(Additional introductory comments are on
The original Andersen and Wolf memorandum on the Beverage antenna was technically
informative but had one or two grammatical errors. It is still quite good,
and reveals that a lot of work and thought went into the construction and analysis of these
systems. The newest articles about Beverage antenna construction include
such things as remotely-variable termination resistance or the use of walkie-talkies
to communicate from one end of the antenna to the other. Evidently these options
(along with spell-checking word processors) were not available to the FCC engineers
In the HTML version below, grammatical errors have been retained but
are marked with an asterisk (*). The formatting of the original is retained as
much as possible; however, the original pages were not numbered as shown below.
By the time the memorandum came to me, it was a copy of a copy (of a copy...) and the quality
was good but not great. There were a few places where the letters were garbled, so I
took the opportunity over the last few days to transcribe the entire thing into text files,
and then into HTML, as a public service. If you would like to check
the pages of the original document, please refer
to Beverage Page 1.
The graphics on this page are in the form of PNG files. If your browser has a
problem with PNG files, the same page is
available here with ordinary JPG files, which
are not quite as clear.
MEMORANDUM ON THE BEVERAGE WAVE ANTENNA
FOR RECEPTION OF FREQUENCIES IN THE
550 - 1500 KILOCYCLE BAND
The theory of the Beverage Wave Antenna is covered in a paper entitled "The Wave Antenna"
by H. H. Beverage, O. W. Rice, and F. W. Kellogg, presented at the
Midwinter Convention of the American Institute of Electrical Engineers at New York, February 14-17, 1923.
This memorandum is intended to outline some of the practical problems experienced in the
erection and operation of this type of antenna for service in the Broadcast 550-1500 kilocycle band,
at the Federal Communications Commission Central Monitoring Station at Grand Island, Nebraska.
Among the desirable properties of the Wave Antenna for monitoring purposes are:
(a) Delivers a stronger signal over the entire band than a good simple antenna.
(c) Atmospherics and industrial electrical interference considerably reduced
especially when the source is in a direction other than that
of the received signal.
(d) Low cost, long life, and unlikely obsolesence.
The optimum length for a broadcast band wave antenna is approximately 1800 feet. This length
delivers peak signal strength at closely 550 kilocycles and again at one-half its wave length in
meters, or about 1100 kilocycles. The peaks are, however, rather broad and the signal delivered is
considerably stronger than that from a good simple antenna throughout the entire band. Near the
antenna peaks the increase in microvolts to receiver may reach more than 400 percent.
Where space is a consideration, the length can be reduced to 1400 or even 1000 feet, but the
signal strength delivery and directivity will be proportionately reduced, and at less than 1000 feet
the slight advantage of a wave antenna for the broadcast band over a good simple antenna does not
warrant its erection.
The surge impedance of the wave antenna is determined by its height above the ground, and by
soil conditions with regard to moisture, etc. By erecting the conductors at a minimum of ten feet
above the ground, the surge impedance remains more nearly constant during all seasons than when
erected at a lower height. If erected at a height greater than fifteen feet the pickup of the vertical
leads at the far and near terminals may considerably reduce directive properties. If a greater height
is required at gates or other passages, the higher poles should be erected at the sides of the opening
and the conductors brought down vertically, proceeding at the selected height after the high point
has been bridged. Reasonably uniform height of the conductors throughout their length is, of course,
The standard practice of telephone line construction is followed. The poles should be spaced
approximately one hundred feet apart and for mechanical strength the conductors should be #12 B & S
hard drawn copper. The line should generally follow the earth's contour, but small knolls are
disregarded and the tops of the poles after planted* may be trimmed for general or even gradient of the
For single, uni-directional reception or single conductor antennas, the conductor can be mounted
on ordinary pole brackets attached to the poles.
For uni-directional reception from front or rear, or both, the two-conductor type of construction
is required. The conductors are mounted on short, standard cross arms without braces, with the
locust insulator pins spaced 16" between centers. This with ordinary telephone line glass insulator[s]
gives the desired or adopted 18" spacing of conductors of the finished antenna.
For maximum efficiency the resistance of the ground system at both the far and near ends of the
antenna should approach zero, or at least be less than thirty ohms. The ground resistance is best
determined by the voltmeter-ammeter, alternating current method, but reasonably satisfactory results
can be obtained by the use of the battery and voltmeter voltage drop method.
For measurement of a single ground the planted conductors are divided into equal halves and
measured, and the result divided by two, or the far and near terminal resistances measured through
the antenna conductors for the combined resistance of the two terminals or the series resistance of
the entire system.
If the voltage drop method of measurement is used and should polarization or other direct current
effects produce absurd readings such as negative resistance, a resistance of 100 ohms or more may
be placed in series with the circuit and two measurements taken with changed polarity. The average
of the two recorded values divided by two in the case of a divided system, less the added external
resistance may then be considered a reasonable approach to the actual resistance. Due to polari-
zation, readings of instruments should be taken at the moment of contact.
If difficulty is experienced in obtaining proper ground resistance values, the constructor is
referred to U. S. Bureau of Standards' Technological Paper #108, issued June 20, 1918.
Single Conductor for Forward Reception
A. Signals from A direction are dissipated either completely or partially in Rb.
B. Direction of reception.
Removal of Rb permits the antenna to be used bi-directionally or forward and rear.
Two Conductors for Forward Reception
All the factors are the same as the single conductor type except that the surge
impedance will be lower as a result of the two conductors in parallel.
Two Conductors for Rear Reception
A. Direction reception.*
The signal builds up until it reaches B where the phase is reversed by
grounding one of the conductors and leaving the other free after which it is reflected
back to T using the antenna conductors as an untransposed transmission line. The
signals from B direction are dissipated either completely or partially in impedance
R connected from the center tap of T to ground. It will be noted that Figures 2
and 3 are the same except with respect to far and near terminal connections. When
it is changed as shown, the antenna can be made uni-directional for either forward or
rear reception, but not for both forward and rear reception simultaneously.
When the wave length of the signal to be rejected is a multiple of one-half wave of the length
of the antenna, it is either completely or largely absorbed in resistances Rb or Ra. Odd multiple
frequencies of one-quarter wave length of the antenna length deliver a greater residual or undesired
signal to the receiver.
In order to balance out an undesired signal originating at an angle of more
than ninety degrees from the source of a desired signal originating in direction of maximum
reception of the antenna, part of the undesired signal is reflected back to T in proper
phase and magnitude to cancel itself out. This is accomplished by the insertion of
a tuned circuit in series with a variable resistor at the far terminal, as in Figure 4 for
rear signal rejection, and as in Figure 5 for forward signal rejection.
Two Conductors for Forward Reception
Two Conductors for Rear Reception with Rejector Circuit
The system of changing directivity of the antennas and remote erasure
adjustments as outlined is somewhat inconvenient because of required travel, terrain
condition, time consumption and the fact that the antenna when made directive in one
direction is not available for service in the opposite direction.
To overcome these undesirable features, a modified arrangement of the
terminal coupling units and rejection system may be employed which permits the use
of a two-conductor wave antenna for uni-directional reception both forward and rear
simultaneously on the same or different frequencies with the band for which the antennas
are designed, and with all variable factors under control at the receiver. The
circuit is shown in Figure 6.
Two Conductors for Simultaneous
Forward and Rear Unidirectional Reception
The signals are delivered from the near antenna terminal coupling transformers
to a terminal rack through two 400-ohm transmission lines connected to a Graybar #223-A,
three-point, switchboard jack. The receiver input lead connects to the terminating jack
through a short length of good quality lamp cord and a Graybar 3-A phone cord plug. A fixed
one-watt resistance unit of proper value is connected across points 1 and 2 of the Graybar
jack which automatically connects across the transmission lines when the receiver
plug is out.
In the event the erection of more than one wave antenna may be
contemplated, the question of inter-coupling may arise. Tests covering this feature
at Grand Island using a standard signal generator indicated that broadcast band wave
antennas may be crossed within a few feet of one another at angles of sixty degrees
or more without detrimental coupling effects or may be safely parlleled when
spaced four hundred feet or more; in either case without noticeable or excess coupling.
When the antenna conductors run parallel over wire fences with non-conducting
supports, the fence wires, to avoid interaction, should be broken with insulators at even
lengths of about seventy-five feet.
Conductors such as fence wires and metal posts rubbing against one another
within a hundred feet or so of the antenna, may be assumed to be a source of noise
generation in the antennas.
During some weather conditions such as snow or dust storms, or summer
electrical storms, voltages sufficiently high to break down unprotected coupling coil
insulation are developed in the antenna system. To prevent transformer damage
from all but direct lightning strikes, L. S. Braasch #270 neon-argon tube arresters
or their equivalent with breakdown range of 200 to 300 volts may be connected to the
two antenna terminals and ground at both far and near ends of the antenna; also to
both terminals of the transmission line from antenna to receivers when the length
exceeds two hundred feet. The ordinary 1-watt I.R.C. metallized resistance units
as used at Grand Island for terminal impedances will almost always be found open
circuited after each electrical storm occurring in the vicinity of the antennas
whether or not protected by arresters.
Where it is not practical to erect the antenna with the near terminal
direct to receiver location, it can be located at any distance up to a half-mile or
more from receiver location and the signals brought to the receiver by transmission
lines without noticeable loss. For long stretches the four parallel #14 B & S conductor
type of transmission line is preferred. For distances of 100 feet or less requiring
no intermediate supports the two conductor transposed line may be employed. Two or more
transmission lines may be mounted on the same poles or other non-conducting supports
when the separation equals or exceeds ten times the spacing of the transmission
A coupling transformer with astatic shield is required at the receiver
to keep the transmission line balanced and prevent possible pickup of the transmission
line getting into the receiver.
When this resistance unit does not itself absorb or reject undesired
interfering signals delivered to the receiver while the antenna is in service in the
opposite direction, a shilded L.C.R. circuit, Figure 7, also terminated with a short
length of lamp cord and a Graybar 3-A plug is inserted in the opposite reception jack
of the antenna and by manipulation of C and R the interfering signal can generally be
largely or completely erased without reduction in strength of the desired signal, when
the interfering signal is more than 90° from the direction of maximum reception of the
antenna. It is not as effective for signals predominately* sky wave because of their
varying phase and intensity.
Some casual experiments and observations in service indicate that by
disconnecting and grounding one or the other of the conductors of a two-conductor
antenna at the station terminal, the forward reception pattern can be changed
sufficiently to permit partial or complete erasure of interfering signals
originating thirty or more degrees from either side of the antenna, depending upon
which conductor is grounded and without noticeably affecting the strength of the
desired signal originating more nearly directly forward or from an angle opposite
the grounded conductor.
This seemingly unreasonable circuit connection is, apparently, more
effective during afternoon hours or for several hours before local sunset. It
is frequently not effective at all.
In some instances forward interfering strong signals can be more
completely balanced out by connecting a simple or general purpose antenna to one
of the terminals of the signal rejector for clear reception of weak signals on the
same frequency originating in the rear.
In order to obtain a better balance at the far end, and to partially
or completely eliminate pickup by the vertical ground lead at both far and near
terminals, a beverage reflection transformer, Tr, Figure 6, is employed and a lead
covered #12 B & S conductor used as a shielded vertical ground connection. The
conductor with lead shield is made water tight at the bottom by soldering conductor
and shield together and connecting the whole to the underground radials. The
transformer housing is grounded to the upper end of the lead shielding and the coil
terminal to the shielded conductor. This arrangement is an apparent improvement
in directivity over the exposed, directly grounded, vertical section of the
TEST FOR TRANSMISSION LINE BALANCE
The transmission line may be checked for signal pickup or unbalance by
disconnecting the two antenna wires and connecting a resistance across the transformer
in place of the antenna and equal to the surge impedance of the two antenna
conductors. Very little pickup of signal or noise should register in the
receiver from a well-balanced line.
TEST FOR TRANSFORMER BALANCE
The center-tapped coupling transformers may be tested for balance by
the use of a signal generator and receiver connected as shown in Figure 8. When
properly balanced the receiver will indicate nearly zero input or a very small tranfer
of energy to the secondary of the transformer.
Circuit for Testing Transformer Balance
Impedances Ra and Rb in Figure 6 should be equal to the surge
impedance of the transmission line as calculated from:
(s equals spacing from center of wires)
(d equals diameter of wire [d and s in same units])
The transformer Ta couples the transmission line to the two-conductor
antenna and the two wires are now acting only as transmission line to carry the
reflected signal from A direction to the receiver. The impedance of the
two-conductor antenna serving as a transmission line is calculated from the
same formula. The Grand Island transmission lines are approximately
400 ohms and the antenna conductors 700 ohms.
The transformer Tb couples the transmission line to the antenna
with the two wires acting as if they were in parallel, because the primary of
Ta has practically no effect on signals arriving in phase at A from B
direction. Therefore it is necessary to know the surge impedance
of the two wires in parallel with respect to ground.
The transformer Tr reverses the phase and sends the signal which
arrives from A direction back over the two-wire line now acting only as a transmission
line. The primary coil is designated* to match the surge impedance of the two-conductor
antenna line in parallel with respect to ground and the secondary to match the same
two-conductor antenna as a transmission line back to the station.
Circuit for Determining Surge Impedance Forward Reception
A signal is tuned in from A direction on the receiver. Frequencies
of multiples of one-half wave length are preferred as the are more completely balanced
out when the proper resistance value is used at Rb. The resistance Rb is adjusted
for minimum signal in the phones and this value used as the antenna surge
impedance. The surge impedance may also be determined by adjusting the resistance
at the opposite end of the antenna.
Circuit for Determining Surge Impedance Rear Reception
In this case signals from B direction are tuned in to the receiver
and balanced out by varying Ra. This method will usually give a lower value
of Ra than that recorded for Rb in the other case. The difference is due to
the R.F. resistance of the primary of the transformer and some residual inductive
reactance, as unity coupling between turns is not possible. If the recorded
value of resistance is low it is an indication of high ground resistance. Values
at Grand Island for two wires in parallel are approximately 300 ohms.
Another and more simple method which has been employed is to make the
resistance Ra, Rb variable in steps of 20 ohms from 230-440 ohms. With an
observer at the reciever, the readings are recorded as changed at the antenna terminal
and as indicated on a meter in the second detector plate circuit of the receiver for
each changed step of the resistance. The observer at the receiver hears a click
when the resistance is switched to the next step. Its value is known according
to a shifting plan as prearranged with the antenna terminal collaborator and the proper
impedance value thus determined.