We're considering using the BMI088 or BMI270 for the measurement of the roll and pitch of a device using a Kalman filter to combine the gyroscope and accelerometer. Our main concern is long-term stability. Current consumption is not important.
So far I've read on this forum that the closed-loop gyro of the BMI088 has better long term bias stability. However, the BMI270 has a self-calibrating gyroscope. Which one would be recommended for our application?
Thanks for your inquiry. Please see the attached "Comparison chart between BMI055-BMI088 and BMI160-BMI270.pdf" for more information.
For your application, gyro sensitivity needs to be calibrated to obtain accurate pitch and roll angles with your Kalman filtering. If you will calibrate your device one by one on a single-axis or tri-axis turn table, then you should use BMI088. Otherwise, you should use BMI270 because BMI270 has built-in gyro sensitivity calibration which is called CRT (component re-trim) feature.
BMI088 accelerometer has built-in high vibration robustness feature for harsh environment applications such as drones and robots (e.g vacuum cleaners). This means that high frequency noise caused by vibration will be filtered out by BMI088 accelerometer. Therefore, the host processor doesn't need to implement extra low pass filter on the BMI088 accel data because it is already filtered. When drones/robots fall or crash, BMI088 accel will not get damaged. But the measured accel data will cause pitch/roll angles to jump. So your sensor fusion algorithm should take care of this.