Somehow, after being in the RF business for four decades, I have
to admit to not being familiar with the term 'acceptance angle'
for antennas. That is after having read scores of articles on antennas.
Maybe I did and just don't remember - embarrassing. Acceptance angle
is mentioned and explained in this article during the description
of rhombic antenna characteristics versus dipoles and multi-element
designs. Although the author focuses on television installations,
information provided on signal reflections, shadowing, ghosting,
multipath, etc., is applicable to radio as well.
Rhombic Antennas for Television
By Woodrow Smith
Author, "The Antenna Manual"

For fringe and bad ghost areas the rhombic antenna will always
outperform an ordinary dipole array.
The rhombic antenna has for many years been a favorite for high-frequency
sky-wave transmission and reception. This is explained by its simplicity
of construction as compared to dipole arrays having comparable gain,
by its broadband characteristics, and by its sharp, unidirectional
directivity pattern throughout its broad frequency range.
The same characteristics which recommend the rhombic array for
high-frequency sky-wave applications make it an ideal TV receiving
antenna for use in fringe areas or bad ghost areas when sufficient
room for erection is available, particularly when several stations
lie in very nearly the same direction.
Even when the desired stations do not lie in approximately the
same direction, the rhombic array often will be found useful in
providing a usable picture on one channel at distances so great
that an ordinary dipole array will not perform, or in providing
a useful picture on one channel in mountainous areas where ghosts
are so bad that a ghost-free picture cannot be obtained on even
one channel with an ordinary dipole array.
Rhombic Characteristics
Unlike an ordinary dipole array using one or more parasitic elements,
a rhombic array exhibits a good front-to-back ratio over a wide
frequency range and also has a very narrow acceptance angle off
the front side. This means good rejection of ghost-producing reflected
signals on all channels, both from the side and off the back.
The wide acceptance angle of conventional dipole arrays makes
them highly vulnerable to ghost-producing echo signals arriving
obliquely from the front side. Contrary to widespread belief, ghosts
are not necessarily produced by echo signals arriving from the back
side of an antenna; they often arrive from the front.
Front side echo signals sometimes are not apparent as ghosts,
because they may have a comparatively short delay time. Under such
conditions a separate image is not discernible; the echo signal
simply degrades the definition of the picture without producing
a separate image.
A high-frequency rhombic array may be designed for effective
sky-wave transmission or reception over a frequency range as great
as 4 to 1. While the vertical angle of maximum radiation or response
of such a rhombic increases considerably with decreasing frequency
over such a wide frequency range, higher angles become effective
as the frequency is lowered. Therefore, the change in vertical directivity
is not particularly objectionable for sky-wave applications.
The situation is different, however, in the case of TV reception
or other v.h.f. ground wave applications. The only effective vertical
angle is that of the angular elevation of the horizon at a point
where the arriving wave passes over it. There is only one useful
vertical angle, and this angle does not change with frequency. The
only useful gain is that which occurs at this angle.
As a result, the useful frequency range of a rhombic array designed
for TV reception or other v.h.f. applications does not exceed approximately
2 to 1, and preferably the range should not exceed 1.6 to 1. This
means that it is not possible to design a rhombic array which will
provide near-optimum performance on both the low and high television
bands. The ratio of 216 mc. to 54 mc. is 4 to 1, and any attempt
to cover this range with a single rhombic array will result in mediocre
performance over much of the range.
Where two-band coverage is re-quired, a high-band rhombic can
be strung inside a low-band job from the same poles. The separation
will be sufficient to avoid undesirable interaction. For short runs
separate feed lines and a suitable switch should be employed. For
long runs a d.p.d.t. relay can be placed at the antenna end and
a single line run from the relay to the set.
For FM
For an excellent DX FM receiving antenna, simply double the dimension
s given in Fig. 1 for the high TV band.
Design Dimensions
Design data are given in Fig. 1 for four rhombics: (1) a long-leg
rhombic for use on the low band for maximum gain and directivity
where space permits; (2) a short-leg rhombic for use on the low
band when space restrictions will not permit a long-leg rhombic
or where less horizontal directivity is desired due to a slight
spread in the station directions; (3) a long-leg rhombic for use
on the high band for maximum gain and directivity; and (4) a short-leg
rhombic for use on the high band when less directivity is desired
or when it is desired to hang the array from a single pole and two
cross arms, or from two poles and a spreader. The latter array is
small enough to be mounted on an amateur beam antenna rotator.
The long-leg rhombics are four wavelengths on a side at their
"design center" frequency, have a gain of approximately 10 db.*
over a matched half-wave dipole (varying slightly over the band),
have a useful beam width or acceptance angle of approximately 8
degrees (varying slightly over the band), and exhibit excellent
ghost rejection (azimuthal discrimination) throughout their frequency
range.
The short-leg rhombics are two wavelengths on a side at their
"design center" frequency, have a gain of approximately 7 db.* over
a matched half-wave dipole (varying slightly over the band), have
a useful beam width or acceptance angle of approximately 13 degrees
(varying slightly over the band), and exhibit good ghost rejection
(azimuthal discrimination) throughout their frequency range.
The low-band arrays employ 68 mc. as a design center, and the
high band arrays employ 194 mc. as a design center.
*When comparing gain figures, keep in mind that the high gains
claimed by some antenna manufacturers in their advertising are highly
optimistic." A very elaborate dipole array is required for a gain
of more than 10 db. over a single matched dipole, and the gain of
such an array falls off at a comparatively rapid rate for departures
from the design frequency.

Fig. 1 - Rhombic antenna design characteristics.
Antenna Patterns and Ghost Rejection
As a measure of vulnerability to ghosts, the common figure of merit
for ordinary dipole arrays having a wide acceptance angle is the
"front-to-back ratio" on each of the channels under consideration.
For an array having an acceptance angle of only a few degrees, however,
we are interested in the relative response in all directions outside
the main lobe, regardless of whether it is off the back or off the
front. For this reason, "azimuthal discrimination" is a more appropriate
term than "front-to-back ratio" when referring to a rhombic array.
Like almost all large, high-gain arrays, a rhombic array exhibits
various "minor lobes," and it is the ratio of the amplitude of the
main lobe to that of the various minor lobes that determines vulnerability
to ghosts. (See Fig. 4.) Generally speaking, as the legs of a rhombic
are increased in length and the included angles are maintained at
optimum values, the minor lobes become more numerous, sharper, and
lower in amplitude (compared to the amplitude of the main lobe).
Strictly speaking the response of a large rhombic to a signal
whose modulation envelope changes very rapidly with time, such as
a television video signal, is not quite the same as for a steady
carrier or a signal containing only low modulating frequencies.
The effect is insignificant for a wave arriving "head on," because
a wave propagated from the fore end to the aft end of the rhombic
via the wires travels not more than about one wavelength farther
than the direct distance between these two points, and even on Channel
2 this difference is only about 18 feet.
For television signals arriving obliquely or from the back the
effect is no longer insignificant, particularly in the case of a
long-leg rhombic cut for the low band. However, the effective azimuthal
discrimination will compare closely to that which obtains under
unmodulated conditions and, for practical purposes, may be considered
to be the same as for an unmodulated signal. It should be pointed
out, however, that if an attempt is made with a low-band, long-leg
rhombic to receive a nearby TV station off the back side, the picture
quality will be poor even though the received signal (due to the
transmitter proximity) is of good strength.
If a rhombic array is placed well above surrounding objects it
is not necessary to "probe" the available area for an optimum location.
This is explained by the sharp directivity pattern (making the array
rather insensitive to phasing reflections from nearby surrounding
objects) and by the fact that the rhombic is spread over a considerable
area as measured in terms of wavelength.
While the exact location of the array is not critical, so long
as it is "in the clear," the direction in which the array is to
be pointed must be determined very precisely, particularly for the
long-leg jobs. It is for this reason that dimensions for still longer
legs are not given in Fig. 1. While even greater gains may be obtained,
the beam width becomes embarrassingly narrow, making the array difficult
to orient. However, if you are sure you can get the array "dead
on" and want to pick up another 2 db., here are the dimensions for
legs six wavelengths long at the "design center" frequencies: Low
band - L, 86' 5"; S, 64' 7"; D, 160' 2". High band - L, 30' 7";
S, 22' 11"; D, 56' 8".

Fig. 2 - Rhombic antenna construction details.
Flat Terrain Installations
If the antenna is to be located in open, flat, or rolling country,
and the angular elevation of the horizon is practically zero, then
the orientation is fairly simple, and the height of the array is
not especially critical (though as much height as practicable is
desirable). The array should be pointed in the exact direction of
the transmitter. This can be determined by means of a suitable map
and an accurate compass. (Obtain the magnetic declination for your
area from a surveyor, if this is not already known.)
If the transmitter is on a peak which can be located with a pair
of glasses: or a telescope, the problem of orientation is still
simpler. The orientation of the short-leg rhombics of Fig. 1 should
not be off more than about three degrees and that of the long-leg
rhombics of Fig. 1 more than about two degrees if maximum performance
is to be obtained.
In open, flat, or rolling country, the higher the array the better
(within the range of practical pole heights) but there is not much
profit in going above about thirty feet for the high-band rhombics
or more than about sixty feet for the low-band rhombics unless it
is necessary in order to get the array well in the clear with regard
to surrounding objects, particularly objects in front of the antenna.
The law of diminishing returns applies, and it is up to the individual
how much pole expense is justified. The higher the antenna the better,
but the higher the antenna the less difference another ten foot
length makes, and the harder it becomes to obtain another 10 feet.
Shadowed Locations
When the receiving location is in comparatively flat country
but is separated from the transmitter by a range of mountains or
high hills some distance away, in such a manner that the angular
elevation of the horizon in the transmitter direction is more than
about three degrees on the high band or more than about ten degrees
on the low band, then one must be careful not to get too much height.
Under such conditions the height at which maximum signal strength
occurs (and above which it falls off) will come within the range
of practical pole heights, and high poles may provide too much height.
If the angular elevation of the horizon exceeds the above limits,
it is a good idea to lower the array on the poles a few feet to
see if the signal strength drops off. If it increases instead, then
the array should be run up and down the poles to find the optimum
height.
When the angular elevation of the horizon in the direction of
the transmitter exceeds approximately eight degrees for a long-leg
rhombic or approximately twelve degrees for a short-leg rhombic,
it also is a good idea to try elongating the array by increasing
the dimension D of Fig. 1 a certain small percentage while checking
signal strength, dimension S being decreased accordingly to allow
for the elongation. (Elongation raises the elevation angle of the
main lobe.) In installations where the angular elevation of the
horizon exceeds the aforementioned limits, the distance between
the fore and aft poles should be made about fifteen or twenty per-cent
greater than the distance D given in Fig. 1, to allow for experimental
elongation of the array.
When sufficient room to permit experimental elongation of the
array is not available, the array should be tipped upwards so that
an extended line through the fore and aft apices of the array would
intersect the horizon. This requires that the front pole be higher
than the center poles and that the rear pole be lower than the center
poles (assuming level ground). The array must be kept in a flat
plane, even though it is tipped upwards.
Because of the sharp vertical directivity of the array, one of
these two expedients is required for maximum performance whenever
the angular elevation of the horizon exceeds the aforementioned
limits. The dimensions given in Fig. 1 are for maximum response
at zero elevation with the array lying in a horizontal plane. Therefore,
for good response at an angle much above zero, the array must be
either elongated or tipped upwards. Elongation gives slightly better
results and is the preferred arrangement.
Hilly Country
When the array is to be located in hilly country or down in a
canyon, it is not safe to orient the array in azimuth simply by
aiming it at the transmitter. The dominant signal may be taking
a devious route. The safest procedure under these circumstances
is as follows.
Using four low, temporary poles and some willing assistants,
determine from which compass direction the main signal is arriving.
During this operation keep the S and D dimensions nailed down by
tying strong string between opposite apices. After the signal direction
is determined, layout the location of the permanent poles, allowing
for experimental elongation of the D dimension if there is room.
Then proceed as before, checking to see if greater signal strength
can be obtained by lowering the antenna. If so, optimize the height
and then try either elongating the array in the D dimension or tipping
the front of the array upwards, as previously described. This may
sound as though a lot of trouble were being taken, but it is necessary
in a hilly receiving location in order to insure maximum performance.
By following this procedure good pictures have been received in
what were considered "impossible" locations.
Occasionally when the terrain is very hilly and the spurious
reflections very bad, the discrimination of a rhombic is not sufficient
to eliminate a ghost coming in on a minor lobe. If this is the case,
try varying the D dimension slightly either way (at the expense
of the S dimension). The numerous nulls can be steered over a narrow
arc in this manner, and usually one can be lined up on the troublesome
ghost without affecting the main signal.
The Terminating Resistor
For proper operation a rhombic array must be terminated in a
substantially non-reactive resistance of approximately 800 ohms.
Satisfactory operation will be obtained simply by connecting in
series two 390 ohm metalized resistors of the insulated, hermetically
sealed type (such as an IRC type BTA), shown in Fig. 2. Two in series
are preferable to a single resistor having twice the resistance,
for reasons which need not be discussed here.
Care should be taken to make sure that the resistors used are
not of the wirewound type. In the low resistance range, 1 watt resistors
are available in both metalized and wirewound types, and the two
cannot be told apart by inspection except by type number. If in
doubt, break one open to see.
The Feed Line
Only about 10 per-cent loss in signal voltage will result if
a rhombic array is fed directly into a 300 ohm line without benefit
of a matching transformer. Therefore, while it is possible to construct
a matching arrangement which will result in a precise match, the
improvement hardly can be considered worth the trouble.
If the feed line must pass through a region of high ambient noise,
and it is desired to employ 75 ohm coax for lead-in, it can be done
with the aid of one or two of the broadband 300/75 ohm balanced-to-unbalanced
transformers now on the market (as manufactured by The Workshop
Associates and the J. W. Miller Mfg. Co.) A transformer is connected
between the line and the array as shown at Fig. 3A. The device should
be made water-tight or protected from the weather. If the receiver
does not have provision for 75 ohm input, another transformer should
be employed at the set end of the line, as shown at Fig. 3B.
At Fig. 3C is shown how 300 ohm ribbon or an open wire line may
be employed with a set having only 75 ohm unbalanced input (such
as certain receivers employing the Du Mont "Inputuner").

Fig. 3 - Three applications of a broadband impedance
transformer and line balancer (T) in conjunction with a rhombic
antenna installation. The device is used at the antenna end (A)
to permit use of a coaxial line in locations where the line must
run through a region of high ambient noise. It is employed at the
set end (B) to match coaxial line to a receiver having only 300-ohm
input. At (C) it is used at the set end to match 300-ohm ribbon
or an open-wire line to a receiver having only 70- to 75-ohm input.
Using Open Wire Line for Lower Losses
Where difficulty is experienced with 300 ohm ribbon close to
the ocean becoming very lossy after a few weeks, or where it is
necessary to run 1000 feet or more of line to put the antenna in
the clear (as in some mountain locations), line losses can be reduced
greatly by substituting an ordinary open wire line constructed of
No. 14 or No. 12 copper wire spaced two inches. Suitable line spacers
are available at radio parts stores.
This type of line is particularly recommended for use on the
low band but is still useful on the high band. Somewhat closer spacing
is desirable on the high band, but suitable spreaders are not readily
available and must be made up. It is best to stagger the distance
between spreaders a little, to make sure that all or several do
not fall upon exact multiples of an electrical half-wavelength on
a desired channel. (This will result in an undesirable effect called
"periodicity.")

Fig. 4 - Horizontal directivity and gain pattern
(voltage across 300 ohms) of a typical long-leg rhombic array and
of a typical stacked dipole array using parasitic reflectors cut
to the channel. The rhombic will maintain substantially the same
gain and discrimination over considerable frequency range. The dipole
array will not.
An open wire line constructed as above will have a surge impedance
of from 450 to 500 ohms, or somewhat above the standard 300 ohm
receiver input impedance. However, this is nothing to worry about,
because as the impedance of the line is raised, and the match becomes
worse at the set end, the match improves at the antenna end of the
line. As a result, it is possible to employ a 450-500 ohm line without
running into transmission line ghosts, loss of picture detail, or
increased loss due to impedance mismatch.
Two examples of the improvement to be expected over 300 ohm ribbon
used under unfavorable conditions can be cited. Substituting an
open line for 160 feet of ribbon exposed to beach weather for four
months resulted in an increase on two low band channels of more
than 10 db. After six months the performance of the open wire line
had not deteriorated enough to notice.
In another case an. open line was substituted for 1100 feet of
300 ohm ribbon newly installed and not providing adequate signal
strength. The location was not near the ocean. The measured increase
in signal at the receiver on Channel 5 was approximately 12 db.
(measured in dry weather), a very worthwhile gain. Wet weather proved
to have little effect upon the performance of the open line, which
is more than could be said for the ribbon line. It should be kept
in mind, however, that for a comparatively short run not near the
ocean, ribbon or tubular 300 ohm line is entirely satisfactory.
Where noise pickup by the line is not a serious problem, the
same order of improvement will be obtained with an open wire line
on the high band when the installation is near the ocean or the
run is very long.
When expense must be kept down, plastic curlers of the type sold
for use with home permanent wave kits may be used for spacers. Get
the clear ones rather than the colored.
Somewhat more care must be taken with an open wire line to avoid
sharp bends and to keep the line away from objects as much as possible.
Nylon cord or fish line may be used to support and position the
line where necessary. The feed line in any case should leave the
antenna symmetrically for at least six feet on the high band or
at least fifteen feet on the low band before making a bend to the
right or left. Contrary to popular belief, little if any reduction
in noise pickup will be realized by twisting or transposing the
line.
Construction Pointers
It is recommended that wood poles be used to support the array
unless they are spaced several feet from the apices. Guy wires should
be well broken up where they run within a few feet of the antenna,
or else rope guys should be used. All joints should be soldered.
Posted September 25, 2015
|