A couple suggestions. Check continuity from PPB relay K1 to the non banded side of D22 to J6 pin 9 to CPU CN12 pin 2 Black/Red wire. Q29 on the CPU drives relay K1 or in other words provides the ground that energizes K1. Q29 may be bad or have a bad solder joint. Check continuity with everything associated with Q29 and pre driver Q21. Test Q21 and Q29. I think you can ground Q29 to energize Relay K1 to confirm continuity in the circuit. Check fuse F6 on PPB and continuity with everything associated with F6. F6 connects to J6 pins 4&5 orange wires. That is where flasher power comes from. It seems as if K1 isn't being energized to send power to the flashers and the non working coils. Or it could be energized and not being DE energized when needed. It is a left right relay being turned on and off by the game program to send power to coils and flash lamps when needed.

Ok I took a look at my Lethal Weapon. It has the exact same board set as R & B. I started a game and K1 energized and stayed energized. Is your game doing that? If not, that leads back to my thoughts about the relay not being activated. If it is being activated then we know that it is working correctly and you should now focus on why flashers and coils aren't getting power or the ground path has been lost.

Shine a flashlight on it. You will be able to see the contacts move.

Ok try this. While the game running in attract mode use a piece of wire and attach it to the metal ground plate that all the boards are screwed to. Then BRIEFLY touch the other end of the wire to the metal tab of Q29.If all is well the K1 relay will energize/click. This will confirm the ground path to K1 from Q29 is functioning.The relay coil needs to be grounded for it to work . This is where the CPU comes into play. It is possible that Q21 or Q29 test good but don't really work. Further upstream are a couple of 7408 chips and a 6821 PIA that drive the transistors in that row and may not be doing their job. I'm not sure how to check them for proper operation though. You may have to replace them if nothing else is resolving the problem. 38v on both sides of D22 in attract mode is correct. When you start a game the relay energizes and the non banded side of D22 drops to zero volts and the banded side remains at 38v.

This may sound like a dumb question but is Cn 13 on the CPU board plugged in?

How about J3 on the PPB

Kinda out of ideas to try at the moment. Seems to be a grounding problem knowing that there is 38v to the relay.

http://www.ipdb.org/search.pl?searchtype=quick You can find a very good copy of the R&B manual here. It has a back box wiring diagram and full schematics. Maybe it will give you some ideas to try and find the root of the problem.

All coils behave in this manner. Flippers, pop bumpers vuk etc. The diode is connected to both ends of the coil winding. Applied voltage flows through one end of the coil wire to the other end of the coil wire. When one end is connected to ground the coiled wire creates a magnetic field. In this instance the Q 29 transistor circuit on the CPU board provides the ground path for the coil in relay K1. The diodes purpose is to prevent back-emf induced voltage from damaging the drive transister (Q29) when the magnetic field collapses. Think of the diode as a one way gate. Voltage can pass through the gate but not return back through the gate.

In attract mode 38v on both sides of D22 is normal. When the relay energizes as it should in game play there will be 38v on one side of D22 only. The problem seems to be the coil winding in the relay is not being grounded/energized/clicks as it should be when a game is started. Focus on this issue first.