The eSMA

"Submm Interferometry with the combined SMA-JCMT-CSO array"

Introduction

Lab tests at the SAO

The eSMA equipment was installed in a dedicated bay in the JCMT cabin during June. Mark Bentum (ASTRON/WSRT, The Netherlands) visited the JAC to take charge of this effort with financial backing from a NWO-M grant awarded to Prof. Ewine van Dishoeck (Leiden) in support of the development of the eSMA. In spite of having to sort out a number communication and receiver issues, this work proceeded quickly and first fringes between the SMA and JCMT were obtained at 267 GHz on July 12, 2005 at 11:07HST, well ahead of the target-date in mid-August. For these inital tests only a single LO was available at the SMA forcing a choice of receivers with compatible LO multiplication factors: i.e. the 345 GHZ receivers at the SMA and the 230 GHz receiver at the JCMT, which have an overlap in frequency coverage from about 260-270 GHz. The photos below show the first fringes with the JCMT configured as 'antenna 8' and the team celebrating their success. Since the exact baseline with the SMA had not been determined yet, the fringes on the JCMT baselines start wrapping with time.

The eSMA interface equipment for the CSO was delivered in September and first fringes with the whole eSMA were obtained on Oct. 31. Illustrating the benefits of the large collecting area the first baseline to show clear fringes was the JCMT-CSO one!

First fringes!
First fringes of the SMA-JCMT interferometer observed on July 12, 2005: phase as a function of time. The JCMT is configured as antenna 8 of the array. Due to an imprecise baseline solution at the time, the fringes on the JCMT baselines start wrapping rapidly as time proceeds.

Commissioning

During the autumn of 2005 eSMA commissioning was carried out on several evenings, first with the JCMT alone, later including the CSO. Initial baseline tests proved to be discouraging, since a proper solution seemed to remain elusive: 'decimeters' uncertainty had to be reduced to 'millimeters' before an attempt could be made to further refine the solution by observing astronomical calibrators. This is less straightforward than it may appear: the baselines are between the axes-intersections of the telescopes rather than the nominal survey coordinates for the observatories or the antenna locations of the GPS receivers. Taco (Ken Young) managed to do the impossible by finding the solution in spite of what later turned out to have been a mix-up of axes in the transformation of 'Hawaiian survey' to astronomical coordinates.

Taco also undertook another major development: a complete revision of the SMA tuning software in order to be able to exactly line up the sky frequencies of receivers driven by two different LOs. During the early autumn the SMA implemented their second LO system primarily intended to support dual frequency observations e.g. 230 GHz and 690 GHz. The same system could in principle be used to obtain interference between the 230 GHz receivers at the SMA and and JCMT which use different multiplication factors and thus require a different LO frequency to tune to the same sky frequency. However, for successful interferometry of the IF coming from the respective receiver the two LO frequencies had to be fine-tuned and the sky frequencies aligned using secondary oscillators in the system. This required a major update of SMA's tuning software, but first spectra at 230 GHz were successfully observed in early October.

A number of basic operations of the eSMA have been checked out successfully and the baselines were refined to mm accuracy. After the JCMT shutdown commissioning we plan to continue with phase-stability and imaging tests and the switch-over to 345 GHz. This will be followed by a 1-year pilot programme. A workshop on the pilot programme is being planned for the autumn of this year (details to follow).

The first spectra with the eSMA of the CO(2-1) line in CRL2688. The figure shows the JCMT baselines (configured as antenna 8), which to a varying degree resolve the source. The total integration time was about 10 minutes.

RxW conversion

Late this year and early next year the SMA will add 400 GHz receivers to the frequency line-up. These receivers overlap with their 300 GHz receivers over most of the 345 GHz window, but observe the orthogonal polarization. For the eSMA to take advantage of this dual-polarization capability, a state-of-the-art dual-polarization receiver is needed at the JCMT with an IF frequency of 5 GHz. HARP by contrast is a single-polarization receiver. RxB3's mixers are no longer state-of-the-art and its mechanics would require a significant overhaul and after a careful analysis it was decided that a better option would be to convert one of the frequency channels of RxW to 345 GHz. In March the JCMT Board gave the go-ahead to convert RxW's 460 GHz channel.

At the time of writing, development of new 345 GHz mixers and a re-design of the optics has started at MRAO. The new mixers are based on the HARP mixers and should provide the JCMT with a superb dual-polarization single-pixel 345 GHz receiver. We are also investigating the possibility of upgrading the 690 GHz mixers with ALMA new junctions from SRON (Groningen, The Netherlands), although this is a more speculative endeavour. We hope to be able to reinstall a converted and upgraded RxW in the autumn of this year, and, once in place, it will become the primary receiver for observing with the eSMA interferometer.

Remo Tilanus,
eSMA Project Manager JCMT

RxW optics inside the cryostat in front of the focusing lens and 460 GHz mixers (top); proposed redesign of the optics to fold the beam into the 345 GHz HARP-style mixer block using a curved mirror (bottom).

The SMA