You should be surprised to get 2A out of this cell at any voltage! First, Ohm's Law doesn't dictate that a supply voltage be stable while current is being drawn. The internal resistance of the hair cell or whatever this [very unlikely] thing is will not be zero, therefore the output voltage will not remain a perfect 8.98 V as you load it. I have a hard time believing that a hair solar cell could exist (and the article did not convince me and should not convince you). I have a much harder time believing that such a thing could start up one of those CFL lights he is holding there.
The conductivity of hair is very low. I know this because I have inadvertently applied 600 V between 3/4" of hair and my (thankfully dry and unsweaty) skull, yet I live to type about it. The possibility of a hair solar cell is, in my oh-so-humble opinion, exceedingly unlikely.
Be nice if they released their Roku Developer Kit. Couldn't find it in 10 seconds of googling. I guess we'd still need a way to get any custom app to talk to the box. It'd be pretty nice to use this box to stream from my own computer.
You have IR cams mounted on your helmets? Seriously? We have one per unit officer. What kind?
Agreed though, it would be nice if this were useful. I suppose I could build an ultrasonic device that would enable some of these capabilities -- IF it would be useful (unlikely) and IF I wanted to carry even more shit around with me.
Intel assets should not be used to spy on our own country. They have too much money to spend on this sort of thing. Imagine the DOD budget being spent to enforce laws. Traffic tickets being issued because a satellite saw you going too fast, or jaywalking. Obviously I'm going for histrionics here, but it's a slippery slope once you take away the absolute prohibition.
In my mind it's enough that Google is able to understand how my thought processes develop by tracking search queries - which is an overt and expectable result when I tell them what I'm looking for. And since they present the results to me, they also get to see which ones are appealing, both by me clicking on the result link (and thereby telling them which one I have clicked) as well as by whether I return for a similar search, or a search which takes me down a related tangent.
But I don't care for the idea that Google or any other company can know which other sites I visit, either as a result of (omg) Yahoo searches or whatever MS calls their search engine these days. Or even sites whose names I know, like facebook or various company sites whose names are typically companyname.com or similar. In other words, if I didn't ask them for it, it's not their business to know. In fact, I see it as my business alone. The fact that there is value to tracking that information, or appending that data (where I surf) to some customer record that contains my real name and address - I should have some level of control over that information. My (sadly unrealistic) opinion is that Choicepoint should be paying me a percentage of their revenue when they sell information about me. I don't care that it's aggregated - there is value to that data, it should not be theirs to sell with no restrictions.
Rather than cool, I'd call this the worst thing about the "web". If you buy a brand new X (car, TV, toaster, etc.) and half its features are missing or broken, aren't you pissed? But it's okay for software. This mentality enables companies to save money on engineering and QA and pushes these costs onto their customers. It's BS...
I remembered that a quick format only writes 0s to the drives over top of the info that is there, so its still underlying.
No, that's not right. A quick format just destroys the FAT table, and doesn't touch the data. It's just that you don't know where to look anymore, sort of like throwing away your address book. Even a full format doesn't erase the data, the OS just verifies that it can read every sector of the drive.
What you're referring to would be a low-level format, where the drive is zeroed out and the HD controller marks out bad sectors. And that, as the one of the uncle posters explains, is pretty tough to recover from in modern drive technologies.
Typically DVRs only disable the display drivers when power is off. Not a lot of power usage there. Going into a true low-power standby state requires a decent amount of work - you want the thing to be awake enough to record your programs, download the latest guide, etc. - and also to turn on fast when you want to watch the tube. Why spend the time doing that design when most consumers have no idea how much power the device uses at any point?
I'm actually coming around to the idea of regulating something regarding power use of always-powered devices. At least (as I posted above) something akin to the yellow tag you get on a dishwasher, boiler or other household device. It shows how much power the device uses in a year of typical use and its annual cost, and compares to "similar" equipment. (I can never find the items on the low end of that scale, though). For most equipment, a scale showing how the device compares to its competitors for power use in operating and standby modes could certainly sway me when buying a new TV or DVR. Assuming all else is equal, that is.
According to wikipedia, California held the 9th largest economy in the world (compared to entire countries) in 2006. Were manufacturers to decide not to sell their wares in CA to avoid regulations would result in their forfeiting sales opportunities the size of Canada or Spain. That would be simply foolish.
These "silly regulations" often lead to positive change. For example, the tag on the side of all of your pillows which scratches my itchy nose at night. Or clean air regulations, which were driven largely as a result of smog in California cities.
I'm no fan of regulation, preferring voluntary programs like Energy Star (which promote buyers to consider energy efficiency by providing a readable energy cost rating). But the free market doesn't always provide consumers with the choices they desire on its own, either. A reasonable compromise - like Energy Star - often works well. I'd love to see standby and operating power usage displayed on the box - like the big yellow tag when you buy a new boiler or dishwasher.
Interesting drawing. I was amazed to see that Cuba has no subsea internet connection at all. It's kind of interesting to see one of the larger countries in the area completely circumnavigated like that. Out of curiousity I looked to see what connection they use, and it looks like a completely satellite-based service from newcom-intl.com, judging from the traceroute responses and the huge delays which occur at that hop. Neat.
Guess that's one way to avoid having your internet connection destroyed by an anchor...
I used to work at a semiconductor fab - basically a big chemical factory. Access control, security and timecards were all kept by a barcode system, printed on the back of your badge. I had a lot of fun making bar codes to see which would get me into places I shouldn't have been, like the spaces between the cleanroom walls, or the tunnel under the building, or the chemical storage area (that was a place I didn't ever like being in). Probably seems worse now than it did then.
Back in elementary school we had a stored-value system for buying lunch, with security based on bar codes on little plastic cards. This was nearly 20 years ago and there was free software available then (on my Commodore 64? Atari? Can't remember) to generate bar codes. I made a couple, based on the ID numbers of friends, and gave them to the lunch lady, telling her that those cards were a bad idea. They never changed anything, though. These days I'd have been kicked out of school for that, though, if not arrested.
According to Helios...
Notice for users of the following display models:
Sharp AQUOS 46XD1E, Sony KDL-40V2500, Sony KDL-40XBR2, Sony SXRD KDS-50A2000, Sony Bravia KDL46"XBR, Sony KDS-R60XBR2, Mitsubishi WD-57732, Westinghouse LVM-42w2, Westinghouse LVM-37w3, other similar models.
There is a known compatibility issue between the H4000 and certain 1080p displays. The issue affects 1080p signals over HDMI connections. Due to different implementations of the Player-Display communication (HDMI handshaking) certain displays can not accept the 1080p signal from the H4000. If you have any questions about your display and compatibility with the H4000, please contact Support.
Unfortunately this includes the TV I got. And I was such a fan of the HDMI cable concept - a decent connector design, a single cable for both audio and video. Ah well. Seems like a decent DVD player, and for $119 it's tough to go that wrong.
I don't know why this is a big deal for the Slashdot crowd. Might be nice for keeping the water out of my sneaks, but for electronics? Conformal coatings (like Humiseal) have been around forever... they're just expensive and make rework miserable.
Don't think this would work all that well on entire systems. Remember, most battery-module interfaces are connectorized. If you overcoat the connector, you will reduce surface contact between the connector sides, which increases electrical resistance. In other words, a little jiggle on the connector and your device resets itself.
If they don't waterproof the connector, no problem, but then you destroy the battery when the phone or ipod or whatever goes in the toilet. Which is better than losing the whole device, but most consumers won't know the difference. Or they use waterproof connectors which are bulky and expensive and don't belong in the middle of consumer equipment.
Re:This makes sense in a lot of ways.
on
Treating the Dead
·
· Score: 2, Informative
Hm, a review of my note reveals that I am the one needing the brain surgery; the OP correctly referred to a Russian technique of inducing hypothermia for cardiac surgery. Although I have clear memories of some sort of media informing me that there were long-term negative CNS effects on these patients, a paper whose summary I have read shows that I am mistaken: http://asianannals.ctsnetjournals.org/cgi/content/ abstract/10/1/3
As an occasional EMT, this story reminds me of one of our favorite maxims - they're not dead until they're warm and dead. Children have been known to survive near-drownings in frigid water (won't bother trying to recall a specific length of time). Perhaps this work will lead to a field technique of cooling patients enroute to a hospital.
Re:This makes sense in a lot of ways.
on
Treating the Dead
·
· Score: 1
I seem to recall that a follow-up study performed on the people who were treated using the Russian hypothermic brain surgery technique demonstrated a significant loss of brain power as measured by IQ testing. No linky unfortunately.
I used to use a system much like you describe. I used to work at a major international airport, which secured some private areas from the public with a cipher lock. It had rocker buttons, five of them, at the bottom of a metal "butter tub". You could stick your hand in there and look inside and see the labels on the buttons, but once you'd seen it once, you didn't need to look again. The rocker buttons were centered and if you press one way it might be a '1' and the other way was a '6', I think.
A more interesting system was on the front door to my office - a 9-digit keypad where the numbers were lit up in a dot-matrix format. You could only read the numbers standing in front of it, and they would change each time you walked up to it. It was very cool, but they stopped using it in favor of ethernet-programmable fingerprint readers.
There are a lot of options. The tougher part is weatherproofing any of these solutions. The more fancy electronics you have, the more important keeping water out becomes. Good luck!
Aah, come on, teacher! What about supply and demand? If you offer more money, you will get more interest in the position. As an electrical engineer who enjoys mentoring [slightly] younger engineers, I would love to teach except for a few reasons - first, the pay! Even when you figure that the salary is approximately 12/10 times mine, since I'd get summers off, it's still too low compared to what I make/can make in industry.
There other reasons not to do it, but pay is tops. You have to pay for talent. That's the way it works. You give me 100k to teach and I'm in. For 50k I'll stay in industry. If it were, say, 70k to start, and I had the non-monetary benefit of knowing I was doing good every day, I'd do it in a heartbeat. Might take 60-65k to start if I had a 20-year pension. Those numbers are in dollars for an expensive US city.
Guess I'll have to go out and win the Nobel prize. Since there's a lot I can do about that. That's some top-notch research, very applicable to my life.
I think the best demonstration that a problem exists is on this story - Slashdot's tags are currently: tags, tagging, no, tag.
A good solution would be to find the linguistic similarities between the selected tags and meld them together. I get no extra information by using the three words "tag" "tagging" and "tags" which I would not have received from selecting only "tagging" or perhaps "tag". Find the similar word roots and use the appropriate one (probably the same as the most often-selected of the lot).
I can't speak from experience with this exact situation, but hey, this is Slashdot, so my mom could write here and say that she knew the answer, so I'll just plod on.
The typical 120VAC silver box converter is basically a forward converter, which means that you could almost call it a DC-AC-DC converter (but that's not what they're called, they're DC-DC converters). These are similar to the flyback converters that make your TV whine and display pretty pictures.
The core (ha, ha) of these converters is a transformer. By going from 120VAC to 380VAC, you're talking about a bigger transformer, with, actually, less wire (fewer turns) but slightly more insulation. That's about it. So the efficiency "should" be pretty comparable to the existing systems. Except since it's new they'll charge a lot more for 'em.
Yes, they will save on resistance losses by about 1/3. (120/380)
The savings will be somewhat in components and somewhat in power. After your UPS system (in which AC power is converted to DC for battery storage) there will be no need for the DC/AC conversion and then AC/DC in the silver box. This deletes one conversion stage, in theory (in theory communism works, in theory...)
AC supply -> UPS AC/DC -> battery storage -> bus converter -> several servers
rather than
AC supply -> UPS AC/DC -> battery storage -> DC/AC -> rack AC/DC -> several servers
But you'd still need an intermediate bus converter at the server rack to drop from 380 to 12/5/3.3/-5. This could be a nice opportunity for an APC or someone like them.
Also there are efficiency savings. If you optimistically quote 95% efficiency for each conversion in the existing topology, the overall efficiency from supply to server is (0.95)(0.95)(0.95)(0.95) = 81%. For the DC-only system, it would be (0.95)^2 perhaps, or 90%. These are certainly optimistic numbers but are useful for scale. For a 300W system the difference is 27W. For 1000 300W systems, it's 27000W, at.08/kWh it's 2$/hour or $19k/year.
Am I assuming too small a server farm? This doesn't really seem worth it.
As far as safety goes, 380VDC and 380VAC are both gonna hurt you. Remember that the AC voltage is spec'd as RMS, so they're basically equivalent in power delivery, which means they should kill you equally fast. My recollection is that DC voltage will cause your muscles to tense, so if you grab an energized part you cannot let go, which is a pretty scary idea. On the other hand, it ought to cause less interference (no more 60Hz).
On wire thickness... no more skin effect, at least.
I dropped a waterproof electromagnet into a fish tank filled with algae. I took samples every day for two months while I generated a (measured) 2 gauss (yes, two gauss) 60 Hz electromagnetic field. It caused the algae to grow much more dense and dark than the control samples.
Did you consider the effects of heating? Certainly you put a bit of current into that big magnet. My [limited] experience with growing algae showed me that heat and light make a big difference in its growth.
As far as the constant magnetic fields go, people are moving through it, so the field strength with respect to those around the permanent magnets you mentioned would not be constant... but this whole article is pretty stupid, so I'll just stop there.
Absolutely. There's a pretty interesting report online suggesting a system which would reform gasoline using electrolysis. This powerpoint presentation (http://fire.pppl.gov/fpa03_cohn.ppt) suggests that the cost could be around $1000 including a turbocharger and result in an efficiency increase of about 20%, as I recall. As much as 40% with a hybrid powertrain.
And this system is demonstrated by a reputable group (MIT, DoE and presumably Princeton), unlike this other guy.
Slashdot Mirror
You should be surprised to get 2A out of this cell at any voltage! First, Ohm's Law doesn't dictate that a supply voltage be stable while current is being drawn. The internal resistance of the hair cell or whatever this [very unlikely] thing is will not be zero, therefore the output voltage will not remain a perfect 8.98 V as you load it. I have a hard time believing that a hair solar cell could exist (and the article did not convince me and should not convince you). I have a much harder time believing that such a thing could start up one of those CFL lights he is holding there.
The conductivity of hair is very low. I know this because I have inadvertently applied 600 V between 3/4" of hair and my (thankfully dry and unsweaty) skull, yet I live to type about it. The possibility of a hair solar cell is, in my oh-so-humble opinion, exceedingly unlikely.
However, I was [for once] inspired to RTFA.
Be nice if they released their Roku Developer Kit. Couldn't find it in 10 seconds of googling. I guess we'd still need a way to get any custom app to talk to the box. It'd be pretty nice to use this box to stream from my own computer.
You have IR cams mounted on your helmets? Seriously? We have one per unit officer. What kind? Agreed though, it would be nice if this were useful. I suppose I could build an ultrasonic device that would enable some of these capabilities -- IF it would be useful (unlikely) and IF I wanted to carry even more shit around with me.
Intel assets should not be used to spy on our own country. They have too much money to spend on this sort of thing. Imagine the DOD budget being spent to enforce laws. Traffic tickets being issued because a satellite saw you going too fast, or jaywalking. Obviously I'm going for histrionics here, but it's a slippery slope once you take away the absolute prohibition.
In my mind it's enough that Google is able to understand how my thought processes develop by tracking search queries - which is an overt and expectable result when I tell them what I'm looking for. And since they present the results to me, they also get to see which ones are appealing, both by me clicking on the result link (and thereby telling them which one I have clicked) as well as by whether I return for a similar search, or a search which takes me down a related tangent.
But I don't care for the idea that Google or any other company can know which other sites I visit, either as a result of (omg) Yahoo searches or whatever MS calls their search engine these days. Or even sites whose names I know, like facebook or various company sites whose names are typically companyname.com or similar. In other words, if I didn't ask them for it, it's not their business to know. In fact, I see it as my business alone. The fact that there is value to tracking that information, or appending that data (where I surf) to some customer record that contains my real name and address - I should have some level of control over that information. My (sadly unrealistic) opinion is that Choicepoint should be paying me a percentage of their revenue when they sell information about me. I don't care that it's aggregated - there is value to that data, it should not be theirs to sell with no restrictions.
Rather than cool, I'd call this the worst thing about the "web". If you buy a brand new X (car, TV, toaster, etc.) and half its features are missing or broken, aren't you pissed? But it's okay for software. This mentality enables companies to save money on engineering and QA and pushes these costs onto their customers. It's BS...
I remembered that a quick format only writes 0s to the drives over top of the info that is there, so its still underlying.
No, that's not right. A quick format just destroys the FAT table, and doesn't touch the data. It's just that you don't know where to look anymore, sort of like throwing away your address book. Even a full format doesn't erase the data, the OS just verifies that it can read every sector of the drive.
What you're referring to would be a low-level format, where the drive is zeroed out and the HD controller marks out bad sectors. And that, as the one of the uncle posters explains, is pretty tough to recover from in modern drive technologies.
Yet another physical phenomenon fits the theory of everything. How about a prediction from string theory for once?
Typically DVRs only disable the display drivers when power is off. Not a lot of power usage there. Going into a true low-power standby state requires a decent amount of work - you want the thing to be awake enough to record your programs, download the latest guide, etc. - and also to turn on fast when you want to watch the tube. Why spend the time doing that design when most consumers have no idea how much power the device uses at any point?
I'm actually coming around to the idea of regulating something regarding power use of always-powered devices. At least (as I posted above) something akin to the yellow tag you get on a dishwasher, boiler or other household device. It shows how much power the device uses in a year of typical use and its annual cost, and compares to "similar" equipment. (I can never find the items on the low end of that scale, though). For most equipment, a scale showing how the device compares to its competitors for power use in operating and standby modes could certainly sway me when buying a new TV or DVR. Assuming all else is equal, that is.
According to wikipedia, California held the 9th largest economy in the world (compared to entire countries) in 2006. Were manufacturers to decide not to sell their wares in CA to avoid regulations would result in their forfeiting sales opportunities the size of Canada or Spain. That would be simply foolish.
These "silly regulations" often lead to positive change. For example, the tag on the side of all of your pillows which scratches my itchy nose at night. Or clean air regulations, which were driven largely as a result of smog in California cities.
I'm no fan of regulation, preferring voluntary programs like Energy Star (which promote buyers to consider energy efficiency by providing a readable energy cost rating). But the free market doesn't always provide consumers with the choices they desire on its own, either. A reasonable compromise - like Energy Star - often works well. I'd love to see standby and operating power usage displayed on the box - like the big yellow tag when you buy a new boiler or dishwasher.
a mother was breastfeeding in a restaurant, and the waitress asked her to stop...
Hey, the sign said "No outside food"...
Interesting drawing. I was amazed to see that Cuba has no subsea internet connection at all. It's kind of interesting to see one of the larger countries in the area completely circumnavigated like that. Out of curiousity I looked to see what connection they use, and it looks like a completely satellite-based service from newcom-intl.com, judging from the traceroute responses and the huge delays which occur at that hop. Neat.
Guess that's one way to avoid having your internet connection destroyed by an anchor...
I used to work at a semiconductor fab - basically a big chemical factory. Access control, security and timecards were all kept by a barcode system, printed on the back of your badge. I had a lot of fun making bar codes to see which would get me into places I shouldn't have been, like the spaces between the cleanroom walls, or the tunnel under the building, or the chemical storage area (that was a place I didn't ever like being in). Probably seems worse now than it did then.
Back in elementary school we had a stored-value system for buying lunch, with security based on bar codes on little plastic cards. This was nearly 20 years ago and there was free software available then (on my Commodore 64? Atari? Can't remember) to generate bar codes. I made a couple, based on the ID numbers of friends, and gave them to the lunch lady, telling her that those cards were a bad idea. They never changed anything, though. These days I'd have been kicked out of school for that, though, if not arrested.
Unfortunately this includes the TV I got. And I was such a fan of the HDMI cable concept - a decent connector design, a single cable for both audio and video. Ah well. Seems like a decent DVD player, and for $119 it's tough to go that wrong.
I don't know why this is a big deal for the Slashdot crowd. Might be nice for keeping the water out of my sneaks, but for electronics? Conformal coatings (like Humiseal) have been around forever... they're just expensive and make rework miserable.
Don't think this would work all that well on entire systems. Remember, most battery-module interfaces are connectorized. If you overcoat the connector, you will reduce surface contact between the connector sides, which increases electrical resistance. In other words, a little jiggle on the connector and your device resets itself.
If they don't waterproof the connector, no problem, but then you destroy the battery when the phone or ipod or whatever goes in the toilet. Which is better than losing the whole device, but most consumers won't know the difference. Or they use waterproof connectors which are bulky and expensive and don't belong in the middle of consumer equipment.
Hm, a review of my note reveals that I am the one needing the brain surgery; the OP correctly referred to a Russian technique of inducing hypothermia for cardiac surgery. Although I have clear memories of some sort of media informing me that there were long-term negative CNS effects on these patients, a paper whose summary I have read shows that I am mistaken: http://asianannals.ctsnetjournals.org/cgi/content/ abstract/10/1/3
As an occasional EMT, this story reminds me of one of our favorite maxims - they're not dead until they're warm and dead. Children have been known to survive near-drownings in frigid water (won't bother trying to recall a specific length of time). Perhaps this work will lead to a field technique of cooling patients enroute to a hospital.
I seem to recall that a follow-up study performed on the people who were treated using the Russian hypothermic brain surgery technique demonstrated a significant loss of brain power as measured by IQ testing. No linky unfortunately.
I used to use a system much like you describe. I used to work at a major international airport, which secured some private areas from the public with a cipher lock. It had rocker buttons, five of them, at the bottom of a metal "butter tub". You could stick your hand in there and look inside and see the labels on the buttons, but once you'd seen it once, you didn't need to look again. The rocker buttons were centered and if you press one way it might be a '1' and the other way was a '6', I think.
A more interesting system was on the front door to my office - a 9-digit keypad where the numbers were lit up in a dot-matrix format. You could only read the numbers standing in front of it, and they would change each time you walked up to it. It was very cool, but they stopped using it in favor of ethernet-programmable fingerprint readers.
There are a lot of options. The tougher part is weatherproofing any of these solutions. The more fancy electronics you have, the more important keeping water out becomes. Good luck!
Aah, come on, teacher! What about supply and demand? If you offer more money, you will get more interest in the position. As an electrical engineer who enjoys mentoring [slightly] younger engineers, I would love to teach except for a few reasons - first, the pay! Even when you figure that the salary is approximately 12/10 times mine, since I'd get summers off, it's still too low compared to what I make/can make in industry.
There other reasons not to do it, but pay is tops. You have to pay for talent. That's the way it works. You give me 100k to teach and I'm in. For 50k I'll stay in industry. If it were, say, 70k to start, and I had the non-monetary benefit of knowing I was doing good every day, I'd do it in a heartbeat. Might take 60-65k to start if I had a 20-year pension. Those numbers are in dollars for an expensive US city.
Guess I'll have to go out and win the Nobel prize. Since there's a lot I can do about that. That's some top-notch research, very applicable to my life.
I think the best demonstration that a problem exists is on this story - Slashdot's tags are currently: tags, tagging, no, tag.
A good solution would be to find the linguistic similarities between the selected tags and meld them together. I get no extra information by using the three words "tag" "tagging" and "tags" which I would not have received from selecting only "tagging" or perhaps "tag". Find the similar word roots and use the appropriate one (probably the same as the most often-selected of the lot).
I can't speak from experience with this exact situation, but hey, this is Slashdot, so my mom could write here and say that she knew the answer, so I'll just plod on.
The typical 120VAC silver box converter is basically a forward converter, which means that you could almost call it a DC-AC-DC converter (but that's not what they're called, they're DC-DC converters). These are similar to the flyback converters that make your TV whine and display pretty pictures.
The core (ha, ha) of these converters is a transformer. By going from 120VAC to 380VAC, you're talking about a bigger transformer, with, actually, less wire (fewer turns) but slightly more insulation. That's about it. So the efficiency "should" be pretty comparable to the existing systems. Except since it's new they'll charge a lot more for 'em.
Yes, they will save on resistance losses by about 1/3. (120/380)
.08/kWh it's 2$/hour or $19k/year.
The savings will be somewhat in components and somewhat in power. After your UPS system (in which AC power is converted to DC for battery storage) there will be no need for the DC/AC conversion and then AC/DC in the silver box. This deletes one conversion stage, in theory (in theory communism works, in theory...)
AC supply -> UPS AC/DC -> battery storage -> bus converter -> several servers
rather than
AC supply -> UPS AC/DC -> battery storage -> DC/AC -> rack AC/DC -> several servers
But you'd still need an intermediate bus converter at the server rack to drop from 380 to 12/5/3.3/-5. This could be a nice opportunity for an APC or someone like them.
Also there are efficiency savings. If you optimistically quote 95% efficiency for each conversion in the existing topology, the overall efficiency from supply to server is (0.95)(0.95)(0.95)(0.95) = 81%. For the DC-only system, it would be (0.95)^2 perhaps, or 90%. These are certainly optimistic numbers but are useful for scale. For a 300W system the difference is 27W. For 1000 300W systems, it's 27000W, at
Am I assuming too small a server farm? This doesn't really seem worth it.
As far as safety goes, 380VDC and 380VAC are both gonna hurt you. Remember that the AC voltage is spec'd as RMS, so they're basically equivalent in power delivery, which means they should kill you equally fast. My recollection is that DC voltage will cause your muscles to tense, so if you grab an energized part you cannot let go, which is a pretty scary idea. On the other hand, it ought to cause less interference (no more 60Hz).
On wire thickness... no more skin effect, at least.
Did you consider the effects of heating? Certainly you put a bit of current into that big magnet. My [limited] experience with growing algae showed me that heat and light make a big difference in its growth.
As far as the constant magnetic fields go, people are moving through it, so the field strength with respect to those around the permanent magnets you mentioned would not be constant... but this whole article is pretty stupid, so I'll just stop there.
Absolutely. There's a pretty interesting report online suggesting a system which would reform gasoline using electrolysis. This powerpoint presentation (http://fire.pppl.gov/fpa03_cohn.ppt) suggests that the cost could be around $1000 including a turbocharger and result in an efficiency increase of about 20%, as I recall. As much as 40% with a hybrid powertrain.
And this system is demonstrated by a reputable group (MIT, DoE and presumably Princeton), unlike this other guy.