Hi "a"

Just a thought -- I think you are concerned with the radiation from the
illuminator adding to (or canceling) the radiation from the reflector.
Normally, the illuminator is designed to establish a pattern that aims only
toward the reflector. But, since you dont wish to shape the radiation
pattern from the illuminator, you might want to consider that the total
pattern from both the illuminator will depend on how big the reflector is
(in wavelengths).
If the parabola is large, the half of the illuminator's radiation in the
"wrong" direction wont seriously effect the total pattern, and can be
ignored. You might assume the radiation from the reflector is concentrated
to be within a 20 degree wide sector (20 dB gain antenna). The radiation
from the illuminator in that 20 degree sector is so weak that it will not be
noticeable.

The fact is - There are many factors that a designer of parabolic
antennas considers. Google

Jerry



Hi Jerry,
Thanks, yes the adding/canceling was what I was trying to address.
I'd Googled on the subject but couldn't find anything that answered that
point. I did come across some interesting sites that described the need
to illuminate the whole of the reflector to make good use of its full
dish size (and hence minimise beamwidth), but I was trying to keep mine
simple.

Treating it simplistically, I guess that, for a given diameter, a deep
parabola (ie the focus well inside the paraboloid) will capture more of
the radiated energy and redirect it in the required direction, than a
shallow parabola would.

I was thinking that I would go for a sheet of aluminium bent into a 2-D
parabola, about 2.5 wavelengths across, giving me a beamwidth of around
0.5 radians Presumably a 2-D parabola would only have gain in 2-D, and
it would be about 2*pi/0.5 = 12, (11dB over a whip).
For a 2.4GHz signal this equates to a parabola 31cm across.