Just my marginally scientific thought on the issue - it could be that near the ground, the air flow from all the props is being deflected out to the sides and causing a small area of low pressure in the middle, which is where the flight controller (and barometer) is located. You try to surround the barometer chip with foam to reduce wind/air currents, but it will still read larger general changes in pressure, which you want, and which this could be.
What's interesting is the inconsistency of the times when it'll bounce and flip on landing. I've auto-landed probably a majority of the times I've flown mine, and the landings have always been soft and solid with barely a bounce, if any at all.
I agree with the above - that flipping over is the result of a lateral move where the legs get caught on something, BUT a hard landing could be enough of an input to the accelerometers to be read as a change in position which the flight controller would then try to react to, which then leads to a lateral move, the legs getting caught and a flip.
What's interesting is the inconsistency of the times when it'll bounce and flip on landing. I've auto-landed probably a majority of the times I've flown mine, and the landings have always been soft and solid with barely a bounce, if any at all.
I agree with the above - that flipping over is the result of a lateral move where the legs get caught on something, BUT a hard landing could be enough of an input to the accelerometers to be read as a change in position which the flight controller would then try to react to, which then leads to a lateral move, the legs getting caught and a flip.