Hall Effect Energy Generation

[tak22]

With permission and encouragement from forum members elias and gyula over at Energetic Forum:

Hi

Most of you know about the Hall Effect: Hall effect - Wikipedia, the free encyclopedia

This effect is widely used for sensing, but has it been thought to use it as an energy generation method? As it seems the the output energy flows perpendicular to the input energy, and seems to be more affected by the energy of the magnetic field, rather than the supplied current.

When we have got a conductor, and pass a current through it, the current can be deflected by using a strong magnetic field (e.g using a strong Neo Magnet) and a potential difference occurs perpendicular to the current flow. So can this potential difference used as an Energy Source? Any ideas? It seems that when we have got a current draw, a force can be applied by the magnetic field to create a potential difference which can provide energy. The question is: How much energy can be generated this way to surpass the energy needed to keep the current flowing. By using a very low voltage source, high currents can be generated by very low power levels.

I think that this system if properly designed can become a COP > 1 system.

To increase the potential difference, an n-type semiconductor is the best choice as it gives a very good potential difference.


I don't have n-type semiconductors at hand, any suggestions for buying?

Elias

Hi Elias an All,

I have done some search on manufactured Hall sensors that are also called as Hall generators. For instance this firm, F.W. Bell produces some type, see this link: fw bell products

At newark.com there is some of the Bell products but with a min 26 days leading time (they do not stock it):
More Test & Laboratory Equipment | Newark.com

Let's examine the most sensitive type, the SH-410 which is US$20.39 at Newark (with 26 days Lead Time).

Here is the data sheet:
http://www.fwbell.com/PDF%20Documents/SHseries.pdf

Suppose we get such device with the following parameters:

Rin=400 Ohm (ranges between 240-550 Ohms)
Rout=400 Ohm (ranges between also 240-550 Ohms)
Sensitivity ranges between 292 - 1120 mV/kGauss, let's pick a 300mV/kG
Nominal input current In=5mA

Now let us use a magnetic field of 1Tesla (10kG), this is an assumption from me that this device is able to operate at 1Tesla, data sheet does not give this data for this device), this gives 300mV*10=3V unterminated output voltage.

Using 5mA DC input current from a current source, input power is

Pin=0.005*0.005*400=0.01W=10mW

Because we want the highest output power, the output must be terminated with the same value of resistor as the output resistance itself, so I assume a linear relationship and the output voltage will be halved: Vout=3/2=1.5V DC
So the output power is

Pout=(1.5*1.5)/400=0.005625W=5.62mW This gives a COP=5.62/10=0.562

If you receive an SH-410 which happens to have the specified 1120mV/kG sensitivity (from data sheet), then the unterminated output voltage at the 5mA input current and at B=1T comes out as 11.2V, half of which is 5.6V.
So assuming the same input and output resistance, the output power comes as:
Pout=(5.6*5.6)/400=0.0784W

This gives a COP=0.0784/0.01=7.84 sounds quite a nice COP number!

Notice that half of the output power is dissipated in the device itself i.e. in the latter case 0.0784W heats the Hall device and 0.0784W is dissipated in the 400 Ohm terminating resistor.

I have not considered pulsed input current operation that gives pulsed AC output voltage.

I hope I did not make errors in the understanding of the data sheet and in calculations. IF you notice errors, please correct it. This Hall sensor so far seems an off the shelf device that has a COP possibility higher than 1.

Remark: In case the input resistance changes to a lower value when you connect a load resistor across the output, then input power should be revised again (though using a true current generator operation, this may be minimised?).

rgds, Gyula

This is a great find Gyula.

I read the datasheet. Because the input and output impedance is the same, and the voltage output of the hall device is dependent on the input current, then the only thing that matters is the magnetic field strength, after a certain magnetic field strength the system becomes COP>1.

For example considering SH-410, in a scenario we would provide the device with 20mA @ 10V (assuming 500ohms of input and output resistance) Pin = 200 mW.
Note that at maximum input current we have maximum sensitivity. So by considering a sensitivity of 1760 mV/kG, for 1 Tesla we have got 17.6 Volts at the output, so Pout = 17.6*17.6 / 500 = 620 mW, so COP = Pout / Pin = 3.1

But, we need to make usable power, so if we use a 500 ohm load at the output, we can produce about 154mW of usable power. The beauty is that we can increase the COP, by simply using a more powerful magnetic field.
if we double the magnetic field strength the usable power at the output would become 620 mW, thus yielding to a real COP of 3.1.

So by using a more powerful magnetic field and a more sensitive device, we have can increase the COP of the system.

This is really remarkable, somebody really needs to experiment with these devices.

Elias

tak

[gyula]

Hi Folks,

To continue this interesting topic, I would like to ask anyone if they are aware of highly sensitive 4 terminal Hall sensor types (like the SH-410 type) to mention them so that a possible hands-on test could be done to justify Elias' idea.
Unfortunately I have not found any other off the shelf 4 terminal Hall sensor other than the exceptional SH-410 that has got as high sensitivity as over 1000-1300mV/kG  when the input current is between 5 to 20mA.   (1kG=100Gauss and 10kG=1Tesla)

Thanks,  Gyula

EDIT: Sorry, I made typo in the convertion rate: 1kG=1000Gauss (not 100G).

[tak22]

Looking around for Hall sensors with high sensitivity and I find lots with their sensitivity listed in mV/mT,
and my conversion skills decode that as being a 10x factor from mV/kG, correct?

So 1300mV/kG = 130 mV/mT?

tak

[ion]

If I were interested in experimenting along these lines, I would consider the Hall effect in an ionized gas or plasma instead of a solid metal or semiconductor.

The Hall effect in an ionized gas (plasma) is significantly different from the Hall effect in solids (where the Hall parameter is always very inferior to unity).

In a plasma, the Hall parameter can take any value. The Hall parameter, β, in a plasma is the ratio between the electron gyrofrequency, Ωe, and the electron-heavy particle collision frequency, ν:

Gas tubes would not be too difficult to construct using this mode.

[gyula]

Looking around for Hall sensors with high sensitivity and I find lots with their sensitivity listed in mV/mT,
and my conversion skills decode that as being a 10x factor from mV/kG, correct?

So 1300mV/kG = 130 mV/mT?

tak

Hi Tak,

1Tesla is 10000Gauss so 1mT is 10Gauss, hence a x100 factor is involved... this means that 1300mV/kG is 13mV/mT.

(I made a typo in my previous mail what I now have corrected.

Notice that when you search for sensors, make sure they are 4 terminal devices because anything with less than 4 is already embedded into a circuit which needs a supply voltage of between 5-30V DC for feeding the circuit, you cannot access the Hall sensor itself in such a sensor with only 3 pins.

Gyula

[gyula]

Hi ION,

Thanks for this info, I did not know this earlier, very interesting.  Unfortunately I cannot make any gas tube for this purpose, I simply do not have the means at my workplace or at home.

rgds,  Gyula

If I were interested in experimenting along these lines, I would consider the Hall effect in an ionized gas or plasma instead of a solid metal or semiconductor.

Gas tubes would not be too difficult to construct using this mode.

[GibbsHelmholtz]

Well, I just had a thought on Hall effect energy generation and it leads me here.

I'm also trying to find out how can this concept achieve COP>1.  The main problem we facing here is to know if the energy deflected from the current subject to action reaction.   I can imagine the current produces voltage in a stationary magnetic field, or a moving magnetic field create a voltage.  If we are to draw current from this voltage, would it cause a reaction.  I think the answer is yes because it seems similar to a motor. 

However, the reason I bring this up because I also read something in addition call persistent current.  It seems that electron circling an atom represent persistent current and the energy is provided by everything in the universe.  If we are to stop the electron, it would comes back up to speed by absorbing energy from ambient.  Just like if we can completely stop an air molecule, it would come back into vibration by its neighbors. 

I believe the energy source is justify and the method is to combine persistent current with Hall effect to extract electron orbit energy. 

[exnihiloest]

...
However, the reason I bring this up because I also read something in addition call persistent current.  It seems that electron circling an atom represent persistent current and the energy is provided by everything in the universe.  If we are to stop the electron, it would comes back up to speed by absorbing energy from ambient.
...

A persistent current is a reality, see http://phys.org/news174222765.html, which is similar to a current in a superconducting loop: the current can flow forever. But if we try to extract energy, either it stops and energy is extracted, or it persists and other phenomena prevent the extraction. For example electronic spins and orbital rotations mean also persistent currents, they are obvious in permanent magnets, nevertheless no one has succeeded in extracting energy from them.
The idea to use the ambient energy to restart a persistent current from which we extract energy is not bad, providing that the cause of the persistence of the current is the ambient heat, otherwise there would be no reason for the electrons to flow again. But in the two first mentioned cases (physorg and superconducting loops), the heat is not the cause of the current: the electrons are like in ballistic motion, like planets orbiting a star, so there is no reason to think that heat would produce current, not even an initial "kick", for the perennial reason that a current is an ordered motion while heat is not.
So we are left with the classical question of a Maxwell's demon. An interesting Chinese experiment http://fr.arxiv.org/abs/physics/0311104 (or similar paper at http://fr.arxiv.org/abs/0904.3188) affirms it's possible and done but I note that the current is so weak that we are not beyond the experimental uncertainty and that even if it worked, we are far from useful energy. For that, the voltage should be multiplied by 10¹⁴ and so the energy by 10²⁸!

[GibbsHelmholtz]

Thanks Exn,

It's good to see someone out there already tried it and saw a little light.  I'm not going to worry about the result being too small in order.  I'll tell you the reason later.  I also don't think ambient heat is responsible for atomic motions.  It could be radiation field like gravity that we are yet define.  I would say that some EM waves are not visible light, so maybe not all heat is sensible in term of hot and cold.  But it's safe to conclude that electron will always be in motion and restart motion if we happens to extract its energy.

To scale up the Hall effect, we just need the large magnetic field to compensate for the small current.  What we working on (nano pulse of ordered watts) maybe proven to be a useful technique to achieve this.  However like you said, the current may persists and other phenomena prevent the extraction, so maybe the changing magnetic field is what needed and not a static field.  In either case, we can also achieve high changing magnetic field with our high watts pulse.   

[GibbsHelmholtz]

I see a clearer view now.

There are two things going on in a coil when current is applied. 
1/ domains alignment
2/ atomic expansion/contraction

Domain alignment is typical and the energy extract from transformer is of this type.  It is like swiping a magnet on a coil.  Atomic expansion and contraction is due to the present of the static magnetic field (Hall effect).  Electrons subject to a force F=qvB.  It then move radially from/to the atom causing pressure pulses.  The pressure energy is taking from electron motion and it ceased when the magnetic field stop changing. 

[F6FLT]

When we have got a conductor, and pass a current through it, the current can be deflected by using a strong magnetic field (e.g using a strong Neo Magnet) and a potential difference occurs perpendicular to the current flow. So can this potential difference used as an Energy Source?

Elias

Of course, because we have a potential difference. But this would be completely inefficient because of the low voltage, and even if it were high, the energy is taken from the current along the conductor and the COP is <1.

The Hall effect is the result of the Lorentz force acting on the electrons across the width of the conductor, which causes the potential difference.
If the Lorentz force were a path for free energy, the Faraday disc would already be spinning by itself.

[Hakasays]

If the Lorentz force were a path for free energy, the Faraday disc would already be spinning by itself.

Tesla discusses this in his 1891 'Notes on a Unipolar Dynamo' paper (attached).  The modus-operandi appears to be using a spiral segmented disk to force current to take an angled, tangental path rather than its natural path.

Of course, if there were non-conservative action inside a Faraday dynamo, it would be exceedingly difficult to detect due to the large friction losses at the periphery.   That, and Faraday dynamos are very rare to see in the modern world (mostly in laser/railgun tech as a rapid impulse power source)

[F6FLT]

Tesla discusses this in his 1891 'Notes on a Unipolar Dynamo' paper (attached).  The modus-operandi appears to be using a spiral segmented disk to force current to take an angled, tangental path rather than its natural path.
...

The modus-operandi  for what purpose?
The spiral does not change the work of the Lorentz force. As the conductor is at an angle a < 90° to the velocity vector, the intensity of the Lorentz force v.B is to be multiplied by sin(a) and sin(a)<1. At an angle of 45°, one has only about 2/3 of the force experienced by a radial conductor (and in addition the resistance increases).
It also doesn't increase the voltage compared to the pure Faraday disc (one can easily infer this from the cut-off flux, which is the same, so the EMF is the same). I have been thinking about increasing the voltage of homopolar generators for a long time, to no avail. I had eliminated this idea of the spiral I had imagined without being aware that Tesla had already had it .
The only possible advantage, weak now that we have neodymium magnets, is probably the increase in magnetic flux when the spiral is oriented in the right direction, which doesn't change the energy balance.

[Orthofield]

Hi Renaissance,

I appreciate you re-opening this long forgotten thread. Although not many of the comments refer to the Hall Effect, I think this is the most important part of it. The Hall Effect is one of the classically nonreciprocal effects. It is a fact that applying current to the Hall generator in a magnetic field produces a transverse voltage. It's also a fact that creating a transverse voltage doesn't produce a current or affect the existing current. So, the Voltage is indeed 'free' in pragmatic terms. In fact, Hall believed that he had found a new energy source, He believed that the magnet was doing work to create the E field. This is described in his journals. I can post the relevant portions if it's of interest.
I've lost it now, but many years ago I drew up plans for a Corbino generator: A disc of Bi has an electrode at center, and a ring electrode at the outer edge. A solenoid electromagnet, part of a tank circuit, is aligned with the axis of the disc. When the changing magnetic field passes through the disc, an AC voltage appears between the edge electrode and center, and this can power a load. If the tank circuit has a high Q there may be enough energy tapped at right angles to power the device and then some.
I used the Corbino effect because it is a bulk effect. and R losses are less than with the planar hall effect.
These two patents are also of interest in maximizing the efficiency of the Hall effect:

https://patents.google.com/patent/US3134082A
https://patents.google.com/patent/US3197651A

Fred

[Orthofield]

I see that I missed the more detailed discussion on the first page of this thread.. reading it now..

Fred

[Hakasays]

I appreciate you re-opening this long forgotten thread. Although not many of the comments refer to the Hall Effect, I think this is the most important part of it. The Hall Effect is one of the classically nonreciprocal effects. It is a fact that applying current to the Hall generator in a magnetic field produces a transverse voltage. It's also a fact that creating a transverse voltage doesn't produce a current or affect the existing current. So, the Voltage is indeed 'free' in pragmatic terms. In fact, Hall believed that he had found a new energy source, He believed that the magnet was doing work to create the E field.

Love the synchronicity   My work has drawn me to parametrics for exactly the same reason.
That varying the parameters of a system (in our case permittivity and permeability) also seems to result in mathematically nonreciprocal effects.

[Orthofield]

Hi Hakasys,

Yes, there is considerable overlap between the study of nonreciprocal circuit elements and parametrics. For instance, you often see them used together in patents of the 60s and 70s. But parametric devices aren't automatically nonreciprocal. Parametric transformers have loading, just loading of a different sort-- the L of the primary core is increasing as current tries to rise there, creating a parametric attenuation and loading of the source.

Still, the whole parametric interaction is poorly defined thermodynamically, and there seem to be a lot more loopholes to exploit than in inductive devices. It's just an endlessly fascinating area...

It does seem possible that these effects between secondary and primary can be made asymmetric in a way that appears nonreciprocal. I just mentioned a patent in Partzman's forum that may show such effects-- when two fluxes are used to vary the L that a third, orthogonal flux sees. From the patent it appears that INCREASING the current in X and Y coils INCREASES the L that Z coil sees-- and that changes  in flux through Z don't appreciably vary the L of X and Y, thus breaking the symmetry usually seen in these devices.
(The inventor doesn't use this for parametrics but for a current control transformer.)

https://patents.google.com/patent/US4210859A
pgs. 8 & 14

Fred

[F6FLT]

In a parametric system, the change of a parameter has an energy cost. The textbook case is the charged capacitor whose capacity is reduced, our parametric effect, so we increase the energy since W=1/2*Q²/C.
The calculation is simple: the work to move the plates away from each other is exactly the same as the electrical energy that will be gained by moving from C1 to C2.

The idea can be winning if the parametric change is done for free by a natural process, but I haven't found one yet.

[Hakasays]

In a parametric system, the change of a parameter has an energy cost.

That was my first impression as well, then I realized the energy required to induce a parametric change can generally be recovered. (in the case of a Mag-amp, via a high-Q LC circuit or Class E amp circuit at the control winding)
I find parametrics elegant conceptually, because one could directly see and quantify parametric gain/loss/destruction by measuring the area of the hysteresis cycle that overlaps itself; a 'negative hysteresis' region in that cycle.  Everything becomes directly engineerable (albeit still a massive pain to engineer).

It also helps explain why so many anomalous/unpredicted effects are being observed studying somewhat-recently-discovered multiferroics and negative-index metamaterials.
https://www.sciencedirect.com/science/article/pii/S1369702117307277

[F6FLT]

That was my first impression as well, then I realized the energy required to induce a parametric change can generally be recovered. ...

I agree. In the case of the capacitor, the Coulomb force of one plate on the other can cause them to return to their original position providing useful work.
But this useful work being only that of the previous change of parameter, the energy balance is null, we have gained nothing.

On the other hand, if we put out two large charged plates and one of them is moved back and forth by the wind, then we can recover the energy of the wind by the parametric variation of the capacity.

So you still have to find what would play the role of the wind in your multiferroics and negative-index metamaterials.

[Orthofield]

Hi F6FLT, Hakasys,

Yes, that is the standard understanding, and well worth repeating to eliminate the idea that parametric amplification is automatically 'overunity'.

Parametric changes due to heat, light and even acoustic energy have been tapped for power in numerous devices since early in the 1950's and perhaps earlier. (As I recall, Edison invented a device that generated energy from a change in L in a heated inductor). This is a well established area.

A very well designed parametric transformer is about COP .8-- somewhat less than a well designed conventional transformer. 

Having said that, actual measurements to test energy conservation are lacking. The same example of a plate capacitor is used in almost all cases, and extended to the inductive case without any change. Simultaneity is always assumed, even though it's rarely true in a magnetic core.

The behavior of a varactor diode whose C changes by a ratio of 20 with a 10 V reverse bias and close to zero current seems to contradict the simplicity of the standard formulation...

There are open questions in this area.

Fred

[Orthofield]

Hi Hakasys,

I don't know if you are currently building in the parametric area, but your posts have stimulated my interest. I'm currently involved in a thermoelectric build so I can't do anything much in terms of experimenting in that area. And I'm a novice builder anyway--been on the theory side for 25 years and only just started building. Well anyway, I had a thought this morning while drinking my coffee. Your typical parametric oscillator might be two toroids of saturable material (like permalloy or supermendur) with reversed windings on the secondaries to eliminate induction. The mag amp primary can be driven by DC pulses taking the toroids into the knee of the BH curve. Normally the load would be in series with the secondaries and a capacitor for tank oscillation. My idea is to put a transformer primary in the tank loop, and put the load on the secondary of this transformer. Now the load is reflecting to the primary and this in turn tends to attenuate the current through the tank circuit, which leads to a reduced 'parametric loading' on the DC pulse supply. Also, parametric transformers are very load sensitive, and use of different turns ratios in the transformer would allow the load impedance to be adjusted. This may be a half baked idea, but I thought I'd put it out there.

Fred 

[Hakasays]

Hi Hakasys,

I don't know if you are currently building in the parametric area, but your posts have stimulated my interest. I'm currently involved in a thermoelectric build so I can't do anything much in terms of experimenting in that area. And I'm a novice builder anyway--been on the theory side for 25 years and only just started building. Well anyway, I had a thought this morning while drinking my coffee. Your typical parametric oscillator might be two toroids of saturable material (like permalloy or supermendur) with reversed windings on the secondaries to eliminate induction. The mag amp primary can be driven by DC pulses taking the toroids into the knee of the BH curve. Normally the load would be in series with the secondaries and a capacitor for tank oscillation. My idea is to put a transformer primary in the tank loop, and put the load on the secondary of this transformer. Now the load is reflecting to the primary and this in turn tends to attenuate the current through the tank circuit, which leads to a reduced 'parametric loading' on the DC pulse supply. Also, parametric transformers are very load sensitive, and use of different turns ratios in the transformer would allow the load impedance to be adjusted. This may be a half baked idea, but I thought I'd put it out there.

Fred

Indeed the process can be very difficult to engineer because of all the variables at-play.
Not only are we balancing on the 'knee' of the BH curve with two cores, also balancing inductance between them, as well as providing both stable input and output impedance to measure input+output loads.  There is also the factor of the core itself , self-inductance vs mutual inductance, and the relaxation time of the material, which may or may not also overlap with magnetostriction.  Then there's also many other ways to achieve parametric effects that do not involve 2-core magamps.
I plan on doing a long thread on this eventually, but not until I have enough details+tips to make it useful for other amateurs.

[Allcanadian]

Hakasays
It's an interesting subject and as always its important to do our own experiments. For example, I was interested in the path a current actually takes under different circumstances. So in one experiment I used a 12" x 12" x 1" thick aluminum plate and tried to track the DC current across it with a hall effect magnetic field sensor. Contrary to popular belief the current never follows a straight path, its nonsense. The current self induces in the material which makes the path form an arc.

So we need to be very careful what we believe because many simply parrot what they have read as second hand information. In fact many theories we find in textbooks are outdated based on 100 year old technology. Later I used modern sensor arrays, a microcomputer and free 3D imaging software. It begs the question why we wouldn't do the experiments with new technology to find out what's actually going on for ourselves.

Personally, I went back to the beginning and repeated many of Faraday, Ampere and Webers experiments. As it turns out even these scientists missed a lot of effects or never noted them. Thus as Tesla pointed out much of our foundation is incomplete or flawed. For example, the Weber longitudinal force is real contrary to what most would parrot. However they never did the actual experiments which is generally always the case.

AC

[Orthofield]

Hi Hakasys,

I wouldn't overemphasize the difficulty of such projects. I've been involved in 7-10 parametric projects (depending on how you count them) since 1997 when I first worked with Jean-Louis Naudin on the subject, and we always got oscillation if not OU. Parametric oscillation is actually a pretty robust natural phenomena. The only project which approached, or was OU (still not sure) was one with varactor diodes. rather than cores.
In any case, when you begin your project work, I will have a deluge of suggestions and ideas. Be warned LOL!

Fred