Finishing a build of a large Octo which had flown out a full battery in Loiter for the first tests.
Next was an Auto Tune flight which started out fine, then suddenly from 5 or 6m up it did a massive flip to the left and hit hard.
3DR Pixhawk, APM:Copter V3.3.2 (7f16e4d6)
I have looked through the logs expecting to find signs of a mechanical failure but what I am finding is the 3 right hand motors being suddenly commanded to go to full throttle.
I cannot see any reason why and this is where I need help.
Can anyone please have a look at the logs to see if there is an indicator of what went wrong?
I am very reluctant to fly this Pixhawk again unless I can find the cause.
Hi sorry I’m really not experienced in reviewing log files, and in fact have my own up here that i’m hoping someone can help me with!
It does seem interesting that it doesn’t look like any auto tuning took place- it went int auto level mode but was still pre leveling- it hadn’t modified any rates.? i’ll try and go through it a bit closer tomorow, where abouts did you spot the motor throttling - which param were you checking?
looking at att - there’s a sudden roll - while desired roll is flat;
yo mentioned it had flown on loiter, did it respond to pilot control correctly too?
my instinct is to think its some kind of mechanical/ electrical failure rather than the flight controller- but like i said I’m quite new to these logs!
Hope you get it sorted and flying again.
guess with an octo yo may well have a fair experience flying so sorry if telling you to suck eggs, but on my current hex had similar couple of flips/ dive to side crashes, after about a min, i ran it tethered down and found one motor was glitching after about a minute and would occasionaly fail- was just a matter of resoldering ESC.
Well, doing an autoanalysis revealed some compass issues.
[quote]Size (kb) 1439.1103515625
No of lines 18444
Duration 0:00:49
Vehicletype ArduCopter
Firmware Version V3.3.2
Firmware Hash 7f16e4d6
Hardware Type
Free Mem 0
Skipped Lines 0
Test: Autotune = NA -
Test: Balance/Twist = NA -
Test: Brownout = FAIL - Truncated Log? Ends while armed at altitude 4.01m Test: Compass = FAIL - WARN: Large compass offset params (X:-124.24, Y:143.09, Z:-425.55)
WARN: Large compass offset in MAG data (X:-124.00, Y:143.00, Z:-425.00)
Large change in mag_field (429.71%)
Max mag field length (2003.36) > recommended (550.00)
Test: Dupe Log Data = GOOD -
Test: Empty = GOOD -
Test: Event/Failsafe = GOOD -
Test: GPS = GOOD -
Test: IMU Mismatch = GOOD - (Mismatch: 0.43, WARN: 0.75, FAIL: 1.50)
Test: Parameters = GOOD -
Test: PM = NA -
Test: Pitch/Roll = NA -
Test: Thrust = NA -
Test: VCC = GOOD [/quote]
Has this vehicle had a compass calibratoin/accel/level done? How about extended compass calibration?
Those compass numbers are HORRIBLE.
Do you have an external compass/gps? If so, where do you have it mounted at?
Also, it appears the autotune switched off after about 35 seconds after starting it. I also see a EKF bad variance error, which is likely caused by the compass issues.
@BigTulsa The compass is mounted on the tray the Pixhawk is on.
It is a Ublox M8N with a high mounted multi-channel antenna.
The Octo is back together and ready pre flight tests tomorrow.
I did notice the compass variances and had calibrated the compass a couple of times.
It might be worth relocating it but there is no where else to really mount it.
I am meticulous about doing all the calibrations before test flights.
The compass mot on these are always a bit hairy
@ninjamonkey the roll was what brought it down.
It had started to auto tune but a slight breeze had drifted it away to I started to bring it back, it seemed hesitant for a second to respond then did this massive roll to the left and went in on one arm.
Prior to that it had flown a battery out in loiter without issue.
No toilet bowling, just not as locked in as I would have liked though, hence the autotune.
Thanks guys for the time to look at this.
If I find anything I will post it.
I have looked at your log and the pixhawk is commanding full opposite roll and I can see it is being output to the motors.
This looks like a hardware or esc failure. This is a little unusual for an octo unless you lost a couple of motors at once or one of your motors went to full power for some reason. Do you have a video or any sign after the crash of a lost esc/motor/prop?
Sorry for the delay in getting back to this and thanks for the time Leonardthall.
I have obviously misread the logs as it looked to me like the right hand motors went to full throttle and the copter did roll left, so it was almost a deliberate act on the part of the FC.
This is what has thrown me as I could not understand what would cause such a roll after stable flight and part way through the auto tune.
I now have the copter back together and ready to test again so I will be able to verify the failed motor theory.
