Quad suddenly flipped and crashed

[DUKE40]

Hi,

today during my first flight, the quad suddenly flipped uncontrolably and crashed, breaking the two rear props and one of the from arms. I was qietly flying in circle in Position Hold mode to check the quad, after I had rebuild it after a crash, due to wrong handling from my side (misestimated the position), and suddenly in about 2m heigth it flipped and went down with no chance to make any rescue attempt on my side.
I have tried to review the log, but could not find any hint about what caused this strange behaviour.
I have even the feeling that some signal was received causing this, despite no such stick movement occured on my side.
I want to make sure where the cause of this strange behaviour is, before I perform the next flight after rebuild.
Perhaps someone can have a look bat the log and find any hint about what caused this.

011-AQL.zip

Log of flight and crash

(3.44 MiB) Downloaded 636 times

My setup is a Spacewalk 550 frame with:
Flyduino SimonK 25A ESCs
T-Motor 2216/11
10x45CF Props
RX GR24
TX MX-20HoTT
Telemetry Mavlink2HoTT

Thanks in advance and best regards,
Jean-Gabriel

[chschmid]

Sad to see.

If you say that no stick input came from you, than a severe radio disturbance reached the AQ. It looks like all sticks went to almost full deflection and that caused a saturating/clipping of the outputs.

Your sensors seem ok to me.

Cheers
Christof

[DUKE40]

Hi Christoph,

sad to hear, but this confirms what I though to have understand from my log. What I am wondering, from MultiWii I know that if no flying in maula mode there is a limitation to the angle the Quad can roll or nick. Is there no such think with the AQ in PH? Obviously not, as this would have hindered the quad to complete flipp over, which was the case, it rolled 180° ending head down.
Unfortunately on the compound I was, no other RC-pilot was in the vicinity, and also the next habitations are about 1km away, so this is really disturbing, how such a radio disturbance could have occured.

Till now I thought HoTT to be quite safe, but this is shaking my confidence in HoTT. I never had such issues with Jeti for several years.
I will have to think about this while rebuilding and look what options I have:
* remain with HoTT, with the uncertainty about the reliability
* move back to Jeti on my old Mx16, but with some limitations, as I do not have any possibility to use two 3 position switches, so loosing the mission possibility, as well a srenouncing to have telemetry
* build an adapter to use Jeti in my new Mx20 HoTT, with renouncing to have telemetry.

[LPR]

Duke

I had a flip and a crash tonight. It's the first time I've seen my octo go to the wild side. I was trying to see how fast I could have the octo race around a small square of WP. I had it going fast by setting the PIDs for navi speed rate to PM=30, PI=30 and OM=50.
Then I was flying in Alt/Pos Hold with RTH switch left on. I then switched to the mission and stopped the mission in the middle by switching to Alt/Pos Hold all with the RTH toggle still on.
Because I was in RTH and stopped the mission the octo flipped three times and fell from 10 meters to the ground. I still can't believe that there was no damage because the one foot tall soybeans cushioned the fall and it landed right side up.
I will be very careful not to leave the RTH switch on in the future.

Larry

[DUKE40]

Hi Larry,

this is interesting, but does not match the situation I had.
I have not used mission or RTH up to now. Not at all.
The only funtionality I have been using up to now beside manual mode is Alt/Pos Hold.
Further I am still flying with the standard PIDs, no changes have been made to them.
And I was flying rather slow, no big or fast stick movements, just getting accustomized to the way AQ is fliying in comparison with MultiWii.
Further I am still more in the hover fraction than really flying, rather trying to control the quad in all situations before moving to the next level.

Jean-Gabriel

[LPR]

Jean

The AQ normally will not tilt further then the limits you set with the PIDs. The saturation that Christof mentioned is what caused the crash. The AQ is sensitive to acceleration which makes it more likely to have it's IMU saturated with unusual RC inputs. I hope you can find the source of your lose of RC link.

Larry

[DUKE40]

Hi Larry,

there was in fact no loss of RC link as through the whole log the quality of the RC signal is at 100%.
In fact looking more intensly to the AQ Log, I can see that the large RC input occured after LOG_IMU_RATEX, LOG_IMU_RATEY and LOG_IMU_RATEZ started to increase without any conscious stick input change from my side. Those large stick movements were probably made by me as I saw the Quad flipping over and getting out of control towards the crash. Panic reaction which might have accelerated the whole process.
I will probably fly the next times with some fear this might occur again, before regaining full confidence in the whole configuration.
I have a few though about what went wrong, but only the future will tell if my assumptions might be right or a totally wrong. I will probably never get 100% certainty about the source of it all.
Anyway, many thanks to all those who had a look at my logs and provided advise or hints about the possible causes of this.

Regards,
Jean-Gabriel

[DUKE40]

Hi,

today when starting to work on my Quad to repare it, I found finally the root cause for the whole issue. The left from prop was loose.
I always check everything during the pre-flight check, so the prop must have gone loose during the flight and it caused the crash. I will think about how to prevent such situation again.

So I a reliefed that the crash was not caused by a techincal glitch with the radio or the FC, nor was it due to wrong stick movement from me.
The cause is something that can be easily avoided with corresponding measures.

[bluuu]

Try to use glue for screws e.g Loctite 243

[DUKE40]

That is exactly what i will do