Unusual crash. Switching from RLT to Loiter

After re-building our H8 and re-tuning it with autotune. (previous post Autotune Crash) I had a few good flights. The rig was flying very nicely, autotue had done an excellent job at tuning the PIDs.

However, we went out on a job today and decided to to use this rig as it was best suited and is / was flying very well.
After a few flights I got lazy and desired to do a RTL. The gig flew back very nicely and started to land. At this point I jumped over to to Loiter as it missed the home point and was going to land in a pond.

As soon as I switched to Loiter the rig started to yaw clockwise and descend. I reacted by applying ccw yaw and waking the throttle full up.
I got no response, the hely just continued to yaw CW and decent into the pond, Unfortunately the camera was not recording at this point so no sound captured.

I have looked at the log several times and my understanding of it is:-
vibrations are low.
https://www.dropbox.com/s/uczdnodfzrdp7u9/Vibration.JPG?dl=0
I bereave the vibrations are low typically between 0.4 and -0.4.

VCC Voltage:-

Still fluctuating too much but there is also power from the rail and USB.

The point when all went wrong is exactly when I switched from RTL to Loiter.

Tracking, looking at DesRoll vs Roll it was racking very well up until the point there it switcher to Loiter around 19.5 mins into the log. The then starts drifting ever so slightly.

https://www.dropbox.com/s/qfvy0di1qi5va0b/RollvsDroll.JPG?dl=0

The point where is diverse off is when the Helli is lifter out of the water.

Yaw:-
You can see the Yaw at the switch over point. is basically starts spinning CW.
https://www.dropbox.com/s/3nbzr998sayaa7n/Yaw1.JPG?dl=0

Now something funny here. I’m not sure how accurate the timing stamp is but looking at the RCIN4 it does look as though I instruct a yaw at around 177478,1502 and the heli starts yawing at 177530,282.81 a little after. This is odd as as I remember it I commanded a CCW yaw to counter-act the CW yaw after it started yawing!!!
Also the DesYaw is matching the Yaw! this is also strange.

Altitude:-
At 178938,1548 you can see I hit the throttle (Altitude) up to stop the heli from descending into the water.
However you can see the altitude droops NOT fast but relatively slowly and is matching DAlt. So it os clear that the NAV is wanting the hely to decent. Question is why???
https://www.dropbox.com/s/071dvgg6o4ciez1/Alt.JPG?dl=0

As for the battery, i’m not sure which data in the log to look at. but they were at 73% when I checked after the crash…
[b]
So two issues, is the system miss decoding the Yaw commands? Is my radio going nuts?

but mainly,
Why did the heli drop and land in the pond dispute me commanding a increase in altitude!!
[/b]

PLEASE could somebody have a look and cast some light into this.

Pixhawk Log:-
https://www.dropbox.com/s/vxbrtca58eh0036/31.BIN?dl=0
Tlog:-
https://www.dropbox.com/s/hc2gsa24mkku2w6/2015-12-16%2011-53-53.tlog?dl=0

Thanks

Anthony

OK Yaw again, looking at the log again I think the Yaw command from me was a bit misleading, well to me it was. The Yaw command input stopped for quite some time whilst is was still yawing CW.
I do distinctly remember trying to counteract the yawing and stopped as it made no difference.
I stopped the input at 179150 and the yawing continued on until 180845 at which point the rig hit the water.

I am still confused as the desired yaw and altitude is wanting the heli to yaw and descend!

Is this some sort of fail-safe kicking in?? or the APM trying to save the heli from complacently falling out the sky?

Just as a note, I have tested this rig with 4 out of the 8 motors dead and it still flies well. So I presume if even up to 4 ESCs / motors failed it would still be controllable and land-able… IM CONFUSED.com

This could be a bug!
diydrones.com/forum/topics/copte … ta-testing

[color=#FF0000]Warning #1: Copter-3.3.2 fixes a bug found in Copter-3.3.1’s desired climb rate initialisation which could lead to a sudden momentary drop when switching from Stabilize or Acro to AltHold, Loiter or PosHold.
Warning #2: Copter-3.3.2 fixes an issue found in Copter-3.3.1 which could lead to hard landings in RTL or AUTO if the WPNAV_SPEED_DN was set too high (i.e. >400 or 4m/s) and/or the WPNAV_ACCEL_Z was set too low (i.e. <100 or 1m/s/s).
Warning #3: a bug was found in Copter-3.3 which could cause a sudden crash if you abort a Take-off initiated from a ground station. Video description is here. The bug is fixed in Copter-3.3.1 so we recommend upgrading.[/color]

Sorry ignore last post wrong version. it’s on 3.3.2 not 3.3.1.

i think I’m just grasping at straws now.

Hello Acpuk, Sorry for the late reply. I was drowning in work and could not have time to visit the forums.

Last time I have the Roll vs Droll behaviour was trying to setup a tricopter. I have posted it in viewtopic.php?f=111&t=14896

To summarize, it was caused by vibrations. Besides the vibrations were on the recommended threshold it was behaving wildly.

How do you dampen the board? Could you do a flight with fast imu logging enabled?