I have been setting up an Octo for some time, a Tarrot 1000 size frame with T-Motor U5’s.
It has flown well with the Pixhawk and UBlox M8N GPS , FW 3.3.3 although with slight yaw variance.
We have just fitted a Septentrio RTK GPS.
Initial tests with the GPS set to normal NMEA were fine with position variance less than .5m
Currently testing it with SBF enabled and RTK injected GPS corrections using Mission Planner.
We immediately had compass problems and after recalibrating several times were able to arm.
The copter showed huge compass variance and what I first thought as a fly away each time I switched to Loiter turned out to be very large diameter toilet bowling once I was game enough to let if fly away for a few meters.
This is curious as it was flying well before using both the Pixhawk compass and the Septentrio internal compass.
I have been trying to analyse the logs but there is just too much wrong for me to digest.
Using Dronekit-la I see “Severe scheduler overruns” that I have never seen before.
It also reports Compass fails and GPS fails, although our GPS readings were varying by less than 5cm on the GCS.
In the log display in APM Planner the log is ‘plastered’ with the message “GPS1: detected as SBF” which doesn’t seem correct and I thought would be loading up the logging functions.
I am hoping someone can have a glance at this log and hopefully point to something stupid I have done that I just can’t see.
2016-09-19 11-12-42.bin.zip (929.4 KB)