Drone flyaway help

Hello Forum,

This is my first post, so I'm not 100% sure if this is the format the group would prefer, however I am stuck in a little pickle.

I decided to build my first quadcopter back in September, which turned out to become relativly simple. With that choice, I choose the APM platform to allow me to explore autonomous operations, as well as its versitility.

My first couple flights were fine, but later on I started to encounter issues.

The day I first tested RTL, the first time I flipped the switch, it worked flawlessly. I went to record the RTL for my video series the second time, but it eventually led to a sudden flip and crash. (Video will be attached).

After repairing the copter, I did a couple of tests in my street. All of them were fine except for one, where I had no control over it drifting and then descending into my neighbors fence. 

The big kahuna of crashes occurred after that, where I once again, intended on testing RTL. The flight was going OK, until I flipped it into RTL, where I felt that it was not working right. I flipped it back into stabilize, where it starting drifting over a neighborhood pool. I'm not positive if I flipped it back into RTL or not, but the copter suddenly shot-up into the sky at full throttle, leaving it out of sight. Eventually we found the copter less than a quarter of a mile away from launch.

I'm still trying to learn the ins and outs of these systems, so I thought posting the video of the last flight, data and kmz logs of the last flight would help me receive some insight of how to fix my quad, as it is grounded until I know this will never happen again.

Raw GoPro video of crash: https://youtu.be/X02X2SxAvTk

All criticisim and comments welcome!

Thanks for the help,

Todd

crash log.kmz

You need to be a member of diydrones to add comments!

Join diydrones

Email me when people reply –

Replies

  • Do you have any photos of the vehicle?  I'd like to see the setup.

    • Haha, it kind of does.

      Actually, the placement of everything was coincidental.

      Also, it may not be in the pictures, but there was a vibration dampener (basically the same one you sent) on the APM.
    • I took a closer look and I see it now..

    • 3702612396?profile=original3702612096?profile=original

    • Looks like a 3D printed copy of an X-525...

      3702731923?profile=original

    • I do have some images, and I will post them once I have access to a desktop.
  • Developer

    This is one of the most horrible fly aways I've ever seen actually and it clearly deserves the term "fly away".

    The problem appears to be the barometer which started going bad 18seconds before the vehicle switched to RTL in response to a radio failsafe.  The bad baro data upset inertial navigation and leads to full throttle being output.

    There's not enough logging turned on to know much more about what's wrong with the baro but it's showing altitude changes of hundreds of meters within just a few seconds.

    3702160929?profile=originalAll I can suggest is to replace the flight controller, ideally with a 32bit board like the Pixhawk which has the CPU power to run Copter-3.3 which has an EKF instead of the simpler "inertial navigation" system which does a much better job of dealing with sensor failures.

    Really sorry for you troubles.  Thanks for the report.

    • Hi Randy,

      Thanks for the support!

      So you think I should just buy another APM?
    • Developer

      Todd,

      Yes, I think that's best.  I'd go for a Pixhawk (or compatible) though instead of an APM2.x.

    • Do we know why there was a radio failsafe?  Normally we would expect the pilot to take over manual control to stop the autopilot from flying away.  But I guess that wasn't possible in this case as there was no radio control.  Seems extremely unfortunate the radio failed just before the baro started glitching?

This reply was deleted.