IRIS sinking to the ground after consecutive roll

I have a problem where the Iris sinks to the ground after consecutive opposing roll or pitch inputs. It can be most easily reproduced by commanding a rapid full left followed immediately by full right roll. The Iris will pitch back to level then sink to the ground.

While this is an extreme example, it has caused a number of crashes while near the end of a battery and attempting to land. If there is a gust of wind or something requiring a quick roll or pitch maneuver, the Iris will do it, but then rapidly sink to the ground. This has broken legs and gimbal a couple of times, as once it starts sinking it is EXTREMELY hard to stop.

Please see log attached, and image below. Also including a video example.

Youtube: …


If this happened near the end of the battery as you say - could it be that it actually was the end of the battery and your sharp roll maneuvers caused a voltage sag which then caused the copter to go down?

You have gotten rid of the lift vector with the rolls so its no wonder it sinks after it rights itself. When you increase roll you must feed power in as well. A good landing is a result of a good circuit. Try practicing a standard way of arriving that works for you and then always do it that way. Once your thumbs are dialed in you will recognize unusual situations early. Use yaw to point the nose and you will loose less height in the turn.

This was a BRAND new fully charged battery. I just did it like this to clearly demonstrate without breaking anything. Towards the end of a battery, while trying to descend, almost ANY moderate roll or pitch input will cause the thing to drop out of the air like a brick.

OK, but doesn’t this indicate it is underpowered? I can do this without the long legs and it will be much better, if I do it without the long legs & gimbal, its almost nonexistent. Yes, I can feed in power while it’s still a fully charged battery, but getting towards “landing time” (3-4 minutes with the gimbal & legs) I assure you there isn’t enough power to feed in to stop this from happening.

Again, this is an extreme example to demonstrate what happens when near the end of the battery and giving pitch or roll inputs. I don’t even have to do two in quick succession like I did here, any moderate pitch or roll input and it sinks to the ground FAST.

Take a look at eCalc. I know it’s not dead-on accurate all the time, but 80% at hover is worrisome, no? Obviously, at “stock weight” it performs great. Perhaps this is too much weight for it to lift?

[attachment=0]Screen Shot 2014-05-27 at 11.04.37 AM.png[/attachment]