deskthority

Hi there!

First post ever here
I usually frequent reddit but I'm trying to branch out recently to other enthusiast forums. (Also here posting shamelessly because I need your help)

Anyhow, I recently came into possession of an IBM 5556 Multistation keyboard:









Full album and some of my blabberings here: https://imgur.com/a/zhYbG

---

I've had some trouble trying to get this keyboard to work with my computers, unfortunately. The connector on this board looks like AT, so I've tried connecting it via an AT-PS/2 adapter + active PS/2-USB converter, and also through native PS/2 but so far nothing works.

I have 2 active PS/2 converters, a blue cube and a Belkin F5U119, which have given me different results in use. Connecting the board with the Belkin converter and native PS/2 will have no effect – nothing happens when I type on the board. Connecting it with the blue cube will result in some phantom keypresses, basically this key combination "=Lr" gets held down perpetually, and randomly changes to some other letters with no action involved. I've tested this on both a Windows PC and a Macbook Pro, both yielding the same results (different random key combi in os x)

---

So to cut straight to the point, will I be able to use a Soarer's converter or Hasu's TMK to get this keyboard working? found a youtube video of a guy who's managed to get it to work, but I've not managed to get his reply yet:
Anyways I'm not too sure what model of the board this is, seems most likely to be a 5556-004 based on your wiki? (I freaking love it btw, it's my favourite wiki ever). I did have a look through Sandy55's archives and found another chinese multistation board, but the part number on his is different from mine. (http://sandy55.fc2web.com/keyboard/5578.html)

Thanks in advance for the help! I found a thread discussing these boards but I didn't want to necro post so I made a new one here.

---

TLDR; Got an IBM Multistation keyboard that can't work. Can I use Soarer's converter/TMK to make it work?

Hmm…

Someone may know. But not me. It might just be XT: AT's predecessor protocol, confusingly used on the exact same pinout! If so, Soarer's Converter ought to handle it. I'd definitely consider getting a Teensy and making a Soarer Converter, as it's a rock solid converter in any case and is smarter with antique boards than blue cubes and the like.

http://deskthority.net/workshop-f7/xt-a ... t2510.html

Possibly but the looks of the layout it may well be the terminal connector form the 122F , soarer's should do nicely

What an excellent find.

Before trying Soarer's Converter make sure you know which pins are power and ground on the DIN connector. You can trace back to the 8048. Use a multimeter to check for continuity and find a data sheet for the 8048 online including pin out.

I think orihalcon got a similar board not too long ago.

Wow, what an awesome board. With Alps greens no less. Congrats on the find and good luck on the operation! Welcome aboard!

A quick tip. to rule out ground is to use continuity between somthing that you know is ground eg the plate or the outer of the connector or the ground plane of the PCB, that rules that one out.

Also to rule out power look for a capacitor and trace back. Chances are if there is one it will be on the VCC for smoothing.

That keyboard is sweet by the way i like it!

Thanks andrewjoy. Sorry for the terse explanation before. I do have a reverse engineering project in the works and will give a more thorough explanation of reverse engineering principles in a future post, complete with photos, diagrams, and techniques.

This is a good video on reverse engineering

https://www.youtube.com/watch?v=7TedIzmguP0

Thanks for all the help, especially regarding the pins! I've ordered a Teensy to build a converter for this keyboard, I will get back with my success/failure once it's done.

Did you ever get it to work?

I got what I think is a version 003 with cylindrical dyesubs. I plugged it in through Soarer's converter and I got some weird behaviour - keys were roughly shifted down to the left and one row down, and when I press the key, it keeps sending it until I press another one. I noticed a bank of DIP switches on the PCB, but I have no idea what any of them do.

This page says it is AT(PS/2) singnal but scan code is different and keys don't send break codes except for some keys.
http://m0115.web.fc2.com/Products/ps2ibmpc_man.html

You will have to write code to translate scan codes and generate psuedo break events.

alh84001 wrote: ↑Did you ever get it to work?

I got what I think is a version 003 with cylindrical dyesubs. I plugged it in through Soarer's converter and I got some weird behaviour - keys were roughly shifted down to the left and one row down, and when I press the key, it keeps sending it until I press another one. I noticed a bank of DIP switches on the PCB, but I have no idea what any of them do.

Didn't you say it was ABS? Did that turn out to be untrue? I have only ever seen the cylindrical models with PBT, myself.

hasu wrote: ↑ You will have to write code to translate scan codes and generate psuedo break events.

Interesting. I'm not above that, but I don't know when I will catch some time for it. Do you think I could use tmk as a base for that?

E3E wrote: ↑ Didn't you say it was ABS? Did that turn out to be untrue? I have only ever seen the cylindrical models with PBT, myself.

I really have too much stuff laying around, so I mixed some keycaps, either on the desk or just in my head Turns out they are actually PBT. I'm in the middle of cleaning the keycaps, so I'll have some pics soon.

Don't mind the necro-ing of this thread, but I'm curious if anyone succeeded in converting this board? Writing the code, as Hasu suggested, is a bit outside my knowledge.

//gainsborough wrote: ↑Don't mind the necro-ing of this thread, but I'm curious if anyone succeeded in converting this board? Writing the code, as Hasu suggested, is a bit outside my knowledge.

Same issues here.

I wonder if this might be a job for an Xwhatsit.

Anyone know how many columns and rows the matrix has?

nice keyboard man

I'm not really sure writing new code is neccesary in this case.
They were 3 scancode sets on IBM machines. Originally Set 2 and Set 3 were translated back to Set 1 by the 8042 microprocessor on the motherboard. Soarer's converter supports all three.
If this board indeed doesn't send break codes it uses Set 3 just like M122 "battlecruiser".
With passive PS/2 adapter it'll work on linux (terminal only, no xwindow) and winXP, win7 (requires hot plugging after system boots). Don't know about win10 though.
With Soarers's converter there's no such annoyances so it's a better option.

purdobol wrote: ↑I'm not really sure writing new code is neccesary in this case.
They were 3 scancode sets on IBM machines. Originally Set 2 and Set 3 were translated back to Set 1 by the 8042 microprocessor on the motherboard. Soarer's converter supports all three.
If this board indeed doesn't send break codes it uses Set 3 just like M122 "battlecruiser".
With passive PS/2 adapter it'll work on linux (terminal only, no xwindow) and winXP, win7 (requires hot plugging after system boots). Don't know about win10 though.
With Soarers's converter there's no such annoyances so it's a better option.

When I tried using a Soarer's, several keys were not showing any output with hid_listen.

just_add_coffee wrote: ↑ When I tried using a Soarer's, several keys were not showing any output with hid_listen.

Hmm then I would try PS/2 passive adapter on linux to check what the keyboard is sending in these instances. Alternatively you could just try to force set1, set2, set3, set2ext in config and check if any of them work.
It may be the case of weird keyboard ID confusing Soarer's converter.

I'll try to give that second option a shot. I don't have a linux computer handy right now.

I'm interested in seeing if an Xwhatsit could work here.

just_add_coffee wrote: ↑I'll try to give that second option a shot. I don't have a linux computer handy right now.

I'm interested in seeing if an Xwhatsit could work here.

As am I!

I have a multistation board myself (though doesn’t have the older/cooler profile caps like the original post shows) that I’d been meaning to convert and wanted to use an xwhatsit, but if I recall, the matrix isn’t very straightforward. We need a master like oldisnew to sort through it. I could try to take a close up picture of the PCB and if anyone wants to try to trace out rows and columns, be my guest! I can wire an xwhatsit to it and post if it works

I’ve exhausted the Soarer’s converter option and I think the protocol is unsupported (ps/55 perhaps?)

I would absolutely kill for one of these with the spherical caps... Especially a Japanese one. The layout is immaculate, possibly the best caps ever made for SKCC mount, this thing is a beast.

ullr wrote: ↑I would absolutely kill for one of these with the spherical caps... Especially a Japanese one. The layout is immaculate, possibly the best caps ever made for SKCC mount, this thing is a beast.

They are super nice - I really want to be able to use it! The mount is SKCM, though.

Oh that’s what I meant, I guess I was getting them crossed in my head the ones on the so‐called ‘Pingmaster’ keyboards which are very similar to my eye but for SKCC.

I just love everything about these, from the off–center pencil holder and the embossed badge to the oddly prescient split spacebar (what other non‐ergonomic keyboard from so long ago echoes today’s tiny keyboard trend like that?)

It is an incredible shame that this thing would be nigh unusable without making a new PCB, even with the converter.

I don’t play computer games but the shift keys and backspace key not being able to be held down is a deal breaker, moving those keys or using sticky keys would work but it just isn’t worth it