I've gotten myself into the QMK world, and was recently thinking about the possibility of something like this.
My thought was to start with a atmega32 or something. Might venture down that path when I get a free weekend or two now that I have access to a 3d printer.
I still have some ADNS2620 sensors in my stash (yes, that was a typo that is now fixed). Look into the readme and see that there is a company called PixArt that seems to be the only source for some of these parts. I did get a bite (mouse bite?) from their rep at the time of the OSHWA Summit 2 years ago, but they went crickets. This project could get resurrected if there is a replacement sensor chip. I kind of like the Arduino Leonardo + ADNS sensor, and not interested in just a breakout of an 'all in one' mouse sensor.
Avago's datasheets were thorough and easy to understand. Pixart datasheets, if you can find one, are incomplete and full of grammatical non-sequiturs and ambiguities. I really miss Avago.
I've been wanting to try retrofitting new guts for my Kensington SlimBlade trackball, particularly making it poll and process at 1000Hz rather than 125, since now all of my monitors refresh faster than it does. I've switched to a Logitech G Pro primarily because of the marked difference in responsiveness.
The one thing that might be difficult is doing the scroll detection, since the SlimBlade produces scroll events when you rotate the ball about the vertical axis. Maybe I could get one of those kendryte thingies to do realtime inferencing on a model to discriminate scrolling from pointing.
15 comments
[ 2.7 ms ] story [ 35.5 ms ] threadAlso, his photo reads "ADNS2016"..
I've gotten myself into the QMK world, and was recently thinking about the possibility of something like this.
My thought was to start with a atmega32 or something. Might venture down that path when I get a free weekend or two now that I have access to a 3d printer.
I still have some ADNS2620 sensors in my stash (yes, that was a typo that is now fixed). Look into the readme and see that there is a company called PixArt that seems to be the only source for some of these parts. I did get a bite (mouse bite?) from their rep at the time of the OSHWA Summit 2 years ago, but they went crickets. This project could get resurrected if there is a replacement sensor chip. I kind of like the Arduino Leonardo + ADNS sensor, and not interested in just a breakout of an 'all in one' mouse sensor.
If yer interested in getting a full kit or an ADNS sensor, please comment on the github issue tab https://github.com/biomurph/Mouse/issues/7
Now all old Avago models are manufactured by Pixart.
https://hackaday.io/project/171196-rx-modulus-completely-mod...
[0] https://www.ploopy.co/
[1] https://github.com/ploopyco/mouse
[2] https://www.reddit.com/r/Trackballs/comments/dk4ha7/ploopy_v...
The one thing that might be difficult is doing the scroll detection, since the SlimBlade produces scroll events when you rotate the ball about the vertical axis. Maybe I could get one of those kendryte thingies to do realtime inferencing on a model to discriminate scrolling from pointing.