[Towertalk] RF relays --aka-- Relay Summary
W0UN--John Brosnahan
shr@ricc.net
Thu, 28 Mar 2002 13:01:53 -0600
Here are few observations to add to the discussion based on
my experiences in radio astronomy.
Gold-plated DowKey coaxial relays are not reliable on reception unless
you take special care when using them. They have very little wiping
action and a film will build up that causes them to make poor contact.
This film is probably only a few molecules thick--so you still have
pretty good reception through the relays due to capacitive coupling--
but a phase measurement through the relay will show random results.
This is a fatal flaw when the relays are used in an interferometer or any
phase-sensitive, RX-only application. You have no idea of the phase
relationships between the interferometer elements.
If you clean the shinny contacts (typically by wiping a piece of file
card paper between them) you will see good performance for a few
weeks but then the phase problems start to happen again. But if you
run a current through the contacts the performance will remain good
for an extended period of time.
We started running our preamp power THROUGH the phase-shifter relays
rather than introducing the bias-Tee after the relays. This solved our
problems, although we ultimately changed the relays to a mercury-whetted
reed relay.
I think that relays with good wiping action, gold contacts, and that
carry some power (DC or RF) will probably remain in good shape
for typical amateur applications.
But if you are doing RX-only switching and phase is important, and/or
if you are using generic power relays, you need to consider going to a
vacuum relay or a reed relay for consistent performance.
NOTE: If you use an ohmmeter to check the relays they will show
that they are good--the ohmmeter voltage is enough to break down the
film layer. But if you use low-level RF and look at the phase
and amplitude with something like an HP Vector Voltmeter you
will see the problem immediately. Sometimes repeated activation
will clear up the problem, or you may need to wipe the contacts before
you see phase-stable signals again.
Although my example is based on experience with DowKey coaxial relays
I think you will see the same sort of problem with any relay that has poor
wiping action and is open to the air.
John W0UN