I am still trying to understand how to tune well for bigger props. I had a dismal failure with my first copter where I had 16 inch props on low kV motors (340 kV) and was trying to tune it, and the result was that it was either not stable enough, or had low amplitude high frequency stable oscillations which allowed a very safe flight, but made the platform almost unusable for camera. And it took me more than two weeks to figure out the best settings for this copter. Shawn helped a lot, and I also have learnt a lot in the process, so probably I could do it now much faster, but still, there seems not to exist a perfect tune.
Now when I am talking about big props, 16 inch is not very big compared to 24 or even 30 inch which are flying around, I am not sure if the criteria is the ratio of prop size to distance of motor from the CG, or the copter weight (disc loading), or both, and how they relate to each other (weight versus distance from CG). The first copter had 16 inch props 2.5 kg flying weight and a diagonal span of 75 cm. Switching to 15 inch improved significantly its behaviour, but unfortunately, the low kV motor mean that I would have to go to 7S or 8S to have enough thrust for high altitude safe flying, and I do not have ESCs nor batteries for this.
I built another copter, this time with 600 kV motors running on 5S and 13 x 5 props, 2.2 kg, and diagonal base of 65 cm, and it flew almost perfectly, with oscillations in the order of 0.2 degrees.
I tried to mount 15 x 5.5 EOLO props, did standard autotune with aggressiveness of 0.1, and that resulted in oscillations of the order of 0.4 … 0.6 degrees, about 2 … 3 times the amplitude of the copter when flying with 13 x 5 EOLO props.
The biggest advantage of 15 versus 13 inch prop is the efficiency – for hover, there is was a reduction of about 10 % of power with 15 inch props, and also much more efficient climb (maybe 15 % more efficient), but probably lower efficiency for horizontal flying, at least at sea level.
My options are:
A. Abandon completely the idea of bigger prop and maximum efficiency in favour of more usable, but less efficient smaller props.
B. Someone has ideas about the tuning of such copters, or even changing the design (bigger span) and could share his ideas about that.
C. Fighting out with LUA scripts, i.e. changing the PIDs on the fly based on the flight phase and potentially on oscillations present. This would involve a lot of fine tuning and testing, and the result would be highly dependable on the concrete copter, so it would be a lot of work for one of a kind, or rather one of a type copter. The concept probably is not too complicated, because in some flight phases (the criteria would be overall power, vertical descent, and attitude), the PIDs would have to be adjusted by LUA, but the exact values to be used both for criteria for changing the PIDs and the PIDs themselves would either involve a lot of experimental work, or theory which I do not master.
Any input?
