42 keys ergonomic keyboard
Loading README and project media…
40%, splittable, mechanical keyboard.
This iteration isn't working. Well, it technically works, but it's a pain to assemble, and the detachable mechanism didn't work out as well as I'd hoped.
The magnets aren't strong enough to keep the two sides connected, so unless you're connecting the two sides with a cable at all times the keyboard is unusable. Also there isn't much space to work with to switch to more powerful magnets.
For now I'm shelving this concept, and might come back to it in the future. As of now the best solution I can think of to fix the issues with this iteration are to:
Always have the two sides connected with a cable. This is the easiest thing to do, but having two cables at all times to use the keyboard kind of defies the purpose of having it splittable.
Make both sides wireless. Last time I checked this seemed very complicated, especially if you want a reasonable battery life. You need one main side communicating with the computer and one secondary side that communicates with the main one.
A mix of option 1 and 2. This is probably the option that gives the best result for the least effort. Have one cable to connect the two sides and then have one of the two connect to the PC via Bluetooth. Need to look more into how to get good battery life though. Not sure if it was QMK or the Feather BLE, but when I tried using Bluetooth the keyboard would only get through a day or two with a 600 mAh battery, which was quite disappointing.
Build – Build log, bill of materials, tips on where to source parts, assembling tips, learning resources I used while designing this keyboard.
Case – Fusion 360 project and STL files for 3D printing the enclosure.
Firmware – QMK configuration. There is a Makefile to build and load the firmware.
PCB – Kicad project and Gerber files for manufacturing the circuit board.
In January 2018 I wanted a keyboard with a specific set of features. The simplest options to get that keyboard were to wait for a group-buy or build my own. Waiting time and costs for both options seemed similar, so I went with the option where I would learn the most.
Needs and knowledge changed over time, so I made different iterations to add new features and account for the lessons learned from previous builds.
If you have some feedback or questions, feel free to create an issue. Your input is more than welcome!
Pictures for the current version will come soon. In the meantime, you can check the previous versions.