3
\$\begingroup\$

The question Noisy data line clears up on wiring up multiple output pins of uC to Logic Analyzer got me to thinking that the best way to see if outside interference is causing problems for a circuit is to test it in a Faraday cage. Big companies will even go to the expense of testing in Anechoic test chambers.

Very few hobbyists could afford an anechoic chamber, but what about a faraday cage? Could an old microwave oven be used as a Faraday cage? It would be cheap, or free, have a nice glass window to look through as well as a nice tight mesh for all openings good up to the microwave (duh!) frequencies. The only issue I would think would be to make sure that the door seal was in good shape and that the door closed properly.

Has anyone done this or is there a better, cheaper solution for use at home?

PS: I'm not thinking of EMP hardening like this guy: http://www.unitedstatesaction.com/emp_and_faraday_cages.htm only routine testing.

\$\endgroup\$
  • 1
    \$\begingroup\$ Is it bigger than a breadbox? \$\endgroup\$ – Chris Stratton Dec 24 '11 at 19:26
  • 2
    \$\begingroup\$ Remeber to disable the microwave tube. 1 kW at a few GHz is useful for heating your lunch, but not so much for longevity of your electronics. Not so good for wet poodles either. \$\endgroup\$ – Olin Lathrop Dec 24 '11 at 20:53
3
\$\begingroup\$

There are no perfect Faraday cages that are useful. Or, to put it a different way, if it's perfect then you don't want to use it. A perfect cage will prevent everything from going in or out-- including the power you want to use and light (used for seeing if the experiment went well or not).

So, what you want is an imperfect Faraday cage! But here's the problem: is it imperfect in the correct way? A microwave oven only has to work at approximately 2.45 GHz, so it is quite possible that it doesn't work at all at 4 GHz, or 100 MHz. A quick experiment with my cell phone (AT&T 4G) shows 3 out of 5 bars outside the microwave, but inside it also shows 3 bars. We could conclude that at whatever frequency my cell phone is using is not at all shielded by my microwave.

Whatever cage you use or build, you will have to come up with some way to show that it works at the frequencies you care about.

Here are some ideas for making Faraday cages:

  1. A simple metalized mylar anti-static bag. I'm talking about the shiny silver bags-- not the pink, black, or clear+black bags. For small things, these work OK. Not perfect, but not bad either. Usually the seal on the open end is the weakest link. Fold the open end over several times and tape.

  2. Wrap whatever in aluminum foil. As before, take care with the seams. Fold the foil over several times to make a nice, tight seal.

  3. A "Screen Room". Build a room/chamber/box out of a wood frame. Then cover the frame, on both the inside and outside, with brass window screen. A simple staple-gun can attach the screen to the wood. Fold the screen seams together then solder them with a torch. Cover the floor with a solid plate of metal. Good doors are hard, but you could kludge something together. Create panels to run cables (with filters) through the walls of the screen room. Because the screen is on the inside and outside of the wood, you effectively get double-shielding.

Many companies will have a screen room of some sort for preliminary EMI and ESD testing.

\$\endgroup\$
  • \$\begingroup\$ Btw, ATT uses 700 MHz, 850MHz, 1900MHz(PCS), and 2100MHz(AWS). Your point is still true regardless. It probably is worth mentioning that wrapping something can actually cause interference issues that you wouldn't have otherwise \$\endgroup\$ – Kellenjb Dec 25 '11 at 3:26
0
\$\begingroup\$

Well, to test if a microwave would work, check to make sure your phone has signal. If it does then put it in the microwave, if you can see through the glass that it still has signal then it is not an effective faraday cage, if it loses all signal then congrats! you have a faraday cage to test your circuits in!

\$\endgroup\$
  • 1
    \$\begingroup\$ Just because a phone looses signal doesn't mean all frequencies will. \$\endgroup\$ – Kellenjb Dec 25 '11 at 3:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.