Interview Gert Kruithof

– Research and Development

Gert Kruithof

“The forefront of radio astronomy instrumentation development”

Accelerate science

The research and end-to-end design capabilities of the ASTRON R&D department are at the forefront of radio astronomy instrumentation development world-wide. The mission of the R&D department is to research, develop and realize innovative world class radio telescopes. Since the completion of the single pixel feeds for the WSRT, the focus of the department has shifted towards phased array technology. LOFAR was the first truly large scale instrument worldwide that was based on that concept and is still the largest of its kind today. The phased array expertise built up in LOFAR has also opened new opportunities for WSRT. The research and development program APERTIF (APERture Tile In Focus) for phased array feeds (PAFs) that was carried out in the department has resulted in the roll out of an upgrade of WSRT in 2016. The uncooled PAFs of APERTIF have a system temperature of only 70K, an increase in survey speed of a factor of 20, allowing a revolutionary approach to survey science.

The department also has a leading role in the preconstruction phase of the Square Kilometre Array (SKA). It leads two consortia for Low and Mid Frequency Aperture Arrays and has a substantial role in the Science Data Processor (SDP) and in the Central Signal Processor (CSP) consortium.

We also have a leading role in the global research agenda towards a space based facility for the lowest radio frequencies (100 kHz to 30 MHz) which can’t be observed with a ground based telescope. The R&D department is now collaborating with Radboud University and the SME ISIS with Chinese organizations to deploy an instrument on the Chang’E4 mission to the moon in 2018. This is considered to be a milestone for radio astronomy since it is the first step towards a facility that will disclose the last unexplored part of the electromagnetic spectrum.

An overview of some of the projects the R&D department has been involved with in 2016:

APERTIF
At the beginning of 2016 the APERTIF-6 hardware was in place at Westerbork according to schedule. About a month later the APERTIF-6 software was also ready and technical commissioning with APERTIF-6 software was started. Whilst continuing with technical commissioning and attempt to also start some of the science commissioning the remaining hardware (APERTIF-12) was installed as well. The last parts were the Uniboard due to longer deliver time, but during the summer, all hardware was installed including the correlator. During commissioning various issues issues were encountered and at the end of 2016 the first fringes, the first real image, the image of Leo T and the first full APERTIF mosaic of 37 beams were produced. 2017 offers a lot of challenges for delivery of software and firmware in order to continue with commissioning and start surveys.

“First APERTIF fringes for baselines formed by 9 dishes

DOME
The DOME project is aimed at creating technologies for energy efficient massive computing and efficient data transport. In support of this, the project developed a series of microserver cards as part of the hot water cooled DOME microDataCentre concept (see figure). The company Variass produced 24+ of these cards for system development and for use in the Users Platform. The start-up company ILA Microservers is currently focusing on bringing the microserver to the market.  In addition, together with industry, DOME supported development of radio over fibre links for transporting antenna radio signals at the SKA telescope station. Following the successful test results, these links have actually been ordered by ICRAR for the Murchison Widefield Array. Other highlights include the support of the SKA designs. The project contributed to roughly twenty SKA (SRR and PDR) design documents. Also nearly fifty scientific papers were written in all fields in which DOME is active.  This includes the development of a flexible beamformer, and a novel way to create images using  radio-interferemetric data. And last but not east, the DOME project celebrated its first PhD defense.

Hot water cooled microserver unit

I-LOFAR
After a long period of preparation, finally in the early start of 2016 while the winter pushes the temperature around zero, ASTRON carried out the site survey at the proposed LOFAR location at Birr Castle in Ireland. This 13th International LOFAR Telescope (ILT) station will be the most west located station, about 1000 km from the LOFAR core. With the “Great Telescope”, built in 1845 at the background, measurements were carried out assess the RFI situation at this new LOFAR location. In addition, the environment was checked for practical roll-out issues. A major problem at this location was the river which flooded the station area during winter time. Several options to make the site suitable for building the LOFAR station were investigated. A solution was found in raising the station area with ground from the surroundings. Groundworks for the station has been finalised during this year. All the electronic modules that are required for this station have been ordered and available for installation now. Construction of the 13th ILT station in Birr Castle is scheduled in early 2017 and will be operational in the second half of that year.


Preparing for RFI test


The “Great Telescope”will soon be taken over by LOFAR


I-LOFAR groundwork

SKA – LFAA / AAVS1
The Low-Frequency Aperture Array (LFAA) element of SKA1-Low is the most visible and in scale the largest component of the telescope: the full realization will consist of 130.000 antennas forming 512 stations. LFAA deals with the design of the antennas, signal transport and the station signal processing. LFAA completed the detailed design of all the component and first prototypes have been tested both in laboratories in Europe but also at the Murchison Radio Observatory, in Western Australia. The ultimate test for the consortium is construction and evaluation of the Aperture Array Verification System 1 (AAVS1), a 400 antenna element system. Late 2016 all components of AAVS1 have been produced and shipped to Western Australia. Early 2017 deployment and test will start.

Besides the realization of AAVS1, the consortium has been busy with the LFAA cost model; what will be needed for the construction of the full SKA1-Low, in terms of hardware expenses but also manpower. Production cost of AAVS1 parts and MWA and LOFAR experience led to a mature cost model, with low contingency and good confidence from the team.


“Initial installation of 96 SKA1-Low antennas”

SKA-CSP
ASTRON, CSIRO and the Auckland University of Technology are working together on the design of the Central Correlator and Beamformer for SKA Low (called Perentie). The collaboration has proven to be highly successful with all parties learning a lot from each other in the process. In 2016, the team presented the system design and the system engineering artefacts to the Delta PDR review panel and successfully achieved the milestone (see picture).

Delta PDR review panel

The first signal processing board prototype has also been completed (see picture below) and the initial tests are highly positive.


The first signal processing board prototype

In the run-up to CDR (to be held in mid-2017) and as a preparation for the construction phase, the team is prototyping the full Gemini line replaceable unit (LRU), which includes a second iteration of the signal processing board, packaging, cooling, as well as firmware and software elements.
ASTRON provides a wide range of resources to the collaboration in the areas of system design, system engineering, firmware, mechanical and thermal engineering, as well as project management.

SKA – SDP
The SKA Science Data Processor (SDP) consortium passed its Delta – Preliminary Design Review (PDR) in May 2016. A relatively mature architecture was reached and discussions with the SKAO led to decisions about SKA Regional Science Centers, clarifying the scope of the SDP. After Delta-PDR the consortium way of working was restructured and moved away from a Waterfall style of working towards a Risk driven, Agile, Sprint based way of working.  The SDP Costing was iteratively refined and the Architecture of the SDP further detailed. ASTRON has a broad contribution to SDP and is leading some of the major work areas.

The SDP Consortium at their yearly face-to-face meeting in Malta, 2016.

SKA – MFAA
ASTRON is leading the consortium that develops Mid-Frequency Aperture Arrays (MFAA) for the second phase of the Square Kilometre Array. ASTRON is involved in the system design, the antennas and analog electronics. All this in close cooperation with international partners, industry and the DOME project.

In 2016, the consortium delivered the science requirements and the associated system requirements for a MFAA based on SKA2 telescope, and successfully passed the System Requirements Review. The next step towards SKA2 is the realization of a demonstrator. Such demonstrator is necessary to demonstrate the MFAA technology at station level, but will also be capable of conducting science. To support the ambition to build such a demonstrator in South Africa, together with a group of leading scientists a white paper has been published (https://arxiv.org/abs/1612.07917).

In 2016 ASTRON further refined the design several aspects of the MFAA antenna tiles. As a result, the cost of the antenna tile decreased by more than 25% compared to the previous design, the energy consumption decreased by 50%, while the sensitivity of the receivers increased. At the same time the simulation models of the tiles have been improved, resulting in a nearly perfect match between the designed and measured characteristics of the antennas.

Details of the new MFAA antenna design


(Part of the) MFAA team at the design review, with in the background thee 76m Lovell Telescope.

NCLE
The Netherlands China Low frequency Explorer (NCLE), is a radio receiver aimed to be launched mid 2018 as a scientific payload on-board the Chang’e 4 relay satellite (see figure). It will comprise of three monopole antennas of 5 m length connected to a digital receiver, supporting dedicated science modes, implemented in a flexible software-defined radio system. These modes for instance perform fast Fourier transforms to create average radio spectra, allow triggering on transient radio events, or allow to retrieve direction of arrival information using beam-forming or goniopolarimetry techniques. Raw time traces can be stored for ground-based post processing and VLBI. The ASTRON team, responsible for the RF part of the antenna and for the low noise amplifier have made and tested initial designs, and are preparing for the PDR/CDR phase mid 2017. The NCLE scientific payload is developed by Radboud University (PI), ASTRON, and ISIS Innovations in Space, and is supported by ESA PRODEX, the Netherlands Space Office (NSO), and by partners from LESIA, TUDelft, UTwente, and JIVE ERIC.

Impression NCLE antennas on Chang’e 4 relay spacecraft

UAV aided LOFAR antenna measurement campaign
An antenna measurement campaign was conducted in April 2016 at LOFAR station CS302 in collaboration with Italian partners from the Instituto Nazionale di Astrofisica (INAF), the Consiglio Nazionale delle Ricerche (CNR) and the Politecnico di Torino (PT). These measurements support the development of an improved LOFAR beam model and help to develop strategies to support the roll-out, commissioning and characterization of the Low Frequency Aperture Array (LFAA) of the SKA. These measurements were done using an Unmanned Aerial Vehicle (UAV) equipped with a radio frequency (RF) transmitter developed by our Italian partners for antenna characterization below 500 MHz in the context of the SKA.

During this campaign, the team shown in Figure 1 performed measurements on all three antenna arrays of LOFAR station CS302, i.e., the Low Band Antenna (LBA) inner array, the LBA outer array and the High Band Antenna (HBA) array with several different flight strategies. In one of these flight strategies, the socalled spin flight, the UAV is hovering in a fixed position above the array while spinning around its vertical axis. Figure 2 shows the ouput of the central antenna in the array, which was right below the UAV. The plot clearly shows that the probe antenna on the UAV alternatingly aligns with each of the two dipoles that make a LBA. With such measurements, we can very accurately measure the orientation of the antennas and verify that the antenna can discriminate the two polarizations very well.

In the meantime, we have made significant progress on improving the electromagnetic (EM) model for the LBA, whose predictions are now in good agreement with the measured response. Efforts to make a similar improvement to EM model for the HBA will start in 2017. We also successfully demonstrated our ability to derive the antenna positions from the RF measurements. This could lead to significant cost savings in LFAA as this pre-empts the need to physically determine the position of each antenna one-by-one.


The team that conducted the campaign(front row, then rear row): Fabio Paonessa (CNR), Paolo Maschio (PT), Andrea Lingua (PT), Giuseppe Pupillo (INAF), Menno Norden (ASTRON), Stefan Wijnholds (ASTRON), Giuseppe Virone (CNR), Pietro Bolli (INAF) and Irene Aicardi (PT).

 


Output of the central LBA during a spin flight.


The UAV in act
ion


Close-up of the UAV