The Correlator for the Expanded Very Large Array (EVLA)

With the recent announcement of funding in the Canada budget, the EVLA correlator being built at NRC’s Herzberg Institute of Astrophysics (DRAO) is able to proceed to completion. This provides an opportunity to contribute to the building of what can only be described as a "new telescope", such is the extent of the anticipated new capabilities of the EVLA compared with the VLA’s present capabilities. The initial improvements will be in the areas of sensitivity (~10x in continuum), wavelength coverage (continuous from the 20-cm hydrogen-line band to ~7 mm wavelength), and spectral resolution (>1000 x). Later improvements will extend the wavelength coverage at the longer end, improve the sensitivity to faint extended emission by adding a compact array-configuration, and greatly increase the resolution by the addition of antennas in an array several 100 km in size (throughout the state of New Mexico). This will certainly provide new scientific opportunities to the Canadian astronomy community, complementary to those now available through the Gemini telescopes, and eventually ALMA. The VLA has been one of the most heavily used and productive telescopes in astronomy. Its range of applications is as broad as astronomy itself, from solar studies to cosmology.

The Canadian "WIDAR" correlator design is responsible for a sizable portion of these improvements. Its 16 GHz of total bandwidth will enable much of the telescope’s sensitivity improvement. The correlator will produce 16000 spectral channels at full bandwidth, and up to 250,000 channels at narrower total bandwidth. The total bandwidth can be sub-divided – each sub-band can be "targeted" for simultaneous observations of different spectral lines with separate spectral resolutions. All modes of observation can accommodate the extraction of the polarization properties of the emission. Its flexibility and dynamic range are one of the keys to being able to increase the frequency coverage beyond the protected radio astronomy bands. An important new pulsar capability will be added, two banks of 1000 "phase bins". Full spectral capability with high time-resolution will permit the acquisition of time-frequency maps of pulsar emission. In addition the correlator will be able to combine all the EVLA antenna-signals into a single-stream output (as if it were one VLBI antenna), and simultaneously process VLBI and connected-element arrays.

These things can be accomplished only through the latest digital signal processing technology, along with engineering techniques that have evolved recently. The EVLA correlator staff at DRAO is now four engineers, soon to be five with the arrival of a new software engineer, assisted currently by a few others. The detailed design phase will happen over the next few years, followed by an intense period of manufacture by industry and assembly at the new DRAO laboratory facility. The current schedule has the first "shared risk" observing beginning in 2007, as by that stage the new correlator will surpass the VLA’s present correlation capability. Installation and commissioning will continue until 2009. The 2007 date, to which our US colleagues are working, is an exceedingly tight schedule.

This technology continues to advance at an astounding pace. The EVLA correlator work will provide a foundation of expertise for even more advanced processing systems for astronomy, especially for the Square Kilometer Array, in which digital beam-forming and correlation will play a very important role.

Peter Dewdney