Update on the JCMT Heterodyne Array Program

For several years, the heterodyne facilities on the JCMT have languished behind those for continuum. Although state-of-the-art at the time, most of the current 1 and 2-polarisation receivers and spectrometer are now almost a decade old! Consequently the JCMT will be undergoing several significant upgrades to the heterodyne system in the next 1-2 years; to include a flexible array correlator (ACSIS), low-noise B-band array (HARP-B), rapid realtime spectral display, queuing and advanced preparation of observing programmes, and data reduction using AIPS++. It is perhaps the latter two features which will directly affect how observers see the system, but it is the former that will provide the great improvement in its capabilities.

So why is all this being done? The first goal is to provide for rapid, efficient mapping in spectral lines; from the astronomers point of view, the system can then be viewed as a spectral line 3-D "camera", to complement the continuum capabilities of SCUBA and SCUBA-2. The second goal is to allow observers to efficiently prepare and queue up observations in advance and to track observation progress; this will take some of the burden off already overstretched support staff at the telescope.

Some of the performance gains of the new HARP/ACSIS system will be the following:

1. The mapping speed (to the same noise level) will be increased by factor of up to 30.
2. Widefield mapping will be possible (eg mapping 1 square degree of sky in less than an hour, with <1K rms/point).
3. Calibration accuracy, error detection and overall observing efficiency will be improved.
4. Several new observing modes will be provided (such as jiggle mapping and fast rastering).

The first phase of the heterodyne upgrade will be installation of ACSIS and a new control system, as well as the heterodyne Observing Tool by the start of semester 03a. The second phase will see the commissioning of HARP later in 2003.

ACSIS

The new array correlator and imaging system for the JCMT is nearing completion. It is being built mostly at the Dominion Radio Astrophysical Observatory (DRAO), in collaboration with the UK Astronomy Technology Centre (UKATC) and the Joint Astronomy Centre (JAC). Correlator boards have been manufactured (see pictures below), and reduction software is being completed and debugged. The final manufacture of other components such as the IF have been delayed due to shortage of effort, but it is planned that these will be available by the end of the year.

Initially, ACSIS will be integrated with the existing receivers, and so will replace the DAS. The highest resolution available will be increased by a factor of two, and maximum bandwidth will be limited mosly by the frontend. The primary bandwidth/resolution configurations are shown in the following table.

1. Exact usable bandwidth will depend on final ACSIS IF filters
2. Maximum bandwidth will be limited by Frontend bandwidth. Typically for existing receivers this is 2GHz

The high-level programmable nature of the new control and reduction system allows for relatively rapid development of new observing modes. It is hoped these will be refined over the first few months of use, and new ideas from observers are always welcome! However, the four primary observing modes available with ACSIS are shown in the following table.

Observing mode	Basic specifications	Typical uses
Raster-nod	Up to 1000 samples/row. 50 millisec/sample. Up to 1000 rows/map.	Widefield mapping
Jiggle-chop	16-point rapid jiggles using smu. Multiple ons per off.	Deep fully-sampled mapping of array field-of-view (120arcsec). Pointing.
Jiggle-frequency-switch	Jiggles with frequency switching	Efficient compact mapping
Grid/stare - nod	Takes samples at individual points	Point by point, or single point observing

In some observing modes, the raw data rate will be as much as 10Mbytes/second. To reduce this to managable levels, ACSIS has built-in realtime data reduction, so observers will normally take home data after it has been through a full reduction process, including tasks like baseline fitting, smoothing and - when imaging - regridding onto a final map. Results will normally be available as a standard FITS cube, or for simple low-datarate modes, as a relatively raw AIPS++ measurement set format. Data can then normally be reduced using AIPS++.

The detailed user manual is being written, and is available on : http://www.roe.ac.uk/atc/projects/harp/acsis_usermanual.doc

The following figures show a fully-populated correlator crate and a single board on the bench at DRAO (images from Tom Burgess).

Is a 350GHz 4 x 4 element heterodyne focal plane array using SIS detectors presently under construction. The work is being carried out by a collaborative group led by the Mullard Radio Astronomy Observatory (MRAO), in conjunction with the UKATC, Herzberg Institute of Astrophysics (HIA) and the JAC. SIS junctions designed by MRAO for HARP-B will be fabricated by the Delft Institute of Micro-Electronics and Silicon Technology (DIMES).

Working in conjunction with ACSIS, HARP-B will provide 3-dimensional imaging capability with high sensitivity at 325 to 375GHz. This will be the first sub-mm spectral imaging system on JCMT - complementing the continuum imaging capability of SCUBA - and affording significantly improved productivity in terms of speed of mapping. The core specification for the array is that the combination of the receiver noise temperature and beam efficiency, weighted optimally across the array will be <330K SSB for the central 20GHz of the tuning range (includes the CO_3-2 line), i.e. - each of the 16 mixers should have a noise temperature better than the current RxB3 receiver.

In technological terms, HARP-B brings together a number of interesting and innovative features across all elements of the design. One important example is the design of the 4 x 4 imaging array module, which has a `state of the art' SIS detector design and a novel local oscillator coupling scheme. The design also uses a cryogenically cooled Mach Zehnder interferometer for sideband filtering to minimize contributions to system temperature. The image below shows HARP-B on the JCMT. You can clearly see the HARP-B `front-end' situated on the RHS Nasmyth platform and the HARP `K-mirror' (a beam rotator to maintain array orientation on the sky) situated in the cabin.

So how are we doing?

Being somewhat simpler than the HARP-B camera, the K-Mirror has always been seen as almost a separate instrument and design and build progressed accordingly. In fact it has mirror surfaces that would allow any instrument operating down to about 450mm to couple efficiently to the telescope when suitably mounted on the RHS Nasmyth. The K-mirror has already gone through preliminary acceptance trials and should be shipped to JCMT and installed during the summer 2002 shutdown. The image below was taken at the initial acceptance test.

The HARP-B `front-end' itself is also progressing well. Most of the experimental prototyping has been done and the design is essentially complete. A full, externally moderated `Final Design Review' is planned for early summer this year. This will trigger the main build phase, which will last approximately nine months. After that comes nine more months of integration and testing in the UK before shipping to Hilo for installation & commissioning - presently planned for late 2003.

The heart of the receiver, the prototype array unit, `looking out' of the MRAO test Dewar as it was being assembled for an LO coupling test. Note the 16 pixels!

A single prototype mixer block constructed using MRAO's `split block' technique showing the corrugated feed-horn, waveguide, SIS junction slot and IF board pocket - about 4 weeks work in less than 30g of aluminium -machined to a tolerance of five microns.

The new JCMT Observing Tool is based on the UKIRT OT, currently being developed initially for use with SCUBA at the UKATC and JAC, and then to be released with ACSIS. However most of its facilities will be common to both heterodyne and continuum observing. The idea is that once the time is allocated, the observer can prepare their PATT programs in advance. The OT relies on the concept of "Minimum Schedulable Blocks" (luckily, shortened to MSBs), which define parts of an observation programme which can be done independently. Thus, once saved to the JCMT observation database, an MSB could be done at the most convenient and efficient time.

The example shown below is a project to make jiggle maps of three Galaxies with Harp-B. On the far left is the overall observing plan, in the form of a nested set of components describing different aspects of the programme. In the midle the components such as map area and integration times can be defined. On the righ the target and map area can be overlayed and adjusted on optical catlaglogue images (or this could be an IRAS 60um map, for example). Although it will be possible to define a programme from scratch, for many observers a common way to prepare their run will be to read in and edit a programmes from a library of sucessfully tested routines.