Gravity wheels with a fundamental difference

[Robinhood46]

when it gets to 10, it will flop over.

Did you not see that it goes all the way to 12 ( ish ) before i stopped recording?

I don't see any reason for it to want to return to the start position.

That is the bit i can't show with Algodoo.

The leg falls, as one would expect it to, from 12 ish to 3 or 4 ish, and in the process of falling it causes the lower weight to be raised from its current position to above the weight currently above it. Taking up the exact same orientation, and position of the R at the start of the cycle.
There aren't 36 ways, the falling weight can cause the R to exist in it's initial state, in it's starting position.

[Tarsier79]

id you not see that it goes all the way to 12 ( ish ) before i stopped recording?

Yes. if it call fall in Z, it wouldn't go to 12 unless you latched it, which is easy enough.

You want to use a falling weight downgoing to lift a weight on the LHS?

[Robinhood46]

The light weight, at the end of the "leg" doesn't need to fall on the Z plane. It mustn't fall on the Z plane, because it would mean the wheel needs to be far thicker than it was.
The light weight falls from 12 to 3 (or 4) as would be expected, passing through 1 2 3.
It's the heavy weight, on the ends of the "C" part of the R. that travels on the Z plane.
When the video stops, you can see the two heavy weights, one each end of the "C", which is back to front ")", the top weight stays where it is on the wheel and the bottom weight travels on the Z plane to be above it's partner. It does this as the light weight falls from 12 to 3.