Phased Array Feed Design. Stuart Hay 23 October 2009

Transcription

1 Phased Array Feed Design Stuart Hay 23 October 29

2 Outline Why phased array feeds (PAFs) for radioastronomy? General features and issues of PAF approach Connected-array PAF approach in ASKAP

3 Why PAFs? High sensitivity Wide field of view (FoV) (instantaneous and contiguous) Survey speed (η ap A phy /T sys ) 2 x Ω FoV Possibilities for discovery (eg transient phenomena) Large frequency range (.7-1.8GHz)

4 PAFs in general Concentrator (optics) <λ min /2 Beamformer (digital) Low-noise amplifier (LNA) +Digital conversion +Polyphase filter Beams (weighted sums of inputs)

5 PAFs vs existing multibeam feeds Beams (colors) and combined correlations (black) Beams (colors) and combined correlations (black) Sensitivity (m 2 / K) Sensitivity (m 2 / K) Angle (deg) <λ min / Angle (deg) >λ max /2 Beamformer (digital) PAF gives complete high-sensitivity sampling of FoV

6 PAFs vs existing multibeam feeds Sensitivity (m 2 / K) Noise temperature (K) Beam T sys Amplifier T min Number of beamformer inputs Number of beamformer inputs PAF has significant mutual coupling effects However beam T sys amplifier T min with optimum (active) noise matching of array/amplifiers and optimum beamforming

7 PAF optics - focal vs image plane focal plane D f image plane #elements α (f/d) 2 and area of FoV Compromise between f/d and η Can beamform subsets of elements Can expand FoV by adding more elements (upgrade path) #elements determined by magnification All elements must be beamformed for good η FoV limited by subreflector size

8 PAF optics - beam stability Beam stability with respect to the astronomical sources is required for high dynamic range image formation Possibly solutions in clear-aperture (offset-fed) optics and electronic beam rotation through beamformer weights ASKAP will use a 3-axis antenna design

9 PAF optcis - beam stability

10 PAF/LNA issues Uncooled LNAs with T min in 2-4K range over.5-2ghz have been developed T sys /η ap < 7K realistic target that would make PAFs competitive Cooling PAFs is not straightforward Many coupled elements distributed over large area However should be further considered Differential LNAs and LNAs with >5Ω noise-match impedance are desired for some PAF designs Reduce balun loss Minimize LNA T min LNA modelling and measurements require further work

11 Connected-array approach to PAF Broaden PAF investigation Previously focussed on Vivaldi (ASTRON and DRAO) Planar connected arrays Alternative viewpoint emphasizing mutual coupling Possible advantages of planar and low-profile structure Cost Loss Integration Noise coupling? Cooling? Other? Prototype connected-patch FPA

12 Connected-array approach Patches Transmission lines Currents Digital beamformer Current continuity enhances bandwidth (element spacing <λ/2) Target.7-1.8GHz frequency range 9x1x2 elements for ~3 sq deg FoV D=12m front-fed reflector with f/d=.5 Ground plane

13 Approach to the design Develop modelling capability Electromagnetic modelling of array Electronic modelling of LNA Numerical and experimental investigations of prototype Resonances and matching to LNA Radiation pattern in chamber Parkes 12m testbed Optimize larger design for ASKAP Power (db) CBFM pol. 1 CBFM pol. 2 GEMS pol. 1 GEMS pol. 2 CBFM pol. 1 (no diel.) CBFM pol. 2 (no diel.) θ (deg) x (m) x (m) y (m) y (m).1.8.6

14 Connected-array matching 7 Patch Patch Impedance (ohm) real Z opt imag Z opt real Z in imag Z in ZSE ZSE A (V i+ -V i- ) V i + V i - Groundplane ie Z C =Z SE / 2 Z D =Z SE x Frequency (GHz) Z diff ~377Ω LNA optimum noise-match impedance in mutually coupled array environment Lower Z diff modifications to array will lessen sensitivity to parasitics

15 CSIRO ICT Centre Stuart Hay Phone: Stuart.Hay@csiro.au Web: Thank you Contact Us Phone: or enquiries@csiro.au Web: