Background/Context
Everything started with a colleague's idea to find a small trackball near the keyboard to minimize the need to move the hand from it to the mouse.
After a brief internet search and finding "mice with trackballs" and the Ploopy project, we concluded that there are not many options or they are relatively expensive. I came across the Dactyl Manuform project and its evolution, the Tractyl Manuform. An ergonomic keyboard with a trackball. I decided this is it. I'm building it.
Project assumptions
- As cheaply as possible (to see if I will like it)
- Using Arduino Leonardo boards from the drawer
- Using components from an old broken mechanical keyboard
- QWERTY layout to ease the adaptation to the new keyboard
- 5x6
- QMK
Assembly
To meet the goal of keeping costs as low as possible, I encountered several challenges. The first major issue was the trackball. Dedicated trackballs are very expensive and there are no affordable alternatives. I had the idea to purchase a shiny 33.3mm bearing ball. The second challenge was the PMW3360 optical sensor, which is only available from China. Additionally, I needed to design or order a PCB and solder the components, which added to the complexity. However, I found a very cheap gaming mouse on a local auction site that included the PMW3360 sensor. I decided to take a chance, and it turned out to be a perfect solution.
Printing the entire structure took about 20 hours and used approximately half a spool of filament. After a difficult struggle with removing the supports, I decided to finish the model by sanding it, applying a primer, painting, and finally adding a clear protective layer. I chose not to use sockets for the MX switches, so I soldered them directly to 2.5mm2 wire, bent to fit the shape of the shell. Yes, I know it's overkill.
After compiling QMK, resolving a few connection issues, and adjusting the layouts, the keyboard was almost ready. The only thing missing was the trackball. Since the bearing ball took a month to arrive, I temporarily printed a substitute. Instead of desoldering the mouse, I brutally cut its PCB. Surprisingly, it still worked.
When the bearing ball finally arrived, I discovered that it didn't work because the sensor couldn't detect movement on the shiny surface. I came up with the idea to patina the ball. After quickly finding a method, I boiled it in vinegar for 20 minutes. The result was amazing. The sensor detected movement, and the trackball gained a unique appearance.
Images
