Hi AquariuZ.
thanks for the explanation.
I personally think you found something. That will work.
You will probably run into problems with the swing of the pendulums, because the whole structure will try to swing/move and that will make a long free run hard to achieve. But I think you will do it.
Good luck building it.

Thanks Oscar. I have found after a day or so it is futile to make a desktop model using a bicycle wheel.

I can clearly see the effect and it is promising, but some drag comes into play as well. I need more mass, and better parts.

The trick is for the magnet to be placed so that it starts pulling the balls between 2 and 3 o clock - or shifting rather - and the balls leaving the field at three. Now I could persist in trying this small model but I feel even though it might work the next step (pickup coils to generate electricity) will not. So ball comes into the far edges of the magnetic field - gets a small jolt - and leaves the field.

Pleace it too close and the magnet may attract and hit a ball and cause breaking due to drag. Place it too far and it has no effect on the ball. Precision wanted.

So, I will have to put my money where my mouth is on this one and scavange for parts. Need a large round table top, bolts, wood to make a stand, pour a concrete base for the stand, some aluminum maybe for a magnet holder and serious uniform weights with a way to make correct pendulums. Bolts with eyes. Rods with double eyes to attach weights and create true pendulums.

Now if that big attempt fails I can easily alter the design to the one Mikhail proposed using a small ventilator motor to displace the weights instead. Can be closed system if started by hand. In any case: a lot of work is needed.

Another idea that came to mind is a powerful electromagnet in that location that is somehow driven and switched by the wheel itself (requires handstart in that case). That would come closer to Force on/Force off that Broli mentioned.

If anyone has tried building Mikhails version I wonder if they could try placing a magnet in the 2:30 region?

I´ll be back.

I took a look but that model is not wm2d. Please remember Mikhail uses a small motor to push the weights out, and I use a magnet to pull the weights outward. A simple difference.

I do feel that a good build of Mikhail can achieve OU without a doubt. Imagine a microventilator running at 0.5 watt with little arms displacing the weights making the wheel turn. I am quite sure you can get more than .5 watt out of a wheel that spins as such.

And I feel quite confident I can create movement using a magnet to displace the pendulums.

That simulation is made with Algodoo, and the model uses magnet, not motor. The problem is that the magnet pull the weights outside, yes.... but the wheel can't turn cause the magnet block the weights near to itself. However, I hope you'll build a working model. Good luck.

Here a few pictures what I am thinking of.

The wheel with bolts will almost look like a Western Saloon Gambling Wheel.

Thanks for this post.. I understand how it is suppose to work now. 

@oscar,

yes, good observation.

But there will be drag in either option which is basically the same as friction 

Even though I think there will still be an effect, the pull and the drag might cancel each other out-

But to be sure correct positioning of the magnet needs to be tried out.

I also was thinking about the pendulums being able to move only in a counter-clockwise direction using a spanner mechanism in the joint. Not sure how to describe it, but what I mean is something with teeth that does not allow the pendulum to swing left, only right.

That coupled with minimal Fx should be enough.

Look at this interesting video of a Mikhail Dimitriyev design:
http://www.youtube.com/watch?v=ODf49HtCNj0


only ¨pendulum" movement in CCW direction while the entire assembly moves CW.
and only two or max 4 weights needed.

If only I knew how to model this... Trying.

Thanks oscar

The mechanism you describe is a ratchet, also called a freewheel clutch.
http://www.ringspann.com/en/Products/Freewheels/Overview__403/

Yes, to have that is a good idea, in my opinion.

Exactly what I meant... thanks oscar.

When i started reading this thread i thought this looks promising. Before doing anything serious on this i would highly recomend a pendulum test as this is what your washers are. In my opinion the laws of physics would dictate that for every action there is an equal and opposite reaction and for each casue there will be an effect. My belief here is that the action of the washers being atracted to the magnet will result in an equal and opposite reaction to said latching that will reduce the weight of the washers and thus negate the weight advantage you are trying to achieve. Just as an asside this is kinda pointless as it can never go fast due to centrifugal force throwing out the whole thing eg every washer pendulum and so if you were to scale this up to a usable level say the size of a feris wheel with 1000lb weights and you were actually able to displace one 1% you would have a total force of 10lbs of usable work from it. The vid with the ball and wheel is awesome but that too is never going to scale past a toy.

Good points Ron.