Faraday Cage

[perplex]

i dunno the exact meaning of that word either, but in school we used the epsilon-0 constant (permiability of air or something) for electrostatic work.

 

[__redruM]

If your going to go beyond a metal hammond case for your amp, only worry about shielding Before and At the Amp. Interference is usually only a problem when it is amplified.

You could drop an Ipod and an amp in a metal case, and ground the case to IG. Honestly though, a hammond case grounded to IG is more than enough isolation from RF.

 

[tangent]

Quote:

Originally Posted by PinkFloyd You obviously haven't evaluated the LM6171 "in depth"

Fooey. Microwave is a few GHz. The GBW of the 6171 is 100 MHz, meaning that it cannot pass any signal faster than that; that's mere radio frequency.

 

[drewd]

Quote:

Originally Posted by stadams Depending upon the frequencies that you are talking about here, which I don't think that anyone has explicitly defined, you will most definitely need to take into account the material's permittivity and permeability. Offhand, I consider these the most important attibutes when initially determining the effective attenuation of EM fields.

Unless your next door neighbor is running a ginormous linear amplifier, the EM fields that we're talking about are very, very small. A sheet of aluminum foil could block them. The sort of noise that you're apt to pick up is something like a vacuum cleaner and even then probably only if it has a motor with brushes and acts as a spark gap generator.

Quote:

Also, if this were going to be constructed as a room, has anyone given any thought to the construction of the door?

I know that we're drifting off topic now, but we addressed that problem by cladding the door in copper and grounding the cladding to the rest of the room's shield. Then...close the door.

-Drew

 

[Born2bwire]

Quote:

Originally Posted by stadams Hey guys, Depending upon the frequencies that you are talking about here, which I don't think that anyone has explicitly defined, you will most definitely need to take into account the material's permittivity and permeability. Offhand, I consider these the most important attibutes when initially determining the effective attenuation of EM fields. Also, if this were going to be constructed as a room, has anyone given any thought to the construction of the door? Later,

That's immaterial in this instance. For metals, the permeability is usually the same as free space but the permittivity will vary. But this all irrelevant because the skin depth of the metal is dictated largely by it's conductivity. The conductivity is so large that you only need to consider the depth of the metal in accordance to the wavelength that you want to attenuate. For a good conductor, the skin depth is approximately 1/sqrt(pi*frequency*permeability*conductivity). If we take copper for instance, and we choose 60 Hz for noise that's originating from a radiating AC power signal, the skin depth is 8.5 mm. If you take like radio frequencies, it's 6.6 microns at 100 MHz. So for your common background radiation from radio waves, terrestial sources, computers, etc, you're looking at microns and many magnitudes smaller in thickness needed. A 60 Hz EM signal is very unlikely to be produced given the fact that the wavelength is about 5000 Km long then the wiring in your house is not going to be long enough to act as an antenna. Like drewd said, a thin highly conductive metal is enough to shield from EM fields.

 

[tangent]

Quote:

Originally Posted by Born2bwire A 60 Hz EM signal is very unlikely to be produced

True. What we're worried about there is electrostatic fields, not EMI. And in that instance, what you need is lots of ferrous material, or lots of space, or both, between the generator and what you want to be protected.

 

[Born2bwire]

Quote:

Originally Posted by tangent True. What we're worried about there is electrostatic fields, not EMI. And in that instance, what you need is lots of ferrous material, or lots of space, or both, between the generator and what you want to be protected.

Well in the electrostatic case a Faraday cage isn't going to do much of anything and like you said you can only just use distance and large amounts of ferrite to try and remove/redirect the static fields.

 

[pDave]

Wouldn't a metal shell already protect the inside from electrostatic?

Shouldn't need large amount of metal/space. Or are we into a/b/g radiation, requiring huge blocks of lead?

If you are really building a cage, remember to think about the cable entry/exit points, and consider the noise that your equipment make themselves (CRT-TV?). The door shouldn't be too difficult. Just imagine your fridge door, but replace the rubber thing with compressible metal mesh (like those you see under the "race car on metal/power track" toys). It doesn't need to be air tight. Another problem is how do you make all this look good...

 

[drewd]

Quote:

Originally Posted by pDave Wouldn't a metal shell already protect the inside from electrostatic? Shouldn't need large amount of metal/space. Or are we into a/b/g radiation, requiring huge blocks of lead?

The metal will protect the inside from electrostatic interference only if there are no openings. Since it's a static field, it consists only of a fixed amount of charge. If a metal box is introduced into the field, its surface becomes charged as well and that's the key point - surface doesn't mean just the outside, it means anything that is a surface that can be exposed to the field. If there is a hole in the box, then the interior space is also a surface. ES is different from EMI in that there is no propagation by wave - it's just a fixed charge. The real problem comes when that electrostatic potential finds a path to ground, say through an interconnect cable. Then that potential is rapidly discharged, you get, in effect, a spark (or literally, for that matter) and that discharge creates interference with the signal. If the potential is big enough, it can also damage solid state gear, as we know from dealing with computers.

It's not radiation because it's not propagating - it's just an accumulation of charge sitting on a surface, not unlike the plates of a capacitor.

-Drew

 

[Born2bwire]

Quote:

Originally Posted by pDave Wouldn't a metal shell already protect the inside from electrostatic? Shouldn't need large amount of metal/space. Or are we into a/b/g radiation, requiring huge blocks of lead? If you are really building a cage, remember to think about the cable entry/exit points, and consider the noise that your equipment make themselves (CRT-TV?). The door shouldn't be too difficult. Just imagine your fridge door, but replace the rubber thing with compressible metal mesh (like those you see under the "race car on metal/power track" toys). It doesn't need to be air tight. Another problem is how do you make all this look good...

Another way to think of it is this: If you have a charge +Q and place a perfectly conducting spherical shell around it, then the +Q charge will induce a total surface charge of -Q on the inside of the PEC shell. Since we're using a shell that was originally neutral, then to preserve charge conservation, the total surface charge on the outside of the shell is going to be +Q. Thus, even though we encased a static electric field and it's source by a PEC, it's the same as if there was no shell at all. If you ground the sphere then you would be able to keep the surface charges at zero. The other main problem is with a static magnetic field. One of Maxwell's Equations stipulates that the amount of magnetic flux leaving a closed surface is equal to the amount of magnetic flux entering. Basically it just means that magnetic fields are continuous, they don't originate from a point and radiate out, they always have to come back to that point. So you can't "ground" a magnetic field like you can with an electric field. The best you can do is try to bend the magnetic field around your object. Ferrite materials have a higher permiability than free space and this causes the magnetic fields to bunch up when in the material. By using large amounts of ferrite, you can create sort of a guide to bend the majority of the static magnetic field away from your object.

EM wave radiation is different in that the fields, though produced from a source, are self-generating. The changing electric field creates the changing magnetic field which creats the changing electric... So by attenuating one of the fields, you effectively attenuate the entire wave. A conductor destroys the field by attenuating and reflecting the electric field.

 

[PinkFloyd]

Quote:

Originally Posted by tangent Fooey. Microwave is a few GHz. The GBW of the 6171 is 100 MHz, meaning that it cannot pass any signal faster than that; that's mere radio frequency.

Well, it almost microwaved the end of my finger when it oscillated on one occasion