Daily Image
02-03-2007
Sculpting the FPA beamshapes
| Submitter: | Tony Willis |
| Description: | In a previous submission (16-01-2007) we showed a MeqTrees focal plane array simulation where we formed focal plane array beams by means of a simple algorithm called phase conjugate weighting. This method forms beams which have a high gain in a specific direction, but the beams are not necessarily symmetric and may have high sidelobes. Astronomers would prefer to observe with a beam having a nice Gaussian shape and minimal sidelobes. So we extended the MeqTrees simulation by providing the weights obtained by phase conjugate weighting as a 'first guess' to the MeqTrees Solver. The Solver takes these weights and adjusts them until the formed beam has a reasonable resemblance to a Gaussian. This sequence of screenshots shows the MeqTrees solver in action. The first frame shows an off-axis focal plane array beam formed by the phase conjugate weighting scheme. Notice the asymmetric shape of the central lobe and the high sidelobe. The remaining frames show the beam being adjusted by the Solver as it attempts to form up a Gaussian beam. It takes 18 iterations to do so but the end result (frame 19) is a beam with a much better approximation to a Gaussian shape. The intensity levels are plotted on a logarithmic scale as we wanted to emphasize the changes that take place at the lower intensity and sidelobe levels. On a practical note one could take the final weights obtained by this type of simulation and use them in a real beamformer mounted on a telescope. |
| Copyright: | Tony Willis / MeqTrees Foundation |
| Tweet |  |