krunchboard
Created by
[slack: krunch]
• Started on June 26, 2025
A 85% ISO mechanical keyboard powered by Raspberry Pi Pico
Total time spent: 26 hours
June 26th - Made the schematic (3 hours)
Click to expand
Made the basic schematic with the ISO105 layout, but without the number pad. Had to use a gpio expander because I was running low on pins I used [keyboard-layout-editor.com](https://keyboard-layout-editor.com) and [kbfirmware.com](https://kbfirmware.com) to help.   **Time spent this session: 3 hours**June 27th - Placed the components (4 hours)
Click to expand
Followed this diagram  Spent quite a while trying to get the key switches to not overlap, but [this video](https://www.youtube.com/watch?v=8WXpGTIbxlQ&t=864s) helped a ton! I switched the diodes and gpio expander from THT to SMT, it might be harder to solder but it's going ot look much better! ## Progress   ## Finished the PCB layout    **Time spent this session: 4 hours**June 28 - Redid the entire schematic (3.5 hours)
Click to expand
I decided to make the function key row closer to the number row, i also switched the oled from 0.96" to 0.91". I added 5 programmable macro keys on the right and moved the volume knob to the top right. Unfortunately I had to redo the schematic and pcb because when I changed the size of the keycaps, the column number also changed as well, which basically moved all of the keys to a new column     **Time spent this session: 3.5 hours**June 30th - Placed the diodes ...again (2 hours)
Click to expand
I rotated all the symbols for key switches in the schematic by 180 degrees so that the wiring would be much cleaner, I originally tried to make a script that does that in one go, but I gaved up and decided to just rotate them all manually
I also placed all the diodes on the PCB, and whilst wiring up the columns, I realised that the row traces will block the connection
And since the hotswap sockets and diodes are both surface mounted, I'd have to create a via at all the column connections
Pretty sure I can just use the footprint that have the holes plated, that way I can just wire the column to the THT pad, I'll figure that out next time.
**Time spent this session: 2 hours**
July 8-9th - Started wiring the rows and columns (1.5 hours)
Click to expand
Like I mentioned in the last journal entry, i ran in to a "issue" where I would need to create multiple vias in order to properly wire the columns which I thought doesn't look as good, so I found a hotswap footprint with the hole being plated, this way, it can serve as a via (through hole) itself, and I won't have to place down multiple vias | Before | After | |--------|-------| |  |  | ## NO VIAS! Since I am now using a different footprint, I had to reassign the 3D models to the switches which took quite a while Once I finished with that, I began wiring up the rows and columns!  It's now 00:15am lol, I'll continue wiring up the rows tomorrow and will replace the neopixel footprint with a reverse mounted one.
**Time spent this session: 1.5 hours**