One kind of filter you can use with antenna analyzers is a "half wave"
filter. This is an LC ladder network which has the delay and impedance
transformation properties of a half wavelength transmission line *at the
design frequency*, so the load impedance appears at the filter input. It
can be made with either a lowpass or highpass frequency response, with
the attenuation beginning not far from the design frequency. While it's
not by any means the sharpest cutoff filter you can build, it has
respectable attenuation beyond the cutoff frequency. It will alter the
impedance if you get very far from the design frequency, but for some
bands you can probably get by with a single filter, provided that the
impedance you're measuring isn't too far from the filter Z0 (see the
next paragraph) -- you should check with known loads at the frequency
extremes you want to measure. The cutoff might not be adequate for the
160 meter/1350 kHz problem at hand. You can cascade more than one for
better attenuation, but then the frequency range limits become more
restrictive and the filter construction more critical.

A "half wave" filter is made by cascading two "quarter wave" filters. A
"quarter wave filter" is simply a pi network whose elements all have a
reactance of Z0, where Z0 is the impedance of the transmission line
which the network imitates. 50 ohms is of course the most commonly
chosen value, but you'll get more extended frequency operation if you
choose a Z0 which is fairly close to the impedance you're measuring. For
a lowpass, use inductors for the series elements and capacitors for the
shunt elements. For a highpass, use the opposite. When you cascade two
"quarter wave" lowpass filters, you have a C-L-C-L-C
(shunt-series-shunt-series-shunt) configuration, where the middle C is
twice the value of the end Cs. A highpass looks like L-C-L-C-L, where
the middle L is half the value of the end Ls.

The inductor Q has to be decent, but not spectacularly good. I use type
6 powdered iron cores for the HF range. Type 2 might be a little better
for MF. Because of the low loaded Q, these filters are quite forgiving
about construction and small value variations. Give one a try -- they're
easy to build.

I live in a high-level RF environment, especially from TV transmitters.
I use my antenna analyzer for a lot of different things -- just about
everything except analyzing antennas, in fact, which ironically is the
one thing it's most poorly suited for. When I really need to know an
antenna's impedance (at least up to 60 MHz), out comes the GR bridge,
with an old Icom R1 receiver as a tuned detector.

Roy Lewallen, W7EL