rodtang wrote:When I first saw this I thought it was gonna be another person who will be here one day to show some pics of his awesome stuff and never be seen again but this is some really great info, keep up the good work.

You gotta make 'em stay, at least those that come bargin' in with stuff like this!

And regarding I2C vs SPI, don't see him needing those two extra lines I2C would grant on a usb bus, but seeing there's already code out might make it interesting nonetheless.

Nice thumbbuttons on the split version of your first (first pic in OP at least) board though ML's (and I assume you didn't make it with so much space that MX's including caps can fit nicely) and on that single right handed design in the bottom, does that frontmost button actually sit slightly lower?
Now i'm starting to get curious, you didn't by any chance make your ML design with MX spacing or did you?
hehehehe, you DID! N1.

just (quite) a few cuts and it can fit switches that bind less

Also, regarding your giveaway, I'd definitely say get them to those that got you to make them in the first place; on second thought, getting them actually used might be best, so make sure they do.

edit1: Oh and Dox, nice to see you around here!
*ofc you don't know me, but I've lurked enough GH threads to know you/r work; nice!

edit2:
[ERGO-RANT_START]
Just to make it clear I'm not begging to nab your designs (nor ergodox/key64); I'm definitely gonna be ending up with some type of ergoboard, but the first step in that is trying to find nice relaxing fingerpositions; ofcourse using the work that others have already done; basically reinventing the wheel while building on the shoulders of giants; so when I read that you'd been going back n forth with colleagues about ergonomics, it really peaked my interest, but didn't yet want to ruin the thread with talk on that; now it's that time though:
Tell me, what have you learned so far from your experiments and all that you've read on those matters?
I'm still even attracted to a convex board for the fingertips; initially thought your latest had that until I saw your hand on it, and decided that having the front slanted for thumbs is more essential; yet it's still alluring to test out, even in the face of the ergonomics that the DataHand set forth, namely, short movement using the fingertip's surroundings.
I know that convex could potentially make it very tricky to actually press down on thing without having to bring your fingertips back to those points (i.e. curling your fingers in).
[ERGO-RANT_END]

Well, I wasn't so sure when you mentioned it but you're right, I first had ML designs with wider MX spacing before shrinking it down and thus making it a lot less desirable.

while you're taking pics, could you do a sideshot of that one?
'cause it's a true mindfsck for me atm; the top pic vs the bottom one 'render' completely different things in my head xD

noted that small (custom?) caps are required for that thumb-block on the middlepic.

Wow... beautiful keyboards totally. Great hacks!
I'm very inspired.

off wrote:while you're taking pics, could you do a sideshot of that one?
'cause it's a true mindfsck for me atm; the top pic vs the bottom one 'render' completely different things in my head xD

noted that small (custom?) caps are required for that thumb-block on the middlepic.

Sure, here they are (please excuse the bad quality): I put a pen under it to show its warp, it was of course intended to be tilted upwards with the pinky row staying on the desk. But in this position, the "X" on the thumb and the center middle finger "#" are level with the ground, while other columns are offset in two directions and the outmost ones also tilted slighly.
All in all a promising idea, but from my experiences overly complicated...



I also made keycaps for both ML and MX, but only very simple flat ones in different sizes you'll notice in some of the pictures - but switched to designs using standard cherry ones as that was another lesson learned: the well-shape and very smooth finish helps a lot in constantly re-orienting your fingers in the right position. It was more difficult (but not impossible) to adjust to the rougher flat surfaces of for example above winged design.

Here's an old blender source with several prototypes - sorry, I haven't got access to the warped, smooth one atm but I'll post it when possible. These are simply demonstrators, no guarantee dimensions are exact and tested for this file...

suka wrote:

uploadfromtaptalk1335475975016.jpg

That one still screws with my head. Thanks for showing the side though, that really shows the way I 'see' the first pic of quoted post correctly.

What do you mean by "the warped, smooth one"? A diy keycap (different version of the one in the top right of that blender capture)?

off wrote:What do you mean by "the warped, smooth one"? A diy keycap (different version of the one in the top right of that blender capture)?

Sorry, getting confusing with all the models : I mean the Black one above with the hole for the thumb

suka wrote:

off wrote:What do you mean by "the warped, smooth one"? A diy keycap (different version of the one in the top right of that blender capture)?

Sorry, getting confusing with all the models : I mean the Black one above with the hole for the thumb

Hihi, guessed you meant a keycap.
And you can just press 'delete' on a post if you feel so inclined, somewhere next to edit in the bottom.

great, i am speechless.

could you give some technical information about the trackpoints you use?
You use IBM trackpoint you stated, but i like to know more about pin assignment and so on.

I used two different trackpoints from scrapped laptop keyboard replacement parts: While different PCBs are used, both employ a design as described in a pdf from Philips "Microcontroller with TrackPoint microcode from IBM" (TPM754) that I found on http://www.datasheetarchive.com/TPM754*-datasheet.html#datasheets. The serial numbers on the PCB did not really help as I found no references to them anywhere, but the TPM754 (or its A revision) seem to be used in most TPs I have come across so that should be a good start

I am not sure 100% about the larger, left one in the pictures, but I think it also had the same pin assignment as the lower one I recently built into my keyboards:

red: RESET
brown: unknown
black: DATA
white: unknown
gray: GROUND
purple: VCC +5V
blue: CLOCK
green/yellow/orange: 3 Buttons

It can then be registered in my firmware as a normal PS/2 mouse, or most likely even directly attached to a PC if you solder a plug to the appropriate pins and attach some buttons.




Excuse the bad picture quality, if you need more details feel free to ask.

Useless bump for justice.

You can get specification of TrackPoint here.
http://wwwcssrv.almaden.ibm.com/trackpo ... nload.html

You may be interested in TrackPoint specific PS/2 commands described in this spec. It seems you can invert axis, calibrate and even configure ADC with the commands!

Thanks, hasu, for reminding me - I already knew about the documents there, just never bothered to delve deeper into the matter as everything was working quite well, or so I thought. But now I couldn't help but try to implement a PressToScroll feature to get rid of my nasty manual scroll wheel emulation, and it works wonderfully!

So what does PressToScroll do? Does it switch to scrolling if you press it harder or something?

Yes, the track point now acts as both horizontal and vertical scroll wheel if pressed down harder than normal - sweet!

Sweet indeed, damn!

That sounds very nice! And you read additional force data with your custom controller to do that?

No, only had to set some register in the TP via PS/2 commands to enable PressToSelect as it is called in the IBM docs, and played around with the related button masks until I received button 4-7 events, which neatly scroll in any app and OS. No further adjustments in my code were needed,quite simple...

For all interested parties, I cleaned my current tree a little and set-up a new repo at github.com - don't expect clean or re-usable code yet, it is not at all in a state I'd like to have it, but rather a snapshot of my WIP.

Thanks for sharing your code, suka.
I cloned repo to my local PC and will look into later!

Outstanding work, outstanding thread, at many levels. Thanks for the many photos and the sharing. The Trackpoint info is very valuable.

I'll dig in the details when I have more time.

I know I'm rather late to the party here, but wow. There's lots of great stuff here. Thanks so much for what you've already shared. Is there any chance you could be enticed into posting (or checking into a github repo) more of your designs?

Long time - no post, but mainly because the BlueCube is still performing nicely and I have felt no real pressure to change anything at that setup during the last months of use. But inspired by the hypermini designs of 7bit I decided I needed another project for the upcoming winter nights, and also a more portable minimal board with the custom AdNW layout. Plus a chance to give MX clears a try - or will it be reds in the end

Anyways, as described in the hypermini thread I feel that the symmetric staggered 7bit layout is a good ergonomic choice, given the constraints that come with simplicity, and will build one the next weeks.

I was going to ask if you have use for some trackpoints...I am sitting on a pile of 100 g81 desko boards with trackpoints, that will most likely end up in the trash.
The actual trackpoint "footprint" is tiny and connects to the pcb with a ribbon cable. The two mouse buttons are on a seperate ribbon cable, attached to that white plastic piece. PCB plug attached.

Ascaii wrote:I was going to ask if you have use for some trackpoints...I am sitting on a pile of 100 g81 desko boards with trackpoints, that will most likely end up in the trash.
The actual trackpoint "footprint" is tiny and connects to the pcb with a ribbon cable. The two mouse buttons are on a seperate ribbon cable, attached to that white plastic piece. PCB plug attached.

Ascaii, I'm currently designing a PCB with a hole in the middle to make use of a track point!

Please don't throw them away...

Well, I was going to use my last original IBM one for this, but yours look much better suited for a custom design. With their smaller footprint I guess you could even fit them between the switches without having to increase the base height for the PCB. Thanks for your offer, PM sent.

7bit wrote:

Ascaii wrote:I was going to ask if you have use for some trackpoints...I am sitting on a pile of 100 g81 desko boards with trackpoints, that will most likely end up in the trash.
The actual trackpoint "footprint" is tiny and connects to the pcb with a ribbon cable. The two mouse buttons are on a seperate ribbon cable, attached to that white plastic piece. PCB plug attached.

Ascaii, I'm currently designing a PCB with a hole in the middle to make use of a track point!

Please don't throw them away...

Yeah, i figured that already
Im still trying to figure out what to charge for them, because it is a bit fumbly and takes some time to take em out. Desko loves hot glue.

As for putting them between switches, you should be able to do it, but only before the switches are soldered in, since they bulge in the middle.

@Ascaii: Used trackpoints cannot last forever. Can we to figure out where this component is produced, maybe is can be bought separately? Hall effect sensors I have found aren't cheap at all ($25 cheapest), while resistive ones cannot detect pressure force.

@suka: Thanks for this great project and sharing it with us. I got inspired by your work.

philpirj wrote:@Ascaii: Used trackpoints cannot last forever. Can we to figure out where this component is produced, maybe is can be bought separately? Hall effect sensors I have found aren't cheap at all ($25 cheapest), while resistive ones cannot detect pressure force.

@suka: Thanks for this great project and sharing it with us. I got inspired by your work.

They were assembled by cheops electronics in germany: http://www.cheops-elektronik.de/