KFI's Tower, and other antenna questions
jef...@math.hawaii.edu
What's the length in wavelengths of KFI's tower? Might it be
a 5/8 wave antenna? In fact, what's more popular for AM
stations: 1/4 wave or 5/8 wave? With either you have to have
an excellent ground plane (visions of a couple San Francisco
AM transmitter sites sitting in the back waters of the Bay
come to mind; salt water is an unbeatable counterpoise!).
And how do you engineers tune your towers for the lowest
possible SWR? At the power levels you're running any
reflected power would be measured in the hundreds of Watts -
enough to cause quite a bit of damage to your PA.
Do you notice any changes in SWR with the weather? Since your
ground plane system is so extensive I would imagine after a
heavy rain the impedance, and thus the SWR, of your antenna
system would change.
Jeffrey
NH6IL
Doc Searls
> a 5/8 wave antenna?
I believe it's a half-wave, about 760 feet in length.
> In fact, what's more popular for AM
> stations: 1/4 wave or 5/8 wave?
Generally, half-wave is considered ideal, but they come in many different
fractions. Many of the stations at the bottom end of the dial, for
example, could not afford to put up half-wave towers, especially in
directional arrays. Quarter waves are more common there. As for
popularity, though, I think most engineers would prefer 1/2 wave towers,
altough maybe Lou Schneider or somebody can come in and correct me on
that.
A 5/8 wave tower pushes more of the energy along the ground and less to
the sky, but at low angles still produces a good skywave. WSM (650) in
Nashville has a tower that's over 800 feet tall that probably fits this
profile.
> With either you have to have
> an excellent ground plane (visions of a couple San Francisco
> AM transmitter sites sitting in the back waters of the Bay
> come to mind; salt water is an unbeatable counterpoise!).
Salt water has the highest ground conductivity available; and the SF Bay
area is blessed with a lot of it. Some of the Bay Area also has high
ground conductivity away from the water, too. This is one reason why AM
stations here do so well.
By the way, none of the big Bay Area stations uses a 1/2 wave tower. The
closest is KNBR, which uses a "capacitive hat" on its 568-foot tower to
achieve an electrical half-wave length. The height is truncated, I
believe, to respect airport approach paths. KGO's towers at the east end
of the Dumbarton bridge are only 220 feet. KCBS's up north of Novato are
around 500 feet. KFRC does extremely well with a 400+ foot 1/4 wave tower
in Berkeley. By the way, KCBS blasts the equivalent of about 200kw in the
direction of Los Angeles, down the length of the bay. The signal at night
in Southern Cal is like a local.
You'll need a better engineer to answer your VSWR questions.
Doc
Ron Cole
: > What's the length in wavelengths of KFI's tower? Might it be
: > a 5/8 wave antenna?
: I believe it's a half-wave, about 760 feet in length.
: > In fact, what's more popular for AM
: > stations: 1/4 wave or 5/8 wave?
: Generally, half-wave is considered ideal, but they come in many different
: fractions. Many of the stations at the bottom end of the dial, for
: example, could not afford to put up half-wave towers, especially in
: directional arrays. Quarter waves are more common there. As for
: popularity, though, I think most engineers would prefer 1/2 wave towers,
: altough maybe Lou Schneider or somebody can come in and correct me on
: that.
In reality 1/4 wave towers are the norm. Of the 100's of AM stations
that I measured for NRSC Compliance last year on Texas, New Mexico,
Oklohoma, and Lousiana there were only 3 that used any thing other than a
1/4 wave tower. Those two were 1/2 to get the height needed for the side
mounted FM antenna. One was a Franklin design that allows you to contol
the takeoff angle of the signal for Sky Wave control.
Ron
WPGC955
science. Virtually all AM towers are tuned at the base with usually a "T"
or "L" network composed of reactance arms (inductors and capacitors). The
most common installation uses a lagging 90 degree "T" network. Its easy to
tune and works well as a low pass filter for harmonic attenuation. AM
towers frequently are not matched very well back to their transmission
line impedance. VSWR levels greater than 1.5:1 or even 2.0:1 are commonly
found, especially in directional arrays to the low power towers. Even
mismatches of this magnitude to high power towers or non directional
single towers usually do not cause any real problems, unless the
transmission line is very long or marginally sized for the power it is
handling. In 30 years of AM field work, I've never seen a transmission
line burned up or even a PA stage damaged because of a line mismatch,
unless the mismatch became grossly high (caused by a failed component) and
the transmitter did not overload and shut down for whatever reason.
Moderate mismatches are easily transformed by the transmission line and
the output "T" or pi network in the transmitter so that the PA stage is
still looking at a comfortable load. At AM frequencies, the wavelength of
most transmission lines is very short, so the losses per hundred feet are
very low. Thus higher VSWR values do not translate into significant system
losses or potentially destructive voltages and currents (not to mention
signal quality degradation) which do occur at FM and TV frequencies.
Ron Cole
: VSWR levels greater than 1.5:1 or even 2.0:1 are commonly
: found, especially in directional arrays to the low power towers. Even
: mismatches of this magnitude to high power towers or non directional
: single towers usually do not cause any real problems, unless the
: transmission line is very long or marginally sized for the power it is
: handling.
Some directionals that I have worked on latley are in deed marginally
engineered. 1 5/8" line for 50Kw Transmitters. Medium power towers with
7/8" line. This is asking for truble and trouble did indeed strike.
Taking out the Referance tower, causing the remaining power some 35 Kw to
flow into a 7/8" line resulting in burned line. At that power level the
line go like fuse. By the time the Mangnphase triped the transmitter off
the damage was done. The transmitter of course was fine but the
undersized 5 tower directional was gone. After being in operation only 3
years.
I replace the 1 5/8" line with 3" and the two medium power towers were
upgraded to 1 5/8". That was 1987 and it's been on since then with no
other problems.
Ron
US 90W Communications
K7qa
the station Ron Cole refers to. The magniphase systems I've worked with
in the Continental 317C transmitters are very fast and almost provide too
much protection during wind storms and high static conditions, if they are
setup with adequate sensitivity threshhold. A properly responding
magniphase should normally have instantly shut down that TX when the 35 kw
from the ref tower got shifted to a mismatched 7/8 line, unless ofcourse
there was lots of isolation between the phasor power divider and the TX
output. I wonder if your station's magniphase sensitivity control was set
too low when the burnup happened?
Tom McGinley
CE WPGC AM/FM
K7QA or WPGC955 @ aol.com
Lou Schneider
: It surprises me a bit that the magniphase circuit did not properly protect
Sometimes you can lose a tower or have another major event that only
causes a small change in common point impedance. KCBS lost one of our 4
towers back in 1988 during a violent windstorm - this was the highest
powered tower in the array - 22 out of 50 Kw. The feedline shorted
against the ATU backplane as the tower went down, and the feedpoint
shifted to about 60 ohms with a little reactance. We were running a MW-50
with fully functional VSWR protection, and it was perfectly happy feeding
this until we were able to get out to the site the next morning.
mst...@delphi.com
Just a little KFI history.....When I visted the transmitter plant in the
1950's they used ballanced line (about 8" spacing on the conductors). This
was feed through a VERY large knife switch that switched between a dummy
load and the line out to the dog house at the base of the antenna. BTW,
they now have a stand-by antenna next to the main. I suppose it is a 1/4
wave but it does not look half as high as the main.
Regards,
Mike
K7qa
anything can happen in a multi-element directional array when any "major
event" occurs. The KCBS experience is a good example of what can happen
when impedance transformations occur because of a dramatically mismatched
load. Depending on the length of the line, a shorted 50 ohm line that is
normally expected to feed 22 kw to a tower could present a rather high
impedance at the phasor end of the line, causing the impedance of the
power divider to shift a small amount, as apparently happened. What then
happens to the 22 kw that got shorted out? Let us theorize a bit. A little
of it no doubt heated up the shorted line and was dissipated, since that
line was now acting as a shorted stub transformer, but was not passing any
real power on to a load. Most all of it was accepted by the remaining
three towers where real resistive components existed to dissipate the
power. Since the dead tower was no longer in the act driving 22 kw into
the array, the driving point impedances of the remaining towers all
shifted due to the changes in the array factor mutual coupling. That then
caused additional impedance shifts and mismatches throughout the system.
Whatever new driving point resistance values then were present at each
remaining tower base determined the new power distributions. One or more
of these could have changed to values well beyond the existing maximum
ratings of the various feedlines, which is what caused the burnout on the
array Ron Cole described in a previous letter. Thats why its probably a
good idea to run larger lines in a 50 kw system than nornally required.
You can never predict what tower might fall or what LTU component will
fail causing a complete reordering of driving point impedances and power
distributions.
Tom McGinley K7QA @aol.com CE WPGC AM/FM