Hey guys, I'd like to take you on my journey of how I got into making keyboards and my findings and conclusions I've come to after experimenting with different layouts. This will be quite a long post and it could be just the ramblings of a mad Romanian that nobody will read but I felt the need to make this post so here it goes...
About five years ago I was in my first year of university studying CS and while trying to learn how to code I realized that I absolutely sucked at typing. All my life I have used a computer but I've mainly used it for gaming. WASD was embedded in my DNA but the less I played on my PC and the more I needed to do actual work, it became increasingly frustrating to have to move my focus from the monitor to the keyboard and back in order to type. Something needed to change. So I started to learn touch typing. Over the course of a year I had managed to get to about 80 WPM through various tutorials and exercises online. I had grown to love typing and I started to do it as a relaxing activity in my past time. In this time, my interest for mechanical keyboards had grown as well. I bought my first cnc'd aluminum keyboard, started experimenting with new switches and had gotten into lubing, soldering/desoldering.
Over the course of a year I had started to grow a collection of various shaped keyboards, switches, keycaps and keyboard parafernalia that would ultimately lead me to designing and building my own keyboards.
You see, after learning to touch type and mainly getting used to the home row, I had started to get more and more annoyed by the placement of some keys on the board. A lot of the keys that I was using while coding were away from the home row and I had to do finger gymnastics that I wasn't fond of. That's when I started to look into ergonomic keyboards. I very much agree that the thumbs should be used for pressing more keys than just space and I was a fan of the split design that lets you position the keyboard after how your hands naturally sit and not awkwardly placing you hands after how the keyboard sits.
Having a 3D printer, a few switch sets and the confidence of soldering some wires, I first started my journey by building myself a handwired Ergodox keyboard with ProMicros. I liked the staggered ortholinear layout it had and I quickly got used to it. Some of the issues I had with typing while coding were gone and at first I was just excited to have a split keyboard, but I quickly started to question some design decisions the Ergodox had. I felt the need for some wrist rests and the original ones are terrible. The thumb cluster felt weird and I could not understand the position of the smaller keys in the thumb cluster which are quite hard to access. Learning to type on the Ergodox was quite easy for me since I had learned to correctly touch type using all of my 10 fingers but my position always felt somewhat awkward when typing.
The Dactyl Manuform got my attention next, so I built one of those as well. I really like the concave placement of the switches but again I felt like I couldn't get comfortable enough, I was still missing some sort of wrist rest and I never really got used to typing on it.
That's when I started playing with the idea of designing and building my own split keyboard and maybe even sell it in a group buy in the future.
I wanted to base my design on the Iris keyboard but with some keys placed differently. Another thing I really wanted and couldn't find on other split keyboards was an aluminum case. I know there are some keyboards, Iris included that can be built up with layers of laser cut aluminum to form a case but I wanted the real deal cnc'd block of aluminum look with high margins that surround the keys.
So I went to work on what would be a two and a half year journey of learning how to design pcb's, 3D modelling and prototyping.
At first I started the Ergodox way by handwiring my prototypes. My designing experience was almost 0. I made a lot of mistakes and unnecessary complications as you can probably see from the next images.
I was happy enough with the feel of the keyboard so I went on with creating my first PCBs with KiCad. I had a lot of fun learning PCB design and I got working PCBs from my first try. In the beginning I kept it simple by using only through hole components and still relying on ProMicros for the brains of the boards. I never thought I'd reach this point when I first started and this was the moment I knew I would actually finish my project.
I had managed to implement a complete split keyboard with magnetic wrist rests. Now was time for polishing the case design.
I liked the design so far and was looking for cnc companies that would take on my project but I was getting really expensive offers for aluminum milling, mainly because my design was pretty complicated on the inside with lots of small radius corners and tolerances that were expensive to achieve. I gave up on the idea of an aluminum keyboard for a while and started looking at wood as an alternative. I came very close to building a wooden keyboard which turned out pretty good, although since the design was unchanged, there were a few defects with the wooden prototype that ultimately made the assembly impossible.
I realised that I needed to adjust my design for the case and the best solution I could come up with at the time was making the case out of two parts instead of one. At first, I didn't want to have a sandwich design. I wanted each half to be sculpted from a single piece of aluminum, but in order to finish my dream, I was willing to make this compromise. I had created a chunkier keyboard which imo looked worse but at least it was easy enough to produce without completely obliterating my budget.
After more than a year of prototyping I was finally holding my first finished split keyboard made out of aluminum.
This thing is absolutely massive. I had it 3D printed in the beginning but I wasn't expecting this outcome. Each half weighs roughly 2kg, without the tenting feet or the wrist rests (which I ultimately made out of wood but never got to use).
I wanted it powder coated and while I didn't get the color right, I loved it. I was about to start a group buy for it but something else always came up and I never got the chance to actually do it.
In the meantime, I got to use it for quite a bit and started noticing some design issues that weren't so apparent at first.
First was the tenting of the board. While it's cool, I didn't quite get it right. I think the tenting shouldn't just go over one axis but actually multiple ones. While some keys were easier to reach, others became harder specifically the ones actuated by the pinkies.
Second was the placement of the thumb keys. I've chosen this design because I thought it was tried and tested but I never felt fully comfortable with it. Not on my Ergodox, nor this crazy aluminum behemoth I was wielding. I never thought of even doing the thumb cluster any other way.
Over the time I've spent typing exclusively on split keyboards, I have come to understand more and more what it means to have a truly comfortable and ergonomic typing experience.
Here are my findings:
1. The most important part of a split ergonomic keyboard - the position of the thumb cluster. While it looks cool, I believe the thumb cluster on most Ergodox style keyboards is counter-intuitive for a split board. Hear me out. This layout with the thumb keys off to the side and angled makes sense from a normal keyboard perspective. If you keep both of the halves straight you have the tendency to move your thumb away from the rest of the fingers.
However, what's the point of having a split keyboard if not to be able to rotate the halves inwards for a more natural position. Look what happens to the thumb when doing so.
You don't even need to have a split keyboard to come to this conclusion. You just need to look at what your thumbs tend to do when rotating your wrists.
This missplacement of the thumb cluster was keeping me from being fully contempt with my keyboard.
2. Tenting is nice but I prefer a good wrist rest much more. Also, tenting takes a lot more fine adjustment to get it right and you're basically stuck in one position for it to feel comfortable (usually sitting upright, looking forward). While it's true, that position is the recommended position at the desk, sooner or later we all tend to slouch while siting in front of the computer, especially after long hours. You're changing your posture many times. A wrist rest is easier to design and implement and it will suit more people without any additional adjustments. It also works in multiple positions at the desk.
3. A split keyboard should stay in place when you want it to but also move easily when you need it to. After typing on 4kg of pure metal that feel welded to my desk I have noticed just how many positions I actually have when typing. Throughout the day I'm constantly making minor adjustments to the two halves of my keyboard. Maybe I'm leaning back a bit more and I'm dragging the halves toward me. Maybe I just need to rotate the halves a bit more to get that angle to feel perfect.
Every time I wanted to do the slightest adjustment, I had to lift up my 2kg per side keyboard, move it and then test if it was in the right position. If it wasn't, I had to do it again. This got to be pretty annoying.
I believe if you get these 3 points right, you are going to get an amazing result.
So that's what I set out to do.
I knew I had to solve these issues somehow, and definitely make the keyboard smaller, lighter and cheaper to manufacture. So here comes my final iteration of my split keyboard. I've completely redesigned it, incorporating the 3 points I was talking about and this is the almost final outcome:
I've ditched the Pro Micros and used SMD components this time.
The thumb cluster is no longer isolated from the rest of the switches and contrary to many people's first impressions, the outer most thumb key is not the main thumb key. The 1.25U key is where your thumb naturally falls in place when placing your fingers on the home row. In fact, I've designed this layout so that you almost never have to leave the home row for any key or key combination you're trying to access.
The wrist rests have a natural curvature that comfortably support your wrists when typing and again, include magnets. An unexpected but welcome benefit this curvature adds is the ability to easily grab onto the keyboard and move it without the need to lift your hands off the keys. I have found that a good weight combined with believe it or not, some mouse skates, keeps the keyboard from moving when you don't want it to, but also lets you quickly and effortlessly adjust the position of the halves without needing to lift your hands off the board. I know this sounds crazy, a keyboard should feel planted in the desk if possible, but it's much better for a split keyboard to be movable when needed imo.
This is the final design for the new keyboard before I build it out of aluminum. I've decided to ditch the tenting as it would just get too complicated and costly to design a tenting solution that would work with the wrist rests too. What I have included though, are some m2 screw holes on the bottom of the keyboard in each corner where people could insert m2 screws of various lengths to give the halves an angle if they so desired.
I have greatly reduced the amount of material needed to machine the keyboard. I have returned to a single piece design instead of the sandwich design of my previous keyboard and the thickness of the case has also reduced by a lot to just 22mm. I'm hoping these changes fill also cut the production cost by quite a lot, making the final product more accessible.
I know this was quite long so thank you to whoever read the whole thing! I might have left out some important details from this post, but I'll be here to answer any questions. I'll also do an interest check for the keyboard once I get my prototypes.
