Stuck switch problem with my Xenon pinball machine

[2PacMan]

I've had my Xenon pinball on craigslist, someone comes over to look at it and buy it.  So i moved it into the garage.  It worked fine before.  Now all of a sudden, it's showing a stuck switch.  Well, this scares the guy off and now my sale is blown.  If I can get it working again, he says he'll buy.  The switch is clearly not stuck, and it doesn't happen all the time.  You'll be playing the game, and right in the middle, the switch will just register as being closed and start accumulating points.  After a while, it'll go off again.  In test mode, the screen will flash 'no stuck switches' and then every 10 seconds or so, it'll flash switch #19 is stuck, then go back to no stuck switches, then #19 again.  I checked marvin's guide, and it says there could be numerous problems causing this.  I replaced the diode, but that didn't help any.  It also says to clip the ceramic cap on the switch and see if that stops the shorted switch.  I haven't tried that yet.  It also says it could be an MPU board problem.

Does anyone know what would cause it to act flakey and register closed sometimes and not other times?  Most of the problems on Marvin's site refer to a switch that is always stuck on, and the problem I have is that it registers as being stuck intermittantly.  Just trying to narrow the problem if i can.

[ChadTower]

It is almost certainly a connector issue.  Trace that circuit back to the connector and replace the pins. 

[2PacMan]

How do you replace connector pins?  I've never had to do that before.  Is it the pins that are sticking out of the circuit board, or the the metal pieces that are at the end of the wire harness?

[ChadTower]

Both.  Replace them in pairs.  It is explained in great detail in the marvin3m guides.

[2PacMan]

Ah crap, that looks like alot of work.  I have none of those tool or supplies.  Is there some kind of cheat that can be done?  Is there a way i can solder the wire directly to the pins to get a better connection?  Or is there a way i can tell for sure it is the connector?

[shardian]

Yes you can direclty solder, but that is considered a hack - not a good idea to start accumulating hacks inside the expansive underbelly of a pin.

It definitely sounds like intermittent shorting of wires. I would think a switch matrix failure would wreak a lot more havok than a single intermittent stuck switch.

Adjust the switch, maybe the contacts are too close and touch a bit after being actuated.

[ChadTower]

Ah crap, that looks like alot of work.  I have none of those tool or supplies.  Is there some kind of cheat that can be done?  Is there a way i can solder the wire directly to the pins to get a better connection?  Or is there a way i can tell for sure it is the connector?

If you can't do this, you're never going to have decently working pins, unless you have a good local service tech that you're willing to pay loads of money on a regular basis.  Pinball machines are high maintenance unless you get a fully shopped one... a game with connector issues is not fully shopped.

Basically, unless you're made of cash and can pay someone to solve all these issues, this is a skill you need to have to own pinball machines.  Repinning a connector isn't all that much work compared to the amount of troubleshooting you'll do if you won't repin.

Don't solder directly - as sharidan says, that's a crappy hack job that would make any competent pin repair person seriously doubt the quality of your games if they saw it.

[2PacMan]

I know this may sound bad, but i'm selling the game, so at this point, a hack would be fine with me, as the game is on it's way out of my hands.

Am I barking up the wrong tree thinking it has anything to do with the ceramic capacitor on the switch?  Marvin's site says:

The most common stuck switch problem relates to the capacitor on playfield switches. Note all playfield switches have a capacitor, but most do. Often these capacitors short, leak, or just break off. With the game on and in switch test, cut one end of the capacitor and see if the number goes away on the display in switch test. If so, the capacitor will need to be replaced. If the number doesn't go away, then reconnect the capacitor, as this wasn't the problem (check the switch diode next, in the same manner).

But would a bad cap register an "always stuck" short and not be intermittant, which is my problem.

[ChadTower]

Not 100% sure, but intuition tells me it wouldn't cause either, it would just stop smoothing out voltage spikes back to the cpu board.

Intermittent is very very likely an actual intermittent connection, probably a short to ground, which is what is closing the switch.  I wouldn't be surprised at all to see that the switch itself is misaligned and the leafs aren't far enough apart.

[2PacMan]

I understand.  I guess i just don't have that much motivation to do a great job on this since i'm not even keeping the machine and the price i'm selling it to the guy for doesn't seem like it's worth me investing a lot of time and energy in this.

I wish i could narrow it down to be definately the edge connector.  If i wiggle the connector around while the switch test is going, and i get it to turn off, is that a way to tell it's got a bad connection?

[2PacMan]

I'm nearly 100% positive it's not the leafs.  I had the playfield popped up, nothing actuating the switch at all, and the self test flashed no closed switches, then all of a sudden, #19 is closed.  And I didn't do anything to the switch.  I bent the leafs out and unscrewed the switches from the board and everything and it was still saying it was closed when i could clearly see the switch was open.

Maybe it was just my imagination, but i could have sworn when i wiggled the cap/diode/wires around the ends the switch i could get it to stop saying closed switch for a while.  But i reflowed the solder around the area with no help.  Maybe it was just coincidence though.

[ChadTower]

re: wiggling connectors

Sure, but if your connectors are sketchy, you'll just cause another problem.

Most issues on pinball machines have their root in degraded connectors.  Refurb the connectors well and the game becomes reliable again once the boards are repaired.  Leave degraded connectors in place and problems will just keep occurring all over the game.

If you don't want to fix it then adjust the price and give the guy your opinion on it.  Personally, Xenon is high on my wantlist, as there is a likely better than NOS repro playfield coming out for it in 2008 and I have a thing for blue translucent themed playfields.

re:  wiggling switch assembly

Two thoughts... you still have a cold solder joint in the assembly... but far more likely, you put a little tug on the signal line back to the CPU board connector. 

EDIT:  fixing because you posted while I was posting.

[2PacMan]

Yeah, cold solder is possible too.  The thing is, it worked before i moved it in the other room, so i'm guessing it's a vibration issue or a component failed (diode/cap).  I would think a cold solder joint on the leaf wouldn't manifest itself like that, just by moving the machine.

Is there a way to bend the pins in the connector to get a better grip on the pins?  Are connector problems caused by the loosening connection, or because of corroded metal?

I'm just wondering, about that cap...marvin says if you cut the cap and the stuck switch goes away, then that's the problem.  Well why wouldn't you just leave it cut then?  Why even replace it?  If the cap fails, is it just all or nothing?  Like it either works or it doesn't?  Would a failing cap cause intermittant problems.

[ChadTower]

Yeah, cold solder is possible too.  The thing is, it worked before i moved it in the other room, so i'm guessing it's a vibration issue or a component failed (diode/cap).  I would think a cold solder joint on the leaf wouldn't manifest itself like that, just by moving the machine.

Failed diodes and caps do not cause intermittent issues.  Those would be constant.

Is there a way to bend the pins in the connector to get a better grip on the pins?  Are connector problems caused by the loosening connection, or because of corroded metal?

Not a realistic solution given the problem... the pin is too short and the solder joint on the board too fragile.  You'd just do more damage.  Connector problems are caused by both vibrations and corrosion, basically because they are designed to last 4 or so years and your machine is almost 30 years old.

[2PacMan]

yeah, it is sounding more like a connection issue now.  Do you know a place that sells premade connectors?  With the wire coming out already?  So all one would have to do is cut the old connector off strip the wires and connect them together?  That seems like an easy solution.

[ChadTower]

It does, but I don't know of anyone who sells those.  It's not that hard to repin a connector and it doesn't cost much in tools/parts.

[2PacMan]

If the connector isn't making good contact, wouldn't it show the switch as not working when the connector isn't making the contact?  If the connection is bad, why would the switch be closed, and why only the one switch?  Wouldn't it be all the switches that the connector pin corresponded with?

[ChadTower]

Depends on exactly how a Xenon switch matrix runs... If you read up on how a switch sends its signals back to the CPU you'll be able to figure that out.  It's all in the guides.

[2PacMan]

yeah, i read the manual, but it was all over my head.  So in theory, if i put the game in test mode, and the completely pulled the connector, that should register as a whole bunch of switches being closed if that is indeed how Xenon reads the switch matrix, right?

[ChadTower]

Are the switches NC or NO?  I'm guessing from your earlier statements they're NO.

EDIT:  A good tutorial on pinball switches in wiring diagrams.

[2PacMan]

they are normally opened i would assume.  The other switches register as a stuck switch when the leafs make contact.

[D_Zoot]

Not 100% sure, but intuition tells me it wouldn't cause either, it would just stop smoothing out voltage spikes back to the cpu board.

While connection issues are common with this vintage machine, the original problem isn't an open switch, but rather a stuck closed switch.  Poor connections would cause an open,  so would poor solder joints.  Those two causes can be ruled out.

"Wire shorted to ground"  You can rule  this out too.  The switch system in this machine uses a matrix.  Short any of the switch circuits to ground and you effect an entire row or column, not a single switch.

Voltage "smoothing" isn't the purpose of switch caps at all.     YES, a shorted cap CAN cause this symptom.  Actually it's internal leakage in the cap, and it can manifest as an intermittant.    Clip one end and see if the problem goes away.

If the cap is the problem, you will need to replace it.  It lengthens the switch "closure" time to ensure the CPU sees it.  Without the cap, the CPU probably won't see the switch close every time.


D

[ChadTower]

Internal leakage in a capacitor?  Where would it be leaking to?  A cap only has two leads - in and out.

Thinking about it, I can see how it would be used to sustain the signal a bit longer than the switch press, in some situations... given the speed of electricity, though, I'm not sure I see why it would be necessary other than bad timing design.

[D_Zoot]

Internal leakage in a capacitor?  Where would it be leaking to?  A cap only has two leads - in and out.

Thinking about it, I can see how it would be used to sustain the signal a bit longer than the switch press, in some situations... given the speed of electricity, though, I'm not sure I see why it would be necessary other than bad timing design.

Yes, leaky caps.    Technically, a "leaky" cap is a cap that "leaks" electrons across the dielectric, or from one side to another.    A cap can have leakage and not appear as a short with a DVOM.   If the leakage across the cap is large enough, it could be read as a closed switch to the CPU.  The switch matrix is a very low current circuit, so even small amounts of current leakage within a cap can effect the circuit.


In regards to switch closure, it has nothing to do with the "speed of electricity" or bad timing design,  but rather the speed of the CPU board and how frequently it can poll the switch matrix while doing the other things it needs to do.  Limitations in how fast the CPU board can run is primarily determined by the CPU chip itself, and CPU's from that era weren't that fast.   In a pinball machine switch closures can happen very quickly and can be very very brief.  If the closure occurs between the slower stobes of these earlier electronics, the CPU won't see it.  The cap was added to electrically increase the time the circuit was "closed", even after the switch itself mechanically went open.  Later machines with faster CPU's polled or strobed the matrix fast enough that this isn't an issue (faster than the physics of the mechanical switch)

D

[ChadTower]

Hrm.  IIUC, you're saying that the extended closure time is necessary because of the slowass matrix polling of 1980... which is a result of 1980 CPU capabilities (and I would bet they were using older CPU than that to save cash).

I wouldn't have used leaky as the word but a Google is showing it to be common usage here... good to know.  There's always more to learn.     Thanks for filling in those blanks.


EDIT:  hrm.  Now that I delve into some of these articles, nearly all of them are describing leaky capacitors as physically leaky, not electrically leaky.  They're actually dropping fluid.  Still, your info was great, and totally makes sense.     I'm not trying to be a pain in the ass about that term... I want to understand what each term means to each component as it affects how to troubleshoot circuits.

[D_Zoot]

EDIT:  hrm.  Now that I delve into some of these articles, nearly all of them are describing leaky capacitors as physically leaky, not electrically leaky.  They're actually dropping fluid.  Still, your info was great, and totally makes sense.     I'm not trying to be a pain in the ass about that term... I want to understand what each term means to each component as it affects how to troubleshoot circuits.

It's one of those terms that gets used a couple of different ways depending on who's using it.   For example, TV guys will use it to refer to physical leakage, since that's what they see, and electronics guys will use it to refer to electrical leakage, since that's what they deal with.

[ChadTower]

Right.  And when I think of leakage I tend to think of voltage going where it is not supposed to go rather than going there a little faster than it should, like a transistor with a fatigued PN junction sending out a small signal when it shouldn't be.

[2PacMan]

Wow, i come back from a meeting and see these new developments.  That would be great if it were just the capacitor, as that will be easy for me to replace.  I'll definitely try cutting it when i get home.

Seems like you could go without it, couldn't you, the switch just wouldn't be as sensitive?

So have we definitely ruled out the connector/mpu board?

Edit: looks like this guy had the same problem I did:

http://www.tomshardware.com/forum/58567-13-help-wanted-problem-bally-embryon

I just hope this will fix it.  Can these be bought at radio shack?

[ChadTower]

We haven't definitely ruled out anything.  We are not in ur b@s3 k1ll1n ur d00dz.  We can't see your machine.

If you read his explanation, you cannot go without the cap and still have the switch be read properly by the CPU.  It will miss many of the signals and act almost randomly.

[2PacMan]

You're correct, maybe i spoke too soon.  I will replace the cap...i don't have any though, and i probably won't be able to get one until Friday.  I was just hoping if I cut the capacitor, at least i'd be able to tell if that was the issue or not.  I don't plan on running the game without it.

[D_Zoot]

You're correct, maybe i spoke too soon.  I will replace the cap...i don't have any though, and i probably won't be able to get one until Friday.  I was just hoping if I cut the capacitor, at least i'd be able to tell if that was the issue or not.  I don't plan on running the game without it.

Yes, cut one lead off the cap to find out if that's the problem.  The switch may or may not register during gameplay without it,  but if your "Stuck On" switch problem goes away you'll know the problem is in fact the cap. 

Post back with the result.. 




D

[ChadTower]

Won't it get "stuck on" randomly after X amount of presses?  It will eventually miss an "off" and get "stuck".  Logically this isn't going to do much more than make it start missing "ons" as well as "offs".

[D_Zoot]

Won't it get "stuck on" randomly after X amount of presses?  It will eventually miss an "off" and get "stuck".  Logically this isn't going to do much more than make it start missing "ons" as well as "offs".

Nope.      Without the cap the game will probably miss "on" if it's a very very quick switch closure.  But for test purposes we don't care about that. 

The original problem is that the switch is reading "stuck on",  which translates to closed circuit,   The contacts have been verified to be open (not bent shut),  so something else is completing the circuit.  If the cap is shorted, or leaky, and we clip the lead to it, and the closed circuit (stuck closed switch) now becomes an open circuit, then the cap is what's completing the circuit.

Remember, the cap is wired parallel to the switch,  the matrix is DC volage, caps don't "pass" DC, so a properly working cap doesn't "complete" the circuit. 

Since all the cap is supposed to do is stretch the appearance of a closed switch, we can test the game without it just fine.  Slower switch closures (slower moving ball, pushing the switch with your finger) will still read since the switch contacts will be closed long enough for the cpu strobe (or poll) to come back around.   And open circuit will always be read as "off"


I hope I'm making sense, sometimes I don't know if what I type makes sense to anybody but me......       

D

[ChadTower]

Remember, the cap is wired parallel to the switch,  the matrix is DC volage, caps don't "pass" DC, so a properly working cap doesn't "complete" the circuit. 

I'm going to have to parse that out a bit, I think... that's a new one on me.  Pretty basic, but I've missed things like that before in my piecemeal learning.

[Q*Bert_OP]

You *could  try* unplugging the connectors on the MPU board that go to the playfield switches...I think you may have a board problem...

[D_Zoot]

You *could  try* unplugging the connectors on the MPU board that go to the playfield switches...I think you may have a board problem...

Board end switch matrix problems generally are manifested as an entire row or column being effected, not one single switch.

Say for example you had a bad input logic IC,  since each matrix input to the cpu board serves multiple switches, a bad logic IC, (or even connection) at the board, will effect each of the switches on the problem row or column, so there would be multiple switch errors.   Remember, in a matrix arrangement, EVERY wire (and the logic it is connected to) serves multiple switches.


D

[2PacMan]

Well D_Zoot, you were right so far.  I clipped the end of the cap and now it is working normally again, except the switch is not as sensitive, just as you said.  Now I am heading to radio shack this afternoon to get a new cap to solder on.  That will be the real test, i'm hoping after i hook it up, the game will play normal again.  I figure, worse case scenario, i leave the cap off, and the switch just isn't as sensitive, it still triggers pretty often though.  Thank you so much for your help so far.  I'll post back when i get the new cap on.

[2PacMan]

Well, got the new cap at Radio Shack and BINGO, it works great now.  Thanks for everyone's help and suggestions in this thread.  Unfortunately, i think my buyer is scared off now, as he doesn't seem too interested in my game anymore after i told him it was fixed....oh well, should be easier to sell 100% working again to someone else.

Hopefully someone else who has a similar problem can look at this thread and solve their problem too. 

Thanks again guys, i really learned alot talking this over with everyone.

[ChadTower]

Awesome.  I love being wrong on this stuff when there is a chance to learn something new from it. 

Post a link to your CL listing... I'd love to see it.  I'm halfheartedly looking for a Xenon, though yours is way too far for me.

[2PacMan]

Sure, here you go:

http://omaha.craigslist.org/for/490008488.html

The pictures aren't great, it's a really decent looking machine, nearly no playfield wear.  Most of the Xenons i've seen have considerable wear.