Calling on model F experts: single key not working

[voided]

Hi all,

I have an IBM Model F PC AT with a single key not working. It's the ESC key, placed at the top left of the number pad (where you would expect NUMLOCK to appear on a standard keyboard layout). This key also doesn't sound quite right. There is a much duller click than any other key on the keyboard.

I recently took all the keys out for washing, and managed to get every other key back in properly (sometimes it took a couple of attempts), but I can't get this one working or sounding right even after at least 30 attempts at reinserting the key.

I also have reason to believe that the problem may arise from something other than the key insertion. Namely, the ESC key rocker/hammer doesn't seem to be moving normally. For all other keys, if I remove the key cap/stem, tilt the keyboard backwards and lightly push the spring backwards, it will easily "fall" towards the rear wall of the key well and stay there. In the case of the ESC key, the spring will not move from its central position. It seems that the ESC rocker is stuck, or at least much harder to move. Is it possible/common for the rockers on a model F to stop moving properly?

Perhaps this is a red herring, though. My main question is whether those who are familiar with the Model F might be able to suggest further steps towards a diagnosis/resolution. Do I need to start opening up the keyboard, or is there more I can do without disassembling it? Any wisdom you might be able to offer would be greatly appreciated.

p.s. I also just noticed that when I was trying to push the ESC key spring around that I would occasionally trigger the NUMLOCK key accidentally (the key immediately to the right). Not sure whether this indicates anything.

[fohat.digs]

Be very careful if you try this:

It may be that the little tabs (or "ears") of the pivot plate have jumped out of their tracks, or gotten mis-aligned, just  slightly.

I have successfully used a small screwdriver, held vertical and pressed straight down, then shuttled around horizontally, to force the pivot plate to slide around, every so slightly, until it re-seated itself.

Just something to try in desperation.

[voided]

Many thanks for that tip...not something I would ever have thought of and potentially much better than opening up the keyboard.

I'll take it gently and let you know how it turns out.

[fohat.digs]

I'll take it gently and let you know how it turns out.

If this is going to work, it will work in 10 seconds.

Don't keep poking and scraping around and potentially chew things up.

[jacobolus]

The slightly more dramatic solution is to take the keyboard apart. It’s not as hard as some folks claim to completely disassemble and reassemble one of these, and it’s kind of fun to see how they work inside. (And that should definitely give you a solid idea what’s going on with that switch.)

[voided]

Thanks to everyone for the input. In the end, I did end up taking the board apart. I agree with the advice from jacobolus that the disassembly/reassembly was not quite as hard as I had been expecting. But, then again, perhaps this was because I enjoyed the benefit of so many people sharing their knowledge here and elsewhere, and was being very careful to document everything as I went along.

When I finally separated the PCB from the rest, I found the problem immediately. The hammer that was at issue actually stuck to the PCB and lifted away with it. The reason? A single tiny bead of water, just slightly larger than a grain of sugar, that was between the hammer and the PCB. This must either have been a remnant from the environment out of which I rescued the Model F a few days ago, or else a problem I introduced myself by not drying the keys thoroughly enough after cleaning.

A couple of techniques I worked out in the process of the disassembly/assembly that might perhaps be of use to someone else later (these might well be common knowledge, or have been suggested elsewhere already):

1) The metal backplate that carries the PCB has holes in each end. By inserting an allen key or thin screwdriver and leveraging against the black metal plate that holds the keys, the two halves of the assembly can be easily slid in opposite directions (as is required to separate and re-join them). No need for any hammering, at least in this case!

2) The hammer for the spacebar can be tricky to get in the right position when pushing down the metal backing plate. I found that using a hacksaw blade to keep the hammer pressed down until the metal back was in position made the reassembly a lot easier.

Thanks once again to the geekhack community for your input.

[vivalarevolución]

I have had the same problem as you when reassembling the Model F.  I always thought it was the result of my mistake when reassembling the machine with my homemade foam mat, but I never consider that the issue was a bit of liquid on the pivot plate.  That makes sense, because I was a bit impatient about putting my springs back in the keyboard.   Perhaps I did not let the components dry properly.

[berserkfan]

The slightly more dramatic solution is to take the keyboard apart. It’s not as hard as some folks claim to completely disassemble and reassemble one of these, and it’s kind of fun to see how they work inside. (And that should definitely give you a solid idea what’s going on with that switch.)

It depends on who you are. Some people are not good with their hands at all. These people tend to use avatars of barbarians.

[fohat.digs]

I agree with the advice from jacobolus that the disassembly/reassembly was not quite as hard as I had been expecting.

The little ones are much easier than the big F-122s. Also, the original compressed mats probably make the sliding fit smoother than any of the firmer replacement mats.

I did 3-4 each of XTs and ATs using my hands, alone, without any worry with clamps and such.

Boy, was I in for a surprise when I started working on the big boys!