I often wonder if Major Armstrong knew what he was giving to the
world when he began developing frequency modulation as a public
service. Before World War Two, when the low-band FM service was a
comparatively high frequency, it seems that not much attention had been
paid to possible long distance interference produced by local VHF
broadcasts. Not long after WW2, it was discovered that the BBC TV
broadcasts were receiving interference from the Chicago Police
transmissions. This led to a reevaluation of the frequency allocations,
and our TV channel 1, which was in the top end of the 40MHz band, was
deleted. Fortunately, no channel 1 stations had been built, although
one channel 1 CP had been issued. This was quickly changed to another
Armstrong's tower in Alpine, NJ, is still being used today, but houses
only one FM broadcast signal.
By the time that the TV problems had arisen, our FM service had long
since been changed to the high band at twice the original allocation
frequency. This was particularly important to me because my first job
at KMBC in November 1945 was to convert its low-band FM transmitter to
99.7MHz. Happily, the new frequency was exactly twice the old one. It
was only necessary to build a doubler stage and a new PA, and then find
a new antenna. The latter was the hardest part.
The station also had an experimental television license and a fine
laboratory on the top floor of the Kansas City Power and Light
Building. This was where I did my development work on the new
transmitter. The station would have preferred to buy a new transmitter,
but none were available at that time (the “don't you know there's
a war on?” attitude still prevailed).
I had the idea that it might be possible to obtain a circular
pattern by means of some kind of long radiator around the top of the
tower because it was not easy to buy FM antennas in those days.
Unfortunately, all the details of my idea have been lost and
forgotten. I do recall, however, that the pattern seemed quite
circular, but in the absence of any measuring equipment, all I could do
was to compare reception at different points on a receiver and a whip
Fortunately, a Federal triangular FM antenna became available
suddenly, and the problem was solved before the FCC looked into my
A simple start
When Major Armstrong was developing FM, he probably used simple
dipole antennas at first. Then, as the commercial application of the
new medium began to appear, he no doubt gave attention to the method of
radiating the new signal.
It must have been obvious as Major Armstrong progressed with the
development of FM that the narrow bandwidth and highly directional
properties of dipoles would hinder FM's growth unless patterns and
characteristics were modified. It was necessary to develop antennas
with broad bandwidths, lower Q and more or less circular patterns. It's
interesting to reflect on the fact that the original humble, simple,
single-dipole radiator has become important today as a major part of
the popular panel antenna. Although the windloading tends to be a
little high, the panel-dipole antenna produces one of the most circular
patterns available in FM.
A circularly polarized antenna, such as this FMA-727 from Armstrong
Transmitters, provides a signal in both the horizontal and vertical
Fifty years ago we didn't have an EPA, and no one worried about the
effects of RF non-ionizing radiation (and no one yet has proved that it
hurts us). However, the FM antenna manufacturer still has to contend
with downward radiation and consider antenna bay spacing in his
Antenna-bay spacing and number of bays still seem to control the
level of downward radiation as modified by various antenna engineers.
Over the years, simple FM broadcasting has matured into a highly
technical operation, in many ways, perhaps, more complicated than the
directional AM antenna field. As FM ERP has increased and stabilized,
matters that probably were not even considered have now become
Some time around 1960, in an effort to improve stability of
reception, a number of FM antenna manufacturers developed circular
polarization (CPOL). The FCC permitted equal power in both horizontal
and vertical polarizations, but vertical power may never exceed
horizontal. Coverage is based on the horizontal signal, and vertical
power is not considered when calculating contours.
It was anticipated that CPOL would improve reception for receivers
using line cords as antennas (with various polarizations), as well as
auto radios and in areas of tall buildings and mountainous terrain. In
many cases, reception did improve; however, it was found that maximum
improvement could only be obtained by using CPOL receiving antennas as
well. Unfortunately, few people can find CPOL receiving antennas.
Nevertheless, there is a definite benefit in most cases because the
average listening location will likely have a wider choice of signal
polarizations, with a better chance of capturing a clean signal.
As CPOL antenna development continued and directional antenna usage
increased, HPOL and VPOL patterns frequently showed large differences
in polarization coverage. An HPOL plot might show excellent coverage of
the city of license, while VPOL coverage is poor and vice versa. This
has led to the introduction of short vertical parasitic correction rods
located near the antenna in an attempt to make the two polarizations
provide overlapping coverage by modifying the vertical signal.
The use of CP requires twice the transmitter power, or twice the
antenna gain, needed for HPOL to provide equivalent vertical and
horizontal coverage. This often poses financial questions of
transmitter operating costs versus antenna support costs.
As the number of bays increases, other problems arise. Neglecting
tower windloading, the greater the antenna gain, the narrower the
vertical beamwidth. Wide bay spacing of up to one wavelength tends to
increase the radiation at extreme angles of elevation. This means that
more power is directed downward (frowned upon by the EPA) and into the
air, where there are no listeners. High elevation radiation can be
reduced by using half-wave spacing, which also reduces the tower space
required, and multiple sidelobes can be reduced by careful bay spacing
and feeder design.
About 50 years ago not much attention was paid to the vertical
patterns of the relatively simple HPOL antennas, unless the antenna was
on a very tall tower or a high mountaintop, possibly resulting in
signals passing over the city of license. Thus, electrical beam tilt
found its way into our engineering vocabulary, together with new
interest in the vertical radiation pattern.
E-mail John at firstname.lastname@example.org.