"a" wrote in message
...
snip


Thanks for the replies.

I agree that the radiating element must be placed at the parabola focus
to give minimal beamwidth, and that this condition is met when the
radiating element is placed at the focus (which is given by D^2/16d).

The point remains that I can still choose the parabola parameters to set
the focal length to whatever is desired.
Should I choose them so that the focal length is an odd or even number
of quarter wavelengths?

I don't think it matters. The gain signal off the parabolic
reflector will be so very much stronger -- except for a small antenna. Do
you have the gain formulas? http://en.wikipedia.org/wiki/Parabolic_antenna


What I really had in mind was a uniformly radiating element (ie a simple
whip) with a parabolic reflector behind it, like this:-
( x
reflector radiating element

To get the right-going signal from the reflector in phase with the
right-going signal from the radiating element I need to choose the
reflector distance correctly.

This is an unusual concern; it's not common practice to
have an "omni" feed for a parabolic.

I have a feeling that there WILL be a phase inversion at the reflector
but I'm not certain.

The reason that I think that there might be a phase inversion is that
the (radiator plus reflector) could be considered to be a (radiator and
its image). At the (perfectly conducting) reflector the voltage will be
zero and the current will be infinite, which implies that, at the
reflector, the reflected wave must be phase inverted wrt the radiated
wave.

Any thoughts?