[Central-Electronics] CE20 oscillator drift

[k6jek]

Have any of you tackled the 9 MHz oscillator drift in the 20A?  

I measured it yesterday. At T+35 it was skidding along at 45 Hz every ten minutes.  It was still moving about 25 Hz every ten minutes after 65 minutes of warm up. It settled down nicely after two hours having moved about 300 Hz from power on +5 minutes.   

Today I tried keeping things cool. I installed one of those Weber 5U4 replacements instead of the real 5U4.  I placed a very slow fan on top of the cabinet drawing out, a 220 Hz muffin plugged running on110.  This made things worse.  The overall drift was less but it was still moving after four hours.  I'd hoped it would stabilize at a lower temperature and do that sooner.  No dice.  I guess all I did was slow down the oscillator heating process.

Incidentally, those Weber solid state rectifiers get plenty hot themselves.  I'd guess there's a big resistor inside along with a couple of diodes.

My thoughts are turning to negative coefficient capacitors and even to crystal ovens.  Either that or to heck with it.

Jon, K6JEK