Adaptive Optics system of the Evanescent Wave Coronagraph (EvWaCo): optimised phase plate and DM characterisation

Abstract

The Evanescent Wave Coronagraph (EvWaCo) is an achromatic coronagraph mask with adjustable size over the spectral domain [600 nm, 900 nm] that will be installed at the Thai National Observatory. We present in this work the development of a bench to characterise its Extreme Adaptive Optics system (XAO) comprising a DM192 ALPAO deformable mirror (DM) and a 15x15 Shack-Hartmann wavefront sensor (SH-WFS). In this bench, the turbulence is simulated using a rotating phase plate in a pupil plane. In general, such components are designed using a randomly generated phase screen. Such single realisation does not necessarily provide the wanted structure function. We present a solution to design the printed pattern to ensure that the beam sees a strict and controlled Kolmogorov statistics with the correct 2D structure function. This is essential to control the experimental conditions in order to compare the bench results with the numerical simulations and predictions.This bench is further used to deeply characterise the full 27 mm pupil of the ALPAO DM using a 54 × 54 ALPAO SH-WFS. We measure the average shape of its inuence functions as well as the inuence function of each single actuator to study their dispersion. We study the linearity of the actuator amplitude with the command as well as the linearity of the inuence function prole. We also study the actuator osets as well as the membrane shape at 0-command. This knowledge is critical to get a forward model of the DM for the XAO control loop.

Inverse problem approach in Extreme Adaptive Optics: analytical model of the fitting error and lowering of the aliasing

“…where d ∈ R n dat is the vector of the n dat ∈ N data (here the wavefront slopes in each sub-aperture), S sy ∈ R n dat ×nw is the WFS synthetic model matrix (here the average phase derivative on each sub-aperture 11 ), w ∈ R nw is the vector of the wavefront described on n w ∈ N spatial knots (it can be innite for a continuous description of the wavefront), and n is the noise on the measurements. Let us also introduce M ∈ R nw×nact , the matrix of the mirror inuence function 5 of the n act ∈ N actuators of the DM, and G sy = S sy • M ∈ R n dat ×nact , the synthetic interaction matrix of the AO system that links the commands c ∈ R nact applied on the DM to their equivalent synthetic slopes.…”

“…The bench was also used to measure the inuence functions of the ALPAO 192DM. 5 These measurements are used to generate the mirror matrix M.…”

“…As a coronagraph dedicated to high contrast and high resolution imaging, EvWaCo implies the development of a dedicated extreme adaptive optics (AO) system. The description of its AO system and its associated characterisation bench aligned in the NARIT cleanroom is the subject of another paper by Berdeu et al 5 We remind here only its main features. This paper focuses on the two rst errors via an end-to-end model (E2E) that was developed along with the bench design and alignment.…”

“…This paper focuses on the two rst errors via an end-to-end model (E2E) that was developed along with the bench design and alignment. 5 The main objective of this tool is to have a numerical model as close as possible from the real bench to develop and test new AO algorithms or predict on-sky raw contrast. 12 Studying the discrepancies between the numerical predictions and the experimental data can be of great help to improve the models and develop more robust algorithms.…”

“…Thus, our E2E model integrates the real inuence function of the DM measured on the bench as well as the turbulence injected in the bench via a custom-designed phase plate. 5,13 First, in Section 2, we present a new analytical model of the tting error. Up-to-date, the only way to study the impact of the shape of the DM inuence function in the tting error is to perform extensive Monte Carlo simulations.…”