Joined
·
680 Posts
I've heard so much about faraday cages that I got a little tired of the misinformation, so I have decided to explain why they're not, as designed, suitable for EMP protection. I first have to explain how this electromagnetic stuff works as it relates to EMP, then I visit faraday cages and make the connection. I've tried to make this as simple as possible, and when I got done, my brain hurt, so I didn't proofread it. Might be a few errors in there, but when discussing my errors, please keep it as layperson-oriented as possible so as to not lose the average person with the theory. I'm also not an expert at any one thing. In my job, I have to know a little about a lot of things, and have to be able to connect a lot of different physics effects to other, ordinarily non-related, effects.
A faraday cage, by its definition, can be made of any conductor, be it silver plated cloth (a lot of transportable EMI enclosures are made of this,) solid copper plate, steel wire, or aluminum window screen. It’s purpose, in classic and industrial uses, is to remove electrostatic and electric field influences from one environment to the next. It tends to have an added benefit of having some electromagnetic shielding benefits, but when devices are made to be EMI (electromagnetic interference) shields, the construction INTENT is a whole lot different.
This can get really complex, and I'm having a hard time getting a deeper understanding of this myself, so if anyone knows the details of this better than I do, feel free to pitch in. I'll also try to keep this as simple as I can make it, to the point where I may overdo it or get some detail wrong. In a discussion of this nature, it’s all too easy to take the topic way over the heads of the majority of the followers of this thread. Those who understand this stuff really well need to be mindful of my request to keep things simple for everyone's benefit, or take it elsewhere.
Okay, here we go. There’s a term called wavelength. It's an inverse of the frequency. Basically for a relatively low frequency it's measured in meters, and the higher you go they use smaller units. You've often heard of the 13 meter band, which is 21450-21850kilohertz, where kilo is a thousand and hertz is per second, so the numberx1000 is how many times per second the signal switches polarity. Another example is 2.4Gigahertz, which is 125 millimeters. In general radio operations, this signal is called RF, for radio frequency, although RF can be sourced from many things other than a radio. Two components of this signal are important, and now we move on to the next bit of the lesson.
The first component is the rate at which the signal crosses from high to low. The slower this signal crosses over, the less likely it is to be attenuated/reduced by other surfaces. Low frequencies have the great ability to go right through a lot of materials, which is why submarines use ULF, or ultra low frequency, communication frequencies. It also tends to be omnidirectional, meaning you can't "point" the signal in a specific direction. The higher the frequency, the higher it crosses over, and because of this high rate, it's is more easily affected by surfaces around it (hence it can be pointed with an antenna or waveguide.) When you get way up there, something else starts to happen. Instead of going right through materials, they tend to travel along the surface, and the higher the frequency the thinner this skin effect becomes. So, in a nutshell, low frequencies go right through many materials and high frequencies run along the surface.
Now this material doesn’t necessarily have to be a metal. Any material has some electrical resistance properties, as well as other properties that relate to electromagnetic I won’t get into. Your skin, for example, is a conductor to some degree. I won’t go into defining what terms are used for what sort of conductor, but a conductor in this respect can be water, soil, skin, tree bark, air, plastic, insulation, steel, copper, or even a superconductor (won’t go there.) Any material that can conduct any amount of electricity we’ll call a conductor. All materials have a property called resistivity, which is an expression of how well a material resists the flow of an electric current. Very good insulators have high resistivity, conductors have very low values, and everything else is in between. There is no perfect insulator, just as there is no perfect conductor. Real smart folks are working on both, but it ain’t there yet. The inverse of a conductor’s resistivity is called conductivity.
Here’s where we start to put it together. For a material of given conductivity and the frequency of an RF signal nearby, we then bring up a concept called skin depth. The better the conductor and the higher the frequency, the less current actually travels through the bulk of the material. Conversely, the lower the frequency and the lower the conductivity of a material, the more the material can conduct through the bulk of the material. An example is the low frequency of household power, at 60hz, (with a wavelength of 5000 kilometers) will travel down the bulk of any conductor, so we generally have solid conductors like wire or bar. At high frequencies like your cell phone, the frequency is around 2.4gigahertz, with wavelengths of 125mm. At these frequencies, a solid copper wire is useless weight and cost, so they use hollow conductors, since the copper in the middle is useless as the current is flowing out the outside surface.
Should I put a hole or slot in that hollow conductor whose size matches, or is a fractional relationship of, that wavelength, I can then allow the current to flow through the slot and into the middle of the conductor or to another area depending on other design factors. If the slot is not the right length, it won’t pass through, as the slot is “tuned” for other frequencies. It could, at specific frequencies, act as an antenna. The described effect is used heavily in waveguides, cell phone towers, and so on. Nifty stuff, and the math gets really complex really, really fast.
Should I feed an RF signal either onto or through a conductor, by the nature of it’s geometry, has a frequency that it really likes a lot. Exactly like a guitar string that only wants to vibrate at a certain frequency. If you feed a conductor a broadband RF signal (like the hiss you hear in the audio band) into a conductor, it will reflect the signals back and forth, making the signal at it’s preferred wavelength stronger, just as a guitar string will start to vibrate when you feed it noise. This is called its resonant frequency. Additionally, this will create frequencies whose amplitude is larger than the applied noise level, just like the guitar string. The simplest way to describe this is pumping, where a broadband noise pumps up the resonant frequencies. An example of this is the Tacoma Narrow bridge, where a gentle breeze of a steady rate pumped on the resonant frequency of the bridge until it tore itself apart.
I hope you get all that.
So, now we’re back to the cage. The classic wire cage is usually made up of a screen whose opening sizes are regular and consistent. Therefore, it has a frequency that it likes, and because the openings are regular and consistent, and made up of wires that are all groups of lengths, there are a few frequencies that it wants to resonate at, and some that it will let right through the openings. Additionally, based on the conductivity of the material and frequency applied, the RF signal at certain frequencies (generally lower) will go right through the wire and on into the cage. It may also excite or pump other frequencies that are fractionally related to that frequency (called harmonics.) Now if the cage were made of a solid material, you wouldn’t have things as bad, but there’s still a lot of stuff going on. Low frequencies will still go through the conductor, as the magnetic component isn’t reduced much (attenuated) by the conductor, yet because each piece of the box has its own geometrically defined resonant frequency, there are certain frequencies that will be amplified and passed on to the other side (passing through isn’t quite the right term, but it’s close enough for this explanation.) Due to skin depth, you need two things: One, you need a conductor which will have a shallow skin depth (no penetration) at a wide frequency range. This is why copper an aluminum are preferred, and the thicker the better. Since at low frequencies, the magnetic field component will pass right through, you also need a material with high magnetic permeability, which is a way to describe a property of a metal that resists a magnetic field. (This is generally a ferrous metal like steel, but there are specific metals made with higher permeability.) Materials that have both high permeability and high conductivity don’t exist, so you need layers of both. However, even these materials “saturate” at very powerful magnetic fields, and any magnetic field over a certain threshold will then get through.
An EMP (electromagnetic pulse) is an event with a very broad band of RF signals, from very low frequency to extremely high frequency, even x-rays, which is out of the electromagnetic spectrum, but it serves to describe how much of the entire RF frequency range is covered. Basically, it’s all in there there. The signal strength is also extremely high, so high that instruments don’t really exist that can directly measure it. They can measure the indirect effects only.
Now another term and concept: Eddy currents. Also called the transformer effect. If I have a very strong magnet and a sheet of copper (this is a cool trick worth trying, I promise!) I can wave the magnet next to the sheet of copper and I’ll feel a slight drag. The faster and closer to the copper I move the magnet, the more I feel the effect. What’s happening is that the magnetic field flux lines cut through the conductor (copper works best, then aluminum, but others are too inefficient at for this to be illustrated) and that moving magnetic lines create an electrical current flowing through the conductor. It’s not necessarily flowing from one end to the other, but it’s literally flowing in circles around the magnetic field lines. With a powerful magnet, these electrical currents can flow well enough that they actually create their own magnetic field. The cool thing is that the eddy current generated magnetic fields actually oppose the magnetic field that created the eddy currents in the first place. This is how the older needle type speedometers on your cars worked, as well as a lot of roller coasters and amusement park rides stop the car. Magnetic braking is what’s working here. This is partially why high frequencies are a surface effect, but I won’t go into that further. If I have a broadband magnetic wave cutting through the conductor, I have all sorts of eddy currents being generated, some large and some small. The higher frequency components will tend to be at the surface and be of a lower magnitude, while the larger magnetic fields on the inside push current to the skin. Then we have lots of circulating currents on the skin, of various frequencies and amplitudes. Since a voltage potential and electrical current flow go hand in hand, there are a number of large voltage potentials all over the surface of the conductor. When there’s a voltage potential, the voltage is looking for the path of least resistance, but with magnetic fields in and around the conductor thwarting the conductor being that path, the air often winds up being the next best thing and you get a lot of really nifty looking arcs and sparks jumping from point to point on the surface of the conductor, as well as from any features that look attractive from an electrostatic perspective, like a bolt head, wire, corner, and whatever.
Now with the extremely broad range of frequencies involved, as well as the extremely high RF and associated magnetic fields it’s not that difficult to imagine that an enclosure of some sort with any openings at all is not suitable for EMP mitigation. The military and electrical power grid companies have been dealing with this problem for a very long time, and it’s still their best guess as to how their plans will work when presented with an EMP event. Since an opening in any military container is inevitable, they must design all sorts of filters and voltage blocks to offset this challenge. If there’s one good thing about the internet’s current design as it relates to EMP, is that the bulk of the internet lines are fiber optic, which Is, for the most part, not affected by EMP, nor will it make things worse like long, long conductors strung across the country can do.
Okay, CarlMc, you’re so darn arrogant, what can I do about my precious electronic equipment in the event of an EMP event? Well, this is really your problem, not mine, but you really do need to get smart on your own, starting out with defining what's the most important to you, and grouping them down in order to lowest priority. I’ve given you more useful information than you’ll get from most places out there.
While Wikipedia is a good source of information for those just jumping into the field, it’s far from the last source. If you’re inclined to learn physics and all that is physics, or even a small part of it, poke around here:
http://hyperphysics.phy-astr.gsu.edu/hbase/hframe.html
For further reading, you can search the web using the following terms:
EMP mitigation
MIL-SPEC EMP
MIL-SPEC EMI
My brain hurts, and I’ve said enough for today.
As the church lady on SNL once said: “Talk amongst yourselves.”
A faraday cage, by its definition, can be made of any conductor, be it silver plated cloth (a lot of transportable EMI enclosures are made of this,) solid copper plate, steel wire, or aluminum window screen. It’s purpose, in classic and industrial uses, is to remove electrostatic and electric field influences from one environment to the next. It tends to have an added benefit of having some electromagnetic shielding benefits, but when devices are made to be EMI (electromagnetic interference) shields, the construction INTENT is a whole lot different.
This can get really complex, and I'm having a hard time getting a deeper understanding of this myself, so if anyone knows the details of this better than I do, feel free to pitch in. I'll also try to keep this as simple as I can make it, to the point where I may overdo it or get some detail wrong. In a discussion of this nature, it’s all too easy to take the topic way over the heads of the majority of the followers of this thread. Those who understand this stuff really well need to be mindful of my request to keep things simple for everyone's benefit, or take it elsewhere.
Okay, here we go. There’s a term called wavelength. It's an inverse of the frequency. Basically for a relatively low frequency it's measured in meters, and the higher you go they use smaller units. You've often heard of the 13 meter band, which is 21450-21850kilohertz, where kilo is a thousand and hertz is per second, so the numberx1000 is how many times per second the signal switches polarity. Another example is 2.4Gigahertz, which is 125 millimeters. In general radio operations, this signal is called RF, for radio frequency, although RF can be sourced from many things other than a radio. Two components of this signal are important, and now we move on to the next bit of the lesson.
The first component is the rate at which the signal crosses from high to low. The slower this signal crosses over, the less likely it is to be attenuated/reduced by other surfaces. Low frequencies have the great ability to go right through a lot of materials, which is why submarines use ULF, or ultra low frequency, communication frequencies. It also tends to be omnidirectional, meaning you can't "point" the signal in a specific direction. The higher the frequency, the higher it crosses over, and because of this high rate, it's is more easily affected by surfaces around it (hence it can be pointed with an antenna or waveguide.) When you get way up there, something else starts to happen. Instead of going right through materials, they tend to travel along the surface, and the higher the frequency the thinner this skin effect becomes. So, in a nutshell, low frequencies go right through many materials and high frequencies run along the surface.
Now this material doesn’t necessarily have to be a metal. Any material has some electrical resistance properties, as well as other properties that relate to electromagnetic I won’t get into. Your skin, for example, is a conductor to some degree. I won’t go into defining what terms are used for what sort of conductor, but a conductor in this respect can be water, soil, skin, tree bark, air, plastic, insulation, steel, copper, or even a superconductor (won’t go there.) Any material that can conduct any amount of electricity we’ll call a conductor. All materials have a property called resistivity, which is an expression of how well a material resists the flow of an electric current. Very good insulators have high resistivity, conductors have very low values, and everything else is in between. There is no perfect insulator, just as there is no perfect conductor. Real smart folks are working on both, but it ain’t there yet. The inverse of a conductor’s resistivity is called conductivity.
Here’s where we start to put it together. For a material of given conductivity and the frequency of an RF signal nearby, we then bring up a concept called skin depth. The better the conductor and the higher the frequency, the less current actually travels through the bulk of the material. Conversely, the lower the frequency and the lower the conductivity of a material, the more the material can conduct through the bulk of the material. An example is the low frequency of household power, at 60hz, (with a wavelength of 5000 kilometers) will travel down the bulk of any conductor, so we generally have solid conductors like wire or bar. At high frequencies like your cell phone, the frequency is around 2.4gigahertz, with wavelengths of 125mm. At these frequencies, a solid copper wire is useless weight and cost, so they use hollow conductors, since the copper in the middle is useless as the current is flowing out the outside surface.
Should I put a hole or slot in that hollow conductor whose size matches, or is a fractional relationship of, that wavelength, I can then allow the current to flow through the slot and into the middle of the conductor or to another area depending on other design factors. If the slot is not the right length, it won’t pass through, as the slot is “tuned” for other frequencies. It could, at specific frequencies, act as an antenna. The described effect is used heavily in waveguides, cell phone towers, and so on. Nifty stuff, and the math gets really complex really, really fast.
Should I feed an RF signal either onto or through a conductor, by the nature of it’s geometry, has a frequency that it really likes a lot. Exactly like a guitar string that only wants to vibrate at a certain frequency. If you feed a conductor a broadband RF signal (like the hiss you hear in the audio band) into a conductor, it will reflect the signals back and forth, making the signal at it’s preferred wavelength stronger, just as a guitar string will start to vibrate when you feed it noise. This is called its resonant frequency. Additionally, this will create frequencies whose amplitude is larger than the applied noise level, just like the guitar string. The simplest way to describe this is pumping, where a broadband noise pumps up the resonant frequencies. An example of this is the Tacoma Narrow bridge, where a gentle breeze of a steady rate pumped on the resonant frequency of the bridge until it tore itself apart.
I hope you get all that.
So, now we’re back to the cage. The classic wire cage is usually made up of a screen whose opening sizes are regular and consistent. Therefore, it has a frequency that it likes, and because the openings are regular and consistent, and made up of wires that are all groups of lengths, there are a few frequencies that it wants to resonate at, and some that it will let right through the openings. Additionally, based on the conductivity of the material and frequency applied, the RF signal at certain frequencies (generally lower) will go right through the wire and on into the cage. It may also excite or pump other frequencies that are fractionally related to that frequency (called harmonics.) Now if the cage were made of a solid material, you wouldn’t have things as bad, but there’s still a lot of stuff going on. Low frequencies will still go through the conductor, as the magnetic component isn’t reduced much (attenuated) by the conductor, yet because each piece of the box has its own geometrically defined resonant frequency, there are certain frequencies that will be amplified and passed on to the other side (passing through isn’t quite the right term, but it’s close enough for this explanation.) Due to skin depth, you need two things: One, you need a conductor which will have a shallow skin depth (no penetration) at a wide frequency range. This is why copper an aluminum are preferred, and the thicker the better. Since at low frequencies, the magnetic field component will pass right through, you also need a material with high magnetic permeability, which is a way to describe a property of a metal that resists a magnetic field. (This is generally a ferrous metal like steel, but there are specific metals made with higher permeability.) Materials that have both high permeability and high conductivity don’t exist, so you need layers of both. However, even these materials “saturate” at very powerful magnetic fields, and any magnetic field over a certain threshold will then get through.
An EMP (electromagnetic pulse) is an event with a very broad band of RF signals, from very low frequency to extremely high frequency, even x-rays, which is out of the electromagnetic spectrum, but it serves to describe how much of the entire RF frequency range is covered. Basically, it’s all in there there. The signal strength is also extremely high, so high that instruments don’t really exist that can directly measure it. They can measure the indirect effects only.
Now another term and concept: Eddy currents. Also called the transformer effect. If I have a very strong magnet and a sheet of copper (this is a cool trick worth trying, I promise!) I can wave the magnet next to the sheet of copper and I’ll feel a slight drag. The faster and closer to the copper I move the magnet, the more I feel the effect. What’s happening is that the magnetic field flux lines cut through the conductor (copper works best, then aluminum, but others are too inefficient at for this to be illustrated) and that moving magnetic lines create an electrical current flowing through the conductor. It’s not necessarily flowing from one end to the other, but it’s literally flowing in circles around the magnetic field lines. With a powerful magnet, these electrical currents can flow well enough that they actually create their own magnetic field. The cool thing is that the eddy current generated magnetic fields actually oppose the magnetic field that created the eddy currents in the first place. This is how the older needle type speedometers on your cars worked, as well as a lot of roller coasters and amusement park rides stop the car. Magnetic braking is what’s working here. This is partially why high frequencies are a surface effect, but I won’t go into that further. If I have a broadband magnetic wave cutting through the conductor, I have all sorts of eddy currents being generated, some large and some small. The higher frequency components will tend to be at the surface and be of a lower magnitude, while the larger magnetic fields on the inside push current to the skin. Then we have lots of circulating currents on the skin, of various frequencies and amplitudes. Since a voltage potential and electrical current flow go hand in hand, there are a number of large voltage potentials all over the surface of the conductor. When there’s a voltage potential, the voltage is looking for the path of least resistance, but with magnetic fields in and around the conductor thwarting the conductor being that path, the air often winds up being the next best thing and you get a lot of really nifty looking arcs and sparks jumping from point to point on the surface of the conductor, as well as from any features that look attractive from an electrostatic perspective, like a bolt head, wire, corner, and whatever.
Now with the extremely broad range of frequencies involved, as well as the extremely high RF and associated magnetic fields it’s not that difficult to imagine that an enclosure of some sort with any openings at all is not suitable for EMP mitigation. The military and electrical power grid companies have been dealing with this problem for a very long time, and it’s still their best guess as to how their plans will work when presented with an EMP event. Since an opening in any military container is inevitable, they must design all sorts of filters and voltage blocks to offset this challenge. If there’s one good thing about the internet’s current design as it relates to EMP, is that the bulk of the internet lines are fiber optic, which Is, for the most part, not affected by EMP, nor will it make things worse like long, long conductors strung across the country can do.
Okay, CarlMc, you’re so darn arrogant, what can I do about my precious electronic equipment in the event of an EMP event? Well, this is really your problem, not mine, but you really do need to get smart on your own, starting out with defining what's the most important to you, and grouping them down in order to lowest priority. I’ve given you more useful information than you’ll get from most places out there.
While Wikipedia is a good source of information for those just jumping into the field, it’s far from the last source. If you’re inclined to learn physics and all that is physics, or even a small part of it, poke around here:
http://hyperphysics.phy-astr.gsu.edu/hbase/hframe.html
For further reading, you can search the web using the following terms:
EMP mitigation
MIL-SPEC EMP
MIL-SPEC EMI
My brain hurts, and I’ve said enough for today.
As the church lady on SNL once said: “Talk amongst yourselves.”
You lost the Layperson before the second paragraph. Good stab at at electrical theory explained in a short post format. My brain hurts and all I did was look at the pictures. 
