Values in ROS driver differs from output of Qinertia

[maxhofidrone]

Hello,
we want to use our QuantaMicro for Realtime UAV navigation. We encountered two issues:

1. In the ROS Message /sbg/imu_data the values for accl and delta_vel and gyro and delta_angle are identically:
   
   we were expecting that these values have slightly different values as delta_vel/delta_angle should be sculling/coning compensated (at least that was the case with our EllipseD and our understanding of the definition of the messages of the Firmware Reference Manual). Does the QuantaMicro performe these compensations? if so why don't we see any difference in the values?

2. we also compared the output of the ROS topics with the postprocessing result of Qinertia. Unfortunatelly these values seem to be different and their difference is increasing over time!
   
   The above picture contains the accelerometer Z values from ROS and the Qinertia output of the same 3 minute long flight path. Which data is used in Qinertia w.r.t ROS? They should be identical?

Thank you very much in advance for any help!

Max

[maxhofidrone]

Also when comparing the output of the ROS topic /sbg/imu_short and the export of Delta Velocity both readings are different! How can that be? Shouldn't be both readings be based on the same sbgEcom imu short message?

Unfortunately this ROS driver seems to be very unstable and not in line with the SBG Firmware reference....