I am using an ADIS 16488 and I am getting the following graphs. The blue data is raw data taken directly from the IMU sample. The red line is the filtered output after passing the raw data through a butterworth filter.
In accel_still_graph.png, I kept the IMU still and the measured axis perpendicular to gravity. The bias measurement is fine, but I was curious if the magnitude of the noise is reasonable.
In accel_graph.png I pivoted the IMU about a joint, and got this result. Unfortunately I don't have a measure of the "true" acceleration, but again I was wondering if the magnitude of the noise is in a reasonable regime for this sensor.
Any input is greatly appreciated. Thanks!