What Is a Faraday Cage? How It Works and What It Blocks

If you have ever been told to sit tight in the car during a thunderstorm, someone has given you Faraday cage advice without using the words.

The advice is sound. The reason people give for it is almost always wrong. Ask around and you will be told it is the rubber tyres, insulating you from the ground.

NOAA, the American weather agency, is blunt about it. The tyres do nothing. What protects you is the metal roof and the metal sides, carrying the current around the outside of the cabin and leaving you in the quiet part.

That distinction is the whole subject. A shield does not work by being thick, or rubber, or expensive. It works by being continuous. Once you have that, the rest falls into place, including why a convertible is no help at all, and why a careful wrap of kitchen foil sometimes stops a phone call and sometimes does not.

What a Faraday cage actually is

A Faraday cage is a conductive shell that keeps an outside electric field on the outside. That is the whole idea. The useful part is how it manages it.

Bring a charge near a conductor and the free electrons in it rearrange almost instantly, moving until the field they create cancels the field arriving from outside. The charge ends up on the outer surface and the enclosed space is left electrically quiet. Nothing is absorbed or destroyed. The shell just gives the field somewhere else to be.

Which is why the inside stays calm while the outside takes a hammering. It is also why the shell has to be joined up. A conductor can only redistribute charge across a surface that is electrically continuous. Break that continuity and you do not have a weaker shield, you have a shield with a hole in it.

How Faraday shielding works, and where it fails

Faraday shielding is measured, not switched on. A shell does not simply block or not block. It reduces, and how much it reduces depends on the shell and on the signal.

Two things govern it. The first is openings. A mesh shields effectively as long as its holes are much smaller than the wavelength you are trying to stop. Your microwave door is the everyday proof. The mesh has holes a millimetre or two across, and it holds in energy with a wavelength of about twelve centimetres, which is how you watch your dinner cook without cooking the kitchen.

The second is joins. A join that conducts is not a gap. A join that does not conduct is an opening, whatever it looks like from the outside. This is why a zip, a fold or an overlap tells you more about a shield than the fabric does.

Frequency matters as well, and this is where people come unstuck. A shell that comfortably stops a phone call can let Wi-Fi through, because the two sit at different wavelengths and a given opening is generous to one and not the other. A shield is never simply working or broken. It is working against something, and you cannot tell which without measuring.

The cube Faraday built, and what everyone gets wrong

You will read almost everywhere that Faraday proved this by lining a room with metal foil. He did not, and the real apparatus is both more interesting and more instructive.

In January 1836, in the lecture theatre at the Royal Institution in London, he had a cube built twelve feet on every side. A light wooden frame formed the twelve edges, braced with diagonal cord, standing on four glass feet five and a half inches long so the whole thing was insulated from the floor. The sides, top and bottom were covered in paper.

The metal was threaded through it rather than wrapped around it. Copper wire ran across the top and bottom to make a net, more copper wire carried down the four corner uprights, another band ran around the lower edge, and each paper wall had strips of tin foil pasted to its inner face, all of it tied back into the wire. His own account of the purpose is the line worth keeping: so that the whole might be brought into good metallic communication.

Then he put a gold-leaf electrometer inside, had the outside charged hard from an electrostatic generator, and watched the electrometer do nothing at all, during the charge and after the discharge. Not satisfied, he got in himself and, as he put it, went into the cube and lived in it.

So the first Faraday cage was mostly paper. What did the work was wire and foil strips joined into one continuous conducting surface. He was not trying to build a thick box. He was trying to build a joined-up one, and he said so.

One more thing worth straightening out, because it is the most common error in print. The famous ice pail experiment, the one with the brass ball lowered into a pewter pail, was not part of this. Faraday ran that in 1843 and described it in a letter dated 4 February that year. A great many articles collapse the two into one date. The cube was 1836. The pail was 1843. They make the same point by different means, seven years apart.

Why a car is safer in a lightning storm

Back to the car, because it is the Faraday cage most of us have actually sat inside.

A car works for one reason: the roof and the sides form a metal shell continuous enough to carry an enormous current around the outside of the cabin and into the ground. The tyres contribute nothing to that, and neither do the windows.

Continuity also tells you exactly where it stops working, and the pattern should look familiar by now.

In every one of those cases the shell is broken, or was never conductive. A broken shell is not a weak Faraday cage. It is not one.

Notice how carefully the official advice is worded, too. NOAA puts a substantial building with wiring and plumbing first. A metal-topped vehicle with the windows up is offered as a decent alternative if no building is available, and you are told to avoid touching anything that leads to the outside, the radio and ignition included.

That hedging is not bureaucratic caution. A car is a rough shell with large glass openings and a great many joins, and it is nobody's idea of a laboratory enclosure. It works well enough at the enormous energies and low frequencies of a lightning strike. Which is precisely the point to carry into the next section, because if a two tonne steel box only rates a decent alternative, it is worth asking honestly what a wrap of kitchen foil is doing.

Can you build a Faraday cage at home?

Yes. That is exactly the problem, because a homemade one usually half works, and half working is the hardest result to detect.

Faraday's own cage was paper, wire and foil strips, so thickness was never what mattered. What mattered to him was that every metal part was joined to every other. A shield is only as good as its continuity, and that is where kitchen materials let you down.

Aluminium foil conducts well and can strip out a serious amount of signal when it is wrapped tightly. It also oxidises within minutes of meeting air, so where you overlap two sheets you often get a join that looks closed and is not electrically joined at all. An overlap that does not conduct is not a seam. It is an opening.

There is a trap underneath that one. The University of Iowa physics department notes that if the foil touches the handset, it can act as an extension of the phone's own aerial, and the call comes through anyway. The device has to be insulated before it is wrapped.

Our call. Look at the crisp packet row, because it is the one that should worry you. In testing it stopped the phone calls and the GPS, and let Wi-Fi straight through, and nothing about that looks like a failure from the outside. Instrumented measurement by security researcher Matt Blaze found the same pattern more broadly: purpose made pouches held up well across the bands he could measure, while improvised methods did not do nearly as well with any consistency, though foil occasionally matched them under ideal conditions that are hard to reproduce.

Where this leaves you

Most people reading this do not need a Faraday cage. They wanted to know how one works, and the answer is more interesting than the marketing around it.

If you do want one, the useful part of everything above is what to judge it on. Not the material, and not the wording on the packaging. Ask what it was tested against and across which signals, then check the closure, because a seal that does not conduct is the same as an opening no matter how well it is sewn. We supply Faraday bags chosen on exactly that basis, and we will tell you plainly what each one blocks and what it does not.

And if you would rather make your own, go in with your eyes open. Test it on every signal you care about rather than assuming a stopped phone call means the job is done.

FreedomTech · The Privacy Experts · freedomtech.com.au