Phased Array Feed (PAF) Design for the LOVELL Antenna based on the Octagonal Ring Antenna (ORA) Array

Transcription

1 Phased Array Feed (PAF) Design for the LOVELL Antenna based on the Octagonal Ring Antenna (ORA) Array M. Yang, D. Zhang, L. Danoon and A. K. Brown, School of Electrical and Electronic Engineering The University of Manchester, United kingdom 1

2 LOVELL Dish Parabolic reflector D=76.2m (diameter of the dish) F=22.9m (focal length) f=1-4ghz (target) theta_0:angle at the rim of the reflector dish is about 79.5 degree 2

3 Gain of LOVELL Dish G = 4π εa λ 2 P = π2 D 2 ε λ 2 ε : aperture efficiency A P : the physical area of the aperture 3

4 ORA Unit Cell Design for 1-4GHz Element separation: 35mm Height of the structure: 42mm H1 (distance between the ground plane and active layer): 29mm H2 ( distance between the active layer and passive layer):13mm Capacitor:0.35pF Dual polarisation: Vertical/Horizontal polarisation 4

5 ORA Unit Cell Design for 1-4GHz contd db S11 S Frequency(GHz) Return Loss > 10dB from 1-GHz 5

6 Size of ORA for LOVELL Main beam direction theta GHz 4GHz 2.5GHz BDF formula Offset(m) With the scanning angle of 0.3 degree, there is an offset of 0.2m from the focus of the LOVELL dish, therefore the radius of the finite ORA (in circular shape) needs to be 0.2m at least. Beam Deviation Factor (BDF), Antenna theory and design by Stutzman and Thiele, pp.341

7 Circular ORA for LOVELL The finite ORA has 17 elements across its diameter with two edge elements loaded for the purpose of impedance match In total there are 117 elements used (white squares)and yellow squares indicate the loaded elements Size of the ORA:0.595 (m) x (m)

8 Radiation Pattern of a Horn Antenna (reference) 0-2 Normalized radiation pattern (db) Angle of incidence (degree) A horn antenna with a taper of -12dB at the rim of the reflector is used in GRASP across the whole frequency range Good illumination efficiency about 75% 8

9 Proposed Excitation of ORA The central single element excited from 2.2GHz-4.0GHz A good illumination efficiency (horn antenna as a reference) when used to illuminate LOVELL Multiple elements excitation has lower illumination efficiency, therefore results in lower gain for the LOVELL 9

10 Normalized Radiation Pattern of the ORA (CST) Frequency (GHz) Aperture taper efficiency (t) Spillover efficiency (s) Illumination efficiency of LOVELL (t*s) Horn

11 Radiation Pattern of (GRASP) Phi=0 Phi=90 11

12 Radiation Pattern of (GRASP) Phi=0 Phi=90 12

13 Radiation Pattern of (GRASP) Phi=0 Phi=90 13

14 Radiation Pattern of (Co-polar) Phi=0 and Phi=90(GRASP) 14

15 Radiation Pattern of (Co-polar) Phi=0 and Phi=90(GRASP) 15

16 Radiation Pattern of (Co-polar) Phi=0 and Phi=90(GRASP) 16

17 Overlapping of the Beamwidth at 2.2GHz, 3GHz and 4GHz For PAF, it is desirable that the HPBW of one beam is overlapping with the HPBW of the next beam This will be able to offer a larger field of view therefore more space of sky can be observed at once, which is one of the advantages of using PAF to feed the reflector dish instead of using a traditional horn antenna 17

18 Commercial LNA(ZX60-P33ULN+, MMIC) 18

19 LNA based on ATF (Avago discreet) 19

20 Steady State Heat Flow (Whole PAF) 30 o C ambient, 400 W/m 2 solar flux Solar Heat, W Convection -303 W Radiation -713 W LNA Heat, +29 W Heat from sides, and support structure, +67 W Radiation, +17 W Convection, +16 W Co-axial cables, +6 W 20

21 Cooling System Installation Cryogenic PAF Lovell Radio Telescope Cryogenerator and Water Chiller Vacuum Jacket Piping 21

22 Future integration work 22

23 Future Work Illumination of the LOVELL antenna by single ORA element excitation is not being optimal and can be further optimized. The whole point is we don t just uniformly excite the array, we will allow the beamformer to use non uniform weights thus optimising the Array feed pattern. This work was not done yet so it is important to say this is preliminary results and not yet optimised. We expect significantly improved results with weight optimisation. For ORA PAF design, the optimal excitation scheme needs further investigation. The goal of the optimization will be to shape the ORA array pattern to get close to a Gaussian beam of an idealized horn antenna. Because the pattern of the Gaussian beam is independent of the frequency, this poses a challenge for the design of ORA PAF. Generally the pattern of finite array will change as frequency: the gain increases with increasing frequency and it results in a narrower HPBW if the number of elements being excited remains the same across the whole frequency range What might be a possible solution to this challenge is to excite different number of elements for the ORA PAF at different frequencies. A general rule is to excite smaller number of elements at higher frequencies than that at lower frequencies. The optimization of the excitation of the ORA PAF could be something which can be taken forward for future work 23