I just had one kick me!

Just spent over 2 hours trying to isolate a problem with a machine running erratically. After searching the entire program (without descriptions, of course) I finally realized that it was a hardware problem. Seems whenever output 4 is turned on on the SLC TRIAC card output 12 comes on with it. Output 4 does not come on with 12 on, just 12 with 4, and yes, the LED for 12 on the card lights when it turns on incorrectly.

And, just for a bonus, the LED for output 7 does not light, but the output works.

Congrats on finding the problem! Some of them are real Doozies, aren't they? Sounds like your output module might have suffered damage from a voltage spike.

I encountered a similar problem once. Outputs were firing randomly, in spite of said outputs being forced off in the processor. I finally tracked the problem down to a film of carbon on the IO modules. The AC had failed in the MCC and the AC unit was temporarily replaced with a fan to cool the room. This MCC building was next to a group of product loadout bins, and the bin adjacent to the MCC contained reject material that was regularly being hauled to a stockpile with an old, non-street legal dump truck that was spewing diesel soot into the temporary cooling fan. I had to wash the modules with solvent and dry them with a blow dryer to get the system back into operation.

You do not give a card number so all this is guess work

This may be worth a check, look and see if that output card
should have a neutral connection on it, and then check
if that connection is present and connected.

I have seen the situation where the output card was eratic
and got worse as more outputs turned on because
the snubbing network on the card relies on a neutral being
connected in order for it to do its job correctly.

+1
What Gil47 said.
I've seen it on a ML1400

I do not know what sort of environment this system was installed in but I work in a food processing plant and this behaviour is typical of a water ingress problem. Even fairly mild condensation on a PLC backplane can cause this type of thing.

I kept the card for a souvenir. It is a 1746-OA16, and no, it does not have a neutral terminal on the card. It was in a dry machine shop, no wet or dirty atmosphere. It does this with only output 4 on, or many outputs (other than 12) on. 12 always comes on with 4, but not 4 with only 12.

And I have put this in a test rack and the problem is consistent on the card, so it was not a backplane or CPU condition. So this has to be a logic problem on the cards interface itself as the TRIACs turn on and off when energized by the card.