"The other John Smith" wrote in message
link.net...
I now have an HP 8405A Vector Voltmeter. I also have a Narda dual
directional coupler.


It's been a long time since I've used a VVM. I see 'em in the "send to
scrap" piles around here. Richard points out much more than I could and
certainly has a more jaundiced eye than I, but this one thing cought my eye

The HP voltmeter is able to zero out the phase angle between channels A
and
B. So, I'll put a 50 Ohm dummy load out there for the Load and set the
phase
angle offset to zero.

I forget how it offsets the phase reading, but I'd start by making the cable
lengths so that the two measurement points are "at the same pointy along the
line". (of course I have a line stretcher to make it easier)

It sure seems to me that the Zero angle 50 ohm load must be a bad
assumption. The 50 could be anywhere angle-wise and is the hardest to
measure since the reflected is so small...Wouldn't it be "better" to use a
short (of course I'm looking at a couple of Narda shorts at my operating
position every night) for a primary 180 degree reference (the open being
"harder" to obtain)? I think you mentioned using several other loads as well
(parallel 50s=25, etc). I would use several different loads like this to
"measure" to see how the measurements look compared to expected (cal-kit
style). I'd measure opens, shorts and several other mismatches to make sure
all available "knows" looks close, then proceed.
I'd also pad the Tx output. Long lengths of miniature coax in lew of
actual power attenuators.
OOPS! gotta re-boot, 73

--
Steve N, K,9;d, c. i My email has no u's.