p01rogue, from your results i'd guess that you need the power of DD to work at such high gains. i say that because i know higher gains require quicker response times to keep up without oscillation (wag). the stock tail definately is at a disadvantage to DD when it comes to response time.

another factor that will affect the overall response of the tail is boom length. john cut his boom to 10.5" (measured from rear of frame to cut), then glued on the tail mount & installed the DD motor. the stock boom is about 11 3/8" long from the frame, which does 2 things: 1) it reduces the moment of inertia (rotational mass) about the yaw axis and 2) it reduces the moment arm of the tail motor.

while part 2 works against us by reducing torque, part 1 (lower inertia) is the dominant factor. when the boom is shortened, the reduced inertia has more of an effect on acceleration than does the reduced torque. this can be seen in the basic formula for rotational movement: T=F*r=a*I~a*r^2. inertia (I) is proportional to the square of the radius (r), and torque (T) changes linearly with radius (assuming the motor is much heavier than the boom, which it kind of is). we can exploit the square radius of inertia, so the shorter the boom, the faster the response. the limiting factor of course, is tail rotor to main blade clearance.

fortunately, the more powerful tail rotor setup (gws DD w/ 3020 prop) uses a much smaller prop. so if you have DD, there's more clearance to work with when shortening the boom.

hopefully i'm not confusing anyone. the moral of the story here is tail holding performance on blades is improved with shorter boom lengths. so perhaps you'd get better results if you shorten the boom too, but you need DD to have clearance to do that anyways (it's all or none).

cheers,
-kev

[edit: this post & those equations should also be a hint on how to place components on our helis too. keeping everything as close to the head as possible will end up helping performance in all areas, including cyclic.]

[edit2: btw, i just noticed john is not making use of governor mode on the phx10 esc. very interesting, and i wonder why? maybe he had problems finding a suitable governor gain, or maybe he likes having excess headspeed going in to maneuvers? i haven't had problems using governor mode yet, and i doubt using this tail setup will change that. fyi, john's t/p curves are NORM: 0,85,100/-11,0,11 and IDL-1: 100,92,100/-11,0,11]