Meg

[Cyphermage]

I noticed Randi asks us to do a Web search for Bearden's MEG (Motionless Electromagnetic Generator) in his current commentary.

The major site describing this device seems to be..

http://jnaudin.free.fr/meg/meg.htm

I didn't know there were any permanent magnet free energy devices with no moving parts, so I found it interesting.

Would anyone care to skim the technical details, and explain specifically what's wrong with Bearden's approach?

I'd like something less general than "It violates several laws of physics which prohibit perpetual motion machines."

[Tanstaafl]

Would anyone care to skim the technical details, and explain specifically what's wrong with Bearden's approach?

I'd like something less general than "It violates several laws of physics which prohibit perpetual motion machines."

Well, while I'm an Electrical Engineer by training, I switched to software a long time ago, so I can't give a very definitive response. But as far I can tell it is simply gibberish. It can be remarkably hard to refute logically claims which are total nonsense. But maybe someone else that's more up-to-date can make some sense out of it.

Also, most of the links on that site don't seem to work.

[Cyphermage]

Well, while I'm an Electrical Engineer by training, I switched to software a long time ago, so I can't give a very definitive response. But as far I can tell it is simply gibberish. It can be remarkably hard to refute logically claims which are total nonsense. But maybe someone else that's more up-to-date can make some sense out of it.

Also, most of the links on that site don't seem to work.

Well, like most devices that claim to extract power from the ether using permanent magnets, the input is applied only during a fraction of the cycle, and it is alleged that the magnet does more work on the load than the applied input power.

The usual problem with such gadgets, which sometimes even fool real scientists, is that meters designed for measuring the power of sine waves often give spectacularly wrong readings when you feed them the complex waveforms used to drive alleged "Free Energy" devices.

If you look at the left of the two diagrams at the top of the Web page, you will see that Bearden's device consists of a rectangular magnetic core, wider than it is tall, with a permanent bar magnet positioned vertically in the middle. The flux from the center magnet goes from one pole to the other through two possible paths in the core, one on the left, and one on the right.

Now the orangey coils to either side of the middle on the top, are control coils, which can be energized to cancel half the flux of the magnet, causing all its flux to travel either through the left part of the core, or the right, depending on which coil you energize. The coils on the sides are power takeoff coils, which generate power as the magnetic flux through them goes between zero, and the entire flux of the permanent magnet.

Operation just consists of pulsing the control coils to switch the magnetic flux between all going through the left part of the core, and all going through the right, which you only have to do during part of the cycle to exclude half the flux from each side. The output coils are connected to a load, and it is claimed that the power out is greater than the power in, the excess energy coming from vacuum energy, through the process of "regauging" the electromagnetic field, a favorite term of the Free Energy people.

A guy in France made one, and tested it, and said it produced more power than is input, although I strongly suspect he ran afoul of the power measuring boo-boo I mentioned earlier. He never managed to "close the loop" and make it run off its own output.

That's about it. Looks like a fun Science Fair Project.

[Ginarley]

Surely the easiest measure would be a meter in and a meter out? Regardless of how it works inside, if you can demonstrate 1kWh in gives >1 kWh out then you are onto a winner... forget about power or electric fields or however it works.

[Cyphermage]

Surely the easiest measure would be a meter in and a meter out? Regardless of how it works inside, if you can demonstrate 1kWh in gives >1 kWh out then you are onto a winner... forget about power or electric fields or however it works.

A kW is a measure of power.

Unless you're running the signal through a resistive load inside a calorimeter, and measuring the temperature rise, you're really at the mercy of the parameters of your meter. A watt-hour meter such as those used by your local energy utility won't measure power consumption correctly that is not around the standard frequency and voltage that they deliver to your home.

Even expensive laboratory power meters can be thrown by complex waveforms, or energy in the RF range.

So proving a gadget doesn't output more power than it consumes is often a very complex measurement problem.

[Ginarley]

What had assumed is that this device would act as an electricity powered generator, taking in electricity and producing more electricity out the other end. I guess where i was coming from is that for one of these devices to be of any use beyond trivial they need to do what they claim they can do on the site, as shown in the passage below - that is take in 1kWh and output > 1 kWh for use in running appliances. Increasing power in a form that is no use to running appliances isn't very useful is it? And if that is what it is trying to do then why won't a simple meter prove/disprove whether this device works?

So at minimum, so far, you can get an OU solid state generator with a gain of 1.75. Just this alone could allow you to reduce your utility bill by about half if you put these devices between your breaker panel and your appliances. You could cascade them also to use 100 watts to get 15KW. It would only take about 9 stages to do this. What Jean-Louis has shown will do the job all by itself with no refinement. You just need bigger magnets, core section, and coils.

From a page on the site linked previously, emphasis mine.

[Cyphermage]

Increasing power in a form that is no use to running appliances isn't very useful is it?

I think demonstrating a genuine overunity effect would be quite useful, regardless of whether the output power was suitable for running applicances.

Basically, what Bearden is alleging is that if one sticks a powerful permanent magnet in series with the flux path in a transformer core, it exihibits anomalous behavior.

The rest, multiple coils, multiple flux paths, and other doo-hickeys, is just embellishment of this basic idea.

Obviously the answer to Bearden's claim is either "Yes it does", or "No it doesn't." Since the Maxwell Equations have worked pretty well up until now, my money is on "doesn't" absent a compelling repeatable experiment any scientist can perform.

Publishing nonsense papers full of 12-dimensional tensors in obscure physics journals refereed by Hungarians, and then claiming they have passed "peer review," doesn't exactly sell me on the concept either.

I suspect if we could have endless free power by putting permanent magnets in our AC transformers, we'd be doing it. It's certainly a lot cheaper than fusion research.

[Ginarley]

Agreed. If this was scientific research that was getting published in reputable journals, and was research for the betterment of knowledge then it would have a whole differerent approach to it. This whole thing stinks of a money making scheme. However the best way they can make money out of this is to demonstrate energy benefits as an essentially free electricity resource, and from reading their site I believe this is what they are claiming they have achieved.

Therefore meters in and meters out is the simplest, most believable and certainly most communicable measure of what they are trying to achieve. If it really worked it is by far the most obvious way to show it works. I have a strong suspician it doesn't work, which is why they are hiding behind scientific "research" and complicated physics to shroud their claim and fool potential investors.

[Curnir]

Obviously the answer to Bearden's claim is either "Yes it does", or "No it doesn't." Since the Maxwell Equations have worked pretty well up until now, my money is on "doesn't" absent a compelling repeatable experiment any scientist can perform.


I have a better responce to Bearden's claim:

Show me.

[Cyphermage]

I have a better responce to Bearden's claim:

Show me.

Of course, there have been other things that no one took seriously, because they were absolutely sure they couldn't work.

Noble gas chemistry, for instance. You'd think in the 70 years we were publishing chemistry books which stated as a fact that "The Noble Gases form no compounds, because their electron shells are completely filled," that someone might have put some Xenon and some Fluorine together.

But no one did, because everyone "knew" it was stupid and wouldn't work.

I remember a science fiction story about some scientists who were given the task of replicating an antigravity generator made by another scientist who had blown himself and his invention up while testing it. All they had was a film of him rising into the air and exploding.

After working very hard for a very long time, they finally managed to construct a very large and very cumbersome device that would lift itself a few inches off the floor.

Turns out the original film and story was a fake, because it was thought that if they worked on the project without absolutely certain knowlege that what they were doing was possible, they would just give up and conclude it couldn't be done.

My favorite current example of something people think is very difficult to do, or someone would have done it, but which probably isn't, is factoring the product of two large probable primes.

[ktesibios]

I'm not qualified to ciritique Bearden's physics (Shawn bishop, a Canadian physicist, has done that; see http://www.phact.org/e/z/meg.pdf and http://www.phact.org/e/z/bearden.htm ), but I did spend some time studying Naudin's MEG circuit back when he still linked to it on his front page.

There were several things that just screamed "dumb@$$ mistake".

First, Naudin was claiming that the MEG was delivering from 10 to 28 watts into a load consisting of a 1 or 2W carbon composition resistor. I've experimented (like, I suspect, most electronics geeks) with deliberately over-dissipating various types of resistor to see what would happen. I know that if you subject a carbon comp resistor to an overload of a thousand percent or so, you will get smoke and very likely flames. The absence of these was obvious proof that no such power levels were actualy being delivered.

Second, the digital scope screen caps which showed the MEG output voltage and current contained all the information needed to calculate the load impedance. In each case, the result came out to 10 kohms, +- 10% or so. the load resistor was a 100k carbon comp. This very strongly suggested that Naudin had made a simple order-of-magnitude error in measuring the load current. Losing track of where the decimal point belongs is one of the easiest mistakes to make; add a bit of confirmation bias and Bob's your uncle- you're off by a factor of ten and everyone but you knows it.

That left two mysteries- first, why a circuit which at first inspection appeared to be a garden-variety push-pull switching inverter- which should produce a square wave output- would produce a visually very good sinusoidal output, with the drain currents of the output MOSFETs being half-sines, and second, why the output voltage was nearly twice what one would predict on the basis of the output stage configuration, the supply rail voltage and the transformer turns ratio.

The answer turned out to be in the fact that a real-life transformer is somewhat more complex than the ideal transformers one encounters in beginning electronics books, where the turns ratio is the only important parameter. Real-world transformers have invisible "parasitic" components, the most important being winding resistance, magnetizing inductance, leakage inductance and winding capacitance. The most common way to model the complete equvalent circuit of a transformer is with an ideal transformer and external components representing the resistances, inductances, etc.

At high frequencies the leakage inductance forms a second-order low-pass filter with the secondary winding capacitance, load capacitance and load resistance (assuming a mostly resistive load). The damping of this filter depends on the ratio of source and load resistance to the capacitive and inductive reactances at the filter's natural frequency. If the d of the filter is <1, the magnitude of its voltage transfer function will be >1 near its cutoff frequency (this doesn't violate conservation of energy; if you do all the sums, accounting for input and output voltges and currents, it turns out that the power delivered to the load is always less than the power taken from the source).

I hypothesized that Naudin had wound a transformer with leakage inductance and stray capacitance such that they resonated near the frequency of operation, and that the parasitic low-pass filter was underdamped and had a peak in its frequency response.

Testing this with circuit modeling software, I found that by using a transformer equivalent circuit that included leakage inductance and stray capacitance, estimating the filter cutoff frequency and damping from the amount of voltage boost, calculating the required parasitic component values and doing some cut-and-try tweaking, I could duplicate the output voltage amplitude and waveform, and the same for the MOSFET drain currents, shown on Naudin's MEG page to within a few percent.

Mystery solved. The best explanation seems to be that Naudin is a sloppy experimenter who doesn't know how to make accurate measurements or enough circuit theory to realize that he had just inadvertently re-invented a type of inverter called a series resonant converter.

I still have the EWB5 circuit files, if anyone is interested in playing with the model.

He did apparently realize that his current measurements were highly suspect, because for a little while he was putzing around with a current probe for his scope trying to prove that the MEG was putting out more power than it took from the source. The page describing those efforts quietly went 404 after a few weeks.

That's one of the annoying things about woo experimenters. They throw the failed attempts down the memory hole, not realizing that failures are teachers just like successes.

[teddygrahams]

He "Conditions" his resistors so that they no longer have the resistance indicated, then calculates power using the original resistance.

[Cyphermage]

I'm not qualified to ciritique Bearden's physics (Shawn bishop, a Canadian physicist, has done that; see http://www.phact.org/e/z/meg.pdf and http://www.phact.org/e/z/bearden.htm ), but I did spend some time studying Naudin's MEG circuit back when he still linked to it on his front page.

This is very good. I also note he mentions the claim that such devices gradually demagnetize the permanent magnets which power them.

You'll find this said about overunity permanent magnet motors as well. It appears to be an attempt to get around appearing to be a perpetual motion machine.

Articles on the Takahashi motorscooter also claim it's powered by the magnets weakening.