[time-nuts] 5370A Trigger

[Matthias Jelen]

Hello Time-Nuts,

once again a question to the 5370 experts:

After some time of operation (hours...) my 5370A starts to become noisy. 
This can be seen best when teeing the 10 MHz REF OUT to IN A nnd IN B, 
choosing TI and setting STD DEV, 1k samples. When everything works well, 
the displayed value is around 10 .. 14 ps. After some time, this values 
starts to rise slowly and might grow up to several ns. The effect seems 
to be temperature related, with the lid removed it takes longer until 
the unit starts to fail. I used quite some cooling spray, and the front 
panel board seems to be somewhat sensible, but I couldn´t find a part 
which, when cooled, returned the display to the initial values.  Cooling 
the whole frontpanel with an external blower also helps.

Then I noticed that I am able to bring back the display back to a few 
picoseconds by switching arbitrary around with the trigger slope 
switches - both channels seem to have an influence. After flipping them 
forth and back several times, everything is OK again. I couldn´t find 
any systematic behaviour behind this.

Did anyone experience something similar? (and maybe even found a cure?) 
Any hints will be highly appreciated...

Best regards,

Matthias