Acrylic Cases and EMI

[nx74205]

Hello,

I just recently ordered an Acrylic case to put together a computer with some parts I had just laying around from my last build. However, I perused on the internet today and found out that these cases do no adequately shield EMI and can play havoc with other nearby electronics. How true is this and how can I mitigate the EMI? Will I just have to buy another case?

 

[vvhocare5]

What are you worried about? Are you EMI sensitive?

 

[shovenose]

well just dont put it right next to ur monitor if its one of the old crt types. it may cause interference like maybe if u stick a wifi router right next to it. but since many people put their comps on the floor next to them it doenst really matter. ps:

 

[RetiredChief]

No sweat, give and take - it not only sends out emi (radiated emissions) there is also susceptibility to EMI.

Most likely no problems. Remember the strength of the field varies inversely proportional to the Sqaure of distance.

 

[arterius2]

you can reduce EMI interference by enabling "Spread Spectrum" in the BIOS, usually there is an option for both the CPU and Mobo. Most motherboard BIOS have this option as stated by the FCC regulation.
This option allows the oscillator to spread EMI frequencies over a larger spectrum thus to reduce the spectral density of the electromagnetic interference (EMI) that these systems generate.

 

[shovenose]

thats what spread sepctrum does? cool!

 

[arterius2]

thats what spread sepctrum does? cool!

yes, it's pretty neat... if you work for the CIA or something. (These guys are pretty sensitive about their equipment)

 

[nx74205]

Thanks everyone for your responses...so basically i have nothing to worry about which is cool..

 

[nx74205]

No sweat, give and take - it not only sends out emi (radiated emissions) there is also susceptibility to EMI.

Most likely no problems. Remember the strength of the field varies inversely proportional to the Sqaure of distance.

Do you know the exact equation.....just curious.I don't remember from my physics 2 class..

 

[Paperdoc]

Almost ALL commercial computer cases today actually have an invisible electrically-conductive coating sprayed on the inside. Although it is not as good as a solid metal case, it substantially reduces the penetration of electromagnetic noise signals through the case. This means BOTH noise going out from your mobo and components, and noise entering from outside. If you get a case of whatever material (acrylic or not) that does NOT have this feature, you are risking more problems from such noise.

"Spread Spectrum" on mobo's apparently refers to a way that the clock pulses (and some data) are formed and used. Clock pulses normally are nearly square waves, and this means they contain some significant high-frequency harmonics of the base clock rate. These can manifest themselves as noise sent out of your computer. Setting the Spread Spectrum option limits this effect by sort of "rounding off" the square corners of the signals. The result in some cases MAY be that the system does not work so well if you are trying to run near the edge of performance by overclocking, etc. For most systems run under stock conditions it ought not to create a problem. HOWEVER, this only deals with half the problem - the noise sent out of your machine. It does nothing to protect your mobo from outside noise sources.

 

[arterius2]

Almost ALL commercial computer cases today actually have an invisible electrically-conductive coating sprayed on the inside. Although it is not as good as a solid metal case, it substantially reduces the penetration of electromagnetic noise signals through the case. This means BOTH noise going out from your mobo and components, and noise entering from outside. If you get a case of whatever material (acrylic or not) that does NOT have this feature, you are risking more problems from such noise.

"Spread Spectrum" on mobo's apparently refers to a way that the clock pulses (and some data) are formed and used. Clock pulses normally are nearly square waves, and this means they contain some significant high-frequency harmonics of the base clock rate. These can manifest themselves as noise sent out of your computer. Setting the Spread Spectrum option limits this effect by sort of "rounding off" the square corners of the signals. The result in some cases MAY be that the system does not work so well if you are trying to run near the edge of performance by overclocking, etc. For most systems run under stock conditions it ought not to create a problem. HOWEVER, this only deals with half the problem - the noise sent out of your machine. It does nothing to protect your mobo from outside noise sources.

um... actually no, spread spectrum looks something like this graphically:

Without Spread Spectrum

With Spread Spectrum

In short, Spread Spectrum spreads EMI across wider range of frequencies. your whole "rounding off corners of boxes" concept didnt really make much sense. and spread spectrum usually impede OC capabilities.

 

[RetiredChief]

reply to NX74205
Extract from
http://www.osha.gov/SLTC/radiofrequencyradiation/electromagnetic_fieldmemo/electromagnetic.html
"The inverse-square law is defined as: "A statement that the strength of a field due to a point source or the irradiance from a point source decreases as the square of the distance from the source. Note: For sources of finite size this gives results that are accurate within one-half percent when distance is at least five times the maximum dimension of the source (or luminaire) as viewed by the observer."

There are some other considerations, ie the frequence as it relates to the wavelenght of the "E" field.

 

[nx74205]

Almost ALL commercial computer cases today actually have an invisible electrically-conductive coating sprayed on the inside. Although it is not as good as a solid metal case, it substantially reduces the penetration of electromagnetic noise signals through the case. This means BOTH noise going out from your mobo and components, and noise entering from outside. If you get a case of whatever material (acrylic or not) that does NOT have this feature, you are risking more problems from such noise.

"Spread Spectrum" on mobo's apparently refers to a way that the clock pulses (and some data) are formed and used. Clock pulses normally are nearly square waves, and this means they contain some significant high-frequency harmonics of the base clock rate. These can manifest themselves as noise sent out of your computer. Setting the Spread Spectrum option limits this effect by sort of "rounding off" the square corners of the signals. The result in some cases MAY be that the system does not work so well if you are trying to run near the edge of performance by overclocking, etc. For most systems run under stock conditions it ought not to create a problem. HOWEVER, this only deals with half the problem - the noise sent out of your machine. It does nothing to protect your mobo from outside noise sources.

Thanks for the info...how can I find out if my case was sprayed with this coating?

 

[Paperdoc]

arterius2, you did not understand what I said. If you analyze a square wave using a Fourier transform, you find that it actually is composed of the sum of an infinite series of sine waves. The lowest-frequency component is the fundamental square wave frequency; added to it are progressively smaller proportions of the harmonics of that. In effect, broadcasting a square wave amounts to sending out many frequencies, not one. But if the square wave generated is altered so that its higher frequencies are reduced in amplitude (attenuated), that "noise" at higher frequencies is removed. The appearance of the "square" wave so formed is less square, especially at its sharp "corners" where the transition from one state to another occurs. That is why I used the phrase "rounding off". However, this has two impacts on the signal and on circuits that use it for a clock (timing) signal. One is that the transitions from on to off, or the other way, are no longer almost instantaneous - they now actually take enough time to affect the way a digital circuit changes state. Further, it also means that the square wave spends less time in a single stable state (on or off), so any work a circuit does that depends on how long the "on" state lasts may be affected. As we agree, that can impact operations of circuits that are being pushed to their performance limits, such as when one overclocks the computer.

 

[Paperdoc]

OP, you should contact the manufacturer of the case you want to use and ask them what they have done to reduce EMI interference reaching the computer mounted inside, and its impact on surrounding devices. There are FCC regulations for this, and they may be able to tell you that their case does or does not meet FCC regs for EMI interference. Truth is, you don't care exactly what they've done if the FCC requirements are being met.

Of course, you can just not worry about this question - your choice. You just need to understand the risks, maybe just small risks, of using a case without this feature.

 

[jsc]

arterius2, you did not understand what I said. If you analyze a square wave using a Fourier transform, you find that it actually is composed of the sum of an infinite series of sine waves. The lowest-frequency component is the fundamental square wave frequency; added to it are progressively smaller proportions of the odd harmonics of that.

If it makes any difference.

 

[Paperdoc]

jsc, thanks. I thought I recalled that it was only the odd harmonics, but was not sure if that was square wave or another waveform. Don't have my old reference books here to check.

 

[RetiredChief]

Yes "If we continue with this series, adding a sine wave at 5 times the frequency and 1/5th the amplitude, 7 times the frequency and 1/7th the amplitude and so on up to 31 times the frequency (and 1/31 the amplitude) you get a waveform that looks like Figure 9.4. "

Figure 9.4 = sq wave