we use a sensor near (5cm) of a permanent magnet (45N).
Is the presence of the magnetic field can influence Data of acceleration ? and is that the disabling of the compass allows to have corect data for acceleration ? (still with the magnet)
I ask myself the question because when the sensor is connected via USB, in the "Data chart" , I see that acceleration is not influence near the magnet. Normal? so...what is the usefulness of the MI mode?
I just tested your situation. The magnet I used was small, as it was the best I could find on short notice.
When introducing a sensor to a magnet, I noticed no discernible change in acceleration while all components were enabled. For these tests, I left the sensor sitting still while the magnet moved around it. I tested the magnet both while static and moving around the sensor, both had the same result. While the Orientation of the sensor was impacted by the magnet, the only component that had any change was the magnetometer. After removing the magnet, the sensor would drift back to its original, correct position.
MI Mode is for when the both the sensor and the magnet are moving. The sensor can calibrate for static magnets in the environment and can take them into account in its normal calibration process. However, if both the magnet and the sensor are moving, the calibration biases are useless and the orientation data is skewed. MI Mode attempts to create a usable heading in such a dynamic and unstable environment.
I hope this helps answer your question!