Finally fixed extra 4-/7-pay coins on my 873 (mech hopper)

[frenchmarky]

I have had one really niggling issue with my 873, where it would intermittently skip past the end of the 4- and 7- coin payout traces a partial step, which depending on the reels position could cause it to pay another few coins since it would then be at the start of another active trace.  Example, it would pay a 4 coin pay, after which the contact *should* have stopped on a blank spot and counter should stop - BUT whatever the problem was caused the contacts to advance a partial step PAST that and it would end up on the start of the 7-coin pay.  Even if it was just for a fraction of a second, as soon as it ran enough to pay one more coin, that would result in continual pays until the second trace ended.  So on a 4-coin pay it would pay another 3 coins that shouldn't have because instead of stopping on the blank it blipped a smidge past that onto the active 7-coin trace, causing it to pick up there until 7-coin pay was finished.  This is a particular issue with the 873 vs. say 3-line or 1 line machines, which don't have these tricky single-step blanks between one active trace and another active trace.

I tried *everything* including replacing the moving trace board contacts (which were worn horribly flat anyway), dialing in the coin kicker perfectly, tweaking the stepper arm return stop tab for absolute minimum of overtravel etc. etc.  Every time I thought "Yayy I think I finally licked it!" 10 min later it would give me another one of these stupid false pays.  What ended up being the real solution was adjusting the coin roller bracket up a little, i.e. reduced the stroke of the roller arm which is advancing the stepper arm.  I wasn't having any other over- or under-pay issues besides this weird one so I hadn't messed with the pivot before but figured what the hell I'll try it.  Instead of using the roller arm gap measurement technique from the Geddes book, I decided to measure and adjust it more directly by:
1.  Holding in coin kicker manually
2.  Placed coin in position in hopper to be counted
3.  Rotated hopper slowly by manually turning motor shaft
4.  Observe when stepper steps (clicks) as roller is raised along with the amount of overtravel past that.
5.  Was able to adjust the roller arm up a bit, reducing it's travel.

That way I was able to more precisely adjust the roller to a reliable minimum of travel, which also reduces pressure on it by the coin along with backlash of the counter arm and roller arm.  Anyway, after playing it for two days just to make SURRRE it was fixed this time, I have proclaimed victory!!!  Does anybody else out there use this method of adjusting the roller arm, or is the gap at the roller arm/stepper arm the usual method?
Hope this info is help to anyone with similar problem.

[OldReno]

What a truly exceptional post!  Wow. Superb work.
I will read this over several times, because it is so excellently written.
The 873's are a difficult one to tame.
Did you try loosening the payboard (2 bolts up inside) and rotating the board?  Sometimes that helps, too.
As a very last resort, you could carefully file down the contact points to make them more V shaped rather than flat.  As a last resort. That could help cut down the bridge between the two hot segments.
Nicely done, BTW.

[OldReno]

Just one quick afterthought...the 7 pay should not be hot when a 4 pay is up.  How is the adjustment on your reel wipers?  Are they hitting/bridging two sets of buttons instead of one, thus making the 7 hot at the same time? That might be your original problem.
been awhile since I had hold of an 873 I will admit....

[frenchmarky]

As for the 7 pay being active... with the hopper pulled from the machine and all the counter contacts lifted off the contact board, the 4 measures as connected on the 'outer' 7 segment but not the inner 7 segment.  4 also measures connected to the 14.  Are you sure this is not correct?  Again this is with the hopper pulled out and all the wiper contacts off the board.  I had a darn 873 schematic but I have been unable to figure out where I put it!  : (  This did seem a little funny to me too but figured if this was incorrect, I'd have a lot more problems such as ones that happened consistently on every 4 or 7 pay.  Also pretty sure it is not an issue with the reel wipers, I have gone thru them very well and all positions are lining up good with no crossing over of wiper contacts to board contacts and all other pays are always perfect.

Rotating the contact board was another thing I tried, ad infinitum.  All it would do is either make it worse (occured more often because contacts were stopping even closer to next segment) or would begin to cause even more basic over/underpay problems.  I could never get the problem to *completely* go away.  This was a toughie cuz it was basically impossible to test it by just testing the hopper by itself, or without the reels being in certain positions etc.  Closest I could get to an isolated test was to win a 4 pay, then keep reaching in and resetting the counter unit and see if it paid 4, or did this skippy thing and paid more.  Sometimes would pay correct, then boom it would make the error.

So... while this roller adjustment did do the trick, I still can't convince myself WHY it worked!  Even after this has been fixed it's still bugging the heck outta me as to the actual cause.

[OldReno]

As I recall the 873 was a very busy payboard, and kinda hard to figure out.  It would be very helpful to get a schematic on that.  Any pics might be helpful too.  Maybe OpBell or some other old timer can chime in here who remembers how those 4 pays worked...?
For kicks and giggles you could put a small piece of paper under selected wiper fingers to see if that affects anything?

Also note, to reset your payboard, all you have to do is push back on your clock variator, and close the reel mech B switch.  That will reset your hopper and do another instant pay. (read page 2 of the following post -- http://newlifegames.net/nlg/index.php?topic=19336.25
You've been very meticulous and thoughtful on fixing this so far, and that's a cool thing....

I presume these pays are all cherry combinations?  If so, then is the 4 pay two single cherries on 2 separate lines?  And the 7 must be one 2-cherry pay and one single cherry?  Are your cherry pay combinations 2, 4, 5, 7 and 10?

[frenchmarky]

Single cherry pays can be 2 or 4, double cherries can be 5 or 7.  I haven't found my schematics yet but here's a description of a single cherry, 4-coin payout from the Geddes book that describes how the 7 coin payout is also active (what mine was skipping over to somehow) -- "...Note that a short 7-pay segment is also activated simultaneously with the 4-pay segment.  There is a one-coin gap between the end of the 4-pay segment and the short 7-pay segment so that the hopper will stop on the 4-pay segment..."  It doesn't actually state that the 4 and that one 7 segment are hardwired in the hopper like I'm measuring but the result would be the same anyway.

Anyhoo, I was initially thinking maybe it was an issue where after the 4-pay finished, the 5th coin in the hopper happened to push the roller up enough to advance the coin counter into a phony 7-coin payout, BUT then was successfully kicked.  However that wouldn't make sense because then I would end up with the phony 7-pay giving out only 6 coins, and I was always getting 7 when it happened.  If it was simply the 5th coin not getting kicked because it was already under the roller, I'd be getting extra coins on ALL pays, which I wasn't, plus this last adjustment I made wouldn't seem to have made any difference with that scenario besides reducing the spring load on the roller a little.  So still looking for my schematics before I close the book on this one. 

[Op-Bell]

I have a schematic for the 873, and somewhere in the darkest, most spider-infested corner of the garage I have a brand new, never used, hopper board for one. I'll see about scanning the schematic.

[frenchmarky]

Thanks, still cannot find mine.  One of those things where you KNOW you didn't throw it out but you've looked everywhere.  That's what I get for removing it from the book! 
I had to do a bit of repair on some worn/burned-off trace ends on my board so it's useable but if you are interested in selling your new one and can find it please let me know, I'd take it.

[OldReno]

That's interesting, that the 7 segment is also hot when the 4 is.  Doesn't make sense to me, but then a lot of things don't make sense to me.  I'll have to try to find the Geddes book as I've never seen it and it sounds like a good source of info.
I still don't believe it, since the cherries do the activating, but then I am often wrong about shit....
If anyone can post a copy of that part of the schematic, I'd love to look at it.
Thanks.

[OldReno]

Actually, on a cherry pay, all segments are hot, at least until the payfinger steps off the cherry pay. I still don't get it.
Op Bell, any input on this?

[frenchmarky]

<<That's interesting, that the 7 segment is also hot when the 4 is.  Doesn't make sense to me, but then a lot of things don't make sense to me.>>

I'm guessing there isn't any particular reason that it was done that way other than it was the easiest way to implement the circuits for all the various pays it has to be able to handle, especially combination pays.  It's neat as long as everything is working right... but if the machine has a minor problem like a 1-coin overpay, the whole 'gap' design can result in even bigger overpay by getting past that gap.