Hi,
I was testing out Circle mode and it started well, but ended in the worst possible way. Quad dropped like a rock :-/
Events:
1. Set Circle Radius 0 and rate to something like 20. Did a few circles. No problem. Landed.
2. Switched battery. Attached log starts here.
3. On the ground looked at circle rate and circle radius. My idea was to use ch6 to change circle rate between 0 and 20 degrees/sec and slowly circle a lonely tree in the field. I got really confused with how changing ch6 did affected the circle_rate variable. It jumped between -20 and 780 when I refreshed params. Decided to try it anyway with Circle Radius at 1500 as in 15m. Took off, set mode to circle, quad started making a huuuge circle very slowly. Ch6 seemed to have a really small effect but not what I expected. Abort flight. Land.
4. Decided to disable ch6 and look at this by the bench later. Set Radius to 150 as in 1,5m to see if that made the circle more to what i wanted. Took off, switched to circle, seemed good for a while and then quad dropped like a rock.
Log attached.
My thoughts:
1. Hardware fault that cut the power all-together of just from the APM. Seems unlikely as the logs show the drop.
2. My earlier fiddling with the params messed something up. I assumed that you can change params without booting the APM.
3. Bug that caused APM to crash? Maybe due to unexpected params as I might not understand the circle params...
Running ArduCopter V3.1.5 (3c57e771)
Replies
There's a violent apparently uncommanded roll at the end. Maybe a prop failure or such? I'm no expert log-reader yet so just a speculation. Also, it looks like you don't have vibration logging turned on.
I will say that I tried circle a couple of times under 3.1X w. Pixhawk and it seemed a little sketchy so I waited for 3.2, which I'm now running (Beta) and it works very nicely.
Hi, Thanks Oliver.
No prop failure. Two props where broken and leftovers where found under the quad. If props would have broken mid air I would not have found all pieces. :-)
I had just disabled vibration logging as it was just withing acceptable limits i.e. -+4 or 5. Earlier when I have had more vibration it has shot upwards and never dropped.