Vehicle disarm in air

Flight Testing & Log Analysis v1.13
1

Hi,

I had a weird situation where the vehicle disarmed mid air.
Looking at the log, there is a kill switch signal, but the pilot didn’t do it.

I’d be sure it was faulty signal from the receiver, but in input_rc.values.07, which is for the kill switch, I don’t see any signal change.

The difference on this flight was some different companion computer code, and I see larger CPU usage than usual, and more logging dropouts.

Maybe I’m missing something more obvious, I don’t know.

Log file:
https://logs.px4.io/plot_app?log=cafaf180-3d89-4583-9092-aefebd239ce4

2

Weird indeed. Maybe we got incredibly unlucky with the high sampling rate of input_rc (it only logs with an interval of 0.5s, compared to manual_control_setpoints with 0.025s).
Also other manual control inputs get crazy, .r and .aux1.

image
image1782×518 117 KB

3

Actually, it is seen on px4io_status/raw_inputs.07.
That input should go to 2000us when a switch is flipped, but it only went to 1800us.

And from the start of flight, vehicle behaved kinda weird.
I tried changing the waypoint, and it took a few tries for it to accept my command from the QGC.

Since I don’t have the log from the transmitter, it’s kinda hard to tell where the fault was.

Plot1
Plot11936×1353 74.2 KB

4

did you have any of the failsafes set to disarm? (Safety tab in QGC)

5

Nope. I don’t think it’s even possible to put it do disarm. Not something that should happen with failsafes.

6

Ok. Just FYI, all failsafes have a disarm/terminate option.

7

Yeah, you are right. Lockdown and kill are the same.
But anyway, i don’t have that enabled.
I have parachute in case of termination, but this unfortunately wasn’t that case, so it didn’t deploy.