Getting weird ISM330DHCX behavior during Increasing Temperature Test

[EMoso.1]

Hello.

I am trying to perform temperature calibration on ISM330DHCX (bought from adafruit) for both gyro and accel.

IMU is sampled at 104Hz via SPI bus.

The IMU sits inside a glass container, and is placed in a freezer. When the IMU temperature reaches -10C the glass container with the IMU is taken out of the freezer and is placed in a hot room with air temperature around 40C. See attached temperature profile. Initial temperature increase is very aggressive - about 1deg C per minute, but then flattens out.

My concern is gyro and accel behavior during the first 100000 samples/1000 seconds. Plots attached. It seems that the IMU shouldn't behave like that during varying temperature.

So far I tested 3 different ISM330DHCX sensors this way and they all behave like that during the aggressive segment of temperature increase.

Any ideas on what might be wrong, or does this seem normal given the aggressive temperature ramp?

Any input is appreciated.

Thanks.

[Eleon BORLINI · Accepted solution]

Hi @EMoso.1​ ,

that's an interesting test. Please consider the zero g level and zero rate level variations decalred in the datasheet, p.10:

 

The total temperature shock of the fist period is about 50°C.

Regarding the accelerometer, the expected milli-g variation can be up to +-25 mg (this must be normalized for the steady state level). The 3 axis are inside this range.

Regarding the gyroscope, you might face up to 750mdps on 50 degrees (50°C * 15mdps/°C). The values you are reading appears a bit higher along Y axis, inside this spec for the other two axis. Are you sure you are not moving the board during the transportation out of the freezer, giving to the device an additional mechanical shock along Y axis?

If my reply answered your question, please click on Select as Best at the bottom of this post. This will help other users with the same issue to find the answer faster. 

-Eleon