Diagnosing a confusing switch matrix issue.
On a TZ, the U20 chip (2803) will blow after a few plays.
Hence, with all connectors plugged in, we are finding some overvoltage on a few of the rows (white wires). But none of the columns (green wires)

Most lines, such as the green wire connectors, are at approx 11.5volts but a few of the white wire leads for rows, are as high as 13.9volts
So unpluging a number of connectors to find which ones allow the voltage to drop has narrowed it down to the OPTO switch circuits being suspect.
(Note that NO change observed for disconnecting Lamp, Coil or Flasher circuits.
and all underplayfield micro and leaf switches inspected for shorts.
Also, as information, CPU, Power Driver and 10-opto boards have been serviced.)

So, overvoltage is the problem.
General Question: Is a reading of 12.5volts tolerable?
Apparently so, as I measure some voltages on a STTNG, at least one of the circuits is at 12.6volts.
Can anyone confirm for sure considering the general "plus or minus 10% tolerance" principle for components?

Next, disconnecting all the opto sensors connectors under the playfield, I see a drop to the 12.5volt level and then reconnecting certain ones (2 or 3) will then increase back to 13.5+volts.

Question: So what about the simple Opto sensors could raise the voltage? Only a short or some other failing of components? (Apparently this game took hit during a lightning storm)
No voltage change when plugging the clock opto connector but the Power miniplayfield is one that boosts the voltage back up. I have yet to check for pinched wires, etc under that playfield.

Any helpful thoughts would be most appreciated. Other questions to clarify are welcome to help all learn from the discussion. Thanks.
- Paul

Thanks for the explanations ... so it seems my approach may be flawed, understood.
So after reviewing this, I decided to test if the chip is truly getting damaged. I never pushed it since bringing into my shop ... it only happened before at the owner's home. But as you will see here, the problem remains.

BTW - Isn't J205, mentioned above the dedicated switches for the coindoor test buttons and coin drop switches? I thought the J206/207 was for columns.

In any case, if the game sits powered on for 15 min, apparently it seems OK.
and while it is OK, I was able to test all the J205 lines and ALL 8 read consistently as between v11.77 and v11.8

I decided to be sure of what I am dealing with, I plugged everything back and measured the "relative voltages" on J206/207 and they measured 11.7ish steadily with my Fluke115 DMM.
BUT for J208/209 ... Pins 9 read at v13.8ish and Pin 8 at 14.02 volts! The rest read safely around 12 volts. (I observed no fluctuation of the readings...is that a due to the quality of the meter?)

So I start a game and after about 5 min, things go crazy. ... the 12 volts was lost.
(I never entered the mini powerfield.)
F115 blew and I tried to replace it to see "which columns quit working", as suggested above.
However, the fuse immediately blows when turning on the power and it won't allow entry into the switch test menu.
I then replaced both U20 and F115 and now the game powers on OK...and it's been sitting for over 15min now with no crazy events.

Regarding 'identifying a particular game play' type is not easy to identify. Do you mean triggering certain switches or playing it hard with flippers or something like that?

Keeping on plugging along...

ok ... I unplugged all the Lamp Matrix related connectors (J133-J137)
and just played a 15 min game including Lost in the Zone (yes, I had the glass off
Plenty of flasher and coil action ... all optos worked fine. But I will test more.
So, I will try again tomorrow to see if it still behaves.

Then the fun of tracking this down.
I did measure the suspicious J208/209 I have been commenting on (white wires) where pin 8 was 14volts high ... it is now only 11.5volts.
The pin 9 remains unchanged at 13.7ish.
The rest are down to 11.7ish.

Interesting...

Interesting point on sharing the 12v unregulated ... no issues around the boards/connector area but I may need to look deeper in the harness.
Trying to get back to this project ... among all the others I have.
It's only the one line - J208/209 - pin 9 that remains high at 13.8ish volts in attract mode.
Even with everything plugged back in, all other lines are around 11.5-11.6volts.
So what switches are on that white grey line -
Outhole, Rocker kicker, Left inlane, Lower left 5mil, slot kickout, Lower Rt 5mil, Lower lock, Clock hour

I will assume that the clock has no issue since unplugging the opto connector does nothing to change the voltage. The rest...hmmm...We will see.

ok ... back from a trip and hoping to tackle this further. June 5th

Everything plugged in gives about 13.76v on White-gray.
So, I am testing by unplugging 12v J116,J117,J118 ...
Results: 2 connectors - not the thick grey wires --> the voltage goes down to 13.1v
Pull the thick connector and --> the voltage goes down to 11.76v ... equal to all the other white switch wires
Reconnect the thick 12v --> the voltage goes down to 13.1v

One of the other 2, reconnect and no change.
The other of the "other 2 thin Gray/Yellow wire" , I hear a slight snap/crackle as I connect and the voltage goes back up to 13.76v ... so I will begin tracing that circuit.

(June 20th) Finally spent some time tracing the 12v gray/yellow wire through the harness and saw what devices it supplied voltage too.
These included the clock, the 10opto board, the 2 eddy sensors for the powerball and a couple of other connectors.

Unplugged all and saw that the voltage went down to a tolerable 12.46v on the White/gray switch wire.

Everything plugged in would leave the voltage unchanged EXCEPT for the 10-opto board.
(Great, it just had to be the biggest trunk of wires!)

So, the large looped connectors, slipping it on with power on, I think I heard the same little snap/crackle as mentioned above.
Then with the power off, I unplugged all 4 of the other connectors, leaving the looped input connector and the voltage did not jump, still at 12.46v. I am assuming this means that the 10-opto board is not the cause. This 10-opto board was completely rebuilt by Clive.

Then, with the power off, I tested with 2 sets connected. First 2 right side and then 2 left side...
Results: Each of these tests would raise the voltage measurement to 13.5v.

So more tracing needed to see which opto circuit is causing this situation.
Thanks for any comments.
- Paul

(June 29) Time keeps moving.
Update: I explored each destination OPTO off the 10Opto board to find out which switch is causing the voltage to rise.
Out of the 8 connectors along the circuit ... 3 of them raise the voltage and 5 do not.
Without these 3 plugged in, the reading is 12.5v which I believe is tolerable.
Then connecting one at a time:
1 - the connector to the ball lock area ... 2 optos there - rises to: 12.9v
2 - the ball popper that feeds the gumball machine ... only 1 opto set - rises to: 12.6v
3 - the connector to the mini playfield ... only 1 opto set - rises to: 12.9v

1 and 2 = 12.9 - 1 and 3 =13.1 - all 3 = 13.3
This is with the game in attract mode ... not sure if it would rise during gameplay.
Obviously, I have to disassemble areas to examine.

Now, what might happen with opto transmitter/receivers to make voltage rise?
Hmm...I have heard of a leaking opto, is that relevant?

Thanks.

(July 4) So after fiddling for a couple of hours, I decided to leave the powerfield optos unplugged (2 sets actually...one at the exit where he laughs when triggered and the other at the goal area...just produces some sound fx.) This allowed me to play 3 long games...tolerable voltage at 12.9, I assumed. The game played fine, except he would laugh on exit and at least one pop had to hit on the way down or else because the flippers are disabled while on the powerfield, it would drain but serve the ball back as a ball saver. No real need for the sound fx triggered by the top/goal optos.

HOWEVER, then I measured after the 2nd game and found the white/gray measuring the normal 11.68v.
Then I played a 3rd game and measure after that and saw the voltage back up to 13.46v.
Why? ... I asked myself.

Turns out that if there is ONE ball locked in the lower lock (micro switch, not an opto), then the voltage drops to 11.68 just like all the other switch rows. So, I plugged in ALL the connectors and only put a ball in the lower lock and it maintained a normal ~11.6v ... taking it out made it jump to ~13.5v.
This micro switch is the same type as at the outhole. And it resides on the white/gray row.

SO...I hope to change this switch and maintain the 11.6v.
What kind of explanation would explain this?
There are many spots of corrosion around the game ... I believe it was used
around the shore area for it's lifetime, so possibly affected by salt in the air?
So could something internal to the normally open switch cause this 2 volt jump in the switch circuit?

STRANGE QUESTION: I was curious if it was an reimport and noted that the 2 manufacturing labels specify AMPS = 60 rather than the usual 5 (for export) or 8 (domestic). First time I have seen this. Is it common and explainable?

Comments are most appreciated.

or just the diode? which I will change first

Here is the label inside the cab ... looks like a typo where they switched the Amps and the Hz values.
TZ_AmpsTypo.jpg

Well, it's not the diode and when I take the switch completely out of the game... it remains at ~13.6v.
Both are Doubled leads and remain connected for the daisy chain to be intact.
So, having them separated with no switch installed ... it jumps to ~13.6v
And connecting with a good diode (white wires get non-banded/green get banded side attached)
then it drop to the normal ~11.6v

So it seems that this would indicate that a new switch would make no difference here.

Very interested in what this could mean.
The switch discussed here is #88 Lock Lower.
In testing if some other switches in the column or row make a difference, I found that both optos on the powerfield WHEN BLOCKED would cause the voltage to drop a few tenths and jump back up when light flows across.
So far only this one micro switch seems to make the 2volt difference.
Hmmm...

ok ... turns out that engaging any switch on the white/gray row #8 will lower the voltage to ~11.6v.
(outhole, rocket kicker, left inlane2, 3 standups and lower lock)
Releasing them puts it back up.

So, If I make a stuck closed situation on just one of the standups ... I guess the game would be safely playable. <grin>

ok ... so now I engaged multiple switches on that white/gray row.
All open ...voltage measures ~13.5v.
Pressing and holding one switch ... voltage drops to ~11.6v.
Then holding other switches at the same time.... each engagement lowers the voltage by about 0.1-0.2v further.
So, holding 6 switches simultaneously lowers it to ~10.9v - (3 standups, outhole, rocket kicker and left inlane 2)

Terry - not sure what you mean by 'partial short'...something compared to a 'full short'? or do you mean something intermittent?
I'd like to try what you suggest but by 'closing a physical switch', do you mean a switch on that same row or any other row?

===
Something I tried was to impose the 'stuck switch' condition on one of the standups of row 8.
The game played fine for about 7 games and then U20 chip blew, along with the 3/4A F114 fuse. Wanted to get a look at what columns blew up but putting in a new 3/4Amp fuse just blew right away when turning on.
So a new U20 chip was required to keep the fuse intact.

This condition supposedly kept the voltage normal on row 8 and not having one target work was ok. It was a pleasure to play for a bit but then a stab of pain when it failed.

So, as I said, it blew up. Then I remembered that there was more than one row with the high voltage and I checked to find that row 7 was up to ~14v. And closing any of the switches on that row would have it drop right to ~11.7v with the .1 or .2 drop for each additional switch closure.

So as an experiment, I found another Greed standup that is seldom hit and am shorting that to a 'stuck closed' status. The whole row has dropped and so I will play with it this way for now and see if another U20 blows.

All other rows measure around ~12v for now.

BTW - I did try to disconnect just the Row 8 wire and hit various switches...none affected the voltage like other row 8 switches would.
The CPU and PowerDriver and 10-Opto boards have all been recently serviced by Clive.

ok...taking a breath. Thanks.

I have all 4 U20 - (ULN2803 18-pin) chips that have blown and
did continuity test from pin 9 (ground) across to
pins 10 (+12volt) and 11(column through 18 (column 1) .
Each chip had a different set of shorts:
#1 Pins 10, 16
#2 Pin 10 only
#3 Pins 10, 11, 14, 16
#4 Pins 10, 11, 12, 14
(Tested a new chip and, of course, no shorts exist.)

Hmmm...so what does this indicate with only pin 10 common to all chips?

@Terry - Not sure what you mean by "closing a switch seems to cause problems" ... the game played completely correct for a number of games. So, do you mean like phantom events or double switch registering or unusual lighting effects? none of these happened.

I completely inspected the pop bumper area and it is very organized in regards to lamp leads and switch circuit wires. But I will keep that in mind for other projects.

Two previous events that could be making this difficult ...
1) this game apparently was subject to a lightning strike, I was told. So, I made sure the 3 boards were not a question mark by being refurbished.
2) The game has been in salt air by the shore for a good deal of time...enough that there are many spots of corrosion on various metal parts. Could this lead to some tiny continuity within switches? hard to say it would make a difference to grounding a column if no physical crossing of wires. I would think only effect would be intermittent switch event or poor contact, like inside a light socket.

hmmm...

Been a little while here ... still tinkering. Everyone's thoughts are appreciated!

I got a suspicious feeling about one of the flipper opto boards and installed a new one.
Then I got to play about 25 games! A record before the U20 chip and F115 fuse blew.
Almost had a good night sleep that time
NOTE: This is where I also permanently closed 2 standup switches to knock down the voltage on rows 7 and 8.
So these three changes appeared to extend the playtime.
When I inspected the chip for what circuits were shorted, I noted that ONLY Column 3 was damaged.

Today, I started monitoring the Column 3 a bit and began playtesting ... the chip blew after 1.5 games during Fast Lock multiball.
Inspecting this 2nd blown chip ... columns 5, 7 and 8 were damaged. No consistency!
Terry...I was going to start pulling wires as you suggested but now I am not sure if that would help.
All along, I have been making sure no other contacts by leads or diodes.

@zaza - your test sounds different but I am not sure how to implement.
Are you saying that I am checking for a cross between reg12v and unreg12v?

So then, am I correct in saying that the regulated12v is supplied to the switch matrix
AND the unreg12v is what drives the OPTO switches and that the 2 should not cross?
Hence, if there is continuity between the Gry-Yel and the switch matrix ... it would BEEP. But normally, they should not?

Pictures of the area around the 10opto PCB. Let me know how this might help.

20150718_131903.jpg
20150718_131757.jpg
20150718_131809.jpg
20150718_131819.jpg
20150718_131836.jpg

Getting back to shop projects here ..I care about the problem but it gives me a headache :O
(It is 18 day later! Aug 11)... I believe these last comments from various ones have been enlightening.

Especially Tony's comment about the input voltages being fine for up to 30volts... I needed to know what the limits were.
So, I am more convinced as well that the voltage readings are confusing the issue (per Terry)
And I will study more from zaza's post too.

I will post more over the next week ... thank you all.
This is a stubborn issue.
- Paul

Good questions Maniac - The CPU and PowerDriverboard have been fully serviced, especially after having concern about the effect of the lightning strike which is, I am told, when the problem started.

How did you know I have a Whitewater...one of my favorite games?
It's hasn't been setup in a while but I agree, that it might be a good swap test machine.

The clockboard is a single layer with LEDs installed but don't know what brand.
I believe I saw it get damaged when the clock optos were unplugged but could something there really affect the U20 chip?

@Radius118 - the U20 is not blowing up violently...it just takes a hit somewhere and the game starts acting crazy as the 12v switch matrix fuse blows (F115 3/4Amp SB) ... of interest, if the chip is not changed along with the fuse, it blows immediately because of the shorts to ground.

It has been consistent over the past 5 chips that changing the fuse and chip will allow the game to play a few games.
One time it was over 20 games and I exercise all areas of the playfield with my play/test style.

I decided to at least run the burn-in cycle ... 2 hours and no incident. ...for those that were curious.

ok ... so I decided to swap the power driver board with a freshly serviced one,
and leave unplugged various circuits and disabled the gumball and clock.
The U20 chip and F115 fuse, died after 3 games.
So the serviced power driver board is not at issue. Still same CPU at this point.

My friend and I spent over an hour inspecting various switch areas for contact issues. Not find any recognizable.
The one area touching is the inlane switch mounting base touches the underplayfield support metal, which is grounded but I believe that would make no difference since there is no ground row short.

For my playtest, I left unplugged:
- J120, J121 for the GI ... so does that mean conclusively that there is NOT a short between the GI and switches? I would say yes.

- J138,136,135,134 ...all related to insert lights ... so this feature is not at fault. correct?

- J125 which controls some of the flashers, motor and loop magnets.
(J124 was plugged in as I saw some flashers and the mini powerfield magnets worked.)

It happened during regular multiball ... no balls were on the mini powerfield.
Oh, and I also restored the 2 standup switches that I forced to be stuck closed (this had brought the average voltage down to 11.5ish like all the other white wire rows.) So the white/grey was back up to 13.7v and would drop to 11.5ish with one switch closure.
No fun!

@zaza I will try more to inspect the 12vUnreg shorting to the white/grey wire...it makes sense.