I will do some more testing, but When I fly, I never make full travel inputs. That’s not how helicopter controls work, but I will test and see what happens.
In the mean time, do some testing for me on your end. Tell me what happens if you sit the controller down, and try to control it like an RC helicopter. Use your index finger & thumb on the stick you have assigned for cyclic and make only tiny inputs to control the S300. Just small little nudges of the cyclic stick.
The full travel of the cyclic is intended for different phases of flight. In a real helicopter, you would never make a full travel forward input, but rather small corrections in whatever direction necessary to keep your hover. To transition to forward flight, you would SLOWLY move the cyclic forward and keep it there to maintain momentum. You can’t just jam it forward and expect to maintain control. That would apply A LOT of torque from the main rotor. Which could theoretically be controllable, but would require a collective, throttle and tail rotor input of the same magnitude to maintain control.
I have never made an abrupt, full travel movement of the controls in a real helicopter to be able to see if the virtual response is correct. I would assume a right twist and altitude drop would occur on the same magnitude of the cyclic input I made, if I didn’t touch the other controls.
I really hope we can solve this so everyone can enjoy this helicopter. Or find something repeatable to help Nemeth fix whatever is wrong.