The making of robust and highly performing imaging spectropolarimeters for large solar telescopes

We discuss the requirements, concepts, simulations, implementation, and calibration of two dual Fabry-Perot (FPI)-based imaging spectropolarimeters, CRISP and CHROMIS, at the Swedish 1-m Solar Telescope, and CRISP2, which is under construction. These instruments are optimised for a large field of view and high cadence at the expense of a moderate spectral resolution, and use a combination of a high-resolution and a low-resolution etalon together with an order-sorting prefilter to define the bandpass. The overall design is made robust and stable by tailoring the low-resolution etalon reflectivity to accommodate expected cavity errors from both etalons, by using a compact optical design that eliminates the need for folding mirrors, and enclosing the entire system within a single container sealed by lenses. By using a telecentric design based on lenses rather than mirrors, image degradation by the FPI system is negligible, as shown in a previous publication, and the throughput of the system is maximised. The initial alignment, and maintaining that alignment over time, is greatly simplified. Moreover, the telecentric design allows the full calibration and/or modelling of essential system parameters to be carried out without interfering with the optical set-up or the cameras. We also discuss briefly the polarimeters developed for CRISP and CHROMIS. The high performance of CRISP and CHROMIS has been demonstrated in an earlier publication through measurements of the granulation contrast and comparisons with similar measurements simultaneously made through broadband continuum filters. Here, we focus on the aspects of the design that are central to enabling high performance and robustness, but also discuss the calibration and processing of the data, and use a few examples of processed data to demonstrate the achievable image and data quality. We put forward a proposal for a similar conceptual design for the European Solar Telescope and conclude by discussing potential problems of the proposed approach to designs of this type. Some aspects of these FPI systems may also be of interest outside the solar community.