Radio Astronomy SA

The CALLISTO spectrometers are based on a programmable heterodyne receiver built in the framework of IHY2007 and ISWI by former Radio and Plasma Physics Group (PI Christian Monstein) at ETH Zurich, Switzerland.  The main applications are observation of solar radio bursts and rfi-monitoring for astronomical science, education and outreach.  The instrument natively operates between 45 and 870 MHz using a modern, commercially available broadband cable-TV tuner CD1316 having a frequency resolution of 62.5 KHz.  The data obtained from CALLISTO are FIT-files with up to 400 frequencies per sweep.  These receivers can be configured in various ways to allow extended frequency ranges and/or observation modes identified as Focuscodes, using simple or complex antennas and associated hardware.

Our initial goal was to get something together and working out at Sunnydale, a settlement about 100km N-E of Adelaide, close to the River Murray and within the River Murray Dark Sky Reserve.  The proposed site was in a paddock which owners David & Tique had reserved for astronomical observations and they very kindly allowed us to set up some equipment on their property.  There was no power or buildings on-site so whatever we put in would have to be largely self contained, fortunately there was a 4G tower just across the river in the town of Nildottie.  Having originally picked a transit Log Periodic Dipole Array (LPDA) antenna design for the project,  Blair began work on building a tower with solar tracking capability while Peter started on the LPDA antenna, low noise amplifier (LNA) enclosure & it's contents and a suitable computer.  We purchased a single Callisto receiver off-the-shelf from Switzerland with the generous assistance of the Astronomical Society of South Australia (ASSA).
DC power would be harvested by a pair of 80W solar panels with a 12V lead-acid battery for storage, managed by a Victron MPPT solar controller.  A Raspberry Pi configured as a Callisto-Lx would control the receiver using software purchased from Whit Reeve in Alaska.  Apart from the battery, everything was to be mounted somewhere on the tower.  Most of the electronics was stuffed into a 550mm x 550mm x 450mm SpaceCase to be mounted partway up the tower, to allow the wombats and kangaroos to pass by largely undisturbed by our presence.  The LNA enclosure was located as close to the antenna as possible and the solar panels were mounted above the SpaceCase to provide some shade during our summer months.
With the assistance of ASSA members John, Joe, owner David and a massive hammer drill, some large holes were drilled into the metre plus thick limestone rockshelf below the thin soil surface.  Multiple 20mm galvanized threaded rods were then chemically bonded into the rock for fixing the tower to.  A small concrete base was formed around the steel rods to support the tower and battery box. Several weeks later after the concrete went off, we returned and erected the tower.
Remote communication access via VPN to control the spectrometer was established and we were able to start, stop, check the status of the system at any time from Adelaide or where ever we were when a operational problem occurred.  Hardware issues were another matter and generally required a site visit which meant a round trip of 3 hours travel for Blair and about 5 hours for Peter.

We did try a lower gain configuration for a while, the second LNA was fitted with a 10dB attenuator to provide +26dB gain but this proved to be inadequate and the attenuator was eventually removed.  A few weeks after the tower went up, David & Tique installed a small container close by for their astronomical activities, this provided a Wi-Fi link back to the house and reliable internet access for our activities.  As Solar Cycle 25 began to ramp up the Sun became more active and it soon became apparent that much of the solar radio bursts were occurring below the range of our LPDA antenna.

Rather than miss out on the fun, it was time to develop Phase 2.

We had earlier considered a Long Wavelength Array (LWA) antenna based spectrometer but dismissed the idea because of cost considerations.  Building the LWA antenna was not a difficult task and producing our own version of the Front End Electronics (FEE) wasn't either.  The problem was in order to fully utilise the radio spectrum below the FM broadcast band, we would need an upconverter and they tend to be expensive.  As a compromise, Phase 2 began with a plan to scan frequencies between 45MHz to 160MHz using a single Callisto receiver, LWA antenna, a pair of FEE boards and combine the N/S & E/W outputs at the receiver.  Fortunately, by the time that we had built the antenna, designed and built the FEEs and convinced ASSA that we needed some assistance procuring another pair of receivers, the opportunity to build our own upconverter presented itself at a reasonable cost.  Phase 2 was a short lived but useful exercise, initially carried out at Middleton rather than Sunnydale, which provided a solid foundation for Phase 2.5 and it's subsequent upgrades/enhancements.
Like all good engineering projects, it sounded simple enough but presented some additional issues that needed to be addressed along the way.  The addition of two FEEs, two new receivers and another two Raspberry Pi computers was going to be a significant drain on the existing DC power system at Sunnydale.  In an effort to reduce power consumption, particularly overnight when the spectrometers are idle, Blair designed and built an autonomous GPS Switch which would be used to shut down most of the equipment in the SpaceCase on the tower and powered it back up prior to sunrise.  The two 80W panels on the tower were replaced with two 190W panels located at ground level, just north of the tower footings.  The wildlife will need to adapt to these new arrangements.
As David & Tique developed their astronomical interests within the container, they installed an industrial PC within and we were offered the use of this machine to control all three recievers out at the tower, further reducing power consumption with the replacement/removal of the Raspberry Pis.

The development of Phase 2.5 began to ramp up when a quadrature hybrid coupler with a suitable frequency range and SMA ports appeared on eBay in September 2020.  For a little less than AU$100, it had arrived in Adelaide before the end of October.  This critical coaxial component forms the heart of our new up-converter, along with a pair of mixers (also sourced on eBay) and a 200MHz local oscillator (LO).  Duncan, who had recently moved from NSW to SA, kindly offered to assist with the 200MHz LO so we were soon well on the way to a significant capability upgrade.

Remembering that the Callisto Receiver can only scan down to 45MHz in it's native mode, the second image shows the final configuration of the Phase 2.5 Spectrometer.  The LWA antenna and it's two ASSA PHI (LNA) boards feed signals from the antenna (10 MHz to 100 MHz) into the up-converter.  The primary function of each Bias Tee is to pass DC power out to the the ASSA PHI board.  A quadrature hybrid coupler is used to combine the signals with a high degree of isolation between all four ports.  In our application, two linearly polarized perpendicular antennas - crossed-dipoles - can be used to receive circular polarizations by combining the antenna outputs after phase shifting one of them by 90°.  When a quadrature coupler is used to combine the outputs, the antenna system can discriminate the two rotation directions of circularly polarized radio waves - one output provides a response only for RHCP and the other output provides a response only for LHCP.  The two dipoles in the LWA crossed-dipole antenna are oriented north-south and east-west and, thus, are perpendicular.  Each dipole is sensitive to the linear component of the incident radio wave that is aligned with it.  Both signal outputs are passed on to each mixer where they mix with the high level 200 MHz local oscillator output.

This produces two similar complex signals with different phase components within the range from 210MHz to 300MHz, one for each polarization.  The receivers then scan from 215 to 288 MHz, avoiding the fundamental and other up-converted complex harmonic signal components, effectively receiving signals in the range from 15MHz to 88MHz.  The Callisto software then sorts everything out.
Prior to this final configuration, we were using two of our original ASSA FEE (LNA) boards with the LWA antenna.  These were replaced with the new ASSA PHI boards in May 2024, requiring some modification to the up-converter by removing the 15dB coax. attenuators fitted to decrease the FEE(s) output levels.  The ASSA PHI boards were designed with an onboard attenuator to achieve a similar level at the input to the quadrature hybrid coupler.

The Murchison Widefield Array (MWA) antenna had been of interest to us for some time.  When we initially discussed what and how we could contribute to e-Callisto back in 2017, the MWA was given some consideration but our learning curve would be simpler if we followed the LPDA path initially.  The MWA was parked in a corner but not forgotten.  Following installation of Phase 2, Blair contacted the MWA Project Officer and cheekily asked if it was possible to obtain several of the LNAs used in their array, they were building lots of them and hoped there might be a slim chance to obtain a couple.  He explained what we had been up to and wanted to expand our capability by using their VHF antenna.  The antenna design was published on the web, but the schematics/board layouts for the LNAs were not and given their small size, would be difficult to replicate accurately.  The then Project Officer, Mia, emailed back and said "Sure, I'll send you a couple" and two months later, Qty 3 arrived in Adelaide.  Blair replied to thank her for the donation and asked if there was any chance of getting the schematics and other details.  Mia was able to assist there as well.  We are truly greatfull for the assistance and parts we received as without them, the following could not have taken place.  Peter had started work on making the VHF (70MHz to 300 MHz) crossed polarized antenna and once the request to MWA had paid off, he really got going.  Rather than approach ASSA again for another significant contribution toward the cost of buying the two receivers from Switzerland, Peter had been inquiring about the purchase of the bare printed circuit boards and some "critical parts" that would be difficult to purchase locally.  e-Callisto PI Christian, agreed to sell us several sets of PCB & components and quickly posted them off to Australia.  Having completed the manufacture of the new antenna, it was time to start building up the two additional receivers we would need and two additional Raspberry Pi's in a nice stainless steel enclosure which had been donated for the project, thanks to ASSA member Gerard.  Once the two receivers were completed, aligned and tested, everything was bolted up and crammed into the enclosure, ready for on-site installation.

This new capability gives us a better coverage into the VHF region, using the LNA's donated to us by the MWA Project Office and our MWA prototype antenna.  The word initial is used because Duncan has some hardware suggestions that could provide some provide significant performance improvements over the current configuration.  Everyone, and I mean everyone onsite, was into whatever needed to be done.  While we were slogging it out in the Sun, Tigue was preparing a refreshing lunch for everyone.  David did the heavy lifting with the JCB to clear the antenna site for an oversize ground plane of galvanized mesh donated by Duncan and Peter.  Beverly and Duncan prepared and laid the conduit enclosed dual coax. run from the container out to ground plane and connected the MWA antenna.  The shallow trench was partially filled using the JCB and shovels using sand infused with limestone particulates to order save the HD (orange) conduit from UV deterioration.  The day was looking older as the Sun began to sink toward the horizon, we were all feeling significantly older along with it.  The march flies 'bite' had left a scar on one or two of us and it was time to head elsewhere, or at least home.  When the MWA enclosure was powered up just prior to our departure, only one of the two Pi's powered up correctly.  So for the first few weeks of operation from late March 2023, Australia-ASSA_datecode_timecode_57.fit was able to be sent to, and images observed from the e-Callisto Observations webpages, along with the  60,  62 &  63 spectrometer images.

Blair and Peter did revisit the site and changed out the faulty Raspberry PI so eventually we had all 5 receivers working.  Winter was soon upon us, the increase in power consumption of the MWA system and poor solar power harvesting had caught up with us (again).  We had to turn off the new MWA system until we could return and upgrade the solar installation on the container.

The cause of the failed Raspberry Pi (56) was attributed to the Callisto-Lx software had been imaged onto a 32Gb microSD card, which may have been faulty.  We normally deploy the Pi model B+ for on-site Callisto-Lx because of it's low power consumption, the microSD cards we use have 16Gb storage capacity.

As one can see in one of the MWA images above, the two spectrometers are a snug fit within the enclosure.  Held together with M4 stainless steel fasteners (screws, washers & nuts), experience over some time in the field has shown how easily fasteners can disappear into the dirt/dust/sand onsite during maintenance operations.  Some time ago, Blair suggested that the installation of M4 nutserts would reduce the potential for "lost hardware items" because of "finger trouble" and environmental factors.  In mid-October 2024, the MWA spectrometers were shut down and temporarily removed from the site for some minor modifications.
 

1.   Replace the original M4 nuts & washers with M4 nutserts, where practical
2.   Replace the original Cat5e UTP cables back into the container with Cat8 shielded cables to help reduce induced RFI within both the enclosure and the container
3.  Test and evaluate both receiver's sensitivity (MDS) to help determine the considerable differences in fits data/image responses observed since the spectrometer was originally installed in May 2023

This task was completed by the end of October and the MWA spectrometer was back online.  As usual, there were some complications encountered as a result of these changes and several creative solutions were implemented to resolve these issues.   The results of these activities may take a while to become evident, we will be watching with much interest.

The modified MWA spectrometer at the end of October 2024.

Images and text still to come - site under construction.