is as prevalent today as is was at the beginning of radio communications
- maybe more so since there are typically more reflective obstacles
between the transmitter and the receiver. The evidence of multipath
with digital communications, be it voice, Internet page viewing,
or movies, is slowness in transmission as opposed to the analog
case where garbled speech on radio and ghosting pictures on television
are the evidence. High data rates with digital transmissions typically
mask the packet errors and their necessary re-transmission; it all
happens before the buffered information is presented to the listener
/ viewer. However, there is no such buffering of over-the-air radio
or TV transmissions so evidence of multipath is immediately noticeable.
I remember how sometimes simply having a large metal trash or delivery
truck roaming through the neighborhood was enough to cause the
Leave It to Beaver (and other) program's pictures to lose
sync. Now, as then, one of the best ways to mitigate the issue is
use of a directional antenna.
March 1957 Radio & Television News
Wax nostalgic about and learn from the history of early electronics.
See articles from Radio &
Television News, published 1919 - 1959. All copyrights hereby acknowledged.
Problems in Metropolitan TV Reception
By Simon Holzman
Chief Antenna Engineer JFD Manufacturing Co.,
Fringe areas have no monopoly on difficulties with picking up good
signal. Overload, ghosting, and interference often plague prime-signal
Fig. 1. The JFD "Super Helix" is an example of an antenna
with a narrow polar angle (see Fig. 4B) that weakens ghosts.
While considerable attention has been given to
the reception difficulties encountered in areas distant from TV
transmitting points, one sometimes forgets that television installations
in metropolitan and other primary areas often suffer from problems
peculiar to these locations. Noteworthy among these problems are
those of ghosting, graininess, the effects of too much (or too little)
signal, cross-modulation, and interference from a number of sources.
In some cases, these conditions can be completely controlled. In
cases where they cannot be entirely eliminated, they may be greatly
minimized, as a rule, resulting in satisfactory reception.
In many localities, ghosting is a serious deterrent to acceptable
viewing. This phenomenon is caused by the simultaneous reception
of out-of-phase signals on the same channel. High-frequency TV waves
are capable of being reradiated (as in radar) by metallic bodies.
The body reradiating these waves acts as a secondary transmitting
antenna and effectively "bounces" the signal in a different direction.
The action of two such reflecting bodies is illustrated in Fig.
3. Since the bounced wave and the original signal travel different
distances before arriving at the receiving antenna, they are out-of-phase.
The phase differential governs the degree of separation of the two
pictures on the screen.
To understand the cure, let us investigate
two properties of receiving antennas. The first of these, the front-to-back
ratio, is the ratio of the received voltage from the front of the
array to the signal received from the rear, given equal power densities
in both cases. The higher the front-to-back ratio, the less signal
the antenna will receive from reflections coming in from in back
The second property is the antenna's polar response. This has to
do with the directivity of the antenna with respect to a signal
from one source. It is determined by the angle formed between the
direction or position in which the antenna is most sensitive (forward
pickup) and the direction or position in which it receives 3 db,
or 30 percent, less signal from the same source. This may be measured
by beaming a constant signal toward the antenna while the latter
is aimed directly at the transmitting source, measuring the signal
voltage generated in. the antenna, and then rotating the antenna
without changing the source until the antenna is in a position where
the signal generated in it is 3 db down.
Fig. 2. Some interference can be tuned out with a receiver-installed
high "Q" frequency-selective adjustable trap.
In actual practice,
this property is of interest because it describes the relative sensitivity
of the antenna to signals coming in from different directions. One
signal, for instance, may be an unwanted ghost. In Fig. 4 we see
the effect of a wide polar pattern (A) and a narrow one (B) in picking
up or rejecting such a ghost. In the former case, pickup of the
ghost is almost as great as that of the primary signal. In the latter
case, the angle at which the ghost comes in is outside the polar
response angle, or 3-db point, and is sufficiently weaker so that
its effect on the viewed picture is considerably reduced. This property
of an antenna is roughly analogous to selectivity in the receiver's
front end. An example of an antenna having a small polar angle,
the JFD "Super Helix", is shown in Fig. 1.
narrow beamwidth goes hand in hand with high gain. Often, the greater
signal level, in a primary area, will overload the receiver causing
buzz, fuzziness, and tearing. This excess of signal, however, is
easily compensated for by the addition of an attenuator pad at the
set terminals. This pad may have a fixed value, but it is preferable
to use a variable type since the degree of attenuation required
will vary with the location. Units of this type are readily available
commercially. The use of a pad will also help minimize ghosts, by
bringing the level of the spurious signal down to a non-visible
value. Extreme caution must be exercised in an installation of this
type to orient the antenna accurately and to keep horizontal runs
of lead-in to a minimum. Horizontal lengths of twin-lead tend to
act as signal-collecting devices, and will complicate the original
In many metropolitan locations, extremely high signal levels are
prevalent due to the proximity of transmitting and receiving antennas
and lack of blocking structures. Too high a signal level can and
does cause poor picture quality. The average television set is aligned
for maximum bandpass with an a.g.c. voltage of approximately minus
3 volts. As the signal increases, the a.g.c, level goes up, shifting
the grid characteristics of the r.f. and i.f. amplifier tubes that
are controlled. When this occurs, each of these tubes operates on
a different portion of its curve, and its effective input capacitance
changes. This, in turn, causes deterioration of the bandpass characteristic
of the set, as shown in Fig. 5, and coarse, hazy pictures with poor
definition may result. There is also a tendency to poor synchronization
due to pre-triggering of the sweep oscillators. This condition may
be eliminated by the same attenuator pad mentioned previously. The
pad should be adjusted for a barely sufficient signal on the weakest
channel. If some channels are still overloaded, a switch type pad
should be used, so that it may be switched into the antenna system
only on those channels where it is needed.
Fig. 3. The longer paths of the reflected signals are responsible
Fig. 4. A narrow-beamed, more directional antenna, like
the one in (B), will reduce amplitude of the ghost.
Fig. 5. Receiver i.f. circuits are designed for best response
(center trace) over a certain normal range of a.g.c. voltage.
Abnormally high signal level (right) can deteriorate performance
as much as low signal and a.g.c. values.
Fig. 6. Tuned line is used to locate interference frequency
or act as trap.
or windshield wiper, is another symptom of too much signal. It usually
manifests itself as a dark vertical bar sweeping horizontally back
and forth through the picture. This particular trouble is usually
caused by a single channel much stronger than the others. TV tuners
are rarely selective enough to eliminate this annoying feature particularly
on channels adjacent to overly strong ones.
If all channels
are sufficiently powerful, an attenuator pad may be used to reduce
the overall signal level. A preferable method, however, is to install
an adjustable high-"Q" trap in the lead-in, and set it for sufficient
attenuation of the undesired channel. This is the equivalent of
a frequency-selective attenuator. These adjustments are usually
quite critical, and care should be taken when they are made. The
interfering channel can be identified by turning the channel selector
to the unused channel positions, and noting which channel is most
often seen. A commercial version of such an attenuator that may
be adjusted for rejection on the high v.h.f. band is shown in Fig.
In an urban area, there are many sources of interference.
Interference caused by amateur radio transmitters is usually of
a frequency below 50 megacycles. This, as a rule, can be eliminated
by the use of a high-pass filter. This type of filter should be
installed as close to the set's antenna terminals as possible, so
that there will not be pickup of interference along the lead-in
wire beyond the filter.
The previously mentioned selective
high-"Q" traps will be of great use in removing interference caused
by powerful local FM transmitters.
one form of which is ignition noise is often the most difficult
type to eliminate. In its milder forms, putting a tight twist in
the downlead may be sufficient to minimize it. Care should be taken
in the original installation to keep the antenna and lead-in as
far as possible from sources of ignition noise. Grounded metal screening,
used to shield the antenna from the street or other source of ignition
noise, is often used with considerable success.
noisy areas, shielded 300-ohm twin-lead or coaxial cable should
be used. The outer shield should be grounded at as many points as
possible. When using 75-ohm coaxial cable, a balun impedance transformer
must be installed at the top to match the antenna to the characteristic
impedance of the line. Another transformer must be used at the set
to match the line to the balanced 300-ohm input of the front end.
Both shielded 300-ohm cable and coaxial cable have comparatively
high signal loss. For this reason, a moderately high-gain antenna
may be needed to compensate for these line losses.
is often difficult to determine the source and frequency of an interfering
signal. A convenient gimmick that the technician can carry with
him is a 20-inch length of 300-ohm twin-lead, with lugs attached
to one end and the other end cut straight across and left open-circuited.
A piece of aluminum foil should be wrapped around the twin-lead
for a length of about 2 1/2 inches, as shown in Fig. 6. This forms
a capacitively loaded half-wave line. In use, the lugs are attached
to the antenna terminals of the set in parallel with the antenna
lead-in, which remains connected. The interference is then tuned
in, and the foil moved up and down the line until a point is found
at which the interference is at a minimum.
is rare today, but, when present, is extremely annoying. The best
cure for this condition is to locate the source and notify the Federal
Communications Commission. For several years, it has been illegal
for a diathermy machine to radiate and cause interference.
Most cases of insufficient signal in metropolitan areas are
caused by blocking of the direct signal by a tall structure, such
as an apartment house. In cases such as this, reception may often
be obtained by means of a high-gain, sharply directional antenna,
an example of which has been shown, oriented to receive a bounced
signal from some large structure to the side or rear. When this
means is ineffective, an attempt should be made to get permission
to install the antenna on the roof of the structure that is blocking
the signal. A third alternative is the use of a tower or telescopic
mast, enabling the antenna to be even with or above the edge of
the roof of the blocking structure. A 30-foot height will usually
be more than sufficient.
Reception problems caused by the
use of indoor antennas are difficult, if not impossible, to solve.
Generally, the best cure is an outdoor roof installation.
A final difficulty occasionally observed is phase shift due
to high voltage standing-wave ratio. The phase shift manifests itself
as a ghost which will change position with variations in the fine
tuning adjustment. This is almost always due to faulty installation
techniques. Be sure the lead-in has minimal horizontal runs and
does not run parallel to metallic objects closer than 6 inches.
Often, the condition can be eliminated by removing several inches
of lead-in from the slack left behind the set. This slack should
also be kept as short as possible.
Posted February 17, 2014