RE: ATM Spider Design

["Alan Levin" <ailevin@netcom.com>]

Andy,

Actually it is Alan, not Neil.  You don't have to really understand 
FT.  There are just a few basic principles.

1. You add and subtract the diffraction patterns of the openings and
blockages respectively and then square the result to get the
intensity pattern.

2.  Larger openings and blockages have more narrowly spaced 
diffraction features, and smaller openings and blockages have more 
broadly spaced diffraction features.  Obviously larger opening pass 
more light and larger blockages block more light.

3.  Large and small are measured in terms of the wavelength of light. 
(World class optical telescopes have small apertures relative to 
radio waves.)

4.  The diffraction pattern of almost any shape can be understood to 
first order by taking cross sections or cuts through the shape and 
thinking about the diffraction pattern along that line.

5.  The diffraction pattern of a uniformly illuminated line segment
(hard cut off at the end) is a peaky/troughy thing called a sync
function that looks like sin(x)/x.

6.  The diffraction pattern of a line segment with a gaussian
illumination (smooth roll off at the end) is a nice smooth gaussian. 

When people say that diffraction is due to edges, they probably mean
the fringe patterns characteristic of the sync function.  Based on
5. and 6. above you can get rid of the fringes by smoothing out the
sharp edges, but you still have a diffraction spot related to size
of the aperture.  In fact, the diffraction spot for a guassian
illumination is broader, so you give up resolution when you suppress
the fringes.  In radar this is called sidelobe suppression. 


Good Observing,
Alan