Roy Lewallen wrote:
dykesc wrote:
The manual on the MFJ 259b warns that strong local broadcast band rf
can cause problems with accuracy of the unit. The warning mentions
strong AM broadcast signals. A new tall tower near my house is
emitting a strong FM broadcast signal. When I hooked up my G5RV to the
frequency counter input on my 259b, it immediately locked on at
100.500 MHz. The signal is so strong that I can hear the station at
multiple frequencies when I tune across the FM broadcast band on a
small FM radio. Could this be interferring with my attempts at getting
reliable R-X SWR readings on my G5RV with the 259b?

Absolutely. Ironically, among the many uses for an "antenna analyzer",
the one they're least suited for is analyzing actual antennas -- at
least in a lot of urban settings. What I've done is constructed a number
of "half wave filters" which effectively filter out TV and FM signals
without affecting the measured impedance appreciably on the band they're
designed for. You do need one for each band, however.

A "half wave filter" has a unity voltage, current, and impedance
transformation, and a 180 degree delay, at the design frequency, like a
half wave transmission line, but provides a lowpass filtering function.
They're broad enough that you can use one over a whole band without too
much impedance error. It consists of a shunt C - series L - shunt C -
series L - shunt C, with the reactance of every element except the
middle C equal to 50 ohms (to mimic a 50 ohm transmission line). The
reactance of the middle C is half the value of the others -- you can
conveniently use two capacitors in parallel, each having the same
reactance as the end capacitors. I generally use mica or NPO ceramic
capacitors, and inductors wound on type 6 or type 2 powdered iron cores,
for HF. You should test the filter by measuring several known impedances
with it in series with the impedances, and with it removed.

You can make a highpass version, too, by exchanging the Ls and Cs. In
your case, though, you need lowpass ones.

Roy Lewallen, W7EL

He's getting RFI at one frequency. I was thinking an open quarterwave
coaxial stub tuned to 100.5 MHz would serve to trap the offending signal,
and be easy to implement. Cut it a tad long from the calculated length,
connect it to the analyzer input using a tee adapter, and "tune" it up in
frequency using a pair of wire cutters:

246/100.5 = 2.45' (or 2' 5-3/8") * Vf

Bryan WA7PRC