>Proximity of my various V/U antennas
Hi all: This became an issue at w3kwh a year or so ago after having repeated
failures of the front end HEMT transistor in a TS-2000X on 23cm. In our case,
the culprit was the local ATC radar at 1310/1330 Mhz, which was about 1.5 miles
away. With the 23cm loop yagi pointed at the radar site, a whopping 10mw of
radar signal was present on shack end of the coax cable as measured with an
HP141T/8554/8552. (We also have a cell phone tower on the property, that's
another issue).
At the time, Dave, K1WHS had suggested to me that even -10dbm was danger zone
levels for a HEMT transistor, so we were 100 times over that. To make matters
worse, the club site in an an approach flight path and random reflections from
aircraft would cause 10mw spikes even if the loop yagi was not pointed at the
radar site. This was discovered by watching the spectrum analyzer display on
1310/1330 MHz for a few hours one evening.
The solution was an interdigital filter followed by a "pipe cap wa5vjb" type
filter I made by cutting down an older F195/U twin cavity filter (cavity height
2.0 inches, probe length 1.375 inches. The combined filter bandpass lowers the
1310/1330 response to around -20 dbm while having around 1 to 1.5 db throughput
loss. No further HEMP losses have been encountered with the TS-2000X using this
approach.
As an observation, I ran an LT23S ssb electronics transverter and and outboard
preamps of the MGF-1302 and 1402 variety in years prior at w3kwh and never had
a 23cm preamp device failure from the nearby radar's 10mw poundings. I did
lose an ATF-10135 on 2.3 ghz from the cell tower overload once at w3kwh some 15
years ago.
A preferred solution for a more typical home or rover station is an appropriate
harmonic trap filter on the amplifier outputs versus having to eat a 1db or
more NF deterioration with a reciever bandpass filter. For example, run
separate radios and listen to your second or third harmonic with the preamp
off. If the received harmonic pins your s-meter does the radio's 10db
attenuator bring the signal down off the peg? If not, how many 10,6, or 3 db
pads does it take to get the S meter below S9+60 or whatever the top of the
scale is?? If you assume a pegged S-meter as being close to the danger zone
then the added attenuator pads will give you a clue as to how much harmonic
notch filtering you need as a minimum for rx front end safety.
At w3kwh I decided to put double coax stub 1/4wave harmonic filters (and 1/4
wave spacng) on all the V/U amplifiers just for safety sake. As an example of
the results the second harmonic from our AM-6155 amp on 222.1 was pinning the S
meter on the TS-2000X on 444.2. After installing the double stub filter on the
output of the AM-6155 amplifier the 444.2 signal was down to S5. The two
antennas involved were on the same mast about 4 feet apart and in the clear at
55ft up. Throughput loss was negligible on 222, perhaps a slight tuning change
was noticed. You might have to try different jumper cables on the output of a
solid state amp to a notch filter to find a cable length that keeps the
amplifier happy (i.e. Mirage Power Oscillators, sri I digressed :)).
What I found was easist to fabricate for the double stub fllters was to use a
pair of N-tee female adaptors linked together with an electrical quarter wave
of flexible coax of the desired harmonic frequecy (get as reasonably close as
you can, use an antenna analyzer to estimate desired resonance). For the
quarter wave stubs I used copper water pipe and brass hobby tubing to make
rigid 50 ohm lines with N connectors. A little epoxy will hold the center pin
on place inside the N-connector shell. A (home fabricated) teflon standoff
washer aligns the center conductor near the open end---these rigid lines are
easier to cut precisely with a bandsaw or miter box---rigid 50 ohm cable may
work just as well. One thing I found with the copper pipe coax stubs is that I
could put a pipe cap on the end of the stub and tune it with a screw---which is
faster than several cut and measure iterations to get a deep harmonic notch
effect for the double stub pair. You can (must) also tune each st
ub individually and then re-tweak for the best notch performance with the
assembled pair. I've consistently achieved 70 db nulls with this approach.
With two 10W erp beacons running at home continously on 2M and 432MHz, I've
measured a few -10 and 0 dbm in-band reflection points on the corresponding
chimney yagis with the omni loop stacks about 15 feet away--most likely tree
clutter reflections. I imagine that a wood roof and shingles in near proximity
to attic-installed antennas would have some similar reflection effects.
At this point I should thank Dave, K1WHS once again for his insight and give a
plug to W1GHZ for his low-power RF meter, as it costs less than the two-way
shipping expense for a radio front-end repair these days. 73 Mike wa3tts
_______________________________________________
VHFcontesting mailing list
VHFcontesting@contesting.com
http://lists.contesting.com/mailman/listinfo/vhfcontesting
|