LIS3MDL offset

glory_man · ‎2014-07-09

Posted on July 09, 2014 at 15:47

I received some results from LIS3MDL.

I tried to measure the magnetic field of the Earth = 0.05 mT = 0.5gauss. Full scale setting = �4gauss. If I rotate the magnetometer Y axis towards the north I receive 0xF3E0 (-3104) = -0.45gauss, if rotate Y towards south - 0xECAB (-4949) = -0.72gauss. If I rotate X axis towards north - 0x1288 (4744) = 0.69 gauss, X towards south -> 0x0B2E = 0,41 gauss.

Where did this offset means?

Maybe I missed something?

#lis3mdl-offset

onehorse · ‎2014-08-30

Posted on August 30, 2014 at 22:42

This is usual for magnetometers. Think of their uncalibrated response as a 3D surface. It is usual for this surface to be non-spherical and off center, both due to hard and soft iron effects (look it up on wiki). One has to calibrate the magnetometer. The easy way is to move the sensor all around and note the minimum and maximum values recorded for each axis. The subtract or add that value that brings the average of the two for each axis to zero, i.e., center the surface at the origin. If the surface is non-spherical then a more complicated bias correction will be necessary. But the first simple calibration will suit most applications.