Bally 891-20 Super Continental credit overpay problem

[essmeier]

I took a look at the wipers on the hopper for the 100 pay and the 200/250 pay.

The 100 pay wiper was stepping off of the trace at 84.  I've adjusted it so that it now steps off at 92.

The 200/250 wipers were slightly crooked and not aligned with the traces.  I've straightened them out and they're now properly aligned and stepping off at 192/242.

I just tried three different payouts - 50, 100, and 200.  I got 52, 102, and 209.

Cherries are working correctly, though.

 

[wolftalk]

wish you lived within a couple hundred miles of me ...


did you get around to making a video of the cam 5 switches in action?


have you counted the number of coil activations on the replay counter step-up coil vs the position the wipers wind up at? 


did you jumper wire 61-8 to wire 60-3 on the cam 5 switches (or the odds unit disc) to eliminate a flaky wiper connection on the odds unit?


when you've previously mentioned continuity, are you measuring with an ohmeter or ? 


continuity checking with something that is go/no-go like a test light or buzzer can bite you.  A circuit with an abnormally high resistance due to a poor connection will show continuity just fine, but it may not handle the high current requirements of solenoid coils and you get flaky solenoid operation.

[essmeier]

Well, that was fun.  Typed up a nice, detailed post and lost it all due to an oversized attachment.

OK...

I have jumpered 61-8 to 60-3.  Still get overpays.  For a 100 pay, I got 100, 100, 102, 103, 109.

I have created a video (zipped MP4.)  Video can be downloaded here: https://easyupload.io/bp8fpt

That's a bell-bell-bell-bell 50 pay...that pays 51.  Switches in question are the ones in the middle stack as shown on the attached photo.  Perhaps watching the video in super-slow-motion will provide some insights.

Top switch is the sensitive one - 61-8/58-3 - that's the one that fires the replay register step-up coil.  Adjust it too tightly and the machine overpays - sometimes by a lot.  If the blades are too far apart, it underpays.

As for continuity, I used a digital volt meter.  I'm aware that continuity can be somewhat misleading on electromechanical devices.

I checked:

1. Continuity from 91 wire on cam 5 to pin on Jones plug (with feature unit disconnected)
2. Continuity from 91 wire on socket of Jones plug to female Beau plug at hopper (with feature unit disconnected and hopper removed)
3. Continuity from 91 wire on male Beau plug on hopper to payout counter coil (with hopper removed)
4. Continuity from 91 wire on cam 5 switch to payout counter coil (with everything reassembled)

I'm confident that the connections are all good.

[wolftalk]

the gap on the top switch is a kinda big.  The non-moving blade (the top one) is not being lifted much by the moving blade, and occasionally it seems to barely move ... like there's a shorter/worn lobe on the cam.

the vertical blade (cam 6) to the left in the video is more like what that gap and overtravel should be.

if you made the gap any bigger, you'd get an underpay because occasionally the switch won't close well enough to power the coil.

making the gap smaller should just make the thing reliable and possibly arc less, assuming you don't make the gap so small that it sometimes arcs when open and creates extra step-ups on the replay register.


having said that, at least the replay register appears to be stepping every time the cam 5 stack is lifting, so your problem is not the top switch on cam 5 stack.

that leaves the payout counter.  You getting overpay when playing two coins?  That uses a cam 6 switch instead of a cam 5 switch to pulse the payout counter step-up coil.  If you still get overpay, something is wrong with the payout counter stepping or the connection from the score motor switches to the payout counter.

I don't know if you point your camera at the hopper if you'll pick up the noise of the payout counter stepping instead of the replay register.  You might have to stick a piece of paper in the cam 5 switch to stop the replay register from stepping so the only noise is the payout counter stepping.  Then you'll be able to hear if the payout counter is missing a beat, especially if slowing down the video (I used vlc media player to slow down your score motor video).

wrt to continuity, assuming you've ensured the circuit you want to test is completely isolated so there's no roundabout closed path thru stuff like coils, lamps, or the transformer, then continuity will always tell you of a circuit is open. 

It will also tell you if a circuit is closed, but that doesn't guarantee the circuit will work.  It's better if you measure ohms since a >1 ohm resistance is a problem, but there's still the rare case that some part of the circuit will limit current flow enough to make a solenoid or coil not work.

for example, the wires in the game are stranded.  If strands have broken so only a couple are left holding the wire to where it's supposed to connect to, continuity, resistance and even voltage when no current is flowing will be fine.  Make current flow in the circuit and it won't work ... those two wire strands don't have the capacity to flow the necessary current.  A cruddy connection can do the same thing.

if you jumper the grey/red wire on the payout counter to the white/black wire on the replay register step-up coil, those two units should step up in sync.  If they don't, something mechanical is wrong with the payout counter when it's electrically stepping or there's an issue with the coil - a poorly soldered or broken coil wire that intermittently disconnects or an incorrect/out of spec coil).   

if you want to be really sure, jumper the orange wire on both units together in addition to jumpering 91 and 58-3







 

[essmeier]

making the gap smaller should just make the thing reliable and possibly arc less, assuming you don't make the gap so small that it sometimes arcs when open and creates extra step-ups on the replay register.

That's the switch that's sensitive to adjustment.  If I close the gap, it increases the overpays.  Opening the gap, as you said, produces underpays.  There's a sweet spot that gets me correct pays most of the time, but not always.

I just did an experiment - I grabbed a piece of spare wire and connected it directly from the 91 wire on the cam 5 switch and ran it directly to the 91 wire on the payout counter coil.

I didn't disconnect anything, but I figured if there's a continuity problem at the Jones/Beau plugs between the cam 5 switch and the payout counter, this will bypass that.

With a single coin set for 50 pay, I got 50, 50,50, 51, 53.
With a single coin set for 100 pay, I got 100, 100, 100, 103, 109.A single attempt for a 200 pay got me 201.

I think it's safe to say that we have continuity on the 91 wire and that it's not the issue.

From the day I acquired the machine, I've done nothing with two coin play.  I'll do some tests with that and see what's up.

[wolftalk]

That's the switch that's sensitive to adjustment.  If I close the gap, it increases the overpays.  Opening the gap, as you said, produces underpays.  There's a sweet spot that gets me correct pays most of the time, but not always.

I don't know if the switch stepping the payout counter in single coin is 5A or 5B, so I'll just call if 5X below.

the logical explanation is you have adjusted the 5C (top) switch to not work 100% of the time, and it's offsetting an issue with the payout counter not stepping every time. 


if there is a worn/shorter cam lobe, you've basically matched 5C and 5X so they either both work or both don't work in sync almost all the time, but occasionally 5C works and 5X doesn't, so that makes an extra credit in the payout.


that also explains why 5C is so fiddly ... a slight adjustment either way will make it work more/less often, but there's no way you'll get it to behave exactly like 5X all the time if 5X intermittently fails.


the problem is your video sounds like the replay register is stepping every time the cam 5 switches lift.  It's possible that the coil is powering enough to partially pull in the solenoid and make noise, but not enough to roll the wheel to the next digit.  It kinda looks like that is happening when the video is slowed down.  Easy way to tell is point your camera at the front of the register and see if you are getting incomplete or irregular incrementing of the numbers.  If you are, the 5C switch gap is too big.


I'd start with making sure the switch gaps and contacts on both 5C and 5X are correct ... where correct is like your 6B switch ... then start debugging issues. 


The normally open switch gap is typically around 1/16" - 3/32", and make sure there's overtravel where both blades are moving after the contacts touch.


ignoring everything said previously, I'd have guessed your problem is dirty/misadjusted 5X switch. 


but ... didn't you say previously that you had overpays when playing 2+ coins?  That would make 5X much less likely since a different switch steps the payout counter for each coin played.

[essmeier]

I don't know if the switch stepping the payout counter in single coin is 5A or 5B, so I'll just call if 5X below.

On cam 5:

5A - 91/60-3 wire to the payout counter from the odds unit. (bottom)
5B - 54-3/36-2 switch that steps down the replay register (middle)
5C-  61-8/58-3 wires that power replay register stepup (top)

Careful examination of the lobes on cam 5 show that they all look OK when compared to lobes on adjacent cams.  I don't see any evidence of inappropriate wear.

That switch at 5C is arcing a lot.  I haven't really looked, but I'm wondering if it isn't occasionally arcing in between lobes and that's what's causing the overpays.

I will readjust 5A and 5C and do some more testing, including testing with two coins.

UPDATE: I did some testing with two coins, and I'm getting overpays.  Four cherries should pay 40; it paid 44, for example.

However...when I play with three coins, everything pays correctly.  I tried every pay combination and every single one of them paid exactly what it was supposed to.  One cherry pays 6, as it should.  Bar-bar-bar-bar should pay 750, and it does, and every payout in between paid exactly as expected.

I think this is good news, but I'm not quite sure what it's telling me.

[essmeier]

With further testing, pays for either three coins or four coins seem to be correct.

Pays for one, two, five, or six coins are overpaying.

Still, with some of it working, it shouldn't be too hard to get the rest of it right.

[wolftalk]

5 & 6 coins both use cam 9X switch.


it seems unlikely that 5A, 4X and 9X all have issues, but who knows.  Have you cleaned off the switch contacts in case someone hit them with wd-40?  I usually use a wire wheel on a dremel if the contacts look shellacked or coated with something wet looking, but anything that doesn't scratch the metal is fine.  Dry black carbon is ok to leave alone.


the only other variable in the circuit is the wiper/rivet connection on the odds disc.   If you jumper wire 25 on the odds disc to the appropriate wire for the number of coins played (60-3, 14-8, 15-8, etc.), you take the wipers out of the picture.


arcing when the switch is barely closing can be reduced by making the gap correct so the contacts are firmly scrubbed together when the switch is closed.  Arcing when the switch is open at rest happens if the gap is too small. 


some arcing in switch stacks as the switches open is normal.  Eventually a pit will be blasted into the contact faces and they need replacing if they aren't reliable.  Tungsten contacts can be filed to remove the pits, most people file the other contact types as well under the "got nothing to lose" principle.   


the easiest way to see if the payout counter is not stepping as often as it should is to use beau plug jumper cables so you can operate the hopper out of the cabinet.  Making those is easy enough if you have the spare beau plugs and wire, but not many people do. 


you may be able to wedge a mirror next to the payout counter so you can take a video of it stepping while it's in the cabinet and the see if the wipers are stepping like they should.

[essmeier]

Update:
I just spent four hours cleaning and adjusting switches on the score motor cams.  I took every single stack off, straightened out crooked blades, and cleaned every contact.

Then I adjusted the gap on every switch, first by eyeball, then by rotating the cams manually to check visual contact, and finally by putting an ohmmeter on each pair as I rotated the cams to make sure there was continuity only when appropriate.

Then I put the machine back together and did 4200 credits worth of testing using orange-orange-orange (10 credits pay per coin played) as my test.

At first, only three coins paid correctly, but I eventually got correct pays for 1,2,3,4,5, and 6 coins.

Five coins is particularly tricky, and the credits add up a lot more slowly when playing five coins.  I don't know why, but the replay register counts at about half the speed when paying out for five coins as it does with 1-6.

At various times during the afternoon I jumpered around the odds unit; this seemed to have no effect.  I believe all of this has to do with the score motor switches.

Three oranges and three oranges alone is hardly a definitive test, but until now, I haven't had any combination pay correctly for all coin options.

More testing to come, but perhaps I've resolved this.

[wolftalk]

well, ya probably spend way more time than 4 hours plowing through my unnecessarily verbose posts (not even counting nightmare time :-))


5 coin play uses a cam 8 switch to step the replay register instead of the cam 5C switch used for 1-4 and 6 coin play. 


cam 5 has 12 lobes and cam 8 has 10 lobes, so payout is a little slower.

[essmeier]

After more testing, I think I have it working properly, or at least, I'm really, really close.  Last I checked, pays were working correctly for any and all number of coins.

Bottom line - there are seven switches on the score motor cams that need to be adjusted rather precisely in order to get pays to work properly for any combination of coins from one to six.

It just took me a really long time to figure out what's what, even when I was looking at the schematic.

As there's not a lot of information on this forum about credit play on the 891, I'll go ahead and outline what I've done and what I've discovered (with wolftalk's help).

There are 12 cams attached to the score motor.  Switches for odd numbered ones are on top; switches for the even numbered ones are on the side (or front, depending on your point of view.)  Cams are numbered with #1 being the one closest to the motor.

I've labeled everything on the attached image.

For each coin played, there is a pair of switches that must both be properly adjusted to ensure correct pays, and one of those two switches will be either "A" or "B," as I've decided to label them.

The top switch on Cam 5 is labeled "A".  Switch "A" is critical to pays for 1,2,3,4, and 6 coins.  The solitary switch on Cam 8 is labeled "B".  Switch "B" is critical to pays for 5 coin play exclusively.

So, for any number of coins, you need a properly adjusted switch "A" or "B" and also:

For 1 coin play, adjust the bottom switch on Cam 5 (shown as 1x on the image)
For 2 coin play, adjust the switch on Cam 4 (shown as 2x on the image)
For 3 coin play, adjust the switch on Cam 6 (shown as 3x on the image)
For 4 coin play, adjust the switch on Cam 7 (shown as 4x on the image)
For 5 and 6 coin play, adjust the switch on Cam 9 (shown as 5x and 6x on the image)

Hint: Before starting to adjust switches randomly, do a test pay for 1, 2, 3, 4 and 6 coins (I like using three bells, which pays 20 credits per coin played.  That payout's large enough for overpays/underpays to be obvious.)

If any of those tests produce a correct payout, then your "A" switch is likely adjusted correctly and you won't have to bother with it.  If they're all wrong, then there's a good chance that "A" is out of adjustment.

I started by adjusting for single-coin payouts by adjusting top switch "A" on Cam 5 and the bottom switch on the same cam labeled "1x".  There's a bit of back and forth to adjusting them; too tight can cause overpays and too loose can cause underpays.

If you're going to be adjusting switches for all pays, this will be the trickiest, as it's the only time you will need to adjust two switches at once.

Once you get those two switches adjusted properly and the machine is paying correctly for one coin play, you can can consider your work with switch "A" done, and you won't have to mess with it again.

This is important, as switch "A" is also used for payouts for two coins, three coins, four coins, and six coins (but not five.)

Next, with switch "A" adjusted, adjust the switch on Cam 4 while testing for two coin play.  It goes quickly, as there's only one switch to adjust.  Again, too tight gives overpays, and too loose gives underpays.

Once that one's working, test with three coin play and adjust the switch on Cam 6 until it pays correctly.

Then with four coin play, adjust the switch on Cam 7 until it pays correctly.

Skip five coin play for the moment.

Then with six coin play, adjust the switch on cam 9 until you get correct pays for six coins.

At this point, the switch on Cam 9 should be adjusted correctly, and that switch is also used for five coin play.  Test for five coin payouts.  If they're correct, you're done.  Pat yourself on the back and you can go play your slot machine like a Vegas tourist.

If they're incorrect, then the last remaining switch to adjust is the switch labeled "B" on Cam 8.  It's the only switch on the cam, and with the companion switch on Cam 9 already adjusted (for six coins), it shouldn't take much tweaking to get it right.
----------------------------------------------------------

At this point, I think I have mine all working correctly.

Thanks to wolftalk for the help.

[wolftalk]

looks cleaner than when it left the factory.


interesting you had that kind of sensitivity with the switch adjust.  For a normally open switch, usually if you set the contact gap to be about the thickness of a credit card,  manually operate the cam/relay to verify the contacts are touching partway thru the stack movement, and watch for overtravel where both blades move a bit as the stack reaches the limit of travel, you are good.


in other words, what you did in post 29 normally takes care of flaky switch issues assuming the gap when the switch was open was big enough to prevent arcing.   What did you have to do after that?


I don't see a youtube video on basic switch adjusting.  Anyone find one?

[essmeier]

in other words, what you did in post 29 normally takes care of flaky switch issues assuming the gap when the switch was open was big enough to prevent arcing.   What did you have to do after that?

I just did what I outlined in the post above - adjusted each switch, one at at time, until the pays were correct.    I didn't have to adjust them all, as the switches for three coins and four coins were adjusted properly from the start.

On the others, I'd have cases where a switch looked good, but I was getting overpays.  I'd adjust to slightly open the gap until the pay was correct. 

Fortunately, this is a home-use machine and as with my home-use pinball machines, I've discovered that once you get them right, they're likely to stay that way for a long, long time.  It's not as though the machine is in use 24/7.