[time-nuts] Re: Repairing an HP 5065A Rubidium Vapor Frequency Standard. Sharing experience and advice welcome.

[john at miles.io]

It's a bit of a hack, but I protect my 5065As with an Arduino mounted to the outside rear panel with heavy-duty 3M mounting strips.  It monitors thermistors at both ends of the RVFR and shuts down the party by removing power to both heater windings at J15-4.

But if you're saying that a short at C1 can toast the lamp driver PCB, that's a different circuit than the heaters use, as you point out.  Sounds like a small inline 1/2A fuse at J16 would be a good idea.  Never heard of this happening before, so thanks for the heads-up!

There are some legends and lore floating around regarding the choice of frequency for the Rb lamp exciter.   Some people say it's a matter of importance, but IMHO it's not clear how, as long as it stays clear of the SRD drive.  Going by page 8-56 of 05065-9041, the 90 MHz figure sounds credible.  In this circuit C4 and the lamp coil have the most influence on the output frequency by far, so if C4 is still on target, maybe the lamp coil winding expanded slightly when the overheating event occurred. 

Re: the optical filter, Corby modified two of mine, and I modified a third with the 10nm-wide part from Edmund Scientific.  It was very worthwhile but the improvement wasn't as dramatic as it was with the first two units (which were better performers to begin with.)  I went with a 'temporary' mounting solution in the form of high-temp copper adhesive tape.  Of course, nothing's so permanent as a temporary solution.  It's an ugly kludge and could potentially be causing some light leakage, so fabricating a real lens mount would be better.

-- john

-----Original Message-----
From: Jean-Charles BILLEBAULT via time-nuts <time-nuts at lists.febo.com> 
Any thoughts or war stories appreciated. This one's turning into quite the project.