Medieval Madness repair from hell

A friend asked for help with his Medieval Madness game.. suddenly the left popper (which returns the ball from inside the moat) activated about 10 times when the game was switched on.
I told him it would be the opto that was bad, so at startup the game thinks there's a pinball laying in the moat and tries a few times to kick it out. After about 10 times the machine decides there's a problem (either with coil or opto) and gives up. During gameplay the ball isn't kicked out until the game starts a ball search.
He had already bought a new opto pair and connected it in place, with the same result.

Weird.. I'd think the opto transmittor is burnt (which is usually the problem) so replacing both the emittor and receiver should solve it. When I got there last week I noticed one of the wires going to the connector where the optos plug in was loose. A few drops of solder fixed it, wire made contact again with the pin. Measured with my dmm to verify there was voltage. Checked continity from the pin to where it comes from on the 10-opto board, everything ok.
Measured all pins and voltages on this 4-pin connector, everything was present. So the other wires were also good.

We traced the wires back to the opto board, measured continuity. Everything was good.
But in switch test the left popper opto never worked. It's switch nr 26 btw if I recall correct - I don't have schematics here so could be wrong in the switch numbers. Tested some other switches and optos, they worked fine.

Put in the new opto pair, didn't work. Connected it into another position, there it did work. So the opto emittor and received were fine.
Checked with a digital camera, the emittor gave light. Lit with a flashlight in the receiver - didn't register in switch test. It did when the opto pair was connected in another opto position. So the opto was definitely good.

Checked what pins on the 10-opto board were used for this switch, followed traces on the board, measured some resistors and diodes.. didn't find anything wrong.
Didn't want to spend more time troubleshooting the board (didn't have a manual with schematics) so took the 10-opto board home to test it in my MM. Only way to be sure the board was ok.

In my Medieval Madness pinball machine it worked fine, the left popper opto registered. I did notice capacitor C1 was missing on this opto board.. but it worked fine in my game so probably wasn't really necessary.

The MM cpu was also brought to my home this weekend. Tested it in my game, it worked fine, all optos and switches registered like the should.
So the cpu board also didn't have any problems. No problems with LM339 ic's whatsoever. The problem had to be in the playfield wiring of that machine and not in the boards. A good thing because I didn't want to start replacing components on boards..

So yesterday evening I went back to this MM to further troubleshoot the opto problem.
Put all the boards back in. Measured voltages on optos themselves - got 12 volt, when optos were aligned it dropped to around 1 volt. So the opto did get its voltages and worked but somehow the cpu did not register this.
Again measured continuity to make sure there were no broken wires between the optos and 10-opto boad.

Looked at the switch matrix and tested all switches around this switch 26. So 16, 36, 25 and 27 (top left lane, left inlane, right troll, ball through 4, castle entry opto).
All worked fine. So this indicated there are no wire breaks and correct diodes on the other switches ?

Tested with unplugging the opto and plugging it back in. Suddenly the game registered the opto ?! So somehow it did work.. Yes it works !! Problem solved ???
Turned the game off, switched it on again, go into switch test.. nothing ?!
Damn.. it's broken again.

OK so we somehow have an intermittent contact ?? As it suddenly worked and then stopped again.
Measured continuity on all the connectors and wiring (again).
Wiggled with the connectors to see if it suddenly made contact. Nothing.
Checked on the board if there were somehow broken traces. All looked fine.

Then I remembered the missing capacitor C1 on the opto board. Maybe, just maybe, this would filter out something and make the signal for this one opto just a little bit stronger ? You never know ??
I know it's a weird theory as the board worked fine in my game and other optos should also have the same problem but you never know.

At this point I really had no options left.. couldn't think of many possibilities.
Did a factory reset, even completely remove the batteries.. as it could be a phantom problem.. game thinks at startup the opto is bad and goes in a diagnostic mode and just ignores the switch (even if later it starts to work again). Seen it happen before in my Whitewater when a closed switch was ignored forever by the game..

Factory reset doesn't bring a solution. I now start just guessing what it can be and testing things over and over again.

Tested again with the other opto. Moat enter opto (switch 31) registered fine all the time.
Switch 27 is the castle enty opto which is the next opto in the same column. Maybe there was a problem with this column ?

Back into the switch tests. Drawbridge test (to lower the bridge as the game always puts the bridge in the high position when you switch on the machine), lower bridge, test opto 27 castle entrance.
Works fine. Test opto 26, left popper.. hey suddenly it works ?!

We may have found breakthrough here.. Opto 26 works after opto 27 is activated once ???
Maybe there's a problem with this capacitor missing on the board and one opto needs to be activated before the other one will work ?? (very weird theory but after spending too many hours looking at playfield wiring I start to believe anything you tell me)

We test further.. switch game off, game on, opto 26 doesn't work, bridge down, opto 27 works, opto 26 now also works.
What's going on here ???

Bridge goes up again and opto 26 doesn't work anymore. Hey, wait.. that's really weird. So the problem is not related to opto 27 but with the position of the bridge ??

Ok the motor of the drawbridge also works on 12v, it's even the same grey-yellow wire that goes towards the optos. Maybe the 12v is incorrect (and this C1 smoothing capacitor filters out a ripple caused by the motor ?) Measure voltages for the motor is activated, see nothing special.

Back into the switch test. Test again: bridge up: opto doesn't work. Bridge down: opto 26 works. Test again and again. Same results. OK now we can reproduce the problem we can really look for the source of it.

I put the bridge in a halfway position, go into the switch test.. manually activate both microswitches that register the position of the drawbridge.
Up position: microswitch works fine.
Down position: microswitch works fine. But what do I see on the dmd in the switch test ? Opto 26 disappears when this opto (number 46 btw so same row but 2 columns further down the line) is activated ?!?!?!?!?!?!


Test again, and yes.. somehow if microswitch 46 is activated, opto 26 just disappears from the switch matrix.

I check the wiring on this microswitch as there's has been soldered on it and notice it.. the white wire is in an incorrect position !
There are 3 legs on a microswitch, one has the white wire and the non-banded end of the diode, middle has a green wire, 3rd has the banded side of the diode. I check with other microswitches and the white wire is incorrect, the correct sequence of microswitches is: non-banded end of diode (nearest to the pivot point of the metal arm), green wire, and 3rd pin has both the white wire and banded side of the diode.

The owner tells me that diode once was loose, the bridge didn't move at all anymore, and someone (who knows something about electronics but nothing about pinball repair) soldered it back into place.. Seems this person did solder the diode correct but not the white wire !

I solder the white wire in its correct position and now everything works fine.. YES YES YES !

I'm very glad to have solved this problem.. but it took me about 6 hours in total to solve it. Aaaaaaaaaaaaaaaaargh.. all that lost time just because 1 little wire was not in a correct place because someone who didn't know a thing about pinball machines thought he could do a repair..

Weird thing is that microswitch 46 itself worked fine (so the game never indicated an error with it) and the other switches around it in the switch matrix also worked fine.

With switch matrix problems and diode problems I always expect the problem to be with a switch directly in the next/previous column or row. So with switch 26 the problem can be also with a diode of switch 16, 36, 25 or 27.. but not with a column even further away ?

And the other stupid thing that made us lose a lot of time is that this microswitch 46 is totally hidden below the playfield (it's part of the drawbridge assembly) and the game always immediately raises the bridge when switched on (and thus activating microswitch 46). It's not a switch you immediately test when you're in the switch test (and it already shows as active so why should you test it manually again ?)
as you have to go in the drawbridge test to lower the bridge..