I've managed to find time to restart work on my own version of Bessler's wheel. I say my version, but if it works it will clearly be recognisable as having been derived from Bessler's clues, many of which I think you are all unaware of, as I haven't shared everything to date - unless some of you are working on the same clues in secret!
There are five mechanisms each consisting of two equal weights and some levers. There are no cords or springs, just levers and weights.
The design still relies on the concept of parametric oscillation and I cannot see any alternative if you accept that gravity alone drove Bessler's wheel. I specifically chose the Estonian sport of 'Kiiking' to demonstrate exactly how parametric oscillation works and it's really very simple.
A parameter is a quantity or mathematical variable that stays constant. So if you have an oscillator such as a swing with fixed lengths it will swing to and fro until it stops, because all the parameters such as length, weight and gravity remain constant. But if you alter the parameters at each swing stroke, as a child does on a swing by swinging its legs at the appropriate point, you continue the swinging motion.
To obtain a variable in the parameters of kiiking, theoretically the person swinging has to raise his weight at two points during each revolution - and the same goes for the gravitywheel. However in practice only one lift is required and the return of the person's mass to its former position in readiness for raising it again, can take place with the aid of gravity as long as it occurs in good time time for the subsequent lift.
There is an extra factor or concept which I discovered about 18 months ago, which overcomes the objections to a gravity-only wheel, but I don't want to share it yet. Suffice to say that it throws out the window all the arguments about the viability of such devices.
JC