While optical and radio transient surveys have enjoyed a renaissance over the past decade, the dynamic infrared sky remains virtually unexplored from the ground. The infrared is a powerful tool for probing transient events in dusty regions that have high optical extinction, and for detecting the coolest of stars that are bright only at these wavelengths. The fundamental roadblocks in studying the infrared time-domain have been the overwhelmingly bright sky background (250 times brighter than optical) and the narrow field-of-view of infrared cameras (largest is VISTA at 0.6 sq deg). To address these challenges, Palomar Gattini-IR is currently under construction at Palomar Observatory and we propose a further low risk, economical, and agile instrument to be located at Siding Spring Observatory, as well as further instruments which will be located at the high polar regions to take advantage of the low thermal sky emission, particularly in the 2.5 micron region.
The WISDOM instrument concept was developed at MIT as part of a NASA-NSF funded study to equip the 3.5m WIYN telescope with an extremely precise radial velocity spectrometer. The spectrograph employs an asymmetric white pupil optical design, where the instrument is split into two nearly identical “Short” (380 to 750 nm) and “Long”" (750 to 1300 nm) wavelength channels. The echelle grating and beam sizes are R3.75/125mm and R6/80mm in the short and long channels respectively. Together with the pupil slicer, and octagonal to rectangular fibre coupling, this permits resolving powers over R = 120k with a 1.2” diameter fibre on the sky. A factor of two reduction in the focal length between the main collimator OAP and the transfer collimator ensures a very compact instrument, with a small white pupil footprint, thereby enabling small cross-dispersing and camera elements. A dichroic is used near the white pupil to split each of the long and short channels into two, so that the final spectrograph has 4 channels; namely “Blue,” “Green,” “Red” and “NIR.” Each of these channels has an anamorphic VPH grism for cross-dispersion, and a fully dioptric all-spherical camera objective. The spectral footprints cover 4k×4k and 6k×6k CCDs with 15 µm pixels in the short “Blue” and “Green” wavelength channels, respectively. A 4k×4k CCD with 15 μm pixels is used in the long “Red” channel, with a HgCdTe 1.7 μm cutoff 4k×4k detector with 10um pixels is to be used in the long "NIR" channel. The white pupil relay includes a Mangin mirror very close to the intermediate focus to correct the white pupil relay Petzval curvature before it is swept into a cylinder by the cross-dispersers. This design decision allows each of the dioptric cameras to be fully optimised and tested independently of the rest of the spectrograph. The baseline design for the cameras also ensures that the highest possible (diffraction limited) image quality is achieved across all wavelengths, while also ensuring insensitivity of spot centroid locations to variations in the pupil illumination. This insensitivity is proven to remain even in the presence of reasonable manufacturing and alignment tolerances. Fully ray-traced simulations of the spectral formats are used to demonstrate the optical performance, as well as to provide pre-first-light data that can be used to optimise the data reduction pipeline.
We present a simple seeing-limited IR spectrometer design for the Giant Magellan Telescope, with continuous R = 6000 coverage from 0.87-2.50 microns for a 0:7” slit. The instrument's design is based on an asymmetric white pupil echelle layout, with dichroics splitting the optical train into yJ, H, and K channels after the pupil transfer mirror. A separate low-dispersion mode offers single-object R ~ 850 spectra which also cover the full NIR bandpass in each exposure. Catalog gratings and H2RG detectors are used to minimize cost, and only two cryogenic rotary mechanisms are employed, reducing mechanical complexity. The instrument dewar occupies an envelope of 1:8×1:5×1:2 meters, satisfying mass and volume requirements for GMT with comfortable margin. We estimate the system throughput at ~ 35% including losses from the atmosphere, telescope, and instrument (i.e. all coatings, gratings, and sensors). This optical efficiency is comparable to the FIRE spectrograph on Magellan, and we have specified and designed fast cameras so the GMT instrument will have an almost identical pixel scale as FIRE. On the 6.5 meter Magellan telescopes, FIRE is read-noise limited in the y and J bands, similar to other existing near-IR spectrometers and also to JWST/NIRSPEC. GMT's twelve-fold increase in collecting area will therefore offer gains in signal-to-noise per exposure that exceed those of moderate resolution optical instruments, which are already sky-noise limited on today's telescopes. Such an instrument would allow GMT to pursue key early science programs on the Epoch of Reionization, galaxy formation, transient astronomy, and obscured star formation environments prior to commissioning of its adaptive optics system. This design study demonstrates the feasibility of developing relatively affordable spectrometers at the ELT scale, in response to the pressures of joint funding for these telescopes and their associated instrument suites.
We describe the design of the fiber-optic coupling and light transfer system of the WISDOM (WIYN Spectrograph for DOppler Monitoring) instrument. As a next-generation Precision Radial Velocity (PRV) spectrometer, WISDOM incorporates lessons learned from HARPS about thermal, pressure, and gravity control, but also takes new measures to stabilize the spectrograph illumination, a subject that has been overlooked until recently. While fiber optic links provide more even illumination than a conventional slit, careful engineering of the interface is required to realize their full potential. Conventional round fiber core geometries have been used successfully in conjunction with optical double scramblers, but such systems still retain a memory of the input illumination that is visible in systems seeking sub-m/s PRV precision. Noncircular fibers, along with advanced optical scramblers, and careful optimization of the spectrograph optical system itself are therefore necessary to study Earth-sized planets. For WISDOM, we have developed such a state-of-the-art fiber link concept. Its design is driven primarily by PRV requirements, but it also manages to preserve high overall throughput. Light from the telescope is coupled into a set of six, 32 μm diameter octagonal core fibers, as high resolution is achieved via pupil slicing. The low-OH, step index, fused silica, FBPI-type fibers are custom designed for their numerical aperture that matches the convergence of the feeding beam and thus minimizes focal ratio degradation at the output. Given the demanding environment at the telescope the fiber end tips are mounted in a custom fused silica holder, providing a perfect thermal match. We used a novel process, chemically assisted photo etching, to manufacture this glass fiber holder. A single ball-lens scrambler is inserted into the 25m long fibers. Employing an anti-reflection (AR) coated, high index, cubic-zirconia ball lens the alignment of the scrambler components are straightforward, as the fiber end tips (also AR coated) by design touch the ball lens and thus eliminate spacing tolerances. A clever and simple opto-mechanical design and assembly process assures micron-level self-alignment, yielding a ~87% throughput and a scrambling gain of >20,000. To mitigate modal noise the individual fibers then subsequently combined into a pair of rectangular fibers, providing a much larger modal area thanks to the 34x106 micron diameter. To minimize slit height, and thus better utilize detector area, the octagonal cores are brought very close together in this transition. The two outer fibers are side polished at one side, into a D-shaped cladding, while the central fiber has a dual side polish. These tapered, side-flattening operations are executed with precise alignment to the octagonal core. Thus the cores of the 3 fibers are brought together and aligned within few microns of each other before spliced onto the rectangular fiber. Overall throughput kept high and FRD at bay by careful management of fiber mounting, vacuum feed-through, application of efficient AR coatings, and implementation of thermal breaks that allow for independent expansion of the fibers and the protective tubing.
The Kepler mission highlighted that precision radial velocity (PRV) follow-up is a real bottleneck in supporting transiting exoplanet surveys. The limited availability of PRV instruments, and the desire to break the “1 m/s” precision barrier, prompted the formation of a NASA-NSF collaboration ‘NN-EXPLORE’ to call for proposals designing a new Extreme Precision Doppler Spectrograph (EPDS). By securing a significant fraction of telescope time on the 3.5m WIYN at Kitt Peak, and aiming for unprecedented long-term precision, the EPDS instrument will provide a unique tool for U.S. astronomers in characterizing exoplanet candidates identified by TESS. One of the two funded instrument concept studies is led by the Massachusetts Institute of Technology, in consortium with Lincoln Laboratories, Harvard-Smithsonian Center for Astrophysics and the Carnegie Observatories. This paper describes the instrument concept WISDOM (WIYN Spectrograph for DOppler Monitoring) prepared by this team. WISDOM is a fiber fed, environmentally controlled, high resolution (R=110k), asymmetric white-pupil echelle spectrograph, covering a wide 380-1300nm wavelength region. Its R4 and R6 echelle gratings provide the main dispersion, symmetrically mounted on either side of a vertically aligned, vacuum-enclosed carbon fiber optical bench. Each grating feeds two cameras and thus the resulting wavelength range per camera is narrow enough that the VPHG cross-dispersers and employed anti-reflection coatings are highly efficient. The instrument operates near room temperature, and so thermal background for the near-infrared arm is mitigated by thermal blocking filters and a short (1.7μm) cutoff HgCdTe detector. To achieve high resolution while maintaining small overall instrument size (100/125mm beam diameter), imposed by the limited available space within the observatory building, we chose to slice the telescope pupil 6 ways before coupling light into fibers. An atmospheric dispersion corrector and fast tip-tilt system assures maximal light gathering within the 1.2″ entrance aperture. The six octagonal fibers corresponding to each slice of the pupil employ ball-lens double scramblers to stabilize the near- and far-fields. Three apiece are coupled into each of two rectangular fibers, to mitigate modal nose and present a rectilinear illumination pattern at the spectrograph's slit plane. Wavelength solutions are derived from ThAr lamps and an extremely wide coverage dual-channel laser frequency comb. Data is reduced on the fly for evaluation by a custom pipeline, while daily archives and extended scope data reduction products are stored on NExScI servers, also managing archives and access privileges for GTO and GO programs. Note: individual papers, submitted along this main paper, describe the details of subsystems such as the optical design (Barnes et al., 9908-247), the fiber link design (Fűrész et al., 9908-281), and the pupil slicer (Egan et al., 9912-183).
The WIYN Spectrograph for Doppler Monitoring (WISDOM) was a concept responding to NASA's solicitation for an extreme precision radial velocity instrument for the 3.5 meter WIYN telescope on Kitt Peak in Arizona. In order to meet the spectral resolution requirement of R = 110,000 while maintaining good throughput and a manageable beam diameter, the front end design of the instrument employed a pupil slicing technique wherein a collimated beam is sliced and fed to six separate fibers. This paper presents the optical and mechanical design of the pupil slicer subassembly, a unique method of dealing with thermally induced defocus error, and the methods and results of aligning a prototype.
Originally designed for night-vision equipment, InGaAs detectors are beginning to achieve background-limited performance in broadband imaging from the ground. The lower cost of these detectors can enable multi-band instruments, arrays of small telescopes, and large focal planes that would be uneconomical with high-performance HgCdTe detectors. We developed a camera to operate the FLIR AP1121 sensor using deep thermoelectric cooling and up-the-ramp sampling to minimize noise. We measured a dark current of 163 e- s<sup>-1</sup> pix<sup>-1</sup>, a read noise of 87 e- up-the-ramp, and a well depth of 80k e-. Laboratory photometric testing achieved a stability of 230 ppm hr<sup>-1/2</sup>, which would be required for detecting exoplanet transits. InGaAs detectors are also applicable to other branches of near-infrared time-domain astronomy, ranging from brown dwarf weather to gravitational wave follow-up.
We describe the construction and commissioning of FIRE, a new 0.8-2.5μm echelle spectrometer for the Magellan/
Baade 6.5 meter telescope. FIRE delivers continuous spectra over its full bandpass with nominal spectral
resolution R = 6000. Additionally it offers a longslit mode dispersed by the prisms alone, covering the full z to
K bands at R ~ 350. FIRE was installed at Magellan in March 2010 and is now performing shared-risk science
observations. It is delivering sharp image quality and its throughput is sufficient to allow early observations of
high redshift quasars and faint brown dwarfs. This paper outlines several of the new or unique design choices
we employed in FIRE's construction, as well as early returns from its on-sky performance.
FIRE (the Folded-port InfraRed Echellette) is a prism cross-dispersed infrared spectrometer, designed to deliver singleobject
R=6000 spectra over the 0.8-2.5 micron range, simultaneously. It will be installed at one of the auxiliary
Nasmyth foci of the Magellan 6.5-meter telescopes. FIRE employs a network of ZnSe and Infrasil prisms, coupled with
an R1 reflection grating, to image 21 diffraction orders onto a 2048 × 2048, HAWAII-2RG focal plane array.
Optionally, a user-controlled turret may be rotated to replace the reflection grating with a mirror, resulting in a singleorder,
longslit spectrum with R ~ 1000. A separate, cold infrared sensor will be used for object acquisition and guiding.
Both detectors will be controlled by cryogenically mounted SIDECAR ASICs. The availability of low-noise detectors
motivates our choice of spectral resolution, which was expressly optimized for Magellan by balancing the scientific
demand for increased R with practical limits on exposure times (taking into account statistics on seeing conditions).
This contribution describes that analysis, as well as FIRE's optical and opto-mechanical design, and the design and
implementation of cryogenic mechanisms. Finally, we will discuss our data-flow model, and outline strategies we are
putting in place to facilitate data reduction and analysis.