IBM Model M membrane troubles

[Krunchy]

Hello Geekhack!  I've been lurking and learning around these forums for months now, but finally I find myself in a pickle and I can't seem to find the answer.

After longing for an experience like the old Chicony I grew up with, I got my hands on a 1391401 Model M a bit over a year ago and experienced buckling spring nirvana for many months.  Unfortunately, my basement had a humidity problem and late Summer brought 90% humidity and several keys on the bottom two rows suddenly not working.

Recently, with more time on my hands, I opened and thoroughly cleaned the board.  I was distressed to find rust near the bottom edge of the steel plate, but it wan't too bad.  Cleaning didn't fix anything.  Then I decided to take the plunge and bolt modded it so I could check the innards.  I found that near the bottom of the membrane closest to the black plastic piece, the outermost lead that runs around the numpad and to the EXACT keys that aren't working had some gunk on it and was black in places (sorry, no pics of it at the time).  I carefully cleaned it with distilled water and a great deal of the black gunk came off although the lead wasn't pristine like all the others, but I figured it was enough.

Long story short, the keys feel better than ever after the mod, but those few keys still don't work.  I'm guessing there's a dead spot in that lead somewhere and I'll need to redraw the lead with a CAIG circuitwriter pen or something like that, but here's where I throw myself at the mercy of you geekhack vets.  How can I properly use an ammeter to find the dead spot, is the CAIG really my best bet at fixing the lead, and am I even diagnosing this correctly?  I'm going to start re-disassembly as soon as I finish typing so I can take pictures later tonight, but in the interim, any help to restore my poor Model M would be very appreciated.

Thanks in advance and I look forward to being more active on these boards in the future.

[Krunchy]

Okay, so it took me a day longer than expected, but I have pictures.  This first one is more of a wide shot of the back of the membrane sheet in question.  If you look along the left edge you can see a darker line on the lead and you can see noticeable darkness on the bottom lead.
[ Attachment Invalid Or Does Not Exist ] 37840[/ATTACH]

This second one is a closeup of that bottom section (rotated 90 degrees so I could zoom in better) and this is already AFTER I cleaned it with distilled water.  It was much worse before that.
[ Attachment Invalid Or Does Not Exist ] 37841[/ATTACH]

At this point all I can think to do is check on either side of those areas with the ammeter and then use the silver conductive pen.  Has anyone else had to deal with something similar to this?  Any insight would be appreciated.  I'd hate to go through all this only to have it turn out I solved a problem that didn't exist.

[bbbbqq]

Grab a multimeter and set it to continuity, and probe away. One end at the controller end on the appropriate pin-out, the other on each "pad". Should beep when it makes contact (good) and not beep when you get to the part with the broken contact.

Worst case scenario, call up unicomp, they should have the part to repair it.

[Krunchy]

Grab a multimeter and set it to continuity, and probe away. One end at the controller end on the appropriate pin-out, the other on each "pad". Should beep when it makes contact (good) and not beep when you get to the part with the broken contact.

Worst case scenario, call up unicomp, they should have the part to repair it.

I went and found the exact stretch that wasn't making a connection and redrew it with the pen.  Now I play the waiting game and hope once it dries it fixes the break.

It didn't occur to me to contact Unicomp in case this fails, though.  While I hope it doesn't come to that, many thanks for the idea.

[ch_123]

I have fixed an M's membrane with a conductive gel pen, it works just great now.

[edada]

I once had this kind of problem with a cheap membrane keyboard, some keys weren't working, except when shift pressed !

No rust here but lot of dust. Due to poor design there's around 400 square holes in those membranes. Cleaning them with water did the trick, but I don't understand what happened.

[Krunchy]

Well, I thought I'd give a final update of sorts here.

After waiting a day for my sloppy silver pen redraw of the broken lead to dry, I again tested with the ammeter.  The connection had definitely been re-established, but it showed noticeably higher resistance than the standard white leads.  I didn't know if that mattered, but there was one way to find out.  I reassembled the board (stopping halfway and starting over because I forgot to put the spacers back in) and popped all the keys back in.  Then, the moment of truth...

I carried the keyboard fresh out of surgery and plugged it back in to the computer.  The green lights flashed.  So far so good.  With bated breath and slightly sweaty palms, I opened a notepad window and slowly put my hands on the board.  I typed a simple word using keys I knew worked...success...then I gulped a bit.  I slowly reached my thumb for the space bar...the space bar that had months ago stopped working and had been the first sign of something ailing my monolith of a model M.  Not unlike the steps in firing a gun, I held my breath and pushed down until it clicked...

...how did it turn out?  Well, which board do you think I used to type this post?

I will add a couple of notes here, though: The residue on the clear part of the membrane did seem to indicate that even with the rubbery layer between the steel plate and the membranes, something was able to seep through.  I don't know if it was moisture or what, but I'm willing to bet that's what caused that lead to die.  

Also, to clean off most of the surface rust on the steel plate, I used a mix of baking soda and vinegar to make a watery paste, put it on the rust spots, let it sit for a few minutes, and then sanded it off with 300 grit sandpaper.  For the really light spots, a bit of WD-40 with that same sandpaper worked just fine.  After both of those, the surface was still marred in the spots where the worst rust was, but the rust had been cleaned off.

Many thanks to all for the ideas and insights.  I had never attempted something like this before and having even a bit of support helped a lot.  Here's to happily clicking away into the wee hours of the morning.  Cheers!

[DaemonRaccoon]

Good job! But you shouldn't hotplug PS/2 devices, that's a good way to fry the controller or the port.

[Krunchy]

Good job! But you shouldn't hotplug PS/2 devices, that's a good way to fry the controller or the port.

Ah, thanks.  I usually don't, but I didn't know it could cause damage like that.  Never again.