LWA Station Design S. Ellingson, Virginia Tech N. Kassim, U.S. Naval Research Laboratory URSI General Assembly Chicago Aug 11, 2008 JPL
Long Wavelength Array (LWA) An LWA Station State of New Mexico, USA Technical Goals: 10-88 MHz tuning range Baselines up to 400 km for resolution [8,2] @ [20,80] MHz 52 stations ; (mjy-class sensitivity Each station is an array of dipole-like elements in ~100 m diameter aperture for FOV = [8,2] Access to Galactic Center (low gain antennas)
Astrophysics LWA Science Ionospheric Physics Cosmology High redshift radio galaxies, containing the earliest black holes Evolution of dark matter & dark energy by differentiating relaxed & merging clusters Acceleration, Propagation & Turbulence in the Interstellar Medium Origin, spectrum & distribution of Galactic cosmic rays Supernova remnants & Galactic evolution Pulsars Unprecedented continuous spatial & temporal imaging of the ionosphere Test and improve global ionospheric models Solar Science & Space Weather Radio heliography of solar bursts & coronal mass ejections Solar radar Exploration of the Transient Universe New coherent sources (More GCRT J1745-3009s?) GRB Prompt Emission Magnetar Flares Extra-Solar Jupiters: Detect magnetic field; conditions for life? Poorly explored parameter space new sources
Dual Polarized Antennas ( Stands ) Per Station Conservative Reasonable Each station needs to contribute sufficient collecting area to ensure calibratibility. Estimates of # of dual-pol antenna elements (stands) required per station, extrapolating from VLA 74 MHz experience N a = 256 selected LWA Memo 94 Galactic center at maximum elevation; @ 74 MHz
Station Antenna Array Geometry 110 m (N-S) x 92 m (E-W) 4 m min. stand separation Every element is digitized to allow unconstrained pointing of beams (among other things) Cost N a, so prefer to minimize N a Using 256 stands results in spacings 3 x Nyquist at 80 MHz Therefore, array has to have irregular spacings to mitigate against aliasing LWA Memos 73, 139 Possible elliptical (extended N-S) geometry to reduce variation in beam shape for low-elevation transit beam pointing
Big Blade Active Antenna Front end noise temp required to achieve indicated level of G.N.D. Tied Fork Current front end noise temp (~250K) Goal front end noise temp (120K) Confirmed (approximately) In field measurements
Mutual Coupling Collecting Area Circular 100 m dia station, Irregular geometry, Min. 4 m between stands Simple dipoles, 38 MHz [m 2 ] [m 2 ] Single Dipole, Simple model Single Dipole, Rx-mode NEC2 Single Dipole, Tx-mode NEC2 Array, Rx-mode NEC2 (stand average) Aperture Efficiency LWA Memo 73 Effect of mutual coupling
Mutual Coupling Beam Shape Shown here: Magnitude and phase of current induced at each feedpoint (moment method) Collecting Area, Pointing Zenith Phase, Pointing Zenith Only a small effect on beam shape and sidelobe level for any given pointing Collecting Area, Pointing 45 o Phase, Pointing 45 o Bigger concern is rumbling of beam as a function of direction of arrival Rumbling doesn t stop even if you fix the beam! LWA Memo 67 Actual impact on imaging not yet clear
Station Electronics Architecture Stand 1 84 samples Coarse delay I/Q & Dec by 2 Delay (for BF), Dispersion, Polarization 78 MHz BW, 98 MSPS AB AB Active Baluns Long Coax ARX Gain & Filter A/D 196 MSPS 12 bits T&Z = Tune within passband, filter, and reduce sample rate ( tune & zoom ) FIFO T&Z 3 more beams T & Z 100 khz from passband FIFO 57 ms 98 MSPS FIR FIR Storage (continuous) (TBN) Storage (one-shot) (TBW) 2x2 Matrx Mult. Sum to Form Beam 1 Sum to Form Beam 2 Sum to Form Beam 3 Sum to Form Beam 4 Available to correlator: 4 completely independent beams 2 calibrated polarizations per beam 2 tunings per beam, 0.4-8.0 MHz each, 4096 channels each Other products: Full bandwidth beams All dipoles, full bandwidth for 57 ms ( TBW ) All dipoles, 100 khz continuously ( TBN ) 0.4-8.0 MHz from passband; 4096 channels T & Z T & Z
Analog Receiver (ARX) Gain & selectivity only (Direct sampling architecture) No LOs 4 channels (2 stands) per board (128/station) LWA Memo 121 Gain Control + Reconfigurable Bandpass: (1) 10-88 MHz (2) 41 MHz highpass shelf filter (x 2) (equalizes HF) (3) 28-54 MHz (safe(r) mode)
Direct Sampling A/D Confirmed performance of direct sampling is consistent with LWA specs 200 MSPS, 12 bit A/D prototype board (Analog Devices AD9230-250) Evaluated also in lab; found OK Quick & dirty front end using ETA active antenna + ETA ARX modified for 20-80 MHz; Site near Blacksburg, VA HF ATSC Carrier NTSC Carrier FM Broadcast LWA Memos 127, 130
Frequency Plan A/D output (196 MSPS real) Beamformer (98 MSPS complex) Fs/4 Shift Left Multirate LPF + Decimate by 2 Challenge here is to achieve: Best possible rejection of strong out of band signals, consistent with: Bandwidth (78 MHz) and Low complexity LWA Memo 101 0-10 MHz aliases onto itself 88-108 MHz aliases onto itself
Polarization & Dispersion Calibration Beams should be not only full bandwidth (78 MHz) and fully independent, but also well calibrated. Xpol! Perfect calibration possible, but only for a single frequency and beam pointing, or if FIR filters of infinite length are available Cable dispersion further complicates this: Z=74, φ=45, M=16 Reasonable performance seems possible with M=16 (98 MSPS) FIR filters Z=74, φ=45, M=4 XPD 5-20 db Z=74, φ=45, M=16 (calibrated for Z=0) LWA Memo 138 XPD negl. ~ 10 db XPD negl.
Effect of Fence 120 m x 120 m security fence required around array effect? Biggest impact is for H-plane pattern, when collinear (as shown in these moment method simulations) < 1 db gain variation, but oscillates Effect depends on ground type 38 MHz 80 MHz
Acknowledgements Site infrastructure, cable system, electronic shelter (J. Copeland, A. Kerkhoff, D. Munton, J. York) Program management, systems engineering, analog receivers, civil engineering (J. Craig, W. Gerstle, Y. Pihlstrom, L. Rickard, G. Taylor) Antennas, front end, array geometry (T. Clarke, A. Cohen, B. Hicks, N. Paravastu, P. Ray) JPL Digital electronics (L. D Addario, R. Navarro) Array, signal processing, calibration, monitoring & control, architecture (S. Ellingson, M. Harun, K. Lee) + Many others at these institutions also involved in the LWA project + Many others at other institutions helping out Office of Naval Research http://lwa.unm.edu