Bally S5500 COD - troubleshooting - fixes UPDATE again

[xwarp]

UPDATE: Changed the thread title because I decided to pursue the faults to component level.



Well.....the previous post about the slant top machine I posted about, that was fixed by a functioning MPU.


The other failure that was not detected until the new board is the COD.


It appears that the COD is not sending out the signal for the "coin credit".


Since these seem to be difficult to find, I'm going to look at building a one shot circuit to do just that using the "coin sense" from the comparitor.


Anyone try this?

[Badbaud]

We sell those COD boards tested in a Bally S5500 for proper operation.

[xwarp]

Just a quick update.


Turns out that one of I.R. LED's on the COD is essentially defective. Appears to have overheated a bit at some point in time by the discoloration around the whisker in it.


Replaced them with I.R. emitters from a spare remote control.


Another issue was with the comparitor output being on the low side, and sense it feeds the COD to initiate the credit, I suspect that helped in creating more issues. Found an op-amp DIP in the unit to be a common use type and since I have NOS of those, I replaced it and the coin sense output is a much better signal.


Between these two fixes, I am getting more credits than I am errors, but still getting the random 20 and 24 codes.


I think I am going to have to pull out the logic analyzers and watch the signals now.

[Badbaud]

The comparator sends out a vend signal to the COD board telling it a coin is on the way.
Have you adjusted your comparator sensitivity control fully clockwise? Recommended for home use.

[xwarp]

I thought fully clockwise was most sensitive?


Found Q3 to be an issue on the optic board. Did not seem to fully turn on when biased, so I got that replaced and now have full VDD voltage to the comparitor.


Got the 4 channel scope out and going to check the signals when a coin is dropped.


Will post a pic when I get that.

[Badbaud]

No, counterclockwise is best for home use.
Clockwise makes it reject more coins as the metal quality of the compare coin has to be closer to the coin dropped through.

[xwarp]

Forgot to make sure the sensitivity on the comparitor was set for full clockwise, but this is another update.


So the coin credit signal is a +5 vdc that pulse to 0 vdc when a coin is dropped, but the comparitor sends out a pulse to the COD that is about +15 vdc. That pulse is then used by the microcontroller to create the coin credit signal that is sent to the MPU. Actually, it doesn't send any signal, it simply pulls the +5 vdc sitting on the coin credit signal line to 0 vdc and that is how the MPU sees the drop as a credit.


Since I don't care about "stringing" or any other possible cheats that could be used in a casino, I am more concerned that only quarters pass through, so I still need the comparitor to do that in addition to the coin sense pulse out of it to create the coin credit pull down pulse.


Knowing this, I found a simple fix to bypass the optics decoder board by using a couple of transistors and several resistors.


This bypass board does work to create the credit signal about 98% of the quarters dropped. The other 2% I believe to be related to the pulse width. Unfortunately, not having a good COD board makes it impossible to know exactly what the pulse width should be.


So to sum it up, this Bally S5500 "All American" slot machine from some place in Vegas called the "Imperial" is now working again. Maybe not perfect, but good enough for me!


On a side note, anyone having an issue with their machines should contact BadBaud for parts because I wouldn't build these boards for less than what his shop charges for a good one.


edit to add the picture of the captured pulses from the comparitor and the make shift COD.


Trace #2 is the Comparitor "Sense" pulse, trace #1 is the pull down pulse that the MPU uses to create a credit.

[xwarp]

Here is the final update of the COD bypass circuit I put together.


The previous version needed something to shorten the pulse with of the "credit" signal going to the MPU, so, I built another circuit just as I had, but I added a one shot timer circuit to output a shorter pulse. Turns out that the MPU likes a negative going pulse about 20 milliseconds long.


Dropped about 150 quarters and not one was missed. Dropped 3 at a time as fast as they'd go into the slot and not one missed.


I am content with this setup.

The following pictures are of the circuit.

[Badbaud]

How about posting a schematic?

[xwarp]

How about posting a schematic?

Will do after I redraw it to clean it up. Also need to look at different transistors as the ones I used are overkill on the specs.

[Oddzball]

No schematic for this?