M-60: A 60% Buckling Spring Keyboard (Like a tiny Model M)

[samwisekoi]

Greetings!

The parts have arrived from Unicomp, Mouser, and PJRC, so I am officially starting the custom build of the "M-60", which will be a 60% version of the "Model M" Buckling Spring (B/S) keyboard.  "Model M" is in quotes, because no actual IBM/Lexmark Model M keyboards will be harmed in this project.  I am using a "new" Unicomp Classic 104 PS/2 keyboard as the base for the mod.  (Oddly enough, the plastic case is stamped with "Model M" near the bottom label.)

Anyhow, I will use some of the hidden switch locations, some custom keycaps from Unicomp, a Teensy-based Soarer USB converter, and careful Dremeling to create the layout below:


Mock-up of M-60 Buckling Spring Keyboard

I have everything to make the keyboard as shown except the front-printed keycaps.  I will get those after the rest of the build is done.

I have not decided whether to mod the Unicomp case or to build a much more basic "case" like the Poker and Race use.  That is TBD after I work out the origami pattern for the B/S membrane and where to put the Unicomp controller and the Teensy.

Anyhow, more to come as the build progresses.  Apologies to everyone who said this would be terrible to do because it is their most-wanted keyboard.

Let's see how it goes...

 - Ron | samwisekoi

[mashby]

  You have my full attention

You are combining my two favorite loves: BS switches and 60% layout. If there is anything I can do to help, please let me know.

[Gupgup]

Wow this would be amazing, can't wait so see the outcome!

[The_Beast]

I bought a model M and was going to cut it up to make a 60% but I just never got around to it. Maybe this summer if I have nothing else to do....

[keymaster]

This will be so cool...That little thing producing buckling spring sounds

[CPTBadAss]

It's so cute! 

I'm really interested in seeing how this turns out.

[regack]

please please please document every step... so that I can recreate this at a later date

[absyrd]

Going to be awesome. Will rival hasu's and CC's mini-M's!

[Gupgup]

please please please document every step... so that I can recreate this at a later date

I totally agree! Please

[snoopy]

Awesome project!

Lookin forward to see a lot of pics.

[samwisekoi]

Messages received.  I cannot promise an increased ability to take the pictures, but I can promise to increase their frequency.

 - Ron | samwisekoi

[CPTBadAss]

Messages received.  I cannot promise an increased ability to take the pictures, but I can promise to increase their frequency.

 - Ron | samwisekoi

Show Image

Potato pictures are always acceptable. The alternative of no pictures at all is not!  :p

[inteli722]

This is a definite will-follow thread. I haven't typed on a Model M, but I hear they're amazing, and its a 60% board!

[samwisekoi]

Ok, Day One, Raw Materials in photos...


Unicomp Classic 104, Custom Keycaps, pivot/springs, Teensy


Secret switches exposed


Custom keycaps ordered via support.pckeyboard.com


Teensy 2.0 and pivot/spring assemblies for the secret switches.


Test install of custom keycaps

Next up: Bolt-mod, install pivot plates and possibly rough cut the plate to about 75%.

 - Ron | samwisekoi

[samwisekoi]

Case is open and the initial cut-lines have been decided.

The flat dimensions of the plastic layer will be 11.50" x 4.75".  There is an easy-to see vertical mod line at 11.50" from the left edge, and an almost invisible horizontal line at 4.75" from the bottom edge.  The cross shown in this picture is where they intersect.  You can just make out the vertical line in the photo.


Preliminary cut lines for the M-60

I will make these two cuts in the plastic during the bolt-mod process.   (See IBM_Model_M:Nut_and_Bolt_Mod.)  I will cut the underlying metal 0.25" INSIDE these cuts so I can fold the membrane around the edge of the metal without it being visible under the plastic.

 - Ron | samwisekoi

[SmallFry]

Cool! Already halfway into the mod! Definitely watching this one!

[alaricljs]

I think SmallFry is too use to Windows progress dialogs where the last 5% on the bar takes 95% of the time.

[samwisekoi]

I think SmallFry is too use to Windows progress dialogs where the last 5% on the bar takes 95% of the time.

Indeed.  This is very much at the 5% mark.

 - Ron | samwisekoi

[Elrick]

I think SmallFry is too use to Windows progress dialogs where the last 5% on the bar takes 95% of the time.

Indeed.  This is very much at the 5% mark.

 - Ron | samwisekoi

Looks extremely interesting.  Please keep all of us posted on this.  Haven't seen a custom buckling-spring project like this before.

[calavera]

OMG this is very cool.

[lazerpointer]

This will be a lot of our dreams come true, including mine... If this works, I will try to follow your project and duplicate it.

[hashbaz]

The only thing cooler than this would be an F60.

[samwisekoi]

60% Bolt-mod complete.  All other plastic rivets have been decapitated.


Model M* with 60% bolt-mod and all other rivets removed.

Next up: add spring/pivots to hidden switch locations and possibly cut plastic as defined previously.

 - Ron | samwisekoi


[edit] *I am tired of typing "Unicomp 104 Classic", so henceforth I shall refer to the donor keyboard as a "Model M." 

[guilleguillaume]

I will be following this. Nice project.

[fohat.digs]

Why do you keep saying "possibly" cut plastic?

Everything has to be cut to the same footprint, doesn't it?

[samwisekoi]

Why do you keep saying "possibly" cut plastic?

Everything has to be cut to the same footprint, doesn't it?

Sorry; I mean "possibly today" I will cut the plastic.

I will be cutting the plastic.  Possibly today.

 - Ron | samwisekoi

[rknize]

Keeping an eye on this too.  I may sacrifice my Unicomp EnduraPro to this cause (as I type this on my F-122). 

[Elrick]

I hope everything goes to plan for your Buckling Spring Project.  What your keyboard desperately needs, is an all CNC'd anodised Aluminum Case.

You can just imagine how great that baby would look just sitting there on the desk all nice and heavy.

I have a feeling that most who still like using a BS keyboard would even jump on for an Aluminum Case Group Buy, to help lower the total cost of machining multiple items from one factory.

[mashby]

I have a feeling that most who still like using a BS keyboard would even jump on for an Aluminum Case Group Buy, to help lower the total cost of machining multiple items from one factory.

Oh hell yeah! I'd be all over that GB like a spider monkey.

[regack]

I have a feeling that most who still like using a BS keyboard would even jump on for an Aluminum Case Group Buy, to help lower the total cost of machining multiple items from one factory.

Oh hell yeah! I'd be all over that GB like a spider monkey.

Count me in on this idea...

[samwisekoi]

I made much progress last night, but my pictures are trapped in this "camera" device I am using.  I will upload pics as soon as I can release the demons.

Plastic was cut.  So was metal.  There is now a 60% PS/2 B/S keyboard sitting in a full-sized Model M enclosure.

I am going to do internals and electronics and then figure out the case.   I am definitely folding the plastic or folding the project.  The man who is making traces is mkawa, not me!

Anyhow, lots of stuff will be inside, so the case may have to be large for a 60%

We will see.

 - Ron | samwisekoi

[samwisekoi]

This is where we are so far.  I will post build photos in a bit.


M-60 installed in full-size case; no electronics yet.

 - Ron | samwisekoi

[samwisekoi]

OK, apologies in advance for the length of this post.  I believe I was told in no uncertain terms that poeople following this thread wanted pictures, so here you go...

I should really have taken the keycaps off before I did this:


Metal plate off, exposing matrix and membrane


Backside of keyboard showing all the pivot/spring assemblies.

NOTE: I used an unused plastic barrel plate, so it still had plastic flash in some of the tiny little spots.  If you modify the barrel plate from a working keyboard, this won't be a problem, but if you use a new one, then make sure you remove any flash.  The pivot plates need to be able to move freely, or the switch won't close when the spring buckles.  It is very obvious when there is a problem and when the pivot plate is moving freely.


Front side before any cutting happened.


Unicomp PS/2 keyboard controller, making contact via pressing against the mylar matrix.
All together; top-view.
I am going to have to replicate or re-use the plastic pressure plate somehow.  And I can't test the switches until I figure this bit out.


Reduce, reuse, recycle!  And don't lose any of the tiny bits.


60% of the plastic barrel plate with pivot/springs installed.

This is really hard to do well.  I need a fixture that supports the barrel plate at least 1" above the work surface so the springs can dangle freely.  A 2x6 piece of hardwood and a 1/2" drill may be deployed if I have to do this assembly again.


Membrane and matrix press-fit back in place.  Carefully!


Metal plate cut-down to be smaller than the plastic plate is now attached.  Carefully.


All together; top-view.


And in the big case for now.

Ok, once I figure out how to re-attach the Unicomp PS/2 controller, I will test the switches.

The journey continues...

 - Ron | samwisekoi

[samwisekoi]

Modified keyboard connected to Unicomp controller.  Almost 100% failure to register keystrokes.

Since the keyboard has no NKRO, any key that is compressed by mis-assembly could prevent others from registering.

On to mechanical de-bug mode now.

 - Ron | samwisekoi

[Alessandro]

This is very awesome, count me to be interest.

[fohat.digs]

I need another project like I need a hole in the head.

Now I know something to do with the 4+ Model M carcasses I have laying around.

[samwisekoi]

Friday afternoon project update.

After re-assembling the keyboard twice, I was able to determine the problem and a path to solution.


Unicomp controller with mechanical connection to matrix.

This controller has edge connectors on the underside when viewed from the top.  These edge connectors are pressed against two layers of mylar with traces printed on in conductive ink.  Three layers of connector must make solid contact in order for the controller to get signals from a keypress.  Unfortunately, this connection is enforced by squeezing the three layers between the plastic barrel plate and the metal backing plate.

This pressure connection must be maintained along an edge approximately three inches in length.  Any bowing or shifting means some or all of the traces will not make connection.  Further, this connection is held in place by pieces of metal and plastic that won't be present in the folded M-60.

So I will need to make a joining fixture OR a PCBA containing the Unicomp controller and the Teensy.  Perhaps with a proper edge connector from a PC motherboard.

I will think about a design this weekend.  I will be very busy next week, so this project will slow down for a while.

 - Ron | samwisekoi

[calavera]

[Gupgup]

Soo jealous of this project and your knowledge..

*Gupgup starts chanting "Group B.. Group Buy.. Group Buy.."

[samwisekoi]

Soo jealous of this project and your knowledge..

*Gupgup starts chanting "Group B.. Group Buy.. Group Buy.."

Any knowledge I have was obtained by reading GH posts, talking to GH members, and trying things out slowly and carefully.

I thank GeekHack for making it possible, and am sure that anyone else here could do the same through interest, courtesy, and a willingness to learn through trial and error.  Especially learning from error.

 - Ron | samwisekoi

[mkawa]

i have plenty of samples of the two ssk layers

[samwisekoi]

Howdy pardnersl

I'm going to split this topic into two threads.

- Ron | samwisekoi

[samwisekoi]

Project update. Custom clamp machined out of .250 x .250 steel bar stock, drilled and tapped for a pair of #4-40 machine screws.

Appears to work. I will test and post pictures tomorrow.

- Ron | samwisekoi

p.s. wcass - delete and re-add the post?  mkawa will know.

[wcass]

any pictures?

i had a case idea i wanted to throw out.


sides are CNC milled wood (oak, cherry, maple, bamboo), aluminum, or 3D printed plastic.
bottom is common aluminum sheet 12.75" x 7" x .03".
rear piece is common aluminum angle 1" x 1/16" thick.
front piece is common aluminum angle 1/2" x 1/16" plus semi-common aluminum extrusion
orange
brunnerent

[RougeR]

* RougeR is monitoring this thread

...how is the controller going to be done?

[samwisekoi]

* RougeR is monitoring this thread

...how is the controller going to be done?

Current plan is a Unicomp PS/2 controller to a Soarer and USB out.

Unicomp controller is curently unhappy with the cut-down chassis, and I have been tied up with Retro DSA. The prototype M-60 is sitting on my desk glowering at me.

 - Ron I samwisekoi

[guilleguillaume]

Very intesresting and this could start a new trend with mini-buckling spring keyboards.

[tricheboars]

waiting for OP to deliver...

[samwisekoi]

waiting for OP to deliver...

Me too!

 - Ron | samwisekoi

[phetto]

WOW RON!!! Nice work. You are acctually doing something that most people only dream of.. Looking forward to the final board