defects are always likely to cause excessive minor lobes
equation (6) and then integrated. The integrated values and cross polarization, each of which may greatly reduce
for the several patterns are then added according to equa- the directivity.
tion (6). In the conical-cut method a set of patterns is
obtained by measuring radiation intensity versus o, for a 5. DETERMINATION OF RADIATION
number of discrete values of 6. Each pattern is integrated EFFICIENCY
and the integrated values are each multiplied by the apThe radiation efficiency of an antenna is the ratio of
propriate sin $ weighting factor and added.
the total power radiated by the antenna to the net power
When the radiation patterns are especially simple, there accepted by the antenna at its terminals during the
are simple graphical techniques which minimize the labor radiation process. The difference between these two powers
of integration. For example, the pattern of an omnidirec- is the power which is dissipated within the antenna.
tional antenna may be plotted on special coordinates de- Radiation efficiency is an inherent property of an antenna,
signed so that the pattern area is directly related to the and is not dependent on system factors such as impedance
total radiated power. Thus a simple operation with a match or polarization match.
planimeter will yield the directivity.
A fundamental method for determining radiation effiThe number of patterns that should be measured inciency relies on the measurements described in Section
creases as the pattern shape becomes less uniform. For
4.2 and 4.3. As noted in Section 4.1, radiation efficiency
pencil-beam antennas, it is desirable that the pole of the
is eflual to the ratio of power gain in any specified direcspherical coordinate system of measurements coincide with
tion to the directive gain in that same direction. It is
the beam so that this important part of the pattern is adeusually convenient to take the direction of maximum
quately covered; however, this is not always easy to acradiation for this determination of radiation efficiency;
complish. In addition, an antenna with a narrow major
thus:
lobe is likely to have a large number of narrow minor
radiation efficiency = maximum power gain
(8)
lobes which must be included in the measurement and
directivity
integration. Generally, it is practical to determine directivity accurately only on those antennas having radiation In measuring maximum power gain and directivity, all
the precautions mentioned in Sections 4.2 and 4.3 must
patterns which are not highly directive.
be carefully observed. It is particularly important to inGd max
o_
Z7T
7T
52 R. W. P. King, H. R. Mimno, A. H. Wingr, "Transmission

Lines, Antennas, and Waveguides", McGraw-Hill, p. 184; 1945.
53 H. Jasik, "Antenna Enginering Handbook", McGraw-Hill, Fig.
3-34; 1961.
54 H. Jasik, "Antenna Engineering Handbook", McGraw-Hill, p.

2.14; 1961.
55 J. D. Kraus, "Antennas", McGraw-Hill, p. 25; 1950.
21
elude the power radiated in cross polarisation in the

determination of directivity. Even when this is done, the
results are not very accurate for low-loss highly-directive
antennas, because of the difficulty in calculating directivity
from the measured patterns with sufficient accuracy.
Another method may be available when the antenna is
electrically small and simple. In this case, an equivalent
series circuit can frequently be found in which the real
part of the input impedance, i.e., the antenna resistance,
is equal to the sum of the radiation resistance and a loss
resistance. The radiation resistance accounts for all
radiated power and the loss resistance accounts for all
dissipation within the antenna. For antennas, such as
dipoles and loops, where the theoretical pattern can be
integrated, the radiation resistance is best found by calculation from the dimensions. 56 ' 5 7 The antenna resistance
is obtained from measurements of input impedance (see
Section 7.1). T h e radiation efficiency is then:
radiation efficiency = radiation resistance
(9)
antenna resistance
This method is valid only if the antenna can be accurately
represented as a series circuit. When the dissipation cannot be represented by a resistance in series with the radiation resistance, as in the case of an antenna coated with
lossy dielectric or an antenna over a lossy ground, the
method should not be used. Furthermore, the calculated
radiation resistance and the measured antenna resistance
must be referred to the same set of antenna terminals. It
should also be noted that the input impedance of this
type of antenna may present a large mismatch to the
connecting transmission line, and a matching network
having appreciable dissipation might be used. Such a loss
is not usually included within the meaning of radiation
efficiency, although it is clear that the loss would be
important to the system as a whole.
6.
GROUND-WAVE
MEASUREMENTS
The measurements described in Sections 2 through 5

presumed that the performance of the antenna is desired
in the far-field region. However, as mentioned in Section 1.6, some antenna systems deliver their useful power
to regions where the simplifying far-field relationships do
not apply. An important class of such antennas are
ground-based vertically-polarized antennas which operate
at frequencies which are low, which rely on the ground
as an essential means for wave propagation, and which
radiate their useful power to receiving antennas located
on or near the ground. In this case the ground can be
considered a significant part of the antenna itself, and
cannot be in the far-field region.
decays at a more rapid rate than would a wave in free

space. The concepts of poiver gain, directive gain, radiation
resistance, and radiation efficiency are not directly applicable in this situation, although arbitrary assumptions regarding the extent of the antenna system are sometimes
made which permit a limited use of these terms. Furthermore, the electrical properties of the imperfectly-conducting ground and the type of terrain may vary over the
useful coverage area surrounding the antenna, so that
irregularities in the ground-wave field strength are likely
to exist.
A complete measurement of the ground-wave radiation
by an antenna involves the measurement of field strength 59
at every significant point on the ground within the useful
coverage area. The power supplied to the antenna during
these measurements must also be determined. A pattern may
then be plotted showing field strength throughout the coverage area for the specified value of antenna power; such
a pattern is often plotted in terms of contours of equal
field strength.* 0 -
When certain conditions exist, a relatively small number of measurements can be used, in conjunction with
well-known
propagation
formulas
and
published
curves, 62 - 63- 64 6 5 to estimate the field strengths at various other points throughout the coverage area. For antennas operating below approximately 5 megacycles per
second, the useful component of the ground wave is
usually the "surface-wave" because the "space-wave"
component cancels out near the ground. 66 * 6 7 If the electrical properties of the ground are reasonably constant
along lines radially outward from the antenna, the "surface-wave" field strength can approximate simple functions of distance on each line. These simple functions
usually apply for distances from the antenna greater than
one wavelength and greater than five times the vertical
height of the antenna, and for distances less than
8 X 106 ( / ) ~ 1 / 3 meters (the frequency, /, being in cycles
per second) where the curvature of the earth begins to
have a significant effect.
58 "IRE Standards on Wave Propagation: Definitions of Terms",
IEEE No. 211 (formerly 50 IRE 24 S I ) ; 1950.
59 "IRE Standards on Radio Wave Propagation: Measuring
Methods"; 1942.
60 F. E. Terman, "Radio Engineers' Handbook", McGraw-Hill,
p., 740; 1943.
61 C. R. Burrows, L. E. Hunt, A. Decino, "Ultra-Short-Wave
Propagation: Mobile Urban Transmission Characteristics",
Bell System Tech. Jour., vol. 14, pp. 253-272; April 1935.
62 K. A. Norton, "The Propagation of Radio Waves over the
Surface of the Earth and in the Upper Atmosphere", Proc.
IRE, vol. 24, pp. 1367-1368, October 1936, and vol. 25, pp.
1203-1236; September 1937.
63 K. A. Norton, "The Calculation of Ground-Wave Field Intensity over a Finitely Conducting Spherical Earth", Proc.
IRE, vol. 29, pp. 623-639; December 1941.
64 C. R. Burrows, "Radio Propagation Over Plane Earth, Field
Strength Curves", Bell System Tech. Jour., vol. 16, pp. 45-75
and 574-577; January and October 1937.
65 C. R. Burrows and M. C. Gray, "The Effect of the Earth's
Curvature on Ground-Wave Propagation", Proc. IRE, vol. 29,
pp. 16-24; January 1941.
pp. 674-769; 1943.
67 H. Jasik, "Antenna Engineering Handbook", McGraw-Hill,
pp. 33.1-33.24; 1961.
The useful component of radiation for such antennas

is usually represented by the ground wave.58 This is a
wave that is associated with currents that flow in the
ground, which is an imperfect conductor. Power is absorbed by the ground, and the ground wave ordinarily
56 J. D. Kraus, "Antennas", McGraw-Hill, pp. 136, 143-148,
166-169; 1950.
pp. 787, 788, 793, 795, 797, 806, 814, 820; 1943.
22
antenna under test and of the standard antenna are such

that the relative effect of the ground-reflected component
of the "space wave" is the same in either case. First the
signal voltage Vx received in a vertically-polarized pickup
Four theoretical curves of relative field strength vs.

relative distance are shown in Figure 11 for this ideal
situation. Each curve represents a different set of electrical properties of the ground; in addition, the quantity
p is a function of the ground properties, so that each
curve requires a translation in the scale of relative distance. By making about ten or twenty measurements of
field strength for evenly-spaced intervals along a radial
line within the range of distances mentioned above, and
by plotting these on the same logarithmic scales as those
of Figure 11, a curve may be chosen and located horizontally and vertically so that it represents the best fit to
the measured data; this process is repeated for each
radial line. By interpolation and extrapolation, the field
strength may now be estimated at any point within the
above-described range. If more detailed theoretical curves
are desired, they may be found in both graphical 68 and
tabular form in the literature. 59 - 6 3 The electrical properties of the ground may be determined from the quantities
b and p in Figure 11 and relationships presented in the
literature ; 6 2 ' 6 3 it is wise to do this in order to determine
that the best fits, as established above, represent reasonable values.
It is often desired to estimate the field strength at
distances beyond the range where the field strength approximates a simple function of distance, and a helpful
factor for this purpose is the unattenuated "surface-wave"
field-strength pattern at a standard distance from the antenna. This is a hypothetical pattern that may be obtained from the measurements and the curve-fitting procedure described above. For each radial set, the line
marked unattenuated field strength in Figure 11 is extended until it intersects the standard distance (such as 1
kilometer or 1 mile) and the value of field strength is
read off. The resulting pattern of unattenuated field
strength vs. azimuth angle for the standard distance may
be used, in conjunction with well-known propagation
formulas and published curves, 62 ' 6 3 ' 64- 6 5 to estimate the
field strengths at greater distances where effects such as
the curvature of the earth become important.
P
RELATIVE DISTANCE
Figure 11
Decay of the "surface-wave" component of the
ground wave for a plane earth
For antennas operating above about 5 megacycles per

second, the "space-wave" component of the ground wave
cannot usually be neglected, and the contour of the ground
is a significant factor. However a hypothetical unattenuated pattern at a standard distance is still a helpful concept in estimating field strengths over a coverage area. In
this case the unattenuated field strength may be determined by a comparison method in which the antenna
being tested is compared with a standard antenna. The
comparison method requires that the antenna under test
be removable; this is usually possible in the frequency
range above 5 megacycles per second where antennas are
usually not extremely large. The standard antenna may
consist of either a vertical loop or a vertical dipole for
which the vertical electric field Es expected over a perfectly conducting ground plane can be calculated when
the antenna current is specified; this value is just twice
the free-space value of the vertical field strength. The
method is valid only if the vertical-plane pattern of the
antenna placed at some standard distance from the antenna under test is measured; this distance should be
greater than one wavelength and greater than ten times
the largest dimension of the test antenna. Then the antenna under test is removed and the standard antenna is
placed in the same effective location with due consideration being given to the probable difference in- current distributions of the two antennas. The signal voltage Vs
received in the pickup is now measured for a current in
the standard antenna equal to that assumed in the fieldstrength calculation. The unattenuated field strength Ex
68 Federal Communications Commission, "Rules and Regulations",
vol. I l l , Section 3.184; September 1961.' See also Section
73.184; December 1963. '
23
mitting situations, in order to attain maximum power

efficiency, a mismatch in the opposite direction may be
required. However in many applications, a matched
condition is the ideal. By this is meant conjugate match
between the antenna and the circuits; maximum power
transfer is attained in this case. When conjugate match
does not exist, some of the available power is lost, as
follows:
Piost
\Zant Z*cct| 2
(ii)
of the test antenna at the standard distance is then determined from the following relation:
(10)
Vx
*
Ex =
Vs
To determine the complete unattenuated ground-wave
field-strength pattern in the horizontal plane, this procedure is repeated in a number of directions from the
antenna under test.
P available
In addition to its use in estimating field strengths, the

unattenuated ground-wave field-strength pattern is helpful
as an indication of the quality of the antenna design. With
the unattenuated pattern, some of the variable effects of
the ground are removed, particularly those at distances
far from the antenna. An example of this is the determination of equivalent radiated power for ground-wave
transmission.59
where Zant is the input impedance of the antenna, Zcct

is the input impedance of the circuits at the antenna terminals, and Z*cct is the complex conjugate of the circuit
impedance. This expression may be useful during the
measurement of power gain with a mismatched system,
as mentioned in Section 4.2.
Most antennas are connected to the electronic networks
via a transmission line, and the desired degree of match
or mismatch could be adjusted for at either end of the
line. In practice, however, the antenna is generally
matched to the transmission line on an image-impedance
basis as closely as possible. This minimizes the line
losses and the voltage peaks on the line, and generally
maximizes the useful bandwidth of the system. Imperfect
matching of the antenna to the transmission line creates
a reflected wave in the line; the reflected power relative
to the incident power is:
It is sometimes desired to determine the vertical-plane

pattern of a ground-based vertically-polarized antenna.
This pattern is of interest, for example, when the "skywave"58- 66 radiation characteristics are important. Although the pattern in the far-field region of the antennaground system may be desired for the prediction of "skywave" characteristics, practical limitations may rule out
the far-field measurements described in Section 2.2. In
particular, at low elevation angles and low frequencies
there may be a special problem in the measurements, because of the presence of ground-wave components in the
total radiation field of the antenna. In this case the shape
of the vertical-plane pattern can depend very markedly on
the distance at which the pattern is measured. 62 69 However, is is sometimes possible to estimate the far-field
vertical-plane pattern from a series of near measurements
by subtracting out, with appropriate relative phase, the
"surface-wave" and "space-wave" components of the
ground wave.
Prefl
Pine
__ \Zgnt Zo\*
[Zant ~T Zo\
(12)
where Z0 is the characteristic impedance of the line. This

ratio is related to the voltage reflection coefficient p and to
the standing wave ratio S by the standard transmissionline relationship:
Prefl =
Pine
|p|2 =
' '
IS - 112
IS + l|
(13)
If the transmission line has a characteristic impedance

which is pure real, and if the electronic networks are perfectly matched to the transmission line, this ratio yields
the loss of available power caused by the reflection at the
antenna terminals; if not, this loss must be computed
from (11).
7. MEASUREMENTS OF IMPEDANCE
7.1 Input Impedance Measurements
The input impedance of an antenna at the specified terminal pair (or port) affects the interaction between the
antenna and its associated circuits. Antenna impedance
can be an important factor in consideration of power
transfer, noise, and stability of active circuit components.
Frequently, it is the antenna impedance which limits the
useful bandwidth of the antenna.
Measurement of input impedance is made at a single

port of the antenna. For the usual problems and procedures related to this measurement, reference should be
made to the appropriate Standard.70 There is, however, a
particular problem inherent in radiating structures, since
the input impedance is modified by the environment of the
antenna. For this reason, the antenna should be placed
in a replica of its operating environment before the meas-
The optimum impedance relationship between an antenna and its associated circuits is determined by the
application. In some receiving situations in the interest of
minimum noise figure, a mismatch in the direction of
lower antenna impedance may be desired. In some trans-
70 " I R E Standards on Antennas and Waveguides: Waveguide and

Waveguide Component Measurements", Section 2.8Measurement of Input Impedance, Section 2.9Measurement of Normalized Input Impedance; I E E E No. 148 (formerly 59 IRE
2 S I ) ; 1959.
69 J. R. Wait and A. M. Conda, "Pattern of an Antenna on a

Curved Lossy Surface", IRE Trans, on Antennas and Propagation, vol. AP-6, pp. 348-359; October 1958.
24
urement is made. Usually, this requirement is easily

approximated for narrow-beam antennas that can be
pointed away from reflecting obstacles, but may be more
difficult for omnidirectional type antennas where much
of the surrounding structure affects the input impedance.
At low frequencies, the two components of input impedance may be measured directly by a radio-frequency
bridge.71 These instruments are commercially available
up to about 500 megacycles per second. The frequency
range for direct measurement may be extended to about
3 gigacycles per second, by admittance meters or reflectometers which are arranged to measure the components of admittance or impedance.
At microwave frequencies, nominally above 1 gigacycle
per second, the slotted section is the usual instrument for
impedance measurement.72 With this instrument, the
measurement of standing-wave ratio and position of the
standing-wave minimum is equivalent to an impedance
measurement. Often, this slotted-section data is plotted
directly on a polar chart of complex reflection coefficient;
such a chart employs the relation between reflection-coefficient magnitude and standing-wave ratio given in (13).
If desired, this data can be transformed to input impedance by means of simple graphical methods involving
impedance coordinates drawn on the reflection-coefficient
chart. Two such coordinate systems are in common use:
one73 involves the resistive and reactive components of
impedance, while the other74 involves the magnitude and
phase of impedance. The impedance so determined is in a
form which is normalized to that of the slotted-section
transmission line; this is usually the convenient form at
microwave frequencies, where transmission lines are
common.
7.2 Mutual Impedance Measurements
Certain antenna systems comprise two or more radiating
elements which are generally similar and are excited to
obtain directional effects; such a system is called an
array antenna. There is usually an interaction between
the elements of an array which significantly affects the
behavior of the antenna system, and it is desirable to
determine this interaction. An important measure of the
interaction is the mutual impedance between the elements
of the array antenna.
The mutual impedance between any two elements is defined by the following equation:
7
Vm
^mn
In
(14)
71 F. E. Terman and J. M. Pettit, "Electronic Measurements",

McGraw-Hill, Chapter 3; 1952.
72 C. G. Montgomery, "Techniques of Microwave Measurements",
M.I.T. Radiation Laboratory Series, vol. 11, McGraw-Hill,
Chapter 8 ; 1947.
73 P. H. Smith, "An Improved Transmission Line Calculator",
Electronics, vol. 17, pp. 130-138, 318; January 1944.
74 P. S. Carter, "Charts for Transmission-Line Measurements and
Computations", RCA Review, vol. I l l , pp. 355-368; January
1939.
where In is the current at the reference point of the

driven element n, and Vm is the voltage produced at the
reference point in element m, when all the elements except
the driven element are open-circuited at their reference
points. The value of mutual impedance depends on the
type of elements, their size and spacing compared with a
wavelength, and the geometrical arrangement of the elements and their environment. Generally, both the magnitude and the angle of the mutual impedance decrease with
increased spacing.
The reference points (or terminal pairs) at which the
currents and voltages are determined may be selected for
greatest convenience, but in any case they must be specified. In tower radiators which are the elements of a
broadcast antenna array, the reference points are customarily taken at the base of the towers even though the
towers may be of unequal height, and the effect of base
insulators is often subtracted from the measured values.
In dipole arrays, such as those used at H F and VHF,
the reference points are generally taken to be at the current maximum of the radiating portion of the elements.
In microwave arrays, the reference points are most frequently taken to be at the ports where the waveguides
connect to the elements.
An impedance matrix can be made up from a set of
mutual impedance values Zmn for a given array. This
matrix includes values for the self impedances Zmm,
which occur when the voltage and current are for the
same element. In general, the self or "passive" impedance
of any element in the array will differ from its "active"
input impedance that exists when the entire array is
excited, since the terminal voltage under the latter condition is the resultant of the voltages coupled from all the
excited elements by the mutual impedances. Since the
impedance matrix relates the terminal voltage at each
element to the entire set of applied element currents,75 the
"active" input impedance of each element for a given set
of excitation currents can be determined76' 77 if the impedance matrix is known. In general, these "active" input
impedances will change if the excitation is changed, because the changed relative phases or amplitudes of
excitation produce different resultant terminal voltages.
Such changes occur, for example, when the pattern of an
array of broadcast towers is being adjusted, or when the
major lobe of a large planar array is being scanned.76
The mutual impedance between two elements of an
array is sometimes determined by a calculation.78 Such
calculation is usually feasible only when all the elements
of the array have some simple, idealized configuration.
For example, if all the elements are electrically short,
thin dipoles or monopoles, the group of open-circuited
elements surrounding the two elements being measured
has little influence on the mutual impedance. In this case
an assumed current flow in one element permits computation of the electric field created at the second element;
this establishes the distributed voltages induced in the
second element as a result of the assumed current in the
first. It is then assumed that this induced voltage in each
differential length of the second element has the same
effect as some other induced voltage acting at the refer25
In U H F and microwave antenna arrays consisting of

a large number of identical regularly-spaced radiating
elements, the mutual impedance between all pairs of elements having the same relative position tends to approximate a constant value except near the edges of the
array. 7 6 * 8 0 In such cases, measurements are sometimes
made on a limited-size array having an element design and
arrangement like that of the full-size array. By measuring
the mutual impedance between elements near the center
of the limited array, a partial impedance matrix can be
determined which has values that approximate the corresponding ones in the full-size array. However, care must
be taken in interpreting the results in terms of various
related aspects of array performance such as the "active"
impedance as a function of electronic beam scanning, or
the gain and shape of the element pattern; the omission
of the effects of the missing elements in the limited array
may correspond to significant errors in these quantities
for the full-size array.
ence point, and the total voltage at the terminals of the

second element is obtained by integration of these equivalent voltages referred to the reference point. Calculated
values of mutual impedance obtained by such methods
are valuable as intrinsic references. However, practical
radiating elements are often relatively long and thick,
and have a variety of shapes; furthermore, there are
usually various attachments such as feeding lines, supporting structures, and insulators, which have modifying
effects. Thus it is generally necessary to resort to a direct
measurement of mutual impedance for an actual array of
practical elements.
The direct determination of mutual impedance from
voltage and current measurements is most commonly
made on antenna arrays in the M F range. A common
example is a broadcast antenna array consisting of several
electrically-small tower elements. A knowledge of the
mutual impedance in such arrays is necessary to determine
the branching and coupling circuit requirements which
will provide the proper current amplitude and phase relations in the array elements to produce the desired directional characteristics. Frequently such antenna arrays
employ tower elements of different heights in non-linear
arrangements with unequal current amplitudes and irregular phase relationships in order to produce a radiation
pattern which best fits the desired coverage and protected
directions. In an antenna array containing only a few
radiating elements, the non-repetitive relative position of
each element in the array may result in a substantially
different mutual impedance between different pairs of
elements.
At much lower frequencies, antenna arrays are not
generally used; when they are, however, the currents and
voltages which establish the mutual impedances are directly measurable as at M F . In the higher radio-frequency region, where currents and voltages are not readily
measurable, the mutual impedances between elements of
an array may be derived from measurements of the input
impedances of the elements at their reference points under
appropriate conditions. Two or more terminating conditions for the coupled elements are required, such as
short and open circuits at the terminals, the latter of
which permits measurement of the self impedance of the
elements. Such measurement procedures are based on
exact analogy between the current and voltage relations
between the terminals in a system of radiating elements
and the corresponding current and voltage relations in a
circuit network. An impedance bridge is often used for
such measurements, and a specific method is described
in the literature.79
For some arrays, especially those at microwave frequencies, mutual impedance is not the most convenient
term to describe the coupling between elements. When
the elements comprise slots in a conducting plane, 81
mutual admittance and an admittance matrix 7 5 provide
the appropriate information in a convenient manner.
Here the measurement involves short circuits instead of
open circuits at the elements, and voltages and currents
are interchanged in accordance with the duality principle.
For other arrays, such as those in which the elements are
horns or other large radiators, it may be more convenient
or significant to measure the incident, reflected, and
coupled waves at the terminals or ports of each element.
Such measurements would yield the coefficients of a
scattering matrix, 8 2 and would involve terminations rather
than open or short circuits at the element ports. If desired, the scattering matrix may be converted to an impedance matrix by a computation. 82 Rather than perform this
complex transformation, however, it may often be preferable to use the scattering matrix directly with the set
of incident-wave excitations to determine the active array
characteristics.
8.
POWER HANDLING
MEASUREMENTS
Antennas are often tested to determine their ability to

handle the power generated by their associated transmitters. These tests may be concerned primarily with limitations imposed by metallic or dielectric heating at high
average power levels, or with limitations imposed by the
arcing, voltage breakdown, or corona discharge associated
with high electric fields at high peak power levels.
75 C. G. Montgomery, R. H. Dicke, E. M. Purcell, "Principles

of Microwave Circuits", M.I.T. Radiation Laboratory Series,
vol. 8, McGraw-Hill, p. 140; 1948.
76 P. S. Carter, Jr., "Mutual Impedance Effects in Large Beam
Scanning Arrays", IRE Trans, on Antennas and Propagation,
vol. AP-8, pp. 276-285; May 1960.
77 L. A. Kurtz, R. S. Elliott, S. When, W. L. Flock, "MutualCoupling Effects in Scanning Arrays", IRE Trans, on Antennas
and Propagation, vol. AP-9, pp. 433-443; September 1961.
78 P. S. Carter, "Circuit Relations in Radiating Systems and Applications to Antenna Problems", Proc. IRE, vol. 20, pp. 10041041; June 1932.
79 G. H. Brown, "Directional Antennas", Proc. IRE, vol. 25, pp.

78-145; January 1937.
80 J. L. Allen, "Gain and Impedance Variation in Scanned Dipole
Arrays", IRE Trans, on Antennas and Propagation, vol. AP-10,
pp. 566-572;. September 1962.
81 S. Edelberg and A. A. Oliner, "Mutual Coupling Effects in
Large Antenna Arrays: Part ISlot Arrays", IRE Trans, on
Antennas and Propagation, vol. AP-8, pp. 286-297; May 1960.
82 C. G. Montgomery, R. H. Dicke, E. M. Purcell, "Principles of
Mictowave Circuits", M.I.T. Radiation Laboratory Series, yol.
8, McGraw-Hill, pp. 146-148; 1948.
26
In cases where the antenna is subject to low atmospheric

pressures, it is necessary to simulate the high altitude conditions by means of a bell jar or some other suitable
chamber in order to achieve satisfactory peak power handling measurements, 83 ' 84* 8 5 since extrapolation from tests
conducted at sea level is not reliable. At very low pressures, the breakdown power level decreases with increasing pressure, reaches a minimum at a pressure that depends
on the wavelength (the glow discharge region), and then
increases with pressure at high pressures. The use of
a radioactive material to create a continuous supply of free
electrons in the atmosphere in the near vicinity of the
antenna under test is a useful means of providing reliability and consistency in breakdown tests 86 , and does not
lower the absolute breakdown threshold.
Because of the great variety of antenna configurations
and environmental conditions, no attempt will be made to
provide a specific procedure for conducting power-handling
measurements. 87 For accurate results, it is important to
insure that the source of power for the measurements has
the same characteristics as the actual transmitter with
which the antenna is to be used; i.e., the modulation, pulse
shape, pulse width, pulse rate, etc. should be the same
and should not change with the applied power level during
the test. Temperature rise may be measured by means of
thermocouples or temperature-sensitive paints applied to
critical surfaces. Care should be taken that these added
components are not themselves heated by the radio-frequency power. Depending on the type, breakdown may be
detected by visual observation, audible indication, change
in signal picked up by a transmission-monitor antenna, or
change in a reflected-wave monitor signal as the applied
power level is increased. It is sometimes important,
as in the case of glow discharge, to determine the power
level at which breakdown ceases once it has been initiated.
Sharp corners, dirt, metal particles, and corrosion are
often major offenders in lowering the breakdown-power
level.
in other sections of this Test Procedure as well as possible

within the limitations imposed by the test chamber.
Whenever measurements are performed in which high
average power levels exist, it is essential that proper safety
devices and procedures be employed for protection of the
personnel in the vicinity. 88
9.
NOISE TEMPERATURE
MEASUREMENTS
Noise power available at the antenna terminals may

arise from many sources; some typical ones are radio
stars, the galaxy, the sun and moon, atmospheric discharges, atmospheric absorption, ground absorption, antenna and feed dissipation, and transmission line attenuation. The noise temperature of an antenna?9 usually expressed in degrees Kelvin, is the temperature of a passive
system having an available noise power per unit bandwidth equal to that at the antenna output port, at a specified frequency. If a wide frequency band is of interest, an
average noise temperature may be employed; in this case
the available noise power is determined in a specified
frequency band. Although the concept of antenna noise
temperature is applicable in general, it is most used at the
higher radio frequencies where the noise is least likely to
fluctuate radically, and where thermal sources of noise are
significant contributors.
Measurements of antenna noise temperature (Tant) are
usually made by comparing the antenna noise output with
that from a standard source. For greatest accuracy, it is
desirable that the noise temperature of the standard (Tstd)
be close to Tant. Values of Tant in practice may range
from 3K to 3000K and further. Several standard sources
covering this range are available. One type comprises a
matched resistive termination, and a means for controlling
its physical temperature, which is also its equivalent noise
temperature. Cooling agents such as liquid nitrogen
(77.4K) or liquid helium (4.2K) may be used; the termination may be air cooled and have a thermometer for
measuring ambient temperature, or artificial heating, for
example, by means of boiling water, may be employed.
The low- and high-temperature standards must be such
that the impedance match is preserved at their operating
temperatures. Another type of standard source involves a
gas tube 9 0 such as an argon noise lamp having an equivalent noise temperature (about 10,000K) much higher
than ambient temperature. In addition to its use in measurements of high values of Tant, a gas tube is applicable
in low-temperature measurements by connecting it to the
system through a directional coupler of known coupling
loss. 91 Its effective temperature can also be reduced by
flashing the tube on a low duty cycle.
It is sometimes desired to maintain transmission from

an antenna in the presence of breakdown, such as when
high altitude glow discharge occurs with a missile antenna.
In those cases, it is usually desirable to measure not only
the breakdown power level, but also the changes in input
impedance, radiation pattern, power gain, and distortion
of the transmitted signal for power levels above the breakdown level. Care must be taken in making such measurements to satisfy the appropriate test conditions discussed
83 W. J. Linder and H. L. Steele, "Estimating Voltage Breakdown Performance of High-Altitude Antennas", WESCON
Convention Record, Part 1, pp. 9-16; 1959.
84 A. D. MacDonald, "High-Frequency Breakdown in Air at High
Altitudes", Proc. IRE, vol. 47, pp. 436-441; March 1959.
85 J. B. Chown, W. E. Scharfman, T. Morita, "Voltage Breakdown Characteristics of Microwave Antennas", Proc. IRE, vol.
47, pp. 1331-1337; August 1959.
86 L. Gould and L. W. Roberts, "Breakdown of Air at Microwave
Frequencies", Jour, of Appl. Physics, vol. 27, pp. 1162-1170;
October 1956.
87 " I R E Standards on Antennas and Waveguides: Waveguide and
Waveguide Component Measurements", Section 2.10Measurement of Dielectric Voltage Breakdown, Section 3.7Measurement of Power-Handling Capacity; 1959.
The circuit arrangements and procedures for measuring

antenna noise temperature vary considerably, depending
on the type of information required. In some cases, a
simple substitution test is made in which two readings of
88 W. W. Mumford, "Some Technical Aspects of Microwave Radiation Hazards", Proc. I R E , vol. 49, pp. 427-447; February 1961.
89 " I R E Standards on Electron Tubes: Definitions of Terms",
I E E E No. 160 (formerly 57 I R E 7 S2), p. 1000, Noise Temperature (at a P o r t ) ; July 1957.
27
power are sufficient. On the other hand, a complex

radiometer system 92 in which many readings are averaged
to achieve a greater precision is often employed in radio
astronomy. Sometimes the circuit includes more than one
type of standard source, and in many of the more accurate
measurements a calibrated precision attenuator is adjusted
to maintain constant output noise power in the two test
conditions. In the latter case, the noise introduced by the
attenuator, as well as its attenuation, must be included in
the calculation. A simple example of this is shown in
Figure 12, in which it is assumed that the attenuator is a
resistive (not reactive) type and that all components are
matched to their transmission lines, which are lossless.
T
ANTENNA
ont
PRECISION
ATTENUATOR
LOSS L
STANDARD
.M
NOISE
POWER
i
INDICATOR
1
Figure 12
Circuit for m e a s u r i n g noise t e m p e r a t u r e of an a n t e n n a

With the switch in position 1 and the precision attenuator
set to some convenient value of loss (ratio of power input
to output) Li, the noise temperature of the noise power
output to the indicator is :
(15)
where Tamb is the ambient temperature of the attenuator.

With the switch in position 2, the attenuator is now adjusted to a new value L2 such that the noise power to the
indicator remains the same; the equivalent output temperature is then:
(16)
Tout = T^
+ I 1 - - M Tamb
L2
\
L2 '
Equating these two relations permits calculation of the
antenna noise temperature:
(17)
Tant =
Tamb T" = (Tstd

L2
Tamb)
It should be noted that with the circuit of Figure 12, an

antenna temperature below ambient temperature requires
a standard with a temperature also below ambient temperature, and vice-versa. N o matter what circuit arrangement is used to measure antenna noise temperature, re-
liable results are obtained only when care is taken to

avoid the pitfalls which are present in any noise measurement of any component. 93
The measurement of antenna noise temperature is complicated by the dependence of the noise on several variables. Two such variables are the direction of antenna
pointing and the antenna location; these determine the
particular external sources of noise which may couple
to the antenna. The antenna noise temperature is also
likely to vary with time, because of changes in the external noise sources.
10. D E T E R M I N A T I O N O F S C A T T E R I N G
CROSS-SECTION
Ordinarily, an antenna radiates power which it obtains
from a circuit of some kind. However, it is also possible
for a radiating antenna to obtain its power from an incident electromagnetic wave. The characteristics of the
antenna radiation are, in general, not the same for the two
cases. In the latter case, which is known as re-radiation
or scattering, the "antenna" may be simply an obstacle
having no connection to any circuit.
The nature of the scattering from an obstacle may be
of great interest, and, depending on the complexity of the
obstacle, may have to be determined by measurement
rather than by analysis. For example, the scattering pattern of the feed for a reflector antenna, when it is blocking a portion of the reflector aperture, may be significant.
Another example is the distribution of power in the wave
scattered from targets in a radar system. In most cases,
it is important to determine the level of the scattering
pattern as well as the shape of the pattern. Often, the
level of power scattered in only certain directions is of
interest.
The quantity defining the ability of an obstacle to
scatter power from an incident plane wave is termed the
scattering cross-section. This may be regarded as the
area from which power would have to be extracted from
the incident plane wave in order to give the same radiation intensity in a specified direction as does the obstacle,
if the extracted power were reradiated isotropically. A
perfectly reflecting sphere which is large compared with
a wavelength scatters an incident plane wave approximately isotropically, 94 except in the forward direction.
(In the forward direction, diffraction created by the
blocking effect always causes an increase in the scattered
field.)95 The power which is scattered isotropically is
equal to that extracted from the incident plane wave in
accordance with the cross-sectional area of the sphere.
Hence for obstacles having a scattering cross-section large
compared with a square wavelength, the scattering crosssection for any direction except the forward one is approximately equal to the cross-sectional area of a perfectly
90 W. W. Mumford, "A Broad-Band Microwave Noise Source",

Bell System Tech. J., vol. 28, pp. 608-618; October 1949.
91 R. W. DeGrasse et al., "Ultra-Low-Noise Measurements Using
a Horn Reflector and a Traveling-Wave Maser", Journal of
Applied Physics, vol. 30, No. 12, p. 2013; Dec. 1959.
92 R. H. Dicke, "Measurements of Thermal Radiation at Microwave Frequencies", Rev. Sci. Instr., vol. 17, pp. 268-275;
July 1946.
93 " I R E Standards on Methods of Measuring Noise in Linear

Twoports", 59 IRE 20 SI, Proc. IRE, vol. 48, pp. 60-68;
January 1960.
94 L. N. Ridenour, "Radar System Engineering", McGraw-Hill,
M.I.T. Radiation Laboratory Series, vol. 1, p. 65; 1947.
95 H. C. Van De Hulst, "Light Scattering by Small Particles",
John Wiley and Sons, p. 163; 1962.
28
rotation, in a manner similar to the recording of an antenna pattern.
reflecting sphere which, if substituted in place of the

obstacle, would scatter with the same radiation intensity
in that direction.
The scattering cross-section of an obstacle is a function
of the incident-wave direction and polarization. The term
"scattering cross-section" is a general one which applies
to any specified direction of scattering, and the crosssection is therefore also a function of the scattering direction. A particular scattering direction of interest in many
situations is that toward the source of the incident plane
wave. In this case the appropriate term is called the
back-scattering cross-section. Alternate terms for backscattering cross-section in common use include back-scattering coefficient, monostatic cross-section, and target
echoing area. When the direction'of interest is other than
that back toward the source, the term bistatic cross-section
is often employed.
The entire system is first arranged without the obstacle.

Adjustments are made on the tuner to produce a minimum
detected signal; by this means, reflections from the antenna and nearby objects may be cancelled. A calibration
is then performed by measuring the detected signal with a
standard obstacle such as a sphere M whose back-scattering cross-section can be computed. Now the back-scattering of other obstacles can be determined by comparison
of the detected signal level with that for the standard
obstacle, on the basis of power into the detector.
In radar practice, that portion of the back-scattering

cross-section which is observed by a radar antenna system
operating with specified polarization characteristics is
called the radar cross-section. It is often given the symbol
a, as in the "radar equation," for example. 96 - 9 7 For the
ordinary linearly-polarized radar, that portion of the backscattered wave having the same polarization as the incident wave is observed; this is usually the major component of the back-scattered wave. For a circularlypolarized radar having opposite-handed transmission and
reception, that portion of the back-scattered wave having
a handedness opposite to that of the incident wave is
observed; again, this is usually the major component of
the back-scattered wave. For a circularly-polarized radar
having the same handedness for transmission and reception, the radar cross-section is often relatively small.
Es
RF
SOURCE
(CW)
HYBRID
TEE
TUNER
IE
MODULATOR
POSITIONING
MECHANISM
DETECTOR
1
RECORDER
In any system involving scattering from an obstacle,

the total field strength Et is given b y :
Et = Ei+
OBSTACLE UNDER
MEASUREMENT
ANTENNA
Figure 13
Measurement system for determination
of back-scattering
(18)
where E% is the incident field and Es is the scattered field,

and provided that the scattered field is not distorted by
the presence of the source antenna. In making measurements to determine scattering cross-section, the main difficulty lies in separating the incident and scattered fields.98
As an example, a simple system is shown in Figure 13
for making measurements to determine back-scattering.
A signal from an R F source is fed into the H-arm of a
hybrid-T. The power divides between the side arms, one
of which is connected to a terminating load through an
adjustable tuner, and the other to a directive antenna.
Some of the power which is radiated by this antenna is
scattered by the obstacle anti part of the scattered power is
then received by the same antenna. A portion of this
received power leaves the hybrid-T by the E-arm, which
is connected to a detector through a modulator which
modulates the received scattered signal. The detected
signal is amplified and recorded as a function of obstacle
96 L. N. Ridenour, "Radar System Engineering", McGraw-Hill,
M.I.T. Radiation Laboratory Series, vol. 1, pp. 21-22; 1947.
97 D. E. Kerr, "Propagation of Short Radio Waves", McGrawHill, M.I.T. Radiation Laboratory Series, vol. 13, p. 33; 1951.
The above measurement is directly applicable to backscattering cross-section when a negligible amount of the
power scattered by the obstacle is cross-polarized to the
incident field. This is the situation if linear polarization
is employed and the obstacle is circular in shape. When a
significant amount of polarization conversion by the scattering obstacle is suspected, a more complicated system is
required in order to measure the power in both the normal
and cross-polarized scattered fields. As in the case of
gain measurements (see Section 4 ) , the sum of the two
orthogonally-polarized powers is employed in the determination of scattering cross-section. The ratio of the two
components may also be of interest.
Extreme care must be used to insure that the R F
source has a highly stable frequency, since the cancellation
procedure required for minimizing the effect of spurious
scattering is very sensitive to frequency change. A CW
source is generally more stable than a modulated or
pulsed source; however, temperature-controlled oscillators
98 R. W. P. King and T. S. Wu, "The Scattering and Diffraction
of Waves", Harvard University Press; 1959.
29
and automatic-frequency-control circuits are often required

in addition.
In order to minimize the need for extreme frequency
stability, it is desirable to have the obstacle in a region
remote from other objects capable of causing spurious
reflections. When the measurement is made indoors, an
anechoic chamber of high quality is required. Great care
must also be taken to minimize the reflection from the
fixture which supports the obstacle, particularly when
rotation of portions of the fixture is involved in the
measurements. The image-plane method can be used
if the obstacle has a plane of symmetry so that one-half
of the obstacle can be replaced by a very large ground
plane or image plane. In this case, only the polarization
perpendicular to the image plane is applicable. The imageplane system has the advantage that obstacle-supporting
structures and instruments which are likely to give
spurious reflections can be below the image plane, and
hence be shielded from the measurements.
Another measurement technique" involves a pulse-radar
system rather than a CW system. The essential feature
of this technique is that the incident and scattered pulses
can be separated in time; hence, there is less difficulty in
evaluating the scattered signal. Furthermore, the received
scattered signal can be range gated to discriminate against
spurious site reflections. The pulse length must be suffi99 C. C. H. Tang, "Electromagnetic Backscattering Measurements
by a Time-Separation Method", IRE Trans, on Microwave
Theory and Techniques, vol. MTT-7, pp. 209-213; April 1959.
ciently short so that excessive distance is not required between the transmitter and the obstacle; however, a measurement error is introduced if the pulse length is too
short compared with the obstacle size.
Bistatic measurements of scattering cross-section can
be made with either the CW or the pulse system. There
must be separate transmitting and receiving antennas, and
the direct coupling from the transmitter to the receiver
must be minimized or cancelled. An additional requirement is usually that the receiving antenna location must
be variable, so that the bistatic anglecan be adjusted as
desired. Otherwise, the essential features are the same as
for the back-scattering measurements.
In the case of scattering measurements on typical radar
targets, model techniques are often used because of the
considerable difficulty in making measurements with the
actual system. For scattering measurements on models,
the same considerations of scaling apply as those described in Section 1.7; in this connection it should be
remembered that scattering cross-section is an area, and
scales in proportion to the square of the wavelength.
Also, the other principles of good practice in making
antenna pattern measurements, such as those discussed in
Sections 1.6 and 2.3, must be observed in scattering measurements. For example, the separation between the obstacle and the illuminating antenna must be sufficiently
large to assure that an essentially plane wave illuminates
the obstacle. For most accurate results, this separation
should be about twice as great as that given in Section 1.6
because a round-trip phase distortion is involved.
30
IEEE Standards of Particular Interest

to Antenna and Propagation Engineers
IEEE
Std
Title
145-1969 Definitions of Terms for Antennas

146-1953 Definitions of Terms for Antennas and Waveguides
147-1955 Definitions of Terms for Waveguide Components
148-1959 Measurement of Waveguides and Components (Reaff
1971)
149-1965 Test Procedure for Antennas (Reaff 1971) (ANSI
C16.11-1971)
187-1951 Open Field Measurement of Spurious Radiation from
Frequency Modulation and Television Broadcast Receivers
197-1962 Measuring Sine-Wave Unbalanced Radio-Frequency
Voltage and A Program to Provide Information on the
Accuracy of Electrical Measurements
211-1969 Definitions of Terms for Radio Wave Propagation
284-1968 Measuring Field Strength, Continuous Wave, Sinusoidal
285-1968 Measuring Phase Shift at Frequencies Above 1 GHz
287-1968 Precision Coaxial Connectors
291-1969 Measuring Field Strength in Radio Wave Propagation
299-1969 Recommended Practice for Measurement of Shielding
Effectiveness of High-Performance Shielding Enclosures
302-1969 Methods for Measuring (Below 1000 MHz) Electromagnetic Field Strength
For a free catalog of IEEE Standards write the
Institute of Electrical and Electronics Engineers, Inc.

345 East 47th Street, New York, N. Y. 10017, U.S.A.

IEEE Standard: Test Procedure For Antennas

Uploaded by

Copyright:

Available Formats

IEEE Standard: Test Procedure For Antennas

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

IEEE Standard: Test Procedure For Antennas

Uploaded by

Copyright:

Available Formats

IEEE Standard

Approved March 1, 1971

American National Standard Institute

Approved March 1, 1971

American National Standard Institute

The Institute of Electrical and Electronics Engineers, Inc.

2. MEASUREMENTS FOR DETERMINATION O F A M P L I T U D E

3. MEASUREMENT O F O T H E R PATTERN CHARACTERISTICS

4. MEASUREMENT O F MAXIMUM POWER GAIN AND

7.2 Mutual Impedance Measurements

10. D E T E R M I N A T I O N O F SCATTERING CROSS SECTION

Test Procedure for

P r o c e d u r e , which supersedes the p r e " S t a n d a r d s on A n t e n n a s M e t h o d s of

electric field strength, volts per meter

frequency, cycles per second

distance from antenna to observation point,

One category of environmental factors may be defined

1.2 Symbols and Units

ratio of full-scale dimensions to scale-model

An antenna can be considered adequately tested only

Throughout this document, certain terms are italicized where

half-power beamwidth, degrees

In Section 1.2, symbols and units to be used in this

Tests of radome effects are often made by observing the

environment. For instance, rapid-scan antennas employing

In many cases the electrical environment of an antenna

Polarization is a property of electromagnetic radiation

wave is said to be "elliptically polarized". This is the most

out by the tip of the E-vector (see Figure 1 ( a ) ) . The

1.6 Field Regions

Beyond the reactive near-field region, the radiating field

where E is a complex quantity defining the amplitude and

In the radiating near-field region, the relative angular

Another method employs a graphical technique in which

7 H. Jasik, "Antenna Engineering Handbook", McGraw-Hill, p.

12 G. A. Deschamps, "Part IIGeometrical Representation of the

(a) Electrically small dipole

Scale-modeling techniques are often used when the

(b) Electrically large reflector

where D is the largest dimension of the aperture. The

14 S. Silver, "Microwave Antenna Theory and Design", M.I.T.

Operating frequency and conductivity of the materials

Electric permittivity and magnetic permeability of

In the measurement of radiation patterns, it is likely that

In the above, n is an arbitrary number (generally, but

In the case of an input-impedance measurement using

In a practical model it is usually not feasible to satisfy

In constructing a scale model, not only the antenna

The radiation pattern of an antenna may represent any

A system of space coordinates often employed for the

18 N. Marcuvitz, "Waveguide Handbook", McGraw-Hill, M.I.T.

discussed in Section 1.6, the angular coordinates are the

Two angular coordinates which may be employed are

Measurements of amplitude patterns are generally made

A wide variety of techniques have been employed for

System of spherical coordinate's

Figure 4 indicates the essential parts of the typical

In the lower-frequency class ( H F and V H F ) , the

tion, controls the position of the pen of the recorder. This