Question on HP 3456A Multimeter CPU Board Interchangability

[DavidG]

I've been attempting the repair of an HP 3456A with a 2201A serial number prefix, which I believe would have originally had a 03456-66504 rev. C board. At time of purchase, it was missing all the boards from its upper half, including the 66504 rev. C.  That unit came with a 66503 rev. B installed in its lower half.

It was purchased in a lot that included another 3456A with a 2015A prefix, which I have been using as a parts unit. The 2015A unit was missing its 03456-66503 (the board under the 66504), so I have no idea which rev. it would have originally been built with. It had a complete right side and and a 66504 rev. B. Its fan had (literally) burned out, but its voltage regulator and all of its boards except its 66504 have been verified to be in working order. It's presently unclear to me if its 66504 is good, as I have no way to test it with its original 66503.

I have been trying to make one working 3456A from these two units. I replaced the right half of the 2201A prefix unit with the right half from the 2015A prefix unit. With a known good 66504 rev. C, it work perfectly. However, I need that rev C board for another 2201A unit. So I'm trying to get it to work with the 66504 rev. B. I have been told this should work, but with the rev. B installed, it generally displays garbled junk, which may or may not disappear a few seconds after it's turned on, or nothing at all. Usually, the garbled junk is completely incoherent.

So far, the following parts on the 66504 have been verified good by trying them on a 66504 rev. C:
* Mostek ROM set
* Mostek RAM
* Main processor (1820-2137/MC68A00P)
* Intersil (1820-2132/7218A) chips
* 1820-2036, crystal, and associated capacitor
* 1820-1759
* 1820-1931
* 1820-2309

I have also tried swapping out all the 74xx series logic as a complete set. My next step will be to swap a few of its chips at a time into a rev. C board to see if any of them cause that board to stop working. But I'm beginning to doubt that will reveal any new issues. So, what I'm starting to wonder is, is it not possible to use a 03456-66504 rev. C with a 03456-66503 rev. B.

So do any of you know for certain if this simply will not work? Or is it possible to do it by making some change to the 66504 rev. B?

Thanks.

[MadTux]

Have you tried the RevB CPU board on the 2201A?

If that works, why not use them this way. No cal data on CPU boards, so you can change them as you like.
I never had problems with swapping boards on the 3456A for diagnostics/repair. AFAIK, the revisions mostly changed the board from 3ROMs over to 1ROM, 2RAMs => 1RAM and and changed/eliminated the U9/U12 delay line (A10 line is not directly connected to ROMs in early versions)

Some of those old red 6800 chip sockets were known to be unreliable. Try tapping on the 6800 and the chip on the bottom board and see whether that changes anything.

[saturation]

I don't know the answer specifically but there are a lot of 3456a users on volt-nuts post there may help.

[DavidG]

I ended up trying each component from the bad board one-by-one on a known good board. The working board still worked every time. So I concluded that, barring some horrendous condition where two of the parts work separately but not together, the components are all good. Therefore, the PCB itself must be bad.

I guess I needed the desoldering practice anyway.