# Does a Torus dynamo actually work?

1. Feb 1, 2016

### Jonathan Swift

Hi all, in short does a Torus shaped generator actually generate any useful power? If so, is there any formula/calculator that can predict its output in watts? Ive seen some claims online but I'm skeptical. They are basically like a shake torch with a solenoid mechanism in a torus shaped device. Thanks in advance. Jon

2. Feb 1, 2016

### Staff: Mentor

Welcome to the PF.

Can you post some links to what you are asking about? I'm not familiar with torus style generators, but as long as they only claim a reasonable efficiency from mechanical input energy to output electrical energy, they may work. Standard generators can have some pretty high efficiency numbers, especially for large ones like in hydroelectric generation stations at dams.

3. Feb 1, 2016

4. Feb 1, 2016

### Simon Bridge

A quick google shows these devices widely touted as "free energy" or "overunity" devices - as such you are correct to view them with a lot of suspicion. Clearly any energy is not going to be free - we can expect the output to be lower than the input ... so you question seems to be more along the lines of:

Can we make this setup turn mechanical work into electricity?
More broadly: what would we have to do with this setup to make it turn work into electricity?

You could build one to see... sounds like someone has already discussed this here:
... didn't end well.

You can imagine a section of a toroid coil with a magnet inside on wheels, rocked back-and-forth like a rocking horse ... that would do it by the same principle as the "Faraday torch". The wheels mean that the magnet tries to stay at the lowest point while the torus-bit moves about it.
Care would need to be taken about the coupling between the magnet and the coil or they will just end up in sync. (The current induced by the magnet sets up a magnetic field inside the coil which opposes the motion... tending to drag the magnet with the coil.)

But that's just a quick naive impression off the top of my head... I think we'd need a detailed proposal to examine, and I have probably missed something.
To get a feel for the way crackpots misuse electricity and magnetism, and charlatans exploit common misunderstandings, you should look at D Simanek's impossible devices page.
https://www.lhup.edu/~dsimanek/museum/unwork.htm
... he has examples of magnet and electric "machines", and puzzles, that illustrate problems.

So much for my own resources...
If the toroid design were good for generating power, we'd expect to see some in industrial use ... so it is a good idea to see what others have tried.
Unfortunately the field is dominated by the overunity/free-energy crowd.

I found two patents:
... this has two solenoids with the magnet in a torus tube that passes through both.
... probably closer to what you are thinking about.
The existence of a US patent does not mean the devices work as advertised - both these designs feature on overunity websites so it's safe to dismiss them right away.

http://web.mit.edu/kirtley/kirtley/binlustuff/literature/electric machine/TORUS.pdf
... description of a generator/motor that uses a toroidal stator: with mathematics.
Seems these guys have published in IEEE Vol139 Iss6

Intregued, also found a paper (same journal) studying toroidal generators...

Comparing these designs with those in the patents: they are much more complicated, with much tougher maths.
If the patented designs were any use, wouldn't these engineers studying the design principle have come up with something similar and saved themselves a lot of work?

... so the short answer seems to be "yes and no" - a torus coil can be used as a component in an electric generator but the kinds of configurations used in youtube free-energy vids are not useful as normal generators either.

5. Feb 4, 2016

### Tom.G

Note that the three "patents" referenced in the above posts are actually Patent Applications, not issued patents. And the one referenced by the OP is over six years old. I wonder why?

6. Feb 6, 2016

### Jonathan Swift

I remember seeing a video which I think was from the Open University around 10years ago. A guy had built a plastic torus tube with a spherical magnet inside. Then it simply had some insulated copper wire wound all the way around but obviously terminating with a positive/negative at the top. He seemed to think it was enough to charge a mobile phone. It did seem to work although granted charging a phone doesn't require that much power.

If I were to make a small example I'm not sure what thickness of cable to use or how many windings I should have. If I do, I will certainly post a video up here. Even if it only produces 10watts at full pelt.

7. Feb 12, 2016

### Jonathan Swift

I have made a solenoid shaped tube with enameled wire. About 40cm in length and 20mm in diametre. It produced 0.2volts! Perhaps the tube I used to wrap the wire is too thick? Perhasps I didn't wind the cable neatly enough? Perhaps it just doesn't produce that much power?

8. Feb 13, 2016

### Simon Bridge

Not enough information to tell what you did.
Note: it is unlikely to have produced any steady voltage without considerably more that what you described.

I did cover the design in the patent application earlier tho... have you tried making a more standard generator to compare it with?

9. Feb 24, 2016

### Jonathan Swift

I have bought a shake lamp which works with a similiar principle but just in a tube. It generated 4 volts approx. and had a resistance of 700ohms approx....Not sure how accurate my multimetre is though. By my calculation I make that 0.02 watts! Does that sound right? Are there any online calculators that can give you expected power outputs of larger scale solenoid type devices?

10. Feb 25, 2016

### Simon Bridge

$P=V^2/R$ is for the power dissipated by load R when dropping potential V ...
A lamp is not a good load for this approach, its resistance changes with temperature.
How you do the measurements is important too... so, yeah, you probably did it wrong.
The power generated is VI ... where these are rms values.
It will vary with the frequency you shake it.

11. Feb 25, 2016

### Jonathan Swift

Measuring current is a bit tricky with my multi metre. For resistance it gave a reading of between 700ohms (which I think is the more accurate) upto 1 mega ohm. Yes, the voltage did vary and 4 volts was the highest reading. Is there any online calculator or even an equation that can give you likely power outputs if you type in rpm, number of coils, strength of magnet etc.

12. Feb 28, 2016

### Simon Bridge

Why would you think one reading is more accurate than the others?
How did you measure the resistance, and what exactly did you measure the resistance of?

Looks like your multimeter may be sensitive enough to register the fluctuations ... did you have it set to AC volts?
If you want to investigate generators seriously you should think about getting an oscilloscope ... you can get audio-frequency oscilloscopes that work through a desktop computer's mic socket.

No. The result is too sensitive to the design details.

13. Feb 29, 2016

### Jonathan Swift

I wouldnt expect the resistance reading to be large as there isnt much cable athough it is thin. I measured both ends of the insulated cable although this probably included the rest of the circuit (bulb etc.) I measured it with my multimetre on all the different resistance settings from 200 to 20mega ohm. The 200 range gave a reading of approx 700ohms, the 20 mega ohm gave a reading of 1 mega ohm approx.

Can you recommend one on eBay? Fairly sure I had it on the DC setting. Is there some reason why the length of the solenoid generators is quite short? Around an inch long? Is it that longer and larger ones would have more resistance and therfore not as productive? I'm guessing there is some correlation between amount of copper and power generated?

14. Feb 29, 2016

### Simon Bridge

Audio frequency oscilloscope: It's the sort of thing you build yourself.
http://www.instructables.com/id/Use-Your-Laptop-as-Oscilloscope/
... there are companies which manufacture kits and modules but I've never used them.

Looks like you can get smartphone oscilloscopes too. Google is your friend.

On DC settings, a voltmeter will show a fluctuation, and you want the peak voltage. You'd need a pretty responsive voltmeter though.
The number of coils in the solenoid is related to the amount of current you get with a particular rate of change of flux The relationship is not a simple one because it depends on the size of the magnet (it's physical size)... but the more current that can be generated, the harder you have to work at it so there's a trade off. These devices are usually used to trickle charge a battery so you get 10mins charge out of 2-3hours shaking it or something. Possibly the small size is to make sure all the coils have approximately the same rate of change of flux as the magnet passes through ... well, maybe by order of magnitude.

You realise that the coil is producing a current - the voltage you measure is dropped across a load.
In your case you are likely measuring the resistance of the coil and bulb in parallel, and the volt-drop across the bulb ... and anything else that is in parallel with the coil+magnet.

Beware - the resistance of a coil depends on how rapidly the current varies. Look up "impedance".
You will need to know about inductance and impedance, and also read up a lot about permanent magnets ... there is probably an engineering course online about generators, probably at senior level: there's a reason the torus generator paper above was post-grad.

I did find a paper that uses matlab to model a generator very similar to what you are considering.
http://systemdesign.illinois.edu/publications/Niu13a.pdf

15. Mar 1, 2016

### Jonathan Swift

http://systemdesign.illinois.edu/publications/Niu13a.pdf

So a wave generator. I don't understand why it is three phase though as I thought it would be producing direct current. Has a four sided permanent magnet linear generator got something to do with it? Sounds like this is a bit more complicated than what we have been discussing. 6 x10power5 watts (600,000 watts?) is a fair bit of power though although granted this is peak. Is there not some way of roughly estimating expected output without needing a degree in maths/engineering?