The schematic shows that the top center rollover has two switches below it. One always fires the C/500 point (white drop targets) relay and the other sometimes fires the E/3000 point relay. The path to the E relay has two other switches in series after the top center rollover switch as shown in your schematic above. One Normally Closed (NC) switch at Score Motor 4E and another NC switch at Score Motor 3E (assuming 5 ball play).

I think the top score motor cam has just one long pin rising from it that reaches the E level switch stacks. Since there is just one long pin it can come to rest at any one of the three score motor index or home positions. So roughly 1/3 of the time the score motor opens the 4E switch when it stops, and another 1/3 of the time it opens the 3E switch when it stops. That means that the remaining 1/3 of the time the center top rollover switch should fire the E relay (because both score motor switches are closed) and score 3000 points instead of 500.

The way that's done is shown here from a bad copy of the schematic:
Atlantis C E relays (resized).jpg
When the top center rollover switches close the C relay fires which starts the score motor turning. The C relay also flips the Make/Break switch in the left red box. About 2/3 of the time the E relay does not fire so the five pulses generated by the score motor 1A switch follow the top path in this schematic to the M/100 point relay which awards 500 points. About 1/3 of the time the E relay also fires which flips the Make/Break switch in the right red box which in turn sends the five pulses from the score motor 1A switch towards the L/1000 point relay instead. Two of the five pulses from the 1A switch are blocked by the score motor 4C (?) and 1B (?) switches so only 3000 points are scored instead of 5000.

If your top center rollover is getting 3000 points most of the time I'd start by checking that the E relay fires reliably when score motor switches 4C and 1B are both closed and the Make/Break switch in the red box on the E relay.

/Mark

What about that switch? Many of your symptoms could be explained by that Make/Break switch not working properly.

Which switches?

Try blocking the bottom path in the schematic from reply #2 by putting a folded piece of paper between the contacts of either of the normally closed score motor (4C or 1B) switches. That should disable the 3000 point circuit by blocking any pulses from the score motor 1A switch from reaching the L/1000 point relay. Then see if you ever get an unwanted 3000 points instead of the expected 500 points. (Note that you won't get expected 3000 point awards either.)

If blocking that path prevents the error from happening that would indicate that the error is going through that path. I still suspect the Make/Break switch on the E relay but more experimentation should help clarify.

It sounds like the E relay always fires when you get the advance and 3000 points, is that right? That would imply that the E relay switches are doing what they should be doing, and that perhaps the E relay is firing when it shouldn't.

You show the schematic for the E relay in your original note. I noticed that there is an identical circuit that controls the lights:
Atlantis rollover lights (resized).jpg
These two circuits should behave identically, but I wonder if they differ when you see the problem.

The score motor comes to rest in one of three index positions. Because of the long pin on the upper cam described in reply #2 each of those three positions will put the switches in the circuit above in a different state:
- pin not against either 3E or 4E switch stacks - the circuit would be as drawn in the schematic
- pin against the 3E switch stack would flip the score motor 3E Make/Break switch
- pin against the 4E switch stack would flip the score motor 4E Make Break switch

I wonder if you always see the error when the score motor is in the same position. If you do it might point to a 3E or 4E switch as a problem.

Since the circuit you posted and the one I posted are the same but use different switches, it could be that they're getting out of sync. For example if the make/break switch at score motor 3E used by the E relay circuit, and the one used by the lamp circuit are somehow in different positions (like maybe a switch stack spacer is missing) then the lamps circuit might light the left rollover while the E relay circuit would award the center rollover.

So three things to check:
- Does the error always happen with the score motor in the same position?
- Do the make/break switches on the score motor 3E switch stack behave identically as the long pin passes by?
- Do the make/break switches on the score motor 4E switch stack behave identically as the long pin passes by?
(You might want to shut off the power and turn the motor by hand to see.)

I thought you swapped playfields and the problem didn't change. If so, then the target switch isn't likely to be the problem.

Did you ever check the things I suggested in reply #11? It might be important to figure out if the Score Motor position affects the problem. If the problem only happens in one of the three Score Motor positions that might help narrow it down.

There are three Score Motor positions:
1) with the 3E and 4E Score Motor Make/Break switches inactive/as drawn in the schematic
2) with the 4E Score Motor Make/Break switch activated/flipped
3) with the 3E Score Motor Make/Break switch activated/flipped

Check to see if each of the values inside the two black boxes matches what your game is doing.
pasted_image (resized).png

Assuming that the lights all match what's in the chart I think your problem is that part of the 3E Make/Break switch isn't opening reliably when the long Score Motor pin pushes in on the 3E switch stack:
Atlantis E relay 2 (resized).jpg
If the normally closed switch in the red box fails to open you'll see behavior that matches the table the way you filled it out.

Since the behavior is intermittent you'll want to examine that switch carefully, looking at the solder tabs, insulators, maybe even the screws. Somehow there's periodically a path through that open switch that shouldn't be there.

If you unplug your game you might be able to find the problem with your meter. Clip the probes onto the blue-white-red and slate-black wires on the 3E switch and put your meter on the lowest resistance setting. When the 3E stack isn't activated you should see resistance of an ohm or less. Then rotate the score motor by hand. When the pin activates the 3E stack the resistance should change to something much higher, probably even open circuit. Try that a few times.

If you can find a spot where the 3E stack is active but the resistance is still low, leave the motor there and poke around the switch to see if you can make the resistance jump up to the high value. That might give you a hint where the intermittent short is.

Continuity isn't a reliable measurement in many EM circuits. Did you get an ohm of resistance or less?

I suppose the problem could be that there is an intermittent short between the blue-white-red wire and the slate-black wire that doesn't involve the 3E switch. The short would only be apparent when the 3E switch is active or flipped.

To test that theory you could desolder the slate-black wire from the Score Motor 3E switch. With the wire removed the two positions you marked as incorrect in the table in reply #14 should always score 500 points. If they ever score 3000 points, the intermittent short is not in the switch but somewhere else.

Note that a couple other positions in the table will fail to score 3000 points with the wire removed, but it's really only the two positions that were getting 3000 points when they shouldn't that are interesting.

That's more evidence that the path with the intermittent short goes through the normally closed side of that 3E make/break switch.

That too points to the 3E switch.

I don't follow this. The center target sometimes scores 2100 when the lower target is lit (not hit)?

That's fine. Putting the meter on the lowest resistance range makes it more sensitive to small values. The problem is that meters can beep for continuity when the resistance is anything up to 20-30 ohms. So what can happen is that the meter beeps because it sees a path through a coil or the transformer (at maybe 10-15 ohms) and the user assumes that it's beeping because the switch is closed (which should be less than an ohm). Without checking the resistance you can't know which path the meter is beeping about.

When the lower/right target is lit, does the center rollover show the same problem as the center target? They're in the same circuit except for an extra switch on the P relay for the target compared to the rollover.

Either way, it sounds like the game starts scoring 500 points (C or D relay fires) and after the first 100 point pulse the E relay fires late to change the 500 point award into a 3000 point award. Can you verify that the E relay fires later than the C or D relay in these cases?