Your pendulum and lever are not moving in way that Milkovic describes it. Masses are not calibrated to get proper movement. Pendulum angle is too small.
You should get something like this: http://www.youtube.com/watch?v=3CepH6OWze4

When the pendulum is at 0° gravity and centripetal force are both pushing down on the y axis and the mass on the lever is lifted up.
When the pendulum is at its maximums mass on the lever is pulled down by gravity force because on the pendulum Y part of centrifugal force is very small, it is 0 when deflection is 90° so its the ideal case. Minimal deflection is 60° in order to achieve milkovics principle.

I know this, and try to make something better. It was just a first look to refute the idea that a loop was not possible as some were saying. So the hammer can drive a flywheel from which a feedbackpulse can be timed properly at the positions you mentioned. I am also studying the flat inertial oscillator (disc in stead of a pendulum) which has an adjustable speed more suited for different types of flywheels and feedback constructions. The two phase loses its attraction in construction because of the complication of cutting the speed in half because of the double pulse. There is an essential difference between the flat inertial oscillator and the two phase gravity pendulum. Next to the gravity wheel and the magnet motor we seem to have a third oscillator option now that equally might fail to prove OU. But maybe a combination of these matters leads to a result. I personally believe in a threefold time, space and material mode definition of basic forces and principles of reality. Marjanovic says both oscillators are viable as long as there is a moving pivot point in both. Testing all of this takes a lot of time. Theories are too easy. What I know for sure is that it is an inspiring option, adding the oscillator to the recipies in our OU cookbook.

All experiments I have done with similar pendulums, does require more energy input if the lever are moving up and down. This movement will act like a break in the way a unbalanced wheel will viberate when it spins. If we manage to keep the wheel stable, it will spin freely, but if we allow the wheel to viberate (Because of the spinning inbalance) it will slow down the rpm's quite quick. The pendulum will also "swign forever" if the lever doesn't move up and down. This break works in this way: When the pendulum swings, the movement of the lever will always be delayed so it will counter act the movement of the pendulum. In an opposite scenario, we can activate the lever by hand in advance of the pendulums position to make it swing more.

So physics will therefor prevent more output than input, but the pendulum will be a very efficient way to provide as much output as possible compared to the energy we put in. However, never more output than input.

Vidar

Input/output logic applies to a closed system. We are looking for conversion of energy from outside the system. A kind of space energy rectifier. What you say concerns the presently sitting paradigm, but consider this: the oscillator might deliver a tension in a circuit that by complementary action combining frequency and amplitude might add up to the energy of time in the universe (or OU). We as yet have no scientific knowledge or experimental evidence of such a time-law of nature, but to my opinion we are looking for it by experiment. This oscillator tension as we can see in Finsrud's version of his rolling ball on track, is derived from a flat oscillator representing the expansion of the universe by centrifugal inertial force, while the pendulum function is of the vertical dimension to fix the frequency by the gravitational contraction of the universe. Finsrud practically implements the slow down of his oscillator as you mention. Thus we have two types of oscillators, gravitational and inertial, as I mentioned above.  On itself the slowdown is no objection to OU. A transfer of energy through a magnet motor (stator/rotor setup) might fix the amplitude of cyclic time in the feedback loop so as to achieve a proper control of natural free energy. This is the logic: control the wild energy of nature, not create it. Thus you have expansion, contraction and rotation as the basic ingredients for a mature time concept. This is what I am testing at present. It's a complicated design in which factually Milkovic is split in an expanding aspect of oscillation and a contracting aspect.

See my (flat-oscillator) Finsrud study on YouTube:
http://www.youtube.com/watch?v=E9GucVwc36Q
See also my oscillation by rotation and vice versa video at:
http://www.youtube.com/watch?v=lYAKzW0H2yY

Solenoid Powered Pendulum

http://www.youtube.com/watch?v=8oBteZ5_sOc
 
http://www.youtube.com/watch?v=zslTYb-HLf0

 
 
since the beginning of this thread (  01 December 2006, 00:11:41 )
 
<<  12 times more output than input, >>
 
and
 
nobody could give a proof of OU for this device !
 
so for me it is a scam !
 
so many people on this device why ?
 
what are you waiting for ?
 
what is your hope ?

Replica by Matthew Jones - Bouncer2_2.MPG

http://www.youtube.com/watch?v=LRkBamhlK8w

So, why should we be interested in these battery-powered devices?

because you can replace batteries with large capacitor which is being charged by generator connected to milkovics device.
these are only early tests with no overunity, classical mechanics.

I did some tests with 2 coil magbot pendulum, also powered with batteries. I get larger deflection by increasing voltage and lowering current through coils. So total integral of energy stays the same (power through time). Voltage is 4.8V, effective value of current is approx 60mA. Total energy over 1s is 1.152 Joules for pendulum with mass 36 grams. Large losses are in lego bearings, too much friction.

With better bearing and mass 15 kilos you need approx 500 joules of energy per second to run it continuously if pivot is fixet. Next test is to determine how much energy is transfered from input to output directly and how much is deflected back to input.
 

because you can replace batteries with large capacitor which is being charged by generator connected to milkovics device.
these are only early tests with no overunity, classical mechanics.

I did some tests with 2 coil magbot pendulum, also powered with batteries. I get larger deflection by increasing voltage and lowering current through coils. So total integral of energy stays the same (power through time). Voltage is 4.8V, effective value of current is approx 60mA. Total energy over 1s is 1.152 Joules for pendulum with mass 36 grams. Large losses are in lego bearings, too much friction.

With better bearing and mass 15 kilos you need approx 500 joules of energy per second to run it continuously if pivot is fixet. Next test is to determine how much energy is transfered from input to output directly and how much is deflected back to input.

So, then, post only when you have achieved a self-runner. Otherwise, posting results form a Milkovic non-OU you're reassuring everyone, critics the most, that self-sustaining Milcovic device cannot be made. Of course, I'm encouraging you to go on with the studies but am asking you to restrain from posting things that people have already seen -- a non-OU Milkovic device.