(NASA CR OR TMX OH AD .NUMBER)
A VERTICAL TEST RANGE
ANTENNA RADIATION MEASUREMENTS
GODDARD SPACE FLIGHT CENTER
A VERTICAL TEST RANGE FOR ANTENNA RADIATION MEASUREMENTS
John Steckel and William Korvin
Spacecraft Technology Division
Spacecraft Electronics Branch
Goddard Space Flight Center
In order to facilitate the measuring of satellite antenna radiation patterns
under a controlled RF environment the concept of a "Vertical RF Test Range"
was devised. The vertical test range configuration is shown in Figure 1.
Of primary importance was the test range's capability to adequately absorb
and/or suppress undesirable electromagnetic energy as low as 125 Mc which
would normally be reflected from the chamber walls and floor. The satellite
antennas of concern are of the broad-pattern, near- omnidirectional type (i. e.,
dipole, turnstile, etc.).
The preliminary study, investigation, and l/9-scale chamber model meas-
urements indicated that a feasible reflection coefficient level of the order of
20 db to 25 db can be expected in an electrically small chamber of approximately
2K x 2k x l\in size. Subsequent full-scale measurements in the completed
chamber have verified these reflection coefficient levels.
At higher frequencies of operation it was anticipated (and also indicated by
1/9- scale chamber model measurements) that the reflected energy level would
be lower. The dominant controlling factor at higher frequencies is the structural
weather protective radome. At the lower frequency the radome appears as a
very thin wall (approximately 0.05 k thick). But as the frequency is increased,
the presence of the radome becomes increasingly significant.
Figure 1 -Vertical Anechoic Chamber
It should be remembered that an anechoic chamber is by no means free
space. Therefore, it is of utmost importance that the RF anechoic chamber be
"calibrated" not only as a function of frequency, but also as a function of illumi-
nating or source antenna characteristics. Then, indeed, the confidence level at^.
which the anechoic chamber may be used is defined.
ce level at
I. INTRODUCTION 1
A. Antenna Test Range 1
B. Definition of R-f Anechoic Chamber 1
C. Testing Criteria as a Function of Type of Test:
Impedance, Back Scatter and Radiation Pattern 2
H. DESCRIPTION OF VERTICAL RF TEST RANGE 12
A. Physical Description 12
B. Electrical Description 15
C. Advantages 19
m. ANALYSIS OF CHAMBER 19
A. Scaled Chamber Measurements 22
B. Full Scale Chamber Measurements 25
IV. CONCLUSIONS 44
V. RECOMMENDATIONS 44
LIST OF ILLUSTRATIONS AND TABLES
1 Vertical Anechoic Chamber ii
2 Effect of Reflected Energy on Typical E-Plane Pattern
of Horn Antenna H
3 Photograph of Wall Mount Inside Chamber 13
4 Photograph Illustrating Size of Absorbent Material 14
5 Photograph of 9 Panel Radome 16
6 Photograph of l/9th Scale Chamber 23
7 Equipment Set-Up for l/9th Scale Measurements 24
8 Standing Waves in l/9th Scale Chamber Measured from
Aperture of Chamber to Floor 26
9 Standing Wave Curves Measured Across l/9th Scale Chamber
10 Photograph of Track and Cart Arrangement 29
11 Interpretation of Standing Wave Curves 30
12 Full Scale Chamber Standing Waves at 125Mc, 400Mc 31-34
13 Radiation Plots Showing Effect of Radome on Reflection
Coefficient at 1200Mc 38
14 Effect of Radome on Parallel and Perpendicular Polarization . 38-39
15 Pattern of Horn Antenna (E -plane) Measured Outside and
Inside of Chamber 42-43
A. Antenna Test Range
The most important feature of an antenna test range is the control and
reduction of reflections to the extent that they do not introduce significant
errors in the measurement of the system under test. In order to achieve
this, most test range sites are chosen in open fields clear of reflecting ob-
jects. Satisfactory operation is obtained as long as the system under test
is sufficiently directive to discriminate against ground reflections and re-
flections from the test tower. Antennas for small scientific spacecraft are
operated in the UHF refion and designed to be as nearly omnidirectional as
possible and thus are a "worst case" for conventional test ranges. The ideal
solution is to provide a reflectionless invironment for the system under test
that still permits convenient operation.
Recent advances in r-f absorbent material has improved the perform-
ance of r-f anechoic chambers to where they are comparable to the best
outdoor range sites.
B. Definition of R-f Anechoic Chamber
A r-f anechoic chamber may be defined as an enclosure, suitably lined
with an electromagnetic energy suppressing (absorbing and scattering)
material, which may be used to measure such electrical characteristics as
impedance, radiation pattern, and back scatter of a body within the chamber.
Further, the degree of suppression of reflected undesirable electromagnetic
energy shall be such as to approximate free space. The allowable divergence
from free space is a function of the type of test and the tolerable error
which the disigner places on the test. For instance, a higher level of reflec-
tions may be tolerable when measuring impedance than when attempting back
C. Testing Criteria as a Function of Type of Test: Impedance, Back Scatter,
and Radiation Pattern
A primary requirement for measuring "free space" impedance of an
an antenna is the exclusion of all external, foreign sources of reflection.
This requirement at first thought appears extremely stringent, if not im-
possible. To meet the requirement explicitly would mean that an impedance
measurement would have to be made at distances miles from the surface
of the earth and from within the antenna system. Fortunately, the require-
ment can be relaxed to a degree governed by the tolerable error permitted
in the system impedance expressed as a Voltage Standing Wave Ratio.
Consider a perfectly matched antenna system radiating energy. Under a
condition of no electrical mismatch from external sources there would be
no standing wave and VSWR = 1.00:1. In practice, all antenna systems ex-
hibit some mismatch from reflection resulting from electrical discontinuities.
The vector addition of the incident and reflected voltages results in a stand-
ing wave, the magnitude of which is expressed in the familiar form:
VSWR db = 20 log 10 -^
Ej = incident voltage
E 2 = reflected voltage
This VSWR is an indication of the power transfer of an antenna system.
For instance, a system with a VSWR of 6.0.1 would reflect 1/2 of the avail-
able power which is intolerable in most applications. Below is a brief table
indicating the percentage of r-f power transmitted as a function of system
VSWR R-f Power Transmitted
1.00:1 100 %
1.04:1 99.6 %
Since a perfect impedance match is seldom achieved the question then
becomes how large a reflection due to the environment of the test site is
tolerable while still permitting acceptable measurement of reasonable accu-
racy of the system impedance. Consider first, a worst case condition, i.e.,
all the incident energy striking a reflecting surface is reflected back to the
antenna under test. If the power radiated P Q from a perfectly matched an-
tenna of unity gain, travels a distance R to a reflecting wall and is completely
reflected back to the antenna a standing wave will result. However, the
inverse square law for the attenuation of electromagnetic energy will reduce
the amplitude of the reflected wave such that the SWR induced will be less
than infinite. For example, the attenuation of a 136 Mc signal to a reflecting
wall 8 feet away can be computed from:
a = 37 + 20 log f + 20 log R
a = attenuation (db)
f = frequency (mc)
R = distance (feet)
<* = 23.53 db
Assuming perfect reflection from the wall, the signal will be further
reduced by 6 db in the return path to the antenna. Thus, the returned signal
will be about 29.5 db below the incident signal. The resulting mismatch to
the perfectly matched antenna is computed from:
r - 1
|T| = reflection coefficient = 29.53 db
••• VSWR = r =1.07:1
From the analysis above it is concluded that the impedance of a low
gain antenna system may be measured with confidence if the walls of the
enclosure are at least eight feet away and are less effective as a reflector
than a metallic wall. Thus, the requirement on the absorption of unwanted
reflections from the environment is easily met when making impedance or
VSWR measurements on low gain antenna systems.
By definition, a , the scattering cross section of a body is the ratio of
the power scattered per unit solid angle to the power incident per unit area.
In terms of the radar equation,
P R 16t7 2 R 4
„ square meters,
P T G T G R \»
k = the wavelength in meters.
R = the distance between the source antenna and the scattering cross
section body in meters.
P = the transmitted RF power.
P D = the received (reflected) RF power.
G T = the transmitting antenna power gain with respect to an isotropic
G R = the receiving antenna power gain with respect to isotropic radiator.*
The problem which presents itself in measuring scattering cross sec-
tions in an anechoic chamber is the suppression of the energy from chamber
*An isotropic radiator is a fictitious antenna, used as a reference, which radiates energy truly omnidirectional.
Further, assuming this antenna is radiating an RF power of one watt, the magnitude of the electric field
strength measured at a radius of 1 mile from the source is equal to 3.40mw/meter.
walls reflected back to a receiving antenna. The necessary chamber per-
formance, as in the case of VSWR measurements in a chamber, is relative.
For example, a large scattering cross section would reflect an energy level
which could be many orders of magnitude greater than energy contributed
by the reflected energy from the chamber walls (including ceiling) . In this
case a percentage tolerance error can be established. Unfortunately, many
radar cross section measurements are made of objects which have a small
scattering cross section. Now, although the walls of the chamber absorb and
suppress the RF energy, the fact that the walls are many orders of magni-
tude in size larger than the object being measured results in a chamber wall
reflection coefficient which is as large as or larger than the object to be
measured. In other words, the scattering energy of the body to be measured
is hidden by the reflections from the chamber.
There is one technique which is described by Elery F. Buckley (Emerson
and Cuming, Inc.) which allows a calibration of a chamber for measuring
radar cross sections in a controlled manner.
Very briefly, a transmitting and receiving antenna are placed at one
end of a chamber. Two or more conducting spheres of known physical size
(and hence known radar cross section) are placed at the opposite end of the
chamber one at a time. The spheres, one at a time in the chamber, are
rotated eccentrically about an axis producing an in-to-out of phase response.
There will be two amplitude components for each of the different size spheres
which contribute to the recorded interference pattern. One of the amplitude
components is from the sphere; the other is the constant field equal to the
energy return from the chamber plus electric field from transmitter- receiver
cross coupling. The voltage ratio of these magnitudes produce an ambiguity.
That is to say, the chamber scattering cross section will be either of two
values, hence, the second measurement is made using a different conducting
sphere with a known a. The second measurement results again in two val-
ues of the chamber cross section. The true chamber reflecting cross section
is that cross section which is common to both of the measurements.
An example which illustrates the magnitude of energy suppression which
an anechoic chamber must exhibit for the measurement of radar cross
R,~ .9R 2
R2 = 10 meter
<r, <r 2
Let the scattering cross section be that produced by a sphere of radius
a = 0.1 meter when the wavelength ^ = 3cm (i.e., a A > 1). Then the scat-
tering cross section o- j = na 2 = 3.14 x 10~ 2 meter 2
Now, the return power received by an antenna G r from a scattering
body of cross section o- when the transmitted power P T is radiated by an
antenna G T is,
_ P t Gr Gt ^
Pr 16tt 2 R 4 a
Consider first the return power received from the spherical scattering body
and letting P T = 1 watt, R x = 9 meters,
G x = G D = 63 and \ = 3 cm.
P P T G R G T ^
Pri 16tt 2 R 4 ^
P D , = .109 x 10" 6 watts
Next, assume the wall w-w in the figure on page 7 to be a perfectly re-
flecting surface 3 meters on a side so that the wall radar cross section is
Arr A 2
" 2 = —
where A = area of wall and X. = wavelength at the operating frequency.
477 A 2
a 2 = 12.56 X 10 4 M 2
The return power received from the back wall w-w is (the side, top and
bottom walls may be neglected in this case since the predominant contribu-
tion to erroneous back scatter measurements is the back wall),
and, P p = .284 watts
By comparison P R /P R = -64.2 db
Or, the energy normal to the back wall W-W must be suppressed by
64.2 db to allow the spherical body (.1 meter radius) to present an equal
amplitude to the receiver antenna G R . Of course, further suppression of the
back wall energy must be attained before the energy from the spherical body
is discernible from (i.e., above) the back wall scattered energy.
The example outlined above now allows a feeling for the nature and
magnitude of energy suppression and/or absorbing characteristics that may
be required of an RF anechoic chamber designed to measure radar cross
Measurement of antenna radiation patterns generally requires a better
anechoic chamber than one for the measurement of antenna impedance. But
the requirements are not as critical as for radar cross section (back scatter)
As an illustration, let us examine a typical E -plane pattern of a horn
antenna and then show the effects of various magnitudes of reflected energy
interfering at various aspect angles in the chamber. Figure 2 shows the
E -plane pattern as measured in a controlled environment at an outside
antenna range. The pattern approaches the theoretical pattern for a horn
antenna exhibiting uniform distribution across the aperture. Figure 2 also
shows a ray outline of reflected energy when measuring the E -plane pattern
of this antenna in an anechoic chamber. Let sources of reflection be loca-
ted at angles defined by 8 = 50° and 60°. Then the resulting perturbations
from these reflecting sources are shown in Figure 2 as dashed lines. The
deviations in the pattern are identified as (T) , (2) , and (Jj) . The magnitude
of the perturbation at 1 is 1.6 db. The magnitude of the reflected energy
from point I which produces this 1.6 db is equal to 21 db since the reflection
coefficient in db = 20 log (S - 1/S + 1), where S = standing wave ratio.
Refering to the figure, it can be seen that the perturbations take place at
-14.2 db on the radiation pattern. Therefore, the actual reflection coefficient
(magnitude of the interfering reflected energy) is equal to 35.2 db. At
point 2 the deviation from point II is 1.1 db or a reflection coefficient of
25 db. Since the radiated energy is down 15.5 db at this point, the actual
reflection coefficient is 40.5 db. Now consider point (§) (i.e., db down on
the pattern). A refection coefficient of 40 db will cause a 0.175 perturbation
in the pattern at this point.
From the above exercise one can now see clearly the effect of reflected
energy on a typical pattern from arbitrarily chosen points in an anechoic
chamber enclosure which is 35 db to 40 db down from the incident energy.
H. DESCRIPTION OF VERTICAL RF TEST RANGE
A. Physical Description
The vertical test range is as shown in Figure 1. The walls and floor of
the structure are reinforced concrete. Attached to the chamber through a
common wall is the control room which houses all the electrical measuring
devices. Opposite the control room side are two doors which when opened
allow the chamber to be used as one end of a horizontal antenna range in
conjunction with available antenna towers. The roof of the chamber is an
A- sandwich type RF transparent radome. Also, shown in this figure is the
outside azimuth - elevation mount and fiberglas mast approximately 35 feet
above the chamber controlled from inside. This facilitates the changing of
source antennas. Inside the chamber (see Figure 3) is a wall-mounted an-
tenna mount capable of rotating and allowing measurement of a satellite
antenna system at or near the center of the chamber quiet zone.
The walls are lined with pyramidal absorbent material 70 inches in
length on 2 feet square bases (Figure 4). In the area of the side wall mount
smaller pyramidal absorbent material is used allowing the rotation of an
off- set arm containing the fiberglas mast. The floor of the chamber is
lined with foam structure 70 inch pyramids on top of which is cemented a
smooth floor - decking material of 1/2" thick sheets of semi-rigid vinyl
foam which can be walked upon in setting up experiments within the chamber.
Figure 4-Photograph Illustrating Size
of Absorbent Material
The radome is a structural body of the A- sandwich type. It is approxi-
mately 4" thick and the 9 sub-pieces are assembled into a continuous weather
protective roof (Figure 5).
Uniform chamber lighting is achieved by directing four flood lamps
toward the radome. The four lamps are located at the four upper corners
of the chamber such that they are well hidden by the absorber on the walls.
This technique of lighting results in two very favorable conditions; the
lighting is uniform and avoids the visual problem of looking into the lamps;
a minimum of RF reflection from the fixtures is obtained since they are
well shadowed by the 70" absorbent material.
Finally, the chamber is temperature controlled to prevent large varia-
tions of temperature which may effect antenna measurements.
B. Electrical Description
Electrically the chamber was lined on 5 surfaces with an RF absorbent
material and the 6th side enclosed by an RF transparent (radome) roof.
The RF absorbent material was supplied by the B. F. Goodrich Com-
pany, Shelton, Connecticut. Each individual piece of the material, VHP-70
is a 70" high pyramid shaped absorber on a 2 feet square base. All of the
absorbent material exhibits a minimum reflection coefficient of 28 db at
120Mc, 40 db at 400Mc and 50 db at 1,000 to 10,000Mc.
The tests on the absorber at the manufacturer' s plant were made in a
closed loop technique at 120Me and 400Mc. At 1,000, 5,000 and 10,000Mc an
open loop technique was used. The closed loop technique consisted of meas-
uring four 2'x2' absorber pieces at a time in a flared waveguide system in
which the absorbent material serves as a load under test. The absorber is
moved inside the waveguide resulting in a standing wave that moves in con-
junction with the physical movement of the absorber. A fixed probe inside
the waveguide located between the absorber and RF oscillator detects the
standing wave which is converted to power reflected.
The open loop technique of testing consisted of a horizontal version of
the NRL type arch method. A reference was used which consisted of a flat
2'X8' metal plate. The reflection level of four 2x2 pieces was measured and
compared to the flat metal plate. The reflectivity of the absorbent material
is the difference in return power in db between the metal plate and the
The radome was supplied by Raymond Development Industries, Inc.,
Huntington Park, California. The electrical characteristics of the A- sandwich
radome material were specified as:
(a) Transmission loss ^ 7.5% over the frequency range of 120Mc to
(b) Refraction less than 10% for angles of incidence from 0° to 45°.
(c) Loss tangent of glass laminate = .005.
(d) Dielectric constant of glass laminate = 4.0.
(e) Loss tangent of foam core = .0005.
(f) Dielectric constant of foam core = 1.12.
Although tests at the manufacturer' s plant confirm the meeting of the
specifications on a flat 40"x40" sample of the radome, measurements indi-
cate that the actual radome does exhibit less than specified performance
particularly at frequencies above l.OOOMc. This can be easily explained
since the actual completed radome psssesses a curved surface and the nine
pieces comprising the total radome are assembled with beefed-up glass
laminate flange sections. These two conditions increase the detrimental
effects (increased diffraction and reflection).
As previously stated the frequencies of primary interest are in the
120Mc to 400Mc region. But some tests were performed in the chamber at
Normally, in evaluating the chamber a quiet zone is defined. That is, a
volume within the chamber in which known (measured) reflectivity levels
exist. Then in this zone antenna systems can be evaluated being fully aware
of the limitation of the chamber. Therefore, by definition, the quiet zone is
a 10 foot diameter sphere which is tangent to the chamber floor and centered
elsewhere within this chamber.
C . Advantages
Ground reflections are the most serious source of error when using a
conventional test range for measurements of low frequency, low-gain an-
tennas. The error may be refuced to a degree by additional height to the
towers, but towers over 100' high are expensive and rather inconvenient
to use. Likewise a conventional anechoic chamber designed for operation
at low frequency is expensive and requires a large building to provide even
a modest size test range. The vertical test range provides an attractive
compromise between tall towers and a large anechoic chamber. The cham-
ber portion need only be large enough to prevent reflections from the ground
and nearby reflecting surfaces and the tower need only be tall enough to hold
an antenna out of the near field of the antenna under test. Furthermore, the
length of the test range may easily be varied by adjusting the height of the
outside antenna. Operation of a vertical range is especially convenient. The
model under test, the test and control equipment, and personnel are all at
ground level and thus avoid the need for hoists and elevators.
HI. ANALYSIS OF CHAMBER
The magnitude of unwanted reflections that can be tolerated in an antenna
test range have been shown to be a function of the parameter being measured.
Since the site is never perfect, the results obtained may be interpreted in terms
of the known site imperfections provided the reflection coefficients and in some
cases phase are accurately specified. However, measurement of reflection co-
efficient of absorbent material is difficult and techniques for its evaluation have
not been standardized. Currently the Pattern Comparison Technique 1 and the
Free Space VSWR Technique 2 are favored in evaluation of r-f anechoic chambers.
In both techniques the significant result is the comparison of the incident to the
reflected energy from an absorbing wall to determine its reflection coefficient.
Briefly, in the Pattern Comparison Technique, the pattern of a directive
antenna (15 to 20 db gain) is measured successively at closely spaced points
along the radii of the chamber quiet zone. The quiet zone may be defined as the
volume within an anechoic chamber in which an antenna under test will be meas-
ured. Then the patterns are superimposed on each other with pattern peaks
coincident. The deviations in the patterns are read at different aspect angles
and VSWR curves vs aspect angle are constructed. The curves may then be con-
verted to reflection levels within the chamber.
The Free Space VSWR Technique is a method of continuously recording the
amplitude variations produced by reflections. Two directive antennas are used,
with one being moved continuously across the chamber quiet zone at a discrete
aspect angle for each recording. The amplitudes recorded are reduced to inci-
dent and reflected energy levels, thus allowing the reflection coefficient vs aspect
angle of a chamber to be determined.
From the brief summary above (and more so from the referenced literature)
it can be seen that both techniques rely on use directive antennas. This is not
a disadvantage in the usual situation where the chamber is large in terms of
wavelength and directive antennas are usually evaluated. However, in the
vertical test range more emphasis is placed on measurement of nearly omni-
directional antennas in a termination chamber that is electrically small,
i.e., approximately 1A. in depth and 2X. on a side. Although evaluation of chamber
performance using directive antennas will indicate both a direction and reflection
coefficient for sources of reflections, unless considerable effort is made to
integrate the reflection levels from all directions the chamber will appear better
than when used with an omnidirectional antenna. Therefore most of the analysis
of both the scale model and the full scale range was made using a dipole antenna
to probe the energy levels within the termination chamber. In this method, a
dipole antenna is moved throughout the quiet zone and energy levels vs position
recorded. The difference between the levels recorded and calculated free space
levels are converted into VSWR and finally a reflection coefficient computed.
It is common practice to try to reduce the large number of different reflection
coefficients that are measured in evaluating a chamber to a single number that is
then used to define the chamber's performance. In general, it will be found that
this number is not really a common denominator and that the relative perform-
ance of two different chambers should not be judged on this alone. For instance
the quiet zone of chamber A may be only 1/3 the volume that was included in
evaluation of chamber B, yet chamber B may be considerably better than A over
the same quiet zone. Unfortunately, the performance rating of anechoic chambers
is like the evaluation of radio receivers in 1940, i.e., not complete unless the
test conditions are known as well as the results.
A. Scaled Chamber Measurements
Reflection coefficient measurements were made in a 1/ 9th- scale chamber.
A photograph of the scale chamber is shown in Figure 6. Figure 7 defines
the test set-up. No attempt was made to design and test a l/9th-scale ra-
dome because the radome would appear as a thin wall structure in the fre-
quencies of most interest (up to around 400Mc) in the full scale chamber.
Although no scaled radome measurements were made in this particular
chamber design it is felt that general comments on scaled radome measure-
ments are in order.
Scaling a multi-panel, sandwich radome is, in general, difficult. Extreme
care must be taken in scaling the ribbing, flange design and radome curva-
ture. Serious errors can be expected if this precaution is not taken. Thin
wall structural radomes of dielectric constants of approximately 3 or 4 may
be scaled for measurements with confidence and also thick wall foam radomes
of low dielectric constants ( e n = l . 1 to 1.4) are practical for scaling purposes.
Measurements were made at frequencies from near l,080Mc to 3,600Mc.
This corresponds to full scale chamber measurement of approximately 120Mc
Figure 6-Photograph of 1 /9th Scale Chamber
to 400Mc. It is to be noted that a dipole probe is used inside the scaled
chamber to deliberately prevent discrimination against any reflected energy.
A directive type antenna would have been too selective.
The measurements at l,080Mc in the l/9th-scale chamber indicated a
reflection coefficient of 20 db and at 3,600Mc 30 db reflection coefficients
Figure 8 is typical of the standing waves measured by probing the
scaled chamber with dipoles. The dipole probes were moved in small
increments in terms of wavelengths from the aperture of the chamber to
the back. Figure 9 depicts standing wave curves resulting from probing
across the chamber in small increments. This data is typical of that ob-
tained in the 1/ 9th- scale chamber.
It was from these measurements that the full-scale chamber perform-
ance was predicted.
As was previously mentioned no attempt was made to determine the
effect of an A- sandwich type radome on the full scale chamber at scaled
frequencies. That the radome effect would be negligible at the low fre-
quencies of interest was predicted on the basis that the radome would
appear as a thin wall (approximately k/8 thick or less) at 400Mc and lower.
B. Full Scale Chamber Measurements
The quiet zone of the chamber was probed with a dipole using an
illuminating antenna which was directional in nature.
1/9 SCALE CHAMBER
f * 1080 mc
1/9 TH SCALE CHAMBER
S/A GAIN STANDARD
POSITION NO. I
REFL. COEFF ~ 21 db
POSITION NO. I
REFT.. COEFF ~ 31 db
POSITION NO. 2
POSITION NO. 2
REFL COEFF ~ 32 db
5 432 I 01 2345
POSITION NO. 3
POSITION NO. 3
Figure 9-Standing Wave Curves Measured Across 1/9th Scale Chamber Aperture
Specifically, at 125Mc and 400Mc the magnitude of the reflection coeffi-
cient was determined by moving a dipole horizontally inside the chamber in
directions normal to each wall and across the diagonals. With the dipole
fixed to a cart made of low dielectric constant foam material, the cart was
moved along tracks which could be oriented as desired within the chamber.
The track, cart and dipole support are shown in Figure 10. There horizontal
measurements were made at discrete heights of 2 (and/or 3), 4, 5, 6 and 8
feet above the floor of the chamber. The reflection coefficient in the verti-
cal direction was determined by measuring the standing wave as a function
of vertical movement of the dipole from 2 to 10 feet above the floor. Result-
ing standing wave curves are converted to an equivalent reflected energy
level (reflection coefficient). Figure 11 is a typical curve. The dashed
curve of Figure 11(b) represents the probed energy level as measured in
Figure 11(a) in the absence of reflected energy. The solid curve super-
imposed on the dashed curve represents the effect of the reflected energy.
Peak to peak value of the standing wave is 0.5 db. This standing wave of
0.5 db results from a reflected energy level of 30 db below the incident sig-
nal level. Or, Reflection Coefficient = 20 log VSWR-l/VSWR+1 . Measured
curves are shown in Figures 12(a) through 12(d).
At l,200Mc reflected energy levels greater than expected were measured
(approximately 32 db average). Also, the reflected energy levels were polari-
zation sensitive (i.e., E -perpendicular vs E-parallel) with differences in
Figure 10- Photograph of Track and Cart Arrangement
CHAMBER CROSS SECTION
r .5db (Q)
Figure 11 -Interpretation of Standing Wave Curves
UJ ii ii UJ
> m- Q> CC
1 ; i |
, liii I jl 1
! ' i !
Ii i !
1 i i I [ j ;
: I 1 i
3 1 I
1 ■ ■' ; "Tf" ^
3 ! 1
- 1 1| |
l| | !|
I i \\
ii 1 \\\\\\'\h
! ' I
i i 1
ul 1! Jl
Il II 1
! ' ' I
1 ! 1'-
1 :lli N 1 I
i i imjii!
i ! MM
I'm ' i ■
, ! :i!jiL !
! MM 1
1 ! '
I 1 ' 1
! 1 !
! 1 1
! ! !; ! Ii !
Hi 1; 1 !
i ii 1
« T -»0'
fl T -0*
8 D »90*
~" :"■.. v3-=^r*:---.
---"' i^""*^ --*-..
i ." ■
.- -|--- _.: .- .
z : . ■_■■ =■
:r-. z_t ,.-. 4^ ■=,-■ :T " "
I ~t- -\-L=. .. . _■
^.,,-^=- " ^ "
. -_:.:: :
_... ~ .^j::: z."
Z ty :
7.: --: /FfrT-J-r --."- ■ ■ "
> -"'t""^-k_ i .- —
^ .JJ- :
-'-■ X i _i__L" "tit
, : i -^ -; i ' ■ . ■** ' +■-
._ >i— ^~ ■ L* >»-* :**?!
!'--;.- : i i i i '
-- -j-- ■ : J-; -1- j- |
■■ -t-~-".^^f- " j "-P -^"
":. "i_T. t^=: n---: : p. ~:^-" i_. ___
r= >*r j-*^- 4"" T- *t^ "" :
; .=, =. =
-.- .-.|~ if? -.1- J^ :rl". r.
= =frr -4= =i=. J.| |
4 5 i
• d'IM- CMfF
is 7 ! --1--1 ""'' " i=J--=
-4- r j I [ 2 — u--J-h — i —
I 23456 78910
t* 125 mc
LOG PERIODIC SOURCE ANTENNA
ORIENTATION FOR S T -« D
(PLAN VIEW OF CHAMBER)
Figure 12b-Full Scale Chamber Standing Waves at 125 Mc, 400 Mc
UJ " 11 UJ
' • ! 1
i i :
, ' i ,
' ; ' i
1 , I
i ' '
. ; i 1 :
i i iii;
' ' i '
I 1 !
i 1 !
: !!i \\\
' i :
■■':\\ 1 J
: ' 111
1 1 I
! ;i ! !
! ' :
1 ; i
1 ! l i
!h! 1 I
ill 1 i 1
; i i 1
ill! i hi
i : ! ■
i ■ ■ 1"
'. ; j 1
'. i ■
, ; ■
I 1 :
; ii- :?*
' ! :
1 1 ' '
; 1 ,
j 1 ; '
: I II ' : '
■ i ! ■
j ! 1
: 1 i
i ' ;
i;i ! h
1 j -
n 1 ;ii|
l\ ! !
! 1 i '.
■ ! 1 !
; '! : hi;
i j! '
; ! j . ;
i i ■
■ ' ! i
. ; i ! j : i ;
] , i ;
1 : ; ' ' :
1 i ' l
! ; j !
; . ' ' ! !
' ' : ' ; !
! M \.
ill; !l! !
hi 1 1 1
! 1 1 i
i > 1
i j !>!,;!
8 T -90*
REFL. COEFF ~ 31 db
REFL. COEFF ~ 32 db
REFL. COEFF'- 31 db
-- t .-^.-+.- t -,-
7 - :-; = = e e
REFL. COEFF ~ 31 db
■■ ~ .
REFL. COEFF -> 39 db
REFL. COEFF'- 31 db
REFL. COEFF ~ 35 db
REFL. COEFF ~ 39 db
- :- l-
REFL COEFF — 32 db
REFL. COEFF ~- 30 db
REFL COEFF ~ 32 db
REFL. COEFF ~3ldb
8 T '9C*
REFL. COEFF ~3tdb
LOG PERIODIC SOURCE ANTENNA
ORIENTATION FOR S T ■ •„
(PLAN VIEW OF CHAMBER)
Figure 12d— Ful I Scale Chamber Standing
Waves at 125 Mc, 400 Mc
reflection coefficients between 6 and 15 db as a function of aspect angle.
This obvious deviation from the 45 to 50 db levels anticipated was attributed
to the radome for two reasons:
(1) The chamber at l,200Mc is electrically large (15 X. x 15X. 10\) and
therefore the chamber would be expected to be much better than at
400Mc (reflection levels %32 db).
(2) This A- sandwich radome (4 inches thick, with solid fiberglas panel
flanges) appears as a relatively large discontinuity.
Measured magnitudes of reflected energy for various linear polarization
orientations when probing the chamber at 125Mc and 400Mc are tabulated in
Tables 1 and 2. Some inconsistency is apparent, especially in the higher
coefficient values measured at 125Mc. This is attributed to instability of
the measurement equipment; however, retests of selected orientations indi-
cate no serious differences from the data presented or the conclusions
drawn from the data.
125Mc Chamber Reflectivity Levels
Dipole Height Above Floor
*Measurements performed at 123Mc using NASA
battery operated signal generator.
400Mc Chamber Reflectivity Levels
Dipole Heights Above Floor
Data taken at l,200Mc are not presented because later tests prove
that the radome is the major contributor to standing waves within the
chamber and therefore does not truly represent the capability of a vertical
To verify the fact that the radome was the major contributor to the
problem (i.e., larger reflection coefficients than anticipated and polari-
zation sensitivity) measurements were performed with the source an-
tenna inside the chamber just under the radome and compared to the
pattern taken through the radome (Figure 13(a)). Thus, an interference
pattern (see Figure 13(b)) between incident and reflected energy was
produced by essentially removing the radome from between the source
and receive antenna. In both cases a horn antenna was moved across
the chamber at a 45° angle (from the vertical). The +45° refers to the
horn antenna being tilted toward the chamber wall containing the wall
mount. And -45° refers to an angle toward the wall without the wall
Plots in Figure 14 compare the standing waves of the A- sandwich
radome for the two polarization conditions E-perpendicular and E-parallel
as defined below.
FREQ. - 1200 mc
PATTERNS MEASURED THRU RADOMEl
Figure 13a-Radiation Plots Showing Effect of Radome on Reflection Coefficient at 1200 Mc
- r --
-— 1 — 1
- < .. ~
;- : - ■;.
r E :
-" . ■ 7
-- _ "
LOG PERIODIC OVERHEAD
GAIN STANDARD HORN @ 45«
; E :
: : --
^J : IE
: "- : -
l7 : ;
— -_■- — .
i: :' ::
— _". -
D - [
-" _ -
. : -
— -.— -
- — -
--.: I :■.-
PATTERNS MEASURED UNDER RADOME
Figure 13b-Radiation Plots Showing Effect of Radome on Reflection Coefficient at 1200Mc
■"■ : ;
' , : .
' :: "
*- • - ■
-_- : :
- r ; :
— 1- -i
. : -- :
VSWR = 6 db
LEVEL = 9 db
VSWR = I '/ 2 db
LEVEL =12 db
REFL. COEFF = 33 db
f = l200mc
LOG PERIODIC ANTENNA OVERHEAD
Figure 14-Effect of Radome on Parallel and Perpendicular Polarization
€ >€ 2 >€\
Except at angles far off normal incidence reflected energy from E-
perpendicular is nearly always greater than from E-parallel because of
the zero at the Brewster angle occurring only with parallel polarization.
From Figure 14 it can be seen that the difference in reflection coefficients
is 14 db. The Brewster angle (where the reflection value goes to zero) is
defined as follows:
1/ e wher
e £ > 6
An important point which is illustrated in Figure 15(a), (b) is the degree
of accuracy with which the antenna radiation petterns can be measured
through a radome which exhibits significant reflection characteristics. In
this case the radiation patterns of an antenna (gain standard at l,200Mc) was
measured first in a free space pattern range method and then the measure-
ment was repeated inside the vertical chamber. To measure the pattern
" : -
Figure 15a— Pattern of Horn Antenna (E-plane) Measured Outside and Inside of Chamber
I1I!!!|! | ! || !!1I | I| | ' | II| 1T!I M^
ii I : J i i i tttrrttTrr i i'ii m i ni l mi hi i ii iii h iiiii i n il m i nnnm
inside the chamber, the antenna was rotated about a fixed axis and not moved
across the chamber. Therefore, the "available" reflections which exist
across the chamber are not probed and therefore, a rather respectable
pattern can be achieved. As a matter of interest Figure 15(b) shows the
effect the wall mount can produce in terms of reflected energy.
Measurements taken from the 1/ 9th- scale model of the test range are in
reasonable agreement with the measurements of the full scale range. The
structural A-sandwich radome definitely reduces the performance of the facility
at higher frequencies, but does not affect operation in the frequency range of
primary interest (125— 400Mc).
The concept of a vertical test range composed of an electrically small
termination chamber with a r-f transparent radome has been found feasible
and provides at moderate cost a convenient, quasi-all-weather, facility for
accurate measurement of antenna characteristics.
Considerable confidence may be placed on the results of measuring scale
models of anechoic chambers. Since the instrumentation is not difficult and the
cost is small it is recommended that more extensive use be made of scale models
to check chamber and anechoic material performance.
A multipanel sandwich radome has definite frequency limitations when
used as part of an antenna test range. Therefore it is recommended that a
thick, low-dielectric- constant foam radome be used.
1. Buckley, E. F., "Outline of Evaluation Procedures for Microwave Anechoic
Chamber", Microwave Journal, August, 1963.
2. Emerson, W., "Chamber Information", unpublished report of B. F. Goodrich
3. Jasik, "Antenna Engineering Handbook", New York: McGraw-Hill Book
4. Harvey, "Microwave Engineering", New York, Academic Press, 1963.
5. Electronic Space Structures Corporation, "Ground Radomes", 1964.