deskthority

Hi all, please advice me.... I'm still new in the keyboard world..... I am using a custom layout, a special ADNW version, which is a bit like, but not quite, Dvorak (*) I also use a few layers. I'll be using this keyboard at home (with Linux computers) and at work (with Windows computers). Later on I will build my own (symmetrical or matrix like) keyboard, but for now I'll settle on an "off-the-shelf" keyboard. The Leopold 660c looks extremely interesting. I have questions on 3 things however, before I "pull the trigger"

I'd like to program my own layout in the keyboard. I doubt that the Leopold firmware has the flexibility for that. Which means: putting a Teensy in it and writing new firmware. My main question: will Teensies work with Topre boards like the Leopold 660c?

Then I have two smaller questions, now that we're at it....could a Leopold 660c get somehow a smaller spacebare and more modifiers, 'Japanese style' ? And, are there keycaps for this keyboard on the market?

Thanks a lot. If you have other suggestions, fire along!



* After lots calculations & experiments, going from Qwerty to Dvorak, Colemak, passing through Carpalx, staying longer in the MTGAP-world , I have (for now) settled on a great custom ADNW-layout. ADNW is made for German & English; I used the ADNW software to come up with a version for my use case, which is Dutch & English. My layout is: [/size]

The Leopold FC660C is a great wee keyboard. Some Topre fans around here insist it's the best of the lot! I prefer the HHKB, myself, but the 660C has a smart layout for those who want to keep their navigation keys. Replacement caps aren't so plentiful for Topres, though. I haven't looked into the Leopold, but Realforces and HHKBs have very limited choice in custom caps. The Leopold is a unique layout among Topres, so I don't expect good news.

I presume you want to change the physical QWERTY layout to your own? That's a tall order. The reason is row profiles. Topres have caps that are different shapes on each row of the keyboard:



When you move a cap from one row to another, it doesn't match. Almost all mech keyboards are like this. Two exceptions come to mind. One is IBM buckling spring, where the caps are all the same shape but the board itself is curved. They're great for swapping to Colemak etc. And the other is flat profile caps, like DSA here on my NovaTouch:



Far as I'm aware, the only Topre switch keyboard that can take flat caps is the NovaTouch. Its quite different to the rest, as it takes MX caps like these and so many others. Including 7bit's group buys with their non-QWERTY packs.

The HHKB can itself be hacked with a Teensy. Hasu has some details on the process here.

You could do this to a HHKB Pro 2 or HHKB Pro JP, depending on which better suits your needs.

HHKB JP:

Oh yeah, I missed the Teensy part of the question. USB keyboards are typically bad news for programmability, but Hasu has a "USB to USB converter" somewhere in his firmware collection. I'm not sure it's particularly solid, but worth exploring if you want to go deep with layers and macros especially.

If all you're doing is changing the base layer (moving alpha keys around to match your preferred layout, and leaving other stuff alone) then you're better off simply using a software remap in your OS. Depending on whether work lets you install it…

Thanks Muirium and Scottc. OK, now I have a few options:
- the HHKB Pro JP + Teensy + Hasu firmware. Or the Leopold 660c. Or other Topre based keyboards.
- the Unicomp, If I would silence it a bit it could be great, thanks for pointing me at that
- or just go all the way and build my own. I would use MX switches, and maybe the Filco doubleshots (all the same row profile, SA)

Choices, choices

the leopold 660c is indeed hackerable but we haven't figured out yet how. I already have the schema of the various chips



but I need a logic analyzer to read what the controller does at startup. If we get that part sorted out we could use the same HHKB firmware that Hasu is using... with an added benefit: we could fine tune the actuation point that in the 660c is not hard coded.

@matt3o - "we could fine tune the actuation point that in the 660c is not hard coded" I thought the actuation point depends on the physics of the switch, that is: on the hardware ? Could you explain how this works in a 660c ?

The capacitive switch is a potentiometer, theoretically it would be possible to change the actuation point (with a 64 points definition). Practically the actuation point is factory hard-coded.

There's a realforce prototype with a gauge that you could use to vary the actuation point but it never turned in a real product.

The HHKB has the actuation point hardcoded in the main PCB, while on the 660C the actuation point is initiated on startup from the daughter board. Since we are going to completely replace the daughter board to hack the 660c it would be possible to initiate the actuation point in any of the 64 points.

Hope that makes sense.

Yes, makes sense. Grazie mille Matt3o.

BTW, this yet another "mysterious closed source case", where I don't see any profit at all for the manufacturers (Leopold, FPU) in keeping the code secret. What would they lose if they opened the code? Nothing!

the leopold 660c is indeed hackerable but we haven't figured out yet how

I laud the efforts you and others are making Unfortunately, I don't have the skills (yet) to help in this field. Guess I'll restrict myself to a keyboard that has been 'hacked' already, the HHKB - or probably the HHKB JP since I like thumb style keys.

The HHKB JP has a 0.25/ 0.25/0.25 stagger. But I don't care, to me all staggers suck, especially on the left hand. Only a symmetrical stagger would be acceptable like the μTron. The AdNW guys thought the same about stagger, since AdNW layouts have no important keys on the left side of the bottom row. Stock AdNW has xßöjä ,my variant has xßöjä

I do see one or a few custom made keyboards in the future though Thanks for enabling this madness