I was under the impression that they were pre-calibrated. After an initial hand
calibration, the heading resolution in the EKF is much better. However I think
a more robust method is required because my hand calculation was based on
rotating the sensor 180 degrees about an axis, then taking both measurements
and using that calculating the offset in that axis.
With 3 turns I calculated the HIF offsets for the sensor but I still needed the
scaling factors which require a bit more. I found this document which I will
implement through matlab:
It is attitude independent and will generate both HIF and SIF corrections with
a non-iterative algorithm. Do you think that information like this could be
included in the ADIS sensor datasheets? I found the information about HIF and
SIF offsets in the ADIS16400 datasheet to be really minimal.
Here is an article that may help them.