This is the second time i have tested loiter, first time was perfect, second time the conditions were not as good (5mph winds).
The sequence of events went as followed:
Flying in stabilize switch to loiter
Hover in loiter
begin to fly forward
Yaw 180 to come home
during the yaw the quad tips right then left before landing on the left 2 arms.
I have checked the motors for sync issues and none are apparent,
props did not break in flight,
props were tight and not freewheeling
Compass was calibrated and position hold was working flawless
good sat count: 10
the only discrepancy i can find from my logs is my z-axis vibration. During the yaw the vibrations raise from -10 to 0 which are otherwise always a solid -10
I have analysed the logs to the best of my ability and though you guys would be my best chance at figuring out this problem.
My spec
tarot 650 iron man
Turnigy 390kv motors
Emax simonk 40a 6s esc
6s 8400mah
apm running 3.2 external compass and gps mounted high above motors and escs.
14x4 props
awg 2.4kg https://www.youtube.com/watch?v=GDLBvsFg6Fs&feature=youtu.be
This is just a classic lost of thrust crash. ESC Sync, slipping prop, bad bullet connectors, I can’t say. Arducopter just didn’t get the thrust it needed to control the attitude.
The change in vibration you saw where it moves to zero, that happens after the loss of control, and is because the machine was falling.
The change in vibrations seemed to happen before the quad fell. No chance of prop slip or bad connectors as all soldered (not dry), esc sync is something that keeps popping up altho with the current escs i am running i haven’t experienced any issues even when trying to induce sync issues on the bench.
I did look at the log and it does show that it rolled Right and then extreme Left which powered it into the ground. The problem is that the desired roll was in the opposite direction. Don’t know why it did that except for some hardware issue. Didn’t notice any from the video. Logging was minimal so can really get anything else from the log like what were the motors doing or what the RC input looked like.
I know this is apm but logging should be turned up until this can get figured out.
I do have telemetry logs which i forgot to include which have raw rc inputs albeit at a slower refresh speed. Thanks again for helping me with this, hopefully i will be able to regain my confidence with the multi.
I can see discrepancy in my raw-rc logs, all sticks were centered apart from yaw, occasionally i would adjust altitude, but other than that sticks were calm. The logs show some strange activity on the raw channels, its making me wonder if this could be a problem with my receiver.
I don’t think that would explain why the copter wanted to Roll left but instead Rolled right.
This is where RCOUT would tell us what the motors were told to do matches the direction.
I don’t really see much movement on NTUN.DPosX and DPosY, which would seem to indicate that there were no stick inputs movement. Well, except of course towards the end where you were very clearly flying it.
The small changes in ATT.DesRoll and DesPitch are due to the actions of the Loiter controller, not stick inputs directly.