Low-Frequency Aperture Array


The consortium, together with the SKAO, is working on the completion of the Critical Design Review material, but in the meantime engineering work continues. Prototypes of the improved SKALA antenna have been built and are currently being tested in Europe (photos below). These prototypes will be tested at the MRO as well.

Aperture Array Verification System 1: initial evaluation

The evaluation of a full size SKA1-LOW verification system: the Aperture Array Verification System 1 (AAVS1) has seen its initial start. An international team of astronomers and engineers are working their way through the AAVS1 commissioning test plan.

A very successful site trip during November 2017 resulted into the finalisation of the AAVS1 main station. The main station closely resembles future SKA1-LOW stations: 35-meter diameter in size, hosting 256 log-periodic dipole-antennas. The antenna signals are transported to the Central Processing Facility (CPF) using Radio Frequency over Fibre (RFoF) technology. Within the CPF the antenna signals are digitised and beamformed using Tile Processing Modules (TPMs), after which the beamformed data is sent to the correlator.

Obviously, after realising the main station, the team is very keen on evaluating the system. Guided by the AAVS1 commissioning test plan, several tests are being performed, ranging from straight-forward power spectrum measurements to complex station-calibration. At the time of writing, only a subset of tests were able to be performed, focusing on power spectrum measurement, drift scan testing and stability tests.

Measuring (and recording) the power spectrum received by an antenna results in plots such as shown in figure 1. One can easily identify and troubleshoot outlier antennas, after which the test is repeated on all installed antennas. Features in the spectra are all understood: the peak around 68 MHz and surrounding wiggles is a feature of the SKALA2; the peak around 137 MHz is from Orbcomm and other satellites and is frequently observed; the peaks between around 240-280 MHz are persistent interference from geostationary satellites; the peaks between 360 and 380 MHz are persistent and have been observed before but are of unknown origin, presumably satellites.

A drift scan is a measurement of the total power received by an antenna or array over a long period, typically 24 hours or more, with the antenna/array pointing in the same direction (often at zenith). At low frequencies, a drift scan for each antenna should display a rising and falling total power measurement over a day, consistent with the system temperature of the antenna being dominated by the background radio sky. An example of a drift scan taken over several days is shown in Figure 2. The figure demonstrates very good repeatability and stability of the system response on timescales of days. One can note that the spectrum is free of terrestrial RFI, in particular for the FM and TV bands, highlighting the quality of the radio quiet site.

Based on previous experience with dipole antennas at low frequency, we expect roughly 3 dB changes in received total power in a SKALA2, depending on whether the Galactic centre is overhead or below the horizon. The variations seen in Figure 2 are generally consistent with expectations.

The drift scans generally follow the same pattern over all antennas for several days. Figure 2 shows all frequencies for a single antenna over several days. It also indicates that the antennas produce qualitatively repeatable results over several days.

Near future tests are focused on instrument calibration and beamforming.

Figure 1: normalised total power measured for TPM 1, X-polarisation

Figure 2: Waterfall plot of antenna 0 from TPM 0 over several days

SKALA Antenna development

The development of SKALA4 (the latest iteration of the log-periodic antenna for SKA1-LOW with improvements in spectral smoothness, polarization and sensitivity) has continued. In April 2, prototypes were deployed by LFAA in Western Australia (see Figure 3) to do RFI measurements and early on-sky tests. Mark Waterson (SKAO) conducted these tests, which have been very helpful in the development of the antenna.

Figure 3. SKALA4 prototypes at the MRO, WA.

Currently, the mechanics of the antenna center the efforts on its development. Presently we are working on the antenna support components. Detailed Computational Fluid Dynamics simulations of wind loading are being performed to address the effect of wind on the antenna structure (see Figure 4). This work is being conducted in collaboration with INFRA AUS. Further work is still needed to finalise the design of SKALA4 including tests of the latest prototypes and array tests. This work will take place in the coming weeks and months.

Figure 4. CFD simulation of wind loading on SKALA4 (Credit: Nicolas Fagnoni, Cambridge University).

Furthermore, LFAA is currently working on full electromagnetic simulations of station beams with the newest station diameter (38 m) and SKALA4 using HARP (software developed by the group led by Prof. Christophe Craeye at the Université catholique de Louvain in Belgium in collaboration with the University of Cambridge). Early validation of these simulations with smaller arrays (easier to simulate with commercial codes) has been performed. Dr Hardie Pienaar from Cambridge University has performed the simulations shown in Figure 5 using HARP and the commercial code FEKO. These simulations will help the consortium in the preparation of the CDR documentation and the finalization of the antenna design. Furthermore, these simulations will be useful for any further work on the station layout and in the calibration of the arrays.

Figure 5. Array simulation validation with SKALA4. Comparison between HARP and the commercial code FEKO. Top: 16-element array used for the validation (FEKO model). Bottom left: Beam cut in the E-plane. Bottom right: Beam cut in the H-plane.

Report provided by the LFAA consortium