Performance of MEMS-based visible-light adaptive optics at Lick Observatory: Closed- and open-loop control

TL;DR

This study evaluates the on-sky performance of MEMS-based visible-light adaptive optics at Lick Observatory, demonstrating that open- and closed-loop controls achieve similar image quality and that MEMS technology is robust in an observatory environment.

Contribution

First on-sky comparison showing open- and closed-loop MEMS AO control are equally effective for visible-light correction.

Findings

Open- and closed-loop control yield similar Strehl ratios (~7%) in I-band.

MEMS deformable mirrors operate reliably over three years in observatory conditions.

Error budget analysis identifies internal static error as the main contributor to wavefront error.

Abstract

At the University of California's Lick Observatory, we have implemented an on-sky testbed for next-generation adaptive optics (AO) technologies. The Visible-Light Laser Guidestar Experiments instrument (ViLLaGEs) includes visible-light AO, a micro-electro-mechanical-systems (MEMS) deformable mirror, and open-loop control of said MEMS on the 1-meter Nickel telescope at Mt. Hamilton. In this paper we evaluate the performance of ViLLaGEs in open- and closed-loop control, finding that both control methods give equivalent Strehl ratios of up to ~ 7% in I-band and similar rejection of temporal power. Therefore, we find that open-loop control of MEMS on-sky is as effective as closed-loop control. Furthermore, after operating the system for three years, we find MEMS technology to function well in the observatory environment. We construct an error budget for the system, accounting for 130 nm of wavefront error out of 190 nm error in the science-camera PSFs. We find that the dominant known term is internal static error, and that the known contributions to the error budget from open-loop control (MEMS model, position repeatability, hysteresis, and WFS linearity) are negligible.