Turning a typewriter into a 60%!

[Masterchief79]

Hey guys,
I got a Triumph Adler 8008L electronical typewriter with nice thick beige-on-black doubleshots keycaps and Cherry M9 switches (clear 10° angled stem, linear, not MX compatible). In fact, I like it so much, I decided to turn into a 60% PC keyboard - and you are here to help me with that x)

I don't think I will be able to keep the original PCB (cause there's all kinda stuff on there - voltage regulators, ICs, memory etc. for the typewriter), and it's way too big with all this, too.

So, first step will be: Find a PCB!
Problems:

• maybe but not necessarily: the height (M9 are about 1cm higher than MX)
• The switches really have different pins. You couldn't just desolder one from the QWERTY-row and put it besides the spacebar, they have different soldering points.
• therefore not MX-compatible either (coudn't just fit them on a MX PCB without fiddling around alot - they're platemounted though, so maybe I could glue the switches into the plate and connect them with the PCB underneath with little wires?)
• If I go ahead and design my PCB myself - no idea how to make a proper controller and/or software for it.
• case and everything will be custom, too - but all that stuff comes later, so let's stick with the PCB first shall we

No Problems:

• MX and M9 have pretty much exactly the same distances from key to key. Stems are both about 20mm's apart and keycaps are both 18mm's wide. So they're comparable there at least.
• everything which has to do with soldering. I have decent soldering equipment at home and already soldered voltage modifications on so many graphics cards I can't count them anymore
• new PCB = no fiddling around with the old one, such as sawing it in pieces, repairing them soldering paths etc. pp.
• the Layout: it has pretty much a perfect 60% layout since on a typewriter, you basically just had the typing field + modifiers. Even came up with proper FN layers.
• funcionality: over 30 years old but every switch is working, so I won't be forced to replace any.

I shot a little video about it for the Deskthority wiki, you can get an idea about the whole thing there (not M6 I know that now):

feature=player_embedded[/u]

That's it, hope it's not too much for the beginning, I'm counting on you guys

[mkawa]

this is an _awesome project_. it's like half a teletype. i'm 100% behind this. let me know if there's any way i can help.

here's how i would start: carefully cut the traces between the passive and active components on the existing PCB and the switch matrix. be _extremely_ careful. i would personally not replace the PCB, because it looks so darn good. instead, i would tap into the switch matrix at appropriate points and wire that to a separate PCB which is around the size of an arduino uno or smaller that does all of your electronics, and leave the existing PCB with gigantic passives for looks. but completely unpowered and unused. to make this really safe, it is best to cover the capacitors in a layer of conformal coating in case the cases degrade and there is some danger of the electrolyte leaking. everything except for the capacitors is basically inert and can just be left where it is. i love old boards and their gigantic passive components though. in addition to looking cool, they look FREAKING COOL.

note that you may not need to cut any traces. the existing power supply and controller on that pcb may be set up in such a way such that they do not interfere with your circuit when not powered. you will have to do some experiments to determine whether this is true. you don't want, for example, for those gigantic capacitors to sit around storing up current that you're trying to flow across the keyboard matrix with your tiny modern electric devices.

i would also do the following hilarious thing: take the print head out of the typewriter case (it has nasty toner/ink in it), but leave everything else. and use your 60% board in a 1000% case

YESSSS!!!!!!!!!!!!!!! <-- this is the only reason i have for doing something so ridiculous.

[BlueBär]

i would also do the following hilarious thing: take the print head out of the typewriter case (it has nasty toner/ink in it), but leave everything else. and use your 60% board in a 1000% case

Why not put a RasPi or Cubieboard inside and make a sort of USB splitter, so you can use it with the RasPi as a terminal or sth. similar and, when the RasPi is powered off, just as a keyboard?

[Halvar]

If you don't want to keep the full PCB and you're good at soldering, since the switches are plate-mounted, you could do without a PCB. Take a look at this for inspiration:

http://deskthority.net/workshop-f7/brownfox-step-by-step-t6050.html

[Masterchief79]

@mkawa really glad you like it so far.
What I forgot to mention: This will be rather a long-term project since I'm in university at the moment and only visit my parents once a month (that's where all the equipment is) - so I will have 3 weeks of planning and getting stuff followed then 1 week of getting the stuff done Next visit will be around 24th so let's see what we can do till then

At the moment, I'm trying to figure out how the PCB is working. The capacitors are connected to the switches in one way or another, so I guess I can't just saw them off. I followed the conductor paths as far as I could (pretty difficult cause a lot is underneath the plate). I think, the problem is, that the speaker, memory etc. and the keyboard itself all have the same power source. On a PC keyboard you would have power supply over the PS2 port or anything. This one has a 6 Pin just for sending signals, and Power comes directly from the wall socket (230V in Germany - that's the red plug top left) and is spead all over the keyboard afterwards. There are some voltage converters though, so I guess the whole thing still operates with 12V. I'm not in expert in these things so this is important to understand for me.

So, I think we need a plan on how to build an alternate power circuit with 12V input over PS2 or USB port. Or would a Teensy or something already take care of that?


By the way, of course the big capacitors and things look awesome, but on the other hand, I don't want a 60% keyboard which is wider than my fullsize Qpad So I don't know, let's see how the project goes first, but at the moment I would rather cut them off.

@Halvar: Wow, that's a possibility I would have never thought of, thanks
First thing that comes to my mind though when I see that, is that wiring under the keyboard would make the whole thing even higher when done. As I said, even without a case, it's more than 1cm higher than my QPad. I see it coming, in the end I will have to cut a hole in my desk for sure  But maybe I could place the controller somewhere between the switches and not underneath (there is about 6mm space between PCB and plate)... I'll keep it in mind and think about it, would be kinda the ghetto solution.
Edit: Just asking, I realised those diodes are kinda expensive (1€ each? If I'm not mistaking...) I could use the ones from a G80-1000HFD though.. hmm...

[BlueBär]

Edit: Just asking, I realised those diodes are kinda expensive (1€ each? If I'm not mistaking...) I could use the ones from a G80-1000HFD though.. hmm...

More like 1€ for 100: http://www.conrad.de/ce/de/product/162280/Universal-Silizium-Diode-1-N-4148-Diotec-1N4148-Gehaeuseart-DO-35-IF-150-mA-Sperrspannung-UR-75-V
Those should do.

[Masterchief79]

Lol. Thanks, nearly thought so. ^^

Well, at the moment I will have to decide for a direction I want to go with this to.
I'm seeing the following possibilities:

• Keep the original PCB including fat capacitors, speaker etc. and "just" strap a teensy to the matrix of the original controller (if I got it right till now).
• Desolder the switches and go "Brown Fox"-style with the plate (what Halvar linked here).

First one: Not so much my favorite, many of you seem to like it, but I actually wanted to get this as small as possible (for no special reason, I just imagine the "usual 60%" style but with M9 switches and those gorgeous caps to be so cool). Making a case would be even more complicated with all the extra parts still on there.

Second one: Seems to be easy and quick, and I get about what I imagined when starting this project. I am still a bit unsure about the height, especially with an extra case + wiring under the plate but I'll find a solution for that, too.

Just got a case-idea for the first way though: What about making a standard 60% case (wider of course, so it covers all of the parts), but not increasing the height. As you may have noticed, the capacitors and stuff are sticking out maybe 1cm from plate level. So usually I had to lift the whole top of the case so it goes above them. But: I could leave the top of the case maybe 2-3mm above plate-level as normal, and cut holes in the top of the case so the capacitor stuff sticks through a little bit? That would be pretty cool indeed. Have to think about that.

For the second way (going hardcore 60%) I imagined something like this without an upper case part at all (standalone plate? No idea how you call this style): http://i.imgur.com/3G9kusd.jpg

[dorkvader]

this is an _awesome project_. it's like half a teletype. i'm 100% behind this. let me know if there's any way i can help.

here's how i would start: carefully cut the traces between the passive and active components on the existing PCB and the switch matrix. be _extremely_ careful. i would personally not replace the PCB, because it looks so darn good. instead, i would tap into the switch matrix at appropriate points and wire that to a separate PCB which is around the size of an arduino uno or smaller that does all of your electronics, and leave the existing PCB with gigantic passives for looks. but completely unpowered and unused. to make this really safe, it is best to cover the capacitors in a layer of conformal coating in case the cases degrade and there is some danger of the electrolyte leaking. everything except for the capacitors is basically inert and can just be left where it is. i love old boards and their gigantic passive components though. in addition to looking cool, they look FREAKING COOL.

note that you may not need to cut any traces. the existing power supply and controller on that pcb may be set up in such a way such that they do not interfere with your circuit when not powered. you will have to do some experiments to determine whether this is true. you don't want, for example, for those gigantic capacitors to sit around storing up current that you're trying to flow across the keyboard matrix with your tiny modern electric devices.

i would also do the following hilarious thing: take the print head out of the typewriter case (it has nasty toner/ink in it), but leave everything else. and use your 60% board in a 1000% case

YESSSS!!!!!!!!!!!!!!! <-- this is the only reason i have for doing something so ridiculous.

speaking of which a selectric terminal recently sold on eBay:
http://www.ebay.com/itm/121189196399

I thought it'd go higher, given the rarity and relative awesomeness of these. Still, he forgot to put IBM or selectric in the title, so most people probably didn't take notice.

Well, at the moment I will have to decide for a direction I want to go with this to.
I'm seeing the following possibilities:

• Keep the original PCB including fat capacitors, speaker etc. and "just" strap a teensy to the matrix of the original controller (if I got it right till now).
• Desolder the switches and go "Brown Fox"-style with the plate (what Halvar linked here).

First one: Not so much my favorite, many of you seem to like it, but I actually wanted to get this as small as possible (for no special reason, I just imagine the "usual 60%" style but with M9 switches and those gorgeous caps to be so cool). Making a case would be even more complicated with all the extra parts still on there.

Second one: Seems to be easy and quick, and I get about what I imagined when starting this project. I am still a bit unsure about the height, especially with an extra case + wiring under the plate but I'll find a solution for that, too.

so

about this mod: it sounds really great! At first I thought you were after doing mechanical typewriters, but converting an electric will be just as fun, but with the added bonus of possibly making it into a printer / teleprinter. You could do this with a wheelwriter by just intercepting the matrix, I think. How exciting!

So, if I remember right, your typewriter is PCB mounted, so you can't just intercept the matrix as easily. I'm drawn to the first option of yours, but then I like the challenge, and I prefer sticking close to the authentic design. It's up to you, but #2 is a fundamentally different type of mod. I think perhaps people prefer #1 because you can just make a 60% using normal cherry mx switches and caps. You can't just make a typewriter any day, so the conversion method is pretty cool. #2 is great, but only if you really like those switches.

Another idea for #2 is that you can potentially "tap" the matrix out of your custom wire job and smd pigtail to the PCB and add a connector in the middle, and plug your KB into the typewriter again and it'll still work. This is, if you keep the matrix the same.

[mkawa]

logic level on teensies and in general are either 5v or 3.3v. 12v is to power motors, like the one your print head is connector to, hence the 12v input into those big capacitors. don't have time to read more, but do NOT try to run 12v over ps2 or USB. you will have a bad time.

[Masterchief79]

so

about this mod: it sounds really great! At first I thought you were after doing mechanical typewriters, but converting an electric will be just as fun, but with the added bonus of possibly making it into a printer / teleprinter. You could do this with a wheelwriter by just intercepting the matrix, I think. How exciting!

So, if I remember right, your typewriter is PCB mounted, so you can't just intercept the matrix as easily. I'm drawn to the first option of yours, but then I like the challenge, and I prefer sticking close to the authentic design. It's up to you, but #2 is a fundamentally different type of mod. I think perhaps people prefer #1 because you can just make a 60% using normal cherry mx switches and caps. You can't just make a typewriter any day, so the conversion method is pretty cool. #2 is great, but only if you really like those switches.

Another idea for #2 is that you can potentially "tap" the matrix out of your custom wire job and smd pigtail to the PCB and add a connector in the middle, and plug your KB into the typewriter again and it'll still work. This is, if you keep the matrix the same.

Thanks for your reply. I will definitely keep those switches and caps. Firstly they feel great, secondly I will have a very unique keyboard, and thirdly the SOUND - it sounds soooo awesome typing on them. This is proably because the caps are so thick and massive. I in fact have different caps which also go on M9 switches (got them in a grab bag) and they don't sound near as awesome. And they are already thick DSA 2shots.

Deciding for one of my ideas:

1. Keep the original PCB including fat capacitors, speaker etc. and "just" strap a teensy to the matrix of the original controller (if I got it right till now).
2. Desolder the switches and go "Brown Fox"-style with the plate (what Halvar linked here).

As I said, the second one was kind of my favorite and it still is. Keeping the wheelwriter and the matrix intact, so I can still use it as a teleprinter, would be pretty awesome indeed. But I think I won't do this in my first project. An idea would be: Doing my second idea as I said, with a wirejob. I could keep the actual PCB intact though and reinsert it in the original typewriter. Then I could make a cable, so I could keep the 60% keyboard on my table and on my PC, and switch over to teleprinter mode if I want to. At the moment I don't see a reason why this shouldn't work, but as I said, if I do something like this it will be later when the actual mod is done.

So guys, I know the PCB is quite the show, but nevertheless I'm afraid I will have to replace it - sorry . I will stick with my second idea and do a wirejob similar to the brown fox method which is linked above. It seems much more simple (keep in mind it's my first real mod), plus I would be able to finish it pretty quickly (hopefully). Making a case is still pretty far away but I will figure out a way for sure. Keeping the PCB, reconstructing the original matrix, and making a case for the whole thing sounds cool, too, but is much more abstract and I have no idea how to do most of the work steps there.^^

For now, I am pretty glad if I can substantiate my main idea and turn this into a nice little 60%.


Next steps:

• Order the necessary parts to do all the soldering work on the electronics. Diodes, small Teensy (something else?). Soldering equip, wires, shrink tubing and the plate and switches of course are all present.
• Getting the stuff together. I will most likely just use the Brown Fox project as a tutorial and will ask you guys if I have questions.
• Software. No idea how exactly to do this yet, but won't be that a big problem. Just need a few tutorials and probably your knowledge.

Do I need special equipment for getting software on the Teensy? Don't think so, right?

[Masterchief79]

Update:

Desoldered all the switches today, took like 1-2 hours, pretty okay. Looks like I didn't damage any of them.
There are funny ones which have 4 pins instead of two, a second spring on top of the actual one, two leafs and two sliders. So, pressing it down gives out two signals. Somebody ever heard of that before?

Now cleaning the plate and coloring it - decided that steel grey isn't hip enough. Thinking about red or beige (contrast color or color going with the caps).

Case idea (and cool because I could actually do that and I wouldn't need months of CAD work and a 100000$ CNC milling machine): Ima make a wooden case with a black leatherette coating. No idea how this works out in the end though. xD

I order the teensy tomorrow and hope I can finish soldering work next weekend (only have the weekends to work atm).

Gonna take some pictures and edit them here stay tuned, won't take long

[CPTBadAss]

Pictures?

[BlueBär]

There are funny ones which have 4 pins instead of two, a second spring on top of the actual one, two leafs and two sliders. So, pressing it down gives out two signals. Somebody ever heard of that before?

Yep, Haata has found some of those in vintage boards. Very rare.

[MTManiac]

Keeping a close eye on your progress, I have an old TA Alpha 610 with M9 switches I need to turn into a 60%

Now I feel the need to check and see if I have any 4 pin switches... hahaha

[Masterchief79]

Here you go:

[BlueBär]

That double clicker. Want.

[Masterchief79]

There were four of them on the typewriter if I'm not mistaking I don't know what awesome things you can to with those...?!

PS: Also got the diodes today.

[BlueBär]

There were four of them on the typewriter if I'm not mistaking I don't know what awesome things you can to with those...?!

Sweet! Maybe if you put one on the backspace you could make it so that if you press it harder it actually sends the signal for Ctrl + Backspace to the PC?

[Masterchief79]

I didnt realize the two leafes were actuated _after_ each other and not at the same time. But now as you say it... Definetely makes sense. I just measured those, and, yes indeed, that would be possible and opens up awesome new possibilities (as long as if I doesn't fail on coding these things. No idea how the restrictions are there).

Ctrl+Backspace, FN and FNLock on the same key, Ctrl+Enter, maybe Alt+Tab even?

[BlueBär]

(as long as if I doesn't fail on coding these things. No idea how the restrictions are there).

Ctrl+Backspace, FN and FNLock on the same key, Ctrl+Enter, maybe Alt+Tab even?

Just wire the second activation as an extra key of that row and everything should be fine.

I don't know how easy the second click happens, so think twice before you put it on a key that you tend to "smash" (I'm thinking about the enter key here). Also keep in mind that the 2nd activation happens after the other, so on pressing enter it would actually send "Enter - Ctrl+Enter" to the PC.
Maybe Tab and Shift+Tab?
With Fn you could also do 2 function layers on 1 key which could be nice

[Masterchief79]

Of course the second activation goes off if you "smash" the key, but I think you could hit backspace pretty well without doing Ctrl+Backspace by accident.
Mind explaining the "two FN layers on one key" idea?

I thought about a layout and I think I'm gonna do it this way ("Druck" is printscreen in German). It's basically a German ISO layout. It's all pretty complicated, because I have (no ****) 6 different kind of stock M9 switches on this thing. On top, the heavy switches have a lower stem for the modifier row so I can't use them where I want (I hope I can find a solution for the Esc key in the top row). Changes I've made and thoughts behind them:

• Esc in the numer row (usual 60% stuff).
• There is another key on the right of the right shift, so I made that my FN key. FNLock is underneath, cause that's an LED switch.
• Wasn't sure about those Track fwd and Play music keys on the bottom right. But I don't need another Winkey or Ctrl, so I thought I'd move play music and change tracks from the FN layer to the standard layer
• Backspace, Printscreen, Alt and Space get a Double-Activation switch. Backspace: Like the Ctrl+Backspace idea a lot. Printscreen: Alt+Printscreen should work, so pressing the key harder will give you a screenshot only of your current window. Alt: Alt+Tab combined on one key will be neat.^^ Space: This is the Double-activation switch with the non-angled stem which can't really go anywhere else.

Excel file is in the attachments, feel free to edit.


Edit: Just thought about something: When i Press Backspace for example, I get a backspace signal. If i press it harder, I get a Ctrl signal additionally. But that's Backspace+Ctrl and not Ctrl+backspace. And Backspace+Ctrl doesn't work.^^

[CPTBadAss]

Whoa, those pictures are really cool and I'm fascinated by the progress being made here. The switches look really interesting and I've love to get my hands on some. And using Excel to make a layout is genius. I can't believe I never thought of that before! Can't wait to continue following your progress .

[BlueBär]

Mind explaining the "two FN layers on one key" idea?

Pressing the key -> FN layer 1
Press harder -> FN layer 2

Edit: Just thought about something: When i Press Backspace for example, I get a backspace signal. If i press it harder, I get a Ctrl signal additionally. But that's Backspace+Ctrl and not Ctrl+backspace. And Backspace+Ctrl doesn't work.^^

I don't exactly know which code you will use but it is possible to send a different keycode when two keys are pressed at the same time - thats how a function layer works. It should be possible to adjust the code to do what you want. If you send me the code I could look into it.

[Masterchief79]

Wow okay, that's surprising, especially because I really suck at Excel
I'm glad you like it, I hope I can start soldering work today and finish it next weekend with the teensy and everything.

Thinking about using a Double activation switch on the Winkey, so I get Win+D = Desktop if I press it 100%. Dont really need it on Printscreen and I have my doubts anyway (as I said, if you want to go Alt+Printscreen, Alt has to be pressed first... ) I just have to figure out if I can use the heavy switches in the top row without problems (remember, the stem is ~2-3mm shorter).


Edit@BlueBär:
Ah right, interesting, so I could basically use one FN key for two function layers. Have to think about it. This idea would carry another advantage: I could use a regular heavy M9 switch for the FNLock key, and depending how hard I press the FN key itself, the FNLock locks either FN layer 1 or 2. Interesting! Just that I have no idea what second FN layer one could possibly need. xD

I have no idea what code I will use either, that's something I will definitely need your help with. I got the base codes from hasu's github, but I dont even know how to open any of those files, not speaking of compiling or anything.

Edit2:
New pics from soldering. I hope I did everything right so far. After finishing I realized I have to swap two switches but the rest seems okay.
The wires of my diodes were too short, so I had to connect everything with enamelled wires. Those are just usual wires, coated with a thin layer of paint which isolates them. So, thats not copper you see there. You can basically remove the insulation (coating) with solder. I usually use those for my voltmods on graphics cards.

[Masterchief79]

Bump for new pics and progess.
Maybe have to glue the switches in, plate not very tight there. Nevertheless satisfied how this goes.

[BlueBär]

Nice work so far, I really dig the orange plate

[Masterchief79]

So guys, update, I wired all of the undone diodes today and have to connect all the stuff to the teensy now. I have no idea how to do that, the Brown Fox tutorial doesn't help a lot (he's talking about like B0 and F7 and all that stuff and the pinout on the Teensy is completely different lol). Also keep in mind I have the Teensy 3.0 version. Would be nice if you could help me with connecting the rows and columns to the teensy.

[Neebio]

Disclaimer: I've never worked with a teensy or wired up a keyboard like you are doing.

The teensy is a microcontroller with pins exposed to allow the user to do with it what they wish.  It shouldn't actually matter which pins you connect your rows and columns to, so long as you avoid reserved pins like +V and GND, etc.  Reserved/special pins will be listed on a teensy spec sheet.

The pins you've connected your rows and columns to become relevant when you program the teensy, since it needs to know which row and column it's looking at as it scans the matrix.  So really, you could attach your rows and columns to whichever available pins you want, so long as you set up the teensy programming accordingly.

That said, it would be wise to consider what firmware you are using, or if you are writing your own for the teensy.  If you are using pre-existing firmware, you may have your question answered for you by simply wiring the rows and columns to the teensy as the firmware is expecting them to be.

[Halvar]

I agree -- you should probably first decide what software you will be using on the teensy and read its documentation to find out what pins to use.

A new alternative to Hasu's code would be Soarer's -- which isn't Open Source but probably easier to use and configure.

Hasu:
https://github.com/tmk/tmk_keyboard

Soarer:
http://geekhack.org/index.php?topic=50437.0

[Masterchief79]

So guys, time for me to dig up this one. I temporarily gave up on this KB as I found out I most probably can't use my Teensy 3.0. I stuffed the whole thing in the closet and concentrated on other things, like expanding my graphics card collection (which now counts more than 100 cards).

Whatever, so, I'm back. Short conclusion of work so far: I dismounted the Cherry M9 switches from my old Gabriele Adler Typewriter. They were plate-mounted, so I also dismounted the plate and out everything back together. Pics are a few posts up. Diodes and the matrix is soldered, so next step would be connecting this stuff to the teensy.

First of all, lets make sure this Teensy3 I bought is really not usable for me. If I got this right, the prob is that it uses an 32bit ARM CPU instead of an 8bit AVR like Teensy2. So every code on the web ever for Teensy2.0 = useless on 3.0, right?
Would be nice to make this clear before I order something new

Greetings

[Grendel]

Yes, the 3.0 and 3.1 are ARM based. If you want to dip into what's available on the net it's probably easier to use a Teensy 2.0 or Teensy++ 2.0 (more I/O lines, more flash/RAM over the plain 2.0).

[Masterchief79]

Yeah I'm afraid I'm not much of a programming nature... software has never been one of my talents. I'm rather the hardware guy. So it's difficult for me to do a Teensy 3.0 Keyboard programming from the ground up all by myself. A plain Teensy 2.0 should be enough for my purposes though, 60% KB doesn't have that a big matrix. Guess I will see if I can get one and then we can get down to the real stuff

[jacobolus]

I believe HaaTa has some keyboard firmware which will work on a Teensy 3.0, but you probably just want to get a Teensy 2.0 (or similar controller) and use something like Hasu’s tmk_keyboard firmware.

[HaaTa]

Replied to your PM Masterchief79.

Yes, my firmware supports 3.0/3.1 (also 2.0, but hasu's and Soarer's firmware is much more tested than mine). I haven't done any matrix scanning with the 3.0/ARM code yet. But since I already have code for the 2.0 that does this, it should just be a bit of porting work.

The 2.0 will be easier, but the 3.0/3.1 is completely doable (just will take a bit longer). I also need more excuses to expand my ARM support

[Masterchief79]

You've got some excuse right here.
You got a reply too. No prob if it takes a little longer for you to help some guy with his little keyboard modding project. However I got a lot of freetime atm and should be able to make things happen quite quickly.

[Masterchief79]

... if I can find a cable for the Teensy 3. Will update when I have one. >.<

[Masterchief79]

Okay so beware, this is pretty random: Last post 9 months ago said I would update when I'd find a cable. I found one in the parking lot a few days ago, so here we go.

Custom mechanical made from an old electronical typewriter. It uses linear, super-smooth, ca. 30 year old Cherry M9 switches wired to a matrix and hooked up to a Teensy 3.0. I used HaaTas "Kiibohd" controller to get this badass 32bit ARM chip running and I only annoyed the Senpai himself for about 6 hours everyday for the past week (nah for real, could not have made this keyboard without major help by HaaTa. All credit for the code goes to him and thanks again for showing this little newb here how compiling in Linux works - also to dfj ).

Keycaps are really thick doubleshots which I took from the typewriter obviously. Keyboard is functional at this point, I'm just typing these lines on it. Little problem with double keypresses but will figure that out later. As the typewriter was called "Gabriele Adler 7007L", and this is a 60% keyboard, I call it the "Adler60".

KB has no case yet and is really just plate and switches to this point, but some keyboards are barely more when finished, and I think it's already looking pretty yummy. I'm tempted to make a wooden case for it but I will decide that later^^ Concerning the cable, I will hide itof course, in the pics it's really just to test the whole stuff. Layout is ISO-DE.

Open for comments, questions, anything. Pic-time:

[Oobly]

Hot diggity dang! Those keycaps are amazing! Great project and nice to see it's functional. I'd love a chance to try typing on it, even though I'm not such a fan of linear switches.

I'd love to get hold of a bunch of those double contact switches, though, for velocity sensing when typing.

[CPTBadAss]

I really like those keycaps. Someone pointed out in another thread about Cherry MLs that they're shaped to reduce off-centered strikes. I think it's clever and they look cool . Now post a video of you typing on that board!

[Masterchief79]

Yep they look awesome, and they're the most massive keycaps I ever laid my hands on Will do a vid, too.

Velocity Sensing when typing is a really cool idea, too.

[Nover]

Looking thick, solid, tight. Great job on that one, have stumbled over a couple of those Adler 7007 thinking theyd look nice as a keyboard. It looks even better than imagined. 

[Masterchief79]

Thanks a lot
Unfortunately it's rather hard to take good pics with the camera I have (it needs tons of light to take decent ones, and now in winter it's almost impossible without flash).
Made a little video as requested by CPTBadAss^^

www.youtube.com/watch?v=u5Dh5pF354I&

[dorkvader]

Thanks a lot
Unfortunately it's rather hard to take good pics with the camera I have (it needs tons of light to take decent ones, and now in winter it's almost impossible without flash).
Made a little video as requested by CPTBadAss^^

www.youtube.com/watch?v=u5Dh5pF354I&

I usually have "too much" light in the wither reflecting off all the snow. I only use a phone camera, but I have to set the ev and wb manually. Really annoying.

But you're right: when it's overcast, it's like right in the middle where using the flash is worthless but there's not enough light to get a good picture otherwise. I used to use a 1500 lm LED lamp in this case 'cause I'm more interested in getting all the details than accurate colour.

typewriter sorta looks ilke this one. I might have missed it but is it cherry switches?
http://imgur.com/a/JGEOB

[muxe]

Hi, I was very inspired when i saw your photo first time, so... I bought my own Gabriele to build my own "Adler60" and it is great!
The keys are very smooth and looking really good.
Thank you very much! The only problem I have is: the case

Is anybody here with an idea for a simple case protecting at least the bottom of the keyboard? My first approach was to simply put my keyboard on wooden sticks so the wires and the controller don't touch the ground, but a plate on the bottom would be nicer, I'm just not sure which material to use and how to attach my keyboard on it...

now i am reusing the original typewriter case, but it's very big and hiding the red color of the plate

[Masterchief79]

Wow, I just logged back in to ask some more questions and saw your posts, that's so great! Looks awesome, the red fits even better imo, and it's fantastic I inspired someone with my idea

So, the reason why I came back to dig this thread up:

After I got the keyboard into the state it was at my last post here, I kinda discontinued work on it and got into other projects. I used it a bit but as it wasn't really finished, it was still a bit impractical. Then unfortunately my teensy 3.0 broke.
I want to get the thing repaired and finished now. Questions:

• Which teensy version should I get? They are up to 3.6 now?
• Are they compatible within each other? Will KiiboHD run on every one? (It's already 32bit-based as the teensy 3.0 had an ARM CPU)
• Any tips on how to recreate all the software environment stuff to flash another teensy with the 32bit KiiboHD software I used back then? Still have it around, just would need to create this linux VM I guess, hook up the new teensy and solder it back into my Adler60. I did have some problems building and flashing the software the first few times so that might still be tricky. Would be cool if we could minimize the difficulty on this, I'm NOT a linux pro.
• Is there maybe a better software solution or a more advanced version of KiiboHD to start with now?

Things on my to-do-list, just for the record:

• Rewire the thing, I had some signal issues (might have been the cable) and removed the wiring between the switches
• Get new teensy, flash, solder in and reprogram if needed (note: keymap was alright but not 100% finished back then)
• Find teensy position that's not putting force on the USB plug and fix it in there properly (note to self: DO NOT use doublesided tape again)
• Build proper case (note: I should have a variety of CNC machinery and stuff at my disposal as I'm studying mechanical engineering now - not sure if needed)
• Add some LEDs for capslock etc, maybe lighting

[orpheo]

I used this set and did a 60% with it with old cutting the space bar, it has spare stems for 7U spacebar, what a luck! Mine is French Gabriele 9009 though. The trick is to turn it into ANSI enter and shift, and you need two 9009 keycaps set to achieve this. Let me find a photo for you.

EDIT : I am silly, I have the Gabriele 9009 ! And it has old Cherry switches that are OK compatible with MX

Here is my Satan PCB, but spacebar, and cut stabs stems for Enter and Shift keycaps.