I programmed the following mission:
It is a quad, 3DR frame, nothing fancy, APM running V3.0.1.
It executed the auto mission until waypoint 3 then I saw the quad go down in an apparent controlled manner (it is not like i lost a prop) still moving toward waypoint 4, as if the quad wanted to land or something.
This is the google earth picture of the auto mission until interrupted at waypoint3:
I think I then switched to stab mode to recoup its descent but it crashed flipping upside down. I broke the telemetry module (antenna connector) in this crash.
Can anyone help me diagnose what happened and caused this strange crash ?
I see no reason : motors ok, propellers ok, battery was far from empty, I do not think I broke the geofence, ...
I looked at the onboard video and at the moment of the incident I hear a drop in motor power. Looking at the battery voltage it is not below the ESC soft power cutoff though. So I do not know what happened.
here attached the log and kmz files
Replies
You know what happened
http://diydrones.com/group/arducopterusergroup/forum/topics/optimiz...
This is why it is important to isolate vibrations from the APM, otherwise dangerous things can happen resulting in loss of craft or worse harm to operator or bystanders.
I will look at the log files, but what you describe is a controlled descent, very similar to an RTL, but without actually returning to "Home".
I experienced this last night on my quad, not in AUTO, but at the time it happened I was in STABILIZE. The quad stopped responding, and landed.
Everything was okay, I couldn't (at first) get it to Re-Arm. I had to switch the flight mode to LOITER then back to STABILIZE to re-arm.
What this is a "LAND" this will happen if the quad loses a GPS fix. This is on the GPS FAILSAFE parameter.
It is ENABLED by default, and will initiate a LAND if the GPS fix is lost for 5 seconds; see the manual extract here:
http://copter.ardupilot.com/wiki/failsafe/
I will have a look at the logs for you though, but that is what I suspect happened,