Autopilot fighting for control

Mxsmanic wrote:
In simulation, if I set the autopilot to maintain the altitude, and
then try to change pitch with the controls, the control surfaces
barely move, and I can hardly have any effect on pitch.

I think this is a limitation of the simulator. My guess is that, in
real life, the autopilot continuously adjusts trim, and so, if you try
to change pitch with the yoke, the A/P retrims to cancel out your
efforts, and the net effect is that you can change pitch, but you must
exert very high control pressures to do it (in order to overcome the
trim that the A/P is applying to neutralize your efforts).

Is this correct? I can see why this is difficult to simulate with
controls that don't provide feedback, but I want to make sure that I
understand the difference between the sim's way of doing it and the
real thing.

It takes pretty minimual control to over power the autopilot. However,
if you disrupt it a great deal (make a big pitch change for instance)
the autopilot may overcorrect a few times before settling back to the
correct pitch. The ability to return to the correct pitch with minimal
over correcting is a function of the quality of the autopilot.
Sometimes when you make changes such as lowering flaps you need to
"help" the autopilot but holding the yoke in the right place for a bit
until the autopilot figures out what is going on. This even with the
more expensive autopilots (like the new KAP140).

However, it is very easy to hand fly the plane while the autopilot is
on. The Cessna training instructors told us that they routinely hand
fly procedure holds on approach w/o changing the mode on the autopilot,
just over powering it.
In my plane I'll hand fly through turb but I'll leave the autopilot on
because it gives me a nudge back to course (i.e. I can feel it trying
to pull me back on course). If I take my hands off during turb though
it will make the turb worse.


-Robert