❯ I'm working on a Sopranos (Whitestar) game.
The right flipper is sticking up occasionally. But the coil is staying energized. It's not mechanically binding. It's not a dirty coil sleeve, mushroomed plunger, or too tight to the bushing. The coil seems to release after a couple of seconds if you mash the flipper buttons. If you don't mash, it seems to stay up energized indefinitely, although I haven't timed it.
A small downward pressure on the stuck flipper seems to release it, and the energy goes away. The flipper button and cabinet switch looks normal. It's not sticking or misaligned. I hit the switch contacts with a file to lightly clean it.
The end of stroke switch that tells the game to reenergize the coil looks ok. I can get it to stick up by flipping it manually with the coil test in the diagnostic menu. So it doesn't seem to be an issue with the button/cabinet switch. What else can I look at?

⏺ Classic stuck-on coil symptom — and the fact that you can latch it in coil test (where the CPU only sends a brief pulse) is the smoking gun. That tells you the coil drive is being held active after the CPU stops asserting it, which means the issue is downstream of the CPU on the power driver side.

Priority order to check:
1. Right flipper driver transistor on the power driver board.

This is the #1 suspect. On Whitestar power driver boards the flipper drivers are TIP102s (some revisions/rebuilds use logic-level MOSFETs like IRL540). A partially-shorted C–E (or D–S) junction can intermittently latch on after being pulsed, and mechanical perturbation (mashing buttons → rapid back-EMF; pushing the flipper down → mechanical kickback voltage from the coil) can knock it out of conduction. That matches your symptoms exactly. Pull it and bench test — don't trust an in-circuit reading. While you're there, check the base resistor.

2. Pre-driver IC for that channel.
The flipper transistor base is driven by a buffer (depending on board rev, ULN2803 or a 74-series). A flaky pre-driver output can hang high. Easy to swap if you suspect it.

3. Flyback diode across the coil.
Look at the 1N4004 (or equivalent) across the coil lugs. If it's open or marginal, every flip has been hammering the driver transistor with back-EMF spikes, which is the textbook reason TIP102s end up partially shorted. Even if the diode tests fine now, replace it when you replace the transistor — they're cheap insurance.

4. Cracked solder joints around the flipper driver.
Whitestar power driver boards are well-known for cold/cracked joints at the high-current legs (transistor leads, connector pins, ground bus). Reflow the area around the right flipper driver and its connector pin.

5. Coil connector pin at the power driver board.
Burned/loose Molex pin on the flipper output can produce weird drive behavior under load. Inspect for discoloration; replace pin + header if suspect. Quick test to localize before pulling parts: with the game on and the flipper stuck up, measure voltage at the coil's "switched" lug (the one going to the driver board). If it's near 0 V, the driver transistor is definitely conducting. Then probe the transistor's base/gate — if it's high while CPU isn't commanding the flipper, pre-driver is latched. If base is low but transistor is still conducting, transistor is shorted.

My money's on the TIP102. Replace it + the flyback diode together.

✻ Brewed for 1m 24s