This project started with Linus Tech Tips and an excitement for HIDs (Human Interface Devices). There is an entire series of videos on Linus Tech Tips exploring different ergonomic keyboards, ranging from very usable with lots of benefit to truly wacky stuff designed for very specific needs but all of which are insanely expensive. A 'cheap' ergonomic keyboard might be $200AUD with most of the popular designs closer to $500.
Knowing that my career will likely see me sitting in an office in front of a computer for much of my future I figured it'd be worth becoming familiar with a weird layout now while I complete University in order to maximize both my efficiency in the workplace but also reducing my chances of developing an RSI (repetitive strain injury).
Having wired up small macro boards before I felt confident that I could hand wire this keyboard. This would mean that I could design whatever layout I wanted rather than using someone's predesigned circuit board. Although not to foreshadow but I wished I would have ordered a circuit board rather than hand wiring but we'll get there.
So I started designing the layout in AutoCAD. My two biggest considerations were that it would be a split layout meaning it is two separate halves to keep your shoulders apart and wrists at the right angles and thumb clusters which allow your thumb to use other keys in addition to space, in my case enter, delete and backspace. I also knew I wanted to make the layout ortholinear (in a straight line) because its just better for you if you learn it. Not that it ended up being much adjustment.
I spent a lot of time in CAD designing the keyboard and I learned a lot throughout the process. First thing I learned was that AutoCAD sucks for this kind of 3D modelling and it was the last project I designed where I used AutoCAD, I now use Fusion 360. Note that both of these programs I get free by telling Autodesk what Uni I go to. So if you know someone that goes to Uni get them to help you sign up *wink*.
The second thing I learned was my 3D printer is really small. It only has a print bed of 140x140x140mm which ensured I would have to print each half of the keyboard in several parts. You can see how bad the warping was on my first attempt before making some adjustments to my printer, particularly in that top right joint.
Before committing to the big print I had done extensive testing to ensure a tight fit for the key switches. So after trying again with the prints and fitting all the keys in I had this. The screws are for attaching the baseplate to the top plate.
Then came to the burdensome task of wiring this thing. Wiring all up was probably between 6-8 hours of soldering, trouble shooting and planning. Planning and documenting I knew from the start was important so I kept a good record of what I was planning on doing and what I had just done. I used the CAD layout file to keep track.
Okay now two things. One don't make this project like I did and if you do find someone else's help wiring and programming they can and have explained it much better than I'll ever be able to. What I did was not great. Secondly I should have thought about how I would program this damn thing before I built it.
Now I intended to use QMK (I'm not explaining that rabbit hole here) however the default layouts and even the make your own layouts don't support split keyboard where the two Arduino Pro Micros in each side talk to each other and only one is plugged into the computer. I spent probably a good ten or so hours trying to get this thing to do anything and realised with the 15 git hub tabs open in my browser that I was in big trouble. So I called up my friend Tim who study's computer science, promised him drinks and dinner if he could crack my problem. Six hours in front of my computer later and a lot of swearing and tears and she works!!! IT BREATHES LIFE. Thanks Tim <3. He did something real janky and time consuming and I would in no way recommend putting yourself through what he had to go through.
Oh but I did have to name it something to call in QMK and I noticed most split keyboard have super aggressive names. So I present to you the Fractal Breach.
Okay so this project was actually more troublesome than I remember as I started writing this down but I think I only remember it better because having used this keyboard for about 2 years now I know I am in love. I will not go back to not using thumb clusters and I won't go back to non split.
You'll also notice I have palm rests, this has been so necessary as the keyboard is very tall with its 18mm base plus the height of the switches and keycaps. I have not found them to be annoying or irritating for extended use even though they are just 3D printed PLA.
The good:
This keyboard works and lets me type accurately and faster than I would on a traditional 104 key layout.
I think the layout is perfect for me and has room to grow, there are several keys I don't use at the moment but I could absolutely see binding a macro there in the future for repetitive tasks in an office. The wrist rests are super comfy and overall the keyboard is a dream to type on with no issues during operation.
This keyboard is so convincing that I will be making another at some point to allow one to be left in an office. And if I do make one again well here's what I'd change...
The bad:
Obviously we'll start bigger, the wiring and the programming. What a nightmare that was, I will in no way be doing that again in a hurry. When I build another one I will be using someone else's layout that is similar enough to my own and buy a PCB, saving myself the time wiring and programming it for roughly $70AUD for most boards. Using a PCB will also solve many of my other worry's like long term durability as it will provide some support, it will help make the project much thinner as it doesn't have to accommodate for so much wiring. It will also create an easy place to start designing a 3D printed case from.
For the redo I will use low profile key switches as they have gotten much better, look much smarter and will just keep the entire keyboard lower to the table to reduce wrist strain.
I now am familiar with how to use heat set inserts and now have a heap and this project would've definitely benefited from using those just to keep everything tidy. If you design things like this definitely consider using heat set inserts, they really are magic and cheap as chips.
Now as to the aesthetics, I quite like the scrapped together hacker vibe of the keyboard but I have definitely grown a littttle bored of it. As this was my first major 3D design challenge I'm super happy with what I created but to design again I will definitely spend more time creating something more elegant and nice to look at. Maybe with more than just an accent colour of red on the wrist rest because I ran out of filament haha.
Again I need to stress for all the changes I'd make to it, this keyboard is amazing and was such a great learning project and has been and will continue to be the most used project I've ever built.
Comments