Double-firing switches

[erw]

I have a Kinesis Advantage with MX browns. After a flight (with the board in my checked-in baggage  ) it started acting weird:  Keys sometimes double-fire (or double-trigger or double-register) key presses. I don't know if there is an established term for this, but I didn't find anything while searching for switches double-firing. Apologies if it is already answered elsewhere.

I have considerably increased my OS delay for repeating key presses when the key is held down and it still happens maybe once a sentence and only with this board. Especially the space key likes producing two spaces, but also letters mmay get typed twice. My keyboard being a Kinesis, it could be a firmware issue rather than hardware. But has any of you experienced something like this?

[boost]

Are the same keys registering twice all the time or is it random keys?

[damorgue]

Does it happen to all switches? If so, then it is most likely not a mechanical thing as all the switches would be required to have the same problems which is unlikely.

I'd guess something wrong with whatever measures they have put in place against debouncing. Could the part of the code that handles the debouncing algorithm have been corrupted and leaving everything else working?

[dorkvader]

At a guess, it could be a loose connection internally. Have you opened it up to look?

[tp4tissue]

This certainly gets rid of the double tap complaint people have.

[erw]

It doesn't happen equally to all keys. Space is by far the most common, then D and E (G and K on QWERTY, resp.). But it also happens to F12 which is a rubber key  (well, probably also the other F-keys but I use F12 a lot so that's where I notice it).

So it appears to be somewhere in between one mechanically faulty key and faulty firmware with all keys misbehaving.

[krazyderek]

this started happening to my kinesis contour classic after a couple years of use, i've been searching for a solution for a while since it's outside the warrantee (grrrr short warrantee's on expensive keyboards)

I have considered getting a teensy board to test the keys to see if it's a firmware issue related to debouncing as discussed here:
http://michael.stapelberg.de/Artikel/kinesis_custom_controller/

[CPTBadAss]

Could be something called chattering. I've solved chatter in the past by making sure the switches don't touch the back of the case by putting drawer liner between the PCB and case. Or swapped out the switches. Or fixed the solder joint. I'd try all that before swapping controllers.

[Findecanor]

It sounds like some connector inside the keyboard would have got unseated when you travelled with it, but it is weird that it is happening to keys in different key groups.

[krazyderek]

It doesn't happen equally to all keys. Space is by far the most common, then ... So it appears to be somewhere in between one mechanically faulty key and faulty firmware with all keys misbehaving.

Did you first notice this on your advantage with browns? Has it happened on your advantage with reds?

Could be something called chattering. I've solved chatter in the past by making sure the switches don't touch the back of the case by putting drawer liner between the PCB and case.

I cleaned it this weekend, i'm pretty sure the contour has a healthy amount of room between the pcb and case, but i did remove any hairs that found their way back there, cleaned the front of the plastic plate, and washed the keys (they just needed it), not really a big difference.

Or swapped out the switches. Or fixed the solder joint. I'd try all that before swapping controllers.

Do you mean literally just heat up the solder, and add a drop as required to "fix" the solder joint? There doesn't appear to be any calcium build up on any of the solders.

I might try replacing one or two switch on the "worst" keys since that would indeed be simpler then going on a new controller adventure.

The repeating might be because i'm not fully bottoming out the keys all the time, and the signal might be oscillating between actuate/release thus registering multiple key presses. But i would think that the firmware would have some type of minimum time between keypresses / signal threshold difference to register a new keypress. There is a built in repeat rate in the keyboard you can program, but like erw, i jacked it up when the problem first started happening with little help.