If you don't have a grounded system, as a stopgap measure, electricians USED to install a GFCI, since it will provide shock protection if it senses an imbalance between hot and neutral (doesn't need a ground). If you can't use a grounded outlet, consider buying a GFCI extension cord so you can still have the protection. There's a good reason why there's a ground pin!
This only helps if the neutral is grounded somewhere, which often isn't the case in ungrounded systems.
The leakage current comes from capacitive coupling between the windings (and anything else). With higher frequencies used in switching power supplies this leakage current is considerable.
On the moon you have two fantastic resources that could change the game completely: high vacuum and strong predictable solar. You can thus achieve temperatures approaching the blackbody temp of the sun (6kK) using lightweight reflective films. Forget about PV and design your systems to use photons directly.
What you have sloshing around is not liquid Helium, as this page's phase diagram shows: http://quench-analysis.web.cern.ch/quench-analysis/phd-fs-html/node45.html you can see that liquid stops being meaningful above 5K and helium becomes a strange not-liquid, not-gas fluid known as a supercritical fluid. I am not certain, but it is possible that a super critical fluid makes sloshy noises due to turbulence from surface interaction.
In any case, making helium a liquid is not necessary to make it more dense than air. Hydrogen is even more convenient to work with, as metal hydrides store more hydrogen by mass than even liquid hydrogen! (and they can be safely handled)
Compressing He (4) to 8 atm makes it more dense than air (29). That is trivial (800kPa). Helium gas tanks, for comparison, operate at 1000atm (100MPa). A more interesting question is how much energy such compression and decompression would take.
If you are lifting 1e6 g of stuff using helium, you need at least 1e6 / (29-4) mol of helium, compressing that by a factor of factor of 8 requires -nRTlog(V1/V2) work, which is 90MJ per tonne. In practice you would need considerably less compression than that because you don't need to completely remove the bouyancy, only reduce it enough to make it managable at ground level. Some energy might be worth recovering with a gas engine.
In practice only those photons that exactly match the bandgap are able to be converted with this efficiency, limiting Silicon cells to about 30%. Using multiple layers of decreasing bandgap can produce higher efficiencies (and hence the interest in higher bandgap materials such as those based on Gallium). This lower efficiency is called the Shockley-Queisser limit which increases with increasing illumination to about 40%: http://www.pv.unsw.edu.au/Research/3gp.asp
Similarly, virtual photons corresponding to the 'flame temperature' or 'temperament' (derived from the Gibbs free energy) limit the maximum efficiency of the fuel cell to the carnot ratio: http://www.benwiens.com/energy4.html#energy1.17
I'm not convinced by the article. A big claim against ethernet is that it is non-deterministic, yet CAN bus is too - if you have packet in transmission it blocks a higher priority message just the same. Furthermore, by having to listen for the start bit, the longest network is limited to the bit size * speed of light. This means that a faster network is shorter, but more importantly, ethernet will get a considerably longer network for a given frequency * length because it uses a preamble of lots of bits.
Regarding noise immunity, ethernet, like CAN, is differentially signaled, but furthermore, uses proper transformer isolation rather than differential comparators. This means it can withstand hundreds of volts of common-mode noise.
Comparing star to multitap networks: multitap is notoriously picky about reflections, standing waves and the like. I remember the old coaxial ethernet (10base2) and the endless pain they created.
In a home automation setting I expect the lengths of CAN will be limiting. You can get the same effect as a multidrop topology with 10baseT by putting switches in each room. These days a switch can cost $10 and use a few W of power and can be powered using PoE.
Having said that, I am reconsidering CAN for some projects.
I was not aware that CAN bus systems were salvageable from cars. Which models are worth looking at?
I was not suggesting HTTP, TCP, or IP, but rather ethernet. A very simple and robust protocol, and one with many devices (PIC18F87J60, for example) with direct support. It doesn't require a separate controller, instead you can look at the network directly from a PC.
Anyway, I'll look into CAN bus a little more and see if it offers anything over ethernet for my projects. Thanks.
CAN bus has exactly the same problems as ethernet as regards to Multiple Access, but without transformer isolation, off the shelf components in the $10 range, 10Mbit minimum signaling or sharing of multiple speeds. Ethernet is also a very simple protocol to use and can be shared with IP and higher networks.
A solar cell does not work when the junction temperature is the same as the blackbody temperature of the light source. There is inefficiency due to thermal carrier recombination. When the junction temperature goes up more thermal tunnelling results in a drop in efficiency: http://en.wikipedia.org/wiki/Recombination_(physic s) Indeed you can make heat pumps this way.
Fuel cells similarly have a Thot due to Gibb's free energy. Indeed you can make an electrochemical reaction such as the decomposition of water into hydrogen and oxygen run at more than 100% electrical conversion at high temperatures, which is the basis of electrothermal reduction, giving exactly the same sort of COP as a heat pump. http://en.wikipedia.org/wiki/Gibbs_free_energy
Unfortunately a heat engine that would use even a significant amount of that 60% heat is going to require temperatures greater than the operating temperature of the solar cell. The solar cell itself is in fact a heat engine operating at roughly 42% of the theoretical maximum efficiency, which compares well with all but the biggest fluid based engines.
However, if you want heat, rather than work, you should be able to collect all of that 60% - thermal desal, domestic hot water, space heating - all are easily doable.
Why can't they make a mechanical totem pole/push pull driver for the valves? My understanding of valve design is that in fact the pressure of the gas holds the valve shut once a certain point is reached?
It looks like we're not cooling the house by electricity produced by a turbine, but by expansion of the cool, compressed gas. Is that theoretically possible? Sure. Is it likely to be controllable by any sort of thermostat? That's going to be harder. And, oh yeah, when you're not using that jet of cool air, you also don't get any electricity.
Not to mention danged noisy. It does work though. There are proposals to use high pressure air to power military computing bunkers for both energy and cooling needs.
Slashdot Mirror
You don't understand the definition of Directed Acyclic Graph.
http://en.wikipedia.org/wiki/Directed_acyclic_graph
If you don't have a grounded system, as a stopgap measure, electricians USED to install a GFCI, since it will provide shock protection if it senses an imbalance between hot and neutral (doesn't need a ground). If you can't use a grounded outlet, consider buying a GFCI extension cord so you can still have the protection. There's a good reason why there's a ground pin!
This only helps if the neutral is grounded somewhere, which often isn't the case in ungrounded systems.
The leakage current comes from capacitive coupling between the windings (and anything else). With higher frequencies used in switching power supplies this leakage current is considerable.
On the moon you have two fantastic resources that could change the game completely: high vacuum and strong predictable solar. You can thus achieve temperatures approaching the blackbody temp of the sun (6kK) using lightweight reflective films. Forget about PV and design your systems to use photons directly.
no. You might like to read up on it:
http://en.wikipedia.org/wiki/Exponential_growth
http://en.wikipedia.org/wiki/Cube_(algebra)
The energy requires to maintain your speed goes up exponentially with an increase in speed.
Actually, it goes up as the cube of the speed.
What you have sloshing around is not liquid Helium, as this page's phase diagram shows:
http://quench-analysis.web.cern.ch/quench-analysis/phd-fs-html/node45.html
you can see that liquid stops being meaningful above 5K and helium becomes a strange not-liquid, not-gas fluid known as a supercritical fluid. I am not certain, but it is possible that a super critical fluid makes sloshy noises due to turbulence from surface interaction.
Another possibility is that in fact it is the nitrogen that has liquified - balloon gas is in fact mostly nitrogen to save money. However, looking at this page:
http://www.astro.washington.edu/larson/Astro150b/Lectures/Fundamentals/fundamentals.html
it seems that the critical point for Nitrogen is still well below room temperature.
In any case, making helium a liquid is not necessary to make it more dense than air. Hydrogen is even more convenient to work with, as metal hydrides store more hydrogen by mass than even liquid hydrogen! (and they can be safely handled)
Compressing He (4) to 8 atm makes it more dense than air (29). That is trivial (800kPa). Helium gas tanks, for comparison, operate at 1000atm (100MPa). A more interesting question is how much energy such compression and decompression would take.
If you are lifting 1e6 g of stuff using helium, you need at least 1e6 / (29-4) mol of helium, compressing that by a factor of factor of 8 requires -nRTlog(V1/V2) work, which is 90MJ per tonne. In practice you would need considerably less compression than that because you don't need to completely remove the bouyancy, only reduce it enough to make it managable at ground level. Some energy might be worth recovering with a gas engine.
http://en.wikipedia.org/wiki/Raven_paradox
/me nods in mystifed agreement with the cool scientific complexity of the future.
You are correct, The carnot efficiency comes about due to the phonons:v let?prog=normal&id=JAPIAU000084000002001109000001& idtype=cvips&gifs=yesl -white-paper/solar-limitations.php2 73/o t-solar-cells.php :(
http://scitation.aip.org/getabs/servlet/GetabsSer
http://www.evidenttech.com/applications/solar-cel
http://www.springerlink.com/content/u8854u2241825
http://www.evidenttech.com/applications/quantum-d
http://eprints.soton.ac.uk/29499/
(some of these require account to access
In practice only those photons that exactly match the bandgap are able to be converted with this efficiency, limiting Silicon cells to about 30%. Using multiple layers of decreasing bandgap can produce higher efficiencies (and hence the interest in higher bandgap materials such as those based on Gallium). This lower efficiency is called the Shockley-Queisser limit which increases with increasing illumination to about 40%:
http://www.pv.unsw.edu.au/Research/3gp.asp
Similarly, virtual photons corresponding to the 'flame temperature' or 'temperament' (derived from the Gibbs free energy) limit the maximum efficiency of the fuel cell to the carnot ratio:
http://www.benwiens.com/energy4.html#energy1.17
I'm not convinced by the article. A big claim against ethernet is that it is non-deterministic, yet CAN bus is too - if you have packet in transmission it blocks a higher priority message just the same. Furthermore, by having to listen for the start bit, the longest network is limited to the bit size * speed of light. This means that a faster network is shorter, but more importantly, ethernet will get a considerably longer network for a given frequency * length because it uses a preamble of lots of bits.
Regarding noise immunity, ethernet, like CAN, is differentially signaled, but furthermore, uses proper transformer isolation rather than differential comparators. This means it can withstand hundreds of volts of common-mode noise.
Comparing star to multitap networks: multitap is notoriously picky about reflections, standing waves and the like. I remember the old coaxial ethernet (10base2) and the endless pain they created.
In a home automation setting I expect the lengths of CAN will be limiting. You can get the same effect as a multidrop topology with 10baseT by putting switches in each room. These days a switch can cost $10 and use a few W of power and can be powered using PoE.
Having said that, I am reconsidering CAN for some projects.
Email me - I'd like to pick your brains a little more.
I was not aware that CAN bus systems were salvageable from cars. Which models are worth looking at?
I was not suggesting HTTP, TCP, or IP, but rather ethernet. A very simple and robust protocol, and one with many devices (PIC18F87J60, for example) with direct support. It doesn't require a separate controller, instead you can look at the network directly from a PC.
Anyway, I'll look into CAN bus a little more and see if it offers anything over ethernet for my projects. Thanks.
CAN bus has exactly the same problems as ethernet as regards to Multiple Access, but without transformer isolation, off the shelf components in the $10 range, 10Mbit minimum signaling or sharing of multiple speeds. Ethernet is also a very simple protocol to use and can be shared with IP and higher networks.
Take a chill pill before you answer.
A solar cell does not work when the junction temperature is the same as the blackbody temperature of the light source. There is inefficiency due to thermal carrier recombination. When the junction temperature goes up more thermal tunnelling results in a drop in efficiency:c s)
http://en.wikipedia.org/wiki/Recombination_(physi
Indeed you can make heat pumps this way.
Fuel cells similarly have a Thot due to Gibb's free energy. Indeed you can make an electrochemical reaction such as the decomposition of water into hydrogen and oxygen run at more than 100% electrical conversion at high temperatures, which is the basis of electrothermal reduction, giving exactly the same sort of COP as a heat pump.
http://en.wikipedia.org/wiki/Gibbs_free_energy
You need a physicist to understand these things.
Unfortunately a heat engine that would use even a significant amount of that 60% heat is going to require temperatures greater than the operating temperature of the solar cell. The solar cell itself is in fact a heat engine operating at roughly 42% of the theoretical maximum efficiency, which compares well with all but the biggest fluid based engines.
However, if you want heat, rather than work, you should be able to collect all of that 60% - thermal desal, domestic hot water, space heating - all are easily doable.
There is no good fix for the sprawl.
I expect rising costs of fuel would help?
So right you are, affirmative is k9.
Affirmative!
http://simple.wikipedia.org/wiki/ATP
Why is it that the moderators on /. always post these silly contrarian articles and ignore the relevant scientific discussion?
Because it creates more ad hits.
Why can't they make a mechanical totem pole/push pull driver for the valves? My understanding of valve design is that in fact the pressure of the gas holds the valve shut once a certain point is reached?
The moral is: Be a fireman. I figure they get more women, anyway.
I explained your post to my wife, and she said "It's because they've got big hoses".
It looks like we're not cooling the house by electricity produced by a turbine, but by expansion of the cool, compressed gas. Is that theoretically possible? Sure. Is it likely to be controllable by any sort of thermostat? That's going to be harder. And, oh yeah, when you're not using that jet of cool air, you also don't get any electricity.
Not to mention danged noisy. It does work though. There are proposals to use high pressure air to power military computing bunkers for both energy and cooling needs.