Random switch triggering

[ignotus]

Ok, so I did some more testing. There was a strong chance that, as it's a hand-made board, something could have got disconnected/shorted/broken/whatever, but when it was last in use, it worked fine.
So I loaded firmware 20140809 into it and configured it with an old version of MDM. The problem has disappeared entirely. It's been on for the past 20 minutes without a single false trigger, so I suspect some kind of bug has crept in since then. I also suspect that there aren't many atmega644/1284 users left, and the ones there are probably haven't updated their firmware in a while. So unless anyone else complains about the problem, I'm fine using this older firmware since it does what I need.

[Kabonfaiba]

My ears perked up at the title of this thread.

I'm not on atmega board, but you've given me an idea to revert to a previous STM32b firmware. (The latest one that supported the old MDM prehaps) I hope I can do this without risk of bricking.

[airflamesred]

I think the atmega and the arm versions should be considered two different beasts.
I'll try to dig out my 1284, but I don't remember any problems.

[Kabonfaiba]

Well I downgraded to 20170503 but the switch behaviour remained the same.

Putting the switch in piezo mode is interesting. It seems to me from playing with the gain that I'm not getting a very strong open reading. But contrary to that, I don't get a strong open reading from my hi-hat either, and that doesn't have any of the issues with self triggering.

[ignotus]

I'd guess that my issue is related to the board being atmega-based and something happening in the firmware since 2014 that affects it. It wasn't occasional, random triggers but a constant stream of them when all inputs were enabled - it would make a drumkit completely unplayable. Then again, I find it strange that no one else has reported it - I mean there must be someone out there who has an atmega-based MD who updates their firmware. So it might also just be my board that has some weird idiosyncrasy...

[ignotus]

Hi again,

I hooked it up again to do more testing and the problem is still there... The random triggers are few and far between but it still poses a problem for what I want it for. I've been trying to troubleshoot the board swapping out components, testing continuity, etc but to no avail. Could the fact that there is no load (piezos) on any of the inputs somehow induce this problem? It's the only thing I can think of at this point...

[dmitri]

Do you disable those inputs which have nothing connected?

[ignotus]

If I disable the bow inputs, the edge inputs won't send program change or aftertouch messages. Is this supposed to happen?

[dmitri]

I don't remember it now without testing. Instead of disabling the piezo input you can try to shorten it to the ground.

[ignotus]

Cheers, I'll be gone a few days but I'll try it when I get a chance.