Hi,

I have my Mavic since April, all good so far. I also have flown a P4 and have built a couple of NAZA M2s.

I always keep everything updated, and have the latest iOS 4.12 GO4 App, and firmware .800. Everything working as advertised so far. No crashes, no strange behaviour, all perfect.

This flight however was a little "glitchy". Soon after takeoff, after moving just up vertically I heard the motors spin louder for a very short time, like a gust of wind, but shorter and more intense.
I tried to move slightly the aircraft in all directions, and commands were responsive, but the start recording button would not respond. (Yes I had a card, and had just reviewed photo/video settings and formated the card in GO4).
At one point, the aircraft made a yaw jerk (maybe 20 degrees) and I decided to land to avoid any trouble. Landed without incident, turned everything off, restarted and the next flight everything was well.

All of this got me curious, so I've downloaded the DAT using Assistant2 and tried to analyze it using CSVView (https://datfile.net/Doc/intro.html). Could not find anything conclusive, but while looking at individual compass mods, rather then the agregated one, the variations looks odd to me. Also "direction of travel" seems to jump up and down very sharply.

Other than the small yaw jerk and video record button not working, everything was ok. I've checked IMU and Compass status prior to take-off and was the same as always, low values and on the green. The location is where I make a lot of flights since is near my house with no issues before.

The Dropbox link for the file (41 Mb) is:

https://www.dropbox.com/sh/i3bbo ... wHTbzwQPGe_Fqa?dl=0

If someone has the time to review the file it would be greatly appreciated.

Thanks.

Not sure when you heard the motors change. But, maybe the change in motor sound happened when the the throttle went from full to neutral? I know my Mavic will do this. See attached #1

The directionOfTravel signal requires some explanantion. The noise you're seeing is while the AC isn't moving, or moving very slowly. DirectionOfTravel is computed by DatCon, not the AC and is derived from the difference in successive GPS coords. If the difference is small then the directionOfTravel doesn't make much sense. Take a look at attachment #2. Maybe what I should do is change the algorithm so that it doesn't record a value if the difference in successive GPS coords isn't large enough.

Re the observation about the 20 degree yaw jerk. Now, that's weird and I don't have a clue why it did that. There was no rudder or ailerion input when this happened. It can be seen from the motorSpeeds that it was commanded; i.e. , it wasn't caused by an external input such as a gust of wind. See #3. The jerk was clockwise which is consistent with the leftBack and rightFront increasing and the leftFront and rightBack decreasing.

Thanks BudWalker for your help and insights. Much appreciated.

Understood on the directionOfTravel. I'm glad that it may lead to improvements in DatCon/CSVVIiew.

The attachments are very clear. While I sill may have commanded the throttle from full to zero in attach #1, I try to avoid just letting the sticks go, and try do do it progressively as I did from 52' to 54' in the picture. The step change just before 52 is odd, but I may have done it without noticing it.

What really freaked me out was the yaw jerk that you were able to confirm looking at the data. I was sure that I have not touched the sticks and I was looking directly at the aircraft. It was good that you confirmed it was clockwise from the data, because I have did not mentioned that in my first post.

I have made other flights since, and nothing abnormal. Interference, some strange voltage glitch during power-up, ghosts, aliens?
Guess we'll never know.  

Again, thanks very much for the help.