Taxi Primary Display - several segments always lit

Ok hopefully some of the guru's around here can point me in the right direction!

Here's the deal - I got a killer Taxi, she's in great shape and I'm finishing the resto on her right now. I've got everything in 100% shape, except for this:
2014-04-28 21.08.58.jpg

Here's what happened... Display was working great, and then I hit up a HS of 10.6 million. While the machine was whooping and ringing and flashing (massive amount of voltage going on!) these segments suddenly lit up and have not gone away since. It appears that ones on the right are always on no matter what. The ones on the upper left (hard to see in the picture because of the score display) only seem to show up when the Jackpot is flashing - and then they "flash" along with the secondary display - which also displays a weird lit segment in the same place. Don't worry about the bottom row, it's fine, those are just some segments starting to light when I took that pic. I'm hoping that gives you guys a clue as to which chip may have burnt. The displays themselves are fine, dark black with no gassing - and I'm really hoping that I can find/isolate the correct part and solder in a new one.

Any ideas, O Great Collective??

#2 9 years ago

JAY

dewdude82

Inactive member

MonkeyGrass: I'm sure you've seen (or will see) my response in the thread about Taxi High scores.

For the sake of prosperity for anyone who may find this thread in the future; I'm going to post my response here as well.

The stuck segment in the pictures is "p".

Segment P comes in on the ribbon on J3 and first travels through U9, input on 9 output on 10. From there it goes to U15, a UDN7180; pin 4 input, pin 15 output. It then goes from pin 15 through R40 to pin 82 on the display.

I've seen the 4050 chips fail beore, causing this same problem. I would start with U9, look at the output to see if it in fact changes any or is locked on (logic probe or an oscilloscope if you've got one). If you can't disagnose it, then you could try to shotgun replacing U9 and U15. The problem is U15 has become very difficult to find.

#3 9 years ago

NAD

MonkeyGrass

Inactive member

Yes, yes I did.

Thanks for putting this on here, I hate cross-threading stuff - which is why I created a topic. Let's continue on this one so we don't create a bunch of confusion, cool?

Ok I got a source on a 4050 and UDN7180. I'm going to try to test those now and confirm, and will order them up if I can determine that's it. Even if I can't narrow it down 100% to those two chips, (I don't have an oscilliscope only a DMM) I'll probably order them anyways and, as you say, "shotgun replace" them. $25 is a small price to pay to get my pristine OEM display back!

Seriously man, thank you for all your help. You've given me more legit info in 15 minutes than 2 days of fumbling around schematics and searching the internets high and low!!

#4 9 years ago

CAH

hassellcastle

FUN! Superstore

Mesquite, TX

12y 82,550 2,075 8

I'm going to post an obvious soltution I hope you have tried.
Did you make sure the ribbon cable to the cpu is secured?
I say this because my Taxi will throw up some random segments and I know if I wiggle the ribbon cable they go away. I need to order a new ribbon cable but keep forgetting everytime I place an order for parts.

#5 9 years ago

JAY

dewdude82

Inactive member

Quoted from hassellcastle:
Did you make sure the ribbon cable to the cpu is secured?

In the previous thread we had been conversing in; I mentioned checking the ribbon first as dirt data will cause the same thing. I just forgot to add it to this one.

I am assuming you tried reseating the cable MonkeyGrass. If that doesn't work, try actually flipping the cable over on both ends. In some WPC machines I'd get driver board issues caused by a flaky ribbon cable that would often be solved by flipping the cable.

#6 9 years ago

JAY

dewdude82

Inactive member

OP
Quoted from MonkeyGrass:
Even if I can't narrow it down 100% to those two chips, (I don't have an oscilliscope only a DMM) I'll probably order them anyways and, as you say, "shotgun replace" them.

Not having an oscilloscope I've found does hinder some troubleshooting, especially when you're trying to look at logic levels from chip outputs. You could *also* do it with a DMM by looking at the voltage levels. Generally it's pretty easy to identify if something is stuck high. You have several outputs that seem to be working you can compare to. But you're only going to get a voltage reading and not be able to see how the data is pulsing, so it's very limited.

I will say with some confidence that the P segment data only travels through those two chips and resistor before going to the display. If it's not either one of those or a ribbon; then you'd have a CPU problem...but I'm pretty sure it's a relatively "simple" fix. I've never seen glass fail in that fashion, ever...so it's a logic problem.

You said you had a Police Force; if you swap displays and the Police Force display is fine in Taxi; then you know your problem is going to be one of those two chips (or the ribbon).

_{Post edited by dewdude82: it's too early....I misread about having a DMM.}

#7 9 years ago

NAD

MonkeyGrass

Inactive member

Yes, I've already swapped the ribbon cable (back and forth from both machines) - it works fine in PF and the one from PF does the same thing in Taxi. CPU is testing out fine, everything else looks good. Judging from the way it "went" I think it's a fairly safe bet that one or both of those chips blew out. I'm going to swap the entire display board from PF over to Taxi, just to verify that it's not a CPU issue. Once I do that, I'm pretty much down to those two chips, right?

#8 9 years ago

JAY

dewdude82

Inactive member

OP
Quoted from MonkeyGrass:
I'm going to swap the entire display board from PF over to Taxi, just to verify that it's not a CPU issue. Once I do that, I'm pretty much down to those two chips, right?

Yeah. This was what I was getting at the end of my last post; swap the entire display board around to find out if the CPU is sending bad data or not. I don't think it's the CPU and leaning more toward those two chips; but you can't say for sure till you swap display boards.

If the problem stays with the displayboard, it's a really safe bet it's one of those two chips (or both).

#9 9 years ago

NAD

MonkeyGrass

Inactive member

Ok this is interesting... Just took the whole board out of Taxi and put it in PF.

It worked perfectly - no lit segments, just like it should. I'm going to pop the PF board in Taxi just to double check, but it looks like it may actually be a logic/control problem and not a burnt 7180.

I'll post back in a few with the results.

#10 9 years ago

NAD

MonkeyGrass

Inactive member

WHY did Williams insist on using these backless, push-in style receivers on half their stuff? Grrrrr...

PF's display is secured with 9mm nuts and black standoffs on the front of the backboard, with a slotless bolt on the back of it. Two turns on the nut, and the back starts to spin freely with nothing to grab or push on to loosen the nut!

I'm just going to wait until I hear back from you dewdude, instead of ripping apart a perfectly functional, original display board unless I have to. I'm not afraid of putting in the work, I'd just rather not risk something old and original breaking while moving stuff around testing it, if we already have the info we need. If it was a chip on the display board - no way it works 100% on PF. Right?

Do you think it's necessary to put the PF display in Taxi or should I just move on to troubleshooting the logic end of things?

#11 9 years ago

JAY

dewdude82

Inactive member

I would put the PF display in Taxi just to see what happens? If the Taxi display worked fine in PF, then I'd lean more toward some other logic issue on the board; which could be traced out by maybe working backwards on the CPU from the display connector.

OP
Quoted from MonkeyGrass:
If it was a chip on the display board - no way it works 100% on PF. Right?

Correct. If the Taxi display worked just fine in PF...and you already established it's not a ribbon...the problem isn't on the display board. If those chips were blown, you'd see the SAME problems on the display regardless of machine.

I'll take a look in a little bit at the 11B schematic and see if there's anything obvious. But I'd still take the effort to carefully put the PF display in Taxi (or just disassemble it, lay it on some cardboard and test it; no need to constantly swap it in to the insert just to test..just don't let anything short out or touch anything).

#12 9 years ago

JAY

dewdude82

Inactive member

I just had a flashback from 7 years ago and I forgot one other obvious thing. Check the solder joints on the ribbon connector on the display board. It's possible one is cracked and the process of swapping it caused it to make contact again.

#13 9 years ago

NAD

MonkeyGrass

Inactive member

OK here are some more pics of what's happening. The segments on the right hand upper side (player 2) are on constantly. The segments on the left (player 1) flash in accordance with the jackpot display. If that secondary display isn't flashing, like while in meter mode, the segments are not lit.

Here you can see - the secondary display is scrolling (not flashing), and the P1 segments are dark.
2014-04-29 13.35.19.jpg

Here it is again, this time flashing the jackpot value - and the segments are lit. You can see they are just coming on, they flash with secondary display.
2014-04-29 13.35.21.jpg

This is part of the same sequence - here the secondary display is totally dark, as are the P1 segments. This is between flashes.
2014-04-29 13.35.16.jpg

A split second later, the jackpot amount lights up again and so do the P1 segments.
2014-04-29 13.35.08.jpg

Thoughts, gentlemen? I'm all ears!! Maybe this is tied into one of the secondary display power circuits?

#14 9 years ago

NAD

MonkeyGrass

Inactive member

Possible culprit? I just went thru and pulled, checked, and reseated every ribbon and molex connector between the display, PSU and secondary power. It all looked pristine, as it should since it just got a professional shop job a few months ago.

In pulling the display ribbon from the control board, I noticed this just above the batteries. Now, this whole area was worked on and the battery holder replaced and some work done to repair minor corrosion. This is directly above the batteries, where the ribbon connects. Look under SCR4.
2014-04-29 13.58.57.jpg

Thoughts? If I'm not totally mistaken, that sure looks like 9 on the ribbon cable, which would be the segment in question.

#15 9 years ago

THC

yotaxi125

Pinside member

Los Angeles, CA

10y 26,150 135 5

Hey guys,

I don't mean to highjack the thread, but I have a similar problem on a EATPM. Except it's the opposite in that the upper left "hash" on the left screen will not illuminate.

For example, a "9" looks like a backwards "F" and an "8" looks like a backwards "6".

Quoted from dewdude82:
You said you had a Police Force; if you swap displays and the Police Force display is fine in Taxi; then you know your problem is going to be one of those two chips (or the ribbon).

I also have Taxi and Police Force and I actually did swap the displays (sorry Police Force, Taxi gets priority haha) so this test will definitely work for you. Yeah just go ahead and do it. They aren't as delicate as you think, but obviously use care when swapping!

#16 9 years ago

JAY

dewdude82

Inactive member

OP
Quoted from MonkeyGrass:
The segments on the right hand upper side (player 2) are on constantly. The segments on the left (player 1) flash in accordance with the jackpot display. If that secondary display isn't flashing, like while in meter mode, the segments are not lit.

This is where we start getting in to some real issues. The glass for player 1 and 2 is just one long piece of glass; they're not actually seperated at any point. The segments for all the displays come in and the digits are strobed. But with the jackpot display behavior...you have have a bizarre strobe issue.

OP
Quoted from MonkeyGrass:
Look under SCR4.

That resistor pack looks pretty gnarly. Someone did a lot of board work to that area. Did they replace it? Without being able to inspect it in person with a maginfying glass, I don't know. But they are pull ups for the data for the displays. If that wasn't cleaned up and installed properly, it could cause some issues. The segment data for the player 1/2 (display 1) run through SCR5 and SCR2, pull up resistors/cap packs pulled from the 6821 at U41.

But I'm starting to wonder if you've got some kind of strobe issue...which I'm not sure offhand where the digit strobe data comes from...or why all the *p-segments* would be locked on. If the driver was shorted; they'd always be on. I really need to see what the jackpot display and display 1 have in common. The player 2 area is always locked on but player 1 only locks when jackpot is displaying. Feels like a strobe issue, but I've really gotta look and figure out what they have in common. It's been a while since I've actually looked at these schematics; I'll study them for a while and see if I can't figure out some ways of narrowing this down. It's possible that's a red-herring and one of the chips is just "failing" and the excess load isn't affecting it. But the jackpot display isn't alpha.....my head wants to explode. A HV supply that's gone too high can cause this too; as the segment driver chips have 90V going in to them.

The thing to remember is that *every* segment in a display is connected together. If you light up one segment the power flows through ALL of them. What it does is very rapidly strobe digit locations; constnatly changing the segment strobe data to match what segment it's on. So the fact you have P lit on one side constnatly and sometimes on the other really starts to throw a loop...which is why I said I need to really study segments.

I just have to sit and figure this out without being able to throw a logic probe/DMM/scope on things.

Quoted from yotaxi125:
For example, a "9" looks like a backwards "F" and an "8" looks like a backwards "6".

If you swapped displays and the problem stayed with the display; then you probably have a segment driver not working. The segment that would do what you're describing is "f". It it's the top display; the F segment comes in on pin 5 of U10, out of pin 4, then goes to pin 6 of U16 and out of pin 13 through a resistor to the display. If you know it's not solved by plugging in to another machine or wiggling the data cable...then U10 and U16 would be suspect.

_{Post edited by dewdude82: I'm getting p-segments and commas confused again.}

#17 9 years ago

NAD

MonkeyGrass

Inactive member

dewdude, now you see why I'm starting to lose my mind over here!!! :/

I just pulled the receipts. Brady Distributing went thru this thing top to bottom, re-worked the battery backup and surrounding board areas, and put in a new power control board. Receipt is dated 4/15/14, and I think I'm gonna give them a call and see what they say. There has been some major work done to that board and I don't have a breakdown of exactly what was done, making it double hard to troubleshoot. If it's still under warranty from the work the previous owner had done, I may have to bite the bullet and take it down to them. If I knew exactly what the problem was and I could replace a couple of components and be done with it, I wouldn't bother but unless you can come up with something that I can test over here, I'm pretty stumped. If you need a copy of the wiring I can send it to you. I have a hard copy (missing the last few pages with the critical schematics I need, of course) and I pulled a pdf down from ipdb which has all the pages. Just hard to read those split pages on a 14" laptop monitor!

Thanks again man, can't tell you how much I appreciate you putting some thought into it!!

#18 9 years ago

NAD

MonkeyGrass

Inactive member

Just got off the phone with the technician who worked on the machine last month. He essentially confirmed everything that I thought... battery corrosion in the board traces leading up to the display output. He said he did a bunch of work on it and got it the best he could, but that he couldn't get it 100% and that's as good as he could do without a full board replacement.

So... yeah. Looks like I'm in the market for a new Sys11b board. Unless there is a way to fix it? I don't mind spending time and working on delicate stuff if there's a way to rework it.

#19 9 years ago

JAY

dewdude82

Inactive member

OP
Quoted from MonkeyGrass:
He essentially confirmed everything that I thought... battery corrosion in the board traces leading up to the display output. He said he did a bunch of work on it and got it the best he could, but that he couldn't get it 100% and that's as good as he could do without a full board replacement.

Sounds like it was a good tech for admitting that. Battery corrosion in that area is something I've dealt with for years, and it generally never ends pretty. I wasn't sure who had replaced that, but a tech for Brady is one I'd trust; so the damage must be severe.

OP
Quoted from MonkeyGrass:
Looks like I'm in the market for a new Sys11b board. Unless there is a way to fix it? I don't mind spending time and working on delicate stuff if there's a way to rework it.

Anything is possible. I have had to rebuild switch matrix circuits on WPC boards after years of sitting idle lead to corrosion just eating the traces. It took a lot of time, a lot of looking at the schematic, and a lot of tiny rework wire. It looked like hell, but 5 years later and the machine it was in had a perfectly functioning switch matrix.

You could...in theory...replace all the traces with wirewrap. sixteen pieces of wirewrap between U41 and SR2/5, and sixteen between SR2/5 and the ribbon connector; but that's a lot of work and it'd be really hacky. If you've never done a lot of board work; that much wirewrap gets really difficult to work with. Still, I would often times do that kind of stuff in the field just for the sake of not wanting to take the board back to the customers house 2 hours away.

#20 9 years ago

JAY

dewdude82

Inactive member

OP
Quoted from MonkeyGrass:
dewdude, now you see why I'm starting to lose my mind over here!!! :/

This....almost every day for 10 years. I know it all too well.

#21 9 years ago

NAD

MonkeyGrass

Inactive member

Quoted from dewdude82:
Sounds like it was a good tech for admitting that. Battery corrosion in that area is something I've dealt with for years, and it generally never ends pretty. I wasn't sure who had replaced that, but a tech for Brady is one I'd trust; so the damage must be severe.

Anything is possible. I have had to rebuild switch matrix circuits on WPC boards after years of sitting idle lead to corrosion just eating the traces. It took a lot of time, a lot of looking at the schematic, and a lot of tiny rework wire. It looked like hell, but 5 years later and the machine it was in had a perfectly functioning switch matrix.
You could...in theory...replace all the traces with wirewrap. sixteen pieces of wirewrap between U41 and SR2/5, and sixteen between SR2/5 and the ribbon connector; but that's a lot of work and it'd be really hacky. If you've never done a lot of board work; that much wirewrap gets really difficult to work with. Still, I would often times do that kind of stuff in the field just for the sake of not wanting to take the board back to the customers house 2 hours away.

Yeah the rest of the machine is in spotless shape. He's a trusted tech and I take him at his word. He did a fantastic job overall. It's definitely those SR2/5 connections. I slightly wiggled them to see what was up and when I rebooted, the left side is not flashing, and two of the segments are constantly lit (instead of four or five which flashed with the jackpot display). I may try to replace those traces at some point down the line. Theoretically, if I replaced just the 2-3 traces that are fried with wirewrap, that *should* do it without re-wiring the whole thing, right? The game plays perfectly and I just spent a bunch on new ramps/plastic/rubbers so I'm probably just going to deal with it for the time being. I'll be placing a rottendog board on my list. Heck, a replacement display was going to be $200+, $350 for a whole new board on a keeper pin isn't really that horrible in the grand scheme of things, right? (I just keep telling myself this, over and over...)

#22 9 years ago

JWJ

johnwartjr

That Pinball Place

Columbus, IN

12y 77,400 6,479 14

That rework does look pretty good considering what the tech had to work with. I've worked on probably 10 corroded Sys11 MPUs in the last 6 months, and some of them get just downright ugly.

#23 9 years ago

AAA

pintechev

Inactive member

11y 91,300 4,437

You could send the board out for repair.

#24 9 years ago

NAD

MonkeyGrass

Inactive member

Quoted from johnwartjr:
That rework does look pretty good considering what the tech had to work with. I've worked on probably 10 corroded Sys11 MPUs in the last 6 months, and some of them get just downright ugly.

Yeah, the game is 100% functional with no glitches or errors, except those segments that want to stay lit. Not bad, and I know the previous owner had over $700 in it with Brady last month. The tech gave it his all, he knows what he's doing, and if I have to live with a couple lit segments that don't effect the gameplay until I can save up for a new board, so be it. To me, this table is well worth it. Even factoring in a new board, between the ramps, plastics and rubbers, I'll still have less than $1500 invested in it. I have no problems ending up with a gorgeous, completely shopped Taxi with a new MPU for that price.

Quoted from pintechev:
You could send the board out for repair.

Just got back from the shop - this board been given all the life it's gonna get. With another Sys11 in the house (Police Force) I'm going to be better off in the long run, picking up a new rottendog and keeping this one as a working backup with a wonky display thing that I can always fall back on if something happens to either one down the road.

#25 9 years ago

AAA

pintechev

Inactive member

11y 91,300 4,437

OP
Quoted from MonkeyGrass:
Yeah, the game is 100% functional with no glitches or errors, except those segments that want to stay lit. Not bad, and I know the previous owner had over $700 in it with Brady last month. The tech gave it his all, he knows what he's doing, and if I have to live with a couple lit segments that don't effect the gameplay until I can save up for a new board, so be it. To me, this table is well worth it. Even factoring in a new board, between the ramps, plastics and rubbers, I'll still have less than $1500 invested in it. I have no problems ending up with a gorgeous, completely shopped Taxi with a new MPU for that price.

Just got back from the shop - this board been given all the life it's gonna get. With another Sys11 in the house (Police Force) I'm going to be better off in the long run, picking up a new rottendog and keeping this one as a working backup with a wonky display thing that I can always fall back on if something happens to either one down the road.

My Taxi has a Rottendog in it and it's working great.

#26 9 years ago

NAD

MonkeyGrass

Inactive member

Quoted from pintechev:
My Taxi has a Rottendog in it and it's working great.

Good to know, schonb25! It'll likely be a few months, but I think a rottendog is definitely in my future. I'm planning on keeping this pin for a (life) long time, so putting some money into it here and there doesn't freak me out, I'm not trying to flip it and long term, this is the better fix. It's worth it, IMO.

#27 9 years ago

NAD

MonkeyGrass

Inactive member

Weird. I started it up last night, and no lit segments. After a couple games, the slowly started to fade back on. For the rest of the night, they randomly would pop back on for a game or two, then go back off.

Pretty sure something the tech re-traced on the board got hot enough to slightly loosen/soften the solder and create an intermittent short - and I think is has something to do with that strobe circuit dewdude was referring to. Hopefully, I can get that board out and clean up the excess and get this fixed for the time being!

1 month later

#28 9 years ago

???

onsights

Inactive member

Sf, CA

9y 3,250 1

Hi team-

I read with great interest this thread, as I have an Earthshaker that has the same two segments "on" in the the entire bottom row stuck in the lit position. This machine has been is storage a long time. When I opened her up I discover a corroded battery connector which I cleaned up enough to get it working. The stuck segments seems to be the only issue, but it sounds like I'm gonna have to pull the main board and replace the battery holder so I can clean and "repair" the traces, unless any of you have alternative suggestions. I flipped the data cable and jiggled connectors as suggested.

This is what it looks like.

Thank to you all for the participation in the forum

Michael

gogirl display.jpg

1 week later

#29 9 years ago

NAD

MonkeyGrass

Inactive member

http://www.maaca.org/viewtopic.php?f=22&t=16154

Yeah, seems like you are having the same issue. Battery corrosion right where the display traces lead to the ribbon cable, underneath the and around the battery holder.

Good luck. I haven't pulled the board yet, my segments fade in and out randomly and I'm not going to risk messing something else up attempting to re-trace something that minor which doesn't really affect gameplay. Eventually, I'll grab a rottendog and be done with it. In the meantime, I just deal with it whenever it starts up.

#30 9 years ago

TOE

Toeckeder

Pinside member

Blainville, QC

10y 25,650 42 16

I had a similar issue, maybe not exactly the same but worth a try:

#31 9 years ago

NAD

MonkeyGrass

Inactive member