This is on a WMS driver board that I've been repairing. It came in a batch of bad boards and I don't know its history. It was missing A LOT of parts, including all three PIAs, many other ICs, random transistors, resistors, etc. If you look at the attached image all socketed ICs are new. After re-populating the board and doing my usual WMS driver board work (replacing the 40 pin connector, replacing all headers, upgrading the lamp driver section to MOSFETs, and replacing the switch matrix resistors with zero ohm jumpers) I put the board on my test fixture. Honestly after replacing all of those parts I wasn't expecting much but it booted into attract mode. It's paired with a System 7 MPU with Joust ROMs.
On the test fixture it boots into attract mode normally, except after a bit less than a second solenoids 1-8 lock on. Here's the interesting part, if I go into solenoid test the solenoids behave normally. As it cycles through the solenoids it turns off each locked on solenoid, and then subsequent passes of the solenoid test pulse each solenoid normally. Exiting test mode and going back into game play locks solenoids 1-8 on again.
Tested the solenoid PIA (and the other two PIAs as well) on my PIA tester and all tested okay. All are new. I replaced the solenoid PIA just to be sure, no change. Started poking around a bit with a data probe and saw that the PIA's outputs for the 8 solenoids are all high in attract mode. In solenoid test the PIA's outputs behave as they should.
So I'm thinking that maybe there's a problem with one of the data lines. All 8 seem to be pulsing normally on the date probe, and there's continuity on each data line between the three PIAs and the 40 pin connector. Anyone have any ideas as to what to look at next?