Patrick @ Stern has been helping me via email, but he's on vacation for a few weeks. I thought I'd throw this out to the problem-solving crowd. Picture 1 shows the "active switch" test with the 4 balls in the trough (the 4 lit dots together). Picture 2 shows what happens when I activate the right return switch: all of the ball trough switches are "ghosted" from the 2nd row down to the 4th row.

4balls.jpg 4balls+right_outlane.jpg

This causes all sorts of problems, like Grave Marker MB starting when you drain the ball.

Here are all the things Patrick and I have tried:

1. Test all the switch diodes.
2. Replace the drop target opto board.
3. Disconnect the opto board for Sparky and Grave Marker
4. Reseat switch connectors on CPU board.

I'm at a dead-end now. Tomorrow I'll try switching out the trough assembly with CSI, but I'm just shotgunning it at this point.

Thank you! That's the procedure I used on my Space Station about 10 years ago, and I'd forgotten it. Much better than a shotgun approach.

When I do the row/column probe test, do I need to put a diode in series with the test wire? If so, does the band go toward the column or the row? I just want to be sure I don't zap my SAM board.

I tried the switch matrix probe test and didn't see any ghosting. It strikes me that the ghosting is very specific: it only happens when the right outlane switch is activated. I've already replaced the diode on that switch (no help). But I wonder if I should instead be looking at all of the diodes for row 7 (where the ghosting appears).

If I understand the ghosting correctly, what's happening is that J1-P3 (green/red) is somehow shorting to J1-P5 (green/yellow) when I activate the right outlane switch. So I should be able to SEE that short at the connector with my DMM. I'll try that next, which would really confirm that the problem is on the playfield somewhere.

Unfortunately I do not see a short between the 2 switch drives when I activate the right outlane, so I'm back to square one. I'll call Stern Support on Monday and see if they can send me a new CPU board.

I finally figured this out last night. The diode on the Right Outlane switch was facing the right way, but the switch was wired backward! Meaning, the return wire was mounted to the "common" lug instead of the "no connect" lug. I somehow missed that when I tested the diode and checked the relative orientation of the 2 wires and the diode band. They LOOKED okay...

The lesson I learned from this is to always go back to the most likely cause periodically, even when I think I've ruled it out. I could have saved myself hours of board-swapping and other troubleshooting.

I also learned that it's best not to buy a game in the first few weeks it's available. I'll wait a few months next time.