The Large Lenslet Array Magellan Spectrograph (LLAMAS) is an Integral Field Unit (IFU) spectrograph under construction as a facility instrument for the 6.5-meter Magellan Telescopes. For each pointing, LLAMAS delivers 2400 optical spectra (λ =350-970nm) over a 37”x37” celestial solid angle with a resolution of 2000 through a densely packed microlens+fiber array and replicated low-cost spectrographs. One of our main science goals is to study circumgalactic gas through Lyα emission. To achieve the required signal-to-noise ratio for these observations, LLAMAS must minimize stray light reaching the detector: diffuse scattered light must stay below 0.25% of sky flux and ghost images must not exceed 0.1% of the source signal. We present a non-sequential ray tracing analysis of a simplified LLAMAS model using Photon Engineering’s FRED Optical Engineering Software. We focus on stray light resulting from the volume phase holographic grating and from focal ratio degradation of the fibers. The analysis feeds into a discussion of the design and fabrication of baffles to mask the primary sources of stray light. Additionally, we develop a backup system of mounting rings inside of the cameras where pre-made baffles can be quickly added as needed. Finally, we report on the laboratory performance of a 2-camera LLAMAS prototype featuring the aforementioned stray light interventions.
The Large Lenslet Array Magellan Spectrograph (LLAMAS) is an NSF-funded facility-class Integral Field Unit (IFU) spectrograph under construction for the 6.5-meter Magellan Telescopes. It covers a 37" x 37" solid angle with 2,400 optical fibers efficiently coupled by a double-sided microlens-array, producing R = 2, 000 spectra with 0.7511 spatial resolution. Its broad passband from λ = 350 970nm offers access to line and continuum measurements over a wide range in redshift. Light is multiplexed by the IFU into 8 compact, carbon-fiber bench mounted spectrographs utilizing VPH grisms. We employed several trades on cost-performance ratio while optimizing LLAMAS’ system design including: (a) Splitting the passband between 3 fast all-refractive camera systems with modest entrance pupils, (b) limiting the fibers per unit (i.e. slit length) and building more spectrographs to leverage on production volume, and (c) using a commercial CCD camera built around a common detector (e2v 42-40) and thermoelectric + liquid cooling. To boost blue throughput and achieve high-quality sky subtraction the spectrograph cluster is mounted next to the focal plane on a folded Cassegrain port with gravity-invariant support. This also allows the instrument to deploy quickly, and be fully accessible within 10 minutes on any night, serving as a facility unit for observing astrophysical transients. A sub-sized IFU (169 fibers), mounted in a full-sized front end package with a single spectrograph (2 cameras) was delivered to Magellan in March 2020. We present as-measured laboratory performance from this prototype, though on-sky commissioning was unfortunately cancelled because of the COVID-19 pandemic. This contribution therefore focuses on subsequent design evolution and status of the full facility instrument.
A technique for providing neural network-coached alignment of optical systems is described in detail. The goal is to increase the speed of convergence to an aligned optical system. The neural network model is trained with the wavefront errors (WFEs) of thousands of randomly misaligned instances of the lens system that are modeled in Zemax OpticStudio®. The optical specialist measures the WFE of the misaligned system, then a neural network suggests specific adjustments to be made to the alignment fixturing (which adjuster, which direction, and by what amount). The technique has been developed into a MATLAB®-based tool called rapid optical system alignment with neural network assist that is shown to analytically and experimentally increase the speed of alignment. It is capable of achieving WFEs that are within a few percent of the nominally aligned condition. The ultimate goal is to deploy such a tool to enable an optical specialist to more quickly align challenging optical systems employing freeform or segmented surfaces. Analytical and experimental results for spherically symmetric systems are shown along with an outline of future work.
The design study herein analyzes the design complexity of high zoom ratio lens systems in the visible, SWIR, and LWIR spectrums with four zoom groups (two internally moving). The aforementioned 12.5x zoom lens systems have been designed for use in the Coast Guard for maritime safety, security, and stewardship. To begin our comparative design study, the most advantageous solutions for distinct power groupings were found using a first order solution finder tool. The results showed that solutions with a PNNP, PNPP, and NPNP power grouping with the aperture stop in the third or fourth group had the most potential. At the end of the design process, a comparison was done for the three different wavebands to analyze the relative design complexity. Design complexity metrics were as follows: element count, number of aspheric surfaces, system total track length, element diameter, and tolerance sensitivity.
NON-SPIE: PhD Student
Physics PhD student at The Johns Hopkins University performing research within the field of astronomical instrumentation.