SensorFusion: Yaw drifts in 9-DOF GBY?

[James_B]

What would cause yaw to drift in the 9-DOF case? It starts up relatively close to what the 3-DOF code produces for just the magnetometer.

However, from there, it just keeps on drifting, on and on, with the gyro. I have the robot stationary. I've looked at fbErrPl and it's quite small, rarely getting above 1/20 or less of fMaxGyroOffsetChange.

What would cause the gyro zero to not be meaningfully updated, such that the yaw just drifts and drifts?

By the way, I am using the ANDROID (ENU, gravity from accelerometer +) coordinate system, if it helps...

[JozefKozon]

Hi James,

could you please state to which part number are you referring to? Is it the FXAS21002? I have tested the FXAS21002 in our Freedom Sensor Toolbox-CE GUI. Please see calculated offset and noise for each axis. During the calculation the EVB was stationary.

Please see record for the Gyro. I was moving with the Gyro occasionally. Please refer to an AN5087 to Allan Variance: Noise Analysis for Gyroscopes.

With Best Regards,

Jozef