IBM 3278 Keyboard to PS/2 interface

[seebart]

Not sure if this was ever posted here, I did stumble upon this months ago and forgot about it and just now looked at some old bookmars of mine, this looks pretty real to me, certainly interesting;

http://www.ibmsystem3.nl/IBM3278/

[Vizir]

Thought that was old. Looks like it was last updated in 2013

[Halvar]

Yes, I remember reading that, compgeke posted it here:
keyboards-f2/ibm-3278-beamspring-to-ps- ... ibmsystem3

It's a great project and might be an alternative if you can live with only three modifiers (e.g. you only want the keyboard to work for demnstration purposes). What's also great about it is that it should work with all those 3278 compatible keyboards from other companies, like your Micro Switch keyboard, Seebart.

[XMIT]

Just had a look at the source. Not only do you only get two mods, but also, you don't get clean make/break codes. Yes, this could get Micro Switch boards hobbling along and barely usable, but a real hardware solution (like strobing power for the one-shot Hall sensors) is better for NKRO, more modifiers, etc.

[seebart]

Well I'm certainly not arguing for this in any way, interesting what xwhatsit posted about it three years ago:

Very cool. Murium posted that up on the Beamspring USB thread. In many ways this is preferrable to the solution I came up with—no need to source silly 3.96mm pitch connectors etc.

In addition, the historian instinct in me appreciates how you're still using the original hardware. Granted with my solution you're not destroying anything, but the original IBM controller then has to sit in an antistatic bag in the cupboard instead of in the keyboard where it rightfully belongs...

I wonder how different the 3278 and 5251-style protocols are? I think they used similar connectors on the PCB, you might just be able to get away with different firmware

Shame about the make and break codes—funny how a lot of those early keyboard protocols didn't bother—but this would be a sweet link in the chain for Soarer's converter!

What's also great about it is that it should work with all those 3278 compatible keyboards from other companies, like your Micro Switch keyboard, Seebart.

I guess your referring to what snuci is calling the IBM 3277 typewriter keyboard:

wiki/IBM_3277_typewriter_keyboard

It's an IBM keyboard built by IBM with Honeywell internals.

As for getting Micro Switch boards hobbling along and barely usable how much progress has been made here, seems like quite a complex undertaking.

workshop-f7/converting-my-wang-t12379.h ... on#p272789

[XMIT]

As for getting Micro Switch boards hobbling along and barely usable how much progress has been made here, seems like quite a complex undertaking.

workshop-f7/converting-my-wang-t12379.h ... on#p272789

Yes. That's my thread.

[Halvar]

Yes, in which I posted, too. Seebart is talking to the anonymous silent reader I guess.

[seebart]

Yes, in which I posted, too. Seebart is talking to the anonymous silent reader I guess.

No, writing (not talking) to you. Possibly also to the anonymous silent reader we call lurker. Either way both your posts seem rather silly in this case.

[Halvar]

I didn't want to offend you, I just didn't understand your first answer to me. Of course I know that we don't have a modern Hall effect controller yet!

Your 3277 keyboard, regardless of of whether it was made by IBM or not (I didn't know it was made by IBM) should work with the same external protocol as a 3278 keyboard, right? So people could swap keyboards between 3270 compatible terminals? I'm not 100% sure, but I think it's highly likely. These protocols worked by strobing some pins with +/- 5V and reading some pins with +/-5V, and were oblivious to the switch mechanism, while not oblivious to the matrix. All the capsense stuff or the hallsense stuff was done internally in the keyboard. The external protocol is standard digital logic. Any teensy should be able to do it once the protocol is understood. Well. a Teensy plus a -5V source.

And the great thing about this controller is that it works with the external protocol, so there's a good chance that it is able to work with your keyboard and its original driver chips, even though the Hall effect controller problem is still unsolved. Of course a new controller would be much better, with more modifiers and break codes and all, like an xwhatsit is better for a beamspring 3278, but as long as that new Hall effect controller doesn't exist, this might be a nice project to get keyboards like your 3277 to somehow work, don't you think?

[XMIT]

Yeah, an external protocol converter is great for all the obvious reasons, but it is only as good as the original protocol. So you're stuck with very few modifiers and no key release events unless I misunderstood.

As for any other Hall boards - the "hold low" variant is extremely easy to convert, it's the evil "pulse low" variant that makes trouble. If you have a Micro Switch board and a way of checking to see which one you have please let me know which it is! Check the other linked thread for background.

[Halvar]

I did.

post289057.html#p289057

(But I also have another one with pulsed switches.)

[XMIT]

Yes, yes you did, and I am grateful for this.

I have several more Micro Switch boards here at the Keyboard Ranch that may have hold low switches. If I can get my act together I'll see about porting some changes to get my, and possibly, your, board working, on top of hasu's tmk.

But I don't mean to derail seebart's thread. It would be great if we could get this protocol checked into tmk or similar. I don't have any 3277 or 3278 boards here for protocol checking. Does anyone want to mail me one?

Usually around the holidays I have a little bit of time for this sort of project.

[XMIT]

Finally, I found this thread again. I wanted to share it with Engicoder since it could be a quick and dirty way of getting his Hall effect 3278 keyboard working with minimal effort.

[Engicoder]

Thanks!