It's decided then. It's definitely a hand wire.
1N4148 Diodes are on order, along with some wire for the columns.
I've got the Teensy 2.0 already so we are all set to go.
Really appreciate the offer of help with the firmware, that will be the difficult bit to get right I think.
That wire layout is going to look superb!
Documenting this in pictures in case it helps anyone else.
The PCB is now de-soldered and removed.
All the switches out and I've started cleaning them in a bath of isopropyl alcohol to get any dust and rubbish and the coffee out of each of the key switches.
This is a USA built board, so has the SKCC alps rather than the Matsumi versions.
Once they are desoldered, getting them out is super easy. They can be squeezed with fingers or tweezers to compress the clips any they pop right out.
These are the clips that hold the switch in the plate.
IMG_8582.HEIC (2490.3 kB - downloaded 64 times.)
Dismantling for cleaning is easy too. Gently insert something to tease the side clip up so it clears the black edge and the spring will push it up enough to hold it while you do the other side.
IMG_8585.HEIC (2390.3 kB - downloaded 60 times.)
Then you can wiggle the two apart to reveal the glorious switch, slider and spring.
IMG_8588.HEIC (1567.5 kB - downloaded 62 times.)
The slider and spring just fall out, but the switch plate will be retained by the pins going through the plastic casing. My top tip is to make sure you have cleaned all the solder off the pins, then I pushed them from the bottom with the flat edge of the tweezers. This is better then trying to pull them put as the leaf switch can be easily bent.
Switches going for an alcohol swim
Then it out to dry on some paper. Then I put a tiny amount of GPL 205gO lube on the spring. I've ready on the forums that a bit of lube here can reduce the spring ping. And it does. Not completely, its still there, but enough so (for me) it adds a wonderful high note to the thock without resonating on for a while afterwards. Then reassembly and testing.
I just did this with a voltmeter, looking for full resistance when open and no resistance when closed.
There were a couple of switches that I had to 'tweak' to get them to register properly.
I found that a TINY bend here helped if the resistance didn't change. (bend to move the tines away from the switch plate)
Or a GENTLE push of the tweezers in here if they were always open, coupled with a TINY bend in the other direction on the tines of the leaf. (as above but bent towards the switch plate)
IMG_8622.HEIC (1387.2 kB - downloaded 62 times.)
This can take a few attempts to get right and you have to put the switch back together fully to test.
Thankfully there aren't may that need this extra care.
Being far to keen I've started putting them back in the case (cleaned with water and mild soap, then wiped with alcohol on a microfibre cloth) and putting the keycaps back on to see how it sounds.
Looks and sounds beautiful so far.
There is one broken one where the side wall has snapped at some point, Maybe I did something wrong getting it out. But with the age of the plastic, I'm surprised how flexible it still is so I can't see how I did that.
The switch still seems to be working, just has one bit missing.
I may try and glue it back on, but the switch seem to work perfectly well with only one side holding it together.
[ Specified attachment is not available ]
Will keep working on the options with that.