It can be proven as an Overunitiy system but only if

[mikestocks2006]

It can be proven as an Overunitiy system but only if:

The cost (energy cost) of aligning the magnetic domains from their relaxed state to let’s say ½ saturation is greater or equal to the cost (energy cost) of aligning the domains to same saturation level from their orientation under the influence of let’s say 2 same polarity facing magnets.

Let’s take a ferromagnetic material e.g. ferrite rod. The rod maybe easier to visualize, however the principle will be applied to different geometric shape later.
For illustration purposes the  rod stands vertically and it has a rectangular cross section as we look at it from the side.
The rod is  full of randomly oriented magnetic domains each domain having it’s own flux and orientation, but the net result is that the sum of all fluxes is ZERO. The NET Total flux is zero.

We wrap a coil around it, and we energize the coil with enough energy for a given time so that the rod reaches for example ½ saturation. And the net average domain flux orientation is let’s say bottom to top. Let’s call this amount of energy Ei (initial) that is required for this initial transaction.
At the point of ½ saturation the sum total of all domain fluxes is the result of domain forced orientation due to the coil magnetic field and for this example  is ½ of the total achievable flux for that particular rod.

At the end of the given time we stop the current and as the rod magnetic field collapses the magnetic energy can be converted back to current through the coil, as the domains revert back to their initial random orientation where the sum total flux of the rod is ZERO again Let’s call this energy Ec (collapsing field)

For the shake of simplicity lets assume we have a R=0 resistance of coil “perfect” conductor, and perfect inductor/material so we have Zero Losses there too.

The energy balance then is Ei=Ec
Energy Balance and mechanics of transaction:
Electirc energy to magnetic energy as a  result of domain alignment  during the initial transaction. = Energy given back when the field collapses as the domains relax and revert to their original random orientation.

Conservation of energy is preserved, and in the perfect scenario of resistive (heat losses) and core losses equal zero, the best we can achieve is unity. Correct?

Ok now it is where the first statement above come in play.

Lets take the same rod, and above the top of it place in close proximity a magnet with eg. North facing the top of the rod. AND also take another similar magnet bellow the bottom of the rod  with north of the magnet facing the bottom of the rod.

The magnet strengths and position to be such that the rod will not exceed 1/2 saturation at each end.
Now lets see what happens at the magnetic domains:
The ones close to the top are aligned point toward the top of the rod
And the ones towards the bottom are aligned facing towards the bottom of the rod.
Having similar magnets and same locations would result in half of the domains in the rod facing down and the other half facing up. Yes there are gradient of orientation but the net average effect for the top half of the rod is the same as the bottom half in magnitude but opposite in polarity.
The NET Total flux of the Rod is again ZERO.

Now here is where the first statement above comes in play.
The question now stands as follows:

If we now energize the coil with Ei, will it be enough to again saturate the coil to 1/2 saturation? And will the net average domain orientation bottom to top be same as the rod/coil without the magnets?

If yes then the principle applied to a toroid feromagnetic material with wound coil proves overunity.

Substitute the rod coil in the above example with the toroid cross section where it’s facing the two magnets. The cross section is a square or rectangle and let’s say the top magnet is close to the top right corner of the coil and the bottom in close to the lower right corner. (as in many replications using 2 magnets per position both facing the same way)

Next when  the coil is energized will tend to align the domains perpendicular to the cross section, and thus there will be no longer attraction (or to be more accurate, less attraction to the magnets than when the coil was not energized) Now this will explain why for example, replications and the orbo rotates. Less/no attraction after the coils energization compared to before. So rotor gets pulled in and then the attracting force is nullified due to core domain forced orientation by the coil

But to prove overunity, the Energy cost from random orientation NET ZERO flux initial state to ½ saturation, must be more or at least equal to the Energy cost from ½ up and ½ down aligned domains state again NET ZERO flux initial state, to ½ saturation. Remember that we can collect that Ei back as Ec when the field collapses.

If  at worst the transaction costs are the same, then the mechanical motion is the free ride and the Overunity section of the system.

One way to prove this is experimentally, by measuring the energy input during the on pulse, and the energy collected during the domain reversion to random, plus account for all the losses heat, induction, eddies if any including error analysis .etc.

1/2 saturation was used as a safety factor, to assure all the input energy to the coil goes to domain orientation. Anything  much above saturation is wasted for the purposes of the example.

Magnetic viscosity as a time variant property vs rpm / pulse width, and coil/core inductance change (energy stored available) as a result of external magnetic fields will need to be accounted for in a more detailed examination but that’s another topic.  ...

Thanks
Mike

[exnihiloest]

...
Now lets see what happens at the magnetic domains:
The ones close to the top are aligned point toward the top of the rod
And the ones towards the bottom are aligned facing towards the bottom of the rod.
Having similar magnets and same locations would result in half of the domains in the rod facing down and the other half facing up. Yes there are gradient of orientation but the net average effect for the top half of the rod is the same as the bottom half in magnitude but opposite in polarity.
The NET Total flux of the Rod is again ZERO.
...

Hi Mike,

Flux do not vanish. they transversally escape the rod at the center and loop outside of the rod through the air towards the other magnet poles (these not pointing to the rod).

...
If we now energize the coil with Ei, will it be enough to again saturate the coil to 1/2 saturation? And will the net average domain orientation bottom to top be same as the rod/coil without the magnets?

If yes then the principle applied to a toroid feromagnetic material with wound coil proves overunity.
...

You said the magnets at each end of the rod have opposite flux.

Thus the magnetisation from the coil will add the flux of a magnet at a rod end, but at the other end, it opposes the other magnet.

If it saturates one magnet end, it will de-saturate the other one.

There is also an important thing to know when controling ferrite saturation: the saturation depends on the direction of the field that saturates the material. This implies that the source of saturation field will always couple with the circuit (coil) that would exploit the saturation, resulting in forces between the two ones (for the mechanics side) and in Lenz's law (for the electrical side).

[mikestocks2006]

“Hi Mike,
Flux do not vanish. they transversally escape the rod at the center and loop outside of the rod through the air towards the other magnet poles (these not pointing to the rod).
You said the magnets at each end of the rod have opposite flux.”

Yes as far as direction. The rod domains will align so it looks like South on top, and again South on the bottom equal magnitude but different orientation. - Ends of rod facing the North poles of each magnet.

“Thus the magnetisation from the coil will add the flux of a magnet at a rod end, but at the other end, it opposes the other magnet.
If it saturates one magnet end, it will de-saturate the other one...

Yes in the case of a rod, all the vectors of the domains will tend to align in parallel to the rod top bottom axis. The rod was brought up so it maybe easier to visualize what happens inside, looking at its rectangular cross section. But we are mainly concerned with what happens inside the toroid, in the proximity of the 2 magnets. Then looking at a toroid the alignment of the domain’s flux when its coil is energized will tend to be perpendicular to the top/bottom direction. (tangential to the effective diameter of the toroid.) It will not cancel one and reinforce the other, but rather will tens to nullify the attraction.

I hope this helps.
Thanks for the feedback.
Mike

[mikestocks2006]

A bit of along post, but follow till the end before adding any comments if any.
To simplify a bit more:
The heating effect of the toroid resulting to a loss of energy should not be part of the central OU qualification.
Let’s assume that the coil is made from a superconductor at room temperature, does that make the whole device an ou machine? Or any device for that matter?
A superconductor can only theoretically go to unity, current flow with no resistive losses.
So let’s focus on the magnetic transactions for now, as the center of the OU discussion.

Let’s compare the two scenarios (same angular points for on/off switch)
a.   Coil energize, steady state, de-energize, WithOut the rotor magnet present anywhere near it.
b.   Coil energize, With magnet,  when the magnet is at TDC, steady state current, and de-energize when magnet is far away.

What is the interaction between the toroid core and the magnet? And core and coil?
Let’s designate the following:
Hc  the magnetic field created by the coil at Imax, Imax is also current at steady state.
Hm  the magnetic field provided by the magnet at TDC
Bc  the B-field density of the toroid core, as a result of the Hc
Bm  the B-field density of the toroid core due to the Hm

Note that the Bc and Bm do have a vector quantity.
Bc is circumferential, and Bm is in general perpendicular to the circumference, as the domains want to align towards the magnet.

So now lets go back to a and b above and examine the stages of interest.
-For scenario a,
There are 3 steps.
Step1 the current rises from zero to Imax  and an Hc magnetic field is constructed that induces a Bc on the core.
The energy required, function of I=0 to Imax integrated over the time it takes to go from 0 to Imax. This energy will coerce the domains of the core to a circumferential direction and as long as the current flows it will tend to maintain them there.
Step 2 the current is maintained at Imax for the same duration as if the magnet was passing buy (remember no magnet in scenario a) and the domain are at a circumferential orientation
Step 3 the current source is removed, switched off, the Hc collapses, the Bc wants to get to 0 as the domains want to go back to about their original random disorderly state (B-H curve hysterisis accounted etc)
The energy store in the core in this scenario is either dissipated on a spark back over the switching device or dumped over a shorting diode, circuit resistance, or collected etc.
And if the coil wire is super conducting, the most we can expect is energy from the collapse of the Bc is the same as the energy it took to create it.
Note that in the above scenario the domains in the core went from random, disorderly orientation  to a circumferential orientation back to about their initial state

-For scenario b, and as the device operates with the magnet involved in the transaction:
Again there are basic 3 steps but now the initial conditions are different.
Step1 Magnet is at TDC and the coil is energized. Let’s look at the core.
Just before the switching of the current, the core is under the sole influence of Hm
There is a Bm already induced on the core, and the domains have a general radial orientation towards the magnet.
When the current is turned on it again rises from 0 to Imax, but now the Hc is trying to induce a Bc and attempting to orient the domain in a circumferential direction, so the attraction between the core and the magnet is reduced. However; the core now has a B-field that is a result of two external influences Hc and Hm.
Let’s call this resultant B-field of the core Bcm
In addition, just before switching the current on, the permeability and coil inductance are different at this point as compared to scenario a. due to the already influence of the Hm
Here the main question is, how much energy is pumped into the coil to get to that Bc?
Since the inductance and the permeability are now different, the time it takes for the current to rise from 0 to Imax may also be different including the general curve profile.
In addition the Bcm may not the same as Bc (scenario a) neither in magnitude, nor in domain resultant orientation. The core domains are somewhere between radial and circumferential orientation. The more the circumferential the better the “shielding effect” etc.
Step 2 magnet is moving away up to the point that is far enough so it doesn’t influence the core anymore. During stage 2 we have an apparent steady state current flow within the error measurements of the scope traces (flat part of current curve) and at end of stage 2 the core is about back to Bc with no Hm influence anymore. I use the word about because the resultant radiant orientation of the Bm can not fully revert to 0 due to the hysterisis B-H curve etc.
Step3 same as scenario a, and we maybe having about the same energy out of the core.

Now let’s designate some of the energies.
Ec(a) is the energy required to construct the coil magnetic field in scenario a step1
Ec(b) is the energy required to construct the coil magnetic field in scenario b step1

EBc(a) is the energy stored in the toroid at the end of step2 scenario a
EBcm(b) is the energy stored in the toroid at end of step2 scenario b

Emapp is the kinetic energy of the magnet/rotor at TDC, rotor gains rotational speed and rotational energy as it approaches TDC, and it is max at TDC. I’m excluding friction for simplicity.
Emdep is the kinetic energy loss as the magnet rotor departs from TDC.

The device results in a Emapp > Emdep, aiming to reduce the Emdep as much as possible. The net kinetic of the rotor energy per on/off cycle is Emapp-Emdep
Let’s call that Er. Energy of rotor per cycle Er=Emapp-Emdep

Then, this device will be a proven OU but only if:
Ec(a)>=Ec(b), Ec(a) is greater or equal to Ec(b) and the Er is a free ride.
Or
Ec(b)-Ec(a)<=Er, the added energy required to built the field in scenario b is less that the Er

The above two do not consider the flybacks comparisons of EBcm(b) and EBc(a).
If EBcm(b)-EBc(a)>0 then that is added energy output, and need to be added to Er or subtracted from the net required to build the coil field
Vice versa if EBcm( b)-EBc(a)<0

To keep it simpler, all the above do not immediately account for time variant properties, magnetic, viscosities of both the magnet and toroid core, magnetostriction effects, magnetocaloric effects etc.
However; integrating the curves in dynamic situations where speeds are exceedingly high maybe greatly influenced by the above, as intrinsic time constants become significant in comparison to integrated time intervals in the energy calculations. In addition, the interrelationship of permability and inductance as a function of Hc rise alone, is different when Hc rises while there is the already influence of the Hm etc.

The bottom line is that the only real final way to show OU is to Either physically demonstrate TotalEout>TotalEin, (self runner, continuously re-charging a cap or a battery without an external charger etc)
Or
Accurately account for all friction, heat losses and other magnetic losses etc and accurately measure and account all the energies during the magnetic/electromagnetic transactions, and then show after accounting for all the measurement errors and assumptions, that Eout>Ein

[Omega_0]

Good post.

I think there is an error here "Ec(b)-Ec(a)<=Er", should be Ec(b)<=Er.

When Ec(b)<Ec(a), the left hand side will be negative and will not make sense. Also, Ec(a) has no role to play in scenario (b).

The only necessary condition to demonstrate OU is to show that net energy required to build the shielding field is less than net gain in rotor KE, per cycle.

One does not even need to build an orbo or a self runner, one just need to setup an experiment with a toroidal coil and a magnet and some careful measurements. Building a self runner is an engineering problem and will take some time to develop fully, however, the above condition is easy to show for someone with a good lab.

I hope someone with a good lab is reading this, because proving this result will boost the confidence a million times and we will see more sincere replications all over the world. Some of which will become self runners eventually.