# Energy from Brownian motion

1. Dec 31, 2005

### ocunue

I know that this site is not for speculation, but can someone help me in this doubt ?
I know that perpetual motion of secod kind is considerated impossible.
So I would like to know why what are descripted under doesn't work.
I've a small permanent magnet that can remain in suspension on the water, this magnet for the movement of the molecules it will suffer a motion of brownian type.
If we approach a winding L1 to the permanent magnet on the winding there is a Electromotive Force caused by the movement of the magnet M.
Connecting a resistance to the heads of L1 in the resistance there will be a current with casual way that will heat the resistance while M will be braked.
Then we have heated R and cooled the water without work.
R is heated for the motion of M.
The difference of temperature will be very little but …. exists(????)!!! …
We could , for example, insert a diode in the circuit … so we could use the current ..
With a billion of billion of billion of this item may be we can obtain a perpetual battery of 1,5V ?

Thanks a lot

2. Dec 31, 2005

### chronon

The diode is in the same environment as the rest of the system, and so it will be affected by thermal effects, which will mean it won't work perfectly. This effect will be enough to cancel out the possibility of obtaining useful work from the system.

3. Dec 31, 2005

### ocunue

I think that we can make work the system at temperature of enviroment.
If the system in this way is cooled.. (because energy go away with the current generated) we can mantain the right temperature in the system giving to it the heat (from the external enviroment) that it need.
The most important thing is be able to cooling the water making go out energy in something way.

4. Dec 31, 2005

### Danger

Ocunue, there's no violation of thermodynamics if you don't make it a closed system. Brownian motion probably won't have enough force to move a magnet much, but perhaps a piezo crystal would respond well. It still won't likely produce enough electricity to be useable in any practical device, but it should be measureable. As long as the environment can provide energy to the water, it should keep going. What you're describing sounds like a very small-scale version of a wave-power generator.

5. Dec 31, 2005

### Q_Goest

Hi oconue, welcome to the board. Interesting question.

Hi there oh Dangerous one. Hope you have a terrific New Year's eve. Actually, ocunue is correct, the system he's proposed is a violation of the http://en.wikibooks.org/wiki/Second_Law_%28Engineering_Thermodynamics%29#Statement_of_the_Second_Law_of_Thermodynamics" [Broken]. It takes heat from the environment and converts it to work and does so without a cold sink. Seems a bit like a Maxwell Demon.

Also, I'd disagree the resolution has anything to do with the diode. We can put the diode in any environment, and assuming a perfect diode, that doesn't matter anyway. It doesn't reject heat.

I haven't figured out why or how this concept might be in error yet. Still thinkin'…

Last edited by a moderator: May 2, 2017
6. Dec 31, 2005

### Danger

Happy New Year yourself, Q.
You know, I totally misunderstood the original question.
I thought that he meant something like a pail of water, not a closed reservois system. I'd like to blame that on a language barrier, but it was probably the beer.

7. Dec 31, 2005

### Bystander

First glance, it looks like heat is being transported only one direction. In more detail, the inductive loop being used to generate electric power also broadcasts power from Johnson noise in the section of the system being "heated" by the resistor. This coupling of Johnson noise to the Brownian magnet's movement provides the two-way heat flow necessary to satisfy "us crusty ol' thermodynamosaurs."

8. Dec 31, 2005

### Q_Goest

That's always my excuse. The drinkin' shall commence shortly, it IS New Years eve after all!

Hey there crusty thermodynamosaurs. Not sure what exactly you mean by this:

I'm not an electrical engineer, but I did sleep at a Holiday Inn express last night. Are you suggesting that any power created is also rejected by random motion of electrons in the induced electric field, ie: the wire creating voltage, because the current produced would be so small? That sounds plausible, but I think one would need to do the analysis to prove it. The amount of current that could be generated is so tiny it seems reasonable though.

Let's make this just a bit more difficult. Let's move the current creating part of the system inside a superconducting generator. For example, a piston is impacted by ambient temperature molecules and caused to move (vibrate) with a spring or some other restoring force returning it to equilibrium after it moves. The piston has a long rod that goes inside a perfectly insulated superconducting coil and it is this rod that has a magnet on the end that induces a current inside the superconducting windings. In principal it could work if the rod and piston were of low enough mass. In practice, that may be impossible. But in this case, we have a current induced in a superconducting winding so we have no Johnson noise if I'm not mistaken. Also, the piston dimension is obviously going to be a function of the pressure created by the random motion of the molecules. We will need either an exceedingly small piston or a very low pressure in order to induce any movement in the piston.

Alternatively, I suppose one could put the entire apparatus into an environment where the current created was in a superconductor. Helium is still a gas at these temperatures and as superconductivity becomes a reality at higher temperatures, we could obviously use heavier molecules such as nitrogen or oxygen to induce Brownian motion.

The system would thus violate the second law of thermodynamics but I can't figure out exactly where the flaw in this concept is.

9. Dec 31, 2005

### Bystander

Flaws: 1) Brownian motion of the piston; 2) you've still got Johnson noise in the heating element you use to dissipate the generated power; 3) that noise is still broadcasting from your superconducting loop to drive the piston.

10. Dec 31, 2005

### Q_Goest

Bystander, I don't think we have the same concept in mind. From your responce I believe you have something else entirely in mind. It doesn't even sound like you're understanding the OP. From your list of flaws:

1) Brownian motion need not be a 'flaw'. It's real and obviously can drive exceedingly small masses in random directions. If for example, one can find a variation of molecular collisions over time for a sufficiently small piston, then vibration in that piston will be induced. That motion can be dampened in any way that creates power. The OP has a specific method in mind, but obviously one can expand that to include numerous different methods. Please explain why that might be a flaw.

2) There's no heating element or resistance involved in creating this current. That is outside the control volume if we place the CV around the heat in, and power out. You keep refering to resistance and heat where dissipated. Why?

3) The superconducting loop is not driving the piston. The piston or motion associated with random Brownian motion is the driver. Why do you say it the other way around?

11. Dec 31, 2005

### Bystander

No. I am stating that there is a two way exchange of energy between the OP's "cold" (loop plus Brownian magnet) and "hot" (resistor dissipating power generated by motion of magnet relative to loop) reservoirs. The return mechanism is the current arising from the Johnson noise in the part of the circuit, the resistor, in the "hot" reservoir inducing fields in the loop in the "cold" reservoir, moving the magnet and that motion being dissipated as heat.
In this case, your spring constant has to be low enough to accommodate the effects of Brownian motion; it is also so low that it accommodates the Brownian motion of your "low mass piston." The piston goes wherever it wants, whenever it wants. No problem, it's moving, and you're going to use whatever movement you can get to drive the magnet.
You've induced a current in a superconducting loop. The magnetic field associated with the loop current is going to eject your Brownian motion driven magnet. Bottom line at the end of the Brownian impact event? Your magnet hit the superconducting loop and was reflected with the same energy it acquired from the Brownian collision.
Correct. And, it drives all the small masses of the "machines" people postulate in equally random fashion. No net work.
Your "damping mechanisms" are also subject to Brownian motion, Johnson noise, and other random thermal phenomena.
Thought you were modifying the OP with a superconducting loop rather than trying to store the energy in a closed loop.
The loop most certainly does drive the piston; you attached a magnet to the end of a connecting rod. The field you induced in the loop collapses, returning the kinetic energy from the original Brownian collision to the magnet.

12. Jan 6, 2006

### ocunue

With a eletric transformer ?

In first time ... excuse me for the delay .. and thanks for the answers.
I have understood .. that the current generated from a resistor for the johnson noise is of the same
type (random and of the "same intensity") of the current generated from the magnet and the winding .. and then
the system is symmetrical and the heat is not transported in a preferred direction.
But the two current will be of the same intensity ... ? or there will be a current that is "stronger" ...
in this way the resistor will be more hot than the enviroment of the magnet....
BUT SUPPOSE (AS I UNDERSTOOD THAT IT IS) THAT THE TWO RANDOM CURRENT GENERATED IS OF THE SAME POWER
( or better of the same voltage, even if random)
IF WE CREATE A ELETTRIC TRANSFORMER ( BETTER IF MADE WITH SUPERCONDUCTOR ) .... and position it between
the winding and the resistor we can elevate the voltage generated. In this way the system became asymmetrical ..
and the power go from the winding to the resistor...
What do you think ????????
Where is the error in this consideration ?
I hope I have explained this in good way ...
Thanks ..

13. Jan 6, 2006

### Bystander

For a perfectly efficient transformer, power in equals power out. Voltage is stepped up, current is down; voltage is stepped down, current is up. Or, you've changed the source impedance driving your resistance heater, but not created an "asymmetry" in energy flow between two zones in your system.

14. Jan 6, 2006

### DaveC426913

I read somwhere about the concept of Brownian motion driving nanomachines. In a nuthshell, the nano-engine was merely a cog with a ratchet. Brownian motion would dislodge the cog, but it could turn only one way because of the ratchet. Thus, you would get an axle being driven.

Apparently, it was later shown to be flawed, though I don't know how.

15. Jan 6, 2006

### Bystander

The Brownian motion of the ratchet.

16. Jan 7, 2006

### Clausius2

Maybe I have missed some comment, but it seems here nobody has apparently talked about the reaction suffered by the magnet because of the EF generated. I mean, there is a dissipation of momentum of each particle colliding to the magnet, because the magnet is going to act as a wall. To explain it better, I think that once the magnet is delivered on the fluid surface, a bunch of molecules would collide with magnet, generating a net magnet movement and therefore an EF captured by an external EM circuit. This EF causes a reaction force upon the magnet (the same happens with the reaction torque in an electric motor). As each collision is inelastic and the magnet becomes braked by the reaction, there is going to be a dissipation of momentum of each fluid molecule. Eventually, this magnet (as a non perfect wall) would decrease fluid temperature, acting as a refrigerator. But we may not forget fluid is surrounded by non adiabatic walls (as any wall), therefore there would be an input of energy inside the fluid cancelling the refrigerating effect. I don't know if this is correct, but it's my "engineering view".

17. Aug 14, 2009

### StephenWeinst

1. You do not need the water or the magnet. In any conductor, there are randomly moving electrons (the amount of random electron motion is a function of temperature). This random electron movement in the winding will produce the same effect as your magnet, etc.

2. In a conductor not at absolute zero, a very slight amount of thermal energy is always being converted to random electron motion. If R and M are at the same temperature, then R is not heated, because the conversion of thermal energy at R to random electron motion exactly counteracts the resistive heating of R. Similarly, M is not being cooled, because the random electron motion generated at R causes a slight current in L that causes a slight magnet field that interacts with M, so M's temperature is also constant. If R and M are at different temperatures, then you will have heat transfer from whichever is warmer to whichever is cooler, gradually approaching an isothermal steady state.

3. Diodes do not work in the way that you describe. They are not on/off switches. They do allow some current to flow in either direction. They are useful in circuits because a much higher voltage is needed to create substantial flow in one direction than to create a similar flow in the other direction. However, for the extremely low currents produced using Brownian motion, a diode has essentially the same effect on current flowing in either direction and does not preferentially favor one direction over the other by a significant amount. Also, at these extremely low currents, effects of quantum mechanics that are normally ignored may be significant.