Author Topic: AEK II Split  (Read 6700 times)

0 Members and 1 Guest are viewing this topic.

Offline ScarletSwordfish

  • Thread Starter
  • Posts: 20
AEK II Split
« on: Tue, 10 December 2019, 21:33:08 »
Hello,

I've been planning this project for a month or so. This will be my very first custom board and I'm going in the deep end. The plan is to build a custom split ergo board using Apple Extended Keyboard II keycaps and switches.

Here is the current state of my layout. Scroll Lock would toggle the nav layer (red) while num lock would toggle the numpad layer (blue). 0 and enter thumb keys would both be spacebars. The random keys on the far left are mostly just to use up otherwise empty area of the symmetrical shape.

231631-0

Design Goals:
- Use AEK II keycaps and switches (Alps SKCM Cream damped for tactility and relative quietness)
- Split layout with column stagger and thumb clusters
- Full ANSI QWERTY layout with number row and f-keys, so I don't have to remap and relearn any of Photoshop & Illustrator commands I use all the time
- Nav and Numpad on layers
- Adjustable tilt & tent
- Case design should echo late 80s–early 90s Apple aesthetic
- Should be as quiet as possible so I don't annoy my co-workers

Questions:
- (to myself) Can I personally deal with sacrificing right-side keys to make the whole thing slightly narrower? I use most of those right side keys a lot and I want to keep having to learn a new layout to a minimum.
- Should I try a flat Ergodox-style build first to see if I like the layout, or go straight to a dished Dactyl-style build?
- If I go the Dactyl-style route, I want the columns to have a nice smooth curve with no stepping between caps. This will be difficult because AEK caps have stepping built-in with different height stems. The switches will need to be mounted reverse-stepped (i.e. higher rows are mounted deeper in the case.) Can the Dactyl Clojure code be modified to do this, or will this require a whole new CAD model to be designed from scratch?
- What firmware should I use?
- How the heck do I wire this thing?
- How can I make the quietest case?
- How do I get the case the closest color to the keycaps?

I would welcome any thoughts, ideas, and suggestions you all might have. I'd like to finalize my layout and start modeling the case this month.

Thanks,
Scarlet Swordfish

Offline nevin

  • Posts: 879
  • Location: US
Re: AEK II Split
« Reply #1 on: Wed, 11 December 2019, 08:08:27 »
- if you are planning on using the layout you created which is really unique (i like what you did), i would do a custom flat plate and sandwich/layered style case.
     * remember all alps stabs are plate mount
     * remember all AEKII F-keys are rotated

- if you go dactyl-manuform style, stick with 1U or keys that do not need stabilized
     * you can go with non-uniform profile caps and still get a really nice look/feel i've seen at least one example with MDA Big Bones on a dactyl-manuform and it was fine. yes the columns are curved but all it really does is brings the tops of the caps closer together, not stair stepped. (i think you're thinking too much into this)
     * if you want truly even caps, there are dsa alps caps (dsa is a uniform profile, no differentiation between rows)

- if you want to get up & running quickly with a kit that accomplishes 90% of what you want to do, check out the keeb.io viterbi. it's a 5x7 split, ortho, that supports alps. this size/format will also be the easiest transition into split/ortho with very little layout changes. i built one over a year ago and it was an easy transition. this is the layout i've been using (this is the older rev.1 that supported 1.25 mods) i thought i'd miss the larger keys, but now i see how unnecessary they really are.
231671-0

Ditching dedicated F-keys is not a big deal, they just get placed on a layer on the number row. very common and super easy to get used to.

- firmware - QMK there are already a handful of split boards that you can base your firmware configuration on
- wiring - great article & links for handwiring (you'd hand wire if you make a plate for the layout you created or if you build a dactyl-manuform) use 2 pro micros, trrs jacks and serial (unless you really need i2c)
- quiet - they are already dampened. non-metal plate?, lube switches as well (Tribosys 3204)? if you design a 3d case, the roomier the better/deeper the sound.
- case color - if you go with flat sandwich style case you're limited to the colors they have in the material you want (unless you paint it, which really isn't a good idea for something your hands will constantly be all over). if you 3d print a case, get samples of filament to see what's closest to keycap color. check adk64 & adk18 GB threads for examples of some other 3d printed apple cases

resources:
- QMK     https://github.com/qmk/qmk_firmware
- Dactyl-Manuform      https://github.com/qmk/qmk_firmware/tree/master/keyboards/handwired/dactyl_manuform
- Handwire Guide     https://geekhack.org/index.php?topic=87689.0
- Keyboard Layout Editor     http://www.keyboard-layout-editor.com/
- Plate Builder     http://builder.swillkb.com/   (use code from keyboard layout editor)
- Laser Cutting     https://make.ponoko.com/   (laser cut layered case with files from swill plate builder)

...i'm actually currently working on building a dactyl-manuform with dampened alps & AEKII caps as well.
...and i work in creative suite all day every day too...
Keeb.io Viterbi, Apple m0110, Apple m0120, Apple m0110a, Apple 658-4081, Apple M1242, Apple AEK II, MK96, GH60/Pure, Cherry g84-4100, Adesso AKP-220B, Magicforce 68

Offline superbia

  • Posts: 130
  • Location: Republique de Croatie
Re: AEK II Split
« Reply #2 on: Wed, 11 December 2019, 10:43:25 »
After some initial hurdles I'm pretty satisfied with my AEKII powered Iris, it's a weapon.   :thumb:
It may be useful to check out how I programmed mine.
Keebio Iris 2.8 (Alps) X Colemak Mod-DH

Offline 4sStylZ

  • Posts: 120
Re: AEK II Split
« Reply #3 on: Thu, 12 December 2019, 06:47:19 »
I have also thinked many times about a AEK or ASK split,
The keycap profile is a pain to deal with. For me, it’s better to use another profile in complement than using keycaps at the wrong row.

Here is an AEK mixed with XDA.


I would use this set for the thumbs keys.
Bépo user here : AEK64 White linear dampened, XD75 Cherry Blue Jailhoused, TypeMatrix2030 black skin, Lenovo 0B47200 w/ trackpoint, G13, G512. Kensington Expert Trackball & Orbit, Magic touchpad 2.

Offline ScarletSwordfish

  • Thread Starter
  • Posts: 20
Re: AEK II Split
« Reply #4 on: Thu, 12 December 2019, 22:16:06 »
Thanks for the comments. Glad to hear I'm not the first person to try something like this. I printed the layout out to get an idea of the size and both halves pressed together are about as wide as a fullsize. I may cut the far left column and sacrifice symmetry for compactness. Don't know what I'd use those keys for anyway.
231810-0
I think I'm going to try out the Dactyl modeling code this weekend to see if it can do what I want it to. Otherwise I may end up designing a model from scratch in SketchUp.
After some initial hurdles I'm pretty satisfied with my AEKII powered Iris, it's a weapon.   :thumb:
It may be useful to check out how I programmed mine.
That is cool, and thanks for the json file too, it gives me a little better idea of how that works.
- quiet - they are already dampened. non-metal plate?, lube switches as well (Tribosys 3204)? if you design a 3d case, the roomier the better/deeper the sound.
Wouldn't a roomier case make it more resonant and loud? The switches are dampened, but I may try putting some foam in the case for more sound dampening. I wasn't sure if an more dense or foam-filled case would be quieter or louder than a completely open one like superbia's. Thank you for the tips and links too, and I'd like to see your DM when you are ready to share it.


Offline ScarletSwordfish

  • Thread Starter
  • Posts: 20
Re: AEK II Split
« Reply #5 on: Tue, 11 February 2020, 17:41:22 »
It's been a couple months since I last posted but I've done a lot of work on this in that time.

To try to get accurate profiles of the keycaps, I attached a slider to one of each row, taped it to a vertical surface (the edge of a VHS cassette was the handiest one I could find), placed it on my scanner and scanned it at 600 dpi; I then traced the scans in Illustrator. I then drew the switch housings based on dimensions in an old Alps catalog. This let me draw the plate profile with the angles and depths necessary to create a nice smooth contour between the cap tops.
235614-0

I've been creating schematics in Illustrator, exporting them to DXF format, importing them into SketchUp, and extruding them to make my basic plate and bezel models. It doesn't give the most precise result, and there's still a lot of manual tweaking necessary, but it's a quick and dirty way to make the basic shapes to build around.
235616-1235618-2

A couple weeks ago I got a small test print done at my local library. It's really cool seeing an idea of yours take physical form. I desoldered a few switches from the donor board (I pulled switches that I'm NOT using just in case i damaged them, as this was my first time every trying to desolder something) and test fitted them along with the caps. This was just a single column to make sure the switches fit right and my keycaps weren't colliding with each other or the bezels. Form wise it was very good, I just want to make the bottom row switch sit about 1mm deeper so it's not taller than the next row, and raise the top bezel a couple mm. I've made these modifications to my schematic and am working on a revised model. My next test print will be a full left hand upper case & plate.
235620-3235622-4

The holes for the switches were very tight and I needed to scrape the edges down a bit to be able to squeeze them in. Even so, the keycap mounts are even tighter and I was not able to remove the caps without yanking the switch out of the hole. I'm worried that on the finihed board, if I want to remove the caps for cleaning I will end up pulling out the switch and the handwiring along with it. Does anyone have any suggestions on how to keep the switches securely in the plate?
« Last Edit: Tue, 11 February 2020, 17:49:04 by ScarletSwordfish »

Offline nevin

  • Posts: 879
  • Location: US
Re: AEK II Split
« Reply #6 on: Tue, 11 February 2020, 18:03:47 »
NICE WORK! i know that wasn't easy, but necessary to use the the complete set, looks like your using the rotated F-row as well?

- please post your findings, what your measurements were or what the offset is, would be very handy for anyone else using these caps.

thanks @ScarletSwordfish !

a lot of people will glue the switch bases in place (you can always open the top of alps switches, even when in a plate) when placing into a 3d print/plate, but if you're using these caps on these switches, you probably won't be changing cap sets, as these are extremely unique because of the case they originally came in and the varying heights. this isn't the only apple board to do varying cap height but i'm sure the others do not use the same offset (IIgs keyboard is another i'm thinking of )

most other alps caps are not offset like this.
« Last Edit: Tue, 11 February 2020, 23:16:53 by nevin »
Keeb.io Viterbi, Apple m0110, Apple m0120, Apple m0110a, Apple 658-4081, Apple M1242, Apple AEK II, MK96, GH60/Pure, Cherry g84-4100, Adesso AKP-220B, Magicforce 68

Offline Quercus

  • Posts: 13
  • I have not been murdered.
Re: AEK II Split
« Reply #7 on: Tue, 11 February 2020, 21:24:31 »
Very cool! You can epoxy or superglue the switches to the plates to make pulling the keycaps less risky. The keycaps might loosen up  a bit with a few mount and removals, so you can try that before mounting to the plates. I'm excited for you! Keep up that deep dive. This will be awesome.

Offline ScarletSwordfish

  • Thread Starter
  • Posts: 20
Re: AEK II Split
« Reply #8 on: Sun, 31 May 2020, 23:03:45 »
Been another while since my last post. In that time I got my own 3D printer (an Ender 3) and have been slowly working with it. Modeling the curved board turned out to be a lot more complex than I had hoped. I decided to go the alternate route (at least for now) and do a flat design using AppleDesign keycaps. I also narrowed and simplified the layout slightly. I moved =+ to the left of 1! (inspired by the Datadesk SmartBoard) which allowed me to shave 3/8" off each side. I got rid of the filler keys from the numpad and put lock lights on the left side.
244013-0

The ADK uses very different stabilizer mounts than the AEKII does; they are integrated into the slider mounting plate, which is integrated into the upper case. I did my best to take measurements with my not-very-precise calipers and recreate the stabilizer mount. I printed a small test piece out and my measurements were apparently close enough to allow smooth operation.
244005-1

Once I was satisfied with that, I proceeded with a test print of the left upper case & plate as a general fit & feel test.
244007-2244009-3244011-4

The switches and caps all fit perfectly, so I'm very happy with that. I'm not satisfied with the print quality of the top surface, so I will have to try some different slicer settings to remedy that.

My thoughts on the layout: transitioning from row stagger QWERTY to column stagger QWERTY felt pretty natural. The left space space (numpad 0) feels like it's a bit too far in for my thumb, but I think I should print a right case test before I try moving it out, since I normally only use my right thumb for space when typing.

Also, I had disassembled, cleaned, and paper modded these switches over my winter vacation, but they're still not as good as I hoped they'd be. They feel and sound very scratchy. I could try lubing them, but I don't feel super motivated to disassemble and reassemble them all again. I have been thinking about trying some Matias Quiet Click switches instead.
« Last Edit: Mon, 01 June 2020, 07:53:56 by ScarletSwordfish »

Offline nevin

  • Posts: 879
  • Location: US
Re: AEK II Split
« Reply #9 on: Mon, 01 June 2020, 08:08:56 »
Looks fantastic! Great Work! So happy to see some vintage apple/ortho love.
Quote
My thoughts on the layout: transitioning from row stagger QWERTY to column stagger QWERTY felt pretty natural.

Agreed. As long as you go with 7 columns (which are not that common), the transition is very easy. I had the same thoughts when i built my viterbi.

Quote
I'm not satisfied with the print quality of the top surface, so I will have to try some different slicer settings to remedy that.

I don't have a 3d printer, but some say just reorienting the piece on the table can make a big difference. Especially when the top is angled and very stepped from the printer. Try flipping it upside down so that that angle is flat. (if that makes sense) It sounds a little strange but then that angle would no longer have any stair step effect form the printer. Half-stepping and preference in slicers was also mentioned.

Hit up 4sStylZ (on here) for 3d printing apple shells help. He created the ASK60 & ASK10
https://geekhack.org/index.php?topic=103683.0

Or ReDsNoTDeAd (on here) that made the ADK64 & ADK18
https://geekhack.org/index.php?topic=101963.0

Lubing... I just started playing with lubing a couple months back, and tested it on some cream dampened. WOW, what a difference. Even a light coat on the stem changes the switches in a very good way. It reduces the "noise" of the switch and makes it feel and sound so much better. Yes, it's a pain in the butt and time consuming, but well worth it in my opinion.

The Matias switches, other than the cosmetic differences, are simplified alps, not complicated like the older alps switches.

I would love to buy a set of shells from you (regardless of print quality). I've been wanting to try thumb clusters and i have a couple AEKII's lying around, waiting to be reborn.

Wish you the best, building the rest of your board.
Keeb.io Viterbi, Apple m0110, Apple m0120, Apple m0110a, Apple 658-4081, Apple M1242, Apple AEK II, MK96, GH60/Pure, Cherry g84-4100, Adesso AKP-220B, Magicforce 68

Offline ScarletSwordfish

  • Thread Starter
  • Posts: 20
Re: AEK II Split
« Reply #10 on: Sat, 22 August 2020, 11:34:54 »
Hey I'm back. Been tinkering with other stuff for the past couple months, but my right wrist really started hurting again yesterday so that was a wake-up call to continue with this.

I did the right half test print last month. Still having some print quality issues, and I oriented the switch hole for the enter key the wrong way, but think I'm satisfied with the layout overall. I will do a reprint of both top halves.
250107-0

In the meantime, since I've solidified the layout, I suppose I should start planning the circuitry. And I will be honest, I will need a lot of help with this. I really know next to nothing about how electronics work. I've tried reading several different handwiring guides but none of them make any sense to me. So I would like to ask some more experienced folk to help me with which microcontroller(s) to use, which firmware to use, and how to wire this thing up.

Offline nevin

  • Posts: 879
  • Location: US
Re: AEK II Split
« Reply #11 on: Sat, 22 August 2020, 13:19:02 »
- do you have a KLE (keyboard layout editor) layout already created? post the link.
- i'll give you a hand. will do some figuring later this weekend on what controllers are possible with the number of keys you're using. without any matrix voodoo, looks like you're looking at a 6x15 (6x7 left, 6x8 right) [39 keys left half], [43 keys right half]
- so far i've helped a couple forum members with some one-off handwires.
Keeb.io Viterbi, Apple m0110, Apple m0120, Apple m0110a, Apple 658-4081, Apple M1242, Apple AEK II, MK96, GH60/Pure, Cherry g84-4100, Adesso AKP-220B, Magicforce 68

Offline ddrfraser1

  • Posts: 326
  • Location: Smashville
  • 7u > 6.25u
Re: AEK II Split
« Reply #12 on: Sat, 22 August 2020, 13:24:48 »
Great job dude. I was wondering how you'd tackle the taller F keys.  :thumb:

Offline ScarletSwordfish

  • Thread Starter
  • Posts: 20
Re: AEK II Split
« Reply #13 on: Sun, 23 August 2020, 09:30:25 »
- do you have a KLE (keyboard layout editor) layout already created? post the link.
- i'll give you a hand. will do some figuring later this weekend on what controllers are possible with the number of keys you're using. without any matrix voodoo, looks like you're looking at a 6x15 (6x7 left, 6x8 right) [39 keys left half], [43 keys right half]
- so far i've helped a couple forum members with some one-off handwires.

I appreciate your offer of assistance. I haven't done a layout in KLE. Will there be more utility to it than what I've made in Illustrator?

Offline nevin

  • Posts: 879
  • Location: US
Re: AEK II Split
« Reply #14 on: Sun, 23 August 2020, 13:49:21 »
yes. used a couple different ways during the programing (especially for scratch/handwire boards)
here's the start of the KLE: http://www.keyboard-layout-editor.com/#/gists/efb489a54ed546509ee264f777f5fc4a

working on a wiring version

- also, both your scroll lock (arrows) and num lock (numpad) going to be toggles (not momentarys)?
   - if you end up going with momentarys (regylar Fn style key), i'd suggest putting the Fn on the opposite side of the board than on the same side as the alternate/layered keys. (hold the Fn with one hand, other free to type on numpad... etc...)
example.... try holding down "clear" and using the numpad with one hand.....
- you can also set them as dual function keys, let's use "b".... when tapped it's "b" when held, it could be "Fn" (or raise or lower if you're familiar with plank/40% style boards)

controllers... simple promicro's should be fine. (you'd be using 13 pins left, 14 pins right for the matrix) well within the 18 available i/o pins. **you "can" use teensy 2.0's but there a little tricky for split boards, and they're a little more expensive.

hardware wise you'll need:
- wire can be thin, some use the 8 wires of solid core out of network cables, enameled magnet wire (a pain), ribbon cable, or even a normal small gauge wire (i've had some good results with 30ga)
- (82) 1N4148 diodes (just get a pack of 100)
- (2) pro micros
- (2) TRRS jacks (4 conductor 3.5mm jack or similar) some used to use RJ11
- (1) TRRS cable (4 conductor audio cable or similar)
   * i can point you to a couple shops if you get stuck
« Last Edit: Sun, 23 August 2020, 14:23:35 by nevin »
Keeb.io Viterbi, Apple m0110, Apple m0120, Apple m0110a, Apple 658-4081, Apple M1242, Apple AEK II, MK96, GH60/Pure, Cherry g84-4100, Adesso AKP-220B, Magicforce 68

Offline ScarletSwordfish

  • Thread Starter
  • Posts: 20
Re: AEK II Split
« Reply #15 on: Sun, 23 August 2020, 15:14:35 »
I had started working on a KLE layout just before you posted, but you're way ahead of me! Thank you! Are you also doing the wiring layout in KLE? I looked through all the menus on there but I couldn't find any feature like that.

I intend num lock, scroll lock, and fn to be toggles for the layers, not momentary, hence the lock lights on the far left. I'd been debating which key I wanted fn lock to be, but I'm pretty sure now I want it to be F15/pause.

Offline nevin

  • Posts: 879
  • Location: US
Re: AEK II Split
« Reply #16 on: Sun, 23 August 2020, 15:47:45 »
ok. here it is.
looks messy doesn't it.....
250205-0

will make more sense when you look at it this way.
250207-1

of course, this is just my suggestion.
- was trying to tuck the F14 (3 keys) & F15 (3 keys) into the matrix to save some pins as you had some dead space in it anyway.
- this is probably as concise you're going to get it without making it really convoluted.

these were created with KLE with a separate overlay of what the suggested wiring would be.
Keeb.io Viterbi, Apple m0110, Apple m0120, Apple m0110a, Apple 658-4081, Apple M1242, Apple AEK II, MK96, GH60/Pure, Cherry g84-4100, Adesso AKP-220B, Magicforce 68

Offline nevin

  • Posts: 879
  • Location: US
Re: AEK II Split
« Reply #17 on: Sun, 23 August 2020, 15:55:49 »
Quote
I intend num lock, scroll lock, and fn to be toggles for the layers, not momentary, hence the lock lights on the far left. I'd been debating which key I wanted fn lock to be, but I'm pretty sure now I want it to be F15/pause.

that's fine. and if you change your mind, it's just as simple as editing your keymap & reflashing the firmware.
Keeb.io Viterbi, Apple m0110, Apple m0120, Apple m0110a, Apple 658-4081, Apple M1242, Apple AEK II, MK96, GH60/Pure, Cherry g84-4100, Adesso AKP-220B, Magicforce 68

Offline nevin

  • Posts: 879
  • Location: US
Re: AEK II Split
« Reply #18 on: Sun, 30 August 2020, 17:02:48 »
so i've been playing with qmk & via the last two days, managed to get an ergodash working with via (an awesome qmk frontend). I don't see why we couldn't use VIA on this as well. (https://caniusevia.com/)

a couple split boards i was referencing do the rows & columns in the programming slightly different than you would a standard one piece keyboard or how you think a split board may be programmed. so, instead of a 6x16 matrix (rounding up so both halves are 6x8) you program it as 12x8 (stacking the one half above the other instead of side by side). i can give you other examples later.

keep us posted on how it's going.
Keeb.io Viterbi, Apple m0110, Apple m0120, Apple m0110a, Apple 658-4081, Apple M1242, Apple AEK II, MK96, GH60/Pure, Cherry g84-4100, Adesso AKP-220B, Magicforce 68

Offline ScarletSwordfish

  • Thread Starter
  • Posts: 20
Re: AEK II Split
« Reply #19 on: Sun, 30 August 2020, 20:01:45 »
I ordered the parts you listed from Keebio and they should arrive in a couple days. Except for wire, I still need to buy that.

I did a small test print using top layer ironing and that resulted in a much smoother top surface instead of the diagonal ridges I had before. Unfortunately my bottom layer was a disaster so I need to fix that.
250648-0

I don't really know much about how QMK works or what VIA does, but I'm for using whatever will make the board easier to program and use, especially for total non-programmers like me.

Would a 12x8 matrix be wired differently from the 6x16 you diagrammed below?
Which wires will connect the matrix to the to the controllers and to the TRRS jacks?

Once I have the parts and we've finalized the circuitry, I will probably stream the assembly on https://twitch.tv/scarletswordfish if you or anyone else want to watch and give some advice and make fun of my soldering inability. It will likely be a Tuesday at 7 or 8 PM Eastern.

Offline nevin

  • Posts: 879
  • Location: US
Re: AEK II Split
« Reply #20 on: Sun, 30 August 2020, 22:15:35 »
followed on twitch. will piece together a full diagram.

new top looks great. you'll have to share your technique. i've heard of a bunch of 3d printing tricks, even printing upside down so face is down or angling sloped tops so there is less stair stepping. just changing the orientation on the print bed can make big differences.

the wiring of the halves would be the same (again, will send diagram), it's just in the firmware you stack them instead of side by side in the matrix. will show you example then.

as far as general handwiring here's another really good example along with the one i mentioned earlier in the thread
- Handwire Guide (from bafore)     https://geekhack.org/index.php?topic=87689.0
- additional article                        https://deskthority.net/viewtopic.php?t=1067

VIA is really unique as you can change your keymap LIVE! (make a change in VIA and it updates it in the keyboard)
no editing code, recompliling, reflashing..l just to change a key in your keymap.
Keeb.io Viterbi, Apple m0110, Apple m0120, Apple m0110a, Apple 658-4081, Apple M1242, Apple AEK II, MK96, GH60/Pure, Cherry g84-4100, Adesso AKP-220B, Magicforce 68

Offline nevin

  • Posts: 879
  • Location: US
Re: AEK II Split
« Reply #21 on: Wed, 02 September 2020, 14:26:34 »
- working on the firmware - have rough QMK put together
- have the layout in VIA
- working on led (lock lights) implementation (running short on pins, need one for each led, only have 2 available without experimenting with the promicro's onboard led indicators, ..."could" gain another 2 pins but i've never tried it yet)
- have some hardware coming for testing

via
250835-0

matrix
250837-1
« Last Edit: Wed, 02 September 2020, 14:28:25 by nevin »
Keeb.io Viterbi, Apple m0110, Apple m0120, Apple m0110a, Apple 658-4081, Apple M1242, Apple AEK II, MK96, GH60/Pure, Cherry g84-4100, Adesso AKP-220B, Magicforce 68

Offline ScarletSwordfish

  • Thread Starter
  • Posts: 20
Re: AEK II Split
« Reply #22 on: Thu, 03 September 2020, 19:37:26 »
I realized yesterday that shifting =+ to the left of 1! would really annoy me, because i use it and -_ for zooming in and out in artwork files, and having to use separate hands to do that would be really obnoxious. So i've moved it back to its normal place. I can still keep the case the same width by making delete 1u instead of using the normal 2u cap. Hopefully that won't mess up your matrix too badly.

Here's my revised KLE layout: http://www.keyboard-layout-editor.com/#/gists/943c20451b16d34e0054c3c6882f00b7

I got the parts I ordered on Monday, that was really fast!

I figured out why the bottom layers of my prints are getting screwed up, my print bed is warped and is higher in the middle than at the corners. I want to finish these upper cases by next Tuesday so I can start assembly, and I don't have time to order a glass replacement bed, so I'm just going to have to print them with a raft base. It will take more time and use more material, but it should provide a nice even base for the prints.

Offline nevin

  • Posts: 879
  • Location: US
Re: AEK II Split
« Reply #23 on: Fri, 04 September 2020, 00:30:10 »
Quote
I realized yesterday that shifting =+ to the left of 1! would really annoy me, because i use it and -_ for zooming in and out in artwork files, and having to use separate hands to do that would be really obnoxious. So i've moved it back to its normal place. I can still keep the case the same width by making delete 1u instead of using the normal 2u cap. Hopefully that won't mess up your matrix too badly.
no problem. you haven't started wiring anything yet. absolutely agree. i've been playing with split boards for about 2 years now and i don't see myself going smaller than my viterbi (5x7s) specifically for that reason. a normal number row and enough room on the right side to keep everything (except backslash/pipe "\|") where it was on a standard board. i do more shortcuts & key commands than actual typing.

print the layout on paper one more time & go through your normal shortcuts. see if anything stands out or is more of a stretch then you'd prefer.

Quote
Here's my revised KLE layout: http://www.keyboard-layout-editor.com/#/gists/943c20451b16d34e0054c3c6882f00b7
nice. very well formatted

Quote
I got the parts I ordered on Monday, that was really fast!
yes, i LOVE keeb.io

not sure what your planning for mounting the promicros or trrs jacks but here's a couple that you might be able to use or incorporate to make life a little easier
- you could mount to perf board and then screw the perfboard to your print
- couple 3d designs i ran across while contemplating a dactyl-manuform
promicro/trrs holder from manuform https://www.thingiverse.com/thing:4120575
chunky promicro holder https://www.thingiverse.com/thing:3745009
there's a holder in this set https://www.thingiverse.com/thing:4416866

also, got the lock light to work. the layer lights function and are called differently than the normal caps lock style lights. as it's a layer status, not a switch status. so working on that....
Keeb.io Viterbi, Apple m0110, Apple m0120, Apple m0110a, Apple 658-4081, Apple M1242, Apple AEK II, MK96, GH60/Pure, Cherry g84-4100, Adesso AKP-220B, Magicforce 68

Offline nevin

  • Posts: 879
  • Location: US
Re: AEK II Split
« Reply #24 on: Fri, 04 September 2020, 00:42:29 »
... also, the led's you're planning on using, are they the ones from an AEKII? or are you just using the original lenses for the shell and using newer leds?
will need to figure out resistor values for the leds you are planning to use.
Keeb.io Viterbi, Apple m0110, Apple m0120, Apple m0110a, Apple 658-4081, Apple M1242, Apple AEK II, MK96, GH60/Pure, Cherry g84-4100, Adesso AKP-220B, Magicforce 68

Offline nevin

  • Posts: 879
  • Location: US
Re: AEK II Split
« Reply #25 on: Fri, 04 September 2020, 01:21:55 »
on the layout.... if you're moving the += back over to the right side, just shift the 6 over to the left. no need to change the shells. plus, your F & number keys line up again.
250987-0
http://www.keyboard-layout-editor.com/#/gists/b5638eeca492018918c4b762d374c078
Keeb.io Viterbi, Apple m0110, Apple m0120, Apple m0110a, Apple 658-4081, Apple M1242, Apple AEK II, MK96, GH60/Pure, Cherry g84-4100, Adesso AKP-220B, Magicforce 68

Offline nevin

  • Posts: 879
  • Location: US
Re: AEK II Split
« Reply #26 on: Fri, 04 September 2020, 15:04:00 »
got the layer lights figured out & working!

got a couple promicros & such. going to try the extra pin hack to gain an extra 2 i/o pins
Keeb.io Viterbi, Apple m0110, Apple m0120, Apple m0110a, Apple 658-4081, Apple M1242, Apple AEK II, MK96, GH60/Pure, Cherry g84-4100, Adesso AKP-220B, Magicforce 68

Offline ScarletSwordfish

  • Thread Starter
  • Posts: 20
Re: AEK II Split
« Reply #27 on: Fri, 04 September 2020, 20:38:16 »
I've left the TRRS jack and microcontroller mounting situation (and subsequently the design of the lower case halves) open because I assumed that their positioning would be dependent on the layout of the matrix. I do know that there will not me much space if any in the right half for one of the Pro Micros. I was designing assuming a single microcontroller in the left half, to the left of the keywell. I only intend to leave about 1/8" of space between the switch pins and the bottom of the case, which i want to fill with some sound dampening material. Would I be able to put both Pro Micros in the left half?

The LEDs are RGBs from Keebio. Not actually sure how I'll make the windows. Maybe I can find a tiny bottle of resin to fill the holes.

I don't want to shift the number row to the left because I want to maintain traditional touch-typing columns. I've thought about removing F13 and shifting the F row to the right, but I also kind of like having F6 on the left half because in Photoshop I use F6 for Feather Selection and F5 for Fill. I'll think on that one.

I did a raft base print test and that turned out pretty well. So I'll definitely be doing that. I think I'll be able be able to get both upper cases printed by Tuesday.

Offline nevin

  • Posts: 879
  • Location: US
Re: AEK II Split
« Reply #28 on: Sat, 05 September 2020, 06:47:47 »
controllers...
- putting controller under switches only adds 5-7mm (when using the header pins), probably a little less if you ran the leads out the sides
- promicro (including usb port) is only 4mm thick
- if you want a convenient split board, i'd suggest a promicro in each half
- you could use only one controller in the left half, but it would have to be something with more pins like a teensy ++2.0 (https://www.pjrc.com/store/teensypp.html)
   • if using a teensy++ 2.0 in left half, you would need a MINIMUM of 13 wires (probably a couple more) to connect one half to the other. because you would be extending the wires of the rows & columns from the right half back to the left. this makes the cable between them much thicker and stiffer. i've seen it & done it both ways.
   • the beauty of dual promicros is you only need 3-4 wires between halves.
- you could probably add just a tiny bit to the overall height & be able to use 2 promicros. but again, totally up to you.

leds...
- ok, will redo for the rgb ones, i had implemented regular leds, just an on/off, rgbs will need colors specified.
- the original case had 3 windows on it in a single block of plastic. you could easily cut them apart & use them. some high end aluminum boards don't even use windows anymore, jus a hole down to the led.
251055-0

layout...
i get it, it's totally personal, i've had the same feelings about my layout/key placement
Keeb.io Viterbi, Apple m0110, Apple m0120, Apple m0110a, Apple 658-4081, Apple M1242, Apple AEK II, MK96, GH60/Pure, Cherry g84-4100, Adesso AKP-220B, Magicforce 68

Offline ScarletSwordfish

  • Thread Starter
  • Posts: 20
Re: AEK II Split
« Reply #29 on: Tue, 08 September 2020, 19:06:46 »
Well I didn't get an upper printed in time to wire tonight. I realized Saturday that I had forgotten to add a way to join the upper and lower cases, so I had to do a small test print for some screw holes and those turned out okay. I was hoping I could get the right upper printed today but two hours in my print had some problems and I had to abort, and it's approximately a 14 hour print. I'm aiming for Saturday now.

I decided to delete F13 and shift the F row to the right, so esc is above `~. This is the final layout for real this time I promise cross my heart.

http://www.keyboard-layout-editor.com/#/gists/943c20451b16d34e0054c3c6882f00b7

I had read Cribbit's post several times before but I had not seen the one with the M0110. I really like how that just used the resistor leads to connect the pins without additional wire, I think I will try that.

The thing that I have had the hardest time wrapping my head around in this process is understanding how the matrix is wired to the microcontrollers and how the microcontrollers are connected to each other. But I think it's slowly starting to make sense to me. Basically, I just need to wire each column and each row to a separate pin on the ProMicro, and it can be any pin. Then I can just use QMK to tell the Pro Micro which columns and rows are connected to which pin. Is that correct? How does the bridging between the two PMs with the TRRS cable work?

Offline nevin

  • Posts: 879
  • Location: US
Re: AEK II Split
« Reply #30 on: Tue, 08 September 2020, 19:57:17 »
Touch base little later tonight
Keeb.io Viterbi, Apple m0110, Apple m0120, Apple m0110a, Apple 658-4081, Apple M1242, Apple AEK II, MK96, GH60/Pure, Cherry g84-4100, Adesso AKP-220B, Magicforce 68

Offline nevin

  • Posts: 879
  • Location: US
Re: AEK II Split
« Reply #31 on: Fri, 11 September 2020, 09:06:32 »
sorry this week's been a little nuts.
really doesn't matter how the trrs connector is wired but i wanted to use what most have been using for consistency, had to do a little poking with my multimeter....

so for trrs connector:
sleeve  = ground
ring 2 = VCC
ring 1 = (unused in serial, would be pin 2 if using i2c)
tip = pin 3 (SCL serial)
251556-0

push button for reset goes between RST and ground (you're just shorting RST to ground)
here's the pinout of a promicro
251558-1

will have the rows/columns marked once i update to your final layout.
Keeb.io Viterbi, Apple m0110, Apple m0120, Apple m0110a, Apple 658-4081, Apple M1242, Apple AEK II, MK96, GH60/Pure, Cherry g84-4100, Adesso AKP-220B, Magicforce 68

Offline ScarletSwordfish

  • Thread Starter
  • Posts: 20
Re: AEK II Split
« Reply #32 on: Sat, 12 September 2020, 14:52:00 »
I'm going to start assembling tonight at 7 PM Eastern, we'll see how far I get. https://twitch.tv/scarletswordfish

Offline ScarletSwordfish

  • Thread Starter
  • Posts: 20
Re: AEK II Split
« Reply #33 on: Sat, 12 September 2020, 23:12:21 »
I started soldering the diodes onto the switches. For my first soldering job ever, it's... pretty rough. I may have slightly melted a few of the switch housings. It was recommended to me to get a finer tip for my soldering iron, the one I have now is very really big and imprecise. But at least I made some progress.

251630-0

I also tried my hand at laying out a matrix... still not really sure what I'm doing, but am I on the right track?

251632-1

PS. I also hang out in the #kb-ergo channel on the MechKeys Discord, if you wanna chat there.
« Last Edit: Sun, 13 September 2020, 00:18:13 by ScarletSwordfish »

Offline nevin

  • Posts: 879
  • Location: US
Re: AEK II Split
« Reply #34 on: Sun, 13 September 2020, 12:34:34 »
diodes are usually in the rows instead of columns, but we can probably make this work too. think it's just a comment in the firmware but will check.

soldering looks fine, you want it a little chunky for solid connections, too little solder and the connections are brittle

and this is what i was thinking for the matrix. (basically, on the right side, just move the del up one row so you have 8 in row 5 before the thumb cluster in row6) but either way will work.
251653-0
Keeb.io Viterbi, Apple m0110, Apple m0120, Apple m0110a, Apple 658-4081, Apple M1242, Apple AEK II, MK96, GH60/Pure, Cherry g84-4100, Adesso AKP-220B, Magicforce 68

Offline ScarletSwordfish

  • Thread Starter
  • Posts: 20
Re: AEK II Split
« Reply #35 on: Sun, 13 September 2020, 14:53:12 »
I did intend to use the diodes in the rows instead of the columns, I just soldered them pointing perpendicular to the rows as that's how both of those handwire guides showed them oriented. Though I probably should have pointed them down instead of up, as the top row wire would run very close to the top edge of the case.

Offline nevin

  • Posts: 879
  • Location: US
Re: AEK II Split
« Reply #36 on: Sun, 13 September 2020, 15:03:44 »
Oh, good. Sorry, was a little hard to decipher just by looking at the picture. All good then. Will touch base soon
Keeb.io Viterbi, Apple m0110, Apple m0120, Apple m0110a, Apple 658-4081, Apple M1242, Apple AEK II, MK96, GH60/Pure, Cherry g84-4100, Adesso AKP-220B, Magicforce 68