Repro Flash playfields are now available at CPR:
https://classicplayfields.com/shop/pinball-playfields/flash-3/

Got a picture of how you've wired the NMP board?

Did you do any work on the game between the first night it worked and when it later started misbehaving?

The AC legs on the bridges need to be measured with respect to each other, not respect to ground.
So with your meter set to AC voltage, one meter probe goes on one AC leg, the other meter probe goes on the other AC leg. That will give you the AC voltage coming out of the transformer and into the bridge for that circuit.

I'm not a Williams guy so can't comment without looking at the schematics. But I've gotta head out right now so it won't be til later.

Definitely..

If you have a short on the playfield coil supply wire then the solenoid fuse should be blowing. What fuse rating are you using?

Since it's all your special solenoids that are locking on, I'm guessing there is an error in how their activation switches are wired. They have a resistor and capacitor to create a short timed activation.

Please post some clear pictures (from different angles) of one of the pop bumper switch stacks.

While not being the problem, IMO those resistors are too small in size/wattage. There is a fair amount of power going through them. Also the capacitors should've been higher voltage.

The switch stack wiring looks ok to me comparing to similar manuals. How is it that with 19k machines manufactured there's no scanned copy of the Flash schematics?

Well.. The 23-900 coils have a DC resistance of about 4.4 ohms. They're connected together on the power side via the daisy chained red wire.
When you check continuity between the control (non-red) wires on the coils, you are effectively measuring two coils in series via the red power wire. With your meter in continuity mode, it will give the impression that there's a short circuit.

Put your multi-meter in resistance mode; if you measure about 8.8 ohms then you're good.

Note it's the same story with your classic Bally/Sterns except their coils are a little higher in resistance, typically (but not all) 12 ohms each.

Coil_Resistance_DaisyChained_test.png

The reading is from any coil to any coil (probably except flippers). The daisy chain means they are all hanging from the same line.
Different coils will have different resistance. See here for reference:
https://flippers.com/coil-resistance.html

Seems to me your readings indicate a short at the right pop bumper because every measurement is approx 4.4 ohms less than it should be.
And also the bottom pop bumper indicates a short having a near zero resistance reading across the two coils.

To be better clued in, do the same measurements, this time using the saucer coil as the reference.

BTW, the large blue flash lamp insert is also wired to solenoid power so there is another potential source of the issue. Did you make any repairs/mods to it?

I'm not reading much out of the resistance checks. Besides, if any coil is measuring shorted it shouldn't be locking on at power-up. Something else is going on.

On the previous page vec-tor suggested to disconnect both 2J12 (special solenoid coil drives) and 2J13 (special solenoid switch triggers) from the driver board. When you did this you mentioned the bottom and left pop bumpers still locked on at power up. Please confirm I read your response correctly.
With 2J12 disconnected, you are removing the driver board controlled ground paths to the special solenoids so they should have no way to be energised. If they are you'll need to follow the coils control wires into the backbox to find out if they're connected to the wrong location.

The small resistor is originally 1 ohm, 1/2 watt.
The large resistor is 330 ohms, 2 watts.

If the 1 ohm resistor measures open circuit then those flash insert lamps will not illuminate.

If you're using LEDs (presume you have 12V LEDs there) you need to disconnect the 330 ohm resistor. Search warming resistor in this thread for details.

Yes, should be zero volts. Your 4 volts or so may just be noise. But the bottom and right pop bumper coils being different are showing a short to ground like Billc479 said.

The lug board where the resistors are for the flash lamps has two black wires connected to one side of the big burnt resistor. These black wires are ground on the playfield.
With the game off and your multi-meter in resistance mode measure the resistance between the black wire ground and the control wire at the right pop bumper coil and then the bottom pop bumper coil. Compare the resistance to say one of the slingshot coils.

Ultimately though you need to follow the control wires for the bottom and right pop bumper coils to see where they are getting grounded from.

What happens when you disconnect both (black and white) 24 pin block connectors between the playfield and backbox?
Do you still get this low 1.1 ohm reading across the red and black wires at the flash lamps on the playfield?
Do you still get the low (0.4 - 0.8 ohm) readings between the special coil control wires and black ground wire on the playfield?

In other words is the problem on the playfield or upstream in the backbox area?

This disconnects the red 28V solenoid power from the playfield.

With that black 24 pin "8J3" connector disconnected, you are totally isolating all the playfield coils from the backbox.
How much resistance do you measure across each playfield special coil ? i.e. are you now measuring around 4.4 ohms?

BTW just to confirm you're aware, when your multi meter measures in the kilo-ohms or mega-ohms it will show the letter K or M respectively on the display.

Your 3J3 connector at the power board looks correct to me according to the various schematics of the era.
The power board has a 100uF capacitor between ground and 28V solenoid power. The moment you place your multi-meter in resistance mode across these pins the meter will instantly show a short circuit, but the resistance then rises quickly to almost infinity as the capacitor charges.

Disconnect 3J3 from the power board. With machine OFF, measure the resistance between pin 3 (ground) and pin 6 (28V solenoid power) at the power-supply board 3J3 pin header.

Next disconnect 3J1 block connector that comes from the bridge/transformer off from the power board.
Measure the resistance across the - lug and + lug of the solenoid bridge rectifier screwed to the side of the headbox.

Ok, check the driver board for some old fashioned shorted driver transistors.
With 2J12 and 2J13 disconnected from the driver board, measure the resistance between ground and pins 3, 4, 6, 7, 8, 9 of the 2J12 pin header which goes to transistors Q6, Q4, Q8, Q2, Q10, Q12 respectively.
Readings should be very high resistance.

Infinite resistance means the special coil driver transistor outputs are not shorted to ground. For now, step away from the driver board.

When you have the black 8J3/8P3 intermediate 24 pin block connector hooked up you appear to be getting ground shorts on two of those pop bumper control wires.

The below playfield solenoid schematic is from Firepower but should be the same general principal as Flash.
The special solenoid coil control wires go direct to the intermediate 8J3 black 24 pin block connector then onto the 2P12 connector at the driver board. They don't go anywhere else per the schematic.
When you have both 2P12 and 2P13 disconnected from the driver board you eliminate any potential connection to ground on those control wires.

You essentially need to follow say the control wire from the locking bottom pop bumper coil through the harness and find out where it's going after the 8P3 connector. It should go to 2P12. Is the wire pinched somewhere in the harness, is it going from 8P3 to somewhere different other than 2P12, etc.
Ditto for the other special coil control wire that's locking on.

Solenoids.png

Great to hear you worked it out!

For the switches, go into switch test mode and see if any are reported as closed.

Can you do us a favor? In your last post at #930, Pinside scaled down the Flash playfield diagrams you added which causes some loss in quality. If you hover over the images you'll notice at the bottom where the filename comes up they say (resized).

Can you please edit that post, delete those diagrams, change the image option to 'Original size (no scaling)' and re-upload them so we have the loss-least quality you have?

Pinside_NoScaling.png
PinSide_Delete_Picture.png

Much appreciated!

I think you need to perform the following mod on the driver board bringing it up to system 7 standard which increases the current drive on the switch strobe lines when multiple switch closures occur:
https://www.pinwiki.com/wiki/index.php/Williams_System_3_-_7#Wire_Jumpers_on_System_7_Driver_Boards

That would suggest there is a ground loop issue. i.e. the sound board is getting ground from two separate sources with different levels of current.
In your pictures I see the cabinet ground braid behind the metal bracket mounts for the sound board. I would think that is the offending ground source rather than J1 pin 5. J1 pin 5 is the desired reference for zero volts from the centre-tap of the +/- 12V transformer output.
i.e. try isolating the cabinet braid from the sound board brackets instead.

Hmm, there are at least two problems here. The transformer taps to power the sound board amplifier are also used to power the digital logic (i.e. MPU) board. It is a dubious design to mix digital supply that is very noisy with analog supply and like vec-tor said they should be isolated. Secondly the sound board is located far from the transformer so the cables are likely picking up noise in the harness run.

If disconnecting J1 pin 5 from the sound board mostly eliminates the problem then I would just leave it at that. The cabinet ground braid is otherwise providing the transformer centre tap ground to the sound board because it's also wired to the power board ground from the transformer.