Pinball 2000 Tech Problem: Power Driver Board Damaged?

[jasonbar]

Howdy-

I'm having problems with my Pinball 2000 cabinet. More precisely, the power driver board appears to be damaged...and possibly more(?)

1 - Before getting verbose, does anybody offer repair services on these?

2 - Or, are there better prices than $495 + shipping ( http://ksarcade.net/pinball-2000-power-driver-board-04-12329.html ) or $349 + $100 shipping to California ( http://www.pinball.co/Products/25303-04-12329.aspx ) for a replacement?


Having said that, if anybody wants to chime in on reading along & troubleshooting, then I'd be most grateful!

3a - Ever since I got RFM, 3 of the #89 flasher bulbs consistently burn themselves out after only a few flashes (in test mode or while playing): Right Arch (transistor Q67, connector J112-9, Fuse F109), Left Martian (transistor Q68, connector J112-10, Fuse F109), & Right Martian (transistor Q69, connector J112-19, Fuse F109). The other flashers on this same fuse circuit work fine. I replaced these transistors & the bulbs still have almost zero life before blowing. These same 3 transistors drive 3 back panel flashers on SWEP1, & those flasher bulbs on my SWEP1 playfield are just ducky--they don't burn out. This suggests that there's something weird in my RFM playfield & that this problem doesn't lie in the power drive board.

3b - Perhaps a bit of an aside, but what exactly makes a bulb burn out? The filament could blow from too much current/voltage/heat/vibration, I imagine...a quick Googling didn't turn up something conclusive that applies to troubleshooting my problem...A good experiment for me would be to swap some pins on the driver board connectors to drive these problematic sockets/wires with a well-behaved transistor, & vice versa, but I haven't the proper tool to pop pins out of connectors, & I don't want to do any mangling.

4a - I had just finished bracketry for a plunger/auto-launch combo mod & made quite the electrical goof. The way I re-installed the existing RFM launch coil, I wanted the shaft to move in the direction opposite of how it was intended to move. Therefore, I made a short wiring harness adapter to swap the polarity of the wires to the coil, figuring it would run "backwards." (I failed to reverse the diode, though...)

4b - During test of the coil in RFM Diagnostics mode, Auto-Launch wouldn't fire. After trying for a bit, 3 other coils stopped firing also in test mode: Right Jet, Bottom Jet, & Right Lockup, all 3 of which share Fuse F100 w/ Auto Plunge. Before I clued myself in about the diode being reversed, I unhooked the functioning right sling coil's 2 wires & hooked them to my Auto Plunge coil (again, with the diode backwards). After a moment, the 4 coils on fuse F103 stopped working too (Trough Eject, Left Sling, Right Sling, Left Jet Bumper).

5a - Now, after rewiring everything correctly (and putting new fuses in--read more about fuses below), I get the following behavior:

5b - With RFM or SWEP1, the "Health LED2" blinks when the machine is powered on. Is this normal?

5c - With SWEP1, the Read Test Report at the top of the main menu says that there are no errors. However, when I close the coin door interlock high voltage switch, the following 2 items activate & stay stuck activated (but I fear further damage, so I don't let it stay in this state for more than a couple of seconds): Right Sling coil (transistor Q53, connector J112-12, fuse F103), both Upper Hotdog Flashers (transistor Q57, connector J112-17, fuse F109). Despite these 2 items staying stuck energized, if I go through all of the solenoid tests & briefly test each one, they all behave normally (except for the 2 listed above).

5d - With RFM, I still have the same 3 flasher bulbs (see 3a) not working. Upon immediate power-up, with the coin door open & no high voltage, the Diagnostic Report shows that fuses F100 & F103 are blown. These are the 2 fuses associated w/ the 2 dead HV circuits described in 4b. However, the fuses have continuity. After I run the diagnostic coil test for a moment, the fuses do indeed blow. The RFM software is psychic!


Any thoughts on what to try?

Thanks,
-Jason

[lilshawn]

with the coils that are staying on, replace the transistors tip102's or tip107's on the diver board and it would be best to also replace the diodes for those coils at the same time. they are shorted (causing the lock ons AND the fuses blowing. running the coil reversed or having a bad diode on the coil shorts the transistors and shorted transistors often kills the diode. if you don't replace both, the coil will fire a few times and then the transistor burns out because it can't handle the reverse spike caused by the coil de-energizing. and the constant current of a shorted transistor fries a coil and kills the diode. so do both. also check the coil ohms to make sure they aren't shorted. (without them hooked up or a diode on there) they should be a few ohms but not a dead short. compare them to a known working coil of the same # if you are unsure. (Ie 28-1500 to the same 28-1500 ) for example a 28-1500 measures 20 ohms where a 23-800 is 4 ohms and both are good... but a fried 28-1500 may measure 4 ohms. see what i mean?

with your lamps that are burning out, be sure you have the proper bulbs. maybe a #89 or#906 (depending on the base) it's been awhile since i've worked on one of these, we've sold them all off. I'm not 100% on this but maybe there is an issue with the voltage driving the burning out lamps. if it's too high, that could certainly burn out the bulb. i think the flasher bulbs are 13 or 14 volts. they mostly flash them with 20 volts check the voltage. if it's way higher than that like being powered from the 50 volts solenoid power supply, then that would be the time to tell us someone hacked up, or repaired the harness because it's not right. you should be able to test this by grounding the negative lead on the frame and probing one of the wires on the flasher socket. one of them will be the drive voltage. (you may need to enter the test mode or shut the door or something to enable the power to the flashers. (i believe it's powered by a relay)

the main thing with these machines is you have to be SURE you have things fixed 100%, because if you try running something with a failed coil, it's just going to blow on you. then, when you change the coil...and you still have the bad transistor, you fry the coil. everything has to be 100% or things get fried.

4a - I had just finished bracketry for a plunger/auto-launch combo mod & made quite the electrical goof. The way I re-installed the existing RFM launch coil, I wanted the shaft to move in the direction opposite of how it was intended to move. Therefore, I made a short wiring harness adapter to swap the polarity of the wires to the coil, figuring it would run "backwards." (I failed to reverse the diode, though...)

no idea what you're talking about. there is no way this would work in a million years. it's designed in such a way as to only operate in one direction. Pull in. to make it work opposite you'd have to move the coil to the other side of the item you need to move and make a linkage to have the plunger operate in that direction...still pulling in. that's how a coil works. it will not "push out" or any variation thereof.

let's focus on getting this machine back working as it's originally intended and not fart around with hacking up the harnesses or trying to drive things with other lines. It is just going to mess things up worse. specific transistors in a circuit are designed to drive specific items. the coils look the same, but the drive is different, and it's designed for that particular coil setup. You can't switch them around to see if it works, cause it won't. Stuff blows up.

[ChadTower]

Yep... he borked up the driver board transistors.  No two ways about that.  And when he ran the coils without good diodes he might have fried them too.

[jasonbar]

Thanks for the input. I replaced the 2 transistors that drive the SW right sling & back pair of banana flashers. SW is now working 100% again. Hooray. Played it for 5 minutes with no issues. Thanks for pointing out dead transistors. Easy fix.

Next up, checking resistance & diode on RFM Auto Launch coil.

Thanks,
-Jason

[jasonbar]

I'm on a roll!

Progress on RFM:

1 - I put in the RFM PRISM card & ROMs & had no playfield plugged in. Fuses F100 & F103 OK. Normal behavior. Powered down, plugged in RFM PF, unplugged all 8 coils on the 2 circuits that had fuses blowing (F100's Right Jet, Bottom Jet, Right Lockup, & Auto Plunge, and F103's Trough Eject, Left Sling, Right Sling, & Left Jet Bumper), powered up.

2 - One at a time, I plugged a coil in (with high voltage off) & then tested it, skipping Auto Plunge, which was the source of my coil woes. All 7 coils plugged in work fine, no fuse blowing, no drama. Swell. This means the problem is either in the Auto Plunge coil/diode or in its playfield wiring or in its associated circuit on the power board. But, given that it runs off of transistor Q57, & my SWEP1's Upper Hot Dog Flashers are now running fine off of that same Q57 transistor, I figure that the power board is OK. Aside #1--yup, I clearly blew both transistors with which I tried to drive a coil "backwards". Aside #2 on fuses: when I was having fuses blow, I was using fuses of slightly different specs (It's supposed to have 250V 4A slow blow, but on hand at the time I had only 250V 3A slow blow & 250V 4A fast blow. Apparently, these different fuses lead to immediate problems--can't even do a little troubleshooting with them!)

3 - The Auto Plunge coil is an AE1-23-800, of which there is only 1 other on RFM (Right Lockup) & none on SWEP1. Given that I'm pretty good at replacing transistors on the power board by now, and seeing that I'd need to disassemble a bit & then desolder/cut a diode or coil wire to be able to check the coil and/or its diode, I chose instead to hold my breath & I plugged the 2 wires for the Auto Plunge into the already-tested-good right slingshot. Works fine--that means Auto Plunge wiring & corresponding power board circuitry is good. That leaves the Auto Plunge coil/diode as the final culprit. Sure enough, trying to fire that coil (not backwards this time!!!) blew fuse F100 immediately & the coil didn't fire. Culprit found!

4 - I removed the faulty coil, couldn't manage a clean desolder of the magnet wire or diode, so I snipped 1 diode leg & tested the diode--bad diode. I didn't have any other replacement matching diodes. I checked resistance of the coil: about 4.6 ohm. I'd rather not hack up the other matching good coil in the game, so I don't have a comparison value to see if the coil has an issue, but that resistance seems reasonable.

5 - I just ordered a handful of 1N4004 diodes & will replace the Auto Plunge coil's diode & see if that fixes it. I also ordered a new AE1-23-800 coil assembly & will have that as a backup. (I also ordered another coil b/c the bottom jet seems to have been replaced w/ the wrong coil (but of a similar size) in the past.) I also ordered some more fuses. :]  I also ordered 2 more boxes of #89 flasher bulbs in preparation for the upcoming flasher fix...

6 - ...which brings us to the flashers. Yes, lilshawn, they're #89 bulbs in RFM's Right Arch, Left Martian, & Right Martian. BUT!!! Those same transistors & connectors on the power board drive 3 back panel flashers on SWEP1 that have never blown at all. That's the weird part. I'll still try out the voltage test at the bulb socket to see what it's reading. I can also trace these 3 flashers' wires back to the connectors at the back of the playfield to make sure that there's nothing hinky along the way.


Thanks,
-Jason

[ChadTower]

Aside #1--yup, I clearly blew both transistors with which I tried to drive a coil "backwards". Aside #2 on fuses: when I was having fuses blow, I was using fuses of slightly different specs (It's supposed to have 250V 4A slow blow, but on hand at the time I had only 250V 3A slow blow & 250V 4A fast blow. Apparently, these different fuses lead to immediate problems--can't even do a little troubleshooting with them!)

There are two approaches to take here.  Either buy 20 of the fuse in question when troubleshooting or buy a circuit breaker that fits the specs.  The circuit breaker can be very very useful but I tend to only do that with slow blows.  The breakers don't always trip as fast as a fast blow fuse would burn.

Culprit found!

I would swap the coil.  Even if you didn't destroy the coil you may have substantially reduced its lifespan.

4 - I removed the faulty coil, couldn't manage a clean desolder of the magnet wire or diode, so I snipped 1 diode leg & tested the diode--bad diode. I didn't have any other replacement matching diodes. I checked resistance of the coil: about 4.6 ohm. I'd rather not hack up the other matching good coil in the game, so I don't have a comparison value to see if the coil has an issue, but that resistance seems reasonable.

Jump the diode and take the measurement.

6 - ...which brings us to the flashers. Yes, lilshawn, they're #89 bulbs in RFM's Right Arch, Left Martian, & Right Martian. BUT!!! Those same transistors & connectors on the power board drive 3 back panel flashers on SWEP1 that have never blown at all. That's the weird part. I'll still try out the voltage test at the bulb socket to see what it's reading. I can also trace these 3 flashers' wires back to the connectors at the back of the playfield to make sure that there's nothing hinky along the way.

Sounds like you have a wiring break.  Check continuity on the common ground line for all of those lights.  Check to each other in series and then from the last light to the driver board connector. 

[jasonbar]

6 - ...which brings us to the flashers. Yes, lilshawn, they're #89 bulbs in RFM's Right Arch, Left Martian, & Right Martian. BUT!!! Those same transistors & connectors on the power board drive 3 back panel flashers on SWEP1 that have never blown at all. That's the weird part. I'll still try out the voltage test at the bulb socket to see what it's reading. I can also trace these 3 flashers' wires back to the connectors at the back of the playfield to make sure that there's nothing hinky along the way.

Sounds like you have a wiring break.  Check continuity on the common ground line for all of those lights.  Check to each other in series and then from the last light to the driver board connector.

Spent some time on this last night & think I found the culprit. Flashers worked for about a minute w/ no burnouts, which is better than they've been in a long time.

a - Continuity along the red/white common wire to all 7 flashers is good, all the way to the harness connector at the power board.

b - With all 7 flasher bulbs removed, there were no shorts within any of the 7 bayonets. Ergo, wiring appears to be good. I did a quick solder reflow on 6 of the 7 flasher bayonets just in case there was an intermittent cracked solder joint issue. (Right popper flasher socket is not exposed for quick & easy access.)

c - I installed each flasher 1 at a time & tested. All were good. (There are 4 #89 flashers whose bayonets have to be removed & then screwed back into the playfield wood to change the bulb. There are 3 #609 flasher bulbs that can be popped in & out w/out disturbing their socket).

d - I then screwed each of the 4 #89 bayonets back to the pf, testing after each screw-in, figuring that there might be interference or a short to a piece of metal or something, no problems. All's good!


(The previous owner denied any knowledge of these flasher problems, &, indeed, these flashers worked fine for a short while after I got the game.) Here's what I noticed, which I think is the issue. The 2 tabs on the bayonets are free to spin around. If both tabs are lined up with each other & even slightly bent, then they can short. Also, the flasher connector tabs for the right arch are *right* next to the playfield bent steel tubing base & can easily touch it. If my memory serves me correctly, ditto for the left Martian flasher bayonet tabs, which can touch a neighboring piece of metal.

A previous owner had put a bit of heat shrink tube on the 2 solder tabs of the right Martian flasher's bayonet to prevent them from shorting to each other or anything else. Because my wires were already soldered, I opted for a bit of electrical tape on both tabs on the remaining 3 #89 bayonets. I ran the machine for a minute or so with the flashers a-flashing, & there were zero problems. This makes sense--I'm sure that a little jostle or bump or transport of the machine could swing these solder tabs around to short to each other or a neighbor & cause bulb burnouts. Seems reasonable.


So, are these #89 bayonet tabs supposed to spin around freely??!?! I threw 4 new replacements on my current Pinball Life order (only $1.25 ea) to have on hand if I want to swap them out altogether & dress the wires more cleanly.


Thanks for reading along--I hope that my ramblings will be of use to somebody, possibly my future self who will need to go back & troubleshoot again sometime down the road!
-Jason

[ChadTower]

So, are these #89 bayonet tabs supposed to spin around freely??!?! I threw 4 new replacements on my current Pinball Life order (only $1.25 ea) to have on hand if I want to swap them out altogether & dress the wires more cleanly.

Can't remember ever going into my RFM and poking at those sockets.  I'm going to go out on a sturdy limb and say no, though. 

It's pretty common to have some of the smaller parts need replacement for one reason or another.  A good thing to check now, while you're in there, is the microswitches on the outer orbit shots.  That era Williams game tends to wear those to the point that they don't register fast enough for a speedy ball.  They work in test, they work with your finger, but when the ball goes over it at 88mph it doesn't register at all.  Check the switches on both right and left Jet orbit shots for this.  That's the sort of thing that never shows up on diagnostics but does negatively affect gameplay.

All three Williams DMD games I have owned had this issue and my RFM did too.

[jasonbar]

Oh my, I've already gone in & replaced tons of microswitches in this RFM. Yah, many didn't register at all, & many didn't register in gameplay but did in test mode.

I've got a handful more microswitches in stock, but I think that my switches are all behaving now, last time I checked.

Thanks yet again for reading & chiming in. My Pinball Life order should ship out today. I'll report after fixing the Auto Plunge coil & swapping the not-quite-the-same-but-seems-to-work Bottom jet coil. For the sake of science, I'll try the current spinny flasher bayonets as-is & see how they behave--I'll swap for fresh ones as needed. (And I'll see if the new ones are spinny or not--that would be good to know...)

-Jason

[ChadTower]

Sweet!

I learned eventually that you can order a ton of the bare switch and move the armature from the old switch to the new.  Not sure if that is ever put out there as common knowledge.  Up until I realized that I was buying specific part numbers and paying a lot more per switch. 

Thankfully my SWE1 playfield didn't need a damn thing done to it.  It had about 150 plays on it when I traded for it.   

[jasonbar]

Yup, naked switches are the way to go--the armature just jacks up the price!  =O


My SWEP1 was nigh perfect too: 1 broken P-clamp at back of light saber, several bulb sockets needed to be popped & cleaned/bent & reinstalled, & the Naboo ship needed to be adjusted. That was it. NOS, baby!  =D  (Which makes up for the heavy use that this RFM saw--BTW, along w/ the microswitches, I also replaced most of RFM's stationary targets too--foam gone, bent up, ugly, etc.)


While I've got your generous ear, there are 2 more quirks that irk:

1 - SWEP1 saves settings but seems to lose high scores on power down--bug or feature? I know there's a coin battery on the UNIX PC mobo that's easy to replace, & apparently there's a hard-to-replace battery on the PRISM card. Any idea?
2 - My "ATTACK MARS" art on the banana flasher insert in the middle of the RFM playfield is about half gone--any leads on where to buy a replacement? I haven't found a source for a decal. I did find some DIY options on Pinside--if I can get a high-res scan of the art, I can probably have an artist friend print it on clear adhesive-backed "paper." I'm tempted to remove the remainder of the existing art just so it looks cleaner...


Thanks,
-Jason

[ChadTower]

Hrm.  IIRC the high scores battery is on the CMOS card (daughter card to the prism card) and is a 2032.  Doublecheck that, though.  Obviously you want to verify that battery anyway to be sure it's not in danger of leaking. 

I can't seem to turn up any off the shelf decals either.  I'd offer to get a good pic of mine, since the playfield isn't in the game, but it is in deep storage during my gameroom construction project.  I won't be able to get my RFM playfield out for a while.   

[jasonbar]

2032 coin batteries everywhere!!!

1 on the mobo. It was reading 1.3V, so I put in a new one.

RFM had one in the meat of the PRISM/ROM card sandwich. It was reading strong, so it stayed.

SWEP1 was missing a battery in its PRISM/ROM card sandwich. Probably why SWEP1 high scores weren't saving.

Interesting...RFM & related system settings didn't get lost in the battery swap. The only thing needing redoing was the date & time. We'll see if that sticks now.



As for #89 bayonets, I just handled 4 of them on my Congo & none had spinny tabs. Might as well swap them out on RFM with new ones to keep the flasher bulbs safe & sound.


Thanks,
-Jason

[ChadTower]

Nice.  I usually replace all batteries whenever I'm in the head of a game I haven't worked on in a while.  If it even still has batteries.  Batteries are cheap but MPU boards are very much not.  And that Prism card is unobtanium.  The more I think about it the more I want that 2032 out of there.  There has to be some alternative.

[jasonbar]

Some Googleage revealed mixed messages about button batteries leaking, but the general consensus seems to be that they're way less of a risk than AA batteries. I saw no hard evidence one way or the other. However, I do have some spare 2032 holders leftover from when I was refurbing Sega Dreamcasts (which have the battery spot-welded in place, so you need to provide & wire up a replacement holder if you want to be able to change out batteries), so there's no harm in a little remote button battery action.

On the PRISM/ROM cards, there are 2 nice big solder pads that are exposed when the sandwich is sandwiched--these are the + & - tabs for the battery holder. Simply remove the battery & solder two wires to those pads & run them to a little 2032 holder elsewhere in the backbox. I'm going to see if I can fashion a short stubby cantilever beam that picks up the same screw used to secure the card to the PC chassis (plus probably a neighboring unused screw for the card's neighbors--with 2 mounting screws, the diving board will be fully constrained), & hang the battery over the edge on the outside of the PC case.

I'll report back after I figure out materials (might have some thin G10 Fiberglas sheet leftover at work) & details & survey what else is in that neighborhood, so that I'm not dangling the battery over something else fragile & expensive.

-Jason

PS--as for the PC's 2032, I haven't given it any thought. I'd rather not decase the PC to get to the solder side of the mobo--I'll probably just carefully solder to features on the top side of the battery holder & run wires to a new holder secured to the outside face of the PC case.

[ChadTower]

The motherboard isn't nearly the same concern.  There are multiple replacements you can swap in if you can find them.  That prism card, though, is proprietary and very very expensive to replace. 

[jasonbar]

I couldn't find an elegant solution for remote-mounting the PRISM card button battery & having it look good & be robust & be remote yet attached.

So, I simply desoldered the battery holder & installed it on the other side of the PCB, so it can be accessed easily without undoing the ROM/PRISM card sandwich, & if it actually does leak, it will simply booger up (relatively) easily repaired traces rather than precious components.

Attached pic shows 1 of my sets stock & 1 modded.

-Jason

[lilshawn]

the button cell batteries are pretty good at keeping their innards inside. You shouldn't have to worry about them leaking too much. The old lithium and ni-cads on the other hand...

[ChadTower]

While that's true... "shouldn't" is not good enough on that Prizm board.  If that board it'll cost him a grand to replace it.

[jasonbar]

Coda:
Fiiiinally had time to re-attack this problem. (The problem of the autofire coil/diode/fuse for RFM's autofigure.)


Last update from way back when, before I attacked this issue last week:
I'd replaced the diode & transistor for the autofire coil. It still blew the transistor & fuse.


This weekend's progress:
I swapped the old autofire coil for a new one, using a used diode from a working coil pulled from a jet bumper (This bumper had the wrong flavor of coil from some previous repair, so I replaced that bumper's coil to be what it was supposed to be). Success! This concludes that my problem was the old autofire coil, which I just measured to have the same resistance as the fresh one that I used. Puzzling...

No matter, all's well in RFM land & my p2k combo plunger/button launcher is fab. I'm now pretty good at replacing the 3-legged transistors on the power board! :]

Thanks for the help, guys.

-Jason

[lilshawn]

a small occlusion in the enamel of the wire of the coil can test fine with the low voltage used in a meter, but get 50 volts flowing through there and it will spark across.

[jasonbar]

Excellent observation--thanks for bringing up a fine point!