[Moon] isolation relay performance

[peter blair]

 This post is for general information and comment.
Between the receive port and the preamp on my 23cm feed I use an isolation relay which switches the preamp between the feed (rx) and a 50R termination (tx) This is common practice and I do the same on all the eme bands. I use small SMA type relays to minimise losses and discontinuities. I recently had one that I'd used for probably 2 years until it developed an intermittent loss of about 0.2-0.3dB showing itself as a noticeable loss of moon noise or sun noise.  I changed the relay and the problem went but this weekend I decided to take a look at it as there seemed to be no obvious reason for it doing this. 
 The relay is an MA type ( 35x45x12mm) with  a nickel plated brass body  and the 3 SMA female sockets are screwed  into the top, each with a lock nut. They look as though they are stainless steel.  I carefuly unscrewed them and immediately noticed some white deposit in the threads. (I should mention that the feed is only in the dish for a few days at a time but the relay is not shielded from the weather) I cleaned up the threads with a small wire brush ( for suede shoes!) reassembled the relay and it is perfect again. The loss including 15mm 0.141 cable is 0.1dB measured on the PANFI as a change in NF. Repeated operation of the relay has no ill effect. The gold plated contacts on the inner conductors were nice and bright and clean
 I'm amazed that the water could penetrate the locknut and thread, but it clearly did and the intermittency was presumably the mechanical shock altering the corroded contact of the SMA outer. I shall in future shield it from the wx but I thought it worthwhile to distribute this info as its the sort of thing that can go un-noticed as a gradual degradation in performance.
 I wonder if anyone else has experienced this?
 73 Peter G3LTF 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.moonbounce.info/pipermail/moon/attachments/20080218/608865d4/attachment-0001.htm