Years ago I bought a -
hp- 10544a ovenized oscillator with attached adapter PCB thinking I could use it to upgrade my -
hp- 5328A counter which came with plenty an option, but crucially not option 010, the high stability oscillator. Now the service manual states that all one has to do is remove the old oscillator (this 3 pin device in TO-5 case), insert the adapter PCB with the oven assembly, fasten that to the chassis and bob's your uncle. Well, I tried that, but after turning the power on again, the counter didn't count anymore. That's actually not correct -- it
did count events (for which one doesn't need a clock), but not the frequency or period of an external signal. The count remained 0 (occasionally the last digit jumped one, as there seems to be an off-by-one error happening when changing functions). My heart sank, but there was no magic smoke and eventually the oven became warm to the touch. Such encouraged, I poked around with the oscilloscope and did find a perfect 10MHz on pin 1 of the edge connector to the oscillator. So the oscillator was oscillating, just the counter wasn't counting.
A reader familiar with the counter might have noticed that option 010 bears actually the -
hp- part number 05328-60018, not -60038. The riddle's solution lies in the fact, that -
hp- produced the counter also for the military with slight variations. Afaict, the military version has always the Option 010 built in (there called A3A1), automatically senses the presence of an external reference signal and offers additional outputs for the internal 10MHz reference and a by 10 divided reference.
With that tidbit of information and the schematic of the A3A1 PCB (05328-60038) from [1], the schematic of the counter (which thankfully are provided by Keysight) made sense and it was easy to deduce that the wire-bridge, which connects pin 6 of U16 with pin 13 of the edge connector to the option 010 board had to be cut and rerouted so that pin 6 of U16 connects to pin 11 of the edge connector instead.
hp(i)-5328a-with-option-010-(05328-60038)/?action=dlattach;attach=1987642;image)
I did just that and viola, it actually works. Took me a while to gather the necessary info from the Interwebs, so if someone finds himself in the same situation (however unlikely that might be), this hopefully gets him/her on the way sooner.
PS: the -60038 is actually the preferred option, as it allows for electronic frequency control (the 'fine tune' multi-turn potentiometer on the option PCB (it took about five whole turns clockwise to remove the 30pbb frequency error seen in the picture)
[1]
https://radionerds.com/images/9/9a/TM_11-6625-2941-14%26P.PDF