Adaptive optics with solely natural guide stars for an Extremely Large Telescope

In the past decade the ingredients for making real an Extremely Large Telescope with an Adaptive Optics system driven solely by Natural Guide Stars have been conceived, developed, built and proven on the sky. Still, the straightforward merging of these concepts is not enough to fulfill such an ambitious goal. We show here that a combination of the layer-oriented approach, the virtual deformable mirrors concept, and a combined use of different kind of wavefront sensors, some taking advantage of working in Closed Loop and some other characterized by an extremely high dynamic range, make the goal a reachable one. It is remarkable that such an approach requires, on a telescope of ELT class, including a common Deformable Mirror conjugated to the entrance pupil or close-by, a minimum impact on the guide probe units. The last involves the adoption of small Closed Loop AO system with an extremely high dynamic range wavefront sensor looking at the detailed shape of a small Deformable Mirror that allows the use of sensors taking advantage of the Closed Loop conditions. A pyramid wavefront sensor, fed by the Natural Guide Stars light and closing the loop with the mirror, and a YAW wavefront sensor looking at the mirror itself, allow for a natural and efficient combination of the data. The limits in the Field of View covered by such an approach are given by pure meta-pupils superimposition rather than to the spatial frequency of the achievable correction, breaking the limits previously thought for this kind of systems. The overall combination leads to a significant sky coverage, with performances comparable to the ones under discussion for some Laser Guide Stars approaches, without the related hurdle. The small technical impact on the telescope makes this approach not directly in-conflict with a Laser Guide Stars one allowing the designer to keep all the options on the table up to a very late stage.