An overview of the instrument can be found here, this page details the design of the spectrograph.
The spectrograph sits within a cooled cryostat, that measures over 4.0 x 2.5 x 2.7 metres and will weigh just over 7 tonnes; which is close to the limit that can be placed on the VLT Nasmyth Platform during normal operations. The two triple-arm spectrographs are mounted on a central vertical optical bench, whose natural symmetry reduces stresses within the system. The cryostat has a central section that surrounds the optical bench and on either side of this are two large doors covering each of the two spectrographs.
The MOONS optical design, shown in high-resolution mode. The blue dotted circles indicate the two prism-disperser-prism combinations that are moved in and out of the beam to switch between high and low resolution modes. The RI and H channels resemble the YJ layout when in low-res mode, with only a single dispersing element in the beam.
Many of the properties of the optical design for the triple arm spectrographs are discussed below, but the following aspects are also of note:
- The collimator is f/3.5 and is marginally oversized to accommodate any misalignments of fibers in the slit.
- MOONS will use 158micron diameter fibers. This was chosen as a trade off between the optimal resolution of the detector and the tolerance to allowed tilt in the FPU.
- Each of the two MOONS spectrographs has 5 dispersers, 3 low-res and 2 high-res. The two triple-arm spectrographs are mounted literally back-to-back on the optical bench, which makes it is possible to switch between the high and low resolution modes in the RI-band using a single common linear mechanism that passes straight through the optical bench. Likewise, the same is done for the H-band.
Each spectrograph has a slit comprised of 32 slitlets, each of which contains 16 fibers – although note that not all the fibres in the slit are connected to FPUs (there are 1024 fibers in the two slits, but 1001 FPUs), which allows some flexibility matching FPUs to slitlets.
It is possible to move the entire slit block of MOONS by 200 μm, which represents half the fiber separation. This makes it possible to move the imaged fibers on the detector so that they fall on the previously un-illuminated inter-fiber regions. The exact way this will be used in the instrument will depend on the performance of the detectors. The baseline is to switch between these two positions for different exposures; this would assist in the removal of any fixed detector artefacts, such as bad pixels.
MOONS requires fast, large optics: all the spectral channels use similar f/0.95 Schmidt cameras. A detailed trade off was previously made to consider the benefits of refractive designs, but it was found that due to the number of surfaces and the transmission of the thick optics that were required, the transmission for the Schmidt design is not much lower, despite its ~22% central obscuration.
One of the critical features of the camera design is that the field lens that lies immediately in front of the detector is glued into a recess within the square hole cut into the first lens in the camera, as can be seen image below. A full prototype of this L1/L2 lens assembly has been made and has past cryogenic testing.
A significant advantage of the MOONS cameras is that they effectively only have two separate elements to align: the L1/L2 assembly and the mirror cell. The detector will sit in the Detector Adjustment Module (DAM) that is mounted on the front of the camera – as can be seen on the ‘front’ of the camera in the image below. The DAM is attached to the cameras by an adjustment ring that has three actuated points, allowing the tip, tilt and piston effects to be corrected for when the final detector is installed in the instrument.
Each of the three cameras in MOONS will use a 4K x 4K array that is mounted behind the field corrector lens. The two infra-red channels will exploit the new Hawaii 4RGs from Teledyne. MOONS will use the 2.5 μm cut-off, which is technically longer than is required but this maximises the QE over the MOONS wavelength range, although it does require the optical bench to be cooled to a lower temperature.
In the RI channel MOONS will use fully-depleted detectors that have been developed by Lawrence Berkeley National Laboratories (LBNL). These devices give considerably higher QE beyond 0.9 μm than any other currently available 4K x 4K device. The LBNL detector development work was supported in part by the Director, Office of Science, of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231.