I know it's a long shot - but have you checked the pins in the molex connectors on the ramp switch (under playfield) those pins can sometimes break or lose continuity? I would unplug them and see what it looks like. Check both sides, the pins and the receivers.

John

73a1a43ed726ebbf8a612a3df7cc3dc3 (resized).jpg

Don't give up! themadman has too much invested in your issue!

This may be the longest I have witnessed a single problem go unsolved on Pinside.
Always checking in to see the solution.

proposed fix for that ... "Honey I'm sorry I can't get away right now, but I bought tickets for you and the kids to go to (place favorite place here) for Christmas! I'll miss you terribly but thinking of how much enjoyment you will get from this trip will get me through the long nights alone."

The one thing I keep returning to (which is probably not going to work) is unplugging the clock from the PF and disabling the clock from the settings. If you've done this, my apologies for missing it. But if you are going to install a new clock board, hopefully that is the fix.

Is this still true? If not please summarize the current issue (there have been so many updates it's hard to follow the current symptoms). If it is true I would suspect a high-resistance connection in that wire (often at the IDC connector).

Just a thought, since you have it narrowed down to the CPU did you check the connectors below where the batteries are to see if there is acid damage in the connectors?

I'm not just talking about the board but inside the pins of the connectors. Sometimes acid gets in there and you never even see it

For another pinball I had a switch matrix issue that I was able to narrow down to the CPU by removing the switch harness from the CPU and doing a switch test by jumping the pins. For TZ row 8 it looks like you need to remove J205 and J206 and jump J205 pin 9 to J206 pins 1-9. Do this while in a switch test and see if you get the switch signals. If not, then it is something on the CPU and not in the harness or on the PF. In my case it was a decoder chip that I narrowed down. For TZ it looks like it is U20.

Ribbon cables?

I'm still not convinced it's the CPU, it works in switch test with tape and balls, as all you pics show from your testing:
Making the assumptions that your missed that there are three lane switches on the left (what you called left outlane is actually inlane 1 and what you called inlane is actually inlane 2, and the outlane never got tested).
And about picture 15 your one switch less finished in your testing then you say you are ( I thing you duped picture 14 with another and called the second take 15).

Everything looks beautiful in the testing you did, and the CPU testing you did...

With the row 8 short, I'm still on a ball is causing a chain reaction somewhere (you tested with tape, which doesn't conduct, where a steel ball on a switch touching something else metal will). Or the Gumball geneva switch is a lynchpin, and it was NOT active in the tape test you did, but has moved and gone back to active, and is causing a column to row to column to row wiring issue due to a bad diode on the slot kickout causing a matrix "square" fault.

If the row 8 short was there WHILE all the connectors were unplugged then it is a CPU problem, but if it went away when you removed the connectors, then it's a playfield problem (and I'm still convinced it's on the playfield).

That's real good what you did, Anthony real good! Hahah wrong episode but same actor

I think it is the same episode right? Its his birthday and noone wants to piss him off. lol

Haha yeah! But I think it might be the episode where he talks to his dead grandma on the phone though not for sure!

Ah. I think your right! Good call!

The image was from Long Distance Call, where Granny talks to him from beyond the grave. It's a Good Life is the one where he controls everything with his mind.

Love the Blu-Ray season sets of TZ. Many excellent commentaries.

Gotcha. Good Life is one of my favs, thought this pic was from that one. I like the movie version of Good Life too.

WOO. Yeah I thought it might have been long-distance call. I have the definitive collection

Nice matrix there, but to clarify for future reference:
Switch #71 is technically the second autofire kicker, #82 is "Upper Right Magnet" (not 'Middle'), and #86 is named "Clock Passage", not '4-way')

I just found this thread, and have my thoughts, but since it's so far into the process, I'll just sit back and watch.

Plus, bump.

Wait, what? Explain please….

I had a "switch 71 error" when I first got my machine and through fixing a short in that row, updating the game rom, and replacing the ULN2803, it has never returned. But as far as I knew, switch 71 is "not used" according to the matrix.

To keep it short since it's off-topic -
Original designs had the autofire kicker store two balls in it, for multiball release. Two optos were used for this. Early sample games still had the second opto installed. Switch #71 was called 'Big Kick #1', and Switch #72 was called 'Big Kick #2'. When the top opto (for the 2nd ball) was removed, the #72 was renamed to 'Autofire', and #71 was 'Not Used'. In-game, if you activate the opto, the game WILL open the gate and try to clear out the autofire kicker. So it DOES have code wired to it.

I was just being anal and pointing out correct names for the @themadman's switch matrix chart.

Anal, but 100% correct . It was quick paint editing to remove "not used", and not looking up the "correct" names on the matrix. I figured that someone might have added 82 or 86 to a machine, but I've never seen one with 71 (which is the second autofire), not to say someone wouldn't, but the machine will never (on purpose) try to put two balls into the autofire lane, even if it could sense it. The other two switches can be used and recognized for game play.

But if you don't have good documentation, what's the point in having the documentation, so in the interest of serving Coyote's , and everyone else's needs... A corrected matrix is attached...

What though do you have Coyote? While I still think it's a matrix (wiring) issue, and probably a switch that is breaking/shorting row 8 to ground, I (and I'm sure the OP even more) will take ideas.

We've seen the switches activate in switch test, and by shorting on the MPU, so the MPU is probably not it. He's had ground short row 8 show up once or twice, and everything that is flaky is row 8; I'm still in belief it's a wiring issue, probably on the ramp or geneva switch with a diode shorted to ground, or the clock is causing chaos and taking out row 8. It seems to be a not often used switch causing most of row 8 to disappear (wired wrong thus breaking row 8 when not in use or in use, bad diode, etc...) Could also be something on the opto board.

What we really need is for him to get the machine to not be sensing the ball in the outhole and go into test right there and see if the outhole switch is closed, and if not touch the other switches until we find the one that causes it to close (after verifying that it will close via a short on the MPU.)

tzswithc (resized).PNG

Ahh, yes. I remember someone mentioning a Big Kick #2 error a while back. It's all coming together and making sense now.

I do love learning about all the super early/pre-production stuff with this game. I think I would've enjoyed that roaming jackpot shot mentioned in that other thread. That would've been super challenging!

Well, here are my thoughts on it -
He said he tried the Gumball Machine Test, but I don't like his results - or his description of the results.

Row 8 shorted to ground would affect the game, but when the row is shorted, you'll get other issues - the code wouldn't even get to trying to load the gumball machine.

@op didn't say what revision ROM he was running, so going to assume 9.4H. (If 9.4CH, switch to 9.4H. 9.4CH has some adjusted code that could cause issues, and it'd be silly trying to fix something that could be bugged by the hacked ROM..)

9.4H was built off one of the newer revisions of 'OS'. This version immediately flagged a switch as bad if, on power up, it's not in its normal state. This is excepted for certain switches that have kickouts (rocket kicker, lock kickout, etc.) But other switches, like the Mini PF Enter switch, or the Ramp switch, or the ramp opto will immediately be flagged as bad in the test report if on power up it's not in its usual state.

I digress. I would do the following -
Empty ALL balls from the game.
Go into Gumball Test mode, do NOT select any test.
Drop a ball - a single ball - between the right ramp and the left ramp. There's a plastic there, and the ball may roll into the gumball entry hole. (If not, use your finger and push it in.)
What SHOULD happen is that the Gumball Lane switch should activate, then the Gumball Popper. The popper will fire, ball will enter gumball, triggering the Gumball entry. It may try to eject the ball - that's fine. If it does, catch it out of the slot machine.
If that works as intended, then take another ball. This ball, roll up the right spiral. The game should see the ball activate the R Magnet opto, open the gumball diverter. When the ball enters the hole, the process should repeat the first test - Gumball lane, popper, entry. Again, may try to eject a ball from the gumball. If so, catch it out of the slot machine.
If THAT works, now we test hte autofire kicker. Manually open the shooter diverter with your finger (lightly!) and let the ball roll into the autofire kicker. The 'Autofire' switch should activate on the screen. Then, everything should follow as above. (Magnet switch, lane, popper, entry.)

If that works as expected, then our problem is in the trough.

A little different approach to the same thing I've been trying to get him to do... Methodically test each switch with a ball.

The entire gumball sequence actually works, once the forth ball makes it into the trough, since none of the switches for the gumball load and dispense sequence (outside of the outhole) are in row 8.

One of the problems is getting a set point to work from, with a set of known conditions, and troubleshooting from there; the random board changes, and wiring changes, and game plays, just set us back to an unknown state each time, and we don't know what works and doesn't each time.

There is a row 8 problem, each time it fails it takes out the outhole, slot, and rocket switches, and probably the rest of row 8 too, but since the first three are used almost immediately at the start of a game, that's the ones he notices. (and he didn't even notice the outhole, he thought it was a gumball machine problem, I only saw that inside one of his videos).

It really is a row 8 disconnected, or shorting problem.... Just what and where is it failing...

My advice... do this to the broken one, and buy a new one.

ospace (resized).jpg

Are there too many balls in the machine? I know, dumb question. But, it is something I would do. Do you have the "add additional ball" option activated?

Nevermind my comment. Getting this issue mixed up with another JD one I'm following... Good luck man. Frustrating...

So, wait - what is the actual problem? I thought it was the gumball not working, but.. if that's not the cause, you're just trying to narrow down the cause for the ground short?

If the latter, the opto board won't help, as it's not connected to Col 8. (This is the aforementioned 'shot-gun method' which isn't recommended here..)

It's been a long day at work, apparently I'm just confused now.

Still checking in here for the "OMG, I found the problem." Post.

2nd that

It was originally listed as the gumball not working, but that was a mistaken diagnosis from the OP. He had emptied the balls from the machine, done some work, ramp removal, mods, etc. and when put back together, he dropped 6 balls into the outhole expecting the machine to load the gumball machine, since it had always done that in the past. When it didn't, he assumed the gumball machine was broken.

After watching his video, it wasn't that the gumball machine was broken, but that the 6 balls never left the outhole area and got moved to the trough. He kept trying to start a game and couldn't, so he started moving balls over by hand to the trough, and when three got there, the machine started a game (since it assumed the other 3 must be in the gumball machine). It never loaded the gumball machine, simply because there was never 4 balls in the trough to get the gumball load sequence to start. Once four balls were manually put into the trough, the gumball machine loaded.

Since a game could now be played, ball searches took care of moving from the outhole to the trough (the outhole switch isn't registering), and from the rocket kicker on a skill shot (also not registering), and from the slot machine when hit or shot from plunging (also not registering). But he was missing that it was a ball search moving the game along when the ball "sat for a long time".

I don't think we've tested all the switched in row 8 yet, but the three he complained about are in row 8, and he's also had a startup error (before switching out boards, and roms) that gave a left inlane switch 38 error, or a ground short error row 8. All problems are in row 8.

Removing the connectors on the MPU and jumpering between row 8 and each column works fine, and so does every other column to row, so we are down to a switch or switches, or a loose wire somewhere causing a row short/open on row 8.

The problem is, whenever he goes into switch test, he seems to get row 8 to start working. So my guess is a row 8 switch wire is touching a lamp matrix wire somewhere near a switch (thus the confusion between ground short and all row 8 open), and why the switches seem to work in edge test, but not in game play (since the controlled lamps are all off during testing, but on during game play).

It could also be a switch that is solidly in one state or another for a long period of time (geneva, ball lock, clock optos, etc.) that is sometimes changing state and is causing the short.

I was on the diode shorted to ground on a switch when made at first; while still in the same ball park, I'm more thinking a lamp matrix to switch matrix short...

Do you have one or two switches on the coin door (does power drop to the coils/flippers when you open the door?)

-->
One thing we haven't tried is to remove the controlled lamps power and see if the machine acts normal for switches (with the side effect of no lamps). That could help tell if that is the problem.

OP: Disconnect J119, J120, J121, J133, J134, J135, J136, J137, J138 (Any, and all, if present) marking the correct ones and telling us which were populated, and play a game. Did the machine function normally (minus not having any lights that work)?

If YES, then we know it's a lamp to switch short problem.

If NO leave the above unplugged and OPEN the coin door, go into test see the outhole switch working by placing a ball in it. If it registers then Close the coin door, does it still register?

If NO: then a short to a solenoid.

If YES: to be determined.
<--

I've mentioned this before but I'll try again. If the switch matrix works during test but not during a game this can be caused by a high-resistance connection in a row or column (in this case row 8). During play the earlier switches in the row will work and the later switches in the row will not work. The most common cause is a wire that is not properly seated in the IDC connector.

Of course this does not explain the ground short error.

In regards to a short between the lamp matrix and the switch matrix my experience is if that happens it will typically blow the chip on the cpu. Not saying you should rule out that option, just sharing my experience. Removing the power to lamp matrix while in game mode as you suggested is a good test.

I agree with you on the high resistance as a possibility, and it's where I was to begin with, a broken or almost broken wire on a switch where two of the row wires meet to bridge to the next switch. The row seems to get worse the farther away from the head the switch is based on rough distance theoretically past a break.

But the weirdness of the whole thing of everything seems to work in test mode, not in play, and the ground short, there has to be something else in play that changes when in game mode; controlled lights, solenoid power (I don't know if there is a safety switch on the coin door, nor if it's always closed during the game play when it malfunctions, and always open during testing, another possibility), geneva switch moves, etc.

Usually a lamp to switch matrix short would blow something on the cpu, but I'm open to one of the mods he's installed being LED based, and having a limiting resistor on it, so it's not blowing out the switch matrix, just confusing the heck out of it; if the mod is hooked to the lamp matrix for power and then somehow shorted to the switch matrix, we have a very plausible explanation. Add to that that the machine stopped working after he took it apart to add some mods... and the percentage of probability goes up that this is it. Especially since it has to be something that changes between testing and game play...

Actually that's a rather common indication of this problem. When the system is under load it will show up more than in test.

A close-up photo of the IDC connector would be great. As you mentioned this can also be a frayed wire, which could also explain the ground short if some strands are touching ground.

Sweet please report back! What is $10k though? A new fully decked out mint TZ?