All details are here: https://gitlab.com/lpgalaxy/revxlp

The revxlp is using a shift register - that’s the trick

Done. Love it. thx @manna_harbour@lemmy.world and @petejohanson@lemmy.world for your help - and thx goes to frobiac from Munich for jumping on a mental health call and debugging the issues. it works so well now. my fav keyboard!

SOLVED

all solved… I had 2 dead pins due not really following the tape socket explanation in the readme. and when I pulled out the xiao, the socket came half way out. I now did solder a new xiao and checked all pins, now all good… phew! thx again! it chose the 12u miryoku firmware and it works

Just adding the back of the PCB - you can see some missing diodes there. For now I am pretty sure that is OK, as it’s what JLCPCB has assembled as in the BOM/Place file.

@petejohanson@lemmy.world

I checked connections from XIAO to the board and also took off the bottom plate, alu, just to be sure there is no shortcut. When plugging into usb-c and turning on, the green led charge indicator is on, seems to charge. I was previously able to connect via ble, too.

All right, here is my build run, it is a success and should build a basic miryoku for revxlp - details here https://github.com/hansamann/miryoku_zmk/actions/runs/6761313214

When I double press reset to have the USB storage of the XIAO appear, it will bring up an error "can’t complete operation because some data in …uf2 can’t be read or written. But I think it copies the 359KB of uf2 over correctly. Then Finder of course shows an error as the device resets into the keyboard mode… all expected, right?

Unfortunately it still behaves weirdly. The outer three columns don’t do anything. the 4th column top is a ‘w’ then below r and then x… leaves me clueless. I used JLCPCB with the assembly service. On the back of the pcba, I can see some D’s are not places, for example DU D_U2 / D_U4 / D_U3 / D_U9 / D_U8 - but I assume that is all OK. I added the not swap sockets ONLY on the back as the assembly should do the rest on that side of the PCB.

If you could point me to a ready made build for the XIAO-nRF52840 with a default keymap I am happy to try that out.

will try building with defaults… which I pretty much did to be honest. just mac style copy/paste is special. I will take off the alu back and chec the xiao ble connections. thx

Would you mind making it 36 keys, 3 Thumb keys? Then I am in to try it out… Looks like a great board.

Love it.

I am interested in the 3x5+3 version and would love to get Miryoku on it, plus of course the track pad needs to work, too. Is there an easy way, like a qmk workflow build that achieves this? Or another guide that describes it?

I still would love to build it but jlcpcb shows these mounting holes as not available… any better idea than just trying again every week?

Awesome, you deserve thx so much. I will try again soon as some of the components are not available, indeed. Would be nice to get it assembled fully.

Thx again - just printed the right side and was initially a bit worried about the snug fit, but after pushing it in a bit it simply fits perfectly!

Just fyi - bottom plate Gerber’s for FR4 shows a 404 for me…

Would you say 10u PCBs ready to order? Would order with jlcpcb assembly…

Just checking - files/parts/build guide is not yet available, or did I miss it somewhere?

This is it - thx so much, let me know when you think it’s ready and I am ready to build.

let me confirm it was an issue with karabiner-elements… I totally forget, thx for the great comments @Tompi@discuss.tchncs.de and @manna_harbour@lemmy.world I now just realize that the Miryoku Colemak DHK is also not the default, but I’ll just swithc that back to default and it will most likely work!

thx all!

Oh no many others, some a bit over 36 keys, like piantor and revxlp… But just using 36 typically.

Chocofi is awesome, too. Also built crowboard, and a self-wired skeletyl but right side unfortunately refused to work… Need more time to look into it.

Yeah the same and default alphas means Colemak Mod dh normally - I assume it’s my fault and the key mapper… Checking in a few hours and will report.