[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Faraday Cage and 1/4 wavelength sized holes



Original poster: "Jim Lux by way of Terry Fritz <twftesla-at-qwest-dot-net>" <jimlux-at-earthlink-dot-net>

At 05:01 PM 12/18/2002 -0700, you wrote:
>Original poster: "davep by way of Terry Fritz <twftesla-at-qwest-dot-net>" 
><davep-at-quik-dot-com>
>
>
>>>For a Faraday cage to work for some frequency, then
>>>the holes in that faraday cage must have a diameter
>>>less than the 1/4 wavelength of the frequency you are
>>>trying to block.
>
>         That's one rule of thumb.
>         When i was doing EMI for computers to meet
>         'FCC' it was 1/10 wave.  Rather than a single
>         magic number the leakage increases as the
>         slot/opening gets larger.

Really, it's more of a perimeter thing.. a long skinny slot 1/4 wavelength 
long is just as bad as a hole 1/4 wavelength in diameter.  Doors, chassis 
seams, cover plates and panels are the traditional problem areas (hence the 
popularity of conductive elastomeric gaskets, spring finger stock, etc.)

A non-conductive slot in a conductive sheet is (EM-wise) very similar to a 
conductor of the same dimensions in an insulated sheet: Babinet's 
complementarity principle.  Slot antennas make use of this.

For TC's the slots and holes are tiny.

A bigger problem will be wires that penetrate the walls (including "woven 
wire") and carry EMI out through the shield.


>>>If you build your faraday cage out of chickenwire, sure
>>>it will keep in your arcs and lower frequency garbage, but
>>>the higher RF frequencies (Ghz) will pass right through it.



>>Do you or anyone else have an estimate on how many nanowatts
>>the average coil puts out in the GHz range?
>
>         An interesting question, which would be nice to
>         know.  Wish i did, tho i suspect its LOTS of
>         nanowatts.  (hint: in the early years, radar jammers
>         were powerful spark transmitters.  Granted, they
>         were optimized for it, and at modestly lower freqs.)
>         If I Recall, cell phones (mostly?) are in the
>         800 MHz region.
>
>         (A useful approximation for noise output might be
>         to fire up a coil, then walk away with cellphone
>         in hand until signal is acquired.  There would be
>         a huge variation, due to power variation, incidental
>         nature of 800MHz output, etc.  (cut to THAT TV
>         commercial:  Can you hear me now?  Can you hear me
>         now?  8)>>)

Probably not much, in a power spectral density sense.  The damped sinusoid 
doesn't have much power up high, but the actual sparks are pretty broadband 
(being very short duration impulses), as is radiation from the spark gap 
and connected leads.

If you figure that the spark's duration is on the order of a few tens of 
nanoseconds, then the spectrum could easily extend up into the hundreds of 
MHz range.



> > Anyone with a spectrum analyzer?
>         Agreed.  Or a cell phone, or 800MHz scanner?

If you have a computer controlled scanner like an ICOM PCR1000 with the 
ability to read the signal strength, you could fairly easily measure this, 
in a reasonably calibrated way.  There's software on the web for it.