This is what I found:

"This is what happens to all system 11B and 11C games (with the Aux driver board) when you install LED flashlamps into solenoid spots 3,4, 6 and 7.
There are tie-back diodes on the large TIP36 transistors that connect those circuits to the 50V power source. There is a some feedback into these circuits from the flippers that is not apparent until you install LED's.
To fix this, Cut diodes D3, D4, D6 and D7 from the aux driver board. Then add diodes onto the coils for solenoids 3,4,6 and 7 (banded end of the diode goes to the power wire)"

"I believe any of the 1N400X will work. I usually use the 1n4007's"

https://pinside.com/pinball/forum/topic/ww-tech-help-flippers-trigger-flashers

Seems pretty straight forward.

I tried that on my Whirlwind, but I still had errant flashes. I ended up just putting incandescent flashers back in.

20150905_071756_(resized).jpg

Did you remove the warming resistors as well?

As long as the warming resistors remain in place you will have errant "flashes". Step one in installing LED flashers is to remove the warming resistors from the lamp boards.

No, but from what I understand, games made after Big Guns do not have these.

The incandescent bulbs when in the socket are less than 5 ohms and the circuit requires this resistance to be complete. LED's alone have too high of an impedance for the circuit to perform as designed. All you need to do is put a 1k or 100 ohm resistor across the socket terminals when using the LED bulb. That way the circuit is complete and only partial current is needed / shunt thru the LED (a diode). Increasing the resistance across the terminal will result in a higher LED brightness... but 1k is plenty

Of course, if you have warming resistors, remove them...

Derp, sorry brain malfunction. You are correct. Thus the diodes....

I see you cut one end of the diode, but it looks like in the pic you re-attached the cut end to the resistor?

What's wrong with incandescent flashers?

I recently put LED flashers in a BK2K. They work as long as there are only a few leds in the back glass with the rest incandescent. Try adding a few incandescents till they stop flashing. They will eat up the extra power surge.

They're actually new diodes, but you could also move them if the leads were long enough. Basically, they are not placed correctly to work as flyback diodes. Here's a marked up schematic to show what I moved.

diode_(resized).jpg

Be aware, though, this did not fix the errant flashing on my game.

Great, to hear. I have to do this to my SoF. I placed an order from Great Planes last week and am still waiting for the bits. I am glad it worked out for you. I wonder why Herg is still having issues on his.

Good work and thank you for documenting it.

Just as a fwiw, I used full flashers in a Bugs Bunny Birthday Ball (sys 11c) with no issues, no changes, no cuts. It may be a 'system 11c only if you have the problem' type of fix.

Sounds like your just moving the back emf snubber diodes closer to the source(the coil) instead of having it far away on the board. Do the ones on the board even have to come off??

Now that you mention it, no, they shouldn't have to.

I just ran into this with my Rollergames. Can you explain how just putting diodes on the A side coils would solve the problem? I having trouble understanding how it would if it's the flippers causing the issue. I'd rather not modify any boards if I don't have to, but I could add some diodes to coils if needed.

Problem solved!

61cF0F83bCL._SY355__(resized).jpg

This problem intrigued me, so I decided to take a look at the circuit some more to see what is actually going on. I suspected that D3, D4, D6 and D7 where conducting and causing the transistors to activate momentarily. It appears that the game keeps the A/C select relay energized for most of the game unless it needs to fire an A side coil. Therefore, most of the time, you will have 25VDC on the anode side of the diodes, such as D7. Because, W3 is populated, you will also always have 50VDC on the cathode side of the diodes. Normally, when you don't flip, the diode is reverse-biased and it does not conduct. When I connected a scope to check what was going on, it was just as I suspected... Flipping caused the 50VDC to drop below 25VDC forward-biasing the diodes. They conducted, and in my case, I saw the 50VDC drop to 0VDC.

So, herg, the reason it didn't work for you, was because moving them to the other side of the resistors will basically do the same thing and activate the transistor. You would have to cut them out and put diodes on your coils to make it work.

But, I may have a better solution...
I checked the ripple on my 25VDC and 50VDC lines and I'm getting about 1V - 4V of ripple! I'm thinking that my caps need to be replaced. It's possible that once they are replaced, the 50VDC line will no longer drop below 25VDC and the diodes will not conduct solving the problem.

Now, I'm no electrical engineer, but I do like this stuff, so please feel fee to comment away... tell me that I'm on to something or that I don't know what I'm talking about... I like learning from my mistakes.

PS: Anyone know if there is a System 11 schematic manual like Williams did for WPC?

aux_driver_(resized).png

Just kidding, but I was thinking that the cap on the flipper board may also need replacing. You should give this a try and let us know the results.

We think alike... The game I'm working with is a Rollergames (System 11C) which doesn't have a flipper board. They moved those to the Aux Driver board which is the one that I'm starting with. I'm about to place an order for some caps. I'll post back on how it goes.

Yea I miss spoke on that, but was meaning that C1, C3 need to also be replaced when you change C5 and C6. Should be very interesting results. Maybe a small improvement in flipper power?

I hope so. I've also been having what could be a flipper power problem. Should be an easy fix... only 4 caps on this board.

I finally got my head wrapped around this, and you are right.

I had myself convinced that the diode on the board was the same as on the coil, and it would be except the power lug of the coil is disconnected from 50V when the C side is selected. You have to remove the diode to remove the path from 25V, through the flasher, through the diode, to the sagging 50V. Then, you have to add the diode back to the coil to snub the voltage spikes when those are turned off.

I hate to pull the diodes off the board, however, since if anyone is working on the game down the road and looks up info that says the coils in these games don't need diodes, they might end up pulling them off of there. I'm going to look into connecting them to the A side 50V after the relay. It might end up being too much of a hack, though.

As for keeping the 50V above 25V, I suspect you may have a tough time. Those caps are a very small value relative to the current that is drawn when the coils fire.

I was thinking this too, but since I have so much ripple, I think it's best to replace those caps anyway. Looking at the soldering on the bottom of the board, they look original, so it's time to replace. I think I'll eventually do the main power board also, but interesting thing is that my 5V line was rock solid, so I'm holding off on that board for now.

5V is regulated, so as long as it's input stays high enough, you won't see any ripple on it.

Do ALL system 11s need modifications done to use LED flashers?

According to what was said above, it looks like just the ones that were made prior to big guns. And even then, only if you are actually experiencing the described symptoms.

Well, now that you say that... it's obvious.

I believe that Rollergames (my game) was after Big Guns, so I think there is a chance that you might see this problem with later System 11s that use this Aux Driver board configuration.

I'm going to order some flashers from comet today for my mousin around. I'll report back what my findings are.

BTW: if I do indeed have issues, what would they be? Flashers just not working? or going off at bizarre times?

Flickering (not a full flash) at "bizarre" times. More specifically, when coils fire.

What I notice is that when either flipper button is pressed to energize the flipper, there is a momentary pulse/flash of the flashers 3C, 4C, 6C and 7C. The actual count of flashers depends on the game. It's always 3C, 4C, 6C and 7C in both my Rollergames and Black Knight 2000. I saw this in my Whirlwind when I first put LEDs in that and didn't know why. I put incandescents back in those flashers.

So it sounds like the solution (other than putting in incandescents) is:

Remove D3, D4, D6 and D7 from the auxiliary driver board (or at least take it out of circuit - remove one leg).
Install coil diodes across 3A, 4A, 6A and 7A. There's no easy way to do this on the auxiliary driver board without it looking like a hack.

Can someone with much more electrical knowledge than me confirm that removing D3, D4, D6 and D7 *IS* required? I tried putting a single coil diode across a coil without removing the board diodes but it didn't help. With my limited understanding of the above discussion this seems to make sense (i.e. must remove the diodes off the board).

Bump for more info.

Yes, the way the circuit is, it would be required. There is a way to make it work on the board, but it would look like a hack.
You can wait and see if replacing the caps on my board works and then try that instead. If it works it will not be a hack and will be much cleaner. I'll post back once I get the caps and try it. I expect to have time to do it this weekend.

--Steve

Following this thread with interest. I have just retro-fitted flashers to the insert board of my Rollergames and fitted LED's. They are basically lighting up like a Christmas tree when they should be off. Putting the odd incandescent back in helps but I will definitely be doing the moving of the diodes off the aux driver board and onto the 50v coils.

Strangely, different LEDs exhibit different ghosting behaviour.

I'm not surprised. Not only can LED voltage & current requirements vary greatly between manufacturers, did you know they can also vary depending on color?

201407030656022303_(resized).jpg

Well, the results are in... I've replaced the 4 capacitors on the Aux driver board, but I don't see much change. Is it possible that Williams designed a board with this much ripple?

Here are some scope images of the waveforms...
50VDC Before
50VDC Before

50VDC After
50VDC After

25VDC Before
25VDC Before

25VDC After
25VDC After

The new caps reduced the ripple by a small amount, but nothing significant. And, if you look at the waveform below of when I hold a flipper button, you can see that there is about 33V of ripple on the 50VDC line!

Do you guys think this is normal for a System 11 game?

50VDC While Holding Flipper
50VDC While Holding Flipper

Do you think the bridge retifier is on the way out?

I'd say that looks about like what I'd expect. From what I remember, you only have about 200uF on the 50V. By comparison, I put 20000uF on the GI OCD, and there is still a bit of ripple on it at high load. GI lighting is a small load compared to flipper coils.

I don't think the bridge is out because the wave shape looks correct, it's just that there isn't enough capacitance.
I guess I'll just install a few incandescent bulbs to get rid of the blink. (I'd prefer not to modify my board.)
It was a cool experiment though.
I hope the information gathered was useful.

Sorry. More stupid questions. I don't understand electronics too well.

Is this situation like the controlled lamp matrix "ghosting" phenomenon? I understand that is solved by putting a capacitor on the LED to absorb the stray voltage/current.

Can a similar thing be done here with the flashers?

If so ... what kind of capacitor and what voltage and Farad rating?

Is it a one rating capacitor size fits all regardless of the number of flashers on the circuit?

Or will it vary depending on the number of flashers?

Is it possible to just put it across the positive and negative wires that connect to the auxiliary power driver board as a (central) choke point rather than across each (peripheral) flasher socket?

No, I don't think adding capacitors in this way will work. If you want to have LED flashers in those 4 positions that cause blinking, then you would have to remove the 4 diodes from the aux power board (or one side of them) and then add diodes across the A-side coils. This should work.

In my case, I just don't like removing things from boards because I don't want someone in the future to wonder what I was doing.

Last night I cut off the non-banded side of D6 and put a diode across solenoid 6A. The flashes did indeed go away for flashers 6C. So this works if you're prepared to modify your auxiliary driver board.

Two questions arise. Again ... please forgive my electronics ignorance.

1) Can I attach a diode across 5J-11/12 to 5J4-4 and achieve the same result as the diode across the coil for solenoid 6A? If so I assume this can be done for 5J4-7/6/2 for solenoids 3A, 4A and 7A.

2) What is/was the way to make it work on the board that you mention in post #40?

1.) If when you said "5J-11/12" you actually meant "5J12-11/12", then yes, that should give you the same results.

2.) Yes, doing something like this was what I was thinking. Basically as long as the anode side of D6 is connected to the anode side of D39, it should work. (Same for the others.)

Hope this helps.