This one's really got me stumped. About a month ago, one column of our (LED'ed) BSD started flickering. I poked around with my oscilloscope and it looked like U18 on the Power Driver Board was the culprit. I replaced U18 and it still had the flickering column. So I was thinking the problem was probably U9. Then the entire lamp matrix went out - no matrix lights in operating or test modes. So now I'm thinking my repair must have really screwed up the Driver Board. Except then we swapped Driver Boards with a HS2 and the lights were still out. And the board I "repaired" was working fine in the HS2.

So the entire lamp matrix is not working in the BSD, but the Driver Board is fine. The problem has to be in either the lamp harness or the CPU board or the ribbon cable between the CPU and the Driver Board.

I don't see how it could be the wiring harness. All 8 rows or all 8 columns would have to be not connected for the entire playfield to be dark. We tried connecting the harness at different locations (J137/J138 and J133/J134/J135) but we never even got a flicker. So wiring harness seems unlikely.

How about the ribbon cable between the CPU board and the Driver Board? We flipped the cable around and tried a different cable and the problem was unchanged.

Looking at the signals on the ribbon cable, there's the 8 bit data bus and 3 other lines that seem to be related to the lamp matrix: /LMP_ROW, /LMP_COL, and BLANKING. It seems the 8 bit data bus is OK because the Solenoid test works - all the right solenoids fire when they're supposed to. (This might be a bad assumption - it's possible solenoid timing is a lot more lenient than lamp matrix timing.)

So I looked at the three dedicated lamp-only signals on the BSD and the HSD while both were in All Lamp Test mode. They were identical. /LMP_ROW was high with two negative-going 370ns pulses, /LMP_COL was high with one negative-going 370ns pulse, and BLANKING was always low. So those signals seemed to be coming over the ribbon cable fine.

I should also mention that the GI also started flashing intermittently around the same time the lamp problems began. I haven't looked into that yet, but it also points to a CPU board problem since the lamp matrix and GI have almost nothing in common post-ribbon cable.

We had to give up troubleshooting for the day, but having put this all in writing now, my money is now on the CPU board in general and the WPC ASIC in particular. (I wish I'd tried reseating it while we were there!)

I'm still posting this in case anyone has any other ideas (usually after I write up something like this I discover that the problem is that I forgot to plug the machine into the wall ). Will keep thread updated with any progress...

Quick update: Problem does seem to be with U9 on the CPU board - press on it and the lights work fine. Haven't had a chance to dig in deeper.