Magnic Light: contactless bicycle dynamo light

[kaichu dento]

I posted on this topic earlier within the thread. If you care you can go there and read. For the Magnic Light to work with Eddy currents, those currents would have to evaporate the rim. People have good conversations on ghosts and other things.

I'm marginally interested in the concepts being discussed here, and also interested in seeing friendly discourse. Keep it friendly, don't post in an antagonistic manner, and stay on the subject without discussions of ghosts, which belong in a different forum.

 

[Obijuan Kenobe]

The more I see how Kickstarter is used, the clearer it becomes to me that it's a perfect venue for con artists. This one just happens to be of the "free energy" variety.

I agree with this 100%. I am sure there are many with wholly good intentions. However, having seen that wooden flashlight raise a significant amount of money, I gotta believe it at least also attracts this sort of nonsense.

(I also agree 100'5 with your posit that if it works, it would have to result in some noticeable resistance on the wheel used. That's just physics, right?)

obi

 

[AnAppleSnail]

Let us look at the numbers and see what it takes to rotate a magnet by an Eddy current, with a strong enough torque to produce useful power somewhere down the line. Let us say that I want from the rim a magnetic field of B=0.02T which is pretty meager, but let us start with that. The magnetic field of an infinitely thin straight wire is B=mu_0*I/(2*pi*r), where mu_0=4*pi*10^-7 T*m/A. Now the Magnic Light is some distance away from the rim and Eddy currents are spread out over the rim. Let us take an effective distance r=0.5cm = 0.005m. To produce the B above, I need to induce first the current of I=2*pi*r*B/mu_0 = 500A. The rim will go up in smoke with such a current applied locally. With losses, currents not being all ordered etc., the situation is much worse. On top of everything is the issue of how to induce such a current. When you are dealing with coils, you deal with a much smaller current but it circulates many times around with its effect multiplying in an ordered fashion.

I don't know a lot about eddy current, but I know that I have gotten significant force from magnets near metals without vaporizing anything. To me, this says that some assumption you have is incorrect. Where wiki Tesla unit 0.02T is a quite powerful magnetic field.

I'm skeptical of these numbers because of a very simple experiment I've done to demonstrate the relationship between electricity and magnetism. Get a magnet and a metal, non-ferrous pipe of similar size. I use extruded aluminum and (extruded or molded?) schedule 40 PVC because both are very smooth and can't 'snag' the magnet. I usually use a thin bar magnet for this trick, although round magnets work better (Shorter magnet-conductor distance).

Using a stopwatch, drop the magnet through each vertical pipe. [edit: For clarity. I get much longer times in the aluminum tube than the PVC tube. The speed appears to reach a constant value, meaning that the acceleration reaches 0. The magnetic Lentz force must accelerate it upwards at 1G to cancel out gravity]. While not tremendous, the magnet does weigh about 1 newton, meaning that a modest speed of conductor produces 1 newton of force in a magnet at low speed. The experiment works well for metal plates, though there is certainly less force produced due to decreased eddy currents.

It seems to me that if I make a wheel with one magnet in it, the magnet will brake when near a moving metal object. This will rotate my small wheel and bring the magnet around again. I will not have a chance to produce a video of this, but the principle seems to follow directly from the above experiment. Stronger magnets could probably let me get a watt or two from this, but I have not tried it. I don't think I will, but I thought I'd chime in on "Making useful power with eddy currents requires vaporizing metals." I feel that that is a false statement because of producing 1 newton at low speed.

 

[Steve K]

Let us look at the numbers and see what it takes to rotate a magnet by an Eddy current, with a strong enough torque to produce useful power somewhere down the line. Let us say that I want from the rim a magnetic field of B=0.02T which is pretty meager, but let us start with that.

I'll bite.. why assume that the rim needs to generate a 0.02T (200 Gauss) field?

 

[Steve K]

to open up the conversation a bit... has anyone heard any news about the Magnic??? I've poked around the web a few times since the Kickstarter project was big news, and didn't come up with anything. Not that I'm expecting them to have gotten production going yet, but I'd like to know how they are doing as far as funding, finding some place to do their assembly & test, etc. It would be fun to know some of their assumptions about the business case for the project too; expected production volume, tooling costs, selecting distribution channels, and whatnot.

I'd love to see some reviews and tests by independent parties too... any of those out there?

 

[AnAppleSnail]

An alu rim is made out of a straight piece that is bent into a circle and closed by a steel insert that you do not see from the outside. The magnetic image of your magnet in that insert is what in practice causes your magnet to rotate.

Let us look at the numbers and see what it takes to rotate a magnet by an Eddy current, with a strong enough torque to produce useful power somewhere down the line. Let us say that I want from the rim a magnetic field of B=0.02T which is pretty meager, but let us start with that. The magnetic field of an infinitely thin straight wire is B=mu_0*I/(2*pi*r), where mu_0=4*pi*10^-7 T*m/A. Now the Magnic Light is some distance away from the rim and Eddy currents are spread out over the rim. Let us take an effective distance r=0.5cm = 0.005m. To produce the B above, I need to induce first the current of I=2*pi*r*B/mu_0 = 500A. The rim will go up in smoke with such a current applied locally. With losses, currents not being all ordered etc., the situation is much worse. On top of everything is the issue of how to induce such a current. When you are dealing with coils, you deal with a much smaller current but it circulates many times around with its effect multiplying in an ordered fashion.

I'm afraid that I don't see where you answered the question asked. What have I missed? I see you say "What does it take to rotate a magnet by eddy current?" And then you answer "How do I produce a magnetic field of B=0.02T?" later. It seems to me that if the magnet exerts force on the wheel, the wheel exerts force on the magnet.

If a nearby neodymium magnet exerts 0.02T at the bicycle rim (Not induced - just from the magnet!), it seems to me that it would act on every electron in the rim at F=qvB. I may be taking this in the wrong direction, please correct me if so.

1 Tesla is defined as "1 coulomb of charge moving perpendicular to a 1 tesla field experiences at 1 m/s experiences 1 newton of force.

Q of an electron (Assume each aluminum atom has one) is 1.6*10^-19 C. That means it takes ten trillion electrons to move through a 1T field to experience 1 newton, or far less than one mole of aluminum. In our 0.02T field, it takes fifty billion, which is still far less than one mole. Let's assume we can only fit 1% of a mole of moving aluminum in this 0.02T field, then we'd get a tiny force moving at 1 m/s. But the bicycle wheel goes rather faster, aluminum has more electrons, and the magnetic field of a neodymium magnet is rather stronger. This exerted force must come from somewhere, cross product or no. Doesn't it come from the magnetic field's "source?"


Edit: But maybe it DOES just feed off the common steel insert. Doesn't the light flicker in time with the wheel's rotation?

 

[2_i]

I'll bite.. why assume that the rim needs to generate a 0.02T (200 Gauss) field?

Well, if I am to act on the magnet inside the Magnic Light with a magnetic field, there must be some field there of a sensible strength such as of one magnet affecting another. I took a value at the surface of the rim such as of a meager cheap magnet. Inside the Magnic Light, that field would still drop by a factor of few. For others, 1 Gauss is comparable to the field of Earth.

P.S. A strong magnet inside the Magnic Light will try to attract itself to its image in the fork and I should try to beat at least that with the rim, beyond any sensible power generation. Given this, the 0.02T is presumably well too small. Expensive magnets can have 0.3T.

 

[fyrstormer]

There is ALWAYS a way

Yes, and I described what the way is: either the dynamo drains a significant amount of power from the wheel's momentum, or it drains power from a battery hidden inside.

There is always a way, but the way is not always something you'll like.

 

[2_i]

Where wiki Tesla unit 0.02T is a quite powerful magnetic field.

I'm skeptical of these numbers because of a very simple experiment I've done to demonstrate the relationship between electricity and magnetism. Get a magnet and a metal, non-ferrous pipe of similar size. I use extruded aluminum and (extruded or molded?) schedule 40 PVC because both are very smooth and can't 'snag' the magnet. I usually use a thin bar magnet for this trick, although round magnets work better (Shorter magnet-conductor distance).

Using a stopwatch, drop the magnet through each vertical pipe. [edit: For clarity. I get much longer times in the aluminum tube than the PVC tube. The speed appears to reach a constant value, meaning that the acceleration reaches 0. The magnetic Lentz force must accelerate it upwards at 1G to cancel out gravity]. While not tremendous, the magnet does weigh about 1 newton, meaning that a modest speed of conductor produces 1 newton of force in a magnet at low speed. The experiment works well for metal plates, though there is certainly less force produced due to decreased eddy currents.

It seems to me that if I make a wheel with one magnet in it, the magnet will brake when near a moving metal object. This will rotate my small wheel and bring the magnet around again. I will not have a chance to produce a video of this, but the principle seems to follow directly from the above experiment. Stronger magnets could probably let me get a watt or two from this, but I have not tried it. I don't think I will, but I thought I'd chime in on "Making useful power with eddy currents requires vaporizing metals." I feel that that is a false statement because of producing 1 newton at low speed.

OK, you can go to a paper that solves the problem of a magnet dropped down a copper tube mathematically and compares the mathematical results to measurements. Incidentally, their magnets produce B=0.4T. There is a whole issue of experimentation with that magnet drop, outside of the cited paper, that underscores the nonsense behind the Eddy-current explanation of the Magnic Light (in fact poor-man's regular dynamo), but I have no time for this. From the data in the paper above, you can find the power dissipated by the Eddy currents in their drop and it is 0.004W. Where is this to 3W or so that you want to produce with a dynamo?! You can also calculate the magnitude of Eddy currents running around the circumference of the copper tube in their drop and I came up with 13A. In my estimate this is spread over a length of about 2 cm of the tube with walls that are 2mm thick. I can assure you that this tube can take it. There is nothing in that experiment inconsistent with my claim that Eddy currents needed for an operation of the Magnic Light would have to destroy the rim, provided you could produce them in the first place.

 

[Steamdonkey]

...There is nothing in that experiment inconsistent with my claim that Eddy currents needed for an operation of the Magnic Light would have to destroy the rim, provided you could produce them in the first place.

Maybe they're using eddy currents to spin a diametrically polarized magnet. If you think a non-ferrous conductor moving past such a magnet can't make it spin, then there's a gaping hole in your understanding of the subject matter at hand.

 

[2_i]

Maybe they're using eddy currents to spin a diametrically polarized magnet. If you think a non-ferrous conductor moving past such a magnet can't make it spin, then there's a gaping hole in your understanding of the subject matter at hand.

That is apparently the problem in the Magnic Light designers' thinking, in which they are not alone, in that when an effect is principally there, it should explain what they experience. The issue is that when an effect is there at a 0.1% level, it cannot explain 100%. Otherwise, I have to admit that there are big gaps present here and an on-line discussion can only do so much in bridging such gaps.

 

[Steamdonkey]

That is apparently the problem in the Magnic Light designers' thinking, in which they are not alone, in that when an effect is principally there, it should explain what they experience. The issue is that when an effect is there at a 0.1% level, it cannot explain 100%. Otherwise, I have to admit that there are big gaps present here and an on-line discussion can only do so much in bridging such gaps.

Previously you said, "What spun it [the magnet] was the steel insert joining the rim, not the eddy-current nonsense." Now you're sort of saying that's not true anymore?
From what I've seen, all your debunking efforts appear to address the hopelessness of harvesting power from the magnetic fields that arise from eddy currents in the rim. What I'm talking about is completely different. Where are you getting the idea that only 1/1000th of the necessary mechanical power can be transferred from the rim to a rotating magnet?

 

[fyrstormer]

I would like to point out that a bicycle wheel hub with an integrated dynamo costs $100-200, and having the dynamo hub built into a complete wheel costs about $100 more. So unless this thing is going to cost less than $200-300, it makes no sense for a consumer to wait indefinitely for the Magnic dynamo when they could have something that will definitely work in a week or less.

 

[2_i]

From what I've seen, all your debunking efforts appear to address the hopelessness of harvesting power from the magnetic fields that arise from eddy currents in the rim. What I'm talking about is completely different. Where are you getting the idea that only 1/1000th of the necessary mechanical power can be transferred from the rim to a rotating magnet?

An on-line discussion is not good means of teaching EM. You are interested in the effect of a moving conductor on a pole of a magnet and you perceive the harvesting of the energy above to be separate from the effect you are pursuing. Let us analyze the effect of a conductor on the magnet.

Provided the conductor does not drag much air along, to affect the magnetic pole it needs to produce a magnetic field. If the magnetic properties of the conductor are negligible, the only source of that field can be a current in the conductor and we are discussing the effect of Eddy currents induced by the relative motion of the conductor and the magnet. As has been mentioned before in the discussion, energy cannot come from nowhere and in the specific scheme the energy transfer is as follows: mechanical energy of the rim -> electromagnetic energy of Eddy currents -> mechanical energy of the magnet. Every transfer is lossy and having an extra transfer compared to a hub dynamo is already bad.

Now in the transfer of mechanical energy to Eddy currents, it does not matter whether the magnet or the conductor moves - just the relative motion matters. In the magnet drop down the tube experiment you could be lifting the tube up and have the magnet stay put in one position arriving at the situation approximating that of the Magnic Light. In the Magnic Light situation, the relative motion of the conductor and magnet is faster, which should enhance the first transfer of mechanical energy to the Eddy currents. However, the magnet drop down the tube has a better geometry for that transfer. The magnet is completely enclosed by the tube and you typically use thick copper rather than random shaped more lossy alu. In the end I think the transfer of energy will be greater in magnet drop than in the Magnic Light. However, under any circumstances the transfer will be of the similar order of magnitude, i.e. at the level of 0.004W mentioned before. This is the bottle-neck and you cannot turn it then back into 2W or so in the second transfer. Can you pull out some minute fraction of a watt? Sure you can. If you take a conductor moving faster than the rim, enclose magnetic field better, you can pull more, but just that this can be done in principle does not explain the operation of the Magnic Light. Now if you have a ferromagnetic element in the rim, it acts as a mirror for the magnet in the Magnic Light with the mirror image nearly as strong as the original. Provided you take a strong enough magnet you can take out any amount of energy from the rim, even bringing it to a stop.

P.S. In the supposed Magnic Light operation based on Eddy currents there is obviously one more transfer of energy: mechanical energy of the magnet -> electric energy for operating the light, bringing the number of transfers to 3, i.e. 2 more than in a hub dynamo.

 

[Steamdonkey]

An on-line discussion is not good means of teaching EM...

EM theory isn't very useful if you don't know how to apply it. I'm going to cut straight to the chase here... 5.5 watts. How does that grab you? I held a small (1/2" dia x 1/4" thick) diametrically polarized magnet near a spinning aluminum disk and held a coil of wire next to the magnet. The coil of wire was connected to a 5 ohm resistor. My Fluke 287 DMM measured 5.25 volts across that resistor. I also measured the current flowing into the resistor and it was just over 1 amp.

There is no eddy current bottleneck, at least not the way you describe. And while this may not prove the Magnic light will work as advertised, it does much to dismiss the "eddy current nonsense" we keep hearing about.

 

[2_i]

So why don't you repeat it with the surface of the disk moving at 10km/h?

 

[Steamdonkey]

So why don't you repeat it with the surface of the disk moving at 10km/h?

How do you know it wasn't already moving at 10 km/h? Some extra speed isn't going to explain away the 137500% error in your theoretical figures. These results invalidate almost everything you said about using eddy current coupling to spin a magnet. Maybe you're hoping we'll overlook that if you can point out that my test setup can't match the figures claimed my Magnic? Keep in mind this was a quick and dirty proof of concept experiment using stuff I had laying around my workshop, not a design optimized for this application. As for the speed of the 2.5" aluminum disk I used, I don't have a way to measure it but it was nothing crazy. It was powered with an old 1/10 scale RC car motor running about 2.5 volts.

 

[2_i]

How do you know it wasn't already moving at 10 km/h?

I know because it would violate physics and repeatable experiments. Eddy breaking is employed in practice and you do not need to prove it. Efficiency of the breaking depends roughly on the square of the speed. Before you mentioned a Dremel and for a typical Dremel the speed would have been roughly 340 km/h. At this speed even the air drag would matter. I quit this discussion because it is pretty pointless. I tried to address the issues you raised rationally but the target is moving and I do not see any benefit from this discussion.

 

[AnAppleSnail]

2_i, sorry to see you go. Whenever I pull numbers out on the internet,I try to show where they came from. I'd love to know where to get a dremel that goes 180 thousand rpm.1cm disk, perimeter speed of 340000 m/h, or 5670 m/min, at (pi x d) 0.0314 m per revolution. 5760/0.0314 gives 183000 RPM. My dremel only ever did ten to twenty thousand rpm, or 37 km/h on a 1cm disk.

 

[2_i]

I took 28k rpm (off my decades old Dremel) * 60 *pi *.024*2.5 = 335 km/h. Did you mean where pi came from?