You certainly will have to deal with the realities of the building in
modeling the antennas up there.
You haven't mentioned the composition of the roof. If it is metal, or has
relatively dense metal bracing under it, or is concrete with rebar in it,
you might want to model the roof as a near perfect ground underneath.
Certain programs will allow you to place more than one ground structure
beneath it, e.g., near perfect ground for the first 20 meters, and poor
ground 30 meters lower after that. This can be done in EZNEC-2, as one
If the roof is not conductive, or except for fairly wide spaced steel
crossbraces, then any metal content of the building within 30-50 meters of
the antenna can affect the pattern and/or impedance of the antenna in highly
unpredictable ways. That, of course, is an antenna modeler's nightmare.
Try modeling your antenna as if the roof were a perfect ground and see if
the predicted impedance approaches what you measure, if so then try the
first scenario I described.
An anecdotal aspect, when I was K3FKJ in Washington DC in the early 60's, I
lived in a row house. The roof of my house and the neighbor's roofs up and
down the block were tar covered tin and copper. I had a multiband vertical
with radials about a foot above the roof. It was the only vertical I ever
used that really worked well. Done right in good circumstances, the roof of
a building can do very well as an antenna farm.
Good luck, 73, and a prosperous New Year.
Guy L. Olinger
Apex, NC, USA
From: Igor Sokolov <firstname.lastname@example.org>
To: email@example.com <firstname.lastname@example.org>
Date: Tuesday, December 29, 1998 1:40 PM
Subject: [antennaware] Antennas over steel and concrete
>I have to put all my antennas on the roof of a block of flats. This is a
>building made of steel and concrete. The roof is about 30 meters over the
>ground 15 meters wide and 50 meters long.
>1) When modeling antenna over the roof should the height of the roof be
>taken into account?
>2) Should the conductivity and other properties of the roof be accounted
>3)What sort of influence in general there is from the roof onto antennas
>different freq. from 3.5 to 28 Mhz?
>All these questions emerged when I took 40 meters Delta Loop fed for
>vertical polarization that worked perfectly over the ground and put it on
>the roof (the bottom is about 3 meters over the roof and say 35 meters
>the ground). The impedance of the antenna changed from 60 ohm to 160 ohm
>with a lot of reactive component.
>I would be glad to get any comments on the problem.
>FAQ on WWW: http://www.contesting.com/antennaware-faq.html
>Administrative requests: antennaware-REQUEST@contesting.com
FAQ on WWW: http://www.contesting.com/antennaware-faq.html
Administrative requests: antennaware-REQUEST@contesting.com