Bally 891-20 Super Continental credit overpay problem

[essmeier]

EDIT:  Problem solved - Improperly adjusted switches on the score motor cams, which must be adjusted for each coin played.  Excessive detail about that here.

After three months of working on this allegedly "complete" Bally 891-20 Super Continental, I've largely fixed all of the problems it had when I acquired it and I've replaced nearly all of the missing parts.

It now can play in both coin and replay mode.  In coin mode, it accepts and pays out in coins.  In replay mode, it accepts coins and pays out in credits on the credit register.

But...

...while it pays correctly on all payouts in coin mode, it has a tendency to overpay in credit mode, especially on larger payouts.

It's usually correct in the payouts on cherries, but on larger winners, it tends to overpay, but it's not consistent about it.

Example:

3 oranges with 1 coin should pay 10 credits.  In 5 tries, I got 10,10,10, 11, 11.

3 oranges with 6 coins should pay 60 credits.  In 5 tries, I got 60,60,60,66,72.  One time yesterday, it paid 96 for that combination.

4 bells with 1 coin should pay 50 credits.  In 5 tries, I got 51,51,51,51,52

A payout for 3 bar-3 bar -3 bar for one coin should pay 200.  It might pay that, or it might pay 202, or 205, or 207.

I have completely disassembled, cleaned and reassembled the credit register.  I've adjusted and cleaned all of the switches on the score motor cams.

And as I noted above, it pays correctly for coins on every payout.  This is strictly a credit problem, but I don't understand how credit play works well enough to know where to look here.

Schematic here.

Suggestions?  This is literally the last remaining problem on a game that had dozens of them.

Charlie

[wolftalk]

the shortish answer is you likely have an issue with the wipers<->rivets contact on the odds unit.   


If you take off the wipers to clean the rivets, pay attention which way they go on since they can mount on the shaft two ways.  Reset the unit and mark a wiper and contact plate.  Also, hold the white nylon ratchet in place when removing and especially installing the wipers to prevent the ratchet from moving away from the frame far enough to disengage the reset stop barb.  If that happens, you'll need to remove the coil spring and rewind it with typically 1.5-2 turns of tension at reset.


the tedious reason why ....



if I remember correctly (start worrying):

- in coin payout mode the replay counter unit is mechanically stepped up by the action of a coin pushing down a linkage that is directly connected to the solenoid plunger in the replay counter step-up coil.

- in replay mode, the odds motor switches electrically power the replay counter step-up solenoid.


if you are getting correct pays in coin mode, then the adjustment of the zero stop, contact plate and pawls on the replay counter are ok.

the odds disc and the different cams/switches on the odds motor handle large payouts by changing the ratio of replay register stepping (credit adding) to payout counter stepping.  Payout counter stepping is what stops payout.

an example using your numbers:

1] you know that an orange win is handled via the 10 trace on the payout counter because the replay register and replay counter are both stepping from odds motor cam 5 switches when one coin is played.  i.e. the payout is 10 and both the units are stepping at the same time, so it must be the 10 trace and that is true for all orange pays.

and yeah, you can just look at which trace orange is connected to on the payout counter diagram on the schem to get the same answer :-)

2] with 6 coins played on an orange win, the 10 trace is still being used.  A cam 5 switch is still stepping the replay register, but the replay counter is stepped from a cam 9 switch.  The schem says cam 5 has 12 lobes, so you want the thing to spin around 5 times to pay 60.  In those 5 rotations, you want the replay counter to step 10 times.  Conclusion is cam 9 must have 2 lobes and the odds disc selected a 6:1 ratio of stepping the replay register to the payout counter.

that's why your overpay is 6.  A step-up pulse to the payout counter got lost.  The reasons for that could be:

1] poor wiper<->rivet contact on the odds disc

2] flaky odds motor switches - either dirty or poorly adjusted ... but you already cleaned and checked those.

since you're also getting a missing payout counter step-up pulse on other pays, the problem is almost certainly the wiper tension to the rivets, cruddy rivets or if there's little wires going thru a slinky coil from the contact plate to the wipers, one is breaking off.

if you got this far, get yourself a drink :-)

[DavidLee]

The machine is working properly in coin mode. This tells me that the circuit in credit mode is the problem.
If contacts are to close or far apart and or dirty this could effect the payout.
Check all contacts on the credit circuit.

[essmeier]

if you got this far, get yourself a drink :-)

I read your explanation three times.  I think I need the whole bottle.  While I'm not sure I fully understand credit pay, I did now know where to look for the problem.

So...

I pulled the Feature Unit and had a look at the Odds Unit.  Those are the shiniest rivets I have ever seen.  They absolutely sparkle!

It wasn't the Odds Unit wipers.

Pulled the switches on the score motor...again.  Fortunately, the two sets of switches are mounted to a bar that's removable with two screws - the whole bank of switches lifts out.

I'd cleaned them the other day and made sure that nothing looked terribly wrong, but this time I gave them a closer look.  A couple of switches that are normally open were gapped a bit wide and might not have been making solid contact when closed.

Cleaned them again and adjusted a few.   Checked continuity everywhere.  Powered the machine up and...

...4 bells paid exactly 50 credits.

... 2bar-2bar-2bar paid exactly 100 credits...

...but the credit button was suddenly a bit flaky.  When I pressed the button, it would sometimes take two credits off the counter, as though I'd pressed the button twice.

...so I pulled the Feature Unit out again and tweaked the two switches with the 54 wire at Cam 1 and Cam 5.  That fixed the credit button issue.

So, at the end of the day, the problem was improperly adjusted switches on the score motor cams.

All good.  Credits are playing properly.  Payouts are accurate.  I think I'm done with this thing - at least as far as mechanical and electrical work go.

Thanks, guys!

[DavidLee]

Glad your machine is working property.
Yes, it only takes a switch that is to close or to wide to keep the machine from operating as intended.

[essmeier]

OK, I need to reopen this topic that I thought was solved because I'm still having overpay issues and I still can't quite wrap my head around how this credit register business works.

For simplicity's sake, I'll stick to problems with single coin payouts.

What should pay and what is actually paying is shown in the attached image.

Bottom line?  Small payouts are OK.  Payouts involving multiples of 20 or 50 are overpaying.

Reminder - Game pays correctly in coin play mode.   This is strictly an issue involving credit play.

I'm sure it's a simple matter of adjusting a switch or two on the score motor unit.  My problem is that I can't figure out which switches there may be causing the problem.

I'm sure it's the cam switches, as everything is clean and adjusted on the Odds Unit and I see nothing amiss there.

Perhaps Wolftalk can try to explain to me again how this works.

Why am I getting overpays for multiples of 20 or 50 coins?

Link to schematic in the original post.

[wolftalk]

hang on ... if I have to read what I wrote previously, I need to go get the bottle ....


ok, you've simplified things by having an issue when playing only one coin.  In that case, the circuits that are adding credits are electrically stepping the payout counter, and doing it with 1 credit = 1 payout counter step.


on the flip side, you've made life complicated by saying that payouts of 2, 5, 10 and 14 always pay correctly, so that implies the electrical circuit is working fine. 


If the <20 payouts occasionally screw up, ignore the following.


--------------------- proceed only if credit pays less than 20 always work -----------------------


enough disclaimers?


first thing I'd do is pull the hopper and mark exactly where the wipers are after manually stepping the payout counter up 20 times (or after a 20 coin pay).  What you want to see is after a credit overpay, are the wipers at the same spot or have they stepped more times.


IF the wipers are stopping at the same spot on a credit overpay,  then the assumption is payout is always ending when the wipers step the correct number of times, and since lower pays always work, the step up pulses to both the register and the payout counter are being correctly generated.


that means it's time to put on the tin foil hat, because now ya gotta explain why an electrical pulse to the payout counter got ignored.  The most likely reason is heat in the coil is causing the plunger to occasionally stick.


yank the hopper again, disconnect the mechanical linkage at the spring, remove a screw and flip down the payout counter so you can get to the coils.  Operate the step-up plunger by hand, wiggling and seeing if there is any hint of drag or binding...or just take off the coil and check the plunger end for a mushroomed sharp lip and the coil sleeve inside for wear - especially ridges if it's plastic or holes if it's metal.


you can file the plunger to remove the mushroom, but the coil sleeve you have to replace unless the problem is someone put lube in it...then you can just clean it out.  There should never be lube in solenoid coil sleeves.


if you don't have a coil sleeve handy, look for another one in the machine that you can swap with as a test.


so why does coin pay work?  Because the step-up coil is not used, so it doesn't get hot.  In addition, the plunger is likely not pushed all the way to the coil stop by the mechanical stepping, so if the hang up is happening when the plunger slams into the stop during electrical operation - or there's goop on the coil stop - you avoid that in coin payout.

[essmeier]

I stand corrected.  After a lot of testing, it appears that it does also occasionally overpay on smaller payouts, though less often than it does on larger ones.

In ten tries, 3 oranges correctly paid 10 eight times, but one time it paid 11 and once it paid 13.

In ten tries, 2 cherries correctly paid 5 nine times, but one time it paid 7.

Coil, sleeve, and plunger are clean and the mechanism moves smoothly.

I imagine it's an issue with the motor cams.

[wolftalk]

ok, so it's one of:


1] the replay register is double stepping.  Is it adding credits on a regular cadence with no blips where it adds and extra one?


2] the pulse to the payout counter is not reaching the coil due to a flaky plug connection


3] intermittent connection on the odds disc or a switch on odds cam 5


assuming it's not [1], it's jumper wire time and hope the 891-1 schem that I'm looking at matches your game.


go onto the score/odds motor unit and find cam 5.  One switch on cam 5 will have brown/red (not white...oops) wire 61-8 and white/black wire 58-3 on it.  That switch sends pulses to the replay register for all odds except 5 coins.


another switch on the same stack will have brown wire 60-3 and grey/red wire 91 on it.  That one sends the pulses to the payout counter step-up coin for single coin play.


jumper wire 58-3 to 91 and ONLY play one coin.


if payout is still flaky, your problem is likely in the plug connections for wire 91 getting to the payout counter.


if payout is reliable, move the jumper to connect wire 61-8 and 60-3.   

again only playing one coin, if the payout is bad your problem is the odds motor switch.  Got good switch overtravel? 

if your payout is good, your problem is on the odds disc/wipers.  Wiper pressing on the rivets with decent pressure? (lift wiper with fingernail a little and let go and hear a satisfying plink as the finger contact snaps down on the rivet).   You don't want too much pressure or the wiper reset can drag/stall.  The wiper you care about is the F finger on the outer rivet ring on the below diagram.

[essmeier]

1] the replay register is double stepping.  Is it adding credits on a regular cadence with no blips where it adds and extra one?

Yes.  It's adding credits normally.

jumper wire 58-3 to 91 and ONLY play one coin.  if payout is still flaky, your problem is likely in the plug connections for wire 91 getting to the payout counter.

I have jumpered wire 58-3 to wire 91 and it's still overpaying. 

To get from the cam 5 switch to the hopper, that 91 wire passes through a Jones plug in the top of the cabinet and through a Beau plug at the hopper.

I guess my next step is to come up with a really long jumper that bypasses both, connect that wire directly to the payout counter step-up coil, and see if it pays correctly.

[wolftalk]

sounds like you're on top of it.  I guess you could still try jumpering around the jones plug connection without too much trouble.


for the jones plugs, pull the plug and clean off any black crud on the pins.  You care about the inner and outer edge of the pin that actually touches the socket tab.


since it's usually a pita to get to the inner surface of the sockets, what usually works well is to squeeze hard along the length of the socket between your thumb and finger (with the plug out).  You want to bend in the outer tab a little so putting the plug in requires a bit of effort.  That clamping pressure plus the scrubbing action of inserting the plug almost always makes a reliable connection.

[essmeier]

OK, I've checked, cleaned, and adjusted the Jones plug and I've got rock-solid continuity from the 91 wire on the switch on cam 5 all the way to the coil on the payout counter.

I'm still getting overpays.

I pulled the hopper, reset the payout counter, manually advanced it 20 times, and put a pencil mark on the circuit board next to the wipers.

I reset the counter, put the hopper back into the machine, played one coin, and set the game for an orange-orange-orange-orange (20 credit) payout.

The replay register advanced 23 times.

The wipers on the payout board are right next to the mark I made on the circuit board.

For another test, I reset the payout counter.  I manually advanced it 50 times and made a mark on the circuit board next to the wipers.

I pulled the hopper, reset the counter, put the hopper back into the machine, played one coin, and set the game for a bar-bar-bar-orange (50 credit) payout.

The replay register advanced 55 times.

The wipers on the payout board are right next to the mark I made on the circuit board.

I tested this multiple times with identical results.  Payout counter is advancing correctly for pays, but the replay register is advancing more than it should.

I'm sure my testing is supposed to be telling me something, but I'm still not sure what's going wrong.

[essmeier]

Now it appears that I may be close, and might have actually solved the problem.

As the payout counter seems to be stepping correctly, it appears that the problem is the replay register is incrementing more than it should.

The schematic shows two switches that can advance the replay register - a switch on score motor cam 8 (where 81-8 black/red meets 58-3 white/black) and a switch on cam 5 (where 61-8 brown/red meets 58-3 white/black.)

I stuck a business card in between the leaves of the switch on cam 8 and set up a 20 credit pay.  It paid 23.  I didn't expect that switch to affect this, as that seems to be a switch that affects 5 coin play only (if I'm reading the schematic correctly.)

I figured that the switch on cam 5 is likely where I need to look, and I spent some time adjusting it.  Fortunately, the switch is the top one on a stack and it's easily accessible for adjusting on the fly.

If the contacts are too close together, the machine overpays.  If the contacts are too far apart, the machine underpays.

The trick is to get it just right, and for whatever reason, that switch seems to be really sensitive to that.  There's a fine line between getting a pay of 16 or getting a pay of 23 when you're supposed to get 20.

After a bit of tweaking,  I got it to pay 20 credits ten times out of ten and a single attempt at a 50 credit payout was also correct.

I might have solved this, but I'll test it a bunch tomorrow and see if we're good.

Switch photo attached.

[wolftalk]

another possibility is the payout key switch is closed when it shouldn't be and the payout counter step-up arm switch is making an extra pulse.  Depending on if/how it overlaps the pulse from your cam 5 switch, you could get a double step from that thing. 


may be worth sticking paper in it or misadjusting it always open to see if your problem goes away.


just saw your update ..


if you get an extra step-up  of the replay register, it should sound different.  Maybe it's pulsing  too fast to notice?

a poorly adjusted CU 5 switch would do as you describe, but the switch blade movement on the cams should be pretty big and the adjust not that finicky. 


when the switch stack is not on a cam lobe, is the cam follower firmly on top of the cam edge and all the moving blades down as far they they can go ... no gaps between the blades and plastic lifters?   Sometimes due to metal fatigue the moving blade(s) is not moving completely with the stack and may be floating in air when the stack is down ... that's the exception case where you bend the moving blade to adjust.


typically the normally open switch contacts should touch when the stack is a little around halfway lifted.  The remaining lift to get to the top of the cam lobe gives you the scrubbing overtravel.

[essmeier]

Here's a bit of an update, as I've done some testing.

I'm primarily an electromechanical pinball guy, so I've got a lot of switch parts lying around. 

I built a new switch for the top one on cam 5 just to see if it would make a difference.  I built the switch out of new parts, rather than repurposing old ones.

Did not make a difference.  Overpays remain.

I've got a Bally Skill-Score arcade game that has a three digit replay register, so I swapped that out for the four digit unit in the slot just to see if the problem might be a mechanical one with the register itself.  I didn't expect to find one; I bought the four digit register from a guy who sells parts for Bally bingo machines, and he'd disassembled it, cleaned it, lubed it, and replaced the coils before he sold it to me.

It's not a mechanical problem with the replay register.  Overpays remain.

So...I put the original switch back in the machine and reinstalled the four digit replay register.

With a bit of tweaking on the cam 5 switch, I've got it adjusted to where it pays correctly most of the time.

I just ran through the full list of pays for one coin, from 2 to 250 credits.   At the moment, they're all paying correctly except for the 250 credit payout, which is paying 203, 206, or 208.  Correction: 253, 256, or 258.

I got it to pay correctly on a 20 credit payout ten times out of ten.

Just baffled with this.  It can't be that hard to get this thing properly awarding credits.  Credits are money, and I can't imagine that the switches would be super finicky about how they're adjusted.  Bally wouldn't sell many machines if they were all overpaying the customer.

[wolftalk]

if forget ... is your game a 891-1 so it matches the schematic exactly?


while you could  remove the 58-3 wire from the replay register step up coil and from the cam 5 switch, then use a jumper wire between those two points to eliminate other circuits from possibly stepping the replay register, this really doesn't make sense since you'd jumpered 58-3 to 91 in the past and still got an overpay, which means any extra pulse on the replay register would also go to the payout counter.


you've also found that the payout stops when the wipers are at the correct step on the payout counter, and coin pay always works.


the logical conclusion is something is causing the payout counter to miss a step during electrical operation.


have you stuck your hand on the payout counter frame during payout to see if there's hiccups in the step-up cadence and/or count the coil activations?


if the coil activations/cadence is the same as the replay register, but the wipers moved less steps, then take a good look at the ratchet and pawls to see if the pawls are sticky and not always staying down on the ratchet teeth when the coil is slamming away, the stroke of the linkage makes the pawls unreliably catch ratchet teeth ... or something daft like a missing spring on the step-up pawl




today's tin foil hat theory brought to you by Arbeg Distillery ... peat moss is not just for the garden.

[essmeier]

if forget ... is your game a 891-1 so it matches the schematic exactly?

Not a good time for a lengthy reply, but no, my game is an 891-20.  That, from your documentation, is the Hacienda casino variation and I haven't been able to find the exact schematic for that one.

The closest schematic I've found for it is the w-1046-1025d.

It's not perfect; so far, I've discovered that one or two wire colors are different, but otherwise, it's really close.

I'll read your post in detail and do some testing tomorrow.

[essmeier]

Further testing today, and pays were correct from 2-50 credits.  At 100, it might pay 100, or 101, or 103.

It's worth noting that when it overpays, the coil on the replay register seems to be engaging for every credit.

If it's paying 106 for a 100 credit payout, I hear that coil fire 106 times. 

If I can, I'll try to post a video (with audio) tomorrow of an overpay.  That might help.

[wolftalk]

the 100 and 250 pays are a bit different.  They use the "outboard wipers" on the payout counter carriage ... the thing that moves via the spiral disc.


the problem with the outboard wipers is they move a small amount for every step of the payout counter.  It'd be almost impossible to line those things up so they stepped off the end of the 100 or 250 trace on exactly the right step reliably, so that's not how they work.


instead, the 100 and 250 wipers get the payout close, say 93 or 243 credits added, then the carry-over trace on the payout counter is used for the remaining steps.  Since the C.O. trace is swept by the main wipers that move a bigger distance, stepping off the end of that is reliable.


adjust the outboard wipers so the contact steps off the 100 trace around step 93.  That'll allow for slop in the carriage so at step 100 the outboard wiper can't touch the 100 trace if you wiggle the carriage around.

It's worth noting that when it overpays, the coil on the replay register seems to be engaging for every credit.

did you mean the coil on the payout counter on the hopper?   


the replay register coil has to power once for each credit ... there's no way to add multiple credits on a single coil activation unless you have parts missing near the big star cogs on the side of the replay register.

[essmeier]

did you mean the coil on the payout counter on the hopper?   


the replay register coil has to power once for each credit ... there's no way to add multiple credits on a single coil activation unless you have parts missing near the big star cogs on the side of the replay register.

I meant the coil on the replay register.  It's a solid, refurbished unit and I have no reason to suspect that it is not operating correctly from a mechanical perspective.  My point is that the coil is receiving voltage to fire it once per credit, even for overpays.

Yes, that should be obvious, but I thought I'd make that clear.

I will have a look at the 100/250 traces later today, as that seems worthy of a closer look.  At the moment, the other pays are accurate (he says as he knocks on wood.)

I want to be careful, I think, as it does pay correctly for coins and I don't want to mess that up.