Here is a link to the unmanned drone I am developing:
It is technically a lifting body tilt rotor triicopter vehicle. To restate the problem:
the vehicle is basically a 200 lb drone with very heavy arms that give it a lot of rotational inertia. 60 lbs of batteries are located on the front of the side panels, 30 lbs on each side. The 1st test had only 2 batteries installed and the other 3 were outside in the ground. The vehicle hovered very well and was very stable. Below is link to see the 1st hover test:
The 2nd flight had all 4 batteries in the vehicle. It hovered well, but it was not as stable.
Similar behavior in the ~1.5 lb 1/6 scale model required the Roll P value to be increased to its maximum limit of 0.8. But the full scale vehicle is much heavier and needed a lot of batteries to make it balance at the correct CG for horizontal flight
The vehicle hovered at about 60% throttle, and Mission Planner shows that the back motor operates at 50% servo range, and the front motors operate at 60% servo range.
Again, I need to find a way to increase the Roll P value of the PID controller beyond its max limit of 0.8. If this is not possible, then perhaps there is a way to increase the differential control range of the motors? If it is not possible in Mission Planner, then maybe this can be done in the ESC software. In front we are using Fatboy 14S/15S 300A ESCs in front and Hacker Q100-5L 10kW 138KV motors with 32x10.7in Mejzlik 2Blade carbon props. In back we are using a 18kW cont/40kW peak 80KV motor with a APD 28S 120V ESC, and 2 of the same props used in front forming a 4 blade prop.