Faraday cage question

[McBanjo]

Is it enought to get a faraday cage effect if I put wire on each side in a slalom-way with ~1cm between like this:

Code: Select all

    _
 | | | |
 |_| |_|

And does the sides have to be connected?

Edit: Fixed the drawing, looks very weird in the message but proper when posted

[klankymen]

try to put your drawing into

Code: Select all

code

tags...

[McBanjo]

try to put your drawing into

Code: Select all

code

tags...

Thanks

[jaganath]

Where is the thing you are trying to shield in this diagram?

[McBanjo]

Where is the thing you are trying to shield in this diagram?

It's not there. I am asking about how the shield can be constructed rather than the shielded object. Am I explaining it good enought?
It's a tower case out of wood. Each side will probibly not be connected to eachother unless it's very needed.

[jaganath]

AFAIUI the shield should have no holes in it larger than the wavelength of the radiation it is trying to contain/protect against.

http://www.cv.nrao.edu/~rfisher/Papers/rfi_sumschool.ps

The spectrum of RFI from computers and other digital equipment is a mixture of narrow and broadband emissions with the narrowband spikes being spaced at intervals from kilohertz to many megahertz. There is usually some periodicity to the narrowband frequencies, but there will also be a jumble of spikes that don’t fit any well-defined pattern. A common preconception is that computers radiate primarily at their CPU clock frequency, but the intensityat this frequency is often no higher than at a lot of other frequencies in the spectrum. Most digital RFI is strongest below a few hundred megahertz, but strong digital noise can be seen to several gigahertz and above.

[McBanjo]

AFAIUI the shield should have no holes in it larger than the wavelength of the radiation it is trying to contain/protect against.

http://www.cv.nrao.edu/~rfisher/Papers/rfi_sumschool.ps

The spectrum of RFI from computers and other digital equipment is a mixture of narrow and broadband emissions with the narrowband spikes being spaced at intervals from kilohertz to many megahertz. There is usually some periodicity to the narrowband frequencies, but there will also be a jumble of spikes that don’t fit any well-defined pattern. A common preconception is that computers radiate primarily at their CPU clock frequency, but the intensityat this frequency is often no higher than at a lot of other frequencies in the spectrum. Most digital RFI is strongest below a few hundred megahertz, but strong digital noise can be seen to several gigahertz and above.

Heh how long is a computer wavelength in the metric system?
Guess wire isn't a realisticly working solution then.

[floffe]

The wavelength of EM radiation at 2GHz is 15cm (6"), and 1MHz gives 300 metres, but I'm not sure how much smaller it has to be (wikipedia says "significantly smaller")

[McBanjo]

The wavelength of EM radiation at 2GHz is 15cm (6"), and 1MHz gives 300 metres, but I'm not sure how much smaller it has to be (wikipedia says "significantly smaller")

That far? I assumed it was nano-size. Anyone know specificly about the length a compter is putting out and how much that's in "normal" length?

[Reachable]

How about using window screening -- insect screening -- a couple of layers perhaps? That would create smaller holes than in grids I've seen that would block approximately gigahertz or cellphone frequencies.

I don't speak with any authority on this -- just throwing out an idea. The quote was "Most digital RFI is strongest below a few hundred megahertz, but strong digital noise can be seen to several gigahertz and above." How important is this several gigahertz noise -- excluding the noise from the processor -- and what would it interfere with?

[cmthomson]

The wavelength of EM radiation at 2GHz is 15cm (6"), and 1MHz gives 300 metres, but I'm not sure how much smaller it has to be (wikipedia says "significantly smaller")

That far? I assumed it was nano-size. Anyone know specificly about the length a compter is putting out and how much that's in "normal" length?

The typical peak EMI from a modern computer system is in the 300 MHz to 1 GHz range. This is usually suppressed by careful routing of signal traces on the motherboard, and coupling to ground planes within the motherboard. In PCs, the goal is to have essentially no EMI even with no case, since so many consumers compromise the case shielding for various reasons.

[McBanjo]

How about using window screening -- insect screening -- a couple of layers perhaps? That would create smaller holes than in grids I've seen that would block approximately gigahertz or cellphone frequencies.

I don't speak with any authority on this -- just throwing out an idea. The quote was "Most digital RFI is strongest below a few hundred megahertz, but strong digital noise can be seen to several gigahertz and above." How important is this several gigahertz noise -- excluding the noise from the processor -- and what would it interfere with?

I've been looking at net but it was rather expensive (~10-20USD for 1m). Atleast what I've found.

The typical peak EMI from a modern computer system is in the 300 MHz to 1 GHz range. This is usually suppressed by careful routing of signal traces on the motherboard, and coupling to ground planes within the motherboard. In PCs, the goal is to have essentially no EMI even with no case, since so many consumers compromise the case shielding for various reasons.

Great, then I simply ignore the imaginary need of shielding.
My current box is open and I haven't noticed any problem. Saves me a lot of work