Mid-Frequency Aperture Array



The AAMID Consortium, working on the Mid-Frequency Aperture Array (MFAA), an Advanced Instrumentation Work Package, aims to demonstrate the feasibility, competitiveness and cost-effectiveness of MFAA technology for SKA2. The key advantage of AAs is the capability of realising a very large Field of View and sensitivity, which results in an unsurpassed survey speed. Furthermore, AAs are capable of generating multiple independent FoVs, enhancing the efficiency of the system, for calibration and for multiple concurrent observations.

Front-End Design Summary

The cross polarisation performance analysis of the finite C-ORA array at the University of Manchester, including intrinsic cross-polarisation ratio (IXR) is under investigation. The mechanical requirements of the finite array for anechoic chamber measurements have been produced. The finite array for RFI measurement is shown in Fig. 1. The reflection coefficient for the centre element in the finite array (5×5 dual polarised) is shown in Fig. 2.


Figure 1 The finite C-ORA Array for RFI measurement.


Figure 2 The reflection coefficient measurement for the centre element of the finite array.

A differential beamformer board based on time delay has been prototyped by Station de Radioastronomie de Nançay. The 16 channel beamformer board with differential inputs (SATA standard) is shown in Fig. 3.


Figure 3 The time delay beamformer board with 16 channels and SATA inputs. Developed in Nancay, France.

The 10 x 10 finite arrays based on square grid and triangular grid will be measured in the close range anechoic chamber. 4×4 Elements will be beamformed in this measurement. The mechanical designs for the square grid based array is shown in Fig. 4


Figure 4 The mechanical design model of the 10×10 C-ORA finite array for anechoic chamber measurement. To be manufactured by Trackwise.

ASTRON has performed noise temperature measurements on a 16 element array, this array consisted of 16 LNA modules of the previous design and two beamformer boards which are fully operational. ASTRON team has performed tests with different settings and measured the influence of connecting the antennas electrically together (for now with tape). The array has 40 to 45 dB of gain.


Figure 5. The finite array based on Vivaldi antennas, 16 LNA modules have been integrated with antennas.

A hot-cold test on the receiver developed at North-West University of South Africa has been performed in ASTRON in the Netherlands, and got some preliminary receiver noise temperatures (see the Fig. 6). The current plan is to make an array of these receivers.


Figure 6. The noise temperature of the received developed in the North-West University (South Africa).

The development of a sparse, randomised antenna array for MFAA has made a significant step forward by making the log-periodic antenna highly manufacturable. The first version of a volume antenna as a development from initial prototypes is illustrated in Figure 7. The log-periodic antenna for MFAA builds on the knowledge gained with the Low Frequency developments for SKA1-low, LFAA; it is wide bandwidth with forward gain to minimise the number of antennas required. There are still a very large number of antennas required and making it very low cost to produce and install is essential.


Figure 7. First version of a volume antenna as a development from initial prototypes

This version of the antenna consist of only four low cost plastic parts and the four antenna arms. As can be seen the top cover acts as a sunshield, since this antenna design requires no cover arrangements and has a base for simply bolting to a ground plane mesh, in “Tiles” of 16-antennas: thus avoiding the environmental protection issues associated with cases and covers.

The antenna incorporates novel single ended LNAs, for low noise with low cost and power, one for each polarisation. These are housed in two of the antenna arms, which provide the environmental protection required for the electronics and minimises assembly complexity.

The antennas will be connected via coax, which carries output signal and power for the LNAs to a local Tile power and signal consolidation unit, which then transmits the signal using RFoF links to the processing facility.

A demonstration array using 128 of these antennas will be built at Lord’s Bridge Observatory in Cambridge in 2017.

Update of the MFAA Environmental Prototypes

During the March AAMID all-hands in Cape Town, an ASTRON team spent time on the Karoo site, working on the MFAA Prototypes. Besides general inspection, the data loggers were refurbished and maintenance has been executed. Detailed information was gathered on all four prototypes, this section focuses on one of them, the open prototype with foil cover.


Figure 8: Environmental wear and tear of the insulated prototype.

Looking at the picture above, one can clearly see some environmental influences on this prototype. After opening, we find a small attention point at tension strap. The tension strap, used for keeping the POLYMAR foil in place in the middle of this prototype, is pulling on the POLYMAR foil glued connection tag. It is peeling of the tag but not completely, the peeling stops at the point where peeling becomes pulling due to the shape of it. This solution seems to work.


Figure 9: Vivaldi antenna tile, inside the open prototype, sand and dust are clearly seen.

Looking at the photos within figure 9, we can clearly see that this prototype collects quite some sand and dust even under the EPS corners. Looking at the first row we see hardly any sand and dust, it looks like wind is keeping it clear of it.


Figure 10: Deformation of some the parts.

Figure 10 shows the the Vivaldi arrays are buckled and bended. The rubber straps, holding the POLYMAR foil in place, are pulling on the roof of this prototype and this roof lies on top of the Vivaldi arrays which are made of 0,6mm thick aluminium sheets. This pulling on the roof is applying a distributed force on this section of the arrays and these are not strong enough. Therefore, for future roof designs we should not apply any force on the Vivaldi arrays as these are responsible for the main (mechanical) function of this telescope.

ISSA array demonstrator

The fractal octagonal phased array antenna (Fr-ORA) is an engineering prototype developed at the University of Malta for a middle frequency SKA aperture array, MFAA, operating between 300 MHz to 1400 MHz. To achieve an aperture array with a wide bandwidth performance, the elements were tightly coupled, triggering the mutual coupling effect between the elements. Hence, increasing the capacitive coupling effect at the end portions of the elements aids to counteract the ground plane input impedance at the lower operating frequencies.


Figure 11: The fractal octagonal open rings array.

The optimisation problem of the array contained a fractal geometry, where the elements of the array were designed based on the recursive nature of a fractal. It is likely that the Fr-ORA will be competitive in relation to related technologies because of its lightweight construction, its low cost (because of the planar design and the significantly lower metal content in the fractal elements), and its low profile (planar structure), with dual polarisation and the potential steerability of the individual elements. Another unique feature of the array is that the center frequency and bandwidth can be tailored to any specific application by scaling the element sizes in the active layer.

The possibility of using a fractal geometry as a tightly coupled phased array antenna for SKA application was examined. A practical implementation of the phased array antenna was fabricated and measured. The proposed array element has proven to predict well-behaved impedance over a wide bandwidth to a maximum scanning volume of . Furthermore, a method for calculating far-field antenna patterns exploiting a set of near-field data sampled from a rectangular plane over the prototype 9 x 9 element array was used to verify the validity of the design. The measurement results allowed for the evaluation of the antenna array type for SKA.


Figure 12 Fr-ORA 9 x 9 prototype element array.


Figure 13. Simulated and measured passive reflection coefficient for the 9 x 9 element array results.

Given the fact that the prototype needs to meet SKA specifications and requirements, post-experimental results allow for further and in-depth studies. Sponsored by the Technology Development Programme of the Malta Council for Science and Technology (MCST), the Institute of Space Sciences and Astronomy (ISSA) embarked on building large scales of the array. The Malta array demonstrator is an implementation of two antenna arrays and a receiver chain. Each array consists of 5184 elements, an array of 72 x 72 elements covering an area of 100 m2. The elements of the array are supported by expanded foam over a grid ground plane. The main purpose of this project is the calibration and the characterisation of the array in its real operative condition. Furthermore, testing the analog receiver chain in a realistic environment is also considered through this project.

The characterisation of the antenna array radiation pattern is investigated involving a far-field flying source. The verification system deploys two antenna arrays each of 10 m x 10 m distant from each other. The system exploits an unmanned aerial vehicle (UAV) equipped with a continuous wave RE transmitter and a dipole antenna. Accordingly, the reconstruction of the received power pattern from the overall system can be easily compared with the simulation results of the model. For very large arrays, a near-field to far-field characterisation technique using a flying source is also utilised for far-field pattern calculations. By means of the measurement of the near-field pattern of an embedded element in the aperture array, the response of a prototype of a 9 x 9 element array can be easily validated using an electromagnetic software model. In this way, the element array can be improved for better performance, and the array beam can be modelled for calibration proposes. The measured data can also be used for further and more complex numerical analysis.

MFAA Publications and Presentations

In preparation for the System Requirements Review of the Mid Frequency Aperture Array, a number of documents were produced, and amongst them, the Science Requirements is of interest to a wider audience than just the SRR panel members and the SKA Office. This document gives an overview of the science goals of the full SKA in the frequency range 450MHz to 1450MHz, and goes on to describe the capabilities necessary for the SKA to achieve those goals. The emphasis is on the Mid Frequency Aperture Array which the MFAA Workpackage intends to demonstrate clearly to be the best technology in this frequency range. The SKA Mid Frequency Science Requirements is available publicly on the arXiv document service, and can be found at https://arxiv.org/abs/1610.00683

The SKA Mid Frequency Aperture Array was presented at several conferences and meetings since the previous edition of SKA eNews.

MFAA and an overview of SKA pathfinders and precursors was presented in two talks at Fermilab on 17 October as part of the INFIERI Workshop on Signal Processing in Astrophysics, Particle Physics, and Medical Physics. The talks were given by Steve Torchinsky (SKA/MFAA overview) and Tailei Wang (Integrated Beamformer Circuit Board). These talks were repeated a few weeks later at the Astroparticle Physics and Cosmology laboratory in Paris where there is already some interest in joining SKA.

https://indico.cern.ch/event/557734/timetable/ – all.detailed

Also in the USA, the MFAA pathfinder EMBRACE was the subject of a talk by Gregory Hellbourg and Cedric Viou who presented their algorithm for filtering satellite emissions at the conference “Coexisting with RFI”. EMBRACE was used to test the algorithm.


Jess Broderick gave a talk at the 1st Pietro Baracchi Conference in Perth, Australia, and also presented our plans for an MFAA scientifically capable prototype called MANTIS.


And, of course, Wim van Cappellen presented MFAA at the SKA Engineering meeting in Stellenbosch.


Don’t forget SKA2!

Report provided by the AAMID consortium