Are you jumpering the reset circuit on the bench? If not, you need 12v in addition to 5v

If so, time to break out the logic probe and see what’s happening (or not). Check for CPU clock pulsing, check the test points to see if they are behaving properly. This pdf has graphs that show what you should see at each test point (this info is also in the Williams schematics but this is easier to read in the linked version).

http://www.firepowerpinball.com/downloads/CPUBoard.pdf

If you try to measure clock signal at the crystal, your scope will load it too much, causing drop in amplitude and frequency. The clock is probably OK since you get pulsing waveforms elsewhere.

Test point 6 signal is OK even if it is not exactly square.

The 12V is unregulated, so there might be some variation. But I think it should be more on the over 12V side. Check D7 and D8 on power supply (failure in one of those will also cause excessive ripple in +5V) And of course the filter cap C15.

I’d leave the driver board disconnected for now.

12v is only used by the reset circuit. If the board isn’t being held in reset, I’d ignore that for now.

Assuming you are using Leon’s ROM. What are you seeing on the LEDs? What happens when you press the diag button?

Looks like your basics are running. Pick a PIA and probe the data bus pins and address bus pins. Look for one that isn’t pulsing. If you find one not pulsing, trace it back to the CPU and see where the signal dies.

You removed all the ROMs when installing Leon’s rom, right?

If any of the memory or PIAs are socketed, remove them to reduce the possible failure points.

Try a test ROM that checks the PIAs on the driver board and more thorougly checks the RAM on the MPU. Leon's test is just good to make the PIAs cycle. The memory checks in that test ROM have been unreliable. It will report the wrong 6810 as being bad and etc.

The MPU will not boot up without the driver board connected, but you should be able to get far enough to use the diag button on the MPU to see if it reports a RAM ROM issue. I would pull the driver board off, power on to see if it still is totally locked up.

Test ROM only runs from 1 flipper ROM spot. Two more sockets are used with real game software, make sure no opens on the other socket.

The random starting and not makes me think about IC sockets. Did you replace the CPU and at least iC 17 rom socket?

The RAM socket or any of the other sockets will not cause Leon’s program to not start. I’d focus on getting it to start reliably.

Ok, you say D0-D7 on IC16 PIA are strobing. Do they all sound the same? What about the address lines?

Try pulling all the chips you can other than cpu, rom and one pia. Check again.

Do you have any new PIAs to try? Sounds line you might have a bad one. If all the data and address lines are pulsing with Leon’s rom, all the outputs of the pia should be cycling.

Totally possible and probable. That’s why you remove as many ram, rom and pia chips as you can to reduce the failure points. Do not remove other TTL chips as some are addressing etc.

It’s also why you listen to the tone on your logic probe of the address and data bus. You are looking for one that is stuck high or low or is pulsing significantly different than the others.

Check the chip selects on the PIAs to make sure they are pulsing. If they are not, trace up the path to the addressing chip on that select.

Do you have a new CPU chip to try?

At some point, I’d just pull and socket the PIAs. Especially if one is AMI brand. Then you can insert them one at a time to see if one is squashing something.