I've been off flying for several months now, mostly due to weather and schedule. As I look forward to warmer air and some time to tinker, I've been picking up where I left off with my quadcopter. Unfortunately, this is trying to determine whether or not my 3DR quad running APM 2.5 is still safe to fly after a strange event led to a minor injury for me.

On one of my last flights in the fall the quad did something I had never seen before. Upon connecting the battery (and while I was still crouched over the UAV) the four motors kicked on at ~25% throttle. Sensing the vehicle might rise towards my face, I attempted to get my hand out from underneath to guard myself. I doing so, I clipped one of the props, giving me some minor lacerations to my thumb and little finger. Within 3-5 seconds the engines stopped, and I pulled the battery (then went to get help to stop the bleeding!). The quad did not leave the ground, though it was close to doing so.

To clarify, this was within 5 seconds of connecting the battery, and long before I had even touched the transmitter, let alone time to arm/engage the motors. This particular vehicle, while it has been beat up a few times, had many hours of flight time over the course of six months. This particular event was anomalous.

I did several tests to see if the error would repeat before flying a few more times to run down the batteries for winter storage. The quad did not show any similar issues on 20+ battery connections (without props), or on those last two flights, where it flew without issue.

I have attached the (very short) dataflash logs for this event. The only thing I can get from them is that the craft was in stabilize (which is where the transmitter switch was set), and the throttle was non-zero for a short period of time before shutting off, which agrees with observation.

I have had some other issues with the APM unit, which mostly had to do with loss of orientation, and do not seem immediately related. I have telemetry logs for these events which I can dig up if needed.

I would love any information/advice I can get about what might have happened, and whether or not this was a fault of mine, or if I should retire this quad/APM to prevent further issues.

Observed Timeline:

Battery Connection.

(3-5s) Quad sitting as normal, LEDs on/blinking.

(3-5s) All motors up to ~25% throttle, no liftoff.

Motors stop.

Battery is pulled.

2013-11-19 08-34 111 (Freak motor start).log

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

Join diydrones

Email me when people reply –

Replies

  • Did you ever sort this problem out?? because Im having the same issue

    • Iv made a short video to show my problem,

      https://www.youtube.com/watch?v=4febYIhksfY&feature=youtu.be

      • While mine was not nearly that repeatable, I agree that it seems at least very similar to what I experienced. Unfortunately, I was never able to diagnose the problem, and because of that I have retired that drone indefinitely. I would LOVE to get some more information about it if you can figure out what might be the cause.

  • Maybe a dumb (noob) suggestion here...

    I had the same issue with my fixed wing a few months back. It turned out that the TX trim had somehow been bumped up to about half throttle. May be worth a look.

  • I'm as about as far from an expert as you can get with something like this, but your log was interesting.

    We're you running Simple Mode? It seems to think you were and wanted to head off at a heading of 33.6 degrees.

    The first recorded Throttle In value was 550, which suggests the APM was seeing a half throttle input signal from your RC receiver. It happily obliged and matched your throttle in signal with a matching throttle out. But then after a short time, the throttle in was dropped to 0. I can't say how long it was since your GPS hadn't locked yet and there's no time base recorded.

    Your first EV code was DATA_ARMED (which I believe armed the thing), followed immediately by DATA_AUTO_ARMED (from the Wiki, "pilot has raised throttle above zero and autopilot is free to take control of throttle".

    You also had a GPS Failsafe going at the time.

    So part of the answer might be that the APM was seeing a half throttle signal being fed to it. Possibilities could be an unnoticed raised throttle stick on your transmitter, or other RC transmitter/receiver weirdness.  What's not clear to me is why the arming occurred. Your recorded yaw in doesn't show the stick moved to one side. Were you doing any sort of Auto mission?

    • First off, THANK YOU for taking the time to read through the logs and give me some feedback. I can't remember for sure where the throttle was when I plugged the quad in. It might have been midway but I tend to doubt this. The transmitter was on at the time.

      I definitely did not arm the quad (I didn't even have time to!).

      I was not doing an auto mission. The plan was to do a little more flight testing with a new external battery on board to power some LEDs and the gimbal for a night flight that evening. The mission and the mission before were flown in stabilize (NOT simple), with intermittent alt. hold. I had no fences or other auto-enabling features activated (to my knowledge).

      I guess that is the question... How could it have gotten armed? Because even if there was some auto mechanism, it should not have been able to activate unless it was armed beforehand.

      • Was there any chance that the second "external" battery was powering the APM so that it never disarmed from the precious flight, and that subsequently the ESCs only needed the main bat. to be connected to light off the motors? 

        • That's a interesting thought, but I don't believe that the external battery was ever connected for this flight. I had only flown with it a few times, but I had been connecting it after the main battery. This external battery was a separate circuit from the APM.

          • Wow. Hmmm. Just for the moment let's presume that the second battery was indeed left connected to the gimbal after the prior flight. So you have no wired connection of any kind between either the APM or the Rx and the gimbal (or the lights)? You weren't using for example one of the Rx channels to set or control tilt on the gimbal?  I mention this because I have a similar setup, I have a little 360 Mah lipo that powers my gimbal and nav lights (I like the separation of power supplies between critical and auxiliary stuff) and I think the potential is there for back-feeding, as it were, the Rx/APM if the + servo wire is employed. So if you do have anything wired in from the gimbal system to the APM or Rx, then maybe it's possible that it energized that end of things enough to keep it armed? And maybe when the mains were unplugged it failsafed to, say, RTL      but was unable to execute that for lack of power to the ESCs until you plugged it back in? (I wonder what would happen if an RTL is commanded to an APM sitting on the ground and power to the ESCs is quickly disconnected, but maintained to the APM, and then restored to the ESCs a few minutes later; would the APM still be trying to lift off?)   All just idle (and no doubt ignorant) speculation, but something really weird seems top have happened here, so maybe looking at really weird possibilities is in order...

            The only other thing I can think of would be some sort of really gross level elecrto-mechanical failure.

            • Well, this was the first flight of the day, and I'm certain neither of the batteries were connected before I set up in the field, so I think the chance the APM had remained on is negligible.

              Could a gross level electromechanical failure include an APM issue? I'm trying to determine which parts of this are potentially unsafe (so I know what I can re-use). I am going to feel like 10 tons of stupid if I fly again and have a similar incident because I didn't figure out what was wrong! My fingers will likely always be a reminder of what these devices can do if your not careful...

This reply was deleted.

Activity

Neville Rodrigues liked Neville Rodrigues's profile
Jun 30
Santiago Perez liked Santiago Perez's profile
Jun 21
More…