Copter: hard Yaw can cause all motors to idle when yaw released.

[AndKe]

It's best if you try this for yourself, be sure to have idle spedd up on motors, or some might stall.
Fly at least 5 meters AGL (in my case, position hold) , and apply full yaw, You'll get up to something around 200deg/sec (default) let it reach full rotation speed, then release stick.
At least 50% of the time, rotation will break, the all motors will go to idle for half a sec or so,. causing a fall, then they recover.!

Attached screenshot:
from position 26.5, you can see 3 motors speed up, 3 go easier, I let the yaw stick go around pos27.16 you can see the pihhawk counter the rotation - then all motors go to thr_min, and it falls.

And a zoomed in:

[rmackay9]

AndKe,
can you post the original flight logs? My guess is that the copter is in AltHold mode and it's related to centripetal forces throwing off the inertial navigation acceleration estimates.

[AndKe]

Logs:
http://flyfoto-nord.no/stickbang_yaw.zip
I did it in loiter, the GPS/compass is not perfectly centered, it's 3-4cm toward front, making such position hold a greater challenge :)
Yes, this is just one of my crazy tests, I also used the flight to test rate_yaw_i (which got increased after this flight -i notices there's no option to tune it using RC6 ?)

Anyway - With your knowledge, you're most likely right regarding the reason, but I think it's funny it did not happened at the greatest rotation speed, but after the yaw speed were successfully reduced.

I did not test how fast/much rotation is needed,(like, if 180degree at full stick is enough) but I can imagine a fast approach with hard yaw right before touchdown, could result in crash.
Pixhawk is surprisingly fantastic :)

[AndKe]

I guess this user's motor stalled/got out of sync while testing: he's log is not as clear, but I see similarity:
http://diydrones.com/forum/topics/ac-3-1-1-crash-need-help-in-analysis

[rmackay9]

ok great. i'm going to close this issue then. If that's not ok just post a comment below and we can re-open.

[AndKe]

Please consider re opening.

This is pretty dangerous in hard wind, I did some one fast yaw today, and wind flipped the hexa once I released yaw, (because motors went to idle, and the angle toward the wind caused a extra quick flip) - It recovered with great authority and continued as if nothing happened :)
Please note that my hexa do not have sync issues, one that have, would most likely crash.

I do not have a new log for you, because logging stopped at line 220k due to the usual FAT/cluster problem, I am using a sd that have less problem with that, but not perfect.