On Mon, 19 Sep 2005 23:35:12 -0400, "Bob Chilcoat"
wrote:

The receiver is already in a (steel) waterproof box. Tomorrow I will try
moving the receiver as far away as I can. If that doesn't fix the problem,
I'll try the stub antenna notch filter solution. I have a BNC Tee. Can
anyone point me to the 1/4 wave length formula for 120.6 MHz? Does 0.591
meters (23 5/16") sound right?

Wavelength is 300/F in meters. One quarter wave is 75/f in meters.
However your using coax cable and the speed of light (nominally 300 in
the examples cited) is slower depending on the coax used. for foam
types it's about .8 (varies by brand) and for solid types usually
around .66 (still varies some).

So for 120.6 a 1/4 wave section is 75/120.6=.62189M (24.4838").
A section of RG58 (solid dialetric) would be .66 that length or .4104M
(16.159").

To give you an idea of how critical that length is the 1/4Wave of the
same coax for 123mhz is .4024M (15.84").

So even small variations in VF (velocity factor for the coax) or
gutting error can make for big mistuning.

Generally speaking simple stubs work best if the two frequencies
are widely seperated such as 2m band and pagers at 153mhz
(5mhz minimum seperation).

I'm pretty sure right now that the interference is coming in on the antenna.
It's not enough to trip the squelch, but as soon as someone keys on 123 and
trips the squelch, the AWOS is on the audio. OTOH, I guess that doesn't
prove anything...

Sounds like intermod. The front end of the reciever is being grossly
overloaded.

The average airband radio the difference of 120.6 and 123mhz is so
small to the front end that overload is common.

I suppose the wiring to the 88.1 MHz transmitter or its wiring could be
picking up the 120.6, although all that wiring is shielded (one audio line
with shield terminated at only one end, and one 3v power line with its
shield as return). The transmitter itself is in an unshielded plastic box,
but that's mounted flat against the bottom of a 10" ground plane for the
transmitter antenna. The last possibility is the 6v power line going into
the receiver box. It's not shielded and starts at a wall wart
transformer/psu very near the AWOS xmitter. I could put a few turns through
a ferrite toroid just outside the box, I suppose. Couldn't hurt. Gotta
find a suitable torroid.

RF will com into the box anyway it can. Via power, antenna, audio and
even a poor connection between the cover and the box.

FYI: I'd bet anything that a few hundred feet of seperation from the
AWOS cures it or substantually improves it.

Allison, I'm not experienced enough at this stuff to visualize how to make a
bandpass section out of coax. I can follow the stub notch filter, but the
bandpass isn't there. Could you explain a bit more?

If I had a way to post a schematic It would help. Suffice to say its
something that you'd need hardware to tune. The idea is that using
coax sections it's possible to get the equivelent of high Q tuned
circuits and couple them loosely for selectivity. Basically a open
ended 1/4 wave section of coax is a similar to a tuned cavity save for
a cavity at 123mhz is nearly 25" tall making coax less bulky.

So using an open stub creates a notch at some frequency, a shorted
stub is a tuned (resonant) pass at the same given the same length.
Visualize a feed from antenna, T with shorted stub(for 123), a 1/4
wave section for 120.6 section, another T with a open stub for 120.6.

Thats the starting point and it may be adaquate at that. However
to really make it work there is more and It's hard for me to describe
here. Sufficient to say VHF filters with that kind of selectivity are
not trivial to design and build.

Also obvious you need some instrumentation to tune the beast
for best results. Which is why there are people that build custom
filters for tasks like this.

Allison
kb1gmx