Today I was playing with autonomous Takeoff. 

First test was with remote, to see if everything is in order. 

The second and third tests went full Auto without any RC control and were complete success. 

Fourth test was supposed to be 1Km range flight test at 25m, but the quad-copter just went up, and crashed 70m from launch. 

For me, it seems that possibly had mechanical failure on compass fastening (since it is away from apm2.5 because of the motor wiring and currents.) 

Could someone check out my logs and verify? Problem is, that there is not much left of the copter to do after crash analysis of the fixture. 

Here is the flight path of the crash: 

Here is one flight before:

As can be seen from the pictures, something had to happen between these two flights. 

Also attached the log from the crash. 

EDIT: I did a stupid thing and did not re-check logs I attached. Fixed now. 

Tags: 2.9, arducopter, auto, crash, mission, takeoff

Views: 647

Attachments:

Reply to This

Replies to This Discussion

Hi!

I2Cerror seems to go berserk, but i am not the LOG guy, maybe i misinterpret things.

Normally I2C errors occur, if a device is adressed but not found/responding. Perhaps it's a wiring problem with your external mag?

Cheers Kraut Rob

Ps.: My Condolences for your crash !

I am not able to find i2c errors from my logs... What line should i look it from? 

Also, the crash is 50% my fault. Going to ignore rules, gotta be ready for consequences :). Well the location was remote enough for such playing. 

Hi Markus!

I did 2 screenshots. One is showing your I2C status, and an old log (2.8.1) from me.

You will see the difference....

Cheers

Kraut

Rob

Attachments:

Markus,

     I think that tlog is not from either of your crashes.  It looks like a log of just a stabilize flight where it doesn't go anywhere.  Maybe it was while it was sitting on your desk or something.

     Certainly the compass is not working at all in that tlog.  If you graph the xmag, ymag, zmag or mag_field you'll see they're all zero.

     Sorry to hear about your crash.

I uploaded the logs. Some times head-desk is not enough. 

As the correct crash logs were uploaded to original post I did a new check on my logs. The i2cerr is 0. I do not understand the reason for crash... 

Hmm, Savants comment doesn't seem to be very constructive - but my comments are also sometimes crazy...

I tried to look at the logs

First of all your copter seems to have very little vibration. The deviation from the course is also correctly seen by apm. Somehow motor2 seems to have severe problems. The apm gave full throttle on motor2 to compensate without success. The only thing that happened was a dramatic increase of vibration and amp draw without stabilization. So my guess is, that the motor probably worked and your airscrew/motorshaft got loose/propeller broke on that motor displayed as nr. 2 in the logs?

Just my 2cents

Cheers Kraut Rob

When I un-fastened my props (or what was left of them) the nuts were locked pretty strong. Also I am using nylon locked nuts. Since I have seen pretty weird things happen to materials in cold I can not remove the chance of the problem being a loose prop. 

For extra info, it was -8C outside. 

In that case my analysis was probably wrong. But on my little gaui with cheapo 8 Inch props i had one turning free at start in the cold (not -8! perhaps -1 or so), though i def. controlled all screws at home just that very flight before (i do it every 10 flights, regulary, except i hear noise or have other problems, i use those self locking nuts to secure the props). I didn't come across that problem with apc and graupner, but i don't fly at -8degree ...

Anyway Drone Savants hint with the "Stability Patch Induced Crashes" seems to be serious!

Drone savant's comment seems not to be the case here, since I had stable flights without remote the same day and also the day before.Even the wind was really nice and quiet. 

The tests that I did Day before were quite nice except the balance problem that made copter flip after 3 autonomous flights. The day I did the logs was after I balanced the quad. If there would be the code problem that Drone Savant implied to, wouldn't  it have showed when I did liftoff that was really Unstable for multiple tests (liftoff at 45 degrees or more and stabilization kicking in after 3meters of height or about 1-2s)

The more I think about this crash the weirder it gets... 

An interesting finding. After calibrating new compass and everything my copter is unable to maintain loiter with same pids as it was using for flying before. 

RSS

© 2014   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service