Yeah i'm going crazy with the curve controllers now. Tried the PID but i'm not sure what is the control variable (integer ? % as in 10% = 0.1 ? is it relative to fan minimum or absolute 0-100 ?), and not sure what is the 4th variable (the delay for D ? is it a sample time in order to avoid peaks ?). Went through a square impulse measurement (thats a good one hour experiment here for just one impulse) and applied some known PID recipes, but ended up with too slow response and oscillation...