Quoted from german-pinball:
Check the connectors again, check the connector pins of the lamp board.
Also reseat the ribbon cable from the CPU to the driver board on both sides.
How did you test the TIP107?
Initially I found a couple of cold solder joints on the lamp board and reflowed some solder onto them. So the lamp board is 100%. I will recheck the connector and the connector on the power board.
I checked the TIP107 two ways. First, I checked for continuity between the tab of the transistor and TP8. There was no continuity. Second, I used diode test to check each of the transistor's legs. They returned similar values to the TIP107 next to it, which was working fine, so I took that to mean the TIP107 was OK.
Quoted from herg:
If you did this test properly and the column 4 lamps lit when they were supposed to, you can stop looking at the boards. I love this lamp matrix tester, but the jumpered test can get you the same results with a bit more work.
The most likely cause of a column not lighting is a broken wire. For columns, you can test DC voltage, and you will normally see something close to 3V (due to the 18V being pulsed). Test it at the connector to start, then test again at the lamp board. From there you can narrow down where the open would be.
I will check for the 3V and see what I get. However, I know that the wire from J137 to the lamp board is fine. I will check the row return as well while I'm at it.
If I am understanding you correctly, if the socketed lamp test works, then all lamp matrix components on the power driver board should be working 100%. Is that correct?