In your case you mention they are stuck on and your markup of the schematic suggest a pairing of output stages. If we back up through the circuit we see that U2 is driven by a pair of 2 to 4 line decoders in U1. But if we back up one more stage we see that the inputs of both U1A/B are driven by Q17/18. So in theory if one of those 2 2N4403s is bad then that would result in bad inputs on both U1A/B and half of each of the decoders not working. Since each decoder has 4 outputs, that would be 2 output channels each for a total of 4 output channels. So I'm thinking that one of the 2 input trans is bad or maybe bad connections on J2. The only thing that isn't lining up with that diagnosis is that the pairs of stuck outputs from U2 don't appear to be symmetric. So I wonder if you maybe marked up the schematics incorrectly above?

Not specific you what you list above but note:
The lamp board on this game wasn't going to last. The board itself is probably completely cooked (mine is) and the connectors on the ramp lamp boards are probably flakey too. I ended up cutting out all the triacs and resistors in the output section attempting to clean the board. However the heat damage is so bad all sorts of traces and pads just fell off. So I had to make lots of fixes and simply replaced the output section with new which corrected most of my issues. Someday I'm going to start making gerbers up for all these WMS boards so I can make OEM looking replacements. Which in the long run isn't really going to solve the problems but the game isn't ever on for 16hrs a day 7 days a week either.

Yea, you really need to be super careful about removing those parts anyway but the cooked boards make it worse. I destructively removed all the parts (cut all the pins) and then just cleaned each hole out. But cleaning up the pads is what killed stuff. Just peeled back when I was trying to clean them. Anyway, it's an uphill battle even for those of us with a bunch of exp.

Transistor testing with DMM via google is your friend.

Checking the transistors may tell us but it doesn't match your schematic markup, we'd expect symmetry on U2.

I don't undestand this statement???

Don't be too discouraged about your board repair skill.

I can think of a lot of things that require board repair in pinball that are easier than this board.

One thing that you might consider the next time you have a board repair problem.

On the schematic, Q17, and Q18 are connected the same way. So with your DMM, you should be able to read about the same things on them, even when they are still connected on the board. If one transistor's numbers on the meter were a lot different than the other, I would THEN pull the transistor, and test it out of circuit.

When testing electronics, usually in order to get good readings you'll have to pull the component off the board... you take it out of the circuit in order to find out if it tests OK with your meter.

But if you've got duplicate circuits, you can quickly see if things are reading identically.

On this board, you can quickly check Q1, Q2, Q5, Q6, Q9, Q10, Q13 and Q14, which are all 2n4403 transistors, all hooked up the same way. Even though you aren't going to 'transitor test' these until you pull them out of circuit, if any of these has a different reading than the others, that'll be a good thing to desolder, pull out of circuit, and test. If they all read identical, you probably don't have a problem with these transistors.

Finally, I run a lot of 'bug wires' to connect things that have broken traces. Point to point wiring can get your board working again. It's not pretty, but it can get the job done.

Don't be discouraged because this ended up being really difficult. There is a good chance your next board repair will be much easier.

Ah, ok. I kinda thought that's what you meant but I just wasn't reading that. Yes, that would have been a good test to prove if it was on the board. Typically want to do that before you start working on the board though as you can (as you are no doubt aware) create more problems.

As it appears to be a non symmetric error (you confirmed it's not an error in your markup), I'd be going after the U1 or U2 after your swap the inputs test. Those are TTL level signals so you could probe the input/output pairs of U2 and that would probably tell you which of the 2 chips it is/was.

Zero Ohm across R9 is more likely the transistor Q18 shorted between 'base' and 'emitter'.
Easiest way is to lift one leg of the resistor out of the PCB, then measure the resistor (out of circuit).
When it gives 4K7 , the resistor is OK, and transistor is shorted.

chase drv (resized).png

First off, are you doing these tests in via the service menu or just testing while in attract mode?

Do you own a logic probe? The easiest test to perform at this point is to put it in chase test and probe the inputs/outputs of the LS139. Since NONE of the chase is working, simply take the A side of U1 (U1A) and validate the the input signals, U1 pins 2 & 3. Are they toggling? They should be. If yes, move to testing the outputs, U1 pins 4,5,6,7. They should also be toggling.

When one leg of the resistor was removed, you could measure resistance on the solderpads of the resistor to get the resistance between 'e' and 'b' of Q18.
Expect this shorted. Could be Q18 or traces shorted by excessive solder. (looking at this section, this might be possible)
Q17-18 (resized).png

Ok, so something between D1 (J2 pin 1) and U1A input A (Pin 2) is bad.

First: Understand the truth table for the a 139. Here we have to select elements of the table based upon your signal tests above:

pasted_image (resized).png

We're only interested in the section of the table where G (Pin 1) is Low "L". That gives us 4 rows of Select signals (A, B) and Outputs (Y0-3) to evaluate. Now, your testing determined that Pin 2 stays Low (Select A) so that means that only the rows where Select A is listed as "L" are valid options. If you examine you'll note that Y1 and Y3 (Pins 5 and 7) are always High "H" and that Y0 and Y2 toggle with Pin 3 (Select B). This is what your data shows.

pasted_image (resized).png

It's also why only half the lights flash.

U1A Select A (U1A Pin 2) is fed from D1. Your testing says D1 is toggling. So we'd expect Select A to also be toggling but it's not. So there's something between the 2 that's not working. As you already swapped out the transistors and tested the resistors, and thus I'll assume they are still good (and not blown again). Then I'm wondering if you should test for continuity between all the components between D1 input and U1A Pin 2.

pasted_image (resized).png

Off the cuff... Validate that the header, J2, doesn't have any cold solder joints.
And that the wire is sufficiently seated into the IDC connector.

Shouldn't be using Logic probe on the triacs (Or the inputs D1/2) as the probe is meant for TTL based logic levels. Those devices are operating between 0-12v where the probe expects 0-5v.