I was flying today and all the sudden my quadcopter just took off. I was testing altitude hold mode and it took off. I immediately switched it to stabilize mode, but it was totally unresponsive. It was probably up 300 ft. and drifting, so I cut the throttle, still with no response. About 500 ft. away it crashed and I broke all 4 arms. It can be repaired, but it was very unnerving. Could someone take a look at my logs and see what went wrong, so this does not happen again.
You need to be a member of diydrones to add comments!
Replies
Yeah, I know where the code is. Even reimplemented it in Tau Labs for comparison to our EKF: https://github.com/peabody124/TauLabs/blob/filter_architecture/flig...
But code isn't math, and there are a few magic numbers in there I'm not totally convinced of, so I'd like to see the equations behind it.
I am liking the enthusiasm to get to the bottom of this issue. I'm a web programmer, so I don't know how much of help I would be on C control system code, but I'd love to make a web app where fellow Arducopter pilots can report control issues and upload logs. I am really convinced that my fly-away was not a fault of my piloting, but a bug in Arducopter's control systems. I am only going to be in stabilize mode until I see a fix these fly-away bugs.
This is a great answer. I am almost positive vibrations were not an issue. In the 5 seconds or so that the quadcopter was shooting up I lost visual contact with it and needed to bring it down before it got too far away. Also, it didn't shoot straight up but more like a 45 degree angle.
Moving forward, what do you think I should do so this doesn't happen again?
Not trust the barometer on this quad?
Set up a RTH fence?
Anything else you would do?
Why would you not trust the barometer? It recorded what you actually saw. It's the accelerometers that are lying. And that's almost always a vibration issue.
The first thing you should do is enable IMU logging and do a flight to record your vibration levels. At that point you'll know what sort of problem you are dealing with. I wouldn't bother with a RTH fence or anything else until you check vibrations. If they aren't taken care of any automated flying, like RTH, will be very unpredictable.
As previously stated, your barometer reading was accurate.
RelAlt is the variable which started to drop (despite your true altitude rising). Calculation is done via accelerometer, and then compensated via barometer error margin - which explains the extreme drop in RelAlt, followed by its recovery and alignment with BarAlt.
Sorry, I'm not sure how to help more, but I don't think the barometer was the problem.
Ok, I understand.
Sorry, you disabled logging of Accelerometers, (plus more) the crazy altitude could have been vibrations.
From line 5700 your Roll and RollIn have trouble sticking together, same for Pitch and Pitchin.
My guess is that something came loose, a motor/prop, or a prop desintegrated - anyway - huge vibrations can increase altitude.
Your MOT_SPIN_ARMED is 0 , and you say setting throttle to 0 did not help, the logs disagree.
I see some stickbanging before you put it down for good, and then the quad came falling dowm as expected.
Of course, having RCIN logged would help :)
Anyway - I guess something came loose and caused crazy vibrations, and unstable attitude. (which I guess you could see)
I am sure nothing came loose. The Arducopter was mounted to foam board to lessen vibrations and was still mounted when I recovered the quadcopter. I am thinking something is not right with the barometer. I am staying away from alt_hold for a while now.
ok, so you've set your mind onto blaming the pressure sensor.
I don't really mind that.
But if you actually want to find a cause, you should not blame the pressure sensor, which is a precision, but not high-tech component, and have no tendency to such failure, the actual sensor even covered in soft gel.
Your attitude logs prove sudden (huge) trouble keeping level - this is usually caused by something in the chain of propulsion being wrong - which in turn, ofteny causes vibrations.
@Eddie,
Could you please post the version of the ArduCopter firmware that you were using and some information about your flight controller so that we can put you in an assigned category to improve your visibility.
Regards,
TCIII Admin
-
1
-
2
-
3
of 3 Next