An external review of the de-multiplexer on the 3-bit, 4Gsps sampler board was held October 1. The samplers are used to digitize the 8 GHz of bandwidth per polarization from each EVLA antenna. The review was held to evaluate alternative de-multiplexer designs because the current design did not perform as expected. The primary recommendation from the review was to abandon the current design and adopt a new design based upon a high speed Field-Programmable Gate Array (FPGA). Of the alternatives considered, the FPGA-based design was technically the most feasible and had the lowest impact on the project budget and schedule. The layout of the new sampler board is nearly complete. The production order for the boards should be placed by April 2010, and we expect the installation of the boards in the antennas to be complete by September 2011. The change in de-multiplexer design delays the availability of 8 GHz observing capability, but it does not delay the overall completion of the EVLA project.
A review to select the design of the orthomode transducer (OMT) for the EVLA X-band receiver was held October 13. Three different designs were considered. Originally, the selection of the final design was thought to be closely coupled to the ability of a CTI Model 22 refrigerator to cool the receiver. However, an evaluation of the cryogenic and electrical infrastructure at the VLA site indicated that the current versions of these systems would operate at the margin of their capacity, thereby placing the long-term telescope reliability in jeopardy. Thus, for operational reasons, a decision was made to upgrade the cryogenic and electrical infrastructure and use a larger, Model 350 refrigerator on the X-band receiver. The Model 350 is the same refrigerator used on the adjacent C- and Ku-band receivers. The infrastructure upgrades include the addition of a fourth cryogenic compressor and a new electrical power panel on each EVLA antenna, and two new utility transformers at the VLA site. This decision effectively decoupled the refrigerator issue from the OMT design selection. The decision also mitigated the highest impact risk to the EVLA project. A quad-ridge style OMT, similar to those on the EVLA L-, S-, and C-band receivers, was chosen for the X-band receiver. The installation of fully EVLA-compliant X-band receivers will start later than predicted but should still be complete in late 2012 as scheduled. In the meantime, the Ku-band receiver deployment will advance since its design is mature and tested.
An inspection of the azimuth and elevation gearboxes on all antennas was completed in November. The inspection was conducted because a pinion gear in the azimuth gearbox of antenna 5 failed in April 2009, and we wanted to know if other antennas in the array had similar problems. Overall, the inspection results were not as dire as initially feared. The inspection revealed that all elevation gearboxes are in good condition. Gears in an azimuth gearbox on each of antennas 6 and 7 showed extreme wear. They will be repaired as part of routine mechanical overhauls of the antennas once all EVLA antenna conversions are completed in summer 2010. A plan has been instituted to replace the lubricating oil in all gearboxes more frequently.
Twenty-four antennas have been converted to the EVLA design; the twenty-fifth antenna will be converted by 1 January 2010.
The production, installation, and testing of the circuit boards for the new WIDAR correlator continues. The last set of station boards was delivered to the VLA site in early November. These boards are needed for the Early Science observations that will be made with WIDAR beginning March 2010. The number of baseline boards currently on hand is sufficient to accommodate Early Science observations. The last set of baseline boards will be delivered to the VLA site by April 2010. Tests with an initial subset of the correlator, called WIDAR-0, have been hampered by system timing instabilities, which have been attributed to phase lock loops (PLLs) on the station boards. The PLLs were recently removed from the boards.
The configuration mapper is the low level software that automatically configures the correlator for observations. The software has been written and is being tested with WIDAR-0. The correlator configurations for Early Science observations are supported in a recent version of the Observation Preparation Tool, the high level software used by observers to set up their observations.
Science observations with the VLA correlator will cease on 11 January 2010 so that the final transition to the new WIDAR correlator can be made. The hardware modifications to be made during the transition include the transfer of the intermediate frequency (IF) signal path from the VLA correlator to the WIDAR correlator, a rebuild of the IF distribution racks, a rework of the central local oscillator (LO) racks for the antennas, and the overhaul of the master LO rack. The hardware modifications will be complete by 9 February. Science observations with WIDAR are expected to commence in March after a period of test observations.