BHI160B Accelerometer Bias

[jannik_sarissa]

Hello,

I am presently utilizing the BHI160B sensor and have noticed some anomalies in the accelerometer readings. Specifically, there appears to be a persistent bias in the sensor readings for both the accelerometer and linear acceleration, up to around 0.5m/s^2.

I am aware that the BMI160 sensor offers functionality to correct this type of bias. My question is: How can I access or use this bias compensation feature in the BHI160B sensor?

Upon reviewing the BHI160B's datasheet, I found that it doesn't specify any such functionality. I would appreciate it if you could guide me on how to access this feature, or provide an alternative solution to counteract the bias.

Thank you for your time and assistance.

Best regards,

Jannik

[BSTRobin]

Hi jannik_sarissa,

Could we know what is about your hardware?
And how do you do the test?

[jannik_sarissa]

Thank you for your prompt response.

For your reference, the hardware setup involves an STM32 microcontroller, which is connected to the BHI160B sensor. We are using the official library to interface with the sensor. The primary role of the microcontroller at this stage is to transfer the sensor's data to a computer, where I subsequently log and analyze the raw values.

Regarding the testing procedure, an experiment was conducted in which the board, equipped with the sensor, was placed on a flat table in various orientations. Interestingly, the linear acceleration demonstrated a consistent bias error that varied depending on the orientation of the sensor. 

You can find a plot of the linear acceleration attached to this message. 

Thank you again for your assistance. I look forward to your guidance on resolving this issue.

[BSTRobin]

Hi jannik_sarissa,

When you test the sensor, does it remain stationary?
Do you use fusion data? If fusion is used, BHI160B's built-in BSX algorithm will calibrate the sensor and compensate for offset.