Pinball switch closure problem

A schematic of the matrix and everything associated with it would help so I don't have to guess how to recreate it. Hint:

View in fixed-width font, like Courier. _/ -o/ o- ->|-

|| -||- ||

created by Andy´s ASCII-Circuit v1.22.310103 Beta

Putting a R in series with the cap will slow the charge *and* discharge and make it look like a bigger cap. The formula is T=RC which is how long it takes the cap to charge to something like 63% of the voltage applied or drop 63% from it's initial voltage to zero if that's where it's headed. Big R, big T.

If the machine worked before, what could have changed that it doesn't work now? Did you replace the original cap with a new one of the same value?

As far as that "ghosting" is concerned, the diodes should prevent a read of another switch from detecting the switch in question. The worst that could happen is that the switch acts likes it closed on the next read. Set the replay score higher ;)

--
Best Regards,
Mike

This is a switch in a current new pinball model (Lord of the Rings) that has this common problem even when new (many owners report this one particular switch sometimes won't register as it is in direct line of the flippers and gets really hammered with fast shots). This company does not currently put caps on their switches. Other companies have in the past on troublesome switch locations like this. So I am experimenting with caps on this switch to solve problem. A .05 cap as used in the past did not completely eliminating the non-registering (but improved it). Looking to eliminate it if possible by increasing the switch closure seen by the cpu even more with bigger cap + a resistor if possible. Once pin company did do something along these lines but only on their non-cpu-controlled coils by putting a 22 uf cap and a 100 ohm resistor across the switch, but I don't want to assume I can just do that here as now I'm dealing with a switch matrix and another company and don't want to damage anything. Just trying to come up with rough estimate on, if .05 is the usual cap, what resistor would I use for a cap 3x as big, or 10x as big, for longer closure. thanks!

In the case of this pinball switch, it wouldn't hurt if it remained closed for the next scan, it is only going to score once anyway even if you held the switch closed by hand for the whole game. So getting it to stretch to the next scan would be great..Frenchy

..05 ?? please don't round standard values.

If I read you correctly, the RC is a snubber to absorb the inducktive kick-back from the relay coil thus preventing arcing at the switch contacts. Otherwise T=RC so bigger cap = longer T. It's a simple proportion

I'd still like to see the schematic of the switch/diode matrix and the way it's connected to the CPU/MPU/MCU so I'm not taking shots in the dark. I have to go back to your OP (original post) just to remember how you said you think it's lashed up. I can only guess:

.---------------------. | CPU | | in outs | '---------------------' | | | | | | | o o | '\ '\ | \ \ | o \ o \ | | | | | | |--- to in | | from switch | | --- .-. --- | | R discharge | | | | '-' | | +-----+ | === GND created by Andy´s ASCII-Circuit v1.22.310103 Beta

Determining how long you need to hold it high depends on the scan rate. If you don't know that, any equation is useless.

You could just put the RC network at the input (after the diode and it would take care of all the switches. I like that better.

I hope all this helps your understanding. A schematic would've helped a bunch. Next time I'll hint louder.

--
Best Regards,
Mike

I don't know how to draw schematics good so I will just describe the various switch circuits:

Switches - single switch, with one input going thru a diode first to isolate it in the matrix (if it's in one) ..05 cap - put across certain switches in matrix to lengthen time they are read. Cap is just placed across the switch. Non-polarized. Sometimes done by the factories.

22 uf cap and 100 ohm resistor - the company that used this did it for stroke increase reasons, not for arcing or anything like that. Placed in series across the switch like the cap above. Was used on some kickers and pop bumpers that did not have to be cpu controlled (they just had to be smart enough to fire when they were hit). Cap and resistor lengthened the time the coil fired. Coil was still controlled with pre-driver circuitry which is what was being lengthened in activation time. This is a different application than some much older pinballs used occasionally putting a very large cap in circuir with the coil power itself to help prevent arcing on electromechanical switches. I will see if I can point to some schematics of this particular switch system...Frenchy

Come on man! look at the line under my ascii art and get with the program, pun intended.

Again, A cap across a switch will pass spikes when the line is scanned. If the software debounce ignores it, great. I'll have to convince myself that it's a solution. For now, T=RC and all the stuff I said about it still holds.

--
Best Regards,
Mike

In the two applications of switch caps in solid state pinballs there ever were - the .05 ufs across matrixed switches, and the 22 uf + 100 ohm resistor on stand-alone pre-driver circuits for firing certain coils (both to lengthen how long the circuitry behind the switch 'saw' the switch as closed) - I've never heard of anyone reporting, nor encountered, false firings of the coils due to the caps from spikes. Maybe it's just the details of how pinballs typically are designed as per the switch reading that this is not a problem. Or the values used...Frenchy