Curious device

[swansont]

48 minutes ago, Prajna said:

That provokes images of some structure bowing down to hoist a tiny metal sheet up into the air. Sure, to some barely detectable extent a solid structure will distort due to the forces involved in the thought experiment I described but really, is it really fair to say in this situation that the structure is doing all the work and the magnet is doing none?

A large structure does not have to bend very much to account for this energy. As you say, it is a tiny metal sheet. If you lift a 1 g object 1 meter, a 1 kg structure only has to shift 1 mm

48 minutes ago, Prajna said:

That is a very strange way to describe the situation in my view.

Physics is quite successful, but relies on rigor and not hand-waving.

 

[Prajna]

1 hour ago, exchemist said:

It's very simple. W=Fd is all you need to keep in mind. If there is no change in d, the distance in the direction of F, no work is done. Move them apart and you do work on them, causing energy to go into the magnetic field (a form of potential energy). Allow them to move closer and the reverse happens: they do work on whatever is holding them apart, with the necessary energy coming from the magnetic field.

As for the magnet sliding, that is a distraction. It does not in fact alter the strength of the magnetic force holding it to the metal sheet. Any motion perpendicular to the force means d does not change, so no work is done. 

(What it may do, though, is reduce the frictional force parallel to the sheet which resists the magnet being slid. Typically the frictional drag between two surface that are sliding past one another is less that the limiting friction just before the force is overcome and sliding commences. This is to do with asperities on the surface interlocking, which does not happen to the same degree when they are in motion. But this is a tribological digression.)

 

Regarding the sliding of a magnet off a sheet of steel, I guess that what is really happening is that you are not feeling the same force you feel when you try to lift it directly, you're not working directly against the magnetic field. Once you get to the edge of the sheet the area of the magnetic contact reduces, and with it the force holding the magnet to the sheet until finally the magnet is clear of the sheet.

Triboelectric generation is looking more and more interesting these days. Won't be long before the free energy cranks are baffling themselves with that. Hmm, come to think of it, I've done nothing about adding generation to my device, ... reciprocation, triboelectrics, ... hmm ...

[Mordred]

4 hours ago, Prajna said:

Ah, ok, you were talking with regard to the device and I was (at that time) talking with regard to the thought experiment. Cross purposes, sorry for the misunderstanding. Yes, I'm sure that at some point I will need to consider distortion in the fingers/tabs but engineers often just over-engineer things, add s fudge factor, rather than fuss about such details. This particular design is just a proof of concept and the objective was to make it easy for people to grasp the principal of operation. The rotor is designed so that the number and geometry of the tabs can be changed easily in order to test different configurations.

 

No problem, it all depends on how detailed or how far you choose to pursue the concept. One thing to consider however is that in order to look at stress and stain aspects. You require the force/work terms as well. For example far too often I've seen perpetual energy  articles discussing some popularized perpetual device use nothing more than first order equations. 

However when the same setup gets examined using second order relations by others that the energy loss is found exceed the output power.

As Swansont mentioned in physics one cannot arbitrarily choose to ignore this interaction (in this case different forces) or that but should take everything in consideration. Stress tensors are particularly useful in that as all forces are applied with a means of keeping track via the tensors regardless of angles.

Not saying perpetual energy is involved here however the above is also useful for efficiency calculations.

Edited April 17 by Mordred

[Mordred]

On 4/16/2024 at 8:34 AM, swansont said:

Magnetic forces don’t do work. 

 

I would like to touch a bit on this using Maxwell equations but also Lorentz force law divergence and curl of the Electrostatic field 

Gauss Law

\[\nabla \cdot=\frac{1}{\epsilon_0}\rho\]

\[\nabla \times E=0\]

magnetostatic Field

Amperes law

\[\nabla\cdot B=0\]

\[\nabla \times B=\mu_0 J\]

Lorenz force law with Maxwell for the presence of both the E and B field (Maxwell equations fundamentally is another way of stating Biot-Savart Law (with superposition) just a side note.

\[F=Q(E+v\times B)\].

so the electric field diverges away from a positive charge, (Gauss law) the magnetic field curls around a current (Amperes law). Electric fields originate form a positive charge and terminate on a negative charge. Magnetic lines do not begin or end anywhere and form closed loops as they have zero divergence. (though divergence can be forced). There is no point source for B ( not unless they ever discover magnetic monopoles lol). Now something interesting to note the magnetic field specifies an electric current.( A permanent magnet induces an electric current). So with the 90 degree phase shift between E and B using the right hand rule for Lorentz force law. The following statement applies.

The magnetic field does no work....

so take for example a magnetic crane the work isn't performed by the magnetic field but rather the electric field as well as the cranes mechanical energy.

This is something that isn't well known among laypersons unless they studied introductory electrodynamics and the Maxwell equations. Hence why I decided to mention it here as its related.

The above is better detailed in Griffiths "Introductory to Electrodynamics". I've found his simplified approach useful as a reference in many of his books.

 

 

 

 

Edited April 18 by Mordred

[Prajna]

5 hours ago, Mordred said:

 

I would like to touch a bit on this using Maxwell equations but also Lorentz force law divergence and curl of the Electrostatic field 

Gauss Law

 

∇⋅=1ϵ0ρ

 

 

∇×E=0

 

magnetostatic Field

Amperes law

 

∇⋅B=0

 

 

∇×B=μ0J

 

Lorenz force law with Maxwell for the presence of both the E and B field (Maxwell equations fundamentally is another way of stating Biot-Savart Law (with superposition) just a side note.

 

F=Q(E+v×B)

.

 

so the electric field diverges away from a positive charge, (Gauss law) the magnetic field curls around a current (Amperes law). Electric fields originate form a positive charge and terminate on a negative charge. Magnetic lines do not begin or end anywhere and form closed loops as they have zero divergence. (though divergence can be forced). There is no point source for B ( not unless they ever discover magnetic monopoles lol). Now something interesting to note the magnetic field specifies an electric current.( A permanent magnet induces an electric current). So with the 90 degree phase shift between E and B using the right hand rule for Lorentz force law. The following statement applies.

The magnetic field does no work....

so take for example a magnetic crane the work isn't performed by the magnetic field but rather the electric field as well as the cranes mechanical energy.

This is something that isn't well known among laypersons unless they studied introductory electrodynamics and the Maxwell equations. Hence why I decided to mention it here as its related.

The above is better detailed in Griffiths "Introductory to Electrodynamics". I've found his simplified approach useful as a reference in many of his books.

 

 

 

 

Thanks, @Mordred, I recently read a headline suggesting that someone has found a magnetic monopole but can't remember where I saw it. I seem to recall reading that Maxwell's equations are not actually Maxwell's, that Maxwell expressed his equations as quaternions but didn't publish them before he died. It was Heaviside who published Maxwell's equations, having converted them all to vectors, so really when we talk about Maxwell's equations we are talking about Heaviside's equations and neglecting Maxwell. But then it might have been some renegade astronomer who wrote that, I can't recall exactly.

[exchemist]

12 hours ago, Prajna said:

Ok, the situation in the device though is that the magnets are not moved closer nor separated by the operator, rather the magnets move under the effect of their changing state of repulsion or attraction governed by whether there is a finger or no finger between them. So, in my view, energy is being exchanged between the magnetic fields on one side of the device and the fields on the other. Is this a reasonable description? The operator is not moving the magnets, not even by poking them with a stick, he is merely turning the rotor that determines which side of the device has a finger in the gap and which doesn't. Maybe I should rename my 'fingers' to 'tabs' to avoid the idea that there is any poking of the magnets going on.

I don't see where there is any direct link between the work or energy or effort required to turn the rotor and the work or energy or effort produced by the switch state of the magnets, unless there is some reason to suggest that the eddy current drag is directly related to the output. The drag will certainly depend on the strength of the magnetic field on the side of the device that the finger/tab is passing through and the speed of the finger/tab through that field and, I guess, on the area of the tab that's in the field, perhaps even other things I have yet to consider.

OK, now we get to it.

In your proposed machine, the magnets first repel one another, doing work and lowering the stored energy in their respective magnetic fields. You believe that when you insert the steel finger between them, they will then be attracted towards it, doing more work and further lowering the stored energy in their fields. And then, when you move the finger out of the way, they move apart again due to repulsion, extracting yet more energy from their magnetic fields.

This obviously cannot be the case.  So there is something wrong with your assumption. Either you will find the magnets are not attracted together when the steel finger is interposed, or you will find the finger resists being inserted or removed, such that the operator has to do work against the field, thus supplying the required energy. At the moment (being a chemist rather than a physicist) I am not sure which of the two it is, but logically it must be one or the other, it seems to me. My suspicion is that the magnets will not be attracted to the finger. If you consider the path of the flux lines when the finger is in between, they have to turn sharply horizontal within the finger and pass outward to each side. This I think means the dipoles within the finger will not be able to align themselves with either field in the way that you (implicitly) suppose, as they will be perpendicular to the fields,  and so no attractive force will result.

If that's right, you would be able to twiddle the finger wheel as fast you like and bugger-all will happen!  But perhaps you should build it to confirm exactly how it fails to work.    

[Prajna]

1 hour ago, exchemist said:

OK, now we get to it.

In your proposed machine, the magnets first repel one another, doing work and lowering the stored energy in their respective magnetic fields. You believe that when you insert the steel finger between them, they will then be attracted towards it, doing more work and further lowering the stored energy in their fields. And then, when you move the finger out of the way, they move apart again due to repulsion, extracting yet more energy from their magnetic fields.

This obviously cannot be the case.  So there is something wrong with your assumption. Either you will find the magnets are not attracted together when the steel finger is interposed, or you will find the finger resists being inserted or removed, such that the operator has to do work against the field, thus supplying the required energy. At the moment (being a chemist rather than a physicist) I am not sure which of the two it is, but logically it must be one or the other, it seems to me. My suspicion is that the magnets will not be attracted to the finger. If you consider the path of the flux lines when the finger is in between, they have to turn sharply horizontal within the finger and pass outward to each side. This I think means the dipoles within the finger will not be able to align themselves with either field in the way that you (implicitly) suppose, as they will be perpendicular to the fields,  and so no attractive force will result.

If that's right, you would be able to twiddle the finger wheel as fast you like and bugger-all will happen!  But perhaps you should build it to confirm exactly how it fails to work.    

This is exactly the conversation I hoped to have here, @exchemist. Yes, my preliminary experiments showed that with magnets in repulsion a steel sheet, in this case a steel rule, inserted into the gap caused the magnets to be attracted to the rule. My guess at what is happening there is (if you'll indulge my less-than-physically-exact language) the magnets in repulsion, because their competing fields offer a very high reluctance to the other's flux path, try to complete their circuit by adopting (and attracting) the steel rule, a much lower reluctance path, into their circuit. When the rule is removed the magnets again face an unacceptable high reluctance to their circuit and therefore attempt to mitigate it by moving apart.

[exchemist]

1 hour ago, Prajna said:

This is exactly the conversation I hoped to have here, @exchemist. Yes, my preliminary experiments showed that with magnets in repulsion a steel sheet, in this case a steel rule, inserted into the gap caused the magnets to be attracted to the rule. My guess at what is happening there is (if you'll indulge my less-than-physically-exact language) the magnets in repulsion, because their competing fields offer a very high reluctance to the other's flux path, try to complete their circuit by adopting (and attracting) the steel rule, a much lower reluctance path, into their circuit. When the rule is removed the magnets again face an unacceptable high reluctance to their circuit and therefore attempt to mitigate it by moving apart.

OK, in that case, what I think you will find is it takes significant effort to pull the finger out of the gap, as the force of attraction is stronger once the magnets have moved inward, than the force that pulls it into the gap when you insert it. So you do net work on the system that way and this provides the energy that restores the stored energy in the fields to the status quo ante. 

[sethoflagos]

1 hour ago, exchemist said:

... So there is something wrong with your assumption. Either you will find the magnets are not attracted together when the steel finger is interposed, or you will find the finger resists being inserted or removed, such that the operator has to do work against the field, thus supplying the required energy...

Perhaps one way of looking at this contraption is to compare it with a Faraday disk (aka homopolar generator).

In the latter, both motion and induced current are in the plane of the disk with the magnetic field perpendicular. The OP is rotating this so that motion and magnetic field lines are in the disk plane therefore forcing induced current into the perpendicular. However, different portions of the disk will see different current polarities depending on whether they are moving towards or away from the magnetic poles. In particular, the portion of the disk passing directly between the poles will see a sharp switch in polarity and consequent current flow component appearing in the disk plane. This will in turn deflect the magnetic field lines somewhat out of the disk plane as if attracted by a temporary opposite pole.

I don't know whether it's a good picture, but in my mind's eye, I'm seeing this induced temporary pole falling into a potential well only to climb back out as it departs with no nett overall energy change in and of itself. However these circulating currents are a different matter as they will add a time lag to the ideal case making ascent harder than descent, acting as a brake in exchange for simply heating up the disk.

 

[Prajna]

1 hour ago, exchemist said:

OK, in that case, what I think you will find is it takes significant effort to pull the finger out of the gap, as the force of attraction is stronger once the magnets have moved inward, than the force that pulls it into the gap when you insert it. So you do net work on the system that way and this provides the energy that restores the stored energy in the fields to the status quo ante. 

That sounds completely reasonable and is just what I was angling for a competent analysis of.

I do recall the eddy current demonstration of dropping a magnet down a copper tube, a very impressive demonstration of eddy current drag. Certainly it is my estimation that the fingers will get sucked into the gap and resist being pulled out again but I think (may well be mistaken) that these two forces will cancel each other. However the eddy current drag may be what correlates the input force to the output power. It may be that whatever reconfiguration to reduce the eddy drag will also have an equal impact on the power output, satisfying input >= output. Only my gut tells me that there might be an input < output while my head tells me that cannot be.

19 minutes ago, sethoflagos said:

Perhaps one way of looking at this contraption is to compare it with a Faraday disk (aka homopolar generator).

In the latter, both motion and induced current are in the plane of the disk with the magnetic field perpendicular. The OP is rotating this so that motion and magnetic field lines are in the disk plane therefore forcing induced current into the perpendicular. However, different portions of the disk will see different current polarities depending on whether they are moving towards or away from the magnetic poles. In particular, the portion of the disk passing directly between the poles will see a sharp switch in polarity and consequent current flow component appearing in the disk plane. This will in turn deflect the magnetic field lines somewhat out of the disk plane as if attracted by a temporary opposite pole.

I don't know whether it's a good picture, but in my mind's eye, I'm seeing this induced temporary pole falling into a potential well only to climb back out as it departs with no nett overall energy change in and of itself. However these circulating currents are a different matter as they will add a time lag to the ideal case making ascent harder than descent, acting as a brake in exchange for simply heating up the disk.

 

Thanks for that, @sethoflagos, that is a very useful comparison.

[exchemist]

57 minutes ago, sethoflagos said:

Perhaps one way of looking at this contraption is to compare it with a Faraday disk (aka homopolar generator).

In the latter, both motion and induced current are in the plane of the disk with the magnetic field perpendicular. The OP is rotating this so that motion and magnetic field lines are in the disk plane therefore forcing induced current into the perpendicular. However, different portions of the disk will see different current polarities depending on whether they are moving towards or away from the magnetic poles. In particular, the portion of the disk passing directly between the poles will see a sharp switch in polarity and consequent current flow component appearing in the disk plane. This will in turn deflect the magnetic field lines somewhat out of the disk plane as if attracted by a temporary opposite pole.

I don't know whether it's a good picture, but in my mind's eye, I'm seeing this induced temporary pole falling into a potential well only to climb back out as it departs with no nett overall energy change in and of itself. However these circulating currents are a different matter as they will add a time lag to the ideal case making ascent harder than descent, acting as a brake in exchange for simply heating up the disk.

 

One important difference, though, is that in the OP's machine the poles of the two magnets are opposed so that they repel. The region in which the fingers on the input disc move is in principle an area in which the field lines will be squashed outwards in the plane of the fingers of the disc. 

[Prajna]

2 minutes ago, exchemist said:

One important difference, though, is that in the OP's machine the poles of the two magnets are opposed so that they repel. The region in which the fingers on the input disc move is in principle an area in which the field lines will be squashed outwards in the plane of the fingers of the disc. 

It's not so unusual though for magnetic fields to be sharply bent, a keeper on a horseshoe magnet does this when placed across the poles, providing the shortest possible low reluctance path for the field and containing it, effectively neutralising the magnet.

By the way, I've just taken a look at the Ferromagnetism page on Wikipedia and am finding it very helpful in understanding magnetism. Soon I may be able to drop my naive model and speak intelligently about it.

[sethoflagos]

4 minutes ago, exchemist said:

One important difference, though, is that in the OP's machine the poles of the two magnets are opposed so that they repel. The region in which the fingers on the input disc move is in principle an area in which the field lines will be squashed outwards in the plane of the fingers of the disc. 

Different words, same thing. It's the repulsion of like poles that causes the field rotation I took as a given for sake of brevity.

[exchemist]

6 minutes ago, sethoflagos said:

Different words, same thing. It's the repulsion of like poles that causes the field rotation I took as a given for sake of brevity.

Oh I see. 

20 minutes ago, Prajna said:

It's not so unusual though for magnetic fields to be sharply bent, a keeper on a horseshoe magnet does this when placed across the poles, providing the shortest possible low reluctance path for the field and containing it, effectively neutralising the magnet.

By the way, I've just taken a look at the Ferromagnetism page on Wikipedia and am finding it very helpful in understanding magnetism. Soon I may be able to drop my naive model and speak intelligently about it.

Well that's a win, then! 

It's been an interesting discussion. 

[Prajna]

4 minutes ago, exchemist said:

Oh I see. 

Well that's a win, then! 

It's been an interesting discussion. 

Sorry, I don't understand. Unless you mean it's a win that I'm belatedly reading about magnetism and you don't need to be involved in analysing the device.

By the way, I stumbled back on the article about monopoles, for anyone interested, it was in Popular Mechanics: https://www.popularmechanics.com/science/a60079037/magnetic-monopole-hematite/

[sethoflagos]

1 hour ago, Prajna said:

Thanks for that, @sethoflagos, that is a very useful comparison.

In return, perhaps you could clarify something for me. What is the phase relationship between magnet pole separation and finger position?

If we define zero degrees for the disk when a finger is directly between the poles, and zero degrees for the poles as minimum pole separation, then what phase difference between the two should we consider for optimum performance?

And how is that optimal phase difference maintained?

[exchemist]

28 minutes ago, Prajna said:

Sorry, I don't understand. Unless you mean it's a win that I'm belatedly reading about magnetism and you don't need to be involved in analysing the device.

By the way, I stumbled back on the article about monopoles, for anyone interested, it was in Popular Mechanics: https://www.popularmechanics.com/science/a60079037/magnetic-monopole-hematite/

Both: we’ve solved the conundrum presented by your machine, I’ve revised some magnetism I haven’t looked at since school, and you’ve become motivated to learn more about it. And for me, another perpetual motion machine bites the dust, which I can add to my tally. 

[Prajna]

3 minutes ago, sethoflagos said:

In return, perhaps you could clarify something for me. What is the phase relationship between magnet pole separation and finger position?

If we define zero degrees for the disk when a finger is directly between the poles, and zero degrees for the poles as minimum pole separation, then what phase difference between the two should we consider for optimum performance?

And how is that optimal phase difference maintained?

In the current design the tabs or fingers are arranged every 20 degrees around the rotor. There are nine tabs and the rotor axle is co-planar with the centre line of the magnets, so when there is a tab central to the gap between the magnets on one side there is a space on the opposite side. The magnets are 10mm x 2mm neodymium (N52?) and the tabs centres are at approximately a 35mm radius. The 'bulb' on the tabs that lies between the magnets is 5mm radius to match the area of the magnets. This may be more or less optimal as far as effectiveness in switching the flux and suffering eddy current drag, I don't know yet.

This is somewhat arbitrary and is just my first best guess of what might work.

4 minutes ago, exchemist said:

Both: we’ve solved the conundrum presented by your machine, I’ve revised some magnetism I haven’t looked at since school, and you’ve become motivated to learn more about it. And for me, another perpetual motion machine bites the dust, which I can add to my tally. 

I wish you guys would stop with the perpetual motion slur, I'm rather hoping for over unity!

[Mordred]

37 minutes ago, Prajna said:

Sorry, I don't understand. Unless you mean it's a win that I'm belatedly reading about magnetism and you don't need to be involved in analysing the device.

By the way, I stumbled back on the article about monopoles, for anyone interested, it was in Popular Mechanics: https://www.popularmechanics.com/science/a60079037/magnetic-monopole-hematite/

Monopoles is an interesting study for example it's potential would fall off at 1/r as opposed to 1/r^2 for dipolar, 1/r^3 for quadrupolar ie the combination of two dipolar fields. As opposed to quadrupolar in gravity waves. 

 Boit-Savant law can be uses to solve for the above if I recall.

[sethoflagos]

23 minutes ago, Prajna said:

In the current design the tabs or fingers are arranged every 20 degrees around the rotor. There are nine tabs and the rotor axle is co-planar with the centre line of the magnets, so when there is a tab central to the gap between the magnets on one side there is a space on the opposite side. The magnets are 10mm x 2mm neodymium (N52?) and the tabs centres are at approximately a 35mm radius. The 'bulb' on the tabs that lies between the magnets is 5mm radius to match the area of the magnets. This may be more or less optimal as far as effectiveness in switching the flux and suffering eddy current drag, I don't know yet.

This is somewhat arbitrary and is just my first best guess of what might work.

You're not answering the question I asked.

When there is a tab directly between the magnets, are the magnets at minimum separation, maximum separation, or somewhere in between.

Just to be absolutely clear on where you want to push and when you want to pull. It makes a difference. Rather like the ignition timing on a combustion engine.

[Prajna]

14 minutes ago, sethoflagos said:

You're not answering the question I asked.

When there is a tab directly between the magnets, are the magnets at minimum separation, maximum separation, or somewhere in between.

Just to be absolutely clear on where you want to push and when you want to pull. It makes a difference. Rather like the ignition timing on a combustion engine.

Sorry, I'm not sure of the dynamics yet. The rotor is rotated at whatever angular velocity, the tabs are spaced at 20 deg so the spacing will be varying. At the starting position one or the other side of the device will have a tab/finger centralised between the magnets and the other will have tabs spaced evenly above and below the gap. The magnets are prevented from closing completely on the tabs by a cam groove on the output flywheel. So, at the starting position the magnets with a tab in the gap will be at minimum separation.

[exchemist]

1 hour ago, Prajna said:

In the current design the tabs or fingers are arranged every 20 degrees around the rotor. There are nine tabs and the rotor axle is co-planar with the centre line of the magnets, so when there is a tab central to the gap between the magnets on one side there is a space on the opposite side. The magnets are 10mm x 2mm neodymium (N52?) and the tabs centres are at approximately a 35mm radius. The 'bulb' on the tabs that lies between the magnets is 5mm radius to match the area of the magnets. This may be more or less optimal as far as effectiveness in switching the flux and suffering eddy current drag, I don't know yet.

This is somewhat arbitrary and is just my first best guess of what might work.

I wish you guys would stop with the perpetual motion slur, I'm rather hoping for over unity!

Over unity is the same thing as what is traditionally known as a “perpetual motion machine of the first kind”, i.e. one that claims to break the 1st law of thermodynamics.  So it’s not a slur. 

There have also been ideas for perpetual motion machines of the 2nd kind, which claim to break the 2nd law of TD instead.

As I have mentioned, it can be good sport to spot the flaw in the logic of the designer. 
 

A rule of some patent offices, e.g. the US one, is patent applications for perpetual motion machines will only be accepted if accompanied by a working model. Which they never are, of course. So recognising perpetual motion machines is something patent office examiners (as Einstein once was,incidentally) and patent agents have to be able to do.

[Prajna]

6 minutes ago, exchemist said:

Over unity is the same thing as what is traditionally known as a “perpetual motion machine of the first kind”, i.e. one that claims to break the 1st law of thermodynamics.  So it’s not a slur. 

There have also been ideas for perpetual motion machines of the 2nd kind, which claim to break the 2nd law of TD instead.

As I have mentioned, it can be good sport to spot the flaw in the logic of the designer. 
 

A rule of some patent offices, e.g. the US one, is patent applications for perpetual motion machines will only be accepted if accompanied by a working model. Which they never are, of course. So recognising perpetual motion machines is something patent office examiners (as Einstein once was,incidentally) and patent agents have to be able to do.

I did add a rather conspicuous wink but thanks for the clarification.

[exchemist]

1 minute ago, Prajna said:

I did add a rather conspicuous wink but thanks for the clarification.

OK.

36 minutes ago, sethoflagos said:

You're not answering the question I asked.

When there is a tab directly between the magnets, are the magnets at minimum separation, maximum separation, or somewhere in between.

Just to be absolutely clear on where you want to push and when you want to pull. It makes a difference. Rather like the ignition timing on a combustion engine.

As I understand it, the idea is inserting the tab, or finger, causes the magnets to be attracted to it, instead of repelled from one another as they are in the previous phase of the motion. 

If we describe the operation in terms of an engine cycle, there are 4 phases:-

1) magnets close together no tab inserted, high energy of the field

2) magnets have moved apart due to mutual repulsion, reduction in field energy. Work imparted to output shaft

3) tab or finger inserted into the gap, causing magnets to be now attracted towards it, with further lowering of field energy. More work output to the output shaft (and some work output to the input shaft as well, due to the attraction)

4) tab removed from the gap between the magnets, which are now close together. This replaces the force of attraction to the tab or finger by mutual repulsion of the magnets, which are now at close separation, i.e. back to (1). It is this step that requires the substantial work input which returns the stored energy in the field to its stating value. Failure to realise the work need to do this is what can lead the incautious designer to think he has an over-unity machine, as the other steps all involve extracting work from the magnetic field.

At least, that is my energy-based analysis of this machine. 

 

[sethoflagos]

55 minutes ago, Prajna said:

Sorry, I'm not sure of the dynamics yet... So, at the starting position the magnets with a tab in the gap will be at minimum separation.

Pretty sure there's a x-post here with @exchemist so briefly:

If we're starting from your declared position of maximum attraction, we're moving against an attraction force for 90⁰; then with a weakened repulsive force (poles wide apart); then against the same repulsive force; then finally with the mirror image of the attraction of the initial power stroke.

In the absence of a proper mathematical analysis, by symmetry we have a nett zero sum. 

And then there's cam friction and the hysteresis braking mentioned earlier.

Granted I've ignored secondary effects of the movement of the magnets themselves but frankly, that's beyond my pay scale.

Suffice to say, if there was anything to see here, Faraday would have found it back in the day I think.

39 minutes ago, exchemist said:

OK.

As I understand it, the idea is inserting the tab, or finger, causes the magnets to be attracted to it, instead of repelled from one another as they are in the previous phase of the motion. 

If we describe the operation in terms of an engine cycle, there are 4 phases:-

1) magnets close together no tab inserted, high energy of the field

2) magnets have moved apart due to mutual repulsion, reduction in field energy. Work imparted to output shaft

3) tab or finger inserted into the gap, causing magnets to be now attracted towards it, with further lowering of field energy. More work output to the output shaft (and some work output to the input shaft as well, due to the attraction)

4) tab removed from the gap between the magnets, which are now close together. This replaces the force of attraction to the tab or finger by mutual repulsion of the magnets, which are now at close separation, i.e. back to (1). It is this step that requires the substantial work input which returns the stored energy in the field to its stating value. Failure to realise the work need to do this is what can lead the incautious designer to think he has an over-unity machine, as the other steps all involve extracting work from the magnetic field.

At least, that is my energy-based analysis of this machine. 

 

Looks right enough, so you've got the 180⁰ phase shift covered. Shall we leave the +/-90⁰ phase shifts to the OP?

Edited April 18 by sethoflagos
Typo