I should add that the clone high-stability OCXO seems to have drifted about 12Hz in the past two months. This makes me think that the OCXO is dead/dying and can't be trimmed back into spec.
This is likely the problem - not that it's dying, but that it has drifted too much to be tuned to spec, or to be calibrated.
However, doesn't this particular OCXO have a trimmer for a rough frequency adjustment? Of course, the presence of a trim pot is a problem since the pot is likely to be the least stable piece, so adding a pot means you'll have to regularly adjust it.
I seriously doubt that's the issue.
dadler - that is a totally different board than the one I installed. Can you do what I did and connect the input ref and 10MHz output (from the back of the counter) to your scope and see what happens during cal? You should see two waveforms, one of which appears to be moving forwards or backwards relative to the other one. During call, the speed of movement should change (if it was working, it would slow and eventually come to an almost complete stop). This will tell you whether the instrument is able to adjust the OCXO at all.
If it's not too late, I would also make an attempt to contact the seller. If you haven't left and can still leave feedback, that's usually an effective tool to get them to respond. I can also provide a link to the OCXO board that I bought, which I know works well.
The fact that I was able to calibrate it once, and it matched my old-school Philips counter w/ OCXO within 1Hz, makes be suspect the OCXO (they now differ by 12Hz, and the Philips matches my two GPSDOs and the Rubidium unit, for the most part).
I purchased this board from this seller:
http://www.ebay.com/itm/Ultra-High-stability-Timebase-for-HP-Agilent-53131A-53132A-53181A-Counters-/111678771404?ssPageName=ADME:X:AAQ:US:1123I believe this is the same seller than Dave/Shahriar purchased their clone upgrades from.
I contacted the seller and told him of the issue, waiting on a response.
I bought this oscillator back at the beginning June, so I already left feedback as it worked and calibrated. Probably should have waited, but in my experience, if I wait a super-duper long time to leave feedback, I seldom get feedback left for me in return.
I will perform the procedure you have mentioned and try to see if the DAC is actually changing the output of the oscillator at all.
I'm also thinking of picking up a less-ratty-looking Morion MV89A on eBay, just as another device to test with.
Thanks again for all of your responses.