EMI, as Electro Magnetic Interference is a bit of problem, when we talking about PC audio card. Of course, there is a hell lot of EMI emiters inside of the PC, for example the HDD and the GFX card just above the poor soundcard.

Now my question is, witch material is most effective to use - eg. install on back of the audiocard (in PC it will be on top, facing the GFX card). I do believe that the best material is reasonably thick, grounded iron plate.
I did that (just place a 3-4mm thick isulator pad - like the thick 4mm thick material some mainbards are placed into in their boxes - and above it screw the plate) once on my SoundBlaster 5.1 Live!, however the card was nowhere near the need quality so I did not actually notice any improvent...
...but with my X-Fi I think it maybe should be measurable at least with the Right Mark Audio.

The material choice question poped out, when I saw some professional Japan Onkyo soundcards and they seems to use cooper as shielding - or what it is?





Now I think about iron, because it is magnetic - unlike cooper - and therefore (I think) should filter the magnetic waves way better that cooper. Sure, cooper is way better in conductivity, but it is not like the EMI create that huge currents in my PC/shielding plate, so I need cooper to conduct them.
Or maybe I'm entierly wrong and the Onkyo use these cooper parts for other reasons, unknown ATM to me?

Neverless I'm curious what do you guys think about best material suitable for EMI shielding of soundcard in PC

Hello Trodas,

How much magnetic based intereference to expect to receive? I imagine most of the interference is electrical/radio based, in which case, copper is probably best. Consider cable manufacturers: The shielding that they use in coaxial cables is almost always copper.

Any reasonably conductive metal will block electrical interference. Magnetic is a lot harder though, and copper won't do anything for that.

Personally, I doubt you will get an improvement this way. Some of the best RMAA numbers ever posted are PCI cards, which implies that the limitation is elsewhere, or else firewire cards would consistently beat them.

I read somewhere around here that the Lynx two card is very well shielded, but I don't know how the shielding is achieved there.
I'd also like to know if there are good ways to shield PCI cards, simply because I'll likely buy a decent PCI card soon and it would be great if I could minimize bad effects inside the tower (usb- and firewire-cards seem to be more of a problem using linux).

FWIW, I use a USB DAC with Linux, and it works perfectly (and I'm quite happy with it too). I'm not sure how your experiences have gone, but the system might have gotten better; opensource moves pretty rapidly like that.

I don't have experiences with usb-devices myself, I took that information from different sources on the internet. My distros wiki has few usb-devices listed as working and the ardour page states that usb isn't suitable for low-latency, on the other hand, there are quite a few usb-devices listed as working on the alsa-page, so I have to appologize for my possibly wrong conclusion.
I don't really know how well firewire-devices work, some are listed as working at the alsa page, but I have no information on latency.

Does anyone know how serious the EMI and or the electromagnetic "problem" inside a typical tower is (relating to soundquality)?

I guess it is in the implementation, but on my Lynx card, the sound quality is awesome. My Dac is about $400 more expensive than the Lynx, and I'll take the lynx sound quality anyday.

I am sure the problem is real, but it can be solved.

Also, if someone could explain why audio would suffer in this environement while no one is freaking out about video cards in such a terrible environement.

Exactly!

Any electromagnetic interference (EMI) that is comming directly through the air is likely to be way outside the audio band so that's why you don't generally see this type of sheilding on sound cards (not even extremely good ones). Now look at TV tuner cards and you'll see even the cheapest ones employ some type of shielding, radiated EMI is far more a problem in the RF band than the audio band. Personally I think the OP is wasting his time trying to retro-fit this type og EMI shield to an audio card.

With audio it's far more important to have a board with well designed power supply layout and decoupling (bypass capacitors) than to have EMI shielding to protect against radiated noise.

Electromagnetic waves consist of an E field and an H field, where the E field can be thought of as electrostatic and the H field as magnetic. The requirements for shielding these are quite different.

To shield a magnetic field you need a material that is magnetic, such as iron, but to shield an electrostatic field requires high electrical conductivity.

However, it turns out that the E field is most significant at higher frequencies and the H field at lower frequencies. To shield against high frequencies you want a good conductor, such as copper or silver, and at very high frequencies the thickness is not important because the conduction is almost entirely at the surface.

All of this has very little importance in your case because the primary source of EMI is likely to be conduction from the back plane into the card, and unless you can insert something like a large ferrite core at the connector to the main board, there is not much that you can do. It is up to the designer of the card to anticipate this problem and solve it.

My appologies if I got any of the terminology incorrect. It has been several decades since I learned this stuff.