TBS Discovery
APM 2.5 with Drotek M8 and external compass.
AC3.2 stable
3s 5000mah
Taranis tx with orange openlrsng
Quad had been flying nicely for quite some time on rc14 and the stable release of AC3.2.
Was flying around a regular spot in PosHold when the quad suddenly started to lose altitude after some fast flying, I tried increasing throttle but it seemed to have no effect at all.
After recovery I gave it a quick once over and as there was no real damage (1 leg extension had snapped it’s cable ties) I took off again. The result was a perfect flight in the same style as the first flight without incidence until the battery was spent.
The only odd thing I can see in the logs is a huge disparity between ALT and Barometer readings. The APM board is cased with a suitable piece of foam over the barometer and as I said earlier, the quad has had multiple good flights before this.
This is more out curiosity now as I have just finished upgrading this quad to a Pixhawk setup, but it would be nice to know if it was something I did that I need to change on the new setup.
I think what happened, is you were honking along at 15-20 m/s, and then pulled back to slow down. Pressure effects on the baro or vibration caused it t think it was higher than it really was, so it was actually trying to descend. It had just started to figure out the height and was turning the power back on when it flipped suddenly. I think maybe you had a sync issue or some other problem with the thrusters, and it fell.
Weird thing to watch was that during the descent, which was almost a 45deg glide slope, the quad maintained a perfectly flat attitude right up to when it hit the ground and then flipped forwards resting on it’s back.
I have a video from the flight but as the gopro is on a gimbal it’s not the most descriptive view.
Crash is around 1:09 and the rest of the video is after I brought the quad back to the car, and checked it over. Nothing was changed, then I landed put in another 3s battery and flew it like a bat out of hell again.
I had noticed the baro alt and alt had diverged quite a bit leading up the crash but wasn’t sure if that was cause or effect.