Upgrading a Stern machine to switching DC problem

[jerzmacow]

So I got a broken pinball machine (Magic by Stern c.1970s) about a year ago. I'm proficient in electronics and was able to fix it. Since then I've been modifying and improving it.

One of the things I noticed was the DC power line for all the boards (except 5V) was really fluctuating like a mad man. The power supply considered of a diode bridge and.. well.. that was it. I replaced the 5V with a switching power supply (but kept the 12V it was based off of because something else was using it). That worked great.

Next thing on my list was to deal with the flickering lights every time a flipper was activated. All I did was add a capacitor to the DC solenoid line. Logically, it being DC, this shouldn't affect it negatively. For whatever reason, all the solenoids stopped working. Odd.

I figured this might have something to do with the rest of the non 5V systems having the ugly diode bridge supply that goes from 0-vcc, as opposed the solenoid line which now has a filtered and smooth DC supply. To test this, I switched everything else to a switching power supply (well, a few power supplies because of all the weird voltages, but they have a common ground).

With all the systems on a clean DC system (except for the 120 AC that goes to the sound board and the other one for the VFD tubes), the MPU boots up and makes a sound, but that's it. The feature lights are all stuck on, and none of the solenoids work. When I manually turn all the lights off (by grounding the light data lines or something), they all turn on in the normal order, but they don't go off.

That points to an error in the digital parts of it, but the machine works great on the original dirty DC system. Is it possible to upgrade it to a proper power supply? These problems are so weird I have no idea what to do.

Thank you for reading!
/)*(\

[DaveBonds]

You've created a break in your solenoid circuit.

Capacitors are generally helpful with radio frequency filtration, as they conduct AC, which is why they are used in noise reduction on AC systems (like in this machine). The problem, though, is that capacitors block DC and allow AC to pass/ conduct, which is a great filtration process, but not on a DC circuit. As soon as the plates in the capacitor charge up, it becomes a break in the circuit, which could be the cause of your system coming on and staying on when you play with the lamp circuit manually, because the system (digitally, as you've noted) is not getting a proper signal from the DC circuit that you've put the capacitor into and inherently, a break.

If you bridge that capacitor with a test lead (proper Ga.), your problem will likely go away, but you have the unfiltered DC issue as before, where everything is subject to dim and fluctuate when there is a draw at a solenoid or anything else heavy.

You need to find a good voltage regulator, rated at the amperage and voltage that the machine reads originally (dirty) with a multimeter. @ 12VDC, you might find what you're looking for in the automotive field of voltage regulators, give maybe a DC resistor if you need it to knock it down from 13.5VDC to 12VDC, but they usually require an immense amount of amps.

Really, if you could just find a reasonable amped and regulated power supply that was rated at what you wanted, that would be ideal. For now, though, I would just remove the capacitor in the DC line.

[jerzmacow]

Is that the same case if the capacitor in parallel with the supply line?

Either way, I already changed all the supplies (except the ones that absolutely had to be AC) with proper regulated supplies. It still has the same problems.

Could it possibly be a problem with the MPU communicating with the control boards? They all have common ground, but are on different power supplies. Maybe some sort of voltage offset voodoo? I remember testing the data lines relative to the case (and control board) ground, and they all peaked at 5V.

[DaveBonds]

I think so, because on a parallel circuit with one DC having a capacitor, it doesn't matter which direction it's in. Capacitors kill DC and allow AC to pass, for clean AC circuits. They don't work on a DC system, because of their construction.

Do you have a schematic or manual for this game? I'm sort of curious as to what the factory amperage and voltage readings are supposed to read, opposed to what they do now, both from age and modifications.

I know that an input signal changed slightly on a digital controller of any kind can stop it from working entirely, with insufficient voltage and amperage. Especially when the MPU relies on signals from SS devices/ control boards. If that MPU doesn't get a proper voltage signal from something, the MPU is flashed to retrieve signals from the controllers, in order to perform specific functions. I don't know what order the game is supposed to retrieve signals from which controllers on your game, but it can't continue it's program operations without all of them.

Most of the time, with the shaky componentry on these games, additional voltage draw is usually the cause for things like dimming lights. Cleaning up power supplies is always a good thing, when the same voltage and amperage is retained, but it has to be exact.

If the game is having trouble with dimming lights when a flipper solenoid coil is activated, chances are, the flipper coil ram, fiber arm, coil sleeve, connecting pin, coil ram stop and return spring need attention and would likely benefit from being rebuilt. All you would need to cause a light dimming problem is added voltage draw. And all you would need to turn that solenoid into an energy gobbling alarm clock is a different spring. The spring tension, distance that the solenoid travels (governed by the mechanical stop), how straight the solenoid travels and how accurately the entire device operates will have an effect on electrical draw.

These machines didn't have much of a problem when they were new, because the electromechanics were solid. Over time, the power supplies seem inadequate, because contacts become fouled, the components begin drawing more current as they wear out and cause brownout situations, due to higher resistance in some ciruits and higher draw.

I restore cars for a living and mess with these things on the side and I can tell you from my experiences, that no amount of correcting power supplies will do as much as correcting the rest of the circuitry in a system. I've seen things like a fuse that read good on a continuity check, correct weak energy throughout the entire car's system when it was replaced with a new one.

Regulating voltage won't do any good if the system doesn't always reach the regulated level and isn't always seeing that level of voltage under use.

I'll bet if you rebuild those flippers, remove that capacitor, check (and correct if necessary) all of your voltage and amperage to stock specifications, you'll be surprised at how well the system works, even with a dirty power source.

[jerzmacow]

So just a sort of update for this if anyone runs into the same problem I did:

It turns out these old machines run primarily off of voodoo magic. Of course by voodoo magic, I mean SCRs. I never would have guessed it, but SCRs need to be taken to ground in order to turn off. From what I understand, they can only be turned on (to turn them off you stop all current flowing through them). That's why the lights get stuck on with the switching power supply. With the barely filtered AC supply, it would go from ground to VCC at ~120hz (as opposed to the one I put in that's at VCC the entire time). When the original power supply would go to ground, that would give the SCRs time to shut off. As I said, complete witchcraft!

Right now, these are my solutions:
1. revert back to the old power supply
B. attach a DC chopper to the switching supply (this kinda works, but the light flicker a lot
III. make a whole new board using MOSFETS and be done with this upsetting 70s tech