First off, that's a cool looking keyboard, I never saw a model M like that.

Next, that connector looks iffy, almost like the pins broke off. Maybe it can be rewired?

the connector is likely fine, pin 2 and 6 are unused on PS/2 ports. Although I have to say I have never seen a connector that actually omits those pins. Cost savings gone crazy?

Yeah this is not broken, it's just made like this...
So after further research I found out that the two leds lighting up is the common symptom of dried up capacitors. I'm going to disassemble the keyboard and check if indeed these caps are rated 2.2uF 50v and 47uF 16v, before ordering them.

Get ready for a nightmare. I actually just finished repairing one of these two days ago and I've had it for two years. When I received it it was pristine on the outside, possibly unused. It had bad capacitors and some blackened traces on the key matrix though, so even after replacing the caps it ended up being a complete nightmare to fix. I left it alone for two years after wasting several hours trying to fix it with winshield defogger repair epoxy (conductive paint basically). Most of the time spent is the disassembly and reassembly. All keycaps have to be removed before you take the keyboard apart. DO THIS FIRST, no matter what you picture in your head... there are no shortcuts here. If you leave the caps on, the second you unscrew the back all of the springs will pop up out of their housings and you'll have to manually set 101 springs back into their holders. It is very tedious and has to be done with the keyboard perched precariously on its edges to allow the springs to stay in place... if it slips down, they'll all shoot out of the housings again. You absolutely must take the key caps off, then disassemble the keyboard upside down with the springs elevated, separate all of the clips set into holes in the back and then remove the back without jostling the springs loose, or you'll have even more work to do. Realistcally, you should just plan on having to reset all of the springs anyway because its inevitable that you'll knock most or all of them loose...

The holder for the PCB inside is also an absolute garbage design and involves a series of plastic clips that were clearly only ever intended to hold the PCB in, not to have it removed. It is very hard to not break one. I had one break and had to replace it with a screw (thankfully there is a screw hole in the PCB in an area close to where the clip broke, and there isn't anything below it). This is absolutely essential because the PCB squashes down onto the contacts of the key matrix, and if there isn't enough pressure (due to broken clips) you'll have whole parts of the keyboard that don't work. I actually found that replacing the original screws on the back (only TWO screws holding this thing together... how cheap can they get??) with slightly longer pointed screws allowed me to tighten it together better which made it more reliable after all of my repairs were complete.

Anyway, the keyboard has the perfect layout, a great classic noisy-mechanical feel and sound, but is really really cheaply built and designed poorly. I can only hope that the repairs I did to the matrix all hold up. I used copper foil tape and defogger repair paint. I sincerely hope yours doesn't have damaged traces. Make 100% sure you use good caps and install them properly.

... have fun! 🙁

I had two of these back in the day, both were destroyed by my brother in fits of anger. The more intact of the two had a few broken Function keys and a few non-working ones (cursor keys, - key and a few more) but I found another "broken" one (a pin broke off) so I'm now in the process of merging the two into a working keyboard.

I did a dry test - for those who are having problems, you CAN test every key WITHOUT having the top on (and thus all the springs and caps back into place) - you can see on the black sheet where the feet were before - simply put one spring there and a random key over the spring and press it. It works this way. Do this 104 times for each key and you'll know if any give you issues. Yes it's tedious but it's a LOT quicker than the alternative.

Oh, fun! I got one of these recently -- appears brand new. Haven't tried it yet though. Not sure if I should just head straight to the purgatory of cap replacement, or try to get a year or three out of it first. (Either way, just wanted to say thanks for the tips.)

SirNickity wrote:

Oh, fun! I got one of these recently -- appears brand new. Haven't tried it yet though. Not sure if I should just head straight to the purgatory of cap replacement, or try to get a year or three out of it first. (Either way, just wanted to say thanks for the tips.)

I'm not going to bother with replacing the caps yet - if it works, it works. From what I read, bad caps won't leak but will just make the keyboard malfunction until you replace them.

Also: my earlier tip can even be made easier: just use a metal soft-tipped item to press the regions to test them.

Whelp I spoke too soon. One or both of the caps must have dried out because half the time the keyboard isn't detected. Worked fine the first 5-10 times I turned on the PC but now it doesn't. Thing is, I know very little about capacitors - it says J1 47 16V on one of them - I suspect 47 = 47µF and J1 states the tolerance (5.0%)?

I'm not an expert at deciphering SMD markings, but I can't imagine they would spec 5% tolerance capacitors. I haven't seen the insides, but in a KB controller, they're almost certainly just bulk power decoupling caps. 20% is the norm there. 47uF sounds reasonable for a low-power device, and would make sense based on the markings you describe. Have a picture?

I cleaned such keyboard recently outside and inside. It works perfectly when connected to already booted PC, both to modern motherboard, PS/2 socket, as well as to 440BX board, AT socket.

But if a PC is started with the keyboard connected, it's not working. Somehow it also makes 440BX to do full RAM scan on POST instead of quick self test! 😃 What's going on?

kolmio wrote on 2024-12-22, 04:46:

I cleaned such keyboard recently outside and inside. It works perfectly when connected to already booted PC, both to modern motherboard, PS/2 socket, as well as to 440BX board, AT socket.

But if a PC is started with the keyboard connected, it's not working. Somehow it also makes 440BX to do full RAM scan on POST instead of quick self test! 😃 What's going on?

Honestly, don't even bother with these things. See my post here: Re: IBM Model M2 not recognized

Not long after I posted that the keyboard was dropping keystrokes again because it is such an incredibly shoddy design... I can't even imagine how many hours I wasted on that IBM (Lexmark) Model M2.

Sorry, but I would have preferred to have the time and sanity back that I lost while trying to repair it. If a little bit of bad news can save someone the same fate, I will gladly deliver it. 😮

Thank you for the empathy! I'm not yet ready to give up 😃 Upon better observation it turned out that at first power on, before any restart, there's a usual sign of bad capacitors - CapsLock and ScrollLock LEDs stay on.

So now I just need to figure out where to get new capacitors.

kolmio wrote on 2024-12-22, 10:40:

Thank you for the empathy! I'm not yet ready to give up 😃 Upon better observation it turned out that at first power on, before any restart, there's a usual sign of bad capacitors - CapsLock and ScrollLock LEDs stay on.

So now I just need to figure out where to get new capacitors.

Check the internals carefully first. If the traces have already burned\blackened on the plastic key matrix or if the plastic clips that hold the circuit board snap off when you try to remove it, then I would probably use the time elsewhere. 🙁

kolmio wrote on 2024-12-22, 10:40:

Thank you for the empathy! I'm not yet ready to give up 😃 Upon better observation it turned out that at first power on, before any restart, there's a usual sign of bad capacitors - CapsLock and ScrollLock LEDs stay on.

So now I just need to figure out where to get new capacitors.

Those symptoms are the same I used to have on my M2 keyboard, just swapped the capacitors and it worked immediately, I even used radial ones I had at hand. If some clips broke it's not that big of a deal, normally the main clips are a bit stronger and can hold the keyboard together in the necessary spots.

In my case the clips that hold down the board were weakened which removed the necessary pressure for the membrane to contact the board. Basically it's just a question of applying some padding on top of it so it keeps pressure when the top half of the keyboard is inserted, give it a good guess and apply a little more than you think it's needed. That's because the padding will probably compress with the pressure when it's all closed and with time will need a new layer on top of the others to keep the correct pressure level. I seem to remember using furniture pads cut to size stacked up.

The real pain of this keyboard is the reassembly which combined with the board weakened clips' problem demanded me to reassemble, test, and after a day or two the pads compressed and needed a new layer. It was a bit tedious but after a few tries to find the right padding amount it still works fine today. If you're careful once you open it a first time the clips that broke are the ones that would break anyway, and all others should remain intact for the remaining tries. Have a look for blog entries about this repair, I remember finding one that had a very useful picture outlining the spots on the keyboard where the springs are not to be inserted, which speeds up the reassembly. Keep the keyboard firmly on top of two mounts on either edges to insert the springs, etc. If a spring is crooked or gets crooked for some reason do not be discouraged, even crooked springs can provide the same clickyness than a normal spring, you just have to insert it the right way. Never try to fix them by twisting or bending!!! In any Model M variation, aways insert the caps by tilting the keyboard up and away from you, the springs will tilt away within the same motion, and then you insert the cap aligned with the tilted spring, bringing it towards you. In essence, insert it diagonally so it catches the tilted spring and clicks down in place. Use a plastic spudger to wedge the cap out (not those damn keycap pullers) and repeat it until you get a properly clicky key. It may even take up to 4 or 5 times on some keys, but all of them will be clicky once you get the hang of it.

This wasn't the blog article I mentioned but it can be of some use, you probably saw it already: https://blarg.ca/2018/04/22/fixing-up-an-ibm- … del-m2-keyboard

I have a Model M and a Model M2, and the M2 is the one I use with my Win95 build. It's quieter, sleeker, lighter, more compact and I love the look of it. It's nice to have a Model M too but it's too loud and bulky for regular use. Good luck with your repair, in my opinion it's well worth it!

Ozzuneoj wrote on 2024-12-23, 08:05:

Check the internals carefully first. If the traces have already burned\blackened on the plastic key matrix or if the plastic clips that hold the circuit board snap off when you try to remove it, then I would probably use the time elsewhere. 🙁

Many traces and contacts on the membranes were indeed blackened. I used isopropyl alcohol and an eraser on the membranes and it work quite good, especially the eraser. All oxydation was gone.

And I also knocked off several clips holding the board, before realizing it's much better to bend out a little the external edge of the keyboard case and so release the board from upper clips. Again, the eraser came to help 😀 attached it on top of the board with a thin double-sided tape.

P.S. You can notice that I placed the black earth wire the wrong way, but I don't think it should cause any problem.

kolmio wrote on 2024-12-25, 08:12:

Ozzuneoj wrote on 2024-12-23, 08:05:

Check the internals carefully first. If the traces have already burned\blackened on the plastic key matrix or if the plastic clips that hold the circuit board snap off when you try to remove it, then I would probably use the time elsewhere. 🙁

Many traces and contacts on the membranes were indeed blackened. I used isopropyl alcohol and an eraser on the membranes and it work quite good, especially the eraser. All oxydation was gone.

And I also knocked off several clips holding the board, before realizing it's much better to bend out a little the external edge of the keyboard case and so release the board from upper clips. Again, the eraser came to help 😀 attached it on top of the board with a thin double-sided tape.

I hope it works well for you. 😀

Similar to andre_6 above, I went through the disassembly process and the "Oh! Yes! It works!.... wait... NO... that key stopped working again..." probably 10 times trying to adjust all of the things that kept malfunctioning until I finally gave up on it.

The problem with projects like this is that once you have several hours invested (which is only a few tear-down-rebuild cycles with a device this tedious to work on) it becomes harder and harder to cut your losses and move on. Thankfully, humans are all wired differently and some find no irritation at all in projects like this. It's just not for me though.

I really do hope yours keeps working! ☺️

andre_6 wrote on 2024-12-24, 17:44:

If a spring is crooked or gets crooked for some reason do not be discouraged, even crooked springs can provide the same clickyness than a normal spring, you just have to insert it the right way. Never try to fix them by twisting or bending!!! In any Model M variation, aways insert the caps by tilting the keyboard up and away from you, the springs will tilt away within the same motion, and then you insert the cap aligned with the tilted spring, bringing it towards you. In essence, insert it diagonally so it catches the tilted spring and clicks down in place. Use a plastic spudger to wedge the cap out (not those damn keycap pullers) and repeat it until you get a properly clicky key. It may even take up to 4 or 5 times on some keys, but all of them will be clicky once you get the hang of it.

Oh, spring crooking was the real pain for me. Somehow I got springs crooked on important keys: [Enter] and [←]. And I did try to "fix" them with twisting and bending. Well, eventually somehow I managed to get back required clickiness and they worked correctly.