Hello Pinside community,

TL;DR
After heavy restoration, pinball seems to go in "ball search" mode immediately at the beginning of a game. All switches and coils are OK in diag mode. What could possibly trigger this ??
/TL;DR

I have a weir dproblem with a Secret Service I'm currently working on and hope I can find valuable input as to where to look for fixing this. Long story short, old pinball that was at my parents' for 20+ years, displays died once, repaired, died again, repaired, and the game stayed way too long unused. I've recently decided to restore it. What i've done so far:
1) Repair the PSB: +/- 100V were missing due to blown diodes and MJE340 transistor these are now fixed and the supply provides a nice regulated +5V and unregulated +/- 12V and +/-100V

2) Repairing the CPU board: the board is a 520-5003-00 and has several jumper wires that were installed in factory (see DE Service Bulletin #6). Apparently DE used remaining Laser War boards for early Secret Service productions, modifying them a bit to allow for two 27256 EPROMS to be used. Originally, the board would not boot: solid +5V led, blinking PIA led (no pattern, just blinking indefinitely). After cleaning corrosion from batteries, I got the board to boot by replacing the ram and cleaning throughly its socket. I have been unsuccessful in converting this board to a single EPROM one (by moving jumper J4 to J5)., which I would have liked (to replace the 2 EPROMS by a single one, and also use a troubleshooting DE rom)... but well, that's not the important point

3) Replace the old, busted display with a PinScore LED one

3) With the board booting, I put it back in the cabinet, and proceeded to re-plug one by one the outputs of the CPU board. Yeah, the game boots, lights up (many, many, many dead bulbs), no bad switches, no bad coil. Time to go to diagnostics: coils fire alright, switches trigger alright (some had to be cleaned but seems OK). Tried a game, glitches here and there. Notably, the game initialized with P & Y lanes (up of the playfield) lit and hitting the right flipper button would trigger the full "SPY" point sequence + a few contacts maybe not working so much. Bit of cleaning the leaf switches and goes better, but the main problem of this topic now arises.

Now to the point: after cleaning all switches, replacing nearly all the #89 bulbs and many smaller ones, I have a weird problem: immediately after a new game is initialized, and even before shooting the ball out of the starting lane:
- Eater UP coil is triggered (06R) and the "Hideout" (eater) compartment comes up
- Eater UP coil keeps being triggered 5 times (it stays up but you can hear the coild being triggered)
- At the 5th trigger of this coil, the "Eject hole" (03R) and Kickbig (16) coils are also triggered (meaning eater up + eject hole + kickbig at ~the same time)
- Go back to first item and repeat.

To me, this looks a lot like a ball search pattern: these 3 coils are the only ones associated with a compartment that could lock a ball and not release it should the corresponding switch be failing. However, this routine kicks in immediately at the start of the game. I am also puzzled because this did not occur at the first games I tried (back when some switches were not detected, just after the CPU was repaired).

Aside from that, the game is playable (score gets recorded, but does not stop this pattern). The only way to stop this pattern is to do the step for a 3-ball multiball play: once the 3 balls are on the playfield, the compartment work as they should. Palying, scoring a bit and losing the ball proceeds to 2nd ball, then third, and then the game ends correctly. Sometimes, losing the ball does not trigger immediately the end of game, with coils firing and still looking for the ball (but this is rare).

I have checked again and again the through switches, and they work properly (1st ball in outhole closes THROUGH 1, 2nd ball closes THROUGH 2, 3rd closes THROUGH 3). i've also reset to factory default sjust in case. In normal condition, I got:
- EATER 2 switch enabled (probably normal, since the compartment is down and closes this switch)
- THROUGH 1, 2 and 3 enabled (all 3 balls installed)
- UP POST switch enabled (circular piece that comes up as part of the game and protect the outhole i.e. losing ball is difficult for a limited time period; switch closed seems normal when this piece is not in its upper position)

In diagnostics, all coils fire properly and the L/R relay is working well (clicking sound + flashers work independently from coils). I am therefore assuming it is not a transistor issue on the CPU board and probably not dead diodes on the playfield. Dead ICs on the CPU board seem unlikely as it would probably fire the coils "at all time", even in diagnostics.

I'm really stuck and puzzled: why is this strange pattern kicking in right at the beginning of a new game, while not causing any problem in diagnostics ?

I'll take all the help you can provide! Thanks!

Small bump ?

Another bump a month later... Does anyone has a clue as to what to look for ?
I've replaced all coil TIP122 drivers with TIP102, the two LS7408 that had bad outputs. All coils fire properly, flasher do too, switch tests are OK: it's only in-game and with the Eater compartment that things go wrong...