I have started this bench as a follow on to my "Generator using a superconductor" as what I have discovered goes well beyond that possibility. This has all come from my work on the "Marinov Generator" discussed on my bench of that name. Others have considered Marinov's claims for his motor that he bizarrely named the "Siberian Coliu". This led them to a derivation based on the so-called "convective" term borrowed from fluid dynamics. Unlike the conventional motional induction that creates a sideways force on a charge moving through a magnetic field, this "convective" vector identity yields a force along the charge's movement direction thus creating either a drag force or an accelerating force. My work on the Marinov motor and that of others have shown that this longitudinal force is a reality. What is now becoming clear is something that has been overlooked by the theoreticians, the fact that this "convective" term can induce voltage into a closed loop. Maybe it has been overlooked because it needs either a weird-shaped closed loop (like hairpin) for induction within known magnetic vector potential fields or a weird-shaped field for induction within a common closed loop (like circular). Whatever the reason Science is missing out on some hugely significant possibilities.
Superconductors are a way of life today, in MRI scanners everywhere. There is a cost involved in maintaining these as they use liquid nitrogen temperatures, and the liquid needs regular topping up. I could envisage a world where such FE generators would take the place of the distribution transformers that exist in urban areas, where consumers paid for their electricity to cover the cost of the installation and the running costs. This is not free energy, but significantly cheaper than what we pay now and certainly better for the planet.
Now things move beyond that possibility. The source of the "free" energy in my superconducting hairpin coil is in my opinion the orbital motion of electrons responsible for the magnetization in the Fe ring cores where they can give up energy if they endure a decelerating force. I argued that the electrons in my hairpin conductor suffer a deceleration as they travel round the sharp hairpin bend and that deceleration creates a form of E field within which the Fe orbital electrons do see a deceleration. What I have now discovered is the deceleration of the orbital electrons can come from the "convective" term. The current in the hairpin loop sets up a "weird" A field pattern such that a circular closed loop obtains a non-zero force around that loop. That opens up a new possibility, we do not need a superconductor to get to high value "free" current, we can get an effective "free" current at the surface of magnetized material. Thus the new generator possibility does not wind the hairpin coil around a plastic tube, it is wound around a ferromagnetic tube.
In the first image below I show the A field around the top of such a magnetized ferromagnetic tube. I also draw the effective surface current around the ring core there. In the next image I show the E field around that closed square current loop determined by the method shown in my "Generator using a superconductor" bench. This clearly shows that there is a non-zero voltage around that loop.
More to come on this later.
Smudge
Are you sure that the mesh chosen for the simulation is fine enough to model the field near the conductor angles?
A simulation software can't show anything wrong with the physics equations on which it is supposed to be based.
Quote from: F6FLT on 2022.11.11, 14:19:51
Are you sure that the mesh chosen for the simulation is fine enough to model the field near the conductor angles?
A simulation software can't show anything wrong with the physics equations on which it is supposed to be based.
Those arrows are on a 1mm mesh. The simulation mesh is 0.2mm. The physical equations for the E field are not in the finite element simulation, they are done outside on a spreadsheet. In my opinion the restrictions due to the mesh size are responsible for the noisiness of that plot of E field around the loop. An improved mesh size will not nullify the result that the integration around the loop gives a non-zero voltage, and that is the important feature.
Smudge
I think the importance of this finding is that within a non-uniform A field not only can there be a transverse force on a moving charge but also a longitudinal force. The magnitude of the force in both cases is proportional to the charge velocity (relative to the source of the A field) and to the nature of the non-uniformity. The Curl function defines one form of non-uniformity and then we get the well-known transverse force for movement through a magnetic field. But we now see there are other non-uniformities that produce a longitudinal force. This latter force is not taught in general science but it might explain why some experiments (like magnet motors) have yielded anomalous results. In the complex field pattern of arrays of rotor and stator magnets where the magnets are conductive eddy current induced into the magnets are an accepted feature. What is not accepted is that these eddy current might obtain additional accelerating forces that provide "free" addition to the magnet's power that ultimately results in free running motors.
Smudge
A while back I did experiment with the Marinov motor and generator. Below you see a diagram of this. I basically rotated a copper ring around a magnet that had its poles split/flipped (basically two half circular magnets stuck to each other) which according to the A field around it should generate a voltage if you followed the charge around in a circle. At first it did generate a voltage and I thought this was a confirmation of the idea. However, as I made the ring thinner and thinner the induced voltage also started to disappear. It was then apparent to me that the measured voltage was due to classic eddy currents and not the longitudinal A-field term.
Quote from: broli on 2022.11.11, 18:15:06
A while back I did experiment with the Marinov motor and generator. Below you see a diagram of this. I basically rotated a copper ring around a magnet that had its poles split/flipped (basically two half circular magnets stuck to each other) which according to the A field around it should generate a voltage if you followed the charge around in a circle. At first it did generate a voltage and I thought this was a confirmation of the idea. However, as I made the ring thinner and thinner the induced voltage also started to disappear. It was then apparent to me that the measured voltage was due to classic eddy currents and not the longitudinal A-field term.
I considered the classical induction as the cause and I got the slip ring turned down to a smaller OD, thus thinned down. My voltage didn't reduce like yours did. My slip ring was large (4 inches dia) and my two magnets were 1 inch placed adjacent to the brushes.
Smudge
Following on from the unusual coupling shown in the first message I think it may be possible to create an OU transformer, if not OU but certainly one with unusual properties. The basic set up as shown in the image below uses a toroidal winding on a thin-walled ferromagnetic tube plus two classical toroidal coils on ferromagnetic ring cores. The two classical ring core windings are connected in series so that the flux around one core is in the opposite direction to the other. This then is one transformer winding, say the primary. Our present theories tell us that there is virtually zero coupling across the transformer. The energized primary coils will drive displacement current through the air but the net amount passing through the elongated core is zero (see this image https://www.overunityresearch.com/index.php?action=dlattach;topic=4382.0;attach=46264 (https://www.overunityresearch.com/index.php?action=dlattach;topic=4382.0;attach=46264) for the A field lines that also denote the displacement current). However there should be this new induction in that secondary coil if it is carrying current and that should be observable. It could be interesting to feed primary and secondary with different frequencies and observe the cross talk in each coil.
Obtaining the thin-walled ferromagnetic core is a problem. This could be made by putting a single-layer close-wound coil of Fe wire onto a tube or cylinder former, binding the turns together somehow (sticky tape over the winding?) then removing the former. Or use a thin walled former and leave it there. The scheme shown allows the ends of the hairpin coil to just fit inside the hole in the ring cores, but this is not necessary. The ends could abut ring core faces.
Anyone up to try this out?
Smudge
Quote from: Smudge on 2022.11.12, 16:29:04
Obtaining the thin-walled ferromagnetic core is a problem. This could be made by putting a single-layer close-wound coil of Fe wire onto a tube or cylinder former, binding the turns together somehow (sticky tape over the winding?) then removing the former. Or use a thin walled former and leave it there.
Nickel battery strip? Good characteristics overall, fairly standardized, and easy for anyone to find:
https://www.amazon.com/Nickel-Strip-0-15x8mm-Battery-Welding/dp/B08RHP8NDD/ref=sr_1_5
@H,
Yes, you could cover a thin-walled former in that tape, winding it in spiral fashion. Thanks for that info.
Smudge
I could actually throw the Marinov generator idea into my weber force simulation software. All the weber forces are completely equivalent to any other force/induction law.
Quote from: Hakasays on 2022.11.12, 17:41:18
Nickel battery strip? Good characteristics overall, fairly standardized, and easy for anyone to find:
https://www.amazon.com/Nickel-Strip-0-15x8mm-Battery-Welding/dp/B08RHP8NDD/ref=sr_1_5
Please note that this is not pure nickel but nickel-plated steel. It remains to be seen whether it is ferromagnetic steel, not all of them are, as the nickel plating is likely not thick enough for significant magnetic effects.
Quote from: broli on 2022.11.13, 21:31:03
I could actually throw the Marinov generator idea into my weber force simulation software. All the weber forces are completely equivalent to any other force/induction law.
I would encourage you to do that and let us know the results.
Quote from: F6FLT on 2022.11.14, 11:07:25
Please note that this is not pure nickel but nickel-plated steel. It remains to be seen whether it is ferromagnetic steel, not all of them are, as the nickel plating is likely not thick enough for significant magnetic effects.
Did not notice that, those cheapskates C.C
Smudge et al; make sure to search for 99%+ pure nickel strip.
Quote from: Hakasays on 2022.11.14, 13:40:26
Did not notice that, those cheapskates C.C
Smudge et al; make sure to search for 99%+ pure nickel strip.
Still don't get the idea fully but you're better off ordering an amorphous Nanocrystalline core. Some of these come in loose casing which you can easily pry open (opposed to epoxied casing) and the core is just wound inside of it. It's very thin and brittle.
Well, I threw a Marinov generator design in my Weber Electrodynamics simulator and actually got a surprising result. In the screenshot below you can see it. The red lines represent the forces vectors on the charged particles and the green ones are the velocity vectors. I only kept the longitudinal component of the force vector for easier analysis.
The surprising fact is that the voltage gradient is divided in quadrants opposed to into halves. In fact, if you would measure across the diameter you would end up with a zero voltage as it would cancel. I'm now very curious as to how true to experiment this would be. I might bring my experimental setup back to life to see if there's a different between measuring between 90° and 180° degrees as I've not explored this.
PS: The simulation show some force lines being longer on one side than the other, this is actually an artifact as they are flickering in realtime, it's a bug somewhere in the software I have yet to find.
EDIT: found the bug, sim looks symmetric now.
Reducing the marinov generator to its bare bones eliminates any specially shaped magnets. Would you agree that the below "mechanical" setup would be a good way to benchmark the vector potential induction viability? Basically, you rotate a smaller conductive loop off center to a magnet. The A-field would go from maximum to 0 from a moving charge's point of view and thus induce an E field due to the changing vector potential. This is also a much simpler problem to solve analytically or numerically (as you're dealing only with circles) in order to compare it to experimental data.
I also added the Weber Electrodynamics sim results as reference.
@Broli,
What you are showing is a significant feature that has been overlooked in current science teachings. Your off-center circular loop has a magnetic flux passing through it. Current teaching says that the only way you can get voltage induction in that loop is if the flux is changing with time. Your flux is from a permanent magnet so is static, and if you integrate the tangential force vectors around that loop you do get zero induction. But integrated over only half the loop from one side to the other, the voltage induction is not zero. The existence of a longitudinal force vector for charge movement within a magnetic field is not recognized in current teachings..
If you now create a loop that is a semicircle plus diameter you have something that can't be a rotating slipring, but it could be a moving belt of conductive material passing over tiny rollers at each end of the diameter (and of course a chain of rollers defining the curved path). When you integrate the force vectors around that loop you get a non-zero answer and depending on the conductivity of the belt it may be possible to get current to build up in that driven loop. The belt would then obtain the classical transverse forces from the magnetic field that would require more rollers to maintain the belt's position. A nightmare to construct but a sure-fire way to demonstrate this new discovery.
Smudge
Quote from: Smudge on 2022.11.16, 11:58:59
@Broli,
What you are showing is a significant feature that has been overlooked in current science teachings. Your off-center circular loop has a magnetic flux passing through it. Current teaching says that the only way you can get voltage induction in that loop is if the flux is changing with time. Your flux is from a permanent magnet so is static, and if you integrate the tangential force vectors around that loop you do get zero induction. But integrated over only half the loop from one side to the other, the voltage induction is not zero. The existence of a longitudinal force vector for charge movement within a magnetic field is not recognized in current teachings..
If you now create a loop that is a semicircle plus diameter you have something that can't be a rotating slipring, but it could be a moving belt of conductive material passing over tiny rollers at each end of the diameter (and of course a chain of rollers defining the curved path). When you integrate the force vectors around that loop you get a non-zero answer and depending on the conductivity of the belt it may be possible to get current to build up in that driven loop. The belt would then obtain the classical transverse forces from the magnetic field that would require more rollers to maintain the belt's position. A nightmare to construct but a sure-fire way to demonstrate this new discovery.
Smudge
I have also been aware of this for a while and what lead me to write the Weber Electrodynamics simulation software. However, I'm still on the fence on this due to my passed failed experiments which were admittedly very crude. Anyone willing to help with that is appreciated. This can be done with mathematica for instance where the A field is integrated across half of the circular loop to measure its change and subsequent voltage.
I have a 100mm disc magnet laying around which is plenty big to confirm this effect but I would like to have a theoretical prediction as well so I know what to expect. If prediction says 50 mV and measured value is in couple mV's range then the hypothesis is disproven.
Edit: I'm thinking out loud here, it shouldn't be that hard to get a ballpark estimate all we need to know is the maximum value of the A field and minimum (which should be as good as zero). And using the RPM we can get the derivate which gives us a good estimate of the voltage. So assuming:
Magnet = 100mm N40 Neodymium
A= 0.00515759 Wb/m (According to FEMM, 2mm near the edge of the magnet)
RPM=1000
T = 0.03s (for half a rotation)
r_loop = 25mm
dA/dt= 0,00515759 Wb/m/0.03s = 172mV
That seems like a ginormous estimate which would be hard to miss. Anyone sees anything wrong with this calculation?
Edit2:
I forgot that this is the E field so the circumference of half the rotating loop should be multiplied by it to get the actual voltage:
E = dA/dt = 0,00515759Wb/m / 0.03s = 172mV/m
V = E*L = 172mV/m * PI * 0.025m = 13.5mV
Which is still significant for only 1000 RPM.
Quote from: Smudge on 2022.11.12, 16:29:04
...The energized primary coils will drive displacement current through the air
...
There is no displacement current around a coil. A closed EMF cannot move dipoles around a loop, it only orients them around their respective centers, with no net displacement of charges in either direction.
If the loop will be made in the form of a Mobius strip?
Just a note
Smudges moderated board at Stefan's
https://overunity.com/19293/smudges-new-discovery-of-voltage-induction/msg571769/#new
Mirrored topic !
Respectfully
Chet
Ps
of course please remove this post if inappropriate or ?
PPS
Just reminder Smudge and Partzman are moderator there
I was just put in as a temp for a specific and separate topic ( helping member TinselKoala)
Quote from: chief kolbacict on 2022.11.16, 17:52:52
If the loop will be made in the form of a Mobius strip?
Explain the reasoning.
Quote from: F6FLT on 2022.11.16, 16:12:36
There is no displacement current around a coil. A closed EMF cannot move dipoles around a loop, it only orients them around their respective centers, with no net displacement of charges in either direction.
Displacement current is not charge movement and it is not dipole movement. It is a time-changing E field. Are you telling me that the closed E field around a transformer core can't create displacement current? If so please explain why.
Smudge
I quickly setup an experiment using the eccentric loop idea. I ran the motor at its rated 24V DC which as labeled makes it spin at 3500 RPM. I also used slightly thicker coper loop opposed to the copper tape I used previously. This proved to be much better as the measured voltages were much more stable.
The result I got was in the range of 0.5mV, essentially a null result. Regardless of how I placed the brushes even next to each other the same value was seen which tells me the this very low voltage was due to eddy currents rather than the change of the vector potential. From my crude calculation I should have seen a value in the 30mv range.
Sadly, this confirms to me that the vector potential does not offer an induction effect when charge is moving through a "static" vector potential changing only due to the movement of the charge.
Quote from: F6FLT on 2022.11.16, 18:45:56
Explain the reasoning.
I don't know,can't explain. Just it came to my mind.
This thing is said to have unusual geometric properties.
Even magical properties.So I thought this might help.
Perhaps I opened my brain, but it turned out to be a trash. :-\
Quote from: chief kolbacict on 2022.11.17, 06:12:25
I don't know,can't explain. Just it came to my mind.
This thing is said to have unusual geometric properties.
...
These special properties only concern the surface topology, not the conduction of electrons, for which an ordinary or Möbius strip is the same.
Quote from: broli on 2022.11.16, 20:35:57
...
Sadly, this confirms to me that the vector potential does not offer an induction effect when charge is moving through a "static" vector potential changing only due to the movement of the charge.
On a closed circuit and static A, one cannot expect an EMF because a force deriving from a potential is conservative.
But I think that a variation of A in time is well accompanied by a gradient in space seen from a moving charge and vice versa.
To do the integral over a circuit in a A gradient, with the variation of A in time seen from the charge, as if it were a real ∂A/∂t seen from an observer at rest with respect to the circuit, is to mix values seen from different reference frames, it can only lead to false results.
If the integral over a circuit is done in the reference frame of the charge moving in a A gradient, then the charge sees ∂A/∂t BUT also a moving circuit, so the calculation cannot be done as in the case of the circuit at rest with respect to the observer.
Quote from: broli on 2022.11.16, 20:35:57
I quickly setup an experiment using the eccentric loop idea. I ran the motor at its rated 24V DC which as labeled makes it spin at 3500 RPM. I also used slightly thicker coper loop opposed to the copper tape I used previously. This proved to be much better as the measured voltages were much more stable.
The result I got was in the range of 0.5mV, essentially a null result. Regardless of how I placed the brushes even next to each other the same value was seen which tells me the this very low voltage was due to eddy currents rather than the change of the vector potential. From my crude calculation I should have seen a value in the 30mv range.
Sadly, this confirms to me that the vector potential does not offer an induction effect when charge is moving through a "static" vector potential changing only due to the movement of the charge.
I have just created the A fields near the surface of your PM and I agree with the magnitude you got near the outer edge. In the image of your experimental set up you don't see any brushes. Could you please explain what you used for brushes and where did you place them? Thanks.
Smudge
Quote from: Smudge on 2022.11.18, 14:59:38
I have just created the A fields near the surface of your PM and I agree with the magnitude you got near the outer edge. In the image of your experimental set up you don't see any brushes. Could you please explain what you used for brushes and where did you place them? Thanks.
Smudge
I used braided wires and my own hands to hold them. I could easily slide the brushes around like that and going from 180° till the brushes almost next to each other gave the same result. What's surprising is that Weber's electrodynamics also predicts such longitudinal field yet the experimental setup shows none.
@Broli
OK, I think I have solved the dilemma. F6 and I have had earlier discussions about the vector potential on my Marinov Generator thread.
We agreed that it should be possible to derive classical induction for homopolar motion within a magnetic field directly from the A field that creates B (B=curl A). When I initially did this I got only 1/2 the correct value. To get the correct value you have to take account not only of the changing magnitude of the A field along the movement trajectory but also its changing angle relative to the trajectory. This is somewhat akin to the Coriolis force for movement within a rotating frame.
For Broli's experiment, treating qA as some form of hidden momentum I have a method for deducing force by taking the change in A over a small distance dx of charge movement as dA, taking the angle between A and dx as Θ then using Ex=-v*dA*cos(Θ)/dx. If Θ changes value over that movement I take the average of the two values in the cos function. I got a voltage induction that agreed with Broli's deduction from Weber dynamics. That is OK but it is important to note that a changing angle represents a rotating vector, and a rotating momentum vector creates a force even if the magnitude stays constant, and that force direction is at right angles to A. So I have now deduced an additional force by taking the change in Θ over the movement dx as dΘ, taking the average magnitude of A over dx as |A|, then using Ex=-v*|A|*dΘ*sin(Θ)/dx as the addition. This cancels out the induced voltage. Note this is for the field close to the magnet where B=curl A. So within that particular A field pattern there is no longitudinal induction along the movement path, only at right angles to it. It remains to be seen whether the same thing applies to a non-curl A field.
Smudge
@Broli,
I have rechecked my method of taking the actual A field at different points around the circle (not the component wrt to velocity direction). Then taking the change in magnitude over the time increment between two points as dA/dt, and that is an E vector pointing in the mean direction of the two A vectors. The component of that E along the velocity direction is then one component of induction. I also look at the change in angle of the A vector over the time increment between points (omega) and that gives me another E vector given by omega*A. This E vector is at right angles to the mean direction of the two A vectors. That gives me another component to be added to the first one. And the two components have opposite signs and cancel each other. So not surprising you found zero induction.
I then took another trajectory for a slip-ring mounted alongside the magnet as depicted in my modification to your image below. I am doing this in a 2D program so my magnet is infinitely long in the z direction, thus there is zero magnetic field external to the magnet. But there is a non-curl A field there and now my two E field components don't cancel. The second image shows that external A field and my integration line. This predicts a voltage across the diameter and I got such a voltage in my Marinov Generator work. I would ask you to modify your experiment to do this test using your large magnet and see what you get.
Smudge
Quote from: Smudge on 2022.11.21, 16:21:36
@Broli,
I have rechecked my method of taking the actual A field at different points around the circle (not the component wrt to velocity direction). Then taking the change in magnitude over the time increment between two points as dA/dt, and that is an E vector pointing in the mean direction of the two A vectors. The component of that E along the velocity direction is then one component of induction. I also look at the change in angle of the A vector over the time increment between points (omega) and that gives me another E vector given by omega*A. This E vector is at right angles to the mean direction of the two A vectors. That gives me another component to be added to the first one. And the two components have opposite signs and cancel each other. So not surprising you found zero induction.
I then took another trajectory for a slip-ring mounted alongside the magnet as depicted in my modification to your image below. I am doing this in a 2D program so my magnet is infinitely long in the z direction, thus there is zero magnetic field external to the magnet. But there is a non-curl A field there and now my two E field components don't cancel. The second image shows that external A field and my integration line. This predicts a voltage across the diameter and I got such a voltage in my Marinov Generator work. I would ask you to modify your experiment to do this test using your large magnet and see what you get.
Smudge
Do you have a ball park figure on the estimated voltage of this as I have calculated before?
Quote from: broli on 2022.11.21, 22:27:21
Do you have a ball park figure on the estimated voltage of this as I have calculated before?
Yes, my figure is 6mV. If you could arrange some crude fixed brushes then you can hand hold the magnet and move it around. You could try smaller magnets. Of course the acid test is with a magnetized ring core as shown in the image below (shown without the toroidal winding). You must have some wound cores lying around in your workshop ;) . Then there is no magnetic field, only the A field at the slip ring. I have been waiting years for that test to be done.
Smudge
Wanted to get back to this topic as I've recently got around to performing the experiment where the magnet was positioned outside the rotating ring.
Now sadly this was inconclusive. The readings jumped around up to 4mV-5mV and sometimes flipped sign too and this was around 1000 RPM indicating a not so ideal brush assembly. I'm not the best builder around.
However I'm still skeptical about the vector potential inducing any voltage here.