Bringing the IBM PC XT into the 21st Century

[wcass]

Yes, backspace can be split to two 1 unit keys; left shift can be split to 1.25+1; and right can be split to 1+1.75. You can use either ISO or ANSI Enter, too.

[Muirium]

Superb! I daren't expect that. HHKB function layer here we come! Thanks for have the foresight, Wcass!

Now to find some caps…

Wait, when you say 1+1.75 I guess you mean IBM style, with the single unit key to the left of Shift? If so, no HHKB Fn. Not a huge deal with a TKL anyway; split Backspace is more important to me. That way I get Backspace down to where it should be, with backslash and backquote in its place up top.

[Hypersphere]

Mu, you anticipated my thoughts on this.

@wcass: Could you confirm the Right Shift split? Going from left to right, is it 1.00 + 1.75 rather than 1.75 + 1.00? I suspect the former, but it would be great if it were 1.75 + 1.00, which would enable a HHKB-like Fn key to the right of the Right Shift.

One more thing (for now at least): Are all ten keys in the Navigation Island 1.25 units? I am going to try for black blanks if Unicomp could be persuaded to produce them in 1.25 unit size.

[wcass]

Right shift = 1x on left, 1.75 on right. And yes, all of the 10 navigation keys are 1.25.

I have had very good results with iDye Polly. I was thinking about dying a bunch of keys black and laser etching legends. Does anyone in our group have access to a laser?

[Hypersphere]

@wcass: Thanks for confirming the keycap sizes for the Xtant.

I haven't yet tried dying keycaps. I got as far as buying some Rit, but the process seemed messy and variable. However, I mainly want blank black; this ought to be the easiest color for me to get right!

No direct access to a laser here that could do etching of legends, but I will check with some engineering colleagues and let you know if I find anything useful. We might want to contact IBM:

LaserEtch.jpg (64.05 KiB) Viewed 24107 times

http://www.businessinsider.com/laser-wr ... air-2015-3
Pretty nifty being able to etch their name on a human hair -- twice!

[Muirium]

Right shift = 1x on left, 1.75 on right. And yes, all of the 10 navigation keys are 1.25.

Actually looking at a 2.75u right Shift cap reminds me why it has to be this way. You can either be compatible with a 1.75u HHKB style Shift *or* a 2.75u IBM Shift. Not both. Not without serious work on the cap!

By the way, those 1.25u keys in the nav islands: would there have been room for 1.5u caps over there or is the XT case too small? And, conversely, how much extra space would have been opened up (or rather left unoccupied) by using regular single unit caps instead? I'm trying to get a sense for how good a fit the 1.25s are. I like them in general, but they are a rare thing on IBMs.

[Hypersphere]

It would have been easier to find caps for the Xtant if the Nav Island had been configured for 1.00 unit keys, but apparently the available space was ideal for 3 x 1.25 units (3.75 units total).

I heard back from Etsy already. He said that it would be much less expensive to get the blanks from Unicomp and dye them, because this is exactly what he would do to get the black blank modifiers.

This leaves me with a question already posed in an earlier post. If I want black blanks, what is the best starting color for the keycap to be dyed? The Unicomp choices are Pebble, Gray, or White.

Note that Unicomp can currently provide a blank black keyset, but they don't have blank black Right Shift keys and they don't know if/when they will have them in stock. So, if you order a blank black set, you need to do this as a special order and place a second special order for a Right Shift in a different color. Green is one of the colors, and some people like to put RGB modifier sets on their keyboards. My first choice is still blank black, so it appears that I will be heating up some dye pots soon. Rit or iDye Poly?

[wcass]

If you are dying black then it doesn't matter what you start with. If it were me, I would start with gray (part way there already).

I recommend iDye Polly. Rit changed their formula a few years ago and they don't work as well.

[Hypersphere]

If you are dying black then it doesn't matter what you start with. If it were me, I would start with gray (part way there already).

I recommend iDye Polly. Rit changed their formula a few years ago and they don't work as well.

Thanks. Looks like iDye Poly in black is available on Amazon. I will probably give it a try. I already have lots of IBM BS keycaps in whatever the native IBM colors are called. In Unicomp parlance, I suppose they are Pearl and Pebble. If you start with a Pebble-colored keycap that has a black legend and you dye the cap black, does the legend become completely invisible, or does it end up looking like the black on dark gray caps on "black" versions of RF or HH keyboards?

[fohat]

If you start with a Pebble-colored keycap that has a black legend and you dye the cap black, does the legend become completely invisible, or does it end up looking like the black on dark gray caps on "black" versions of RF or HH keyboards?

My experience, with old-style RIT dye, is that the keys became completely black and the legends disappeared. But I have never gotten out a high-power magnifying glass and bright light to really intensely investigate below the "naked eye" level.

With old IBM PBT, you can boil them for half an hour in the dye bath without hurting them.

[Muirium]

Be careful with those IBM originals! They're nicer than Unicomp's modern caps. I'd experiment with Unicomp stuff first. And likely only dye modern keys, in fact.

[wcass]

This image is from a flatbed scanner, so high light and magnification. In normal light, i don't see any legend on the black caps (purple is slightly legible). I dyed these a while back but IIRC they were in the bath for only about 5 minutes. My red set were done with Rit many years ago and those were boiling for about 45 minutes to an hour.

I use a "pasta pot" to dye so that the caps never sit on the bottom. Doing a full set at a time makes for uniform saturation. The built in strainer makes it easy to check color and agitate the bath.

caps.jpg (572.41 KiB) Viewed 23991 times

[Hypersphere]

Thanks for the pasta pot tip and the dyed cap pic.

Does the dye permanently stain the pasta pot?

[wcass]

I have a stainless steel pot that i use only for this and it has been true to it's name (YMMV). My wife got it at Marshall's or Ross for about $10

[wcass]

Thanks for the support, guys. I am one of the finalists in Big Blue Saw's "Spring" design contest. Results to be announced Friday. Design details at the link below.
http://deskthority.net/workshop-f7/new- ... 10234.html

[Sigmoid]

Hey

This is an interesting project. I have recently restored an XT, and I really like the feel of Model F keys...

Wcass, have you ever thought of developing a new capacitive Model F clone from scratch? I'm pretty sure the patents have all expired, so the tech should be up for grabs. The capacitive sensing circuitry should be no problem either, there's the Xwhatsit controller that does it quite well.

I see that the Spring you have linked to is a membrane keyboard - what was your rationale in choosing that technology instead of capacitive sensing?

[wcass]

New capacitve Model F are on the way. I have heard (and maybe helped a little) from folks looking to recreate the Kishsaver, an FSSK, and ErgoFox. All of these should be possible.

There is only one thing keeping development from really taking off; the need to cannibalize vintage keyboards to make a new one. The specific parts are Model F barrels and pivot plates. Finding a new source for these is critical.

The XTant was a logical first step in that it tried to re-use as much as possible from the donor board. The new board is a logical next step in that it tries to replace every part. If I could find a new source for barrels, the keyboard could have been 100% new parts. We can get brand new caps and M type pivot plates from Unicomp - we can't get model F pivot plates - period. And I really wanted to try out some of the untried/untested/discarded ideas I came up with while making the XTant. TBH, I am really happy that so much of it is working out (because I didn't expect it to).

[Sigmoid]

If I could find a new source for barrels, the keyboard could have been 100% new parts. We can get brand new caps and M type pivot plates from Unicomp - we can't get model F pivot plates - period. And I really wanted to try out some of the untried/untested/discarded ideas I came up with while making the XTant. TBH, I am really happy that so much of it is working out (because I didn't expect it to).

Well the only way we could do that is to get a mold manufactured, and have a factory make them in China. I wonder if it could be sold in sufficient quantities for a successful Kickstarter...

My gut feeling is that a capacitive buckling spring HHKB-like would sell like hot cakes.

[ramnes]

My gut feeling is that a capacitive buckling spring FC660C-like would sell like hot cakes.

Fixed.

[Muirium]

Fresh, tasty XTant plates…

DSCN4455.jpeg (363.28 KiB) Viewed 23622 times

My board will come alive! (In December when I'm there again.) Thanks Viva!

[vsev]

Real beauties here !

[Muirium]

Yeah, they should look pretty hot with some appropriate TLC on the XT case.

Now I'm trying to source the right caps. I'm going for blank, but it's all about details…

Okay, much to my frustration, Unicomp isn't clear about what keys are in their blank set:

http://www.pckeyboard.com/mm5/merchant. ... _Code=KSET

I have one in "PearlPebble", and a spacebar and blank RGB mod set to go with it, but they're in America so I can't check the exact contents.

Do you know what keys exactly are in there? Perhaps I should ask around the forum.

Anyway, what I want to do is… well, first imagine a blank version of this:



I'd like to split the backspace key (HHKB style) and go (dark) pebble on the edges of the 60% block. So I'd need a pebble 1.5u key for backspace (where \ is on regular ANSI) and a couple of pebble 1u caps for either corner of the top row. The left hand side function keys look good all pebble. I think I'll have enough pebble spares from the blank kit to cover all the 1u stuff. But the bottom row mods are a mystery to me.

So, does anyone know what's in that Unicomp blank PearlPebble KSET kit I bought? I'm teaming up with Vivalarevolución to order the necessary caps. I definitely need 10x 1.25 pebble blanks for the right hand side of the keyboard. But anything else?

[Hypersphere]

Now that this is becoming a reality, I wish I had ordered more PCBs and plates!

[Muirium]

More? Is one not enough for you??

[Hypersphere]

More? Is one not enough for you??

Didn't I see more than one plate in the image you posted? Perhaps they are not all for you.

I'm always thinking of providing for my three workstations. I have one in my home office and two at work. If I really like a keyboard, it is handy to have replicates at each of the three workstations.

My colleagues have described me as one for whom "nothing succeeds like excess."

QUESTION: Do the Xtant plates require painting or do they come already painted?

[0100010]

Those look like bare unfinished plates. Would imagine they need to be dehorned, sanded, degreased and finished.

[vivalarevolución]

More? Is one not enough for you??

Didn't I see more than one plate in the image you posted? Perhaps they are not all for you.

I'm always thinking of providing for my three workstations. I have one in my home office and two at work. If I really like a keyboard, it is handy to have replicates at each of the three workstations.

My colleagues have described me as one for whom "nothing succeeds like excess."

QUESTION: Do the Xtant plates require painting or do they come already painted?

Hey there. The plates in the photo are from me. I posted that photo in the XTant plate thread on GH.

The XTant plates are bare steel. You could throw them into the keyboard without any finishing, but I recommend that you finish them.

My plan is to smooth them out with some steel wool, and then just throw some clear coat spray paint on there, either gloss or satin. Might go with gloss to add just a tiny bit of flash to the XTant underneath all that IBM classic blandness.

As for ordering more PCBs and plates, we can always do that because all the files and specifications were made available by wcass. I've thought about doing another run with all the small parts and controllers. But I would like to change the plate fabricator to a more local company. The challenge is finding a fabricator that will do small runs and bend the plates as well.

[Hypersphere]

@vivalarevolución: The 2 plates and extra PCB arrived today. Thanks! I initially panicked, because I had forgotten that I had received the first PCB from wcass quite some time ago. I eventually found the box among my stash of boxed-up IBM keyboards awaiting restoration. Things at work are very busy now, but I hope to get to the Xtant project in mid-October or so. In the meantime, I may order some caps from Unicomp.

Glad to know that it could be possible to order more PCBs and plates in the future if I got on a roll with putting Xtants together.

Is it possible to powder-coat the plates, or would this add too much thickness? In addition, although the top case of the XT is plastic rather than metal, is it possible to do a low-temp powder-coat process or would it be better to hand-sand and spray paint?

[fohat]

The XTant plates are bare steel. You could throw them into the keyboard without any finishing, but I recommend that you finish them.

I got mine today. It seemed to have a light oily film which is probably necessary for machining, so I washed it well with hot soapy water in the sink.

I stood it up on my work bench to dry, and now, 4 hours later, I see a tiny wisp of rust bloom at every water droplet location. So I get the distinct feeling that these need to be coated properly or you will be sorry.

Rustoleum for the win!

[vivalarevolución]

Yea, I would definitely add a finish to them. The steel is cold rolled steel, not stainless, and it probably will rust without a finish. Maybe I can go with stainless if there is another run.