bmi088 gyro self-test failure with big z-axis gyro value bias

[rgiona]

Hi,

recently in one of our boards we observed a unusual behaviour.

On a bmi088 the reading of the gyro Z is quite off from the normal bias, a negative full scale that sometimes tend to change from 30rad/s to 8rad/s.

I tried to split the reading from multibyte reading to single register, and the value was actually in the register, no SPI bus reading error.

I implemented the gyro self test writing 0x01 to the register, and the bit of the gyro self test was on failure (=1 with flag self test ready=1), the background test on the bit 4 instead is always on 1 with no error detected (maybe is another kind of test?). Tried multiple time, the results were the same.

This component is pretty critical in our system, I would like to understand something more about this so specific failure. No anomalies detected on the accelerometer.

Any ideas how the gyro Z-axis might have been damaged in such a way? And why not the complete gyro?

Is there a way to extract the serial number from this device?

Kind regards

[Vincent]

In this case,  you need de-solder the parts then send back to us for further analyze. 

After deep analyze,  we can tell you if there is any damage in the sensor or not.

Please contact your local FAE for the failure analyzing.