Low-Frequency Aperture Array


The LFAA Consortium had a productive period with successful milestone achievements and a good ramping up to the production of the Aperture Array Verification System (AAVS1), a 400-antenna 4-station system, which will be the ultimate LFAA test platform.

An extensive study on the capabilities of Radio over Fibre (RFoF) has been completed with the conclusion that a range of up to 70km will be possible, which creates the option of an SKA1-low realisation where all the antenna signals (even in remote stations) are connected to one central processing facility. In another significant advance, a new fully detailed LFAA deployment plan and detailed knowledge of AAVS1 production cost formed the basis of a complete update of the LFAA capital expense costs, in effect moving from engineering estimates to industry-based costing and giving much higher confidence in the final figures.

The LFAA Consortium facilitated a visit on November 1 of the Dutch King Willem Alexander and Queen Maxima to Curtin University in Perth, Western Australia, displaying an SKA-low mini-station of 16 antennas and associated electronics. Prof. Peter Hall had the honor to outline to Their Majesties the importance of the long standing radio astronomy science and engineering ties between the Netherlands and Australia, in the 400th anniversary year of the Dutch landing in Western Australia. The Dutch royal family has historically showed great interest in radio astronomy and both the King and Queen showed a keen understanding of the SKA project and the need for extreme radio quietness at the site of SKA-low. A much appreciated side theme on the day involved a youth novel set around the Dwingeloo telescope and featuring former Queen Juliana (the novel was originally commissioned by ASTRON as “het logboek” in 2014). The English translation (ICRAR, “The Journal”, 2016) was presented to the royal couple as a gift for the Dutch princesses.


A visit by Dutch Royalty; Their Majesties view LFAA antennas at Curtin University, November 1, 2016. The King and Queen are accompanied by Dr Michiel van Haarlem (left) and Prof. Peter Hall.

LFAA Antenna Progress: a new Antenna Design

SKALA-2 has been extensively tested both in Europe and in Australia over the last few months (see figure 2 below). Furthermore we have now realised the design of SKALA-3 to improve the passband smoothness (see figures below). The changes on the antenna are a modified input matching network of the LNA and a slightly larger bottom dipole to improve the matching at low frequencies.


SKALA-2 deployed at the MRAO, Cambridge and beam pattern measurements performed with the drone system in Turin, Italy showing a good agreement between measurements and simulations.


The SKALA-3 element designed by the Cambridge team has a modified input matching network of the LNA and a larger bottom dipole to enhanced the matching at sub 65 MHz frequencies.


SKALA-3 passband residuals after a low order polynomial fitting (order 3) as described by Trott in arXiv:1604.03273. Paper in preparation by de Lera Acedo et al. Consortium science and engineering personnel have worked closely with the SKAO and wider science community to develop meaningful specifications and metrics for passband smoothness in the context of SKA-low EoR science.


SKA1-low A/T using SKALA3 elements and SKALA2 elements across the EoR band. The A/T has also become flatter with SKALA-3. Paper in preparation by de Lera Acedo et al.

Drone Antenna Calibration

A second campaign of beam pattern measurements and array calibration has taken place at Lord’s Bridge, Cambridge this September. The British and Italian teams performed different pattern measurements of the near field and far field complex beams of the Pre-AAVS1 array (16 SKALA-2 antennas pseudo-randomly distributed over a diameter of approximately 8 metres). The array data was captured using the Tile Processor Modules designed for AAVS1. The data is now being post-processed and will soon be reported.


Drone flying over Pre-AAVS1


Measured beam at 175 MHz.

Satellite beam measurements of Pre-AAVS1

We have developed a beam measurement system using predicted satellite passes at spot frequencies across the SKA-low frequency band as a way to evaluate our full electromagnetic beam models based on the HARP code. This code is being developed in collaboration between the University of Cambridge and the Université Catholique de Louvain, Belgium, and it is capable of fast and accurate simulations of SKA-low station beams. At the center of the figure below we can see the comparison of the beam measured using a satellite and our prediction for the Pre-AAVS1 array in Cambridge.


Satellite beam measurement of Pre-AAVS1 at Cambridge, UK

Calibration Results with AAVS0.5 and MWA

The Murchison Widefield Array (‘MWA’, an SKA-low precursor telescope) hosted a 16-antenna prototype system from 2013-2016 which was fully integrated with the MWA. The prototype system, called “AAVS0.5”, was a small array of 16 SKALA antennas pseudo-randomly distributed over a diameter of approximately 8 metres. By integrating AAVS0.5 with the MWA, the full sensitivity of the MWA could be used to perform antenna metrology on the AAVS0.5 array, verifying the sensitivity and simulated array beam patterns using astronomical sources, and laying the groundwork for an extensible metrology program for AAVS1 and SKA-low.

The results of the characterisation were published as “Characterisation of a Low-Frequency Radio Astronomy Prototype Array in Western Australia” by Sutinjo et al, 2015. ADS Link.


From Sutinjo et al, 2015. The figure shows measured beam response of AAVS0.5 (blue dots) compared to theoretical models (lines). The overall agreement is excellent, providing confidence in the antenna modelling.

In more recent work, the combination of AAVS0.5 and MWA has been used to test models of calibration systems for heterogeneous aperture arrays. Because the MWA and AAVS0.5 antenna elements are different (bow-tie dipoles vs log-periodic SKALA antennas) the phase centre of the aperture array differs between the MWA and AAVS0.5 as a function of frequency. Work is in progress to show that this difference can be corrected for through a-priori knowledge of the SKALA response patterns, and therefore that heterogeneous arrays can be used just as effectively as homogeneous arrays. This result is important for the impending commissioning of AAVS1 (see below), which will also be integrated with MWA.

The Aperture Array Verification System, AAVS1

The realisation of a full size SKA1-low verification system: the Aperture Array Verification System 1 (AAVS1) is well on its way. Hardware production within Europe is ongoing, as well as software and firmware development. All items are tested and verified at the pre-AAVS1 array at Lords’ Bridge, Cambridge, before being sent to Australia for deployment at the waiting MRO infrastructure. AAVS1 is a comprehensive, end-to-end, prototype demonstrating a wide variety of LFAA and, by virtue of its MWA integration, SKA-low implementation aspects.

The AAVS1 hardware production is at full speed. After factory testing, integration of subsystems has to be done, within the work packages, as well as between work packages. Internal interface control documents are crucial and many challenges are ahead of us. Figure 10 shows an overview of AAVS1.


AAVS1 block diagram.

The SKALA2 antennas have been shipped from University of Cambridge (UCAM) to the International Centre for Radio Astronomy and Research (ICRAR), arriving in Australia in late November. A pyramid‐placeholder has been designed and shipped to ICRAR as well, enabling the MRO‐crew to deploy the antennas without having functional pyramids available yet.

All low noise amplifiers (LNAs) have been ordered by UCAM and will be factory tested, after which they will become available towards the start of the first batch of pyramids to be integrated.


A truckload of SKALA2 antennas, ready for transport to Australia.

After testing the pyramids at UCAM, they will be shipped to ICRAR.

The first 40 FrontEnd (FE) modules have been produced and tested by Instituto Nazionale di Astrofisica (INAF). The remaining FE‐modules (and PRE‐ADUs) will be produced in two batches.

The first batch (240 pcs.) will be produced during the upcoming weeks. After fully testing the FE, the hybrid cable will be connected and the modules will be sent to UCAM to be integrated with the LNAs. The first batch, together with the early 40 FE modules will enable us to create the AAVS1 full‐station.

The second batch of FE‐modules will arrive at INAF, January 2017, of which we will realise the three auxiliary stations.

The Analog to Digital Units (ADUs) are produced in two locations, Italy and The Netherlands. The first production run in Italy has finished, resulting in 33 factory-tested boards, of which 2 boards need some rework. The second production run in The Netherlands will be produced early January 2017, resulting in another 12 ADUs.


Figure 12: Analog to Digital Units (ADU) ready for in depth lab-testing.

After factory testing, the boards will be lab tested (full RF and digital) at INAF Medicina, after which they will be integrated with the PRE‐ADUs, creating Tile Processing Modules (TPMs). The TPMs will be put into 7 subracks, each holding 4 TPMs. The subracks will be assembled and tested at INAF, with the help of Oxford and Malta staff.

One of the subracks will be used for the European EMC‐prescan, either at ASTRON or at Bologna University.

With the establishment of AAVS1 at the MRO, the Consortium will begin a range of verification testing using the instrument in both stand-alone and MWA-integrated modes. These “on-site on-sky” measurements using functional, astronomically-capable interferometers are an essential part of the path to SKA-low and are, deservedly, a highlight of the SKA project.

AAVS1 Infrastructure Advances

Recent work in Perth and on the site has been focused on deploying, commissioning and characterising the infrastructure required to support the AAVS1. As well as being necessary enablers of AAVS1, these activities are important in their own right. They provide invaluable insight into the construction and deployment considerations that will be applicable to the deployment of the LFAA and SKA-low. Assumptions about the sequence and timing of the myriad processes and procedures involved in installing, connecting and testing the LFAA can be tested and refined or rejected. The limits of practicality can be explored, and lessons learned injected back into requirements, budgets and designs.

The work is being executed by ICRAR-Curtin, its Consortium partners and a range of Perth and Geraldton based manufacturing, services and logistics and handling companies. Capturing the feedback from the various contractors engaged in the process is a focus of the ICRAR-Curtin group. This feedback ensures that the estimates and plans developed for LFAA, construction and deployment in particular, are informed by relevant commercial considerations and expertise.


Figure 13: Locating and installing the AAVS1 antenna bases

Figure 13 shows the location of the antenna base positions using commercial Differential Global Positions System and laser based surveying equipment. With so many antennas to deploy, every second counts and, with a sufficiently dense reference network in place, this system enables antenna positions to be located with the precision required relatively quickly.

Once the locations have been marked, the bases are handled into position. At LFAA scale, plant and equipment such as mobile cranes and ground handling apparatus will reduce the manual component of this procedure.


Figure 14: Positioning and installing the AAVS1 ‘Antenna Power Interface Units’ (APIU)

Figure 14 shows the positioning and population of an AAVS1 Antenna Power Interface Unit (APIU). The APIU is a weatherproof enclosure housing ‘Fibre-Optic Breakout modules’ (FOBOTs) and power supplies configured as standard ‘rack units’. Functionally, the APIU is the fibre aggregation and power distribution node for an AAVS (and LFAA) station.

The AAVS1 APIUs feature novel, low-RFI, power supplies developed by, ICRAR-Curtin pre-construction industry partner, Balance Utility Solutions. These inherently low-emission power supplies were developed specifically to comply with the stringent MRO RFI regime with only minimal shielding. Balance’s significant contribution to the realisation of the AAVS1 in support of the LFAA Consortium is an example of the kind of meaningful industry engagement required to efficiently realise the aspirations of the SKA Project.

The integration of the APIU into the AAVS1 will likely yield lessons that result in changes to its form and function. Indeed, that is precisely the point of the AAVS1. By building, commissioning, operating and maintaining the AAVS1, the LFAA Consortium will ensure that the LFAA procurement specification is as well informed and as risk free as possible.


Figure 15: Deployment and characterisation of RFoF trunk fibre for AAVS1

Fibre-optic cabling is central to the LFAA architecture. The signals received by the antennas in the field travel several kilometres over fibre before being digitised and processed. In LFAA this fibre will, for the most part, be buried. A variety of cost and practical constraints precluded the burying of the fibre-optic cabling for AAVS1. As a result, the environmental conditions that the AAVS1 signal path will be subjected to represent the most stringent possible test of the proposed LFAA architecture.

The left panel of Figure 15 shows 576-core ribbon-fibre cable being deployed between the AAVS1 site (co-located with the MWA) and the MRO central processing building, a total distance of 5500m. The right panels of Figure 15 show the test setup currently being used by ICRAR-Curtin engineers to characterise the performance and stability of the AAVS1 fibre–from APIU to processing building–under dynamic, often extreme, environmental conditions.


Figure 16: AAVS1.1 (full station) infrastructure complete and awaiting antennas

Figure 16 showcases the fruits of many tens of hours of labour; in the workshop, lab and field by the ICRAR-Curtin engineering group and its Consortium and industry partners. The AAVS1.1 station infrastructure, complete and awaiting installation of antennas.


Figure 17: AAVS0.5 antennas deployed in 2012 provide valuable data on component life on the MRO

The AAVS0.5 system was deployed by a subset of the AADC Consortium partners in 2012, before the formal kick-off of SKA pre-construction. The 16-element array, featuring an earlier iteration of the SKALA antenna allowed the Consortium to make a variety of measurements that have helped to guide subsequent development.

Despite being taken out of regular service, the LFAA design having evolved significantly, the AAVS0.5 continues to offer valuable insights. The various components, materials, connectors, and joins that make up the system have now been subjected to five years of the harsh MRO environment. Investigating and understanding their reaction to this exposure is another important way that AADC Consortium engineers are ensuring that the LFAA procurement specification is as well informed as possible. Figure 17 shows the settling of red MRO-dirt on an LNA within the electronics housing ‘trumpet’ at the apex of an AAVS0.5 antenna.

Report provided by the LFAA consortium