I was trying to conduct a test recently to show that the system I am developing behaved as expected when it can't connect to the Wi-Fi network it is configured for, and then subsequently to test that it resumes normal behavior when it can connect to that Wi-Fi network again.
I have a USB cable, a temperature probe, and a 5V power and GND wire connected to the system.
I started by wrapping the system in aluminum foil in hopes of killing the Wi-Fi signal, but my system was able to connect happily. I did a bit of research and found some suggestions on the web that the seams have to be good to block Wi-Fi because the receivers are very sensitive. I was dubious, but went out and bought some double-sided copper tape and wrapped the whole situation in a box covered in overlapping copper tap strips. I went so far as to even cover the entire temperature probe under the tape. So the only things that are exposed to the 2.4GHz radio waves in the world, in principle are the power wires (wired to a Benchtop DC power supply), and the USB cable (connected to a desktop PC). The whole situation looks like this:
I am startled that my system can still connect to Wi-Fi inside that box. It's nominally slower in getting there, but it powers up, connects to Wi-Fi, and sends a packet to my server without batting an eye otherwise.
Is this a common experience? Can anyone explain how this is possible? Faraday cage says what?
Best Answer
Copper has a skin depth of 1.3um at 2.4GHz, so it doesn't take much copper to attenuate a wifi signal. To properly build a faraday cage, you cant have apertures (holes), which you have some where the cables come out, or conductors crossing the boundary.
The connection of the power lines is likely conducting emissions that are allowing RF to pass through the chassis. To adequately block these conducted emissions use a line filter. Line filters are special low pass filter that allows DC through the filter but has a capacitor around the outside (that connects to the chassis) to block RF. I'm mostly concerned about RF getting into the chassis, so I use them to severely attenuate RF (to a level that the sensitive analog electronics won't see it). They should also work well for keeping RF inside the box.
Source: https://www.digikey.com/product-detail/en/tusonix-a-subsidiary-of-cts-electronic-components/4209-003LF/410-1088-ND/2079662
Here is a list of more of them. Select the approprate one for your attenuation needs at the frequency of interest.