BNO055 losing calibration

[simonenascivera]

Hi,

We are currently using a BNO055 as Absolute orientation sensor for navigation in an autonomous boat.
Even if we calibrate the sensor and upload the offsets at boot, the magnetometer keeps losing its calibration, sometimes with even 50-70° off from the calibrated orientation.
There's no metal object that could interfere in a meter from the sensor.

Is the sensor not suitable for our application? Are we doing something wrong?

Thank you in advance,

Simone Nascivera

[Vincent]

Want to check with you on following quesion:

When the magnetic calibration status drop from 3,  is this autonamous boat moving by motor driver or just stay in one position without moving? 

Please also try to plotter the sensor output especially magnetic sensor data here for deep analyzing. 

[simonenascivera]

We are currently doing the test by manually moving the boat. The calibration status of the magnetometer drops only if we move it. If we keep it in one position without moving, the calibration doesn't change.

This is the plot of the magnetometer from status 3 of calibration to 1. The violet line is the calibration status (from 30 = 3 to 10=1). The time on X axis is in milliseconds

 

[Vincent]

From the plotter,   there is quite big magnetic value change. 

When magnetic environment changes too fast / too big,  then the calibratio status will drop from 3 to 2 or even 0.  this will let system know that you will need a new calibration movement to bring back the calibration level to 3 again. 

When calibration level dropped,  additional offset found in each axis which will definately cause a heading drift. 

[simonenascivera]

So should we turn slower? We noticed that the magnetometer loses its calibration also when the boat slowly drifts because of the waves. Is there anything we can do in order to keep the magnetometer reliably calibrated?

Thank you for your support