[TowerTalk] Re: TowerTalk Digest, Vol 23, Issue 4

[Tom Rauch]

PS - lumped constants in any filter will change the X Y  Z
reading in the MFJ meter

   Thanks to Bob, K8IA, who mentioned the use of an ICE BCB
filter, and to
  John, W0UN, who mentioned using shorted 1/4 wl stubs cut
for the band of
  interest and T'd at the analyser.>>>

Martin,

While all are solutions to RFI what everything does is
almost exactly what a circuit like the MFJ filter does,
except the MFJ filter is much more transparent and has more
overall selectivity.

The MFJ filter has a series-resonant trap to ground, and a
parallel tuned circuit that band switches. You can null a
single strong BC signal with the trap, and the cancel the
remaining reactance with a parallel tuned circuit. The only
thing it doesn't do fairly well is handle VHF signals.

Since a coaxial stub has a Q in the 20's or so, the tiny
crappy parts MFJ used in the filter actually wind up
perturbing the system less and having better rejection of
out-of-band signals.

An above BCB high pass would not be reliable for SWR or
impedance values because of loss, phase shift, and the surge
Z0 not being equal to the line impedance.

My suggestion would be to simply build an identical circuit
used in the MFJ out of higher quality parts, or buy the MFJ
filter. I have the resources to do anything I want to do
with this stuff, and I use a 781 filter. If I didn't want to
have some idea of X and R, I'd just use a SWR meter and a
low power portable rig.

73 Tom