Maxwell's demon

This is where there are basically no rules about topics, or off-topics. Discuss things here of fleeting interest.
Forum rules
Here, you can discuss anything (well, anything legal and not offensive) you want to. Use this for gassing about any half-baked theories, general getting to know one another, and other things that as someone once said, should be forgotten after awhile. This sub forum is set to auto-remove threads that haven't been posted on for a couple weeks, emptied like the office trash can. Almost anything goes here, the idea being to keep the other forums and threads more on topic but in a maximally friendly way. If anything actually worthwhile should wind up here, let me know and I will make it immune from being removed.

Maxwell's demon

Postby Doug Coulter » Mon Mar 12, 2012 6:15 pm

I've posted a couple of times ideas about how one might violate the oversimplified take on the 2nd law of thermodynamics - all you need is a selection process, or some nonlinearity someplace to pull it off.

For example, it's been observed that a tungsten substrate with a bunch of protruding rods starts emitting visible light before it's even "red hot". In this case, the little rods are tuned dipoles, and are emitting energy at their resonance from phonons at the "tail" of the maxwell distribution, and those "tails" are replaced by random interactions as they are lost.

The one I thought of was based on Brownian motion. Suspend a tiny magnet in something hot, then wind coils around it all three axes - with perfect diodes, you could then sum the energy from the coils, and from multiple cells (you'd need quite a few to make any real amount of energy) to get electricity - while the "battery" cools itself. Taken to an extreme, of course, it would stay warm as any temperature differential on this planet tends to create things like convection.

Now, in one day, there are two more at least plausible articles on IOP about maxwell's demons that might actually be real. I was only 40 years ahead of them.
http://physicsworld.com/cws/article/news/48889 In this one, the fact that electrons have higher speed for lower energy in graphene than anything else is the "nonlinearity".

http://physicsworld.com/cws/article/news/48882 In this one, a led uses phonons (lattice vibrations, or heat) to boost the probability of emitting a higher energy photon. IN this case, the band gap is the "selector".

Could the world be saved already? Are there flaws in the observations here? I think the premises are plausible, but when I ran the numbers for my own thought, it turns out you need a heck of a lot of it to get even a microwatt - now, small energies like that can be useful - especially if they last forever, but of course the world is looking for mega, not pico watts.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Maxwell's demon

Postby chrismb » Tue Mar 13, 2012 5:33 am

'Fraid I see nothing in these that is likely to 'defeat' Thermo-2. Speaking from a 'classical' point of view, I believe the answers would be;

The one I thought of was based on Brownian motion. Suspend a tiny magnet in something hot, then wind coils around it all three axes - with perfect diodes, you could then sum the energy from the coils,
Well, there is no energy in the coils unless they complete a circuit (that is, they create a reactive/subtractive magnetic field of their own). Without an electrical load, they are simply lumps of wire metal in the presence of a magnetic field of constant energy (albeit a moving field). And when they do complete a circuit, Lenz law will lock the magnet preventing free motion.

Now, in one day, there are two more at least plausible articles on IOP about maxwell's demons that might actually be real. I was only 40 years ahead of them.
http://physicsworld.com/cws/article/news/48889 In this one, the fact that electrons have higher speed for lower energy in graphene than anything else is the "nonlinearity".
Sounds just like a chemical cell, to me. Dissimilar metals in an electrolyte? This is a perpetual motion machine claim. It is not interesting [yet]. If they are serious, then they should build these cells into a fully sealed transparent container with the absolute minimum quantity of electrolyte, and some LEDs in circuit to show it is operating, then present it to an independent physics department as the first ever permanent light source who can put it on secure public displace for everyone to come and look at. Then in a few weeks/months/years [however long to remove any chance of it being 'chemical'] we'll see if it's still on and/or whether there is any chemical corrosion going on, indicative of electro-chemistry.

http://physicsworld.com/cws/article/news/48882 In this one, a led uses phonons (lattice vibrations, or heat) to boost the probability of emitting a higher energy photon. IN this case, the band gap is the "selector".
This is interesting, and potentially useful, but does not defeat 2-T. Think of it as a reverse cycle air-conditioner but the output is light rather than heat; you pump in energy and not only get that energy out again, but draw in from the surrounding environment too. Iff a 200% elec-to-light efficiency is proven here, then the corollary is that light-to-electricity can never exceed 50% efficiency, else we're back to a claimed perpetual motion machine.

Could the world be saved already?
Impossible! We are doomed already! :twisted:
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: Maxwell's demon

Postby Doug Coulter » Tue Mar 13, 2012 11:43 am

While I might agree we're doomed already, you're incorrect on some of the other stuff. No, the magnet won't be locked, its motion will merely be resisted somewhat, since until it moves, no reactive force exists and the force increases with the rate of motion. This is precisely the same thing as happens when you load any other generator - the shaft becomes harder, but not impossible, to turn - that's how the energy is put in.
You can trivially demo this to yourself with any small PM motor by testing how hard it is to turn open or shorted (eg loaded by it's own winding resistance).

And since the motion is resisted, the magnet doesn't move as much in response to the occasionally coherent bunch of hits it receives from the molecules in the liquid (by pure random chance, sometimes many are moving the same way at the same time), slowing them down more than normal (inelastic removal of energy from the system) and as a result, the liquid cools.

This phenomenon depends on randomness to work, just as when you toss a coin enough times, you'll sometimes get a bunch of heads or tails in a row. If you only score energy when that happens, you still get some, and since there's no free, it came from somewhere - the heat energy of the fluid.

This one's been looked over by better scientists (at least with more creds) than anyone here - it's not a chimera, the real issue is it's not very good - there's not all that much energy in the "tails" to get at low temperatures. None of these claim to be perpetual motion, they merely claim that via some selection process or nonlinearity, that some low grade heat can be converted to higher grade energy, which most people think violates second law - but it doesn't because the second law only states that *on average* energy moves from higher to lower grade - entropy, in other words.

That "on average" is the key loophole being exploited here via cleverness, in all these approaches. Pure randomness is actually in this case the driver, as it creates high energy tails on any distribution - so all we are looking at here is "scraping off" those high energy tails into, well, high energy quality output, which is not perpetual motion at all - just moving energy from one place to another. Randomness then regenerates the tails by pure chance, so we can keep doing it as long as there's energy there to get - which is always present in anything above absolute zero, just not all that much.

Again, the key is either a selection process that only selects the high energy tails, or some nonlinearity in something (the diode in my example) that accomplishes the same thing. If you had a perfect diode (no forward drop, no resistance or parasitic C) then you could do this just fine with the thermal/Johnson noise in any resistor - it's just not a lot of power per unit.

The posters on the graphene battery thread over there did mention that he should check for electro-chemistry happening, and till he does, the jury's out on that one - but it IS plausible it could work.
It's not going to be hard to test, and frankly the materials involved don't really react well with one another in that sense anyway - you'd be hard pressed to make an electrochemical cell that depended on eating gold into solution of those kinds of ions.

The LED thing can also work. Same set of ideas there. Remember, that at the threshold of boiling, there is even a selection process going on in a cup of tea - only the few water molecules that gain more than average energy through random collisions escape as water vapor - which even happens at room temperature, well below the boiling energy for the bulk of the liquid, yet water dries up. Again, while tossing enough coins will get you a 50-50 distribution of heads and tails in the long run, for short runs you can get wildly different statistics - and that's the essence of what's going on here, no magic required at all.

Two photon processes, while not common, are well known in other branches of physics, where in theory one photon can't "make the nut" of some energetic threshold, but if there are enough of them, sometimes two can.

And the tungsten thing was observed in a completely other context, and went unexplained for awhile, till a crusty old RF engineer pointed out that photons is photons, antennas antennas, and straightened them out on that one - that'd be me in this particular case. They were trying to make a photonic "echo chamber" in the first place, which is why all the dimensions of their "antennas" were right for visible light, and noticed this effect while annealing the thing to get rid of some defects in production when it was noticed. A real Fleming moment.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Maxwell's demon

Postby chrismb » Tue Mar 13, 2012 12:28 pm

Doug Coulter wrote:That "on average" is the key loophole being exploited here via cleverness, in all these approaches. Pure randomness is actually in this case the driver, as it creates high energy tails on any distribution....The LED thing can also work.


Sorry, Doug, but I wholeheartedly disagree with the comment that these individual things are the same. In the gold/carbon/silver battery case there is no outside work being proposed as an input. Maxwell's demon is an energetic demon. He needs to perform work before he can start batting the hot particles in the right direction. Energy must be expended to reduce entropy. In the case of the LED there is a clear mechanism by which the local entropy can be reduced and driven to produce light because there is input work to do it.

In your magnet idea, OK, 'locked' was perhaps not the best word but essentially you are then coupling the thermic agitation of the magnet to that of the wires/components in the loops and circuits. What you might achieve is to thermalise the magnet to the coils through the magnetic coupling, but once the magnet and coils are at the same temperature, there would be no further asymmetric coupling of energy one way or the other.

It's a fool's gold to go hunting for 2-T defeating perpetual motion machines, Doug. Don't go down that route, it'll only send you as mad as those Alchemists trying to manufacture gold! Feel free to try - there is plenty of good science to be had in the trying, but it is irrational to think it is achievable.
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: Maxwell's demon

Postby Doug Coulter » Tue Mar 13, 2012 3:34 pm

Sorry Chris, the actual universe, via observations, says you're wrong about this one. The magnet/brownian scheme requires no temperature drop between its parts to work, just a temperature above 0k for the working fluid, though it obviously works better at higher temperatures where the kinetic energy density per atom is higher. There have now been peer-reviewed real papers on it, 40 years or so after I started taking crap for suggesting it was possible. Now that it's "their idea", it's fine.

http://www.physorg.com/news/2012-03-lan ... rased.html

You're not of course refuting that there is thermal noise in any resistance or impedance that has nothing whatever to do with heat moving from place to place (at least not in and out of the impedance). It's there because heat is there, period, with the jiggling creating non uniform electron distributions and thus, a signal. If you had a perfect rectifier, you could even take energy out of that system without putting any heat back into it - and the resistor would cool as a result. It's just that it's not a lot of power there - enough to be an effect in preamps, but not power a flashlight. It's just not the best way to get power - the brownian is better, and can be optimized re the weight ratio of the magnet vs the atoms in the fluid, as well as for a particular band of frequencies most of the motion energy will land in.

No, Maxwell's demon only needs to expend the energy required to tell between hot and cold. That need not be the entire energy of the system, that's an oversimplification of the 2nd law that is just wrong - but widely disseminated in lower school science books (and self appointed academics who haven't actually looked hard enough at it). It needn't "bat" anything at all - it need only open or close a door, for example. Each of those things theoretically need only take one quanta at the Planck scale, and an atom can have a hell of a lot more energy than that in kinetic energy at room temperature.

Remember that correctly stated, the second law deals only with "on average" - and in the absence of coherence or non linearities. If you have a non linearity or skimming-selector, you're not cheating it, you're getting in the back door. Then the "on average" part actually helps you by restoring the tails of the distribution you stole...starting from of course, a lower total energy since you've skimmed some off.

This happens naturally every time a water molecule evaporates and some in the same glass don't - because in the Maxwell distribution of energies there are tails that extend out forever - they get small far from the middle of the gaussian to be sure, but they are there and non zero nevertheless. And you know what, the remaining water is chilled, that's where the energy came from. Freeze drying comes to mind as something you can try at home in a vacuum system as an example of this. If at some temperature, everything was moving the same speed, none of this would work - but that's not reality, reality has a distribution of speeds.
In this case, the "door" is merely the surface binding energy of the bulk liquid - a barrier you can either tunnel through or not (just like DD fusion). If a particular atom has more than that much energy, it will always break that barrier and free itself, if less, only sometimes. But due to it carrying off energy - the water left behind is cooler than before. Which is where the energy came from. No free.

This is one of the very things you yourself complain about - dogma stopping good ideas in science from being investigated. You're spewing it now. Of course, none of this will ever be major power source - the energy density is just too low to get tons of power out of chilling stuff from room temperature (and depending on the normal direction of entropy to reheat it from environmental sources). But in some cases, a little power is worth the hassle this would entail.

In fact, I tried to get the magnet thing made by a semi fab (LTC's, since I have a friend there). They looked it over and their scientists thought it would work, but not be profitable, so I couldn't get it done.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Maxwell's demon

Postby chrismb » Tue Mar 13, 2012 4:18 pm

Quoting the Landauer experiment is an argument against your position, Doug, not for it! It evidences that work needs to be done when you change even the smallest possible part of the system - just one bit of information changed increases entropy.

Landauer is an experimental outcome that is a disproof that Maxwell's demon exists, whereas you are trying to conjure one up.

I'm not trying to argue a dogmatic position, I'm saying it is irrational because whilst you keep agreeing the average entropy must increase, simultaneously you're claiming that you can reverse it for a whole system, and keep it reversed. That's not the 'average' increasing!! You're saying 'I examine N elements of a system and scoop off the top most energetic 1% of them. I claim this is a process that can be done without performing any work and 2-T is not violated but, lo, I now have less entropy in the whole of my bounded system of N elements than before, and I've done no work to get it!!'

This isn't the LED example above (and is certainly not Landauer). The LED example goes 'I take J energy and N elements. I now convert J into another form of energy (or total energies), K, and scoop P% of the top of the distribution of N elements in the process whilst J-K work is being done. The entropy of my N element is now lower than it was before.' This is fine, because what you'll find is that J-K is always more than the change of entropy of the N elements, thus entropy of the J energy and N elements has gone up, even though it has gone down for [just] the N elements.

I'm not seeking to challenge your chance and desire to propose, design and perform an experiment to prove me wrong. Great! Your forum, your right to set the narrative. I just wanted to let you know that I'm not, and still not, convinced!... Onwards to the experiments!... :P
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: Maxwell's demon

Postby Doug Coulter » Tue Mar 13, 2012 6:22 pm

The Landauer experiment points out what the minimum entropy increase happens to be per bit (or choice) - and it's tiny on the scales involved, so tiny it took quite awhile to even measure it - whereas the kinetic energy of an atom a room temp can be easily measured because it's so much huger - which was my point here, so I feel it supports the idea that you can indeed do something new - without any actual cheating. You're just missing the scale factors here, which is why you think it won't work (I think). I think you're missing out understanding what happens at the boundary where things are neither pure quantum or pure classical, which is where the action is these days. We don't really have great equations for this region at the moment, we kind of splice the pure quantum into the pure classical as an interpolation. Thus, this stuff is hard to predict. It's been a bunch of real surprises to the nano tech guys when things don't act as they expect at this scale.

Other than saying I can't cheat 2nd law "just because" - what specifically will not work in the brownian magnet model? Are you assuming I have more I2R loss in the windings than I could have? Is there a quantum limit on how good a diode I can have? What specifically won't work here, other than just repeating dogma? I'm not chaning the total energy in the universe here, and frankly when most people try to get from physics to information theory, they tend to leave out scale factors again - that nifty planck constant, which kind of changes the story when you use it correctly for the system at hand.

IN the magnet system, you have to choose the particle size right to maximize energy harvesting. IF you make it too big, collisions that push it with most of the atoms pushing the same way are too rare, and it's too heavy to move much in response to lesser ones - so you can't get much generation out of it. If you make it too small, it can't be a magnet, you have to make everything so small it's hard to make, and it moves too quick (microwave frequencies) to get useful DC out of it with rectification. There's a fairly narrow sweet spot here according to the math of random walks.

Don't forget, if we are to believe these self-same-self-appointed authorities, the universe could not exist or have come into being as they claim from a big bang...but here we find ourselves, somehow.
Entropy seems to be tied into the direction of time. Funny thing that physics, standard model, says everything works either direction and the sign bit of time can't be determined any other way, eh?
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Maxwell's demon

Postby chrismb » Tue Mar 13, 2012 6:35 pm

Doug Coulter wrote:Other than saying I can't cheat 2nd law "just because" - what specifically will not work in the brownian magnet model?


OK, so you have already mentioned that with your perfect, zero drop diode you'd see a current on a loop of wire just because of random thermic transportation.

So, how hot do the coils need to be before they generate a current sufficient to generate a magnetic field stronger than the one the magnet itself is making?

It looks to me like you've assumed that process is only going to work one way, whereas what would happen in perfectus is that the magnets may well induce a teeny current on the coils, just as the coils may well induce a teeny magnetic field on the magnet. The magnetic field in this example would be the coupling between what are essentially [schematically] two points at different temperatures. It's a magnetic version of the Seebeck effect, and once the two parts reach the same temperature, no further magnetic current will flow, in exactly the same way that once the two junctions in a Seebeck experiment get to the same temp there is no longer a current that can do real work.

Your magnet system may well generate useful work whilst it is undergoing thermalisation. Just like any other system. Darn, it may even work out as a reasonable heat-recovery scheme (as many companies have been pushing for years with way-out claims they never live up to)! But you'll still need a temperature differential. You cannot recover useful work out of a system that is simply undergoing normal thermic agitation.
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: Maxwell's demon

Postby Doug Coulter » Thu Mar 15, 2012 9:26 pm

I would think the tiny thermal noise from the coil would simply add as noise normally does, and represent some backwards leakage - like thermal leakage through a peltier device does, merely limiting the total efficiency. To some extent, that simple reheats the liquid that you're otherwise cooling. Noise power in a fairly low impedance coil (it'd have to be, having to be tiny) is small, whereas by inspection, the power you see in a thing jiggling due to brownian motion under a microscope is large. Nanowatts vs microwatts if I judge correctly. This would be due to the selection that happens due to thousands of atoms hitting the magnet preferentially from one side vs the other - whenever the system "feels like it", vs the random jiggling that creates the normal thermal noise. It's an issue of scale I think - one being a lot bigger than the other, but the system should work if one scale doesn't match the other.

Guess I'll just have to try it, but it's darned hard to fabricate - which is why no one else has. The scales I've looked at say it's gotta be tiny to work well. I guess you could prove it *could* work by showing that it turned all the random agitation into spin around the magnet axis if you allowed that symmetry to remain - it'd be the only free motion allowed.

The tungsten-rod system violates all the oversimplified black body radiation laws, interestingly. No new energy appears, but it IS interesting that at low temps white light comes out of something "simply mechanical". No phosphors or anything, just a buncha antennas.

But again, rather than analyzing the situation, you say that "you cannot get energy without a thermal drop" - relying on authority, which isn't making me satisfied anymore than "humans weren't meant to go over 25 mph" does. I say that with a selection process, you can - or I still don't see why not.

I have a very interesting and math-laden book on thermodynamics, in which the author spends the whole first half of the book ranting (with math backup) about how all energy eventually equipartition's into all degrees of freedom, and such things can never be. Very convincing until you read the second half - where things like population inversions that utterly violate the equipartition of energy into all available states are discussed in detail - with no apparent looking back to the first part of the book, which it's in direct conflict with. Funny as a very esoteric joke. Even he couldn't seem to see the conflict between "eventually" and "what's happening right now", which is actually pretty close to the crux of my argument here. Same deal with the coin flipping simile.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Maxwell's demon

Postby chrismb » Fri Mar 16, 2012 12:43 pm

Doug Coulter wrote:But again, rather than analyzing the situation, you say that "you cannot get energy without a thermal drop" - relying on authority, which isn't making me satisfied anymore than "humans weren't meant to go over 25 mph" does. I say that with a selection process, you can - or I still don't see why not.


Well, I was running on the basis that you accept the 2nd law of thermodynamics, but it sounds like you don't accept arguments based on 2-T. Well, I can give you an answer couched in 'conventional speak', or 'chrismb speak' [i.e. unconventional notions, but come to the same answer], or a blend of both if it is that you might accept 2-T but are not entirely convinced by the text books approach and feel a little more 'out of box' interpretation is required.

So, do you accept 2-T (and, in fact, 1-T also would enter a complete answer), or need some form of persuading?
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Next

Return to The water cooler -- disscusions about anything

Who is online

Users browsing this forum: No registered users and 4 guests