We are enabling the FIR filter bank for our gyros and accelerometers to reduce the effects of vibration at frequencies higher than the required IMU output sample rate. So, questions have come up regarding the linear g-compensation for the gyros. The questions are prompted over concerns of phase relationships between the gyro and accelerometer samples and whether our choice of filters will affect those. We are also trying to decide whether or not the g-compensation should be disabled, even though we would prefer to leave it on.
First, referring to fig 20 (Sampling and Frequency Response Signal Flow), at what point is the acceleration measurement tapped off to be used for gyro compensation?
Is there some other g-compensation low-pass filter not mentioned in the data sheet?
And if so, what is the cut-off frequency and how steep is that cut-off?
At what rate is the g-compensation updated?
We are in the middle of integrating the device, so I will be hitting you with more questions on different subjects. Thanks for the help.