SparkFun Forums

[jordanstreet]

Hey guys, I'm in the process of calibrating the magnetometer readings for use in an EKF. The graph below shows data for three different rotations.

Rotation about z axis in red
Rotation about y axis in blue
Rotation about x axis in green

I know how to adjust offsets, reshape, and rotate the ellipse... but the problem I see is that the rotations are not centered in the ellipse. I can't think of any transformation that could fix this. Any thoughts?



Thanks!
Jordan

[Mee_n_Mac]

but the problem I see is that the rotations are not centered in the ellipse. I can't think of any transformation that could fix this. Any thoughts?

Are we looking at the raw (uncorrected) data ? If so the ellipses should be offset from 0,0,0 by the offset values. The question I have is whether the values are consistent for the 3 rotations. Or am I not understanding your question ?

[jremington]

Ditto on "Are we looking at the raw...". I don't think you have enough points to say for sure what is going on. Why not collect a full 3D set of points? Unless there are strong nearby intefering magnetic fields, the points should fall on the surface of an ellipsoid, as shown here: http://sailboatinstruments.blogspot.com ... ation.html