I'm not very well versed in this kind of thing, and honestly I couldn't make much sense of the velocity detection code, so I did what any idiot does and mess with the numbers a bit and see what happens. I changed velocity_samples to 6, sumx to 14, and sumxsq to 42 (a best guess based on their previous relationships) and gave it a shot. Results seem minor and mixed. Variance seems a bit higher, but it does seem like I have a slightly wider range of velocity available.
Increasing the sample count to 8 made the main pads unresponsive and required use of the very useful FWUP boot option to salvage, but this sadly doesn't retain calibration data. As I don't want to have to re-calibrate every time I mess up the firmware, I'll probably stop my experimentation here. I'm guessing what I want would require a more advanced redesign of the velocity detection code anyway.
Increasing the sample count to 8 made the main pads unresponsive and required use of the very useful FWUP boot option to salvage, but this sadly doesn't retain calibration data. As I don't want to have to re-calibrate every time I mess up the firmware, I'll probably stop my experimentation here. I'm guessing what I want would require a more advanced redesign of the velocity detection code anyway.
Statistics: Posted by toasty224 — Mon Jul 29, 2024 10:49 am