Joe - I've been flight siming for over 20 years, and have even logged 21 hours in the real fixed-wing thing, but this is the first time I've ever seen it suggested that the longitudinal axis is NOT designed to self-centre. However, I'm not disputing your knowledge as the internal design of the CH Yoke has very limited self-centering around the centre position. I assumed that this was compounded by the friction of the shaft, but going by what you say, any amount of lub isn't going to change that much as it's not intended to have a strong centring tendency.

The first two analog axes, those on the yoke itself, are rather obvious; they are intended to be used to control aileron and elevator motion (i.e. roll and pitch). The rear of the yoke grips are molded with contours that fit each individual finger. Both axes have centering springs, although only the roll axis returns to the same physical position every time the yoke is released. Steve, a CH Products representative, has explained the pitch axis mechanics in a post at the CH Hangar:

"The Y axis… has a general center area, it does not come back to the same center every time unless moved by hand. This is a design element that was requested a long time ago by users of the Virtual Pilot and derivatives, in response to the fact that those controllers had a center detent for the Y axis and the users felt this was unlike a real aircraft."

"So the Flight Sim Yoke and the Eclipse yokes have general center areas for the Y axis and will not pop back into a detent or specific center unless moved by hand. The center area is approximately an inch and a half of travel, some of which is taken up with the null zone (some times called the "dead zone"), an area of electronic non-response which is adjustable in most sims or through Control Manager."

But surely in the real world (fixed wing) the yoke self-centres because of the airflow over the elevator? IF it doesn't it would mean there was no use for the trim...