We’ve been trying to debug an intermittent GPS Yaw failure (works most of the time, but breaks at times). However, what we noticed is that during the times when the GPS yaw fails, there is a second weird symptom: the dataflash logs are missing (completely) during that period. I doubt that it’s a coincidence. Logically it sounds like the autopilot gets too “busy” to forward RTCM updates between the two GPSs, as well as to write logs in the dataflash. Fortunately, it’s not too busy to fly the drone.
Here is a dataflash log showing the occurrence:
In there there is a manual flight (LOITER) followed by landing - during all this, all is well. However, a few seconds after landing the controller shows GPS Yaw failure and the dataflash logs show … nothing at all (missing logs).
Any idea on what’s going on - namely, what can be leading to missing logs (and possibly missing RTCM forwarding) - is appreciated.
A bit of extra info: the autopilot is a cube orange - the load on the cube is about 37%-39%, not so onerous I’d think. The drone is “big” (15kg AUW) with a 12s setup (30Ah) with 23" propellers. The two GPSs in discussion are F9P on ArduSimpleRTKLite connected at 38kbps on UART3 and UART4 for the cube (and UART1 on the F9P). I’m not sure if I’m missing any other relevant info. Ah… 4.3.6 for Ardupilot.
I have other similar logs (with missing portions) when the GPS Yaw misbehaved, so I think that they are related.
Thank you,
Mihai