I've recently been editing the memoirs of Dr. J. A. Pierce, an American
engineer who was responsible for much of the design work on the Omega and
Loran-C navigation systems and I thought some of his remarks might strike a
chord with this Group. See for yourselves!!!
His full memoirs run to some 350 pages and will be published as soon as
I've finished editing.
Walter G3JKV.
Excerpts from the memoirs of Dr. J. A. Pierce
(He is talking about setting up Omega in 1956)
Omega Antennas (Omega frequencies were 10.2 to 13.6 kHz).
Transmitting antennas for the low frequency constituted a major problem. We
designed and ordered
for use in the Pacific three 625 foot lattice towers to be used in an
"umbrella" configuration with
heavy top-loading cables. These could not be produced quickly so we settled for
balloon-supported
wires for the experimental stations set up in the United States. The balloons
were little "VLA" (very
low altitude) models about 35 feet in length and with a diameter of 114 feet.
They had stabi1ising
fins so that the kite effect gave them some additional lift from the prevailing
winds. Each balloon
supported an antenna of about 1300 feet of copperweld wire with large
insulators at the top and
bottom. The Army Air Force provided the balloons and helium to fill them, and
also assigned small
crews to fly them for us. We had three stations at East Brewster,
Massachusetts; Cape Fear, North
Carolina; and Key Largo, Florida.
The balloons flew more reliably than we had dared to hope but, of course, were
less than perfect.
They had a maximum load-carrying capacity (in still air) of 58 pounds while our
antenna wire and
insulators weighed 48 pounds. It was therefore essential to keep the balloons
well inflated. It
surprised me that, even with the balloons flying in three quite different
weather patterns, all three of
them were in the air for more than ninety percent of the time. This would not
have been satisfactory
for operational use, where absolute reliability of the signal is highly
desirable, but it was adequate for
the experimental work.
The antenna at Jim Creek (a U.S. Navy LF communications site).
The size of this station was a revelation to me. The antenna consisted of ten
copperweld cables
8,000 feet long strung across a narrow valley between two ridges 3,000 feet
high. The centers of
these strands were connected to downleads that were brought together into a
sort of transmission
line that carried them back to the transmitter building. The antenna was
actually separated into two
halves, each excited by its own transmitter, so that in case of accident or the
need for maintenance
the station could operate at half-power for a time. The transmitter building
was a concrete box a
hundred feet or so square without windows and with access to the area of the
transmitter itself only
by elevator from below. As befitted a station with a transmitter whose
component sections were
mostly of the order of cubes ten feet on a side, the elevator was so big that
we simply drove our
truck into it for the ride up to the operating level,
We spent two or three days setting up our equipment and erecting a whip antenna
for receiving the
signal from Criggion. As the transmitter building was the only possible site
for our gear in the
immediate vicinity, the whip was installed on the roof about fifty feet from
the "lead-in" which
carried about 700 amperes of radio-frequency current. It was in setting up this
antenna that we discovered the falsity of the common statement that "r.f.
doesn't shock; it simply
produces surface burns". This may be the truth for small quantities as
high-frequency currents tend
to flow only on the surface of a conductor, but it fails by a wide margin to
explain the behavior of
large currents at such a low frequency as Jim Creek's. Our rough calibration of
the field strength near
the transmitter lead-in was as follows: a bit of metal up to five or six inches
long (such as a
screwdriver or a pair of pliers) stings like a nettle; rubber gloves are a
necessity for handling metal
objects a foot or two long; and touching a conductor five or six feet long can
knock one down.
The minor pain we encountered in setting up this antenna was wasted, as we
never detected a signal
from Criggion at that site. Two or three days passed while we searched for the
signal. This was a
slow process as the only indication of its presence would be the tracking
behavior of our servos over
a half-hour or more. The search was complicated by the fact that our
oscillators had completely lost
calibration in the trip across the country, while the signal from WWV which we
had expected to use
to find the correct frequency was received so poorly as to be essentially
useless. At that date, the
only real access to precise frequency was through the signals transmitted for
the purpose by WWV
from near Washington, D.C., and also from WWVH in Hawaii. Neither of these
signals was received
well enough for the very accurate calibration we had to make. It therefore was
a painful and erratic
search, moving our oscillator in small steps through what we hoped was an
adequate frequency
range, and watching the servo record for symptoms of proper tracking.
Frequently random behavior
would look real for a few minutes and lead us to erroneous corrections because
our patience was
under such strain.
In the end we gave up trying to operate at the transmitter site. In the search
for an alternate we
found that there was a little "microwave hut" at the top of one of the
mountains that supported the
large antenna. This hut received signals, from Seattle I suppose, and relayed
them down to the
station in a telephone cable. The hut was not much more than a mile in a direct
line from the
transmitter, but was reached by seven miles of mountain road. The hut was near
the
southwestern-most "stub" tower that supported one of the strands of the big
antenna. This tower
was about 200 feet high and made an ideal point to which to tie a fairly long
wire receiving antenna.
At this site we set up our gear in an odd corner and, without too much
difficulty, detected the
Criggion signal. Without any proof, I still believe that our failure down below
was due to the
weakness of the Criggion signal at the bottom of the narrow valley - the signal
was none too strong
at the top of the mountain. We estimated the one microvolt per meter that I
mentioned above from
the degree of sluggishness of our servos. In other words, if the signal had
been stronger the tracking
would have been faster or better.
The antenna at Haiku (Hawaii)
The antenna at Haiku worked very well, partly because the site was in a crater
with a bottom a mile
or more wide, so that the outer ends of the antenna cables, where the voltage
was at a maximum,
did not hang too close to the conducting earth at the mountain top. The similar
antenna at
Jim Creek in the Cascades in the state of Washington had a disappointing
efficiency. I have always thought that the difference was that this Jim Creek
site was in a valley with
a very narrow bottom. The large sag of the heavy cables brought much of their
length too close to
the slopes of the mountains on both sides. Jim Creek was indeed very useful as
a large input to the
antenna made it one of the more powerful stations in the world. The amount of
power actually
radiated, however, was only about a third of what its designers hoped.
The first Omega-like transmissions from Haiku were made from the "small" TCG
antenna, named for
the type number of its transmitter. This was the antenna used for my slow-speed
experiment.This
antenna was a single strand of cable across the same crater but at a slightly
lower height and with
only a thousand-foot downlead. It was used for early tests by NEL, but later
the work was
transferred to the main antenna, which had not been in use for years. The Navy
Electronics
Laboratory crew working at Haiku were allowed to use the large antenna under a
curious agreement.
They could transmit what signals they pleased provided that they would rotate
the armature of the
Alexanderson alternator about 90 degrees once a week, to prevent it from
developing curvature of
the spine, or shaft.
Even the large antenna at Haiku was somewhat inadequate at ten kilohertz. When
it came time to
promote the station from experiments to full Omega operation, I believe that
the four strands of
antenna were increased to six, thereby achieving ten kilowatts of radiated
power. This was an early
estimate of what was needed, but operation of Trinidad at low power had shown
the desirability of
such an increase.
Apart from these "valley-span" antennas, the second type of antenna used for
Omega is the "umbrella".
This is a single central tower
surrounded by radial cables extending from the top of the tower toward the
ground at considerable
distances. These many radials have insulators at such a distance from the tower
that they hang at
about half its height. The remainder of each radial is interrupted by several
more insulators, so that
the antenna part is well isolated from the ground system below. For use at the
Omega frequency
such a central tower should be about as high as any man-made structure, I
believe that the one at
Tsushima is at least 1500 feet in height. In that case, the central tower takes
the form of a steel tube
a dozen feet in diameter. The antenna at Monrovia is supported by a triangular
lattice mast 1457 feet
high. The whole umbrella antenna usually rests on massive porcelain insulators
that are strong
enough to support the whole weight plus serious strains from wind loading, and
insulated for a
couple of hundred thousand volts. This is probably the most economical type of
antenna to use when
"ready-made" mountain sites are not available, as seldom happens in convenient
places.
The base-insulated umbrella is the kind of antenna we designed for the LF Loran
stations at the end
of World War II. We surely did not originate the idea, but such antennas were
certainly not used very
often before then.
I have only recently learned that a variant of the standard umbrella
configuration is used at the
Omega station at La Reunion and perhaps in Australia. This is the umbrella with
the central tower
grounded at the base and insulated at the top from the radial top-loading
cables. A great advantage
of this construction is that the tower is not electrically "hot", and it also
permits some simplification
in conducting to ground the strokes of lightning that invariably hit such a
tall structure. Because the
grounded tower offers a short capacitive path to ground from the high voltages
on the radials
(somewhat like the situation in a valley antenna with too small a height of
parts of the radials) the
efficiency is sure to be reduced in comparison with the same antenna with an
insulated base. The
large and expensive porcelain insulator, however, tends to be mechanically the
weakest part of the
structure and limits the weight that can be placed on it. The base-grounded
antenna can therefore be
extended to a greater height and thus permit radiation of as much power as a
somewhat lower tower
with an insulated base. The tower at La Reunion, for example, is no less than
1,600 feet tall.
|
