I set up two experiments for BMI088 as follows to analyze a relationship between the sensitivity drift error and the operating temperature.
1. Let the IMU be cool to -40°C in a cooling chamber.
2. Record the IMU data on a rate table at 300dps in the ambient environment until the operating temperature reaches 70°C.
1. Test the IMU on a rate table at 300dps in the ambient environment.
2. Keep recording the data while cooling down the IMU.
The figure below shows the comparison. I was expecting two experiments would give me the same result, but obviously, the rate is slightly lower in #2. Since I do not see related data in the datasheet, I have no idea If this what I could expect. Is there any related experiment being done on your side so that I can compare mine with it?
Solved! Go to Solution.
Thanks for your inquiry.
In BMI088 datasheet there is a parameter called TCS for gyro which is the gyro sensitivity change over temperature. The typical value of TCS is +/-0.03%/K and the min/max value of TCS is +/-0.07%/K.
In your #1 case the temperature changes from -40C to +70C with delta of 110C. So the sensitivity error should be within +/-0.07%/K * 110C = +/-7.7dps. At 300dps rate table speed you should see gyro outputs within [292.3dps 307.7dps].
In your #2 case the temperature changes from +65C to -15C with delta of 80C. So the sensitivity error should be within +/-0.07%/K * 80C = +/-5.6dps. At 300dps rate table speed you should see gyro outputs within [294.4dps 305.6dps].
From your plot #1 and #2, BMI088 gyro sensitivities are within the range. So the results you got are within the specifications which are good.
Yes, we have BMI088 gyro sensitivity test data and results from our lab. That is how we characterize the TCS parameter in the datasheet. If you need more information, then we need to place NDA with your company for those confidential information.
Thanks for the reply.
My main concern is about repeatability. Two experiments were done with the same IMU, but they have a slight difference. Does that mean the rate is different at the same operating temperature from two ways (cooling and heating)? Is this how the IMU works? Why is that? I am curious about this issue.
This is expected and it is so called temperature hysteresis. Please see more info at https://www.analog.com/en/technical-articles/gyro-mechanical-performance.html. You can also search online for gyro temperature hysteresis articles.