So in a quatering wind, any delay in pedal position results in a cyclic input. Since the cyclic input was not really required then when the pedals are moved the cyclic needs to be re-corrected. All this means that cyclic effectively sees a disturbance, so the helicopter starts to oscillate.
Any thoughts?
My thoughts are that all the talk of quartering tailwinds, and student pilots losing control while hover taxying, has nothing to do with the subject.
A classic demonstration is a high power low IAS climb - say around Vy - with a pulsed collective input, or rudder doublet, and then observe the behaviour with the controls fixed. A divergent Dutch-Roll-like motion will ensue if no SAS enagaged. As I said in a previous post, the mechanism is different to FW - there's no equivelant to swept wings for instance, which is where the problem really appears generally (hence yaw dampers) - but the motion appears the same.