Chet asked of my knowledge of this device in a call last night and I was unaware of it.
It has started to bug me, but not to the point of disclosing personal thoughts on a more public forum, because i'm perhaps way off base.
All is good, nothing negative...just an idea that presented itself and that i'm not sure of, hence posting here.
If I may direct those not aware of the device, to the Energetic Forum thread:
http://www.energeticforum.com/renewable-energy/12341-mikhail-dmitriyev-input-1000-w-output-near-3000-w.htmlThe basics, as I understand them, are that a wheel turns, with a small bike wheel flinging metal pieces outwards at a certain rotation spot. The increased momentum of the more distant weight pulls the wheel faster, using gravity as the prime driver for energy gain.
It linked inescapably to a project i'm presently working on, along with a friend down in Florida. His pendulum is set up on a much larger scale than mine. His build has OU possibilities from various means and has decided to keep it all hush hush, my project has gone in a different direction and so now finds itself unwittingly linked to the Dimitriyev device.
Here's my pendulum rotor concept video (1min 57sec):
http://www.energeticforum.com/renewable-energy/12341-mikhail-dmitriyev-input-1000-w-output-near-3000-w.htmlIf it can be envisaged that the rotational energy will be in effect for a vertical rotor, mounted at the correct height, then the principle can be extended to that shown by Mr. Dimitriyev.
Using an egg shaped cam (as found on car engines), the rotor would turn and a magnet would be flung outward by the cam, on the opposite side to the pendulum magnetic interaction.
Such a force may then increase the speed of the rotor. The increase in speed will affect the pull and push of the magnets of the pendulum and so swing it too faster and/or further per swing.
Or - instead of the rotor increasing in speed, the extra energy is fed from the rotor to do electro mechanical work. A generator motor being driven by the axle on which the rotor sits (perhaps with an engaging/disengaging simple clutch). The rotor speed would then remain constant, but energy directed to power actual loads,
Being as the pendulum power source can be via any means, the output rotor energy is not directly related to the input energy required to run the pendulum.
By careful positioning, the rotor of my current setup appears to form little to no drag on the pendulum.
The pendulum is powered by a solar cell from a garden light, but will also run on tap water. An actual useful device would be powered by stronger means, but it shows how little power such systems need to begin running and then could have such power needs met to....run itself

So there we are, is such a thing feasible ?