I am flying a DIYdrones quad with the APM 2.5. Mission planner and APM are both up to date on firmware. This happened in a mission I flew yesterday.
Flew around for a couple of minutes doing Loiter and RTL. Both worked rather well.
Landed and set up a waypoint mission.
Took off, pointed the quad in the direction of the first waypoint.
Engaged automode.
Waypoint was to the west but the quad veered north and took a wide arcing path to the first waypoint.
Rest of the mission went fine.
My question is why did the quad decide to take such a weird path to the first waypoint when I already had it pointed in the correct direction?
Auto compass declination and auto mag offsets are both turned off. Compass declination was entered manually and mag offsets were done using the log telemetry option.
Logs are attached. Waypoint mission starts at around 13:06:00 on the mission time.
Replies
Your declination is zero on the vehicle log. Looking at the KML version (attached), the spiral pattern does "scream" compass issue. Though, LIRC, there was a time limit to reach the loiter point. If it times out, it circles around the loiter point--not sure if this feature is still around in 2.9+ (IMU controller). Also battery undervolt can cause spiral behavior (longer flight times)--which should have caused the copter to land, if you have a APM reg installed).
Looking at your offsets for that lat-lon, did you recalibrate them (manual flight for >5min or compass dance) for the area?
out.kml