60%/65%/TKL Beamspring "Thousand MonKeys"

[Monster-Toys]

Huhus ( haven't been around for quite some time )

Boring intro to the project:
In december i fell victim to arson (my floor neigbour set his appartment on fire and a good part of the whole house burned down - he now is in a psychiatric institution) and i was not allowed to enter my appartment for a long time, hardly knowing how much it was damaged.
I also was told, that, because of toxins in the soot, all my remaining household stuff will be destroyed.
I soon realized that, if i could choose what to rescue, it would be (after my computer hard drives) my small keyboard collection...
finally in April after quite some fights over it i was able to retrieve some items from my appartment including my three beamspring boards i bought when they were not that high priced (and my computer and my 3D printer).
I already restaurated one of them back then (only the paint job was not good and i will need to redo it some day ). The other two were in really bad shape (water damage - especially the bigger one was more rust than actual keyboard) and i never got to work on them.
As it will be some more month before i can get back to my appartment (living in my fathers house until then) i thought i could just as well start a project i had in mind for quite some time: modding the most damaged beamspring to be a more "modern" and compact TKL board.
(if it works well i will make a dactyl-manuform inspired beamspring after that)


The Project:
Today i took apart the two beamsprings retrieving the items i want to use:

Spoiler:

001.png (1.16 MiB) Viewed 5374 times

The next stepp will be (a lot of) cleaning and derusting and then cutting the perforated plate and adding three additional holes:

Spoiler:

002.png (1.12 MiB) Viewed 5374 times

I have some ideas concerning the electronics (straying away from capacitive sensing) but need to do some testing before i can write more on that.

[matt3o]

ah the misery! they lasted a fire! why destroy such wonderful keyboards!?

[Monster-Toys]

I know that such a project will hurt the feeling of some beamspring lovers quite bad (and i actually feel a little bit sorry ). Thing is i love the beamspring switch and would love to have a portable (and (more) reliable and usable) beamspring keyboard.
I thought of making a whole new perforated plate and let the original remain so everything i do is reversible but that would be quite a bit more expensive (and money is kinda scarce atm).
---
Now that i think of it: I could try to 3d print a plate... not sure if it will be sturdy enough tho (it might be with some properly placed ridges), the plate is 1,5 mm thick ... hmhmhm, i will give it a shot. (Would that make you feel better, knowing that everything i do is reversible?)

[matt3o]

I know that such a project will hurt the feeling of some beamspring lovers quite bad (and i actually feel a little bit sorry ).

just kidding. I myself destroyed quite some keyboards, generally only if seriously damaged or not-at-all-uncommon/rare (such Model Ms).

(Would that make you feel better, knowing that everything i do is reversible?)

If reversible it's always better I guess, but again, I won't feel any better or worse

[Monster-Toys]

I am relieved

[adamcobabe]

Like a phoenix rising from the ashes!

[darkcruix]

I'd love to see exactly the design you have in mind, but there are several hurdles that I encountered and halted my project. The items I broke the project down to, were:
Switchboard rebuild with the exact same holes as you highlighted (relatively easy)
Bottom Plate from cast Aluminium with mold (expensive, very time consuming)
Two layer capacitive sensing PCB (pad) re-design and small batch creation (high LOE + costly)
Case creation from Aluminium (expensive)

I roughly calculated that I would need to spend at least 10k to get a prototype done and even needed to get the switches / keycaps from an existing Beamspring to get there. That is when I halted the project.

[Monster-Toys]

I need to do it "the cheap way" (as i do not have much money to spend atm)

Switch plate:
Not sure yet.

• Easiest would be modding the existing plate (but it is true.. it would be nice to keep it original if i ever want to go back).

• 3D printing would not be strong enough (the distance between the holes would be a predetermined breaking point.
  I would need to make it from many parts, too, as the printer build plate is not big enough and warping would be a problem if the parts are to big etc.
  Maybe i will try it but i am not confident it will work.

• 1,5mm thick aluminum, a drill, a file and patience might be the best way (but i will first ask how much lasercutting would cost... maybe it is not that expensive and i am quite good with CAD software).

Bottom plate:
For the bottom plate i will use nothing fancy like molded stuff. Have not thought about it yet but most likely i will be using aluminum or aluminum dibond.

The PCB:
There will be no PCB, only hand soldered stuff: I have an idea not to use the switches as intended (capacitive switches) but to slightly mod them (reversible) to be "normal" switches using copper foil tape and wick wire (have done something similar in another project, but not with beamspring switches, so i am not yet sure if it will work... but i do not see why not atm. If it does work the board will be muuuuch less squeamish with dust etc.)

The Case:
Not sure yet.

Electronic:
Most likely a teensy

Further Ideas:
I like this: https://getfreewrite.com/
So maybe i will add a rhaspberry pi with a basic texteditor and an oled screen or something like that.

[PlacaFromHell]

Man, what a piece project! I can even listen Dragonforce music in the background.
Okay, talking seriously I got an idea to make the switch plate a little less tricky. The biggest problem is found an inexpensive way to make those dots that hold the switches in place. You could drill just normal holes in a SAE 1000-something sheet (drilling itself is less expensive than get a CNC cut of the right holes with the dots) and then glue over a 3D printed plate with the holes and the dots. You can also make the body by layers instead of get the sheet fold by that way. I'll draw a quick example for you:

[Monster-Toys]

These "dots" are, as far as i understand only there to prevent the switches from rotatating. It is sure nice and better to have them, but as beamspring switches are very tightly packed there are easy workarounds if the dots are not there. So simple circular holes are no big deal. The tightly packed switches lead to another "problem" tho: the drilled holes must be quite precice as there is hardly a margin for errors. Working on a CAD atm to get some quotations for laser or waterjet cut aluminum.

[PlacaFromHell]

They put them for some reason, so should be better avoid that risk as much as you can, more considering that IBM hate (or at least hated) waste money. A good example is the tabs in the model F's instead of using bolts.
The point of make two layers is avoid laser/water cutting and just send it to drill. At least we agree that it's simpler and cheaper just make a hole than a complex figure and that's the idea of the two ways.

[Monster-Toys]

So, i have two options now:
1. Lasercut (and maybe bended) aluminum (or steel) plate (it is actually not that expensive, i was surprised...).

2. Printing a "rack" with the switch housing imbedded (plate only would not be stable enough) and only transfer the innards of the beamspring switches.

I did some testprints with a single housing and now, with revision number four, i have something that works surprisingly well:

Spoiler:

print.png (282.62 KiB) Viewed 5091 times

The whole rack would look like this (but needs to be printed in two or better 3 parts):

Spoiler:

cad.png (661.21 KiB) Viewed 5091 times

[Monster-Toys]

I settled for lasercutting some steel and started working on the CAD files. I now and need to make a final decision concerning the layout, too... in the beginning i wanted to make a TKL but a 65% or 60% beamspring board would be nice, too^^.
ATM i like the 60% most i think.

Spoiler:

layout.png (907.24 KiB) Viewed 5003 times

[adamcobabe]

I settled for lasercutting some steel and started working on the CAD files. I now and need to make a final decision concerning the layout, too... in the beginning i wanted to make a TKL but a 65% or 60% beamspring board would be nice, too^^.
ATM i like the 60% most i think.

Spoiler:

layout.png

Is the ISO-like enter key on the 60% and 65% a beamspring/selectric mount keycap? Which keyboard is it from? I don't recognize it.

[Monster-Toys]

It is from my german "IBM 196C" ("Selectric II" in English-speaking areas i think) and compatible with beamspring switches.
I need get replacement somehow tho, as the typewriter is almost mint condition inc. cover and has all extras etc. ... too nice to be torn apart for keycaps^^.

[adamcobabe]

It is from my german "IBM 196C" ("Selectric II" in English-speaking areas i think) and compatible with beamspring switches.
I need get replacement somehow tho, as the typewriter is almost mint condition inc. cover and has all extras etc. ... too nice to be torn apart for keycaps^^.

Nice. They're cylindrical not spherical though, right?

[Monster-Toys]

Yes, they are. As the original return key has a different shape, too (concave instead of convex) it is no big deal i think (the bottom row with just enough room left to not fit another key is more annoying^^)

[SneakyRobb]

Yes, they are. As the original return key has a different shape, too (concave instead of convex) it is no big deal i think (the bottom row with just enough room left to not fit another key is more annoying^^)

If I were you I would use this sort of layout that wcass helped me with. Uses all original keys you should have them all. It is quite usable.

hhkbeam.jpg (222.4 KiB) Viewed 4834 times

[Monster-Toys]

Cool, i did not know somebody else did something like that... did you make a thread about it? (I am back to tkl atm. I recognized that i use the arrow keys quite a bit when editing texts and using another layer for them would annoy me). The CAD files for the lasercutting are almost done (would like to have more time to work on it, feels horrible slow atm).

Btw.: Even though the selectric keycap does fit the switches, it is slightly bigger then the other keys and therefore not very usable.

[SneakyRobb]

Cool, i did not know somebody else did something like that... did you make a thread about it? (I am back to tkl atm. I recognized that i use the arrow keys quite a bit when editing texts and using another layer for them would annoy me). The CAD files for the lasercutting are almost done (would like to have more time to work on it, feels horrible slow atm).

Btw.: Even though the selectric keycap does fit the switches, it is slightly bigger then the other keys and therefore not very usable.

I don't have a thread post yet but I think I will make one now. One thing I would recommend is making the PCB. Once you get the hang of it, it is not the hardest thing to do. Also Placas suggestion above of a 3d printed card is a very good idea. I have actually used this idea in some of my variants. I only have a cnc machine and not a drill press so it doesn't matter too much, but the little tabs are not the hardest to get.