IRI - 2014 Team Invitations

[Aren Siekmeier]

Originally Posted by hrench

I imagine you could extrapolate crash g's (decelleration) from the graph of the wheel encoders, as long as the crash was in the direction of travel. How accurate could that be?

The dt of a drive encoder is on the scale of 0.1 ms (360 ticks/rev encoder 1:1 on 4 inch wheels at 20 ft/s), so the largest acceleration you could meaningfully measure would be something around 20 ft/s / 0.1 ms = 200000 ft/s^2 which is about 6000 G's. However, this response would have to take place in the FPGA, which the cRio is not currently programmed to do (the encoder rate is also computed in the FPGA, built-in to the image provided by FIRST - WPILib just pulls this value). Outside the FPGA, the response time is more like 10 or even 100 ms, limiting your accel to 60 G's, or perhaps even just 6 G's.