Difference in IMU characterization - 42688 p on voxl

[Jagatpreet Singh Nir]

I was trying to understand the noise characteristics of the IMU - ICM 42688p on Modal AI M500 drone that has a VOXL board . I am trying to find the best IMU that I can use for my VIO application. I have another board with the same IMU - ICM42688p on it.

I collected stationary data for several hours on VOXL board and DK42688 board that I ordered and then plotted the allan deviation plot as shown in the figure. I am getting very different results for Allan deviation between VOXL and DK42688 board.
For comparison I also plotted Allan deviation of VN100 IMU (https://www.vectornav.com/products/detail/vn-100).

Can someone explain why the VOXL-IMU0 has more noise? Are there any settings which need to be done so that I get a similar behavior from DK42688p board?

[Chad Sweet]

You should compare the settings as configured in the code here. We provide the code so you can fork and implement your own configuration if needed