On Tue, 01 Dec 2009 03:42:13 -0600, Lostgallifreyan
wrote:

why is it
often ok for a Faraday cage to have holes in it? Braided screens, meshes,
perforated metal sheets, etc, I've seen many shields that are not a complete
'seal'... UHF TV cables especially seem to be very loosely shielded but they
work.

This can be explained at super high frequency and at DC as easily.
However, before that it should be pointed out that the coverage (the
ratio of what is conductor to what is not - the air space) defines how
"good" the faraday shield will be. Not surprisingly, coverage is
wavelength dependant. To cut to the chase, a wide mesh will allow
increasingly higher frequencies (shorter waves) through.

Now, as to the how. With a separation in the mesh, and for very large
wavelength (in proportion to the opening size), you will have a very,
very small potential difference across any of the mesh openings. Very
little potential voltage across the mesh opening means very little
current flow around the mesh opening that is specifically due to that
potential difference.

This is not to say there isn't a very, very large current flow by
virtue of some very, very long wave. No, there's no denying that, but
to get through the mesh you have to satisfy local conditions that
demand what amounts to leakage (and this is exactly the term that
correlates to coverage when discussing coax weave). If that huge
current cannot induce a significant voltage across the mesh opening,
then the mesh opening loop current cannot induce a field through to
the other side. Now, if you examine the context of "huge current" in
a resistive conductor, then obviously a potential difference can
occur. Point is that reality (and science) allow for poor grade
shields, but as a one knock-off proof you can summon up any failure,
ignore simple contra-examples and create a new theory.

However, returning to what is well known. If you increase the
frequency applied to the mesh, then at some point wavelength will
allow a situation where the general current flowing through the whole
structure will naturally exhibit a potential difference across some
small scale. By this point, abstraction may be wearying.

Let's say you have a 10 meter-on-a-side cage with 1 meter mesh
openings. If your applied field were exciting the cage at 75MHz (4M),
then any spot on the cage could be at a very high potential difference
from any spot adjacent and 1 meter away (a simple quarterwave
relationship). This works for a solid conductor, it works for a mesh
conductor.

The 1 meter mesh openings can thus exhibit a substantial potential
difference across the opening, and a local current loop associated
with that potential difference. The mesh opening becomes a
quarterwave radiator (aka slot antenna) and can couple energy from the
external field into the interior of the cage (now possibly a resonant
chamber, aka RF cavity). In practice and literature, the mesh opening
loop exhibits a radiation resistance of 10s of Ohms. That compared to
its mesh loop Ohmic path loss, makes it a very efficient coupler of
energy.

Take this very poor example of mesh, and lower the frequency to 750
KHz. The mesh opening - if we originally likened it to an antenna, we
should be able to continue to do that - is now 1/400th Wave. A
1/400th wave radiator has extremely small radiation resistance. The
exact value would be 751 nanoOhms. As we are examining a poor mesh,
it becomes clear that it must have some resistance over that 1 meter
distance (this is a real example, after all).

Being generous and constructing that cage out of rebar will give us a
path resistance of, luckily, 1 milliOhm. This figure and that of the
radiation resistance yield the radiation efficiency (that is, how well
the exterior RF will couple into the interior) which reduces to
0.075%. The cage works pretty well, but not perfectly (it was, after
all, a poor example).

Now, repeat this with a poorer conductor, or a tighter mesh and
imagine the shielding effect. The mesh has an opening radius
squared-squared relationship driving down the radiation resistance
compared to the linear relationship of conductance.

*************

Now, expanding the topic to allow for the contribution of ALL openings
in the mesh, we must again return to the physical dimension compared
to the wavelength dimension. If the cage is truly large, larger than
the field exciting it, then you have miniscule radiators along it,
each very inefficient. However, each of those radiators is out of
phase with a distant neighbor (not so with its close mesh neighbors).
Those two wavelength distant mesh radiators will combine somewhere in
the interior space and build a field. This is very commonly found in
inter-cable cross coupling through leakage that is exhibited in very
long cable trays with tightly bound lines. This doesn't improve the
efficiency, but sensitive circuits running parallel to power drives
can prove to be a poor combination. What to do when conditions
condemn the small signal coax to live in proximity to the large signal
supply?

This introduces the foil shield. The foil shield is a very poor
conductor over any significant length, but over the span between mesh
openings (e.g. coax shield weave), the resistance is sufficiently low
to close the conductance gap.

73's
Richard Clark, KB7QHC