Since you asked... inside a flourescent light tube is argon at a pressure of 3 Torr and mercury at a pressure of 1 Torr (for reference, atmospheric pressure is at about 760 Torr).
A electric discharge creates a plasma such that a fraction of the argon and mercury become ionized (it is a very small fraction). As a result, lots of free electrons are running around. Some of these electrons cause excitation of mercury (either directly or indirectly) which after some radiation transport magic is converted to visible light. Some of the electrons cause further ionization which keeps the discharge around.
For ionization and excitation to occur, the electrons have to be at a high temperature. Argon ionizes at 15eV and to have enough electrons that hot you need electron temperatures over 10,000K (typically 40,000K+). The conversion is roughly 1eV to 11,600K.
The catch is that the electron mass is about 70,000 times less than that of argon. To picture what is going on, electrons are ping-pong balls and argon / mercury are bowling balls. Even if you throw a ping-pong ball really really hard, a bowling ball won't notice it.
As a result, the electrons are able to heat up to very high temperatures. Meanwhile, the glass tube at room temperature keeps the Ar/Hg mix cool. Thus, even though the electron temperatures are high, the heat conduction is incredibly low and the tube feels cold to the touch.
Since this site is interested in computers, these types of plasmas are used in almost every step of semiconductor processing. Because the electron energies are so high, exotic high temperature chemisty can be performed without melting your wafer. And because there are charged species, etchant flux can be electrically manipulated (which is why you have microchips which small features nowadays; look up plasma enhanced anisotropic etching).
As for dangerous experiments, I can think of a few but rather than get sued... I'll leave it to you to think of household devices which have high energy density.
The idea of fusion in sonoluminescence is nothing new. I sat through a talk on it by some computational hydrodynamics experts from Lawrence Livermore National Lab in 1997 at a the Gaseous Electronics Conference in Hawaii (if you really care, you can probably look up the conference proceedings at http://www.aps.org).
The talk was pretty good. Their models were able to explain most of the features reasonably well without having to resort to exotic physics (i.e. quantum electrodynamic weirdness). I mostly remember sitting at this talk because the presenter made a reasonable witty comment (remember, talks like this are usually dry and boring with many audience members nodding off because they are always scheduled after lunch): `Scientists at LLNL have an innately superior understanding of all physics... [pause during the palpable bristling of the audience]... it's either an implosion or an explosion.'
However, the talk did run into a credibility problem when the presenter said the next step was too look for fusion. Several people in the audience correctly pointed out that the temperatures were several orders of magnitude too low. The presenter's response was that the... yes the temperatures are very low compared to fusion. However, a minuscule amount of fusion (think in terms of one or two atoms per microsecond) would occur and thus there would be measureable neutron flux (in theory). However, in practice, the neutron flux would be so low that it would be nearly impossible to distinguish from the background noise.
Without seeing the paper from the ORNL people, I really can't say if they have upped the sophistication or not though.
By the way, the temperatures at the surface of the sun are only ~6K (except in the wispy corona). Not nearly hot enough for fusion... that happens in the core. In fact, it is hotter inside a flourescent light tube (~50K-100K) than at the surface (but the heat conduction is so low that it isn't a safety issue).
About the free electron laser part... it is well beyond present FEL technology. And the technology you describe would have difficulty making anti-protons.
Suppose you want to create electron-positron pairs via counter-streaming FEL lasers. For the physics buffs out there, the reaction would be similar to the Compton backscattering of light off virtual electron-positron pairs (this non-linear vacuum light interaction was demonstrated at SLAC a year or so ago).
The FEL laser would have to operate well into the hard gamma (photon energy exceeding the rest mass of the electron). Current multi-pass FEL technology has been demonstrated up to the ultraviolet (~250 nm I think is the current record). Multi-pass X-ray FELs are near impossible to make because of the difficulty of producing high quality laser cavities for X-rays.
Single pass X-ray FELs (which rely on an electron beam instability instead of a cavity) have been proposed but not yet demonstrated. If I recall correctly, the SASE-FEL program at SLAC to build a $100M dollar X-Ray SASE-FEL (with a 100m long wiggler) did not receive funding.
That is not to say we are incapable of artifically making hard gamma rays. The aforementioned non-linear light interaction obtained the photons for the experiment by Compton scattering of low energy photons off an ultra-relativistic electron beam. But this would probably be pretty inefficent method to try to create antimatter on a large scale (inefficiencies in electron beam acceleration and cross section issues for both the Compton scattering and the non-linear interaction).
The other possibility would be to try to do a multi-photon interaction to create the electron-positron pairs. In this method, an incredible high electric field is created such that it becomes energetically favorable for electrons-positron pairs to form to shield out the field. I think this has also been demonstrated with some of the extremely high intensity chirped pulse amplification lasers. However, the effectiveness isn't anything to write home about yet.
And given the protons mass is 1836 times that of an electron, to create them on a large scale (i.e. micrograms) is not anything I expect to see in the near future.
Thanks... but I'm pretty familiar with skin effect. And I don't need to look it up on Google I may not know much but I do know my electromagnetics.
Actually, skin effect is a lot more complicated that you seem to think it is. The depth of penetration of the current is a function of frequency, of material conductivity and of material thickness. Conductor geometry is also important. For cylindrical geometries (i.e. wires) you'll need to lookup things about "ber" and "bei" functions (see Abramowitz and Stegun to learn about "ber" and "bei"); for planar geometries an exponetial decay is sufficient.
Also, without getting into too much detail, skin effect doesn't really apply to the scenerio I gave in my post. Here is a simple question for you to think about... how does the current know which side of the conductor is the outside? In the Statue of Liberty example, it is not clear that the Statue forms a closed conducting cavity (as posited in my example with the electrical disconnected plates through rust).
Also other posters have pointed out other issues involving sharp metal points and the fact various parts might not be a well grounded whole. Thus, I standby my statement that the Statue of Liberty is not great lightning protection (but probably better than flying a kite into a thunderhead).
That's okay. I frequently butcher spelling and grammar while typing in a frenzy to get precious tasty karma.
Kevin
Re:Your opinion of the X-Men movie?
on
Clear Hard Drive Mods
·
· Score: 4, Interesting
Hmmm... well, the short answer is, unless it is a controlled condition, it is not advisable to summon lightning bolt nearby.
A lightning bolt forms a conductive path from the clouds to the ground. It essentially a capacitor discharging through a short circuit. Given the rapidity of a bolt, the EM radiation covers the whole frequency spectrum. In terms of danger, the lower frequency stuff (this is what transports the charge) is what I would worry about. (When we talk about the low frequency parts, we can use the language of currents and grounds and potentials.)
Ideally, your EE friend would be correct. Being inside a giant metal statue would protect you from the bolt. And assuming the Statue of Liberty is still a good conductor (minimal rust and what not), the Statue will still reflect the high frequency radiation.
However, low frequency concerns make using the Statue of Liberty as a lightbolt protection inadvisable. How well grounded is the Statue? Are all the metal components at the same potential?
For example, suppose you are standing near where two metal panels are abutting. Rust has formed between common edge of the panels. From an electrical standpoint, the two panels are equipotentials electrically connected by a resistance.
When the lightning strikes, current will flow through the panels to ground. You better hope that the current flow doesn't find it easier to jump through you than through the rust to get to ground!
On the other hand, I imagine the New Jersey Parks Department (oddly, the Statue of Libery is in New Jersey... a matter of some annoyance to New Yorkers) has probably attached lighting rods and special cabling to ground to protect against such things.
I can think of other concerns, but this should give enough fodder for your friends to come to a resolution in their dispute.
To get really technical, the mesh deal is a Nyquist related phenomena. So the spacing is half-a-wavelength. The reason why I used 1.25mm to 12.5mm spacing is that I was making an analogy to microwave ovens. The principle harmonic of a microwave is 2.4 GHz, but magnetrons inside such an oven will generate many harmonics of 2.4GHz too (they aren't clean communications type RF sources). So, going significantly below the wavelength is a generally good idea.
This is also true in computers. For example, take an RDRAM bus at 400 MHz (clocked on both edges), a square wave on this bus will consist of domaintly odd harmonics (400, 1200, 2000, 2800... MHz). So, for EMC shielding, 400 MHz shielding is pretty easy to achieve. However, it is a wise idea to actually shield appropriately for 3GHz or so to protect against the EMC issues with the higher harmonics.
Well given what you said I don't think you did EMC testing for the FCC and UL. Your description of how a _Faraday_ cage works is stunningly wrong.
A good conductor reflects incident waves very very efficiently. Very little power is absorbed by the metal itself. If you surround a region with metal, all incident radiation from outside the box is scattered and does not enter the box.
If you want add a transparent window to the box, all you have to do is integrate a metal wire mesh fine enough so that the gaps are much smaller than the wavelengths of the frequencies you want to filter out. So, to filter out all frequencies below 2.4 GHz (lambda = 12.5 cm), you want a mesh much finer spacing on the order of 1.25mm - 1.25cm. (How do you think your microwave oven window works?)
Only if you are talking very low frequencies, would even talking about "goes to ground and out" have any meaningful content (like 60Hz which is essentially the same as DC from any electromagnetics theory standpoing unless your devices are the size of the continential U.S.)
Kevin
P.S. By the way, my Ph.D. background is electromagnetics and I had an office inside a Faraday cage at a former employer.
I am working in the MEMs area these days. So here are some shameless plugs.
Here is an general interest article from the group in which I work with some details oriented towards these types of mesoscopic MEMs.
Here is a neat picture of a Mesoscopic MEMs device (an acceleratometer resting on top the middle part of the "8" in a 1998 penny.
And though my research at Berkeley wasn't MEMS oriented, Berkeley MEMS is pretty active. Here is a link to that.
As the article points out, MEMS are finding applications in cell phones because it is easy to make very small RF filters using inertial effects to provide inductive-like impedences. (In the past, the inductive like parts of a cell-phone filter would either be done with spiral inductors, which are unwieldly or via other microwave circuit voodoo.)
However, beyond cell phones is a grab bag of MEMs applications already at or beyond the prototype stage:
- Car air bag detectors (the above accelerometer)
- Laser gyroscopes
- Projection displays (pixel mirrors arrays)
- Optical fiber switches
- Medical applications (microfluidics, bio-chips,...)
- Remote sensing (minaturized microphones, or in the future, smart dust)
(1) His paper is more a thought experiment. Assuming the author does not have an auto-mechanic background, I doubt he is even aware of the intricate details of exhaust design (I know I'm pretty oblivious to it). However, the standard thermodynamic treatments of such matters don't consider these details either. Nevertheless, the thermodynamic treatment of an idealized engine cycle allows you to put limits on the performance of any engine (regardless how nifty you make the exhaust design).
(2) The extra work is in the form of laser energy. It is not obvious what to do with it in a practical sense. It is "useful" in the theromodynamic sense that the laser energy has a higher equivalent temperature that the engine's cold temperature reservior. Thus, you could theoretical use the laser to perform additional work. How best to do it is difficult to say (a reheat cycle maybe?)
So, I agree, the proposal is not something you can just bolt onto an existing engine design. However, the proposal is interesting as it does give a way to beat the standard Otto cycle (apparently without violating any cherished laws... like Carnot). And Otto cycles are pretty important.
As far a gas turbine is concerned, your guess is a good as mine. If I recall correctly, turbines run a Brayton cycle, not an Otto cycle. I'm sure in theory you could apply the technique to get laser extraction off the exhaust (assuming a suitable working fluid / pressues / cavities) but it is not obvious that it results in an overall improvement to the theoretical Brayton cycle efficiency.
Sources please... he is not describing a standard three level laser system.
I've checked through my laser physics texts (mostly Yariv's "Optical Electrics in Modern Communications", Yariv's "Quantum Electronics", Shen's "Nonlinear Optics", Loudon's "The Quantum Theory of Light") and don't see anything exactly like it... but these texts are oriented towards semiconductor lasers. Also, PRL is a peer-reviewed journal; I would think that if the cooling scheme in the paper is blase that it would be caught. (The abstract upfront brags about how it is a novel technique for inversion... such claims usually get the smack down in a widely read journal like PRL if not accurate.)
However, that is not to say the scheme is original. I can see how you could get it out of the standard laser rate equations and I can see how certain pumping schemes might be superficially similar. So, if you have sources for the technique (using a cool thermal radiation distribution in a enhanced radiation cavity to create a population inversion), I would like the source for my own edification.
and it is a pretty interesting idea. I'm not sure about the practical feasibility of the concept for reasons I'll get into below. But, it shows that quantum effects might be usefully exploited to make better engines and will probably prompt a fair amount of thought and experiment into the matter.
Warning: Ph.D. punditry follows.
Suppose a molecule has three possible states ("a", "b" and "c") with energies E_c, E_b, E_a respectively (E_c is the ground state and E_b is the between E_a and E_c... thanks lameness filter... less than signs could never be useful).
Suppose further, microwave (maser) energy transitions are possible from state "b" to "c". Optical (laser) transitions are possible from "a" to "b".
For lasing to occur, you must have a population inversion... more molecules must be in one of the upper states than in the lower states. However, in a gas at thermal equilibrium, this is usually not the case... the probabiliy of finding a given quantum state in state with energy E is proportional to exp(-E / kT ). Here, k is Boltmann's constant and T is the ambient temperature. At low temperatures, the ground state will be where most of the molecules are.
If the hot exhaust gas is first passed through a maser cavity tuned to the "b"-"c" transition containing a radiation field at the temperature of the cold reservior, the "b" and "c" populations will quickly come to thermal equilibrium with the low temperature radiation field... "b" molecules to preferentially transistion into the ground state (state "c"). However, the "a" population won't be able to come to equilbrium that fast (provided the spontaneous emission rate is sufficiently low and the maser cavity isn't tuned to enhance the transition rate out of "a" state). This net impact of the maser is to create a population inversion between the "a" and "b" states. By passing the non-thermal maser cooled gas into a laser cavity tuned to the "a"-"b" transition, this inversion can be extracted as laser energy. This is the quantum afterburner part.
From a quantum standpoint, nothing is particularly new here. Using rapid cooling of a selective population to create inversion is pretty unique but nothing that can't be explained with the standard laser rate equations.
From a purely statistical mechanics standpoint, the net effect is to extract extra useful work from internal degrees of freedom of the working fluid. Statistical mechanics is not my forte so I can't really say if this is particularly out there.
From a practical standpoint, it might be hard to find gases at engine temperatures and gas pressures where the low spontaneous emission lifetimes necessary to sustain the inversion is possible. My intuition says that collisional de-excitation (high temp and pressures) would wipe out the inversion. Also, the exact scheme discussed in the paper is more complicated... involving passing the gas back and forth through two pistons. I'm pretty sure that materials and a simplified engine design could be made to validate the claims though.
As a thought experiment, though, this shows that it may be possible to improve the efficiency of an Otto engine. (By the way, the paper notes that a Carnot cycle efficiency doesn't get a boost from the technique.)
The site seems to be down. However, last week, I contacted nVidia about this problem on my two dual Ahtlon MP workstations (random hangs when OpenGL is invoked). So the quick answer is you can
Boot your system with following option on your kernel command line: "mem=nopentium"
or
Disable AGP in XFree86 config (i.e. Option "NvAGP" "0" in the "Devices" section).
nVidia clued me into the first approach about a week and a half ago. It made my system completely stable. However, there was still some texture flakiness in some OpenGL applications. Since my workstations are number crunchers (and thus Quake FPS don't matter to me), the latter option eliminated both the stability problems and the texture flakiness (at the expense of some graphics speed).
By the way, nVidia mentioned the same issue exists on Win2K / Athlon boxes.
The comments I've seen all seem oriented to nanoscopic devices. That technology is still in blue sky phase (lots of potential, but nobody really knows what to do with it and it is still not practical for large scale manufacturing). Mesoscopic MEMs (devices on the order of microns in size instead of angstroms) are already used in commerical products. In fact, chances are, you already own a few and didn't even know it.
Here is an article with some details oriented towards mesoscopic MEMs.
Here is a neat picture of a Mesoscopic MEMs device.
(Bias warning: the supervisor of my research group was co-author of this article.)
I saw lots of comments here adovocating HDTV if you have the money. I would like to offer a dissenting view.
Some background... several years ago, I plunged most of my summer job income and left over money from scholarships into my stereo system. It is all high end equipment that you won't find at BestBuy. For example, my CD player has a 2Hz-20kHz frequency response +/- 0.3 dB. (Yes, 0.3 dB --- not 3 dB... and, yes, you can hear the difference a CD player makes if you do a side-by-side comparison with a mainstream CD player. The difference is really obvious on jazz tracks with walking bass lines.)
As I assembled my system, I never upgraded the original 21" monitor (not a really a TV as it doesn't have speakers or a tuner). Why? Because, over 21", you can see the crappy resolution of an NTSC signal even at a distance. So, for me, large TVs are expensive and only serve to remind me how crappy a signal is being transmitted.
(A side note: most people with large TVs wire up their systems in brain dead ways --- cable to vcr to tv, all with coax, such that the TV signal is decoded three times and re-encoded twice. This makes TV viewing that much more painful.)
Now that I am gainfully employed and have a wife who doesn't appreciate the dorm wiring look, I was in the market for an entertainment center. However, I had a dilemma...
Most entertainment centers are designed around a 4:3 aspect ratio big screen TV. However, the FCC has been threatening to go to 16:9 HDTV... so, do I spend over $1K for a piece of furniture that would be obsolete if HDTV becomes commonplace? And if I buy the entertainment center, do I upgrade my 21" screen?
Here is the compromise I came up with for my wife and the reasons for it. We bought a smallish entertainment center which did not require upgrading my screen (after some fun with a drill and jigsaw). Why?
HDTV is not just around the corner.
- Consumer motivation is not there. See above... most people wire their systems to get an even worse quality signal than NTSC... do they care they can't see Jay Leno's pores on the Tonight Show? Do they want to upgrade perfectly good equipment or buy converter boxes? No.
- Cable operators are not required by the FCC to send HDTV signals --- only free space broadcasts. Don't forget, cable has roughly a 70% market penetration. (However, I'm sure the cable company would be happy to rent you converter boxes at a monthly rate if required.)
- Many cable operators are encouraging Digital Cable. (This absolutely sucks... _every_ Digital Cable system I've seen has worse picture quality on average than regular cable for a variety of technical reasons including: original signal is NTSC, original signal is broadcast in a different digital format, cable companies compress the hell out of the original signal assuming customers won't know the difference... my in-laws are now quite pissed about their Digital Cable after I showed them the quite obvious artifacts on their large screen TV screens over the holidays. Perversely, most people assume that since digital artifacts are different that NTSC artifacts that it is some kind of sign of quality.)
- Their is still bickering about standards (modulation formats... the plethora of resolutions... digital "protection" schemes...) Don't buy a technology if it might be dropped like a hot potato in the next few years. HDTV has such a low market penetration that it is not entrenched.
- The stuff is expensive for what you get.
I personally am waiting until the standards settle, the prices drop, the equipment becomes more widely adopted and there is an obvious quality improvement.
I'm not going to pay several thousand dollars to see MPEG artifacts from an over compressed signal blown up life size in my living room. (Watch any shot of the rippling surface of the ocean on Digital Cable to see what I mean.)
Kevin
Re:Forget fiber ... conduits w/ CAT5 and 802.11
on
Wiring A New House?
·
· Score: 3, Insightful
The CD player I bought in 1996 had "fiber optic" outputs. (It was a high end model and I still haven't seen a consumer D/A converter with comparable specs.) And now in 2001, the fiber optic output is still unused. A/V equipment with fiber optic output is nothing new. Even before then, the security system at the electronics store I worked at in high school (circa 1993) consisted of a "fiber optic" loop. A red-LED would send a pulse every through the cable every second (you could see it looking at the fiber end-on). If the cable was broken, the pulse wouldn't be received and thus you would know that somebody was trying to walk off with a floor model.
However, in both these applications the type of cabling and what not is not what people usually think of when they are talk about fiber optics (hence the quotation marks). In both applications, you are only moving the signal a couple of feet and the signal has very low bandwidth. As a result, you can get away with a lot of slop and do most of the implementation with plastic fibers, normal LEDs and no fancy couplers.
If you want to talk moderate bandwidth computer use (1 Mb/s - 100 Mb/s), I can't think of any advantages of fiber over CAT5 and wireless (except possibly EMI). All I can see are drawbacks (price, fickle connectors, comparatively little support... ).
If you are talking high bandwidth (over 1 Gb/s), then you want to consider fiber. However, I would like to know what a home user is doing that needs 1 Gb/s. Someday, users might be running multi-Gb/s home networks. When that happens, I don't expect a DVD optic patch cable to be up to snuff. Furthermore, I don't expect multi-Gb/s home networks to happen before this guy sells his home anyways (5 years is the rule of thumb used by real estate agents).
So why spend the money laying fiber when:
- CAT5 and wireless are good enough for the foreseeable future.
- Standards for fiber home networks are practically non-existant such that picking the right underlying technology becomes a gamble.
- The home's resale value is better enhanced with conduits.
Kevin
Forget fiber ... conduits w/ CAT5 and 802.11
on
Wiring A New House?
·
· Score: 5, Insightful
I just bought a house and instead of wiring it up, I just use 802.11b for the bedroom computers. However, my 802.11b access point / firewall also has a switched hub, so my workstations are hooked up with CAT5. This is more than fast enough for any kind of internet connection you are likely to get in the near future.
Don't listen to the... ahem... "experts" telling you to install fiber. In my day job, I work in research on fiber optics technology (mostly for 40 Gb/s+ DWDM long haul and metro networks). Fiber equipment that I am familar with is not made for the consumer market.
Would you even know what types of fiber to buy? (multi-mode / single-mode, C-band / L-band / Extended-L band,...) Or what kind of connectors you would need? Or do you have the access to the equipment necessary to splice fibers (it's not cheap to do it right)? Do you know what kinds of equipment to attach to the end of the fiber (modulators, switches, splitters, NICs...)? And exactly what are you going to hook up that requires fiber's speed?
If you are worried about an upgrade path, the smart thing to do is install conduits. When fiber goes to the consumer market, you will be ready.
I wonder about the scientific literacy around here when this comment is at a +5.
Most notably, particles don't helix around electric field lines. Particles helix around magnetic fields. This is high school physics people!
Now if I substitute electric for magnetic, the description makes a bit more sense. With this substitution, it sounds like the poster is describing a variant of a magnetic mirror machine.
This is so old it seems new.
(However, the variation might be original... but the description is too vague to tell.)
At Livermore in the 1960s, a giant mirror machine was built and mothballed on the opening day (funding had dried up when the fusion community began its quest for mythical Tokamak fusion reactor). When I last saw the mirror building (three years ago or so), the equipment had been thoroughly picked over by experimenters.
Simple magnetic mirrors always had problems with the mirror loss-cone. Magnetic mirrors work really well except for particles travelling near normal to the current rings (the loss-cone).
Given that particles have to collide for fusion to occur, some collisions end up scattering particles into the loss-cone and thus, plasma confinement is not that great.
Also, the velocity space distribution functions resulting from a loss cone leads to a whole class of plasma instabilities (surprisingly known as "loss-cone instabilities").
More complicated mirror devices have been designed which allieviate some of these problems but they have not received much attention from funding organizations.
Please don't waste other people's time talking authoritatively ("Classical EM shows...") if you don't in fact know what you are talking about.
IANAP, so I am talking out my ass here, but it seems to me that the interior surface of the container of a reaction might just somehow be able to more directly collect energy, similar to the way solar-panels collect light. Of course, I have no idea what kind material might be used to accomplish this. How do solar-panels work? Silicon? Is it just dumb luck that the elements of a solar panel happen to convert light to energy, or is it a man-made composite, built specifically for that purpose?
Yes, you are talking out of your ass.
Note: I did my Ph.D work in plasma physics but now I work in quantum and optical electronics. I am probably one of the better qualified people here to answer your question.
Conventional fusion reactors fuse deuterium and tritium. Or, if you breed tritium from a lithium blanket surrounding the reaction, you can do fusion using deutrium-deutrium full burn (this is tougher to do than D-T reactions).
However, the by-products of D-T and D-D fusion are mostly high energy neutrons (and some gamma rays and neutrinos and alpha particles...). High energy neutrons are not easy to convert into electricity because neutrons are not charged. In fact, the neutron flux of a large fusion reactor would be deadly and thus a fusion reactor needs to be heavily shielded while operating. (Watch "Chain Reaction" and laugh as Keanu and his advisor walk around the operating reactor after it stabilizes.)
A typical approach for the conversion consists of letting your neutron flux heat a block of lead (or other material) and then running a standard steam cycle.
This sucks on many levels.
First, you end up throwing away much of your power from the inefficiency of the steam cycle. That is the theoretical thermodynamic efficiency of a fusion reactor which can somehow do direct conversion is effectively 100% (hot reservior at millions of degrees, cold reservior at room temp) while a steam cycle is limited by how hot you can heat your materials (hot reservior at thousands of degrees, cold reservior at room temp).
Second, the neutron flux will activate (i.e. make weakly radioactive) the walls of your reactor and steadily degrade the structural integrity of your vessel. As a result, current estimates are that the core of a fusion power plant will need to be replaced every couple of years (which makes energy providers frown... fusion reactor vessels will likely not be cheap).
Other approaches are to use fusion reactors as breaders for fission plants (i.e. use fusion neutrons to enrich fission reactor fuel). Many estimates already put current reactor technology beyond breakeven for this type of design. However:
- Design is not politically feasible (some alternative fission fuel cycles might be possible). Fission suffers from NIMBY and fusion-fission breeders have a massive proliferation risk.
- Fuel for fission reactors is not particularly rare and running a fusion plant will likely not be cheap. Thus, economically, currently there is no compelling reason.
Solar cells on the other hand rely on photons exciting electron-hole pairs in a semiconductor. The light from the sun partially consists of photons in the visible and near infrared range which are suitable for conversion by a solar cell.
You might be wondering why fusion reactions produce high energy neutrons and gamma rays and other generally nasty things while our sun shines a whole lot of light. You should remember that the sun is big and the products of a fusion in the sun take hundreds of thousands of years to reach the surface (random walk... collisions...). By that time, all that the sun radiates is a near perfect blackbody radiation spectrum.
11 MW to 10 Mpeople is a bit off. You probably meant 11 GW per 10 Mpeople.
At 11 MW / 10 Mpeople gives 1.1 Watts power consumed per person. Hell, power consumption in the 1800s per person was probably higher through candles alone.
It is more like 1 kW per person (your mileage may vary outside the US).
Thus, the ~40 MW power from a square mile of solar cells could supply a small town of 40,000 people. A conventional power plant (gas, nuclear, coal,...) filling the same area could provide power for a small country (several million poeple).
Good point. Making an oscillator wouldn't be too hard but I wouldn't know the efficiency loss off the top of my head.
This is why I took the generous 30 W/m2 estimate and hacked off another 50% to account for real world engineering issues such as:
- Storage efficiency
- Transmission efficiency (I would consider the DC-AC conversion part of the transmission issue).
- Grazing incidence.
-...
Basically, the point is, that per unit area, most of the time solar cells are not a very attractive technology. Per unit cost, most of the time solar cells are not a very attractive technology.
To try to slap some realism into the idealism floating around here: People keep talking about new fancy solar cells with a 30%+ efficiency but they seem to be forgetting that solar cells have to be a large commodity industry if it to compete with other power sources. A commodity industry means the cells are easy to manufacture, cheap to manafacture (in both money and energy), high volume and low profit margins per unit. And by easy and cheap to manufacture, I mean you have to be able to produce hundreds of square miles of PV material. Fine nanostructured materials with complicated anti-reflective layers with bandgap engineered hetero-junction quaternary compounds doped with exotic rare earths is not going to be a commodity product for the forseeable future. Boring amorphorous silicon based cells with its lower efficiency have a much better chance (which is what I would bet is being used in these third world countries).
Environmentally solar cells may have some advantages but solar cell manufacturing and disposal needs to be accounted for too.
If you think that solar power (as generated by PV solar cells) can satisfy a large part of the power load of an industrialized country, you are dreaming. (If you give me a sqaure mile to be used exclusively for power generation, I can give you giga-watts conventional instead of mega-watts solar).
If you think that solar cells should be used to reduce demand during heat waves and can decentralize power generation somewhat, you are thinking more realistically. Such uses are already supported with tax breaks for deploying solar cells (which are slated to be expanded).
A while ago, I calculated how soon deploying solar cells on the roof of a house would pay for itself. The short answer I would cost several thousand dollars and it would take roughly 10 to 20 years realistically (as energy costs change, this figure bobs around too). And of course, it all depends on where you live, what you doing with the solar energy (i.e. solar cell PVs are not the only use of radiant flux) and whether you can sell power onto the grid in your locale.
If you need to make home improvements, you can get a much higher return doing other things (the afforementioned tax breaks help). If you are an ardent supporter of the technology, you would be better off to invest the money into a solar cell company conducting research to bring down the manufacturing cost to make the technology competitive.
If you more interested in large centralized power plant solar installations, it would probably be more cost effective and environmentally friendly to use a helio-concentratator instead of PVs. And such a plant would really only be feasible in certain areas anyways (I don't expect to see helio-concentrators anytime soon in say Seattle).
So the short answer:
Solar cells will be used when the technology is competitive. Most of the time it is not.
The best solar cells generally have about 30% efficiency, relative to the total flux of sunlight hitting the earth in the given area. To output 100 MW requires about 1 square mile of cells ---------
Interesting maths there.
You get ~1kW/m^2 of sunlight, so that's 2560000kW of raw sunlight, which at 30% efficiency is:
2560000kW * 0.3 = 768 000kW == 768MW ---------
Point of view from an applied physicist:
Sunlight from the earth is 1.2 kW/m^2 (higher than your estimate). However, 30% efficiency is a pretty big overestimate of how much you can get from a presently economical solar cell. There are solar cells approaching 30%, but these tend to be made out of more exotic materials and are not as easily manufactured.
Starting from an incident flux of 1.2 kW/m2:
Half the time it is night. So averaged over a day gives: 600 W/m2.
But in most places, not every day is sunny. So, hack another half off that: 300 W/m2.
Also, the efficiency of a solar cell is strongly dependent on the angle of incidence of the sunlight. Assuming a low cost installation where the cells aren't pivoting (expensive and prone to break down) to catch the rays gives a loss in available flux. Using a generous a cos^2 depenenced takes us down to: 150 W/m2
(Note: It is probably much worse as the cos^2 only accounts for effective reduction in solar cell cross-section area as the sun rises and sets. For grazing incidence light most of the sunlight will reflect off the solar cell. And yes, exotic solar cells have been designed to reduce this, but this adds to cost and manufacturing difficulty.)
Now, apply a realistic 20% solar cell efficiency: 30 W/m2.
Thus, a typical solar cell can be expected to yield on average (a generous estimate):
30 W/m2.
Of course, this ignores the efficiency of any storage system you might have if you want to make use of the power generated when the sun is not directly overhead on a clear day. So to try to get a more realistic feel, hack off another half to account for efficiency and grazing reflection:
15 W/m2.
So, a square mile array of solar cells could make an average contribution of:
1600 m * 1600 m * 15 W/m2= 38,400,000 W
This is one-twenth the value of the previous poster and a bit closer to the original post, but 38.4 MW can power a fair number of homes.
However, as the previous poster pointed out, in most cities, you could get more bang for your real-estate via other means.
However, when the sun is directly overhead, on a sunny day, you will get a peak performance of roughly 240 W/m2.
Since this is the time when power is most needed anyways, this points to solar cells being used to offset peak power demand when everybody's air conditioning kicks on simultaneously. I don't expect solar cells to be the primary source of power anytime soon except in special situations.
So, how do you broadcast a single photon everywhere? That's the key. If you send the message everywhere, you are obviously not sending single photons. If you can send a single photon reliably from point a to point b, you have figured out how to make sure it doesn't get lost in between.
Though it is too late for this response to make any difference, I'll waste my breath.
Quantum mechanically, a photon is an eigenmode of Maxwell's equations for the system under consideration. A photon is commonly thought of as a localized particle of light. It is not. It is most analogous to a wave (a plane wave is an eigenmode of free space; in a complicated system, the eigenmodes are less straightforeward).
A photon is not localized. A superposition of photons may be localized. Such a superposition is best called a wave packet; it is not strictly a photon though.
Confusion over this is why very few people can actually make sense of quantum mechanics, especially if explained without mathematics (all that non-sensical jibber-jabber about wave-particle duality with bad philosophy thrown in for good measure).
At no point in any quantum mechanical formalism I've seen (Hamiltonian-based, Lagrangian-based, Heisenberg matrix mechanics, Schrodinger wave mechanics, Feynman path integrals, relativistic field theory,... ) are particles fundamental.(Bohmian quantum mechanics is a quasi-exception.)
Quantum mechanics is about waves (or more precisely eigenmodes of the Hamiltonian). Superpositions of waves makes particle-like excitations.
So, you can send a single photon everywhere. For a quick example, think of the two slit experiment. It still works when the photons pass through the system one at a time (this has been experiementally demonstrated). Thus, one photon passes through both slits and interferes with itself on the other side.
If photons were localized, as you seem to think, the two slit experiement would fail.
However, producing a single photon is not simple. Devices like lasers will produce a spectrum of photons with a certain narrow energy spread and a certain narrow angular spread. Such superposition of photons will be localized in space and are what people often call photons or particles of light. The probability of detecting such a wave packet in two widely separated places is negligible.
However, other devices (like say an antenna) produce wave packets which are not localized.
And in response to another post:
The reason that quantam[sic] encryption isn't used everywehere, is that it's so darn hard to detect the spin of single photons.
Detecting the spin a stream of photons is much easier than you think. Photon spin and photon polarization are closely related (photon spin is a different set of basis vectors to express photon polarization). Detecting photon polarization is trivial (sunglasses anyone?). Detecting a single photon's polarization with a bit error rate low enough to be usable over long distances is more challenging but not impossible (especially if you are just doing key exchange).
Yes, I have a Ph.D. and quantum electronics is my day job.
Slashdot Mirror
Since you asked ... inside a flourescent light tube is argon at a pressure of 3 Torr and mercury at a pressure of 1 Torr (for reference, atmospheric pressure is at about 760 Torr).
... I'll leave it to you to think of household devices which have high energy density.
A electric discharge creates a plasma such that a fraction of the argon and mercury become ionized (it is a very small fraction). As a result, lots of free electrons are running around. Some of these electrons cause excitation of mercury (either directly or indirectly) which after some radiation transport magic is converted to visible light. Some of the electrons cause further ionization which keeps the discharge around.
For ionization and excitation to occur, the electrons have to be at a high temperature. Argon ionizes at 15eV and to have enough electrons that hot you need electron temperatures over 10,000K (typically 40,000K+). The conversion is roughly 1eV to 11,600K.
The catch is that the electron mass is about 70,000 times less than that of argon. To picture what is going on, electrons are ping-pong balls and argon / mercury are bowling balls. Even if you throw a ping-pong ball really really hard, a bowling ball won't notice it.
As a result, the electrons are able to heat up to very high temperatures. Meanwhile, the glass tube at room temperature keeps the Ar/Hg mix cool. Thus, even though the electron temperatures are high, the heat conduction is incredibly low and the tube feels cold to the touch.
Since this site is interested in computers, these types of plasmas are used in almost every step of semiconductor processing. Because the electron energies are so high, exotic high temperature chemisty can be performed without melting your wafer. And because there are charged species, etchant flux can be electrically manipulated (which is why you have microchips which small features nowadays; look up plasma enhanced anisotropic etching).
As for dangerous experiments, I can think of a few but rather than get sued
Kevin
The idea of fusion in sonoluminescence is nothing new. I sat through a talk on it by some computational hydrodynamics experts from Lawrence Livermore National Lab in 1997 at a the Gaseous Electronics Conference in Hawaii (if you really care, you can probably look up the conference proceedings at http://www.aps.org).
... [pause during the palpable bristling of the audience] ... it's either an implosion or an explosion.'
... yes the temperatures are very low compared to fusion. However, a minuscule amount of fusion (think in terms of one or two atoms per microsecond) would occur and thus there would be measureable neutron flux (in theory). However, in practice, the neutron flux would be so low that it would be nearly impossible to distinguish from the background noise.
... that happens in the core. In fact, it is hotter inside a flourescent light tube (~50K-100K) than at the surface (but the heat conduction is so low that it isn't a safety issue).
The talk was pretty good. Their models were able to explain most of the features reasonably well without having to resort to exotic physics (i.e. quantum electrodynamic weirdness). I mostly remember sitting at this talk because the presenter made a reasonable witty comment (remember, talks like this are usually dry and boring with many audience members nodding off because they are always scheduled after lunch): `Scientists at LLNL have an innately superior understanding of all physics
However, the talk did run into a credibility problem when the presenter said the next step was too look for fusion. Several people in the audience correctly pointed out that the temperatures were several orders of magnitude too low. The presenter's response was that the
Without seeing the paper from the ORNL people, I really can't say if they have upped the sophistication or not though.
By the way, the temperatures at the surface of the sun are only ~6K (except in the wispy corona). Not nearly hot enough for fusion
Kevin
About the free electron laser part ... it is well beyond present FEL technology. And the technology you describe would have difficulty making anti-protons.
Suppose you want to create electron-positron pairs via counter-streaming FEL lasers. For the physics buffs out there, the reaction would be similar to the Compton backscattering of light off virtual electron-positron pairs (this non-linear vacuum light interaction was demonstrated at SLAC a year or so ago).
The FEL laser would have to operate well into the hard gamma (photon energy exceeding the rest mass of the electron). Current multi-pass FEL technology has been demonstrated up to the ultraviolet (~250 nm I think is the current record). Multi-pass X-ray FELs are near impossible to make because of the difficulty of producing high quality laser cavities for X-rays.
Single pass X-ray FELs (which rely on an electron beam instability instead of a cavity) have been proposed but not yet demonstrated. If I recall correctly, the SASE-FEL program at SLAC to build a $100M dollar X-Ray SASE-FEL (with a 100m long wiggler) did not receive funding.
That is not to say we are incapable of artifically making hard gamma rays. The aforementioned non-linear light interaction obtained the photons for the experiment by Compton scattering of low energy photons off an ultra-relativistic electron beam. But this would probably be pretty inefficent method to try to create antimatter on a large scale (inefficiencies in electron beam acceleration and cross section issues for both the Compton scattering and the non-linear interaction).
The other possibility would be to try to do a multi-photon interaction to create the electron-positron pairs. In this method, an incredible high electric field is created such that it becomes energetically favorable for electrons-positron pairs to form to shield out the field. I think this has also been demonstrated with some of the extremely high intensity chirped pulse amplification lasers. However, the effectiveness isn't anything to write home about yet.
And given the protons mass is 1836 times that of an electron, to create them on a large scale (i.e. micrograms) is not anything I expect to see in the near future.
Kevin
Thanks ... but I'm pretty familiar with skin effect. And I don't need to look it up on Google
... how does the current know which side of the conductor is the outside? In the Statue of Liberty example, it is not clear that the Statue forms a closed conducting cavity (as posited in my example with the electrical disconnected plates through rust).
I may not know much but I do know my electromagnetics.
Actually, skin effect is a lot more complicated that you seem to think it is. The depth of penetration of the current is a function of frequency, of material conductivity and of material thickness. Conductor geometry is also important. For cylindrical geometries (i.e. wires) you'll need to lookup things about "ber" and "bei" functions (see Abramowitz and Stegun to learn about "ber" and "bei"); for planar geometries an exponetial decay is sufficient.
Also, without getting into too much detail, skin effect doesn't really apply to the scenerio I gave in my post. Here is a simple question for you to think about
Also other posters have pointed out other issues involving sharp metal points and the fact various parts might not be a well grounded whole. Thus, I standby my statement that the Statue of Liberty is not great lightning protection (but probably better than flying a kite into a thunderhead).
But I digress,
Kevin
That's okay. I frequently butcher spelling and grammar while typing in a frenzy to get precious tasty karma.
Kevin
Hmmm ... well, the short answer is, unless it is a controlled condition, it is not advisable to summon lightning bolt nearby.
... a matter of some annoyance to New Yorkers) has probably attached lighting rods and special cabling to ground to protect against such things.
A lightning bolt forms a conductive path from the clouds to the ground. It essentially a capacitor discharging through a short circuit. Given the rapidity of a bolt, the EM radiation covers the whole frequency spectrum. In terms of danger, the lower frequency stuff (this is what transports the charge) is what I would worry about. (When we talk about the low frequency parts, we can use the language of currents and grounds and potentials.)
Ideally, your EE friend would be correct. Being inside a giant metal statue would protect you from the bolt. And assuming the Statue of Liberty is still a good conductor (minimal rust and what not), the Statue will still reflect the high frequency radiation.
However, low frequency concerns make using the Statue of Liberty as a lightbolt protection inadvisable. How well grounded is the Statue? Are all the metal components at the same potential?
For example, suppose you are standing near where two metal panels are abutting. Rust has formed between common edge of the panels. From an electrical standpoint, the two panels are equipotentials electrically connected by a resistance.
When the lightning strikes, current will flow through the panels to ground. You better hope that the current flow doesn't find it easier to jump through you than through the rust to get to ground!
On the other hand, I imagine the New Jersey Parks Department (oddly, the Statue of Libery is in New Jersey
I can think of other concerns, but this should give enough fodder for your friends to come to a resolution in their dispute.
Kevin
To get really technical, the mesh deal is a Nyquist related phenomena. So the spacing is half-a-wavelength. The reason why I used 1.25mm to 12.5mm spacing is that I was making an analogy to microwave ovens. The principle harmonic of a microwave is 2.4 GHz, but magnetrons inside such an oven will generate many harmonics of 2.4GHz too (they aren't clean communications type RF sources). So, going significantly below the wavelength is a generally good idea.
... MHz). So, for EMC shielding, 400 MHz shielding is pretty easy to achieve. However, it is a wise idea to actually shield appropriately for 3GHz or so to protect against the EMC issues with the higher harmonics.
This is also true in computers. For example, take an RDRAM bus at 400 MHz (clocked on both edges), a square wave on this bus will consist of domaintly odd harmonics (400, 1200, 2000, 2800
Kevin
Well given what you said I don't think you did EMC testing for the FCC and UL. Your description of how a _Faraday_ cage works is stunningly wrong.
A good conductor reflects incident waves very very efficiently. Very little power is absorbed by the metal itself. If you surround a region with metal, all incident radiation from outside the box is scattered and does not enter the box.
If you want add a transparent window to the box, all you have to do is integrate a metal wire mesh fine enough so that the gaps are much smaller than the wavelengths of the frequencies you want to filter out. So, to filter out all frequencies below 2.4 GHz (lambda = 12.5 cm), you want a mesh much finer spacing on the order of 1.25mm - 1.25cm. (How do you think your microwave oven window works?)
Only if you are talking very low frequencies, would even talking about "goes to ground and out" have any meaningful content (like 60Hz which is essentially the same as DC from any electromagnetics theory standpoing unless your devices are the size of the continential U.S.)
Kevin
P.S. By the way, my Ph.D. background is electromagnetics and I had an office inside a Faraday cage at a former employer.
I am working in the MEMs area these days. So here are some shameless plugs.
...)
Here
is an general interest article from the group in which I work with some details oriented towards these types of mesoscopic MEMs.
Here
is a neat picture of a Mesoscopic MEMs device (an acceleratometer resting on top the middle part of the "8" in a 1998 penny.
And though my research at Berkeley wasn't MEMS oriented, Berkeley MEMS is pretty active. Here is a link to that.
As the article points out, MEMS are finding applications in cell phones because it is easy to make very small RF filters using inertial effects to provide inductive-like impedences. (In the past, the inductive like parts of a cell-phone filter would either be done with spiral inductors, which are unwieldly or via other microwave circuit voodoo.)
However, beyond cell phones is a grab bag of MEMs applications already at or beyond the prototype stage:
- Car air bag detectors (the above accelerometer)
- Laser gyroscopes
- Projection displays (pixel mirrors arrays)
- Optical fiber switches
- Medical applications (microfluidics, bio-chips,
- Remote sensing (minaturized microphones, or in the future, smart dust)
Enjoy
Kevin
Two things should be noted about the proposal.
... like Carnot). And Otto cycles are pretty important.
(1) His paper is more a thought experiment. Assuming the author does not have an auto-mechanic background, I doubt he is even aware of the intricate details of exhaust design (I know I'm pretty oblivious to it). However, the standard thermodynamic treatments of such matters don't consider these details either. Nevertheless, the thermodynamic treatment of an idealized engine cycle allows you to put limits on the performance of any engine (regardless how nifty you make the exhaust design).
(2) The extra work is in the form of laser energy. It is not obvious what to do with it in a practical sense. It is "useful" in the theromodynamic sense that the laser energy has a higher equivalent temperature that the engine's cold temperature reservior. Thus, you could theoretical use the laser to perform additional work. How best to do it is difficult to say (a reheat cycle maybe?)
So, I agree, the proposal is not something you can just bolt onto an existing engine design. However, the proposal is interesting as it does give a way to beat the standard Otto cycle (apparently without violating any cherished laws
As far a gas turbine is concerned, your guess is a good as mine. If I recall correctly, turbines run a Brayton cycle, not an Otto cycle. I'm sure in theory you could apply the technique to get laser extraction off the exhaust (assuming a suitable working fluid / pressues / cavities) but it is not obvious that it results in an overall improvement to the theoretical Brayton cycle efficiency.
Kevin
Thanks. Since I normally do plain text submissions, I figured it was Slashdot's text to html filter which was eating them.
Kevin
Sources please ... he is not describing a standard three level laser system.
... but these texts are oriented towards semiconductor lasers. Also, PRL is a peer-reviewed journal; I would think that if the cooling scheme in the paper is blase that it would be caught. (The abstract upfront brags about how it is a novel technique for inversion ... such claims usually get the smack down in a widely read journal like PRL if not accurate.)
I've checked through my laser physics texts (mostly Yariv's "Optical Electrics in Modern Communications", Yariv's "Quantum Electronics", Shen's "Nonlinear Optics", Loudon's "The Quantum Theory of Light") and don't see anything exactly like it
However, that is not to say the scheme is original. I can see how you could get it out of the standard laser rate equations and I can see how certain pumping schemes might be superficially similar. So, if you have sources for the technique (using a cool thermal radiation distribution in a enhanced radiation cavity to create a population inversion), I would like the source for my own edification.
and it is a pretty interesting idea. I'm not sure about the practical feasibility of the concept for reasons I'll get into below. But, it shows that quantum effects might be usefully exploited to make better engines and will probably prompt a fair amount of thought and experiment into the matter.
... thanks lameness filter ... less than signs could never be useful).
... more molecules must be in one of the upper states than in the lower states. However, in a gas at thermal equilibrium, this is usually not the case ... the probabiliy of finding a given quantum state in state with energy E is proportional to exp(-E / kT ). Here, k is Boltmann's constant and T is the ambient temperature. At low temperatures, the ground state will be where most of the molecules are.
... "b" molecules to preferentially transistion into the ground state (state "c"). However, the "a" population won't be able to come to equilbrium that fast (provided the spontaneous emission rate is sufficiently low and the maser cavity isn't tuned to enhance the transition rate out of "a" state). This net impact of the maser is to create a population inversion between the "a" and "b" states. By passing the non-thermal maser cooled gas into a laser cavity tuned to the "a"-"b" transition, this inversion can be extracted as laser energy. This is the quantum afterburner part.
... involving passing the gas back and forth through two pistons. I'm pretty sure that materials and a simplified engine design could be made to validate the claims though.
Warning: Ph.D. punditry follows.
Suppose a molecule has three possible states ("a", "b" and "c") with energies E_c, E_b, E_a respectively (E_c is the ground state and E_b is the between E_a and E_c
Suppose further, microwave (maser) energy transitions are possible from state "b" to "c". Optical (laser) transitions are possible from "a" to "b".
For lasing to occur, you must have a population inversion
If the hot exhaust gas is first passed through a maser cavity tuned to the "b"-"c" transition containing a radiation field at the temperature of the cold reservior, the "b" and "c" populations will quickly come to thermal equilibrium with the low temperature radiation field
From a quantum standpoint, nothing is particularly new here. Using rapid cooling of a selective population to create inversion is pretty unique but nothing that can't be explained with the standard laser rate equations.
From a purely statistical mechanics standpoint, the net effect is to extract extra useful work from internal degrees of freedom of the working fluid. Statistical mechanics is not my forte so I can't really say if this is particularly out there.
From a practical standpoint, it might be hard to find gases at engine temperatures and gas pressures where the low spontaneous emission lifetimes necessary to sustain the inversion is possible. My intuition says that collisional de-excitation (high temp and pressures) would wipe out the inversion. Also, the exact scheme discussed in the paper is more complicated
As a thought experiment, though, this shows that it may be possible to improve the efficiency of an Otto engine. (By the way, the paper notes that a Carnot cycle efficiency doesn't get a boost from the technique.)
Kevin
The site seems to be down. However, last week, I contacted nVidia about this problem on my two dual Ahtlon MP workstations (random hangs when OpenGL is invoked). So the quick answer is you can
Boot your system with following option on your kernel command line: "mem=nopentium"
or
Disable AGP in XFree86 config (i.e. Option "NvAGP" "0" in the "Devices" section).
nVidia clued me into the first approach about a week and a half ago. It made my system completely stable. However, there was still some texture flakiness in some OpenGL applications. Since my workstations are number crunchers (and thus Quake FPS don't matter to me), the latter option eliminated both the stability problems and the texture flakiness (at the expense of some graphics speed).
By the way, nVidia mentioned the same issue exists on Win2K / Athlon boxes.
Enjoy,
Kevin
The comments I've seen all seem oriented to nanoscopic devices. That technology is still in blue sky phase (lots of potential, but nobody really knows what to do with it and it is still not practical for large scale manufacturing). Mesoscopic MEMs (devices on the order of microns in size instead of angstroms) are already used in commerical products. In fact, chances are, you already own a few and didn't even know it.
Here is an article with some details oriented towards mesoscopic MEMs.
Here is a neat picture of a Mesoscopic MEMs device.
(Bias warning: the supervisor of my research group was co-author of this article.)
Kevin
Hello,
... several years ago, I plunged most of my summer job income and left over money from scholarships into my stereo system. It is all high end equipment that you won't find at BestBuy. For example, my CD player has a 2Hz-20kHz frequency response +/- 0.3 dB. (Yes, 0.3 dB --- not 3 dB ... and, yes, you can hear the difference a CD player makes if you do a side-by-side comparison with a mainstream CD player. The difference is really obvious on jazz tracks with walking bass lines.)
...
... so, do I spend over $1K for a piece of furniture that would be obsolete if HDTV becomes commonplace? And if I buy the entertainment center, do I upgrade my 21" screen?
... most people wire their systems to get an even worse quality signal than NTSC ... do they care they can't see Jay Leno's pores on the Tonight Show? Do they want to upgrade perfectly good equipment or buy converter boxes? No.
... _every_ Digital Cable system I've seen has worse picture quality on average than regular cable for a variety of technical reasons including: original signal is NTSC, original signal is broadcast in a different digital format, cable companies compress the hell out of the original signal assuming customers won't know the difference ... my in-laws are now quite pissed about their Digital Cable after I showed them the quite obvious artifacts on their large screen TV screens over the holidays. Perversely, most people assume that since digital artifacts are different that NTSC artifacts that it is some kind of sign of quality.)
... the plethora of resolutions ... digital "protection" schemes ...) Don't buy a technology if it might be dropped like a hot potato in the next few years. HDTV has such a low market penetration that it is not entrenched.
I saw lots of comments here adovocating HDTV if you have the money. I would like to offer a dissenting view.
Some background
As I assembled my system, I never upgraded the original 21" monitor (not a really a TV as it doesn't have speakers or a tuner). Why? Because, over 21", you can see the crappy resolution of an NTSC signal even at a distance. So, for me, large TVs are expensive and only serve to remind me how crappy a signal is being transmitted.
(A side note: most people with large TVs wire up their systems in brain dead ways --- cable to vcr to tv, all with coax, such that the TV signal is decoded three times and re-encoded twice. This makes TV viewing that much more painful.)
Now that I am gainfully employed and have a wife who doesn't appreciate the dorm wiring look, I was in the market for an entertainment center. However, I had a dilemma
Most entertainment centers are designed around a 4:3 aspect ratio big screen TV. However, the FCC has been threatening to go to 16:9 HDTV
Here is the compromise I came up with for my wife and the reasons for it. We bought a smallish entertainment center which did not require upgrading my screen (after some fun with a drill and jigsaw). Why?
HDTV is not just around the corner.
- Consumer motivation is not there. See above
- Cable operators are not required by the FCC to send HDTV signals --- only free space broadcasts. Don't forget, cable has roughly a 70% market penetration. (However, I'm sure the cable company would be happy to rent you converter boxes at a monthly rate if required.)
- Many cable operators are encouraging Digital Cable. (This absolutely sucks
- Their is still bickering about standards (modulation formats
- The stuff is expensive for what you get.
I personally am waiting until the standards settle, the prices drop, the equipment becomes more widely adopted and there is an obvious quality improvement.
I'm not going to pay several thousand dollars to see MPEG artifacts from an over compressed signal blown up life size in my living room. (Watch any shot of the rippling surface of the ocean on Digital Cable to see what I mean.)
Kevin
The CD player I bought in 1996 had "fiber optic" outputs. (It was a high end model and I still haven't seen a consumer D/A converter with comparable specs.) And now in 2001, the fiber optic output is still unused. A/V equipment with fiber optic output is nothing new. Even before then, the security system at the electronics store I worked at in high school (circa 1993) consisted of a "fiber optic" loop. A red-LED would send a pulse every through the cable every second (you could see it looking at the fiber end-on). If the cable was broken, the pulse wouldn't be received and thus you would know that somebody was trying to walk off with a floor model.
... ).
However, in both these applications the type of cabling and what not is not what people usually think of when they are talk about fiber optics (hence the quotation marks). In both applications, you are only moving the signal a couple of feet and the signal has very low bandwidth. As a result, you can get away with a lot of slop and do most of the implementation with plastic fibers, normal LEDs and no fancy couplers.
If you want to talk moderate bandwidth computer use (1 Mb/s - 100 Mb/s), I can't think of any advantages of fiber over CAT5 and wireless (except possibly EMI). All I can see are drawbacks (price, fickle connectors, comparatively little support
If you are talking high bandwidth (over 1 Gb/s), then you want to consider fiber. However, I would like to know what a home user is doing that needs 1 Gb/s. Someday, users might be running multi-Gb/s home networks. When that happens, I don't expect a DVD optic patch cable to be up to snuff. Furthermore, I don't expect multi-Gb/s home networks to happen before this guy sells his home anyways (5 years is the rule of thumb used by real estate agents).
So why spend the money laying fiber when:
- CAT5 and wireless are good enough for the foreseeable future.
- Standards for fiber home networks are practically non-existant such that picking the right underlying technology becomes a gamble.
- The home's resale value is better enhanced with conduits.
Kevin
I just bought a house and instead of wiring it up, I just use 802.11b for the bedroom computers. However, my 802.11b access point / firewall also has a switched hub, so my workstations are hooked up with CAT5. This is more than fast enough for any kind of internet connection you are likely to get in the near future.
... ahem ... "experts" telling you to install fiber. In my day job, I work in research on fiber optics technology (mostly for 40 Gb/s+ DWDM long haul and metro networks). Fiber equipment that I am familar with is not made for the consumer market.
...) Or what kind of connectors you would need? Or do you have the access to the equipment necessary to splice fibers (it's not cheap to do it right)? Do you know what kinds of equipment to attach to the end of the fiber (modulators, switches, splitters, NICs ...)? And exactly what are you going to hook up that requires fiber's speed?
Don't listen to the
Would you even know what types of fiber to buy? (multi-mode / single-mode, C-band / L-band / Extended-L band,
If you are worried about an upgrade path, the smart thing to do is install conduits. When fiber goes to the consumer market, you will be ready.
Kevin
Hmmmm ... the description sounds all wrong.
... but the description is too vague to tell.)
...") if you don't in fact know what you are talking about.
I wonder about the scientific literacy around here when this comment is at a +5.
Most notably, particles don't helix around electric field lines. Particles helix around magnetic fields. This is high school physics people!
Now if I substitute electric for magnetic, the description makes a bit more sense. With this substitution, it sounds like the poster is describing a variant of a magnetic mirror machine.
This is so old it seems new.
(However, the variation might be original
At Livermore in the 1960s, a giant mirror machine was built and mothballed on the opening day (funding had dried up when the fusion community began its quest for mythical Tokamak fusion reactor). When I last saw the mirror building (three years ago or so), the equipment had been thoroughly picked over by experimenters.
Simple magnetic mirrors always had problems with the mirror loss-cone. Magnetic mirrors work really well except for particles travelling near normal to the current rings (the loss-cone).
Given that particles have to collide for fusion to occur, some collisions end up scattering particles into the loss-cone and thus, plasma confinement is not that great.
Also, the velocity space distribution functions resulting from a loss cone leads to a whole class of plasma instabilities (surprisingly known as "loss-cone instabilities").
More complicated mirror devices have been designed which allieviate some of these problems but they have not received much attention from funding organizations.
Please don't waste other people's time talking authoritatively ("Classical EM shows
Kevin
IANAP, so I am talking out my ass here, but it seems to me that the interior surface of the container of a reaction might just somehow be able to more directly collect energy, similar to the way solar-panels collect light. Of course, I have no idea what kind material might be used to accomplish this. How do solar-panels work? Silicon? Is it just dumb luck that the elements of a solar panel happen to convert light to energy, or is it a man-made composite, built specifically for that purpose?
... fusion reactor vessels will likely not be cheap).
... collisions ...). By that time, all that the sun radiates is a near perfect blackbody radiation spectrum.
Yes, you are talking out of your ass.
Note: I did my Ph.D work in plasma physics but now I work in quantum and optical electronics. I am probably one of the better qualified people here to answer your question.
Conventional fusion reactors fuse deuterium and tritium. Or, if you breed tritium from a lithium blanket surrounding the reaction, you can do fusion using deutrium-deutrium full burn (this is tougher to do than D-T reactions).
However, the by-products of D-T and D-D fusion are mostly high energy neutrons (and some gamma rays and neutrinos and alpha particles...). High energy neutrons are not easy to convert into electricity because neutrons are not charged. In fact, the neutron flux of a large fusion reactor would be deadly and thus a fusion reactor needs to be heavily shielded while operating. (Watch "Chain Reaction" and laugh as Keanu and his advisor walk around the operating reactor after it stabilizes.)
A typical approach for the conversion consists of letting your neutron flux heat a block of lead (or other material) and then running a standard steam cycle.
This sucks on many levels.
First, you end up throwing away much of your power from the inefficiency of the steam cycle. That is the theoretical thermodynamic efficiency of a fusion reactor which can somehow do direct conversion is effectively 100% (hot reservior at millions of degrees, cold reservior at room temp) while a steam cycle is limited by how hot you can heat your materials (hot reservior at thousands of degrees, cold reservior at room temp).
Second, the neutron flux will activate (i.e. make weakly radioactive) the walls of your reactor and steadily degrade the structural integrity of your vessel. As a result, current estimates are that the core of a fusion power plant will need to be replaced every couple of years (which makes energy providers frown
Other approaches are to use fusion reactors as breaders for fission plants (i.e. use fusion neutrons to enrich fission reactor fuel). Many estimates already put current reactor technology beyond breakeven for this type of design. However:
- Design is not politically feasible (some alternative fission fuel cycles might be possible). Fission suffers from NIMBY and fusion-fission breeders have a massive proliferation risk.
- Fuel for fission reactors is not particularly rare and running a fusion plant will likely not be cheap. Thus, economically, currently there is no compelling reason.
Solar cells on the other hand rely on photons exciting electron-hole pairs in a semiconductor. The light from the sun partially consists of photons in the visible and near infrared range which are suitable for conversion by a solar cell.
You might be wondering why fusion reactions produce high energy neutrons and gamma rays and other generally nasty things while our sun shines a whole lot of light. You should remember that the sun is big and the products of a fusion in the sun take hundreds of thousands of years to reach the surface (random walk
11 MW to 10 Mpeople is a bit off. You probably meant 11 GW per 10 Mpeople.
...) filling the same area could provide power for a small country (several million poeple).
At 11 MW / 10 Mpeople gives 1.1 Watts power consumed per person. Hell, power consumption in the 1800s per person was probably higher through candles alone.
It is more like 1 kW per person (your mileage may vary outside the US).
Thus, the ~40 MW power from a square mile of solar cells could supply a small town of 40,000 people. A conventional power plant (gas, nuclear, coal,
Kevin
Applied physicist, eh? Does that mean you don't ignore gravity, mass, and wind resistance like the normal physicists? :)
Something like that. Also, don't forget economics.
Kevin
There is no emoticon for what I am feeling.
- The Comic Book Store Guy
Good point. Making an oscillator wouldn't be too hard but I wouldn't know the efficiency loss off the top of my head.
...
This is why I took the generous 30 W/m2 estimate and hacked off another 50% to account for real world engineering issues such as:
- Storage efficiency
- Transmission efficiency (I would consider the DC-AC conversion part of the transmission issue).
- Grazing incidence.
-
Basically, the point is, that per unit area, most of the time solar cells are not a very attractive technology. Per unit cost, most of the time solar cells are not a very attractive technology.
To try to slap some realism into the idealism floating around here: People keep talking about new fancy solar cells with a 30%+ efficiency but they seem to be forgetting that solar cells have to be a large commodity industry if it to compete with other power sources. A commodity industry means the cells are easy to manufacture, cheap to manafacture (in both money and energy), high volume and low profit margins per unit. And by easy and cheap to manufacture, I mean you have to be able to produce hundreds of square miles of PV material. Fine nanostructured materials with complicated anti-reflective layers with bandgap engineered hetero-junction quaternary compounds doped with exotic rare earths is not going to be a commodity product for the forseeable future. Boring amorphorous silicon based cells with its lower efficiency have a much better chance (which is what I would bet is being used in these third world countries).
Environmentally solar cells may have some advantages but solar cell manufacturing and disposal needs to be accounted for too.
If you think that solar power (as generated by PV solar cells) can satisfy a large part of the power load of an industrialized country, you are dreaming. (If you give me a sqaure mile to be used exclusively for power generation, I can give you giga-watts conventional instead of mega-watts solar).
If you think that solar cells should be used to reduce demand during heat waves and can decentralize power generation somewhat, you are thinking more realistically. Such uses are already supported with tax breaks for deploying solar cells (which are slated to be expanded).
A while ago, I calculated how soon deploying solar cells on the roof of a house would pay for itself. The short answer I would cost several thousand dollars and it would take roughly 10 to 20 years realistically (as energy costs change, this figure bobs around too). And of course, it all depends on where you live, what you doing with the solar energy (i.e. solar cell PVs are not the only use of radiant flux) and whether you can sell power onto the grid in your locale.
If you need to make home improvements, you can get a much higher return doing other things (the afforementioned tax breaks help). If you are an ardent supporter of the technology, you would be better off to invest the money into a solar cell company conducting research to bring down the manufacturing cost to make the technology competitive.
If you more interested in large centralized power plant solar installations, it would probably be more cost effective and environmentally friendly to use a helio-concentratator instead of PVs. And such a plant would really only be feasible in certain areas anyways (I don't expect to see helio-concentrators anytime soon in say Seattle).
So the short answer:
Solar cells will be used when the technology is competitive. Most of the time it is not.
There is no vast conspiracy.
Kevin
The best solar cells generally have about 30% efficiency, relative to the total flux of sunlight hitting the earth in the given area. To output 100 MW requires about 1 square mile of cells
---------
Interesting maths there.
You get ~1kW/m^2 of sunlight, so that's 2560000kW of raw sunlight, which at 30% efficiency is:
2560000kW * 0.3 = 768 000kW == 768MW
---------
Point of view from an applied physicist:
Sunlight from the earth is 1.2 kW/m^2 (higher than your estimate). However, 30% efficiency is a pretty big overestimate of how much you can get from a presently economical solar cell. There are solar cells approaching 30%, but these tend to be made out of more exotic materials and are not as easily manufactured.
Starting from an incident flux of 1.2 kW/m2:
Half the time it is night. So averaged over a day gives: 600 W/m2.
But in most places, not every day is sunny. So, hack another half off that: 300 W/m2.
Also, the efficiency of a solar cell is strongly dependent on the angle of incidence of the sunlight. Assuming a low cost installation where the cells aren't pivoting (expensive and prone to break down) to catch the rays gives a loss in available flux. Using a generous a cos^2 depenenced takes us down to: 150 W/m2
(Note: It is probably much worse as the cos^2 only accounts for effective reduction in solar cell cross-section area as the sun rises and sets. For grazing incidence light most of the sunlight will reflect off the solar cell. And yes, exotic solar cells have been designed to reduce this, but this adds to cost and manufacturing difficulty.)
Now, apply a realistic 20% solar cell efficiency: 30 W/m2.
Thus, a typical solar cell can be expected to yield on average (a generous estimate):
30 W/m2.
Of course, this ignores the efficiency of any storage system you might have if you want to make use of the power generated when the sun is not directly overhead on a clear day. So to try to get a more realistic feel, hack off another half to account for efficiency and grazing reflection:
15 W/m2.
So, a square mile array of solar cells could make an average contribution of:
1600 m * 1600 m * 15 W/m2= 38,400,000 W
This is one-twenth the value of the previous poster and a bit closer to the original post, but 38.4 MW can power a fair number of homes.
However, as the previous poster pointed out, in most cities, you could get more bang for your real-estate via other means.
However, when the sun is directly overhead, on a sunny day, you will get a peak performance of roughly 240 W/m2.
Since this is the time when power is most needed anyways, this points to solar cells being used to offset peak power demand when everybody's air conditioning kicks on simultaneously. I don't expect solar cells to be the primary source of power anytime soon except in special situations.
Kevin
So, how do you broadcast a single photon everywhere? That's the key. If you send the message everywhere, you are obviously not sending single photons. If you can send a single photon reliably from point a to point b, you have figured out how to make sure it doesn't get lost in between.
... ) are particles fundamental.(Bohmian quantum mechanics is a quasi-exception.)
Though it is too late for this response to make any difference, I'll waste my breath.
Quantum mechanically, a photon is an eigenmode of Maxwell's equations for the system under consideration. A photon is commonly thought of as a localized particle of light. It is not. It is most analogous to a wave (a plane wave is an eigenmode of free space; in a complicated system, the eigenmodes are less straightforeward).
A photon is not localized. A superposition of photons may be localized. Such a superposition is best called a wave packet; it is not strictly a photon though.
Confusion over this is why very few people can actually make sense of quantum mechanics, especially if explained without mathematics (all that non-sensical jibber-jabber about wave-particle duality with bad philosophy thrown in for good measure).
At no point in any quantum mechanical formalism I've seen (Hamiltonian-based, Lagrangian-based, Heisenberg matrix mechanics, Schrodinger wave mechanics, Feynman path integrals, relativistic field theory,
Quantum mechanics is about waves (or more precisely eigenmodes of the Hamiltonian). Superpositions of waves makes particle-like excitations.
So, you can send a single photon everywhere. For a quick example, think of the two slit experiment. It still works when the photons pass through the system one at a time (this has been experiementally demonstrated). Thus, one photon passes through both slits and interferes with itself on the other side.
If photons were localized, as you seem to think, the two slit experiement would fail.
However, producing a single photon is not simple. Devices like lasers will produce a spectrum of photons with a certain narrow energy spread and a certain narrow angular spread. Such superposition of photons will be localized in space and are what people often call photons or particles of light. The probability of detecting such a wave packet in two widely separated places is negligible.
However, other devices (like say an antenna) produce wave packets which are not localized.
And in response to another post:
The reason that quantam[sic] encryption isn't used everywehere, is that it's so darn hard to detect the spin of single photons.
Detecting the spin a stream of photons is much easier than you think. Photon spin and photon polarization are closely related (photon spin is a different set of basis vectors to express photon polarization). Detecting photon polarization is trivial (sunglasses anyone?). Detecting a single photon's polarization with a bit error rate low enough to be usable over long distances is more challenging but not impossible (especially if you are just doing key exchange).
Yes, I have a Ph.D. and quantum electronics is my day job.
Kevin