Toilet bowl/death spiral after switching to RTL

I have a F450 quad running Arducopter 4.0.5 that has flown reliably for quite some time in gusty weather and so on. Until today.

I was controlling it via telemetry dongle telling it to go to a waypoint when the low battery warning came on so I quit the program on my laptop that was controlling the drone and used my transmitter to RTL.

Well, at this point, the drone starts swirling around and around in roughly 10m diameter circles. The circles it flew in were not well defined however the drone did seem to be going downwards slightly. The swirling included a roughly sinusoidal change in altitude.

I used the transmitter to kill the motors at a point when the drone was a metre above the ground (before it flew several metres up again) and I got away with a broken propeller by some miracle.

I downloaded the log but it is 55MB of zeroes!!! The drone was still powered when I got to it. I had previously used RTL several times a few minutes earlier with the same battery.

Would some genius here be able to diagnose what went wrong?

Many thanks and apologies for the zero logs!

Got an update. Here is a video from Mission Planner with the drone stationary on the floor:

Does this mean my gyro is broken? Or is it the compass or accelerometer?


See if you can post a link to the .bin log so someone can check. Otherwise it’s only guesses.

1 Like

Please see attached a static log I took with the Pixhawk plugged in over USB.

log_1_2021-3-12-12-01-43.bin (996.1 KB)

So this Log was just sitting on the bench level?

1 Like

-0.3G on both accelerometer on Z axis while sitting on the desk is definitely suspicious. However is is very unlikely that both accelerometer went haywire at the same time. I think it is a power supply failure. Also vcc is 4.3 which is very low. So overall I think your flight controller is toast.
(Isn’t it a 2.4.8 ???)

1 Like

Actually, the drone was on its side. I forget which side. It was not moving at all though while I watched the GIF seen above in Mission Planner.

Edit: I have noticed that whatever is going wrong is transient. i.e. I have had the drone on my desk and not seen the GIF behaviour. Then again, most of the time it does show the crazy artificial horizon movement.

Edit: it is a cheap Pixhawk 1 clone.

That would have been useful information to supply before we looked at your log…

1 Like

Sorry about that. My main concern is looking for what is causing the animated GIF behaviour. i.e. something changing dramatically despite the drone being still.

As mentioned, in part of the log it could be working fine as the problem is about 90% of the time. I was hoping the craziness would be fairly obvious.

Just to conclude on this matter…

I am pretty certain my Pixhawk was crashed by the dodgy micro SD card in it.

The micro SD card that was in the drone during the crash was unresponsive after inserting it into my computer. Somehow maybe the SD card had a short circuit and caused a hardware problem in the Pixhawk?

I don’t know, but I blame this crash on the no-name brand cheapo SD card.