AO4ELT
Homepage » Program » Contributed abstracts » Wave front sensing » Oral talks » Characterization of the T/T Conditions at Gemini using (...)
Characterization of the T/T Conditions at Gemini using Adaptive Optics Telemetry Data

Abstract 140

Submitted by Jean-Pierre VERAN

Authors

J.-P. Veran(1), L. Poyneer(2)

Affiliations

(1)National Research Council Canada - Herzberg Institute of Astrophysics; (2)Lawrence Livermore National Laboratory

Abstract

Achieving diffraction limited imaging on an Extremely Large Telescope will require correcting tip-tilt (T/T) errors to well below the width of the diffraction limited PSF, that is typically to a few milli-arcseconds (mas) rms. This unprecedented level of correction will be difficult to achieve because a) the amplitude of incoming T/T is huge, typically several hundred of mas and b) the non-atmospheric T/T (e.g. telescope shake due to wind, vibrations created by mechanical or electrical sources) may have significant high frequency components that may be difficult to reject.

This paper summarizes our efforts to characterize the T/T conditions at the Gemini Telescopes. This work was carried out in the context of the Gemini Planet Imager (GPI) design. GPI is an Extreme AO System, and even though it will equip an 8-meter telescope, it requires T/T to be corrected down to 5 mas rms (goal: 3 mas rms), which is in fact well in the regime of ELTs. We have analyzed closed-loop telemetry data from two Gemini AO systems: Altair (Gemini North) and NICI (Gemini South). Extracting information on the incoming T/T from these data has proved very challenging. For Altair, the problem is the well-known internal vibrations; for NICI, it is the spatial aliasing on the wave-front sensor that obscures much of the incoming T/T. We discuss these issues, which in fact would apply to any AO system, and describe the strategy we have developed to be able to nevertheless obtain an estimate of the incoming T/T PSD. Based on the NICI data, we find that GPI would be limited by several vibration peaks occurring between 100 and 300 Hz, and that, at a 2 kHz frame rate, the 5 mas rms residual requirement would be barely met with the baseline integrator servo-controller. We conclude by discussing different control strategies that we are envisaging to better correct these high frequency components.


Documents to download