Pixhawk Sets Crash Bit and Disarms in Flight - Help!

ArduCopter Copter 3.3
1

Hey All,

Hoping for a little help here.
I had an interesting mishap the other day. I have a Quanum 680UC Pro Hexa-Copter from HK, running the latest version of APM Copter (V3.3.2 I believe) on a Pixhawk. I’ve been struggling with getting the PIDs where I like them. I’ve run Autotune a few times, but never really happy with the results. I have a buddy with the exact same setup, and he has had much better luck than I have. I think there may be a flaw in my assembly or I got a wonky Pixhawk.
Anyway, I had reset the PIDs to the “stock” 3DR default settings, and had flown a ~8 min flight in Stabilize mode. I was pretty comfortable with the copter behavior, so I hooked up a fresh battery and programmed an Auto mission to the local school and back (the school is at the end of my street). On the return leg the aircraft was fighting itself a little in pitch. Suddenly while doing 14 knots at 125’ AGL Mission Planner reports “Crash - Disarming” as I watch the front rotor spin down and the copter roll over and head to the ground before losing video.
Fortunately I was running the video through an EzCap to my laptop and recording the video on the Mission Planner HUD. The transmitted video was getting a little weak as the copter made the turn towards home, but you can clearly see what happened. youtu.be/U9Cbz91J7RA
Looking at the log files, I can see a bit of Pilot Induced Oscillation (PIO) from the Pixhawk, but the pitch excursions didn’t go beyond 25° or so. I really don’t know why APM Copter set the “crash” bit and disarmed in flight. I thought there was protections in place to prevent this.
Well, to finish the story, I ran up the road to find my copter upside down in the gravel parking lot you can see in the video, It landed almost perfectly upside down, smashing my GPS mast, FrSky Rx, arm switch and top Pixhawk plate (borrowed from a 3DR Quad kit). The Pixhawk survived, and once I disconnected all the wiring it would boot on the USB cable. I think this Pixhawk is going into an Antenna Tracker project next, because I’m not sure I trust it. The airframe looks to be OK, but I’ll disassemble the body plates before replacing the Pixhawk. The motor and props look a little scratched but undamaged since they were not producing any power upon impact.

Anyone have any ideas on why it “crashed” in flight?

2

rlog is only tlog + some mission planner stuff, not useful for analysis.
Post .bin from pixhawk.
Crash detection happens if desired attitude are not achieved within some time - the absolutely awful way it “flew” - if we can call it that, :slight_smile: , with some bad luck could have trigged that.

3

Andre-K, good call on the BIN file. I believe this is the one from the crash.

4

it seems you forgot to upload it :slight_smile:

5

Trying Again.

Had to Zip file to fit.

6

pitch axis have huge trouble, over 35degree error over time, earlier over 40degree over shorter time.
it’s not strange at all that this got interpreted as crash, (it’s supposed to happen if desired attitude is not reached within time, within a margin of error)

7

Thanks Andre. I was thinking that was probably the issue. I guess I just thought with the current version of copter that there were additional criteria needed to satisfied before it was determined to be a crash.

Hex has been repaired and flown one short flight. Too windy today for an auto tune and the default PIDs are flayable but not good enough to be doing much but slow flight.