Alan Pearce wrote:

> The other thing that needs to be done is to bond the inner cage to the
> outer cage at ONE POINT ONLY.  This is normally done at the point
> where the cables pass through. This means the door has to have 2
> separate connections, one for the outer cage and one for the inner
> cage.

  Dunno about the business of inner and outer cages so much, but the
critical matter of shielding, particularly as it applies to windows, is
that the shielding must be continuous between window or door, and wall,
that is there must be a continuous contact strip all the way around the
window or door, making contact to both the wall mesh and the door/
window mesh at every point.

  This is the most likely point of failure of such shielding systems.
If contact around a window or door is made at only a few points, no
matter how well it is covered with shielding, it will equate to a
complete hole of that size in the wall.

  You should also consider that unless the mesh is covered with an
absorptive layer (a wavelength thickness of graphite-impregnated IC
transport foam would be a likely material), such a conductive cavity
actually *improves* the signal to the phone by funnelling it without
loss out of/ into the opening or openings provided, i.e., the phone is
as effective anywhere inside, as if it were being used at the doorway/
window itself.

  This view of the situation may help to understand the difficulties of
the shielding approach.
--
  Cheers,
        Paul B.