My lagrange build #10
Replies: 2 comments 1 reply
-
Excellent build log! I'll only note, with respect to the thools section, that 0.8mm solder is probably too large for SMT work. I'd recommend 0.5mm or even a bit smaller than that. Additionally having some sort of magnification can help a lot. A digital micrsocope works pretty well, but even a simple magnifying glass should be enough to get the job done. |
Beta Was this translation helpful? Give feedback.
-
Update for keycaps printingI got the 0.25mm nozzle and installed it in my Prusa. There was a lot of fiddling involved to get it working ok-ish in the Prusa slicer:
--- a/src/lagrange_keyboard/core.clj
+++ b/src/lagrange_keyboard/core.clj
@@ -899,8 +899,8 @@
h_0 :mount-recess
:or {h_add 0
w 1.8
- a 4.1
- b 1.19
+ a 4.2
+ b 1.24
r 2.75
δ [0 0]
h_0 0
You may still experience some stringing and a slight build-up of filament at the outer side of the extruder. It'll probably be fine for 3-5 hour prints, but longer prints may cause the filament clog to affect the print. I haven't been able to avoid that altogether. After all, PETG isn't that recommended for detailed prints, but it works ok for short prints. Using The result is pretty awesome and fits nicely onto the keyswitch: |
Beta Was this translation helpful? Give feedback.
-
Hasufells Lagrange build
I hope I'm the first person to rebuild the lagrange 😁 🥳
Many thanks to @dpapavas who provided support throughout all the stages of my build and improved the build guide wherever possible.
Pictures
More pictures
Motivation
I'm not into 3d printing or weird keyboards in general. But last october I started suffering from very bad RSI that caused me to (temporarily, I hope) quit programming in February. Since then I'm looking for ways to improve my health and fitness.
This blog post was also inspiring and I had already spent hundreds of dollars on ergo keyboards (kinesis advantage 2), height adjustable tables and so on. If you're struggling with that too I recommend reading RSI supplements and signing up for the online course at functionalpatterns to improve posture and dynamic imbalances.
At any rate... I started researching about tilted keyboards and ended up in the ErgoMechKeyboards sub, where I stumbled across the lagrange thread.
So, long story short... I built my own!
Challenges
Here are some of the challenges that I personally remember most. There was much more blood spilled, frustration and times where I wanted to give up... but in the end I did not.
Figuring out a complete shopping list
You need a lot of tools and material to get started. The build guide is very verbose, but I wanted an overview of what I need, so I can plan shipping times, cost and so on. I also wanted not just the material, but all the tools I need to actually get the job done.
Here is what I came up with:
Tools
The following tools are needed to print, solder and assemble. These can be useful for future projects as well.
Material
Various materials that are specific to the project. Some of it might be useful for other projects as well.
Controller electronics
These are already fully explained in the build guide: https://github.com/dpapavas/lagrange-keyboard/blob/master/BUILD.md#the-controller
The design phase
The lagrange is an impressively well parametrized keyboard. I had a quick look at dactyl manuform projects and couldn't find the same amount of parameter details. As such, it's only natural to experiment.
Experimentation was very slow. I started with comparing adjusted parameters side-by-side in OpenSCAD. My main complaint with the default build was that I felt some columns are too deep (e.g. middle column) and I didn't like how my fingers had to travel to the upper row. That required some changes to the parameters that were a little less intuitive and required understanding of the geometry concept. I had an extensive discussion with @dpapavas about it here and the result of my build was the following patch:
I printed 4 prototypes (one half of the chassis print is about 16 to 19 hours) and also experimented with increased curvature, which however felt too cramped. I used an existing keycap set from my Ergodox to figure out the build feel.
At one point you just have to stop experimenting and concede that it won't be perfect anyway. The way your hands will adapt to the keyboard won't be easy to predict anyway. Especially since with a split keyboard and a configurable tenting angle, there are (imo) more ways to touch the keyboard.
Keep in mind that you can print stands of different tenting angle.
Printing issues
I bought the Prusa i3 MK3S+ in hope of having less difficulties with details, experimentation and calibration. And so far I think it delivered.
The only issue I really had was with stringing of PETG. Many people seem to suggest to increase the retraction setting massively, but others say that may clog the nozzle, cause even more stringing or jamming. Others say it's wet filaments, so I went down the road of reading about food dehydrators and even bought one. After drying the filament, I can say the difference of stringing was really marginal. The only thing that really improved was the "oozing" in idle mode (when your nozzle is heated up, but you're not printing yet and you still get extrusion).
I decided that it's not worth more investigation, since removing the stringing is pretty straight forward. It mainly happens on the bottom side or inside the keyswitch holes. These can all be filed easily.
I also nearly destroyed one steel sheet (smooth double sided PEI). For PETG filament you really want a Double-sided Textured PEI Powder-coated Spring Steel Sheet or use special glue for the printbed.
Keycaps
The lagrange has a few custom keycaps as was discussed here and printing keycaps with FDM is a real challenge and I'm still in the process of figuring out the right way. The keycaps from this repository are "perfect fit" at the bottom side, so there's no room for inaccuracies during the print. That means you'll generally need to file them or experiment with the parameters as discussed here. Another idea was to use the bottom fit from the KeyV2 project, see here.
Even if you print at 0.05mm layer height, the top layer won't be without gaps after sanding.
I ordered a 0.25mm nozzle, which have been reported to produce much more accurate prints and will experiment with that further.
The last resort would be buying a resin printer, but for some reason I don't feel like getting into that just for a couple of keycaps.
Sanding
I sanded my first "final" print. It took me 1.5 days and I ended up having serious neck and wrist pain. Then I accidentially destroyed my print by making a huge mistake during assembly of the switches. In the end I decided sanding is not worth it and just skipped that step for my next print.
Building the controller
The controller is pretty intimidating, especially with the microchip. There are many videos on youtube on how to do that, but they are missing so many infos (which exact temperature, iron, flux, ...) that reality will be a cold shower of having shorts everywhere between the legs. You really want to spend a lot of time and care on this, testing for shorts with a diameter. The right amount of flux is key here, but keep in mind that some flux (even no-clean ones) can cause stray currents. That's why I cleaned up all soldered spots with Isopropanol.
Also keep in mind to not put too much solder when soldering the crystal. You can easily cause a short that bypasses the clock.
Mistakes at this stage can be pretty frustrating if you have little experience with electronics. It's good to always test every component part (e.g. condensator or resistor, as far as possible) before doing the final plug-in.
I have to say I was also kinda affected by the Suez canal debacle and almost got no microchips.
Wiring the matrix
I believe this took me 3 days. It may look less intimidating than the controller, but it's much more repetitive and annoying, so I needed frequent breaks because I got cramps in my hands and my neck would become stiff.
We discussed several challenges during the wiring here. Especially the crimping is something I initially completely got wrong and @dpapavas had to point out that the result is not what it should look like. I'm still in the process of redoing that bit, so that my build is robust and will last.
Future work
My plan was to use some very custom keycaps to reduce finger travel to the upper rows further for the index, middle and ring finger. Basically a SA2 keycap for the number row and a shortened SA2 keycap below that. Such keycaps can be easily created from the KeyV2 project. For the upper two keys of the pinky finger I wanted to maybe use very short G20, so I can reach them better with my ring finger (you might call this a bad habit, but it really feels more comfortable to me...).
I will also print a stand with a more radical tenting, but I kinda like the default of the build.
I'm also planning to build a small suit case for the lagrange, so I can comfortably travel.
Beta Was this translation helpful? Give feedback.
All reactions