Hi,
I have been trying out a couple of home made resistive bridges to measure
Return Loss/Reflection Coeff/VSWR and have noted that there seems to be a
few basic types and would be interested in collecting a few comments from
others who have more experience with them than I. I should mention that I am
mainly interested in using these at VHF and up so have tended to discount
toroid/current types, and am interested in as wide a usable frequency range
as possible so also have not looked a lot at transmission line style
directional coupler ones either. If anyone has any comments on those
assumptions I would also be interested.

The first and simplest type I have seen in places like the ARRL Antenna
Handbook and places on the web like
http://www.g4wif.fsnet.co.uk/q_tech15.htm
and
http://www.pollak.sulinet.hu/electro/hidak/hidak.htm

These are a basic resistive bridge usually all nominal 50 ohm resistors with
a built in diode detector. The theory/maths behind these is straight forward
and seems pretty obvious by just treating the impedance of the detector as
much higher than the 50 ohms of the rest of the system and just ignoring any
effects from it.

Then there is a variation, the best example of which I have found seems to
be a circuit given be Joe Reisert in his VHF/UHF techniques column in the
old Ham Radio Mag (June or July 1996 I believe) which is basically the same
as those above but has an additional 50 ohm resistor across the centre of
the bridge setting the impedance of the in built diode detector. In the
versions I have built (using surface mount components etc.) this does seem
to work better, or at least be more well behaved, than the more simple ones,
but while I can readily see how this works for detecting balance, the maths
for unbalanced ( ie non 50 ohm load) cases gets quite a bit more difficult.

The final variation is as found in versions such as that of Jim Ford in
September 1997 QST, or as a v/uhf version at
http://www.webx.dk/oz2cpu/radios/swr-bridge.htm
This variation seems to remove the built in diode detector and replace it
with a 1:1 balun either of the transformer or choke type and then use a
separate external (50 ohm) detector. So mathematically I would think it is
equivalent to the Reisert version, but perhaps there is some advantage to
the explicit balance of the detector.

Any pointers to a theoretical treatment of these last two types, or other
comments would be appreciated.

Sorry for the long post and thanks in advance.

Paul VK3DIP