I think you should probably use a CAD instead of graphics software to design...
Nice, I can buy ch582f dev board for 2 dollars. Still very cheap!
If they are using ToF sensors, it is possible. But hell ToF sensors are expensive...
I think the dometyl keyboard (https://github.com/geoffder/dometyl-keyboard) was designed to use hall effect sensors with wiz switches. Not sure what is the PCB design that read the hall effect sensors.
If we just need digital, a comparator should be enough. However, wooting require 1 ADC for each switch, or at least some multiplexer. As it is unlikely for a single MCU to have that many ADCs, the pcb will likely require multiple chips like ADC0809 that can read 8 different analog outputs. Probably some more mux as well to reduce the number of pins required for the MCU.
Yeah WCH make some really cheap chips. Apart from ch552, there is also a family of BLE chips such as ch571f that is much more efficient than ESP32 (from online sources) and is very cheap. You can get the chip for half a dollar (https://www.aliexpress.com/item/1005005974200193.html) and a development board for less than 3 dollars (https://www.aliexpress.com/item/1005005974200193.html). When I bought it on taobao, it is actually around 1.3 dollars.
The downside is that software support does not exist (yet), so will have to try and make a firmware for it. I haven't yet found time as I was working on manifold things.
Just in case it may help: https://lemmy.world/post/3373051
There were efforts trying to do make this simpler, but it was abandoned. See https://github.com/openscad/openscad/issues/4743 and related PR.
I just did it for the wrist rest, as that is the part your wrist will touch regularly. It is too much work to cut many holes if you want to cover the plate, and I don't think it will be much cleaner comparing with bare printed surface. You still need regular cleaning.
Indeed, the major cost is labor cost. I was iterating on the design once a month and do some very simple tweaking (change the screw location for example). Because I want to make sure that the entire thing works, I need to solder everything for each cycle. Due to the lack of components, I have to disassemble old working keyboard to make a new one. I probably disassembled 4 or more in total. I was looking to sell the old one for $100, but no one was interested. I guess a lot of mechanical keyboard users are probably not used to split keyboard and DIY keyboards, so are not willing to buy it.
Not sure about others, I tried to sell me keyboard but without success, maybe split keyboards are a bit too niche and no one near me (HK) is interested. I just disassembled the keyboars to scrap the components for the next build, which is a bit sad considering a significant amount of time is spent on assembling it.
The main problem with SLA is the material and size. I don't think SLA can print ABS?
For wrist rest, you can stick a piece of TU leather to cover the printed surface, so it will look nice, be clean and comfortable. Not sure if TPU is sufficiently soft if you want a soft wrist rest, you would probably need something like silicone.
wondering if it is faster for you?
This looks pretty! Wondering where you guys get those fancy choc keycaps as well, I usually go to aliexpress (taobao in hk) and choc keycaps are much more expensive comparing to the normal kind of keycaps. I can usually get 60 normal keycaps for around $10 USD, but the price easily doubles for choc.
a soldering iron at 360 degree celcius
yeah, they are quite delicate. I wanted to use thicker ones but it is hard to get thick magnet wires. I never tried dropping them so idk, I guess making them less tensioned would help a bit
soldering SMT diode is really fast
![](https://lemmy.world/pictrs/image/bf0b374a-07b5-4393-a872-e8fb978479a2.jpeg?format=webp&thumbnail=128)
![](https://lemmy.world/pictrs/image/bf0b374a-07b5-4393-a872-e8fb978479a2.jpeg?format=webp)
No need to bend the legs, no need to do wire stripping. I did this in 30 minutes casually... I guess the full halve will take around 1 hour
I am a vim user (and use vim keybinding for web browsing), so I rarely need to use both my mouse and keyboard at the same time.
Manifold: Fast, robust CSG (alternative to OpenSCAD)
Geometry library for topological robustness. Contribute to elalish/manifold development by creating an account on GitHub.
![GitHub - elalish/manifold: Geometry library for topological robustness](https://lemmy.world/pictrs/image/050ce7c2-1f95-4857-a7bd-f785a1440a72.png?format=webp&thumbnail=256)
Just wanted to share the manifold library to 3D printing enthusiasts.
It is a fast and robust mesh boolean library in C++ with a typescript API that you can use in your browser! It has most of the operations supported by OpenSCAD, but without import svg and minkowski. The operation is so fast that even OpenSCAD is using it as a backend (experimental option in nightly) and get up to 250x performance improvement for some models.
Wondering what is the bottleneck for the system. Number of requests per second or something?
Not very feasible I think. Our messages are basically broadcasts, so unless every server respect the privacy of their users and implement some kind of rate limiting, otherwise the data will still be public. I think privacy issue will require regulation instead of a technical solution.