So I had another look at this one recently. I'd narrowed it down to the A9 timing board, and had an extender card made up at JLCPCB with 2x22 way edge connectors, so I could probe on ICs directly on the board.
Even with the logic analyser it was hard to see exactly what was wrong. The fault appeared some time from turning the instrument on - up to a couple of hours - and manifested itself with (a) capacitance readings wildly out or overflowing and (b) failure of the ramped C/V part to work. At first I thought it might be temperature related but a good squirt of freezer aerosol on the ICs on the board failed to alter it from working, or not working. When a fail occurred, self test on power on would always give ERR11.
So I started replacing ICs one by one. Fortunately they are not difficult to remove aided by a Hakko FR-301 and a Quick hot air station. I concentrated on the VFD section - U17, U7, U6, U19 then U30/U29/U28/U18. Nothing, the error still appeared after a while.
This morning I replaced U27 a LS245 data buffer and U32 a LS367 control bus buffer. The unit has been powered on and on a ramp up/down CV cycle for the last 7 hours and has worked faultlessly.
I tested the 245 and 367 on my retro chip tester (
https://8bit-museum.de/sonstiges/hardware-projekte/hardware-projekte-chip-tester-english/) which said both ICs are working. But this only does logic tests. I am wondering if I can test this chip any more as I'd love to know what the failure mode was.