Expanded Very Large Array Status

M. M. McKinnon and the EVLA Project Team

Antennas continue to be converted to the EVLA design. A total of 18 antennas are now used for scientific observations. EVLA antennas accounted for 62.5% of all antenna hours used in scientific observations at the array in October 2008. The electronics outfitting of the 19th EVLA antenna is nearly complete, and the mechanical overhaul of the 20th antenna is well underway.

Good progress has been made in the production and installation of EVLA receivers. Four of the wideband, fully EVLA-compliant, C-band (4-8 GHz) receivers have been assembled, tested, and installed on antennas in the array. Six new Ka-band (26-40 GHz) receivers have been assembled, tested, and installed. The RF and mechanical designs of the new Ku-band (12-18 GHz) receiver are complete. The OMT, phase shifter, and square-to-rectangular waveguide transition were scaled from the design of the K-band receiver. Full production of the Ku-band receivers will start in late 2009.

The development of the wideband orthomode transducers (OMTs) for the EVLA receivers continues with encouraging results. Two samples of the final version of the L-band (1-2 GHz) OMT were manufactured and delivered to the NRAO. The design of all EVLA OMTs takes into account the straightforward assembly of each device, an important issue for their mass production. Tests showed that both OMTs easily comply with performance specifications without any tuning beyond the original assembly process. The L-band dewar kits were completed in late November, allowing the cold testing of the L-band OMTs to proceed. Cool down tests of the final version of the S-band (2-4 GHz) OMT showed that the receiver’s refrigerator has the necessary capacity to cool the OMT to the requisite temperature. The OMT is the largest source of thermal mass in the receiver. The RF testing of the entire S-band receiver is nearly complete. The fabrication and testing of a prototype, planar-style OMT for the X-band (8-12 GHz) receiver are underway in Green Bank. An X-band dewar was assembled in Socorro and sent to Green Bank for the tests. As a risk mitigation measure, an alternative, turnstile junction OMT for the X-band receiver is being designed at the Central Development Laboratory. The design of its three primary sections (turnstile junction, stepped impedance transformer, and combiner) is complete.

Production quantities of the 3-bit, 4 Gsps digitizer chips were delivered to the NRAO in September. The production of the sampler modules that contain the digitizers was delayed due to a clocking problem encountered with a serial-to-parallel converter on the sampler board. The production of the sampler modules is currently expected to resume in March 2009 at the rate of about one antenna per month.

A phase instability problem of unknown origin has existed in the EVLA local oscillator (LO) and intermediate frequency system for some time. The problem was recently found to reside in a crystal oscillator in the antenna reference generator module. A solution for the problem is currently under development. The problem was found using a new measurement technique that divides the antenna LO signals down to 512 MHz, transmits them back to the control building, and measures the signals with the recently completed round trip phase modules.

Six station boards for the WIDAR correlator are being tested at the DRAO in Penticton, British Columbia, in preparation for their deployment to the VLA site in early 2009. The station board provides delay tracking and digital filtering for the correlator. These boards, along with the four in the existing prototype correlator (PTC), will be used to assemble an intermediate 10-station correlator in the racks for the final correlator.

The first set of the "stage 3" baseline boards in the WIDAR correlator have been built. The baseline board contains the correlator chips where the correlations are computed. The functional testing of these boards is currently underway in Penticton.

Industry standard, quality control tests of the correlator chips were successfully completed. The tests indicate that the chip is very robust from a reliability standpoint. All of the production correlator chips have now undergone a burn-in, screen test at full speed. The failure rate of the chips was 0.7%, due largely to incomplete scan-test coverage, and partially due to infant mortality induced by the burn-in screen. This low failure rate means that the spare chips on hand will be more than sufficient for the lifetime of the correlator. The chips have been delivered to the manufacturer of the correlator circuit boards.

A critical design review (CDR) for the correlator was held in Socorro on December 2-3.The purpose of the CDR was to determine the production readiness of the major circuit boards in the correlator. The results from the on-the-sky (OTS) tests with the PTC figured prominently in the review. Preliminary indications from the review indicate that board production can proceed provided that a problem with power supplies on the station board is resolved and that the remaining laboratory tests of the stage 3 baseline boards are successfully completed. Given the positive outcome of the CDR, the full production of the correlator circuit boards will likely commence in early 2009.

Modifications to the Executor software of the EVLA monitor and control system allowed OTS tests with the PTC to be carried out using EVLA antennas simultaneously with ongoing VLA observations. The tests included the preparation of scripts that were executed using a test executor. Graphical user interfaces were developed and used to configure and monitor low-level board components. The actual testing required the painstaking debugging of a series of low-level problems in both hardware and software.The OTS tests are now carried out routinely.

OTS testing with the PTC has also included the production of meta-data that conform to the Science Data Model (SDM), as well as binary data that conform to the binary data format (BDF) specified in the SDM. The data from the OTS tests have been loaded into CASA and AIPS for analysis. An ongoing effort to improve this process is underway. Changes to the Executor, correlator backend, and meta-data and capture modules allow SDMs to be produced without hand-editing, making significant progress toward the goal of automating the process of merging SDMs and BDFs. The generation of the SDM/BDF combination enables one to load the data into CASA and AIPS on a routine basis.

Several features were implemented in the Observation Preparation Tool (OPT) to aid the Ka-band observations that will be made in the near future. Among these were the ability to specify scheduling constraints, such as wind velocity and atmospheric phase, the ability to perform Doppler tracking, a richer calibrator search mechanism, and summary listings of the scans that comprise a scheduling block. Support for tipping scans and fixed-date specification of scheduling blocks was also added. The OPT is now ready for the next phase of testing, which will expose the tool to a wider audience and will involve running the execution scripts produced from the information collected in the observation configuration process.

A review of the Virtual Correlator Interface (VCI) was held. The VCI is the protocol through which tools such as the OPT will communicate configuration requests to the WIDAR correlator. The document describing the VCI lists the details of all correlator configurations and operations in all observing modes. Software is being written to implement the VCI. Work began on a user interface for configuring WIDAR. The interface will be the first phase of WIDAR support within the OPT. It will be used in correlator testing in early 2009.

A National Science Foundation mini-review of the EVLA project was held on September 10, 2008. The review committee was impressed with the excellent progress made on the project. The committee commended the NRAO “… for carrying out the EVLA upgrades while continuing operations with the VLA system, resulting in important scientific results by the user community during the interim construction period.” The committee also noted that the scope of software development appeared to have expanded to facilitate access by the broader astronomical community and advised the project to ensure the critical software required for EVLA commissioning and initial science operations be completed first. Finally, the committee encouraged the project to share its plans for commissioning and operations with the community as soon as possible.