Status of VLTI control system: how to make an optical interferometer a data producing facility

Anders Wallander, Javier Argomedo, Pascal Ballester, Bertrand Bauvir, Mauro Comin, Rob Donaldson, Philippe Duhoux, Alberto Gennai, Bruno Gilli, Nico Housen, Alexis Huxley, Robert Karban (+5 others)
2002 Advanced Telescope and Instrumentation Control Software II  
After having established routine science operations for four 8 m single dish telescopes and their first set of instruments at the Paranal Observatory, the next big engineering challenge for ESO has been the VLT Interferometer. Following an intense integration period at Paranal, first fringes were obtained in the course of last year, first with two smaller test siderostats and later with two 8 m VLT telescopes. Even though optical interferometry today may be considered more experimental than
more » ... le telescope astronomy, we have aimed at developing a system with the same requirements on reliability and operability as for a single VLT telescope. The VLTI control system is responsible for controlling and coordinating all devices making up VLTI, where a telescope is just one out of many subsystems. Thus the pure size of the complete system increases the complexity and likelihood of failure. Secondly, some of the new subsystems introduced, in particular the delay lines and the associated fringe-tracking loop, have more demanding requirements in terms of control loop bandwidth, computing power and communication. We have developed an innovative generic multiprocessor controller within the VLT framework to address these requirements. Finally, we have decided to use the VLT science operation model, whereby the observation is driven by observation blocks with minimum human real-time interaction, which implies that VLTI is seen as one machine and not as a set of telescopes and other subsystems by the astronomical instrument. In this paper we describe the as-built architecture of the VLTI control and data flow system, emphasising how new techniques have been incorporated, while at the same time the investments in technology and know-how obtained during the VLT years have been protected. The result has been a faster development cycle, a robustness approaching that of VLT single dish telescopes and a "look and feel" identical to all other ESO observing facilities. We present operation, performance and development cost data to confirm this. Finally we discuss the plans for the coming years, when more and more subsystems will be added in order to explore the full potential of the VLTI. INTRODUCTION After more than ten years of dreams, plans, simulations, design and development the Very Large Telescope Interferometer (VLTI) became reality last year, when interferometric fringes were obtained using stellar beams from two 8 m diameter VLT unit telescopes (UT). Located at the Paranal Observatory in Chile, the purpose of VLTI is the coherent combination of light beams gathered by two or more telescopes 1 . The amplitude and phase of the resulting fringes, together with the known distance between observing telescopes (baseline), give information of the image structure corresponding to an angular resolution of sub milli-arcseconds. In order to maximize the choice of baselines the layout of the Observatory (Figure 1 ) provides four UTs and 30 observing stations for relocatable 1.8 m auxiliary telescopes (AT), giving baselines between 8 and 200 m in different angles. The VLTI near infrared commissioning instrument (VINCI), operating in K band and located in the interferometric laboratory, performs the combination of two beams using fibre optic techniques 2 . Because the delivery of ATs is scheduled only for next year, the photon collection capacity was complemented early last year with two 40 cm test siderostats (SID), which plug into the observing stations and functionally deliver the beams to VLTI as they would have been UTs or ATs, allowing VLTI integration and commissioning to proceed at full speed.
doi:10.1117/12.460963 fatcat:26dg4rlqcrgh7k7zphqckxwfnq