[amsat-bb] 1/2 Wave "cavity" filter design?

[Robert Bruninga]

Given the classic 1/2 wave cavity filter, what is the relationship
between inner cavity and center conductor sizes?  Of course bigger is
better, but what is "OK".

This is a 1/2 wavelength tube with a center conductor shorted at each
end.  The input is loop coupled at one end, an output is loop coupled
at the other end.  In the middle is a tiny variable capacitor (usually
just a screw)  to tune to resonance.

In the limit, as the cavity size shrinks, you can end up with what
could be considered as just a piece of 1/2wave coax.

Im looking for a really cheap Home-Depot plumbing design that 15
people can reproduce to give them good front end antenna filtering
when operating on mountain tops adjacent to other RF souces. (say
within 100 yards, not permanent installations which of course should
spare no expense at getting the best cavities possible).

We just finished our 4th annual Golden Packet attempt from Maine to
Georgia along the Appalachian trail and many stations were plagued
with front end overload. http://aprs.org/at-golden-packet.html

I'd like to come up with a 3/4" copper pipe design that is robust,
provides sevral dB of out of band rejection.  Im trying to understand
the parameters that drive the size of the center conductor.  Normally
bigger is better for better bandwidth, but I think smaller will give
me steeper skirts and better rejection?  I dont mind say 2 or more dB
insertion loss, because as it is, front end-overload is making us
totally deaf and anything would be bettter.

Lastly, I think such a 1/2wave filter will also pass as a 1.5wave pass
filter on UHF.  We need dual band, since we use dual band rigs and
coordinate on UHF voice from the same antennas and coax used for the
VHF packet.