31/05/15 PSR J1141-6545 has been seen orbiting its white dwarf companion every 4.8 hours using the UTMOST telescope.

The system is a binary radio pulsar that consists of a neutron star and white dwarf in an eccentric orbit (e=0.2) that exhibits a wide range of relativistic phenomena. The binary completes five orbits per day and has a large rate of advance of periastron(5.33^o/yr) explained by Einstein’s General Theory of Relativity.

Data were taken by PhD student Vivek V Krishnan on the night of May 30th 2015. The error bars are too small to see in the plot, which shows the change in the pulsar’s frequency as a function of time, over a little more than a complete orbit.

Image credits: Vivek V Krishnan and Matthew Bailes.

31/05/15 UTMOST’s new GPU install took place on May 26th. Eight 4xGPU NVIDIA TitanX servers have just been installed into the right hand rack pictured below, next to our set of 11 of 2xGPU NVIDIA GTX690s (installed in June 2013, and in the left hand rack). The new units were fully operation by the end of the week of the install, with correlation and pulsar observing possible using every element of the array. The system can peak to 258 TFlops of processing power. Observing with UTMOST presently produces 22 GBytes of data per second.

A/C ducting was added during the week to keep these GPUs cool, along with electrical work for the increased power requirements.

First fringes, using the quasar 3C273, measured across the entire array (all 352 antennae) and 15 MHz of bandwidth. 30 MHz of bandwidth will be possible with the new system, and will be tested in early June 2015.

Our new best observation of the bright Southern millsecond pulsar, 0437-4715, is shown above, with data being taken from all 352 modules of the UTMOST array during the week of the GPU install.

The DM (dispersion measure) trail of the millisecond pulsar 1909-3744 – from data taken on 31st May 2015 by PhD student Vivek V Krishnan. This pulsar’s very high rotation rate and particularly narrow pulse permits some of the most accurate timing possible with pulsars.

Congrats to everyone involved in this major installation – Andrew Jameson, Vikram Ravi, Matthew Bailes, Chris Flynn, Vivek V Krishnan, Pablo Rosado, Duncan Campell-Wilson, Tim Bateman and Dave Temby.

19/05/15 The UTMOST project has achieved full connectivity of the 352 individual sections of the telescope (“modules”) on optical fibre to our central building, where all the incoming data are processed and analysed.

The last few fibre connections with stable data transfer were finished in mid May, 2015. The total data rate down the fibres is an impressive 563 Gbits/sec — which needs to be processed for the most part in real time by our GPU cluster.

Congratulations to Duncan Campbell-Wilson and Dave Temby for reaching this major milestone.

28/05/15 The UTMOST project passed a new milestone as the tied-array beam — which makes full use of the power of the MOST array —  was applied to pulsar timing during April 2015. We were able to measure Times-of-Arrival of the pulses for over 30 pulsars in a day. Our timing to date had been restricted to the transit of pulsars across the meridian (when the telescope is pointing directly upwards) but during April 2015 software improvements allowed us to do this anywhere on the sky.

The image shows the pulsars timed in this way. The most up-to-date list can be found here. Each image shows the pulse at the bottom, and its dependence on the integration time. dispersion measure and frequency. The times of arrivals allow us to measure the changing rotation rate of the pulsar, including occasional “glitches“.

All these pulsars are timed individually, one at a time, but software upgrades currently being implemented will allow us to time as many pulsars as are visible in the large primary beam (8 square degrees) of the telescope. This will lead to huge improvements in the number of pulsars we can time, as our “30 pulsars in a day” campaign was dominated by slew times around the sky, rather than actual time spent observing pulsars. There is  substantial room for efficiency improvements in the near future.

Image Credit: Chris Flynn, Fabian Jankowski and the UTMOST team

23 April 2015 : The UTMOST FRB search program has now attained the full area of 8 square degrees of sky coverage – the complete field of view of the telescope. Until now, only a fraction of the telescope beam was being probed for FRBs. The 4×2 degree beam is being “tiled” with 352 thin “fanbeams” — each of which covers about 45 arcseconds East-West, and about 2 degrees North-South.

In one of our search modes, we point the telescope at the meridian at a declination which means the pulsars Vela and 1644-4559 to pass through the search area once every sidereal period. Individual pulses from these sources test our pipeline, as they are quite similar to what we expect to see with FRBs.

The plot shows a long FRB search — lasting almost 2 days — and shows Vela and 1644-4559 passing twice through the search area. Individual pulses from both pulsars show up clearly — Vela (blue) and 1644 (magenta) as they cross through from East to West). Narrow bursts of radio emission have been searched for in the data, and narrow ones (<32 ms) are marked. Grey shows events with low signal to noise SNR<10, while green shows SNR>10, and the SNR>10 pulsar pulses have been isolated out as known sources and marked in blue and magenta. Vertical green lines mark what is predominantly handset generated RFI near the site, and as they are seen spread through many/most of the fanbeams, we can mainly coincidence them out. The distinctively different dispersion measures (DM) of the two pulsars is seen in the upper panel. The pulse properties of these objects — period, width and DM, are all in excellent agreement with expectation, validating the pipeline.

A third pulsar, 1430-6623, has been rediscovered in our FRB search, from its individual pulses, seen in the plot above as a magenta sweep through the 352 fanbeams of the search area (lower panel) and at its distinctive DM of 480 DM units (upper panel). We serendipitously found the pulsar in our FRB search data, when the telescope was left at its declination due to mechanical work. 1430-6623 was originally discovered at Molonglo, and it is nice to validate our burst search procedures in this blind manner with it.

Data like this over a full 8 square degrees of sky has now been collected for most nights and weekends since Easter 2015, and is currently being analysed in the search for FRBs.

27/Mar/2015 An important validator of our FRB search techniques is to test if we can see individual pulses from pulsars — as they are so similar to what FRBs will look like.

To date, our tests have been using the bright Southern pulsar Vela. On 26th March 2015, we set our FRB search running in a new record sky coverage of 88 fanbeams (or about 2 square degrees on the sky), and allowed a considerably dimmer pulsar, J1644-4559 to pass through the telescope’s field of view. This pulsar pulses about twice a second, has pulses about 10 ms wide, and a dispersion measure (DM) of about 500, and is thus likely to be typical of the FRBs we hope to see with UTMOST. It’s an excellent test of our FRB catching techniques.

J1644-4559 showed up beautifully. We detected over 80 individual pulses, with the brightest being at a S/N of about 18. The plot below of detections in different fanbeams as a function of time shows a clear trail of pulses (green) against very low level RFI (grey) and some brighter RFI seen in multiple fanbeams (red).

A selection of these detected pulses is shown below, over a “Pulse Phase” which corresponds to 1 second of time (J1644-4559 pulses about twice a second). The sweep of the delayed arrival time of the pulses as a function of frequency are caused by the ionised Interstellar Medium. FRBs are expected to have similar or greater sweeps of this type, indicative of passage through the Intergalactic Medium. J1644-4559 is thus an excellent test of our FRB sensitivity and entire data reduction pipeline.

UTMOST has now reached a full 2 square degrees of search area on the sky – already a quarter of the final area that will be searched — just a few weeks after our first FRB search which covered a mere 0.03 square degrees.

Image credits: Manisha Caleb and Chris Flynn

Work on the high speed internet link into Molonglo via a microwave link has begun, with the pouring of the foundations for the tower. The new link feeds in from the North-West over Sharrow Hill.

The new link will operate at 150 Mb/sec, with burst speeds of up to 300 Mb/sec, and will substantially improve on our present links in via a longer, lower speed microwave link coming in from the South and our (much slower still) ADSL line.

Image credits: Anthony Goonan, YLess4U.

March 13th, 2015. A 36 hour search has been carried out for FRBs using a single fanbeam pointed at the meridian at declination -44 degrees, as part of our science commissioning for UTMOST.

Individual pulses from Vela clearly show up in the burst search as it transits through the single fanbeam (which is a few seconds across in time on the sky). The brightest pulses have a S/N of about 30.

This test of a single fan beam was successful, in particular for characterising the amount of RFI (radio frequency interference) affecting the system.

We will start a 44 or 88 fanbeam search in the coming week.

UPDATE: Our first 44 fanbeam search was performed on the night of 17th-18th March. Vela was seen transiting through every beam, making the test a great success, as shown in the figure.

UPDATE #2 : Our first 88 fanbeam search was made on 26th March 2015 — giving us almost 2 degrees of search area on the sky. News on our progress with 88 fanbeams can be found here.

We have trialled for the first time live measurements of individual pulses from Vela. The figure shows about 10 seconds of data, for the telescope in “tied-array beam mode”, adding the signal from a quarter of the telescope’s modules (88) coherently.

Individual pulses are seen at phase ~0.9 in the figure, with a range of S/N, as expected. The mean S/N is about 7. About 110 individual pulses were seen. It was gratifying to see such clean pulses, despite a lot of phone call activity going on at the time (several mobile handsets operate in our frequency band) which show the power of having an array!Seeing individual pulses live is an important step on the way to searching for FRBs.

Credit: Vikram Ravi, Manisha Caleb, Chris Flynn.

CAASTRO has issued this press release on a Fast Radio Burst seen in real time at Parkes Radio Observatory. The properties of the burst are published as Petroff et al 2015, and involve a number of people on the UTMOST team.

A major goal of the UTMOST is to find FRBs at the lower frequencies at Molonglo (843 MHz) than at Parkes (1.4 GHz), in our BURST project, taking advantage of Molonglo’s huge field of view (about 8 square degrees on the sky).