This is one EMI problem I have never had the pleasure of working on :-)
You did a good job stating the clear cut ones of possible near-field
effects to your antennas.
If I lived in a house with a metal roof, I would be worried about three things.
Whether they would every happen or not is another matter.
But you know the old saying about the condition of a mechanic's car :-)
The two issues
1) If the sections of the roof are not well bonded, but over time due
to some detail of fabrication or assembly (use your imagination), you
start getting intermittent contact between sections- then you hear
"sparky noise" in your receiver. You could track it down, but it would
be a pain and I think I would rent a man-lift with your roof pitch!
2) At higher frequencies, roof panels could provide capacitive
coupling from noise generated inside your house to your outside
antennas, within a wavelength distance or so. Unless you have holiday
lights on your roof, or other wires up against the inside surface of
the roof (there ought to be insulation there anyhow), this shouldn't
be a big deal. "Grounding" the roof might not help at all, if your
roof is more than a few inches off the ground (ha ha). Even if it were
that low, it would be very impractical if the roof contains many
insulated epoxy coated sections.
3) "Rusty Nail" non-linear mixing action, due to poor contact between
metal pieces and nails, This could be an "interesting problem",
particular in high field strength areas.
#2 should be avoidable. #3 sounds scary #1 could happen right away
with roof flexing due to temperature, wind, etc. I wonder if anyone
has any precise observations of this happening in the field.
TowerTalk mailing list