I understand that HP has used three different
OCXOs, E1938A, HP-10811, and the HP-10544. I
haven't found any info on the first one.

Leland-

My HP counter has a variation of the 10811 oscillator. To calibrate it, I
bought a 10 MHz rubidium-controlled oscillator on E-Bay, and it is impressive
how little drift the 10811 has over a long period of time. I use an
oscilloscope to compare the two 10 MHz signals, and can count the number of
seconds required for the sine wave to drift one division, where ten divisions
equal one cycle.

I also bought the counter on E-Bay. The seller insisted it had been checked
out and worked fine. What I found was that the 10 MHz oscillator could not be
adjusted onto frequency, and it never quite stopped drifting. After an
investigation that included purchasing a 10811A/B manual, I found that the
10811 had an open thermal fuse. Further, the HP/Agilent parts department was
not able to determine the correct replacement part until I had the manual in
hand, and could tell them the original part number!

It seems that the 10811 (and possibly the 10544) oscillator series was
originally designed with too low a temperature rating on the thermal fuse. At
some point a new higher value was chosen. If you happen to encounter the
problem, there is a good chance that if it doesn't have a burned smell, all it
needs is the new part. Mine has been working flawlessly for over a year now
with the newer thermal fuse.

73, Fred, K4DII