The greatest trickery of Bessler: a horizontal wheel tilted in a vertical plane?

[path_finder]

Everybody knows the the classic labyrinth marble maze game, where the operator shall control the pan and tilt.
Like here: http://www.youtube.com/watch?v=0TYD2CGl_f4

The idea is to use this very pertinent idea for the endless rotation of a eccentered mass under the centrifugal force.
Imagine a such as game where the path of the marble is a circle centered on the rotation axis of the wagon (in yellow on the shot below).
The wagon can be permanently located in the inclined part of the plate (the cart behind the horse), this inclined part being always in advance on the wagon within 90 grades.

The way to do that consists to link the axle of the wagon (the rotating pendulum) with a mechanism acting on the pan/tilting plate azimuth.
I built that before (see here: http://www.besslerwheel.com/forum/viewt ... 0421#80421).
But I made a mistake in the linkage, remember here: http://www.besslerwheel.com/forum/download.php?id=8985
My new design should solve this error.

edited: link to the video.

[Rob]

if the tilted plane revolves so that the wheel is always facing a downward slope, "the cart behind the horse", then the wheel will never descend, it remains at the same height and it does not roll. If it rolls, eventually the cart will catch up to the horse and be at the lowest point with no way to return.

[path_finder]

Dear Rob,
Thanks for your comment.
Any rotational move of the wagon (rolling weight) pushes away the downward slope. I verify this in a previous building.
This has nothing to do with the gravity, but with the centrifugal force.
Your comment is right on a static point of view, taking in account only the gravity force and without any rotation. It's like for the bike: if you don't move you fall. The depicted shifting effect exists only in case of sufficient rotation.
The job is made by the rotation inertial momentum.
The title of this thread supposes the gravity has no action on the wheel, tilted in a vertical plane and always balanced on a pure gravity point of view.

Hereafter a shot of the gimbal frame.
The linkage mechanism between the main axis and the both other axis is not implemented.

[path_finder]

Next step: the actuation mechanism of the gimbal.
The main shaft let rotate four gears wich create four alternative linear motions.
(the hole in the plate is useless, due to the recycling of an old part).
The ground plate is fixed and support a bearing where the main shaft can rotate, driving the four cogs.
The dephasing of 90 grades is obtained by an exact positioning of the gears.