A year or so ago I was working on a microwave local oscillator (at
about 2.5 GHz) multiplied up from a crystal oscillator near 40 MHz.
The output was found to jump in frequency by tens or hundreds of Hz
many times as the LO chain was warming up. I was able to reduce this
jumping by replacing all the dipped silver mica capacitors in the
crystal oscillator stage with NP0 ceramics. There is still a bit of
jumping which may come from some silver micas which remain in the
stage following the crystal oscillator.

I have just been observing the same sort of frequent jumping behaviour
(up to a kHz or so at a time) in another local oscillator (output at
about 10.5 GHz, phase locked to a crystal oscillator around 100 MHz).
I note that this one also has dipped silver mica caps in the crystal
oscillator and I wonder if it too would be improved by replacing them
with NP0 ceramics.

The capacitors used in both cases are from unknown sources and were
probably manufactured in the early 1980's.

Has anyone else experienced this behaviour ?

Steve (VE3SMA)