Hello I am using the DMP on the MPU- to calculate quaternions it puts them in the FIFO buffer at a rate of Hz. I use these quaternions to calculate the roll angle (which is the angular position of a point on a wheel). As far as I can tell the DMP is only able to use the accelerometer in the +-g mode, I observe accelerations that are larger than this in my application (clipping) but the angle still appears to be correct even after extended sample times (greater than - hours). As the algorithm the DMP uses to determine the quaternions is propriety I am unsure how to investigate what is going on. Does anyone know where I could find material to help me get to the bottom of this ?