HP/Agilent 1630D full-field graphics display failure - solution found

[Lorenzo_1]

My 1630D logic analyser graphics display failed recently after being intermittent for a while.  Character display of State and Timing data in LIST mode worked fine but no Waveform or Chart display in the lower RHS graphics display area (a 16x56 character display zone within the overall 23x64 character CRT display). Absent another logic analyser to diagnose it, I had to use the 1630D's own State/Timing analysis to find the bug - slow but eventually effective.  Turned out to be a failed SN74LS393 binary counter (U4F on the tech manual's schematics) that is part of the 60Hz horizontal frame frequency generator.  The counter was failing to count beyond 7H and wouldn't cascade to the following counters.  Replacement brought it back to life immediately with no seeming ill effects.

The specific failure cause may not be of special interest to other forum users, but the diagnostic process left me with quite extensive notes on signal patterns in parts of the CPU board that were still working.  They might be useful to others who encounter problems with this subsystem.  I'll scan and store them for a rainy day.

I can now get back to repairing my HP3325A that's been languishing for some weeks - proved hard to debug without a working logic analyser... 



 

[TerraHertz]

I'd be interested in a copy of those notes, since the logic analyzer I used most of my working life (and still have) is a HP 1630G. I have the service manual, and the one time the machine developed a fault it wasn't very hard to fix. But everything helps.

[Lorenzo_1]

Can do. You'll have to be patient as they'll be unintelligible without my reworking them from the original pencil scratchings.  I'll get to it as soon as I get a breather from trying to fix things!

[AdrianoKF]

Sorry to resurrect this old thread, but do have a scan of your notes available by any chance?
I'm working on a HP1630D with seemingly faulty RAM (garbled text output [missing/wrong characters, random self-test fails with either RAM or Acq errors), where some more information about the signal patterns would come in very handy!

Thanks in advance!

[Andy Watson]

Sorry to resurrect this old thread, but do have a scan of your notes available by any chance?
I'm working on a HP1630D with seemingly faulty RAM (garbled text output [missing/wrong characters, random self-test fails with either RAM or Acq errors), where some more information about the signal patterns would come in very handy!

I  had a similar problem with a 1630G - random and garbled messages, self-test fails with garbled error message. I replaced the dynamic ram and that cured the problem. The ram is a tight fit into the PCB, I chopped the legs of the old RAM and pulled the pins individually. Replaced with sockets and NOS ram chips from ebay . It's still working.

If you are going to operate with the case removed note the warnings about maintaining cooling air flow over the acquisition boards.

[AdrianoKF]

Sorry to resurrect this old thread, but do have a scan of your notes available by any chance?
I'm working on a HP1630D with seemingly faulty RAM (garbled text output [missing/wrong characters, random self-test fails with either RAM or Acq errors), where some more information about the signal patterns would come in very handy!

I  had a similar problem with a 1630G - random and garbled messages, self-test fails with garbled error message. I replaced the dynamic ram and that cured the problem. The ram is a tight fit into the PCB, I chopped the legs of the old RAM and pulled the pins individually. Replaced with sockets and NOS ram chips from ebay . It's still working.

If you are going to operate with the case removed note the warnings about maintaining cooling air flow over the acquisition boards.

Mhm, that's what I had feared (looking at the price for NTE2164 NOS in comparison to what I paid for the unit). Before I go ahead and replace all RAM chips at once, I'd like to track the problem down a little further, though.

If you have any chance and interest, would you be willing to capture a trace of the HCRA[0:6] signal and the /CS pins on U6C (the character ROM) on the main CPU board, either in normal or self-test mode? I took a few captures on my unit and would like to see if I could deduce something about the culprit from that trace. In theory it should be possible to reconstruct the screen contents in text mode from the character ROM data, since the service manual claims the character ROM is directly addressed using ASCII codes (or it might not, I'm more of a software guy, so I like thinking in terms of data ).

Thanks for your help in advance!