The idea:

• move to a more  modern/reliable transmitter technology.
• remove the 430 waveguide. This would decrease the weight on the platform and the dome.
  
• One could contemplate hanging phase array feeds from the carriage house
• If these  worked, you might have wider bandwidths (although i'm not sure how then intend to cool something this big).

Simulating the beam patterns.

    Various simulations were run to look at what the beam patterns would look like for various aperture plane configurations.

The setup was:

• Use a 16k x16k floating point grid for the  simulation
• A spatial step of .125 meters was used.
  
• This gave a 2048 meter aperture plane.
• At lambda=.7 meters this gave an angular resolution of lambda/D=.7/2048 =  1.17 Arc minutes.
  
• A uniform illumination was used for the parts that were illuminated (no taper was used).
• There was no attempt at phasing the rings.

The apertures tested were:

• 150 meter filled disc.
• 150 meter filled disc with 91  meter (300 ft) hole in the center of the illumination.
• The hole in the center is from the linefeed shadowing parts of itself ( I think the hole is 300 feet..).
• 4 meter deep continuous ring at 150 meter radius.
• 2 meter deep continuous ring  at 150 meter radius
• 2 meter ring at 150 meter radius with 128 stations equally space about the ring
• 128 3meter panels covers 40% of the radius at 150 meters.
  

The processing:

• create a radius grid of 16k x 16k points
  
• populate the radius grid with the radius from the center using .125 meter resolution.
• Create a 16k x 16k aperture plane.
• using the radius grid, select all of the points in the aperture plane that should be illuminated.
  
• For the 128 equally spaced panel aperture i also used an angular grid that let me also select by angle.
• The imaginary part was set to  0.
• normalize each aperture plane to the power in the 150 meter solid disc
• 2d fft the aperture plane then compute the power giving the beam map.
  
• Extract a horizontal cut thru the beam map and scale to the peak in db's.
• all except the 128 equally spaced stations had complete rotational symmetry.
• Find the beamdwidth (FWHM), the positions of the first nulls, and the strength of the first sidelobe.
  
• compute the power in the main beam (down to the first null) and divide it by the total power in the beam map.

• Page 1: plot the horizontal cut through the beam map on a db scale.
• Limit the plot to +/- 10 deg (the beamap extends  out to +/- 150 degrees).
• Each frame contains the 150 Meter solid disc in black as a reference, as well as another aperture distribution in red.
• Top: Sold disc with 91 meter hole in center
• 2nd 4 meter ring at 150 meters
• 3rd 2 meter ring at 150 meters.
• bottom 2 meter ring at 150 meters with equally spaced stations.
• Page 2: Blowup the plots to show the close in sidelobes.
• the 4 meter and 2 meter rings lie on top of each other.]
• The beamwidths and mainBeamPower/TotalPower  is also listed.

• The image has a db stretch.
  

Summary:

• The beam width gets narrower as you cut holes in the center of the disc
• The 1st null, and first sidelobe are the same for the 4 and 2 meter rings.
• As the ring because thinner the far out sidelobes fall off faster.
• Putting a transmitter ring at 150 meters will only place about 2% of the power in the main beam. So this idea probably will not work.
• I have not tried playing with the phasing the transmitters.
• The phase will allow you to point the beam (as long as you stay away from the cliff!), but i don't think this will be able to put a lot more power in the main beam.
• The current linefeed illumination puts 68% of the power in the main beam (and the receiver also selects only 68% of the return power from the main beam).
  

 home_~phil