How to minimize computer em interference?

August 27, 2009 4:14 PM

What kind of perforated sheet metal should I use to build a computer case?

I'm hoping someone here can point me in the right direction. I would like to build a computer case solely out of thin, perforrated metal sheet. My worry is that the sheet won't stop the electromagnetic interference the computer parts will broadcast. What size and spacing of hole, and thickness of sheet should I use?

I'm concerned about effects on unsheilded speakers, phones, and myself. Let us assume that this will be running 24 hours a day and will be within a meter of me for most of that.

I know that my microwave has holes about 1.7 mm in diameter and center to center spacing of about 3 mm. This must be enough to stop a significant portion of that thousand watts at that frequency. Also, most computers have fan grills with very widely spaced holes/hexagons/grids. I might be overthinking the whole problem, but I hope someone with more knowledge in the area can help me deterimine that.

posted by Ctrl_Alt_ep to computers & internet (15 comments total)

A Faraday cage can be pretty sparse, I'd worry more about the forced air escaping before it can travel past the hot bits.
posted by StickyCarpet at 5:11 PM on August 27

I believe that I have the cooling worked out. I'm curious as to how to design a Faraday cage.
posted by Ctrl_Alt_ep at 5:28 PM on August 27

As long as the holes are smaller than the holes on a microwave door it will contain everything of microwave frequency or lower. This is sufficient for WiFi but not necessarily for high-frequency cordless phones or some cell phone spectrum.
posted by GuyZero at 5:33 PM on August 27

You remember this thing? That certifies that

"This device complies with part 15 of the FCC Rules. Operation is subject to the following two conditions:
(1) This device may not cause harmful interference.
(2) This device must accept any interference received, including interference that may cause undesired operations."

Look carefully on just about every appliance you own, at least those that are not specifically designed to communicate wirelessly, and you'll find that on them or their packaging somewhere.

Every single electronic device, including all computer components needs to be certified by the FCC in this way. The FCC is pretty rigorous about this, running intensive tests before certifying anything and levying significant fines on manufacturers who do not submit their products to be tested.

In short, your computer produces a negligible amount of electromagnetic interference and you really don't need to worry about it. Any detectable effects vanish after inches at most. Furthermore, electromagnetic waves have not been shown to have any real effect on human health, even prolonged exposure. Or at least, so says the WHO. The idea that electromagnetism has any pronounced effect on the human body is literally tinfoil hat territory.
posted by valkyryn at 5:36 PM on August 27

What is the frequency of the EMF you want to block? 1/40th the wavelength is given as a safe figure, I know, not that sparse after all. Also, the metal can't be too thin, and seams, even soldered or welded, will radiate.
posted by StickyCarpet at 5:40 PM on August 27

The FCC doesn't certify motherboards which under normal circumstances are kept in a solid metal box or one that for the purposes of EM radiation is essentially a solid metal box.

In short, your computer produces a negligible amount of electromagnetic interference and you really don't need to worry about it.

Heh. Noooo. Take your modem, motherboard and video card of of the case and it'll radiate a fair amount of noise.

From the FCC:

"The FCC does NOT currently authorize motherboards, cases and internal power
supplies. Vendor claims that they are selling "FCC-certified cases," "FCC-certified
motherboards" or "FCC-certified internal power supplies" are false."

Therefore if you built a PC and just left it out the mobo and psu are free to radiate as much noise as they please.
posted by GuyZero at 5:45 PM on August 27

Also:

The idea that electromagnetism has any pronounced effect on the human body is literally tinfoil hat territory.

yes, but put your GSM Blackberry next to a speakerphone and you'll discover that EM noise is very, very real for other electronic devices.
posted by GuyZero at 5:46 PM on August 27

Modders have been using/making clear acrylic cases for a number of years and don't seem to have any issues with shielding so I wouldn't worry too much. I've run my PC with both sides of the case completely removed for months at a time with no issues.
posted by Rhomboid at 6:08 PM on August 27 [1 favorite]

Thank you everyone for your answers. I think that I'll let the specifics be determined by what is cheap and availible. That 1/40th wavelength is a handy rule of thumb.

I think this all boils down to the fact that I don't know what frequencies a computer radiates.
posted by Ctrl_Alt_ep at 6:28 PM on August 27

Computers have EM frequencies varying from 60Hz (the AC in power) to over many gigahertz (the CPU clock itself). I have to snark a bit: if you don't believe the scientists, physicists, engineers, and doctors who tell you this EM is harmless, why would you believe those same people when they tell you how to shield an EM source?
posted by chairface at 7:03 PM on August 27 [2 favorites]

At 1/40th of a wavelength, you're getting very good attenuation. Even at 1/10th of a wavelength I wouldn't worry much... and as others here have pointed out, plenty of people use no-shield cases (acrylic, etc). I have the side of my case off right now as it happens.

If you want to be picky with wavelengths etc, go by the primary bus or memory clocks, which are probably about 1GHz. Considering they're not sine waves, you've got terms up to about 5GHz radiating, which means a wavelength of 6cm, so 1/4" holes will be fine. Note the size of typical fan grills: about 1/4" spacing but often with slots a couple inches long; the slots will radiate.

If you ever have a look at some RF test equipment, you will notice stuff like it has only circular holes (never slots) and air vents are often covered with what looks like mosquito mesh. It's not so much for safety as the sensitivity of the equipment and keeping all radiation out of the box.

With a microwave oven (usually 2.4GHz), you want really really good attenuation because there's a couple of kW milling about in there, so the 1mm holes are 1/125 of the wavelength. Your PC though will be radiating only a few mW and is not a safety issue unless you stick your fingers in the power supply or drop it all on your foot.
posted by polyglot at 7:13 PM on August 27

chairface,
I know many engineers and doctors, and my time with them has lead me to value evidence over any opinion. I certainly don't think that running a computer caseless is going to do much more than interfere with my radio and speakers, but a quick Google Scholar search turns up that there are many observed biological and chemical responses in humans and other mammals at high em intensities (including death) and that there really isn't enough information at low levels. Many years from now we'll either come to the conclusion that all of these radio sources are just fine or that we should think a little more about what constitutes a safe level. I just figure that it doesn't hurt to be cautious when I don't understand something well.
posted by Ctrl_Alt_ep at 7:56 PM on August 27

Laminate two sheets, with 1/4 inch spacers between them, this would generate moire effects visually.
posted by hortense at 8:08 PM on August 27

"... Many years from now we'll either come to the conclusion that all of these radio sources are just fine or that we should think a little more about what constitutes a safe level. ..."

That massive epidemiological experiment on human populations was done in the 1950s, 1960s and 1970s with the construction and operation of the Bell Systems intercity microwave network, which, by 1960, was already carrying 40% of the telephone traffic in the U.S., via hundreds of high power microwave link stations in city centers, and repeater stations at line of sight distances about every 30 miles. Millions of Americans were exposed to high amplitude microwave signals 24x7x365, for decades, with no obvious mass effects, except better telephone service. School children and pregnant women were irradiated constantly by that network, and failed to get sick, or have horrible mutations. Birds loved to make the towers home in winter time, for the warmth they could generate, but soon the tower microwave horns were fitted with plastic barrier covers that kept that to a minimum.

That microwave network has since been replaced with much higher capacity optical fiber networks, but in its day, it was buzzing us all with much more microwave energy than we now are generally exposed to, with no observable effect. But if you're concerned still, you should get on the horn to all your local TV, FM and AM radio broadcast stations, and have them shut down immediately, as they are generally the strongest source of EM radiation in urban areas, absent high power local sources like welding machines and medical imaging equipment.
posted by paulsc at 4:57 AM on August 28 [4 favorites]

If I may tag an additional, related question which may or may not be of use to the OP: What effect does metal thickness have on the answers already given? If the shielding was made of aluminium foil, or 1/8" steel, would the acceptability of hole size/spacing change?
posted by Nice Guy Mike at 4:33 PM on August 28