FIDA3: A Novel Active Array for the Mid-SKA

Transcription

1 : A Novel Active Array for the Mid-SKA O. García-Pérez FG-IGN oscar.perez@oan.es J.A. López-Fernández, D. Segovia-Vargas, L.E. García-Muñoz, V. González-Posadas, J.L. Vázquez-Roy, J.M. Serna-Puente, E. Lera-Acedo, T. Finn, R. Bachiller and P. Colomer O. García-Pérez (FG-IGN) 1

2 Overview Introduction prototype Radiating structure Bunny-ear antennas Scan anomalies Array measurements Amplifiers LNA design 1 LNA design 2 Conclusions O. García-Pérez (FG-IGN) 2

3 Introduction (FG-IGN Differential Active Antenna Array) is an active array prototype developed by the FG-IGN for the task DS4-T4 of the SKADS project. It should meet the next requirements: Bandwidth: 3MHz-1MHz Low cost Dual polarization Scanning capabilities up to +/-45º Noise temperature as low as possible The proposed solution provides the next advantages: Dielectric-free antennas: avoid the losses and cost of the substrate Differential feeding: avoids the losses and bandwidth limitations of passive baluns O. García-Pérez (FG-IGN) 3

4 prototype Antennas Low noise amplifiers Box structure Feeding network O. García-Pérez (FG-IGN) 4

5 Bunny-ear antennas Bunny-ear antennas: Similar band to classical Vivaldi antennas. Better performance at lower frequencies. Easy to manufacture. 15 ohm reference impedance (in diff. mode). Simulation of an infinite array with HFSS. Differential feeding: avoids the losses and the bandwidth limitations of a passive balun. No substrate: reduces cost and potential losses. O. García-Pérez (FG-IGN) 5

6 Scan anomalies Scan anomalies appear due to the propagation of common-mode currents. The even-mode currents can be dissipated by connecting two resistors (3kΩ) between the feeding lines and GND, and therefore the anomalies disappear. Optimized design: VSWR<2.5:1, scanning up to 45º. Extra noise contribution lower than 1K. IEEE TAP accepted for publication. Anomalies: Resistors O. García-Pérez (FG-IGN) 6

7 Array measurements Array tile: 32 elements per polarization. Passive baluns to convert from differential to single-ended mode. Active impedance calculated from the measured S-param of the array. Reference impedance: 15Ω (diff.) Good measured results. Center element - scanning elements - broadside O. García-Pérez (FG-IGN) 7

8 Noise Temperature (K) Differential LNA #1: DLNA design 1 (I) Avago PHEMTs: ATF Hot/cold test at ASTRON. Good results for 15Ω source impedance: T<52K G>26dB Low Noise PHEMTs ATF Noise Noise (Measured) Noise (Simulated) - Actual 15ohm input load Noise (Simulated) - Ideal 15ohm input load Gain (db) Gain (Avago Tech.) Gain (Measured) Gain (Simulated) O. García-Pérez (FG-IGN) 8

9 DLNA design 1 (II) Mismatching effects: Collaboration FGIGN-ASTRON Poor s 11 due to the high input impedance provided by the FET in the lower part of the band. Mismatching effects over the active antenna impedance. Critical noise increase Active antenna impedance Z =15Ω Swp Max 1GHz Ideal antenna Z Z Z Z Actual LNA Swp Min.3GHz Active antenna impedance.4 Z =15Ω Actual antenna Swp Max 1GHz Swp Min.3GHz Magnitude (db) S11 (Measured) S11 (Simulared) Z =15Ω S Noise temperature (K) 65 6 DLNA - 15ohm antenna 55 Noise DLNA - Ideal antenna 5 DLNA - Actual antenna O. García-Pérez (FG-IGN) 9

10 DLNA design 2 (I) Noise Temperature (K) Differential LNA #2: Collaboration FGIGN-ASTRON Avago PHEMTs: ATF Inductive degeneration. Good results for 15Ω source impedance: T<4K G>26dB Noise Noise (Measured) Noise (Simulated) - Actual 15ohm input load Noise (Simulated) - Ideal 15ohm input load Gain (db) Gain Low Noise PHEMTs ATF (Avago Tech.) Gain (Measured) Gain (Simulated) O. García-Pérez (FG-IGN) 1

11 DLNA design 2 (II) Mismatching effects: Collaboration FGIGN-ASTRON s 11 <-6dB The mismatching effects over the active antenna impedance are not critical. Good noise performance in the band of interest Active antenna impedance Z =15Ω Swp Max 1GHz Ideal antenna Z Z Z Z Actual LNA Swp Min.3GHz.2 Active antenna impedance Z =15Ω Actual antenna Swp Max 1GHz Swp Min.3GHz Magnitude (db) S11 (Measured) Z =15Ω S11 (Simulated) S Noise Temperature (K) Noise DLNA - 15ohm antenna DLNA - Ideal antenna DLNA - Actual antenna O. García-Pérez (FG-IGN) 11

12 Conclusions The design of an active array receiver for the 3MHz-1MHz frequency range of the Square Kilometre Array (SKA) radio-telescope has been presented. The proposed solution provides the next advantages: Dielectric-free structure: reduces the cost and the losses Differential feeding: avoids the use of a passive balun Reduced number of LNAs/m 2 (~ 7.86 lna/m 2 ) However, some limitations appear due to its differential nature: Scan impedance anomalies Noise characterization of differential LNAs Good measured results: Scanning capabilities up to 45º with acceptable active reflection coefficient. LNA noise temperature lower than 4K for 15Ω source impedance. Finally, the matching condition effects between the antenna and the LNAs are analyzed: The LNA input reflection coefficient should be well matched to the antenna impedance. If not, the active antenna impedance will be mismatched, and the noise of the receiver may increase. Future lines: System integration and hot/cold tests with the active array tile. O. García-Pérez (FG-IGN) 12

13 Contributions [1] E. Lera-Acedo, L.E. Garcia-Muñoz, V. Gonzalez-Posadas, J.L. Vazquez-Roy, R. Maaskant, D. Segovia-Vargas, Study and design of a differentially fed tapered slot antenna array, IEEE Trans. Antenn. Propag., 29. accepted [2] O. Garcia-Perez, D. Segovia-Vargas, L.E. Garcia-Muñoz, J.L. Jimenez-Martin, V. Gonzalez-Posadas, Design, characterization and measurement of broadband differential low noise amplifiers for active differential arrays, IEEE Trans. Microw. Theory Tech., 29. submitted O. García-Pérez (FG-IGN) 13

14 : A Novel Active Array for the Mid-SKA THANKS O. García-Pérez FG-IGN oscar.perez@oan.es J.A. López-Fernández, D. Segovia-Vargas, L.E. García-Muñoz, V. González-Posadas, J.L. Vázquez-Roy, J.M. Serna-Puente, E. Lera-Acedo T. Finn, R. Bachiller, P. Colomer O. García-Pérez (FG-IGN) 14