Current instrument sand future plans for new instrumentation for the Keck Observatory are reviewed. As is widely known, both 10-meter telescopes are in full operation and each has a significant complement of scientific instruments. Keck II has a working adaptive optics system and a similar module is under construction for Keck I in preparation for combining the beams in the Keck Interferometer. Recently, Keck II received two new instruments, ESI - which is a CCD-based faint-object spectrograph, and NIRSPEC, a cryogenic high- resolution near-IR spectrograph. The existing LRIS spectrograph is due to an upgrade to provide a blue- sensitive channel to complement the current red channel. DEIMOS, a multi-object spectrograph and NIRC2, a diffraction-limited IR camera are nearing completion. Future plans include detector upgrades, instruments for integral field and wide-field spectroscopy, interferometry and laser guide stars.

The paper deals with the global approach in place at ESO to optimize observing efficiency at the VLT. It involves not only the telescopes and their associated instrumentation, but as a mater of fact the whole observing process from proposals to data distribution. The ten main avenues pursued to reach that goal are presented, and early experience at Paranal Observatory reported. The need for continuous renewal of instrument complements at large telescope facilities is stressed.

The Subaru telescope had its astronomical first light at the Cassegrain focus in January 1999. Fine tuning of the telescope and commissioning of the seven open-use instruments at four foci of the Subaru telescope are in progress. An overview of the updated status of the observational instruments is presented with reference to papers addressed during this series of conferences. General status of the Subaru telescope, the plan for the open use of the instruments, and the scope for the second generation instruments under investigation are also reported.

Building instruments suitable for the new 8-10 m class of telescopes has been a major challenge, as specifications tighten, costs, scientific demands, and expectations grow, all while schedules remain demanding. This report provides a top level description of the status of various elements in the Gemini instrument program, and touches on some of the common problems the various teams building Gemini instruments are having. Despite these challenges, Gemini anticipates harvesting great scientific rewards from the combination of its Observatory facilities and exciting complement of scientific instruments.

We are building a Multi-Object Double Spectrograph for the Large Binocular Telescope. The main themes of our planned research with the instrument are the formation and evolution of galaxies and their nuclei and the evolution of large- scale structure in the universe, although we expect that the spectrograph will be used for many other varieties of programs as well. The science goals for the instrument dictate that it have the highest possible throughput form 320 to 1000 nm, spectral resolutions of 10³ to 10⁴, and multi-object capability over an approximately 6 foot field. Our design is highly modular, so future upgrades should be straightforward.

The Hobby-Eberly Telescope (HET) is a revolutionary large telescope of 9.2 meter aperture, located in West Texas at McDonald Observatory. HET has been constructed by a consortium of US and German universities for about 16 million dollars. First light was obtained on December 11, 1996, HET was dedicate don October 8, 1997, and early scientific operations started on October 8, 1999. The HET operates with a fixed segmented primary and has a tracker which moves the four-mirror corrector and prime focus instrument package to track the sidereal and non-sidereal motions of objects. Here we describe the current performance of the HET relative to its goals.

The presentation summarizes the status of the first generation of instruments for the 10-m Spanish Telescope, GTC. The GTC is planned to have its first light at the end of 2002, and to be operating at the end of 2003. The management plan at the Project Office, PO, for the procurement of the instruments is presented.

We review the characteristics of SALT and discuss the major science drivers, both of which will decide the choice of a fist-light instrument package. Current science drivers call for instrumentation similar to that of the Hobby-Eberly Telescope, on which SLAT is closely modeled, with fiber-fed and direct spectroscopy in the visible/near-IR being the major observational mode delivering the best astrophysical returns. The suite will include multi-object spectroscopic capability over a continuous wavelength range of at least 400 nm to 1300 nm, with possible extended capabilities to the blue and near-IR. Factors which will decide these limits are discussed, which include mirror coatings and fiber transmission. Other specific science drivers, which may influence instrument choice or design, include photometry, imaging spectroscopy and spectropolarimetry. Possibility exist for different or novel approaches to instrument designs. Together with potential enhancements of the telescope design, these will have some impact on instrument choice. The nature of the first light instrument suite is currently under discussion and will probably be decided by late-2000.

Five SOAR instruments are being designed for high-resolution imaging/imaging spectrophotometry across the isokinetic field and queued/remote observing. Wavefront tilt will be sensed in instruments and corrected by jittering M3. SDSU-2 'Leach' controllers under LabVIEW will operate most detector arrays. All optical instruments emphasize high UV throughput, and will use pairs of UV-enhanced MIT/LL 2 by 4 K CCDs. The University of Sao Paulo may provide a 1500- element integral field lenslet array that is fiber-coupled to a compact spectrography. VPH gratings will be used in this instrument and in the multi-slit spectrometer from the University of North Carolina at Chapel Hill. CTIO will provide an optical mosaic imager with 'trombone'-style ADC, and may also upgrade their IRS with its large complement of existing gratings to a 1 by 1K Hawaii array. Michigan State University may build a 4 by 4K near-IR imager with tunable Lyot filter. Most instruments will be clustered at Nasmyth ports where payload totals 4400 kg. Facility units will allow calibration while another instrument is doing science. The baseline for mechanism control is LabVIEW under Linux in a Compact PCI chassis that is fiber-linked by MXI-3 to a dedicated PC.

FORS is an all dioptric focal reducer designed for direct imaging, low-dispersion multi-object spectroscopy, imaging polarimetry and spectropolarimetry of faint objects. Two almost identical copies of the instrument were built by a consortium of three astronomical institutes under contract and in cooperation with ESO. FORS1 was installed in September 1998 and FORS2 in October 1999 at the Cassegrain foci of the ESO VLT unit telescope nos. 1 and 2. FORS1 is in regular operation since April 1999. Regular observation with FORS2 are scheduled to begin in April 2000.

Faint object camera and spectrograph, FOCAS, is a Cassegrain optical instrument of Subaru telescope. It has a capability of 6 arcmin FOV direct imaging, low resolution spectroscopy, multi-slit spectroscopy as well as polarimetry. Only the imaging mode has been available so far. The overall design, the observing functions, and the preliminary performance verifications of FOCAS will be presented.

The Gemini Multiobject Optical Spectrographs were designed to exploit the exceptional image quality anticipated form both the active and adaptive optics systems. High mechanical stability and repeatability and efficient reconfiguration and calibration were emphasized in the design, as well as the usual requirements of obtaining excellent image quality, high optical throughput and low optical distortion. In addition, an active flexure compensation system is used to assist in achieving a primary goal of attaining velocity accuracies of 2 km/s per spectrum in multiobject mode at the highest spectral resolution. Although the field is modest, small pixels are used to fully sample images as small as 0.2 inch and 28.3 million pixels will be recorded by the detector mosaic which consists of three 2048 * 4608 EEV CCDs. In this paper, results from extensive tests made during integration and testing of GMOS N components demonstrate that the design requirements are being met.

LAMOST is a special Schmidt telescope of 4 meters aperture. It will be located in Xinglong station at Beijing Astronomical Observatory, China. The start of science operation is expected in 2004. There are 4000 optical fibers on the telescope focal surface that will feed 16 low resolution spectrographs (LRS) and one or more medium resolution spectrograph and one high resolution spectrograph. Here we present a description of the LRS specification and its optical design. LRS is a multi object fiber spectrograph that is optimized for galaxy red shift surveys. The output bema with f-ratio of 4 from fiber are collimated by a spherical mirror and then are split to red and blue band by a dichroic filter. We use reflective grating as dispersion elements. Spectra are focused onto 2048 by 2048 CCD using fast Schmidt camera with f-ratio of 1. Spectral coverage per exposure is from 370nm to 900nm. The beam size is 200mm resulting resolution 1000 with 3.3 arcsecond fiber slit. The slit with 130mm length picks up 250 fibers, so we need 16 low-resolution spectrograph to observe spectra of 4000 celestial objects in single exposure.

The HROS is a Cassegrain focus instrument for use on the Gemini South telescope at Cero Pachon, Chile. It is of novel design, using exceptionally large optical components, and subject to the normal high flexural and stability requirements of optical wavelength instrumentation. To meet these requirements while subjecting HROS to the infinite number of gravity vectors found at the Cassegrain focus has resulted in a very difficult design and analysis of the support structure. This paper describes the mechanical design approach to meting these requirements and presents the flexural performance predictions for the structure as given by Finite Element Analysis.

HROS will be the key instrument for high resolution spectroscopy for UV to near-IR wavelengths at eh Gemini South telescope. The instrument is unique in providing a resolving power of R equals 50,000 at the Cassegrain focus of an 8-meter telescope. Taking advantage of this location, the spectrograph is optimized for high throughput, particularly for the UV region, and high efficiency. Here we present the final opto-mechanical design of the spectrograph, together with its predicted performance. In particular, we show how our design delivers an overall peak throughput of almost 30 percent and coverage of wavelengths between 325 and 885 nm in a single CCD exposure. We also discus the development of the design form the science requirements and some of the mechanical issues that drove it to this solution. Finally we report on the current status of optics procurement and testing.

We present a preliminary design study for an adaptive optics visual echelle spectrograph and imager/coronograph for use as parallel instrument of the Nasmyth Adaptive Optics System (NAOS) on unit UT3 of the VLT. The spectrograph is intended for intermediate resolution spectroscopy of faint sources. It could be used for observations of late-type dwarfs in distant Galactic clusters and in galaxies of the local group as well as for spectroscopy of extra galactic objects like quasars and Lyman break galaxies down to a limiting magnitude of V equals 22.5. The implementation of an imaging gand coronograph mode increases the versatility of the instrument and its scientific objectives. The instrument takes advantage of Adaptive Optics at visible wavelengths both for imaging and spectroscopy. With NAOS at the VLT, the light concentration in these bands will be above approximately 60 percent of the flux in a 0.3 arcsec aperture for typical Paranal conditions. Simulations show that a gain of more than one magnitude with respect to compatible non-adaptive optical spectrography will be possible for sky- and/or detector limited observations. In addition, the smaller diffraction limit in the optical than in the IR will allow a significant gain in imaging and coronography as well. Finally, the instrument will allow gathering unprecedented experience on the performances of AO at visible wavelengths, which will be fundamental for further development of AO systems, in particular for very large telescopes.

The FORS instruments are focal reducers and spectrographs which are built in two copies for the unit telescopes UT1 and UT2 of the ESO/VLT by a consortium of University Observatories. An overview of the instrument capabilities is given in a separate paper at this conference.

Motivated by a strong concern to keep maintenance work as low as possible the direction of the CFHT had for some times contemplated the possibility to replace the original mirror train f/20 focus feeding their Gecko High Resolution Coude Spectrograph by a more convenient fiber link coupled to the f/8 Cassegrain focus. A decision supporting that idea was ultimately taken two years ago and our group at the OPM was contacted to build such a system according to precise specifications. This telescope facility, baptized CAFÉ for Cassegrain Fiber Environment, has now arrived to near completion and we are able to present here its main characteristics and the technical solutions that were adopted to meet the CFHT requirements and to provide the system with the best performances in terms of robustness and efficiency.

We are proposing a new spectrograph (ATLAS) which would revolutionize intermediate-dispersion observations at the AAT. Based on the new technology of volume phase holographic gratings, and using transmission optics, ATLAS offers high throughput and a wide field. It will be ideally situated to extensive surveys of faint objects. It has been designed with a collimated beam diameter of 150 mm, giving resolution (lambda) /(delta) (lambda) up to nearly 10,000 with a 1.5 arcsecond slit and good efficiency. It will be a dual-beam instrument, to maximize observing speed and allow optimized optical coatings to be used. The project is working towards its concept design review which will occur during 2000.

The fiber link between adaptive optics and oasis will provide at the Canada-France-Hawaii Telescope a unique scientific capability to obtain 2D visible and near-IR spectra of objects in the I, J, H bands with a 200 to 3000 spectral resolution and with a 0.15 arcsec and 0.4 spatial resolution. The specific is to use microlens arrays and fibers to link the Adaptive Optics Bonnette with the near-IR spectrograph OSIS.

This paper describes a fiber fed bench mounted echelle spectrograph in white pupil arrangement for the 1.3m telescope on Mt. Skinakas. A telescope adapter with focal reducer, Zeeman analyzer and on-fiber guider connects the telescope over a set of fibers lamps is fiber connected with the telescope adapter as well. The echelle spectrograph with a resolution up to 30.000 will be used for the determination of radial velocities and the longitudinal magnetic fields of early type stars.

The low resolution spectrograph (LRS) is the first facility instrument on the 9.2m Hobby-Eberly Telescope (HET). The LRS has three operational modes: imaging, long-slit spectroscopy and multi-object spectroscopy (MOS). We present the design and early operations performance of the LRS MOS unit, which provides 13 slitless, each 1.3 arcsec by 15 arcsec, on 19.6 arcsec centers, within the 4 arcmin field of view of the HET. This type of remotely configurable unit was chosen over the more conventional slit masks due to the queue scheduling of the HET, and the instrument's remote location at the prime focus of the telescope. A restricted envelope around the HET focus at the LRS port forced a very compact design. The MOS unit has miniature mechanisms base don custom cross- roller stages and 0.25 mm pitch lead-screws. Geared stepper motors with 10 mm diameters drive the 13 axes at 0.8 micron per step. The precision of the mechanism is far greater than required by the HET plate scale of 205 microns per arcsec, but result in a robust unit. The slitlets were fabricated at the University of Texas by shadow-masking the slit area with a wire and vacuum depositing aluminum onto the silica substrates. Both sides are then coated with MgF₂ which serves as an antireflection coating and a protective layer. Web-based software is available for optimizing the orientation of the MOS unit and the placement of slitlets on objects in the field. These setup scan be down loaded to the unit for configuration outside of the beam while the HET is slewing to its next target in the queue, or while the LRS is used in imaging mode for setup on faint objects. The preliminary results presented here are from one commissioning run with the MOS, where the unit appears to be meeting performance specifications.

We have proposed a new Nasmyth instrument for the William Herschel Telescope which exploits the potential of wide field-of-view correction of atmospheric turbulence to produce a versatile, high spatial resolution, high efficiency, multi-object spectrograph and imager optimized for the 0.7-1.6 micron region. Using a low-altitude Rayleigh beacon guide star to correct the boundary layer turbulence which dominates the atmospheric seeing at La Palma on more than 25 percent of nights, MOSAIC combines the angular resolution gains of adaptive optics with the observing efficiency gains of multiple-object spectroscopy. Additional operating modes could include a narrow-band tunable filter and a fiber feed to a bench-mou8nted high resolution echelle spectrography. The instruments would provide a unique capability on 4-meter telescopes, opening up a wide variety of new scientific capabilities ranging from spectroscopic studies of crowded star fields to resolved studies of the kinematics of distant galaxies.

The Optical System for Imaging and low Resolution Integrated Spectroscopy (OSIRIS) is being designed as a Day-One optical instrument for the 10.4 mts Gran Telescopio CANARIAS (GTC). It will be the first instrument, on such a large telescope, belonging to a new class of tunable spectrographs, implementing last advances in Volume Phase Holographic Gratings and tunable imaging combined with charge shuffling capabilities, covering the optical wavelength range. OSIRIS< to be first mounted dat one of GTC's Nasmyth platforms, is designed to be compact enough to fit at the Cassegrain focus as well. The optical design is devised around the classical concept of collimator plus camera. The collimator is an off axis ellipsoidal mirror, while the f/2.475 camera consists of several groups of all spherical surfaces lenses, forming a unit together with the detector rand cryocooler. A folder mirror prevents interference with the GTC acquisition and guiding subsystem. Several combinations of color and interference filters. TFs and VPHs are available in the collimated beam, near the pupil, to provide the wide versatility of required observing modes and resolutions. Short descriptions of the OSIRIS geometry, specifications, design strategy and the optical design are presented.

We present results of laboratory test of the high resolution spectrograph, that will be soon in operation at TNG telescope, La Palma. These first result shows that the instruments performs according to specifications, providing the expected very high resolution; and that can be operated remotely according to the TNG standards.

The VIRMOS consortium of French and Italian institutes is building the Visible Multi-Object Spectrograph (VIMOS) and the NIRMOS for the European Southern Observatory Very Large Telescope. Inside the VIRMOS consortium the Technology Working Group of the Osservatorio Astronomico di Capodimonte is in charge of the design, realization and integration of mechanics, opto-mechanical interfaces, control electronics and low level control software. This paper is an overview on the VIMOS technical project managed by the TWG.

In this paper we discuss some of the recent developments with optical fibers at the Anglo-Australian Observatory. Firstly we will describe the upgrade to the SPIRAL integral field spectrograph for the Anglo-Australian Telescope. SPIRAL-B uses a crossed cylindrical microlens array to feed 512 optical fibers at F/5.5 providing a field of view of 22 by 11 arcseconds with 0.7 arcsecond spatial sampling. The performance of the fiber bundle, microlens array, and construction techniques will be described. We will also discus the development of prototype optical fiber switchyard as part of the AUSTRALIS concept study. The switchyard provides an 'optical bread' in the fiber, between the telescope and spectrograph, which allows coupling between fibers of different diameters and focal rations. A dichroic can also be incorporated into the switchyard to allow both optical and IR spectrographs to be fed simultaneously. Switchyards therefore provide much greater observing flexibility by increasing the number of possible instrument configurations. We will briefly discuss the merits of fiber switchyards and present the results of FRD and transmission test performed in the laboratory.

PMAS has been designed and is currently being integrated as a traveling instrument of the Astrophysical Institute Potsdam. It is a UV-visual integral field spectrograph, with optimized efficiency and stability for use as a 3D spectrophotometer. PMAS is prototyped for first light at the Calar Alto 3.5m telescope with an option to go to other telescopes. We present the final design layout, details of the mechanics, optics, detector systems, and instrument control. We report on the current status of the integration.

We present SPIFFI, the integral field spectrometer for the VLT. This instrument allows simultaneous observation of IR spectra in more than 1000 image points of a 2D field. With its set of four gratings and a pixel scale that can be varied by a factor of ten, SPIFFI provides high flexibility, and at the same time offers the unique possibility of diffraction limited imaging spectroscopy at an 8m-class telescope, when fed by the adaptive optics system MACAO. We outline the scientific drivers for building such an instrument, the concept of image slicing, the optical design, and the implementation of SPIFFI.

Photon-counting silicon avalanche photo-diodes (APDs) offer very high quantum efficiency, and might eventually replace photocathode detectors in high-speed photometry of astronomical objects. Laboratory studies have been performed on both passively and actively quenched APDs. Peculiarities of APDs include that the dark signal may exhibit bistability, with the count rate jumping between discrete levels. Following any photon detection, the detector itself emits some light, which might be confusing under certain conditions. Deadtimes and after pulsing properties appear favorable, but the small physical size of APDs causes challenges in optically matching them to the entrance pupils of large telescopes.

The photon address digital detector system (PADDS) in its current implementation represents a cost effective and flexible solution for photon imaging. The system is built of an image intensifiers, a position sensitive photomultiplier tube with crossed wire anodes and a digital signal processor (DSP). TO avoid the complexities involved in the development of special devices. All components used are available from stock so as to achieve reliable specifications and low cost. The evaluation of the photon events by a DSP shifts the problem of detection from hardware to software and enables an exchange of the image intensifier to match the wavelength of interest. Design considerations and the status of PADDS are presented, results of an observation of the crab pulsar are shown and future developments are discussed.

For the photon address digital detector system (PADDS), the determination of position and time of photon events depend on the algorithm used for detection. The introduction of a digital signal processor shifts the problem of evaluation from hardware to software. Different algorithms can be tested in one setup. For a point source, the algorithms can be compared by the statistical properties of the resulting distributions. The data flow in PADDS is presented and different algorithms are discussed.

We report on the design and testing of a new readout scheme for Superconducting Tunnel Junction (STJ) arrays. By grouping the electrodes in rows and columns, this method drastically reduces the number of connections and electronic circuits required for reading out a large format array of pixelated detectors. It is a generic scheme in that it could be applied to different kinds of detector arrays. Using charge sensitive amplifiers with junction field-effect transistors we verify that the energy resolution degrades primarily due to capacitance increase at the amplifier's input node. However, since each detector is read-out by two independent circuits, these two outputs can be combined to increase the signal-to-noise level. The measurements reported here were carried out on an array of 6 by 6 junctions. All junctions were biased but only 2 rows and 2 columns read-out. We compare the result to measurement carried out on a similar 6 by 6 array fabricated from the same trilayer but with individual pixel read-out. The measurements show that stable biasing of STJs is possible with the new configuration and that the measured optical spectral line resolutions are consistent with our theoretical predictions.

A family of controllers has bene developed that is powerful and flexible enough to operate a wide range of CCD and IR focal plane arrays in a variety of ground-based applications. These include fast readout of small CCD and IR arrays for adaptive optics applications, slow readout of large CCD and IR mosaics, and single CCD and IR array operation at low background/low noise regimes as well as high background/high speed regimes. The CCD and IR controllers have a common digital core based on user- programmable digital signal processors that are used to generate the array clocking and signal processing signals customized for each application. A fiber optic link passes image data and commands to VME or PCI interface boards resident in a host computer to the controller. CCD signal processing is done with a dual slope integrator operating at speeds of up to one Megapixel per second per channel. Signal processing of IR arrays is done either with a dual channel video processor or a four channel video processor that has built-in image memory and a coadder to 32-bit precision for operating high background arrays. Recent developments underway include the implementation of a fast fiber optic data link operating at a speed of 12.5 Megapixels per second for fast image transfer from the controller to the host computer, and supporting image acquisition software and device drivers for the PCI interface board for the Sun Solaris, Linux and Windows 2000 operating systems.

6dF is a fully-automated front-end for the multi-object fiber-coupled spectroscopy system of the UK Schmidt Telescope (UKST). It is a magnetic pick-place system with an positioning robot, and it will replace the manually- configured FLAIR front-end that has been operated on the telescope in its present form since 1992. The one-hour reconfiguration time of 6dF makes a range of projects possible that could not be contemplated with FLAIR. Most notable of these is the 6dF Galaxy Survey, a two-part survey of redshifts and peculiar velocities that will become the main task of the Schmidt-Telescope once 6dF is commissioned late in 2000. This will eventually lead to a determination of the large-scale distribution of mass density in the local Universe, which will be a significant achievement for a telescope in the 1-m class. The instrument will be capable of carrying out other survey projects, and 6dF spectroscopy will be the UKSTs main science product for at least the next five years. A notable aspect of 6dF has been the very rapid progress made in its construction by the Anglo-Australian Observatory's Instrumentation Division.

FLAMES is a fiber facility to be installed on the A platform of the VLT Kueyen telescope, which can feed up to three spectrographs with fibers positioned over a corrected 25 arcminutes field of view. The initial configuration will include connections to the GIRAFFE and to the red arm of the UVES spectrographs, the latter, located on the Nasmyth B platform of the same telescope, is already in operation as a long slit stand alone instrument. The 8 fibers to UVES will give R approximately 45000 and a large spectral coverage, while GIRAFFE will be fed by 132 single fibers, or by 15 deployable integral field units or by one central large integral unit. GIRAFFE will be equipped with two gratings, giving R equals 5000-9000 and R equals 15000-25000 respectively. It will be possible to obtain GIRAFFE and UVES observations simultaneously. Special attention is paid to optimizing night operations and to providing appropriate data reduction. The instrument is rather complex and it is now in the construction phase; in addition to ESO, its realization has required the collaboration of several institutes grouped in 4 consortia.

This paper presents the design, construction and performance parameters for the optical detector systems for UVES, the echelle spectrograph for the UT2 Kueyen Telescope at the ESO Paranal. A general overview at system level with a description of the individual components is also given. Emphasis is given on the CCD detectors and their performance. The two rams are equipped with large science CCD detectors, one single chip in the blue arm and a mosaic of two in the red. The blue CCD is a 2K by 4K, 15 micrometers pixel size thinned EEV CCD-44. The mosaic in the red arm is made of an EEV chip of the same type and an MIT/LL CCID-20 chip, which features a higher NIR QE and reduce fringing for the redder part of the spectral range. The read noise archived with the CCDs at the telescope is less than 2 electrons for the EEV CCDs and less than 3 electrons for the MIT/LL CCD at a readout speed of 50 kpixel/sec/port. The system s offer a variety of readout modes, binning options and readout speeds up to 625 kilopixel per second per port with linearity variations better than 1 percent peak to peak at a dark current level of around 0.5 electrons/pixel/hour.

The 40 CCDs produced at Marconi Applied Technologies are dispatched at CEA to be tested and characterized before being inserted into the MegaCam camera mosaic. We describe here briefly our CCD test bench, the measurements we perform with it and the results obtained so far. A model has also been developed to interpret the quantum efficiency results.

Fully depleted silicon pn-CCDs with an active thickness of 300 micrometers exhibit a quantum efficiency of 90 percent at a wavelength of 1 micrometers in the near IR. The multi-parallel readout architecture allows for a frame time shorter than 2 ms for a device having a format of 256 by 256 pixel. It can be operated in a full frame mode and in a frame store mode. The pixel size is 50 micrometers by 50 micrometers . The active area is then 12.8 by 12.8 mm². Cooled down to -90 degrees C the electronic noise floor is below 5 electrons at 50 Megapixel per second. Quantum efficiency measurements will be shown as well as the physical models of the radiation entrance window. A camera system with comparable specifications - except for the pixel size - was already fabricated for ESA's XMM-NEWTON satellite mission, which was launched in early December 1999. The active size of the detector is 60 by 60 mm² with a readout time of 4 ms. Future work includes the extension of the active area to 1000 by 1000 pixel, monolithically fabricated on a high resistivity 6 inch silicon wafer. The main driver for this development is ESA's planned XEUS mission, to be launched at the end of the next decade.

A synergy of digital imaging technologies is taking place that is expanding CCD applications and markets in the high speed arena. Discussions in this paper are directed toward requirements for a new class of High Speed Scientific CCDs. This paper will discuss the theoretical limits in readout rate for maintaining high performance in the areas of charge transfer efficiency and dynamic range. In particular, a high sped problem called 'substrate bounce' is discussed which represent an important speed limitation for the CCD, especially for backside illuminated sensors. Theoretical results are compared to experimental data taken from developmental HSS CCDs recently fabricated.

In this paper we present the new optical camera ROSI mounted at the 60/90/180 Schmidt telescope of the Campo Imperatore Station. We have developed a new LN₂ compact cryostat designed to be mounted directly at the internal focus of the telescope and optimized to obtain a very long duration of the cryogenic liquid. The instrument is based on a 2K by 2K thinned EEV cooled down to 180K and despite of the reduced capacity of the vessel the overall holding time of LN₂ is greater than 10 hours, providing a long working cycle. The CCD is controlled by a modified version of the Astrocam DUO provided by LSR that offers both a high readout speed and a low noise. ROSI has been equipped with the same high transmission filter set use din SUSI2 provided by CETEV. The computer design of the entire instrument allows a negligible obscure of the light path, comparable to the traditional one of the Schmidt telescopes equipped with photographic plates.

The Subaru telescope has an excellent performance of wide field of view at the prime focus. A big area of 30 feet times 24 feet is observable at a time with the prime focus camera. Making the best use of the wide view, we are constructing narrowband (NB) filter system consisting of 20 bands. This system covers the wavelengths between 4,000 angstrom and 10,000 angstrom. The band width (BW) varies form 200 angstrom to 400 angstrom depending on the center wavelength (CW). The resolving power of the system is 23. Each filter has a big dimension of 205mm times 170mm and excellent uniformities on CW, BW and peak transmittance. Employing this filter system, spectroscopy for all objects recorded in fields of view is possible at the wavelength resolution of R23. The limiting magnitude would reach 27AB in reasonable observation time even at long wavelength bands. Such deep NB imaging spectroscopic survey should provide huge catalogue on cosmological objects. Especially, photometric redshift analyses with higher spectral resolution of R23 than ordinary broadband system of R approximately equals 4, will revolutionarily develop studies on formation and evolution of galaxies together with search for large scale structures at high redshift, based on enormous statistics, for example, 10⁴ or more galaxies at high redshift of z > 3. Also, a lot of objects having strong emission lines as QSO/AGNs and Ly(alpha) or more galaxies will be discovered, because NB filter is strong in detection of emission line. The use of NB filter is strong in detection of emission line. The use of NB filter system in survey observations is surely quite conservative in concept and time consuming in general. However, combining this method with the wide field of view provided in the largest class telescope, new window to the universe is going to open.

A conceptual design study of a high-speed spectro-photo- polarimeter for the Subaru Telescope is presented. Use of Superconducting Tunneling Junction detectors was compared with the use of Hybrid Photodiode or Photomos arrays with dispersive optics. Expected performances using at the Subaru Telescope were evaluated. This instrument will be extremely useful for a large variety of astronomical studies, especially for the field of mass accretion onto compact objects: white dwarfs, neutron stars and black holes, and low amplitude photospheric pulsation studies of stars.

The recent interest in detecting asteroids and Reentrant Space Debris has developed a need for large field of view, high-resolution focal pane array cameras to provide for mapping and tracking of these object. Ideally, a number of these cameras would be used around the world to provide a complete sky map. The Japan Space Forum is currently constructing one such facility in Japan. This paper describes the construction and features of two cameras to be used at the Japanese facility. One camera will be used in conjunction with a half-meter telescope and the second camera will be used with a one meter telescope. Features of the camera included radiative cooling of the focal pane array to minimize stress on the detector assembly, a specially designed focal pane array mount to maintain flatness, a 24 hour hold time liquid nitrogen dewar, and high speed fiber optic readout. The cameras feature arrays of up to ten SITe ST002 2048 by 4096 backthinned CCDs for high performance and wide field of view.

MegaCam is the new wide-field imaging camera currently being built for the new prime focus of the 3.6m Canada-France- Hawaii Telescope. Among all the challenges imposed by the ambitious specifications of MegaCam, there is the CCD Controller. We chose to realize a new highly integrated CCD Controller and to implement some other original solutions. This paper presents the solutions we retained.

The large Binocular Telescope is currently in the pre- erection phase. The instrument has been already funded and its first-light is expected shortly after that of the LBT. Given the peculiarity of the telescope optics we designed tow prime focus cameras with two five-lens refractive correctors, optimized in the blue-side and red-side of the visible spectrum respectively. This independent coating. Detectors also reflect this choice, being optimized separately. We present the most relevant features of the instrument, the optical design as well as the structural and mechanical layout. Each of the two Prime Focus cameras gather light form a very fast, F/1.14 parabolic primary mirror. The field is corrected over roughly half a degree in size, allowing optical performances in terms of 80 percent of Encircled Energy in better than approximately 0.3 inch. Focal length is slightly increased in order to provide a better sampling using 13.5 micrometers pixel size chips. The CCD array is made up with 4 EEV 42-90 chips, per channel, to obtain an equivalent 6000 by 6000 pixels optimizing the AR coating to the U,B,V and V,R,I,Z bands respectively. The array will be read out in 10 seconds using a 1Meegapixel/second controller with four video channels. The cryostat will use a state of the art dewar to reach an holding time of several days using a limited amount of liquid nitrogen. The whole mechanical design has bene modeled using Finite Elements analysis in order to check for mechanical flexures of the mount tube and of the optical components by themselves. A brief overview of the informative facilities to be provided with the instrument and of a few science case studies that can be attacked by this instrument are also given.

The goal of much recent engineering improvements at the 3.5m WIYN telescope has been to improve imaging performance that utilizes the good intrinsic seeing at Kitt Peak. This direction complements the efforts of high order adaptive optics by maximizing the usable field. The new 'mini-mosaic' camera, which is a mosaic of 2 4K by 2K SITE CCDs is in the final stages of commissioning. With its 0.14 arc-sec per pixel scale at the Nasmyth f/6.3 focus, it is capable of adequately sampling the best delivered images from the telescope, while maintaining a relatively large field of view. We present some early performance results from this new instruments, and demonstrate the excellent image quality over the entire 9.6-minute field.

The characteristics of gallium arsenide junction field- effect transistors (GaAs JFETs) and the performance of cryogenic readout circuits using GaAs JFETs with various gate sizes ranging form W/L equals 5micrometers /0.5micrometers to 200micrometers /200micrometers to systematically measure their static characteristics and low-frequency noise spectra. We found that the low-frequency noise voltage depends on the device size in the saturation region of GaAs JFETs at 4.2 K, and the power density of the noise voltage is inversely proportional to the gate are. These findings allowed us to determine the Hooge parameter of the GaAs JFET at 4.2 K to be 4 by 10^-5, assuming that the carrier mobility is 1.5 by 10³ cm²/Vs. On the other hand, we did not find the obvious correlation between the low-frequency noise and gate size in the ohmic region of GaAs JFETs. Based on these measurements for GaAs JFETs, we fabricated and tested a dual GaAs JFET, a source-follower-per-detector (SFD) circuit, and a 20 by 3 channel SFD circuit array. The Common-Mode-Rejection-Ratio (CMRR) of the dual GaAs JFET with W/L equals 50micrometers /20micrometers at 4.2 K was determined to be 40-60 dB under small power dissipation. The performance of SFD circuits and 20 by 3 channel SFD arrays for 2D far-IR Ge:Ga detector readouts are currently being evaluated.

FEROS is a new fiber-fed bench-mounted prism-cross dispersed echelle spectrograph which has been recently commissioned at the ESO 1.52-m telescope at La Silla. The opto-mechanical concept and performance predictions have been presented by Kaufer and Pasquini. In this contribution we present the test results as obtained during two commissioning runs in October and December 1998. Special emphasis is given to the measured performance sin efficiency, spectral resolution, straylight contamination, and spectral stability. The definite highlight of the FEROS instrument performance is the high peak detection quantum efficiency of 17 percent at 550nm. These measured numbers include the 20mirror telescope, the fiber link, the instrument, and the detector while the whole wavelength range is covered by a single exposure on a thinned EEV 2k by 4k 15 micron pixel CCD and a constant resolving power of R equals 48.000. In addition the FEROS instrument proved its high spectral stability by radial-velocity observations as carried out on the known radial-velocity standard star (tau) Ceti over a time base of 2 months. By recording a calibration-lamp spectrum in parallel with the object spectrum and by the use of a simple cross-correlation technique, a rms of 21 m/s has been obtained for a data set of 130 individual measurements. FEROS has ben made available to the ESO community in January 1999.

A brief overview is given of the data-reduction software (DRS) being developed at Geneva and Paris Observatories for the fiber-fed multi-object GIRAFFE spectrograph - part of the VLT FLAMES facility. The specific aspects of each of three modes is presented and discussed. The localization and the extraction are described in more details, the original features are outlined and the critical issue addressed. We discuss the problem of the PSF variation across the surface of the detector and the impact on the sky subtraction. The strategy of the real-time data quality assessment using the simultaneous calibration exposure is described. Some aspect of the DRS implementation and the VLT environment are discussed. The dirty prototype of the DRS under Matlab including the crude instrument simulator is discussed and some experiments measuring the expected accuracy of radial- velocity are shown.

Our team is designing and realizing the fiber links from which two spectrographs will operate at the second unit of the Very Large Telescope. GIRAFFE is an intermediate resolution spectrograph which has access to the entire field of the Nasmyth focus and is equipped with three fiber links producing alternately 132 spectra at the same time, 15 tri- dimensional spectra or 300 spectra for a single object. The last fiber link is intended to provide simultaneously with GIRAFFE, the observation of 8 objects at a higher resolution with UVES. All these observation modes will be available at VLT by the end of 2001. The specific development of the fiber links includes the systematic use of simultaneous calibration. This paper sums up the design of each mode and describes the specific developments required to optimize the performances of the fiber system. A full description of GIRAFFE, of its fiber link and of the related scientific programs is available at http://giraobs.obspm.fr.

The Potsdam Multi-Aperture Spectrophotometer (PMAS) is a flexible UV-visual integral field spectrograph designed for operation at different telescopes. It is based on a dedicated fiber spectrograph with a novel, fully dioptic collimator-camera system. The optical system was specifically optimized in terms of efficiency and stability for operation with a fiber input. The final optical design is described with remarks concerning the manufacture and acceptance test results.

The Echelle Spectrograph and Imager (ESI) is a multipurpose instrument which has been delivered by the Instrument Development Laboratory of Lick Observatory for use at the Cassegrain focus of the Keck II telescope. ESI saw first light on August 29, 1999. The optical performance of the instrument has been measured using artificial calibration sources and starlight. Measurements of the average image FWHM in echelle mode are 22 microns, 16 to 18 microns in broad band imaging mode, and comparable in the low- dispersion prismatic mode. Images on the sky, under best seeing conditions show FWHM sizes of 34 microns. Maximum efficiencies are measured to be 30 percent for echelle and anticipated to be greater than 38 percent for low dispersion prismatic mode including atmospheric, telescope and detector losses. In this paper we describe the instrument and its specifications. We discuss the testing that led to the above conclusions.

We describe the design and construction of the ESO UV-visual echelle spectrograph and the performance that was measured during its commissioning 1999. UVES is a dual-beam, grating crossdispersed echelle spectrograph. The resolution for a 1 arcsecond slit is 40,000. With narrower slits, resolutions of up to 80,000 and 115,000 are achieved with adequate sampling. UVES provides order separations of minimum 10 arcseconds at any wavelength between 320 and 1050 nm. The wavelength coverage is 100 nm in the blue arm and 200 or 400 nm in the red arm, with possibility to use a dichroic. Some concepts pioneered in UVES are now increasingly being used in other echelle spectrograph for large telescopes: a white pupil design, very steep replicated mosaic echelles, and large refractive cameras with external focus. Regular observations are starting in April 2000 at the Nasmyth focus of Kueyen, Unit Telescope 2 of the VLT array.

The VIRMOS consortium of French and Italian Institutes is manufacturing 2 wide field imaging multi-object spectrographs for the European Southern Observatory Very Large Telescope, with emphasis on the ability to carry over spectroscopic surveys of large numbers of sources. The Visible Multi-Object Spectrograph, VIMOS, is covering the 0.37 to 1 micron wavelength domain, with a full field of view of 4 by 7 by 8 arcmin² in imaging and MOS mode. The Near IR Multi-Object Spectrograph, NIRMOS, is covering the 0.9 to 1.8 microns wavelength range, with afield of view 4 by 6 by 8 arcmin² in MOS mode. The spectral resolution for both instrument scan reach up to R equals 5000 for a 0.5 arcsec wide slit. Multi-slit masks are produced by a dedicated Mask Manufacturing Machine cutting through thin Invar sheets and capable of producing 4 slit masks approximately 300 by 300 mm each with approximately slits 5.7 mm long in less than one hour. Integral field spectroscopy is made possible in VIMOS by switching in the beam specially build masks fed by 6400 fibers coming form a 54 by 54 arcsec² integral field head with a 80 by 80 array of silica micro-lenses. NIRMOS has a similar IFS unit with a field of 30 by 30 arcmin². These instruments are designed to offer very large multiplexing capabilities. In MOS mode, about 1000 objects can be observed simultaneously with VIMOS, with a S/N equals 10 obtained on galaxies with I equals 24 in one hour, and approximately 200 objects can be observed simultaneously with NIRMOS, with a S/N equals 10 obtained don galaxies with J equals 22, H equals 20.6 in 1h at R_eq equals 200. We present here the status of VIMOS, currently under final integration, with expected first light in the summer 2000, together with the final design of NIRMOS presented at the Final Design Review. The VLT-VIRMOS deep redshift survey of more with the final design of NIRMOS presented at the Final Design Review. The VLT-VIRMOS deep redshift survey of more than 150000 galaxies over the redshift range 0 < z < 5 will be undertaken based on 120 guaranteed nights awarded to the project.

We are building the Kyoto tridimensional spectrograph II and are planning to mount it on Subaru telescope. The spectrograph has four observational modes: Fabry-Perot imager, integral field spectrograph (IFS) with a microlens array, long-slit spectrograph, and filter-imaging modes. The optics is designed to be used in wide wavelength range from 360 nm to 900 nm. The design well matches with high spatial resolution of Subaru: 0 inch .06 pixel^-1 in Fabry- Perot mode, for which we actually will use binning before adaptive optics at optical wavelengths becomes available, and 0 inch .1 lens^-1 in microlens array mode. These well sample image sizes obtained by Subaru, which are about 0 inch .4 in relatively good conditions. We have evaluated a point spread function of our cylindrical microlens array and found that it consists of a diffraction pattern and more extended component which probably comes from border regions between microlenses. With a suitable mask at the micro pupil position, the crosstalk between spectra will be limited down to a few percent. With a suitable mask at the micro pupil position, the crosstalk between spectra will be limited down to a few percent. We have succeeded in synchronizing frequency switching of Fabry-Perot etalons with the movement of charge on the CCD. This technique enables to average out all temporal variations between each passband.

A design for HROS has now been developed which is able to meet the science requirements as specified by the community. This design calls for a Cassegrain-mounted échelle spectrograph with a resolving power of R=50,000. Efficiency is maximised both by the location at the Cassegrain focus, and by the adoption of prism (as opposed to grating) cross-dispersion. All of the transmitting optical elements will be made of fused silica, to maximise to efficiency in the UV. The design relies on the concept of an 'immersed' échelle, in which the ruled surface of the échelle grating is 'immersed' in a fused silica prism. This paper outlines the working of the HROS instrument and summarises the progress that has been made on the project, on a modular level. The reader is also refered to 3 additional paper in these proceedings. Paper 4008-14 discusses the flexural performance of this instrument, paper 4008-15 discusses the optical design of the instrument and paper 4008-98 describes the Active Flexure Compensation (AFC) that is to be used on the HROS instrument.

HARPS is a high-resolution spectrograph dedicated to the search for extra-solar planets by means of precise radial- velocity measurements. It will be installed on the ESO 3.6-m telescope at the ESO La Silla observatory and should start its operation son end 2002. The observations will provides for many solar-type stars of the solar vicinity a high number of RV measurements with 1 ms^-1 accuracy and will allow the detection of Saturn-like planets. The instrument is a fiber-fed and cross-dispersed echelle spectrograph. It has been designed and optimized to measure the Doppler shift of the stellar spectrum by means of cross correction of the spectrum with a numerical mask. For this purpose the entire spectrum in the wavelength range from 380 nm to 680 nm is imaged on a 4k4 mosaic CCD detector at a spectral resolution of about R equals 90,000. Wavelength calibration is provided by a ThAr lamp simultaneously imaged on the CCD. The HARPS spectrograph is also characterized by its high optical efficiency and the high thermo-mechanical stability. Indeed the whole instrument will be temperature controlled and operated in vacuum, in order to assure its short and long-term stability. This paper summarizes the present status of the project and describes the optical and instrument design of HARPS. Also a short presentation of the scientific objectives pursued with HARPS is given by the author.

The National Optical Astronomy Observatory is developing a new, wide-field, imaging spectrograph for use on its existing 4-meter telescopes. This Next Generation Optical Spectrograph (NGOS) will utilize volume-phase holographic grating technology and will have a mosaiced detector array to image the spectra over a field of view that will be something like 10.5 by 42 arc-minutes on the sky. The overall efficiency of the spectrograph should be quite high allowing it to outperform the current RC spectrograph by factors of 10 to 20 and the Hydra multi-fiber instrument by a facto of fiber to ten per object. The operational range of the instrument will allow observations within the optical and near-IR regions. Spectral resolutions will go from R equals 1000 to at least R equals 5000 with 1.4 arc-second slits. The large size of this instrument, with a beam diameter of 200 mm and an overall length of nearly 3 meters, presents a significant challenge in mounting it at the Cassegrain location of the telescope. Design trades and options that allow it to fit are discussed.

BEAR is a high-resolution imaging FTS in operation on the CFH Telescope, obtained by the coupling of a step-scan FTS and a near IR camera. TIGER is a integral field spectrometer (IFS) for the visible, which was in operation at CFHT, in which a microlens array segments a small entrance field. A new version (OASIS() has been made to be used behind an adaptive optics system, leading to an even smaller field of view. By adapting a TIGER-type field segmentation on an imaging FTS, the instrument remains an IFS more versatile than each instrument taken separately. Such an instrument would be characterized by the access to the same spatial resolution, but on a field larger than with a TIGER instrument, a larger spectral coverage, a continuously chosen spectral resolution as in an FTS, and a better sensitivity than a BEAR instrument. For that, a mosaic of microlens array, made in an IR glass transparent between 1 and 5 microns, is paving one input port of a large field dual-output port interferometer which has a provision for a mirror travel of 5 mm, giving the choice for a resolution up to 10⁴ at 1.7 micrometers . A prism on each output beam disperse the entrance points which are imaged on InSb array. A camera on each output beam records an image at each step of the FTS. This concept can be of interest for the instrumentation of a large ground-based telescope or better for a NGST.

In order to enhance the spectroscopic capabilities of the William Herschel Telescope (WHT) we have recently completed an integral field unit comprising 1000 elements. Integral field units maximize the efficiency of a spectrograph by re- formatting a 2D field in order to match the entrance slit of the camera. Such techniques enable high-resolution spectral data to be obtained over the whole field simultaneously, and are particularly suited for use with adaptive optics systems. TEIFU is an optical fiber system employing microlens arrays for input and output coupling. The field is divided into two halves, permitting object and background to be derived during the same exposure. In addition, the fields can be optically re-positioned to form a larger, single field for greater object coverage. Thus the observer can choose between different observing modes to emphasize background subtraction or contiguous field. The fore-optics can be changed to alter the image scale and to interface to the NAOMI adaptive optics system which is currently under construction. TEIFU in its present configuration as tested on the WHT, gives a spatial sampling of 0.25 arcsec with a total field of 7.8 by 7.0 arcsec, but a 0.125 arcsec sampling option may be provided. We are also considering an option to upgrade TEIFU for near IR operation. This paper will outline system design, operation and preliminary results.

The Optical System for Imaging and low Resolution Integrated Spectroscopy (OSIRIS) will be a Day-One instrument of the Spanish 10.4 m telescope Gran Telescopio Canarias, whose first light is planned for 2002. GTC will be installed at the Observatorio del Roque de los Muchachos in La Palma, Spain. OSIRIS three primary modes are imaging and low resolution long slit and multiple object spectroscopy. The instrument is designed to operate from 365 to 1000 nm with a field of view of 7 by 7 arcminutes and a maximum spectral resolution of 5000. Among the OSIRIS main features are the use of tunable filters for direct imaging, the use of Volume Phase Holographic Gratings as dispersive elements for spectroscopy, and the implementation of an articulated camera to provide maximum spectroscopic efficiency and versatility. Here we present a general description and an overview of the main instrument characteristics.

We review recent progress in the field, using a s a framework a partial list of present limitations and problems: CCD and mosaic size, packing fraction in mosaics, red response and fringing, and intrinsic point-spread function due to lateral charge diffusion. Related topics such as orthogonal-transfer CCDs and the special requirements of adaptive-optics wavefront sensor are also discussed. Only cursory attention is given to other relevant issues, such as readout speed and anti-blooming techniques.

S-Cam is a cryogenic camera for ground based astronomy designed around a 6 by 6 array of Ta-Al Superconducting Tunnel Junctions. The camera has been conceived as a technology demonstrator, aiming to provide the potential of this new generation of single photon counting detectors at a ground-based telescope in La Palma, an improved version of the camera has been developed and tested. In this paper we provide an overview of the latest camera performance, a description of the up-dated S-Cam 2 system and a summary of the main test results. An example of the novel astronomical data obtained during the test campaign conducted in December 1999 are also shortly described.

MegaCam is the new wide-field imaging camera currently being built for the new prime focus of the 3.6m Canada-France- Hawaii Telescope. The camera will offer a 1 square degree field of view and is built around a mosaic of 40 2K by 4.5K CCD devices. The delivery of the CCDs is proceeding along the schedule, the project passe dits final design review and the realization phase started, for an expected delivery to CFHT in Summer 2001.

This paper presents the VLT-VIMOS Integral Field Spectroscopy Unit. This unit allows to observe a very large 54 inch by 54 inch field on one edge of the VIMOS instrument multi-object field. This unit contains 6400 sets of (mu) lenses-fibers-(mu) lenses, producing the equivalent of a 72 arcminute by 0.67 arcsec slit projected on the sky. Two spatial resolution are offered, coupled with the low and high spectral resolution of VIMOS. The design philosophy, technological choices and the first test result of the assembled unit are presented.

The OH airglow emission lines are the dominant source of background emission in the near-IR J and H bandpasses. In principle, these emission lines can be avoided by observing at sufficiently high spectral resolution, rejecting pixels contaminated by OH lines, and rebinning to the desired resolution. Two trade-offs to this approach are non- negligible detector noise per pixel and the added expense of instrumentation with higher resolution. In this contribution, we simulate various observed and desired resolutions as a function of detector noise and target brightness to develop a set of guidelines for the optimal resolution in a variety of observing programs. As a general rule, observing at a 2-pixel resolution of 2000-4000 provides optimal OH rejection for a wide range of detector noise and source signal.

We describe the current status of the technical aspects of the GOHSS project. It consists of a fiber-fed NIR spectrograph for faint objects. It will be a second-light instrument for the Nasmyth focus of the 3.5m Galileo telescope located on La Palma. GOHSS is an innovative instrument which accomplishes OH night-sky subtraction, differently from the hardware solution used by other devices; it provides a multiechelle design with software OH subtraction capable of yielding about 25 spectra in the z,J and H bands at an effective spectral resolution of about 4000, which is necessary to strongly reduce the impact of atmospheric OH lines. The GOHSS design is completed and the operative phase is already started through the procurement of the most important components. We have also started to develop the data reduction package for the instrument and the first result of the 1D approach as presented.

We describe the Nordic Optical Telescope's facility short- wavelength IR instrument, NOTCam. The instrument will be capable of wide-field and high-resolution imaging, long-slit and multi-object grism spectroscopy, coronography, and imaging-and spectro-polarimetry. First light will be in mid- 2000. Current progress is summarized and some problems we have encountered and overcome are discussed.

At the 1998 SPIE meeting we described a cryogenic, high- resolution spectrograph for use in the 1-5 micrometers region. At that time Phoenix had been used at Kitt Peak for about a year. In the intervening two years we have worked extensively with the instrument and have modified a few aspects of the design to bring the operational characteristics more closely into agreement with the original specifications. Changes to the instrument since 1998 that resulted in significant improvements in performance will be discussed. We will review the current operational characteristics of the spectrograph. Phoenix is a facility instrument of the National Optical Astronomy Observatory with use planned at Gemini South and CTIO.

Gemini have funded a design study to investigate the technologies needed in a versatile multi-object spectrograph for IR astronomy. We report on our investigations into wide- field spectroscopy using multiple integral-field units (MIFUs) to match particular areas of interest to the available detector(s). Such technologies enable integral field spectroscopy of several targets over a much wider field than can be covered with a single IFU. A brief overview of the scientific rationale for a multipel0IFU capability matched to multi-conjugate adaptive optics, and with its wider uncorrected field, on Gemini is given. A proposed method of deploying MIFUs is then described along with the optical consequences of the method.

This paper presents the design of a near IR camera for the 9.2 m Hobby-Eberly Telescope (HET) Low Resolution Spectrograph (LRS), which will cover the wavelength range 0.85 to 1.35 micrometers . The LRS-J, an upgrade to the existing LRS, replaces the optical camera with an f/1 camera optimized for the J-band. The instrument design is strongly motivated by the desire to observe galaxies at 1 < z < 2, where the principal strong spectral features used to measure redshifts are shifted into the J-band. Since we are primarily interested in wavelengths up to 1.35 micrometers , mating the cryogenically cooled camera to the warm LRS spectrograph does not result in enough thermal background emission to compromise its performance. LRS-J represents a rapid and cost-effective way to enable multi-object near-IR spectroscopy on a very large telescope.

We present the main characteristics and astronomical results of SWIRCAM, a NIR imager-spectrometer mainly devoted to the search for extragalactic Supernovae, in the frame of the SWIRT project, a joint scientific collaboration among the Astronomical Observatories of Rome, Teramo and Pulkovo. The camera is currently at the focal plane of the AZT-24 1.1 m telescope at the Observing Station of Campo Imperatore, operated by the Astronomical Observatory of Rome. SWIRCAM saw its first light during December 1998 and it is currently employed for both the SWIRT operative phase and other institutional projects.

Wide-field IR cameras, operating on the ne generation of large telescopes, offer unprecedented gains in the detection of faint sources and in observing efficiency for both direct imaging and spectroscopy. PISCES, a near-IR wide-field camera designed for the f/9 secondaries of the Steward 2.3m and 6.5m MMT, is one such instrument that has been operational for over one year. Equipped with a 1024 by 1024 HAWAII HgCdTe array, PISCES offers an 8.5 arcminute field at the 2.3m and a 3.1 arcminute field at the MMT. Here we present our design to upgrade PISCES with a low resolution grism for single and multi-object spectroscopy. The design allows J, H, and K-band spectroscopy in orders 5, 4 and 3, respectively. The combination of 6.5m aperture and multi- object capability will make PISCES a powerful tool for extending our knowledge of the low-mass regime of the initial mass function as well as star-formation in 0.5 < z < 1 galaxy clusters. We discuss design and fabrication issues and simulate the performance of the grism system.

LUCIFER is a full cryogenic NIR spectrograph and imager to be built by a consortium of fiber institutes, Max Planck Institut fuer Astronomie in Heidelberg, Max Planck Institut fuer Extraterrestrische Physik in Garching, Astronomisches Institut der Ruhr Universitaet Bochum and Fachhochschule fuer Technik und Gestaltung in Mannheim. The instrument has been selected as one of three first-light instruments for the Large Binocular Telescope on Mt. Graham, Arizona which first mirror becomes available to the community in early 2003. The second mirror and a second more or less identical spectrograph/imager follows 18 months later. Both LUCIFER instruments will be mounted dat the bent Gregorian foci of the two individual LBT-mirrors and include six observing six observing modes: seeing and diffraction limited imaging, seeing and diffraction limited longslit spectroscopy, seeing limited multi-object spectroscopy and integral-field spectroscopy. The detector will be a Rockwell HAWAII-2 HgCdTe-array with a pixel-size of 18(mu) .

We are developing a high-resolution cross-dispersed echelle spectrograph for installation at one of the coude foci of the new AEOS 3.67 meter telescope, operated by the Air Force Space Command on Haleakala, Maui, Hawaii. The spectrograph will consist of two major subsystems: an optical arm for the wavelength range 0.5-1.0 micrometers and an IR arm for the range 1.0-2.5 micrometers . Both arms of the spectrograph use a white- pupil collimator design to maximize grating efficiency and to limit the size of the camera optics. The optical arm of the spectrograph will use deep-depletion CCDs optimized for operation near 1.0 micrometers . The IR detector will be a 2048 by 2048 HgCdTe array that has bene developed by the Rockwell Science Center for this project. Both the optical and IR arms of the spectrograph will be equipped with slit-viewing cameras for object acquisition and control of a fast guiding tip-tilt mirror located in a pupil image in the spectrograph fore optics.

One of the major problems in trying to observe a faint companion from the ground comes from atmospheric turbulence. We present here a new camera capable of subtracting the effect of atmospheric turbulence on the PSF by taking simultaneous images of a star at three different wavelengths centered around the 1.6 micrometers methane absorption band typical of brown dwarfs and jovian planets. It is to be used at the adaptive optics compensated foci of the Megantic 1.6m, the CFHT 3.6m and the Gemini 8.2m telescopes. We present its design, which results from experiences at Megantic and CFHT with various techniques. Simulations predict there will be a 15 magnitude increase in sensitivity for detection of faint companions.

EMIR is a near-IR, multi-slit camera-spectrograph under development for the 10m GTC on La Palma. It will deliver up to 45 independent R equals 3500-4000 spectra of sources over a field of view of 6 feet by 3 feet, and allow NIR imaging over a 6 foot by 6 foot FOV, with spatial sampling of 0.175 inch/pixel. The prime science goal of the instrument is to open K-band, wide field multi-object spectroscopy on 10m class telescopes. Science applications range from the study of star-forming galaxies beyond z equals 2, to observations of substellar objects and dust-enshrouded star formation regions. Main technological challenges include the large optics, the mechanical and thermal stability and the need to implement a mask exchange mechanism that does not require warming up the spectrograph. EMIR is begin developed by the Instituto de Astrofisica de Canarias, the Instituto Nacional de Tecnica Aeroespacial, the Universidad Complutense de Madrid, the Observatoire Midi-Pyrennees, and the University of Durham. Currently in its Preliminary Design phase, EMIR is expected to start science operation in 2004.

Although the interest in PtSi IR focal pane array has waned due to its low quantum efficiency compared with InSb and HgCdTe arrays, it is very potential in observing brighter celestial objects. We explored the possibility of applying it to the observation of IR solar spectrum. In the paper, the methods of the simulation and calibration in our observation are introduced and discussed in detail. Using this kind of camera, a new observational band is added to the 2D Multi-band Solar Spectrograph at Yunnan Observatory. The dispersion for FeI 1.56 micrometers of the new IR solar spectrograph is 0.0722 angstrom per pixel, and each vertical pixel represents 0.51 inch of solar disk. It is specially suitable for 2D spectroscopic observation of the deepest solar photosphere. Some primary observation results are also presented.

We describe the design, characterization and performance of the IR Camera for Adaptive Optics at Lick (IRCAL). IRCAL is a 1-2.5 micron camera optimized for use with the LLNL Lick adaptive optics system on the Shane 3 m telescope. Using diamond-turned gold-coated optics, the camera provides high efficiency diffraction limited imaging throughout the near- IR. IRCAL incorporates optimizations for obtaining high dynamic range images afforded by adaptive optics, coronagraphic masks, and a cross-dispersed silicon grism for high resolution spectroscopy.

We present in this paper a preliminary design of a near IR wide field imager for the prime focus of the ESO 3.6 meter telescope. Such a design was developed in response to the increasing demand in the astronomical community for a fast and efficient Near IR survey facility taking profit of the small pitch 1k² and 2k² Hybrid Focal Plane Arrays recently made available. The prime focus is a privileged location for high throughput instrument since it allows to design imagers with a relatively small amount of optical elements . At the prime focus however the plate scale is constrained to a narrow range of possibility by the F/number of primary mirror since the need to correct for aberrations does not leave much freedom for focal elongation. The ESO 3.6 telescope has a F/3 primary mirror. Unlike most of the new technology telescope of the same size, e.g. NTT or TNG, that have F/2.2 primaries, the longer focal length of the ESO 3.6 allows to reach plate scales of the order of about 15.2 as/mm, matching the scientific requirement for wide field NIR surveys. The Camera consist in a prime focus corrector made of 4 spherical lenses in fused silica providing a corrected field of 45 arcmin 0. The limited number of optical elements maximizes the throughput but does not allow to re-image the telescope pupil to position a cold stop. Therefore a detailed mode of the telescope thermal background has been sued to optimized the baffling system. Only a non-contiguous fraction of the corrected field is covered by 4 Rockwell HAWAII II 2k² HgCdTe chips.

First laboratory test result of CONICA are presented for the variety of observation modes: using the final ALADDIN- Detector, IR images in direct, spectroscopic and polarimetric modes are compared to theoretically expected diffraction limited point spread functions. In long slit spectroscopy, wavelength calibration and spectral resolution is demonstrated for the different grism, slit and camera combinations.

We have developed the mid-IR Camera and Spectrometer (MICS), which optimized for ground based observations in the N-band atmospheric window. The MICS has two observing capabilities, imaging and long slit low-resolution spectroscopy. The major characteristics of the MICS are nearly diffraction-limited performance, both in imaging and in spectroscopy and the capability to take a spectrum of the whole N-band range with a spectral resolving power of 100 under one exposure. The MICS employs a 2D array of 128 by 128 Si:As BIB detector, an aberration-corrected concave grating, and a high-speed read out system of a compact design with high sensitivity. In mid-IR observations form the ground, there is a large background radiation from the telescope and the sky. The fluctuation of the background radiation is not well understood so far. We measured the sky fluctuation in the N- band region with the MICS on the UKIRT. These measurements have revealed that (1) the sky noise was dominant below 0.5 Hz when the sky condition was good, and (2) the sky noise has strong excess at the positions of atmospheric water vapor lines than those without water vapor lines. In this paper, we describe the design of the MICS, including optics, cryogenics, and electronics, and its performance when used on the UKIRT. We also discuss sky noise measured by the MICS in the N-band region.

MIRTOS, Mid-IR Test Observation System, is a high spatial resolution mid IR (MIR) camera for the Subaru Telescope. It consists of two IR imagers. One is for MIR bands with a Si:As array with 320 by 240 pixels. It has 21 by 16 arcsec field of view (FOV) with a pixel scale of 0.067 arcsec. It also images the pupil of the telescope. The other is a near IR camera. A 256 by 256 InSb array with 0.028 arcsec/pixel is used to image 7 by 7 arcsec FOV at one corner of the MIR FOV. We apply Shift-and-Add (SAA) technique; a technique that shifts images detecting the displacements and adds them to cancel seeing. However it is often difficult to shift and add MIR images using a reference within because of low sensitivity in MIR for short exposure time. We solve this problem utilizing NIR images taken simultaneously as position references. We call this method two-wavelength shift-and-add (TWSAA). In this paper we show result from the test observations. 1) Pupil image was taken. It shows hot structures around the secondary mirror that are now planned to be covered by reflecting plates to direct the beam to the sky. 2) Correlation of motion between MIR peak position and NIR centroid position shows that NIR images can be used as TWSAA reference for MIR observations. 3) On a standard star and the core NGC 1068, SAA method was applied to reconstruct images. Resulting images show higher spatial resolution than previous observations.

The Advanced Camera for Surveys (ACS) is a third generation science instrument scheduled for installation into the Hubble Space Telescope (HST) during the servicing mission 3B scheduled for June 2001. The instrument has three different cameras, each of which is optimized for a specific set of science goals. The first, the Wide Field Camera, will be a high throughput, wide field optical and I-band optimized camera that is half-critically sampled at approximately 570 nm. The second, the High Resolution Channel (HRC) has a 26 inch by 29 inch field of view, it is optimized for the near- UV and is critically sampled at approximately 630 nm. The third camera, the Solar-Blind Camera is a far-UV, photon counting array that has a relatively high throughput over a 26 inch by 29 inch field of view. Two of the three cameras employ CCD detectors: the WFC a mosaic of two SITe 2048 by 4096 pixel CCDs and the HRC a 1024 by 1024 CCD based on the Space Telescope Imaging Spectrograph 21 micrometers pixel CCD. IN this paper we review the performances of the devices baselined as flight candidates.

We describe the current status of photon-counting cameras developed at Observatoire de la Cote d'Azur for ground-based interferometry. The development aims at improving both the time and spatial resolution of our detectors using fast intensified CCD chips, real-time pattern analysis techniques for super spatial resolution implemented on multi-DSP boards. The on-line processing includes a quick look-up of estimated interferometric visibilities in order to dynamically drive the interferometric parameters during the observations. Our cameras will be used in priority on the GI2T/REGAIN dual channel spectrograph which allows 'spectral bootstrapping', i.e. blind fringe acquisition in blue wavelengths, while tracking fringes in red wavelengths. The other applications concern adaptive optics coronography in the visible at Observatoire de Haute Provence and the SPID speckle-interferometric camera at Observatoire de Lyon.

Superconducting tunnel junctions (STJs) have been demonstrated as photon counting detectors in the UV-NIR wavelength range. They combine a modest wavelength resolving power with fast response and high detection efficiency over a broad wavelength band. This makes this type of detector an interesting alternative to the present generation of detectors used in UV/optical astronomy, such as CCDs and micro-channel plates. Practical applications require imaging detectors with large sensitive area and good spatial resolution. While the feasibility of small arrays of closely packed STJs which are individually biased and read-out has already been demonstrated, the development of large format arrays is limited by the large number of electronics chains and wire connections to the cold detector which would be required. An alternative approach is to use a large area absorber combined with a few STJs at the edges or corners. A photon's energy as well as its absorption position in the absorber can be derived form the signal amplitudes measured in the STJs. In this paper the performance in terms of wavelength resolving power and position resolution of four different linear geometries of Ta absorbers, read out with Ta-Al STJs, is investigated and compared with single STJs. The UV and optical spectra obtained with the absorbers show resolving powers within a factor of two of the theoretical limit. In particular, a measured resolving power at (lambda) equals 300 nm of approximately 16 with a position resolution of approximately 9 micrometers is achieved with a 100 by 50 micrometers ² absorber in between two 50 by 50 micrometers ² STJs.

The Cassegrain location of the high resolution optical spectrograph (HROS) for the 8-meter Gemini telescope presents a difficult challenge in controlling mechanical flexure. This is especially the case for a high-resolution spectrography, which requires large and heavy optical components. In HROS, to achieve the required spectrum stability of 2.0 micrometers /hr, we developed a closed-loop monitoring and correction system capable of measuring the spectrograph flexure in as it happens and actively compensating for image motion.

Two digital television complexes with three small telescopes have been successfully used for photometric and spectrophotometric observations at the Crimean astrophysical observatory. All data have absolute calibration, recorded to a personal computer and reduced for flat field and distortion. The high sensitive pickup tube I-isocon of the first complex has mounted at the 500-mm meniscus telescope. The second tube coupled to the 200-mm telescope used as a television guide. The second complex can be used with 380-mm telescope. Two TV complexes can work simultaneously observing one object or different objects. The complex is equipped with filters reproduced photometric B, V, R bands and two spectrographs. The BVR observations of stars with magnitudes up to 19^m are carried out with the accuracy of 0.05^m with exposure time about several minutes. Simultaneous observations in the BVR bands can be carried out also. The original afocal slitless spectrograph allows us to receive the absolute spectral distribution in spectra of point sources up to 14^m with spectral resolution about 50 angstrom and exposure time of several minutes. The slit spectrograph allowed us to receive unique Jupiter observations during its collision with comet Shoemaker-Levy 9.

The international environment of astronomical research and the worldwide distribution of astronomical institutes and observatories create a high demand of flexibility on the designers and engineers developing and building scientific instruments for use in astronomical research programs. In particular with respect to the safety performance characteristics of the scientific instruments a harmonization process among the various safety requirements could lead to more safety awareness and understanding of these requirements, and also to some kind of standardization concerning the methods and means used during the acquisition period of the instruments to reduce safety risks to an acceptable level.

This paper reviews transmission anomalies of freestanding frequency surfaces (FSS), which are commonly used as filter and guiding structures in the far IR region. The effect is characterized by a sharp breakdown of the transmission factor in the passband and was first revealed by measurements. The paper demonstrates the numerical confirmation of these effects by computer aided calculations based on the method of moments combined with the Floquet theorem. The scattering characteristics are examined for different sets of parameters like the geometry of the structure, frequency, angle and polarization of the incident wave. The modeling with electric and magnetic currents allows the characterization of both patch and slot arrays.

To assist in determining and tracking the transmitted wavefront error of the Canada-France-Hawaii Telescope filter stock, a method for determining the wavefront error, in transmission, of the filters using a white-light source has been investigated. The method uses the ef software from Laplacian Optics which determines a wavefront based on curvature sensing. The filter is placed in a collimated white-light beam. After the filter, the beam is imaged onto a small CCD which is placed on a translation stage. The CCD is moved to either side of focus to produce the two out-of- focus images required for curvature sensing. The results of this investigation compared to results using a n interferometer will be presented.

The Gemini North Adaptive Optics Instrument is currently under construction at the Herzberg Institute of Astrophysics, National Research Council Canada. In the Altair design, many optical components must be positioned relative to one another within micron-level tolerances. These relative positions are required to be stable over the life of the instrument, all optical components will be mounted to an optical bench. The choice of materials and manufacturing processes are key to the long-term dimensional stability of the optical bench and represent a significant engineering challenge. The candidate materials and manufacturing processes are key to the long-term dimensional stability of the optical bench and represent a significant engineering challenge. The candidate materials and manufacturing options will be discussed. In addition, the final design, and the manufacturing and heat-treatment processes for the Altair optical bench will be presented.

In this paper a preliminary design for a prime focus corrector to be mounted at the Telescopio Nazionale Galileo (TNG) is presented. The telescope is located on La Palma and has a primary mirror of 3.5 m with a hyperbolic sag. Two optical designs have been considered in order to exploit detectors with 1k by 1k and 2k by 2k pixels. Each design makes use of four lenses, tow of which are aspherical. The first lens diameters are respectively, of 140 mm and 320 mm for the two kinds of detectors. For both designs the telescope pupil is deliberately not re-imaged, and therefore it will not be possible to insert a cold stop. For such a reason particular care has been dedicated to the telescope thermal background study, in order to optimize the baffling system. The optics is able to correct fields of 11 feet by 11 feet and of 60 feet by 60 feet, depending on the design. It should be considered that the particular geometry of the focal plane array mosaic does not permit a full exploitation of the entire field, being based on a combination of four detectors. The selected chips are HgCdTe manufactured by Rockwell, i.e. Hawaii I in the 1k by 1k case and Hawaii I in the 1k by 1k case and Hawaii II in the 2k by 2k case.

OzPoz is a multi-fiber positioner which will feed Nasmyth spectrographs on one of ESOs VLT unit telescopes. Its concept follows that of the positioner for the two degree field facility on the Anglo-Australian Telescope. Thus its fibers will be fed from prims housed in buttons which attach magnetically to steel focal plates; a robotic system will position the buttons; and the plates will be interchanged so one can be re-configured while the other is gathering starlight. However, OzPoz has a number of novel features, most notably the use of a pneumatically operated gripper which relies for its accuracy and friction free rotation on air bearings. The robot motions also employ air bearings, with vacuum preloading. The mechanism which exchanges focal plates has been carefully designed to ensure it will survive the maximum likely earthquake on Paranal without significant damage.

LAMOST is a 4m class wide field of view telescope special design for multi-object spectroscopic observation. For the observing mode of low spectral resolution, 4000 fibers are employed. 4000 fibers is a challenge to the positioning of multi-object spectroscopy nowadays. We are going to use a method that each fiber is controlled by independent positioner, like MX. But the working area of every fiber is limited in quite small a region comparing with whole focal pane to reduce the size of the position device so that 4000 positioners could be homed inside focal plate. The first problem of the method we have considered is the usable rate of fibers. Because 4000 fibers employed by LAMOST and a couple of times of targets in a field of view of LAMOST, a reasonable usable rate of fiber is expected. But there are still a few problems we have to solve before the method could be used. Paralleled control of 4000 positioners, the electronic and mechanical interference between positioners may be a trouble. And the stability of bases plate should be considered. Some models of positioners has been proposed and made for test in past two years. They have been examined in laboratory and in observatory.

This paper describes a dichroic beam splitter design for convergent f/8 telescope beams with high throughput, excellent image quality and a large field of view. Ray tracing and laboratory test demonstrate the workability of the concept. The beam splitter design is applicable in tip- tilt systems and multicolor imaging system as well in the wavelength range from 360 nm to 1000 nm. Initial observations with a prototype of this new beam splitter design are planned in the second half of the year 2000 at the 1.3m telescope on Mt. Skinakas.

The paper describes the optical design for the Multi Object Dual Spectrograph (MODS) for the Large Binocular Telescope (LBT). MODS is designed to cover the entire spectrum accessible to silicon CCDs from a ground-based telescope with the highest possible throughput. Multi-object capability is available using 0.6 arc-minute in diameter with reduce image quality. Under the very best seeing conditions and with the LBT adaptive optics in operation, slit widths of 0.3 arc-seconds can be used to enhance the resolving power and/or reduce the background. The optical path is divided into blue and red channels by a dichroic beam splitter following the slit masks. The blue channel covers a wavelength range from the atmospheric cut-off at approximately 300 nm to approximately 550 nm while the red channel covers the range from approximately 550 nm to the limit of useful sensitivity of silicon CCDs. This approach allows the optimization of transmissive and reflective coatings to provide the very highest throughput for each channel. The design is conventional in the use of reflective parabolic collimators. However, the cameras are designed as decentered Schmidt/Maksutovs with large aspheric coefficients for the inner surfaces of the correctors. This approach enables the field flatteners and detector to be positioned outside the beam entering the camera where it will not vignette. Figures are presented showing image quality for imaging and spectrographic modes.

The Convex Camera Structure (CCS) is a welded Invar structure that supports of the final two optical elements in the Altair science path. The CCS assembly holds the large camera mirror, a 295 mm diameter off-axis conic, in close relation with the smaller convex mirror, a 146 mm diameter off-axis conic. The image quality requirement imposed stringent spatial tolerance between the two mirrors: +/- 45 micrometers in axial separation, +/- 70 micrometers in decenter and +/- 30 arcsecond in allowable tilt. These tolerances include contributions from initial alignment error, geometrical and dimensional changes due to gravity and temperature. This paper gives an overview of the CCS and its associated components. Flexures are integral part of the CCS assembly, they are utilized to connect the CCS to the aluminum main optical bench, the mirror to the six degree- of-freedom mount, and the 6DOF mount to the CCS end plate. The design considerations and engineering analyses of the CCS assembly are outlined.

A small prime focus corrector for the new 6.5 m MMT telescope has been built using small off-the-shelf optical components. Its purpose is to aid in initial test of the primary mirror. At f/1.25 the paraboloidal primary is by far the fastest of any large telescope, and without correction the field and the plate scale are both inconveniently small. The corrector design is based on the Maksutov camera, with a concave spherical relay mirror and a refractive meniscus corrector to subtract the coma introduced by the primary mirror and balance out the spherical aberration from the relay mirror. The result is a 48 arcsec field at f/7 with 80 percent encircled energy within 0.25 arcsec diameter. The instrument includes a CCD with 500 by 500 20 micron pixels, a guide camera and provision for a fast wavefront sensor to measure mirror figure and atmospheric turbulence. The instrument was tested at the telescope before implementation of the active control of mirror support forces or the mirror air conditioning system. Nevertheless images with FWHM 0.53 arcsec were recorded.

Altair is the facility adaptive optics system for the Gemini-North 8-meter Telescope. This paper gives an overview of the key mechanical systems and feature of Altair including the optical bench assembly mechanics and opto- mechanics, the instrument structure, and the electronics enclosure.

Mechanically Actuated Reconfigurable Slits (MARS) provide a remote means of creating and reconfiguring multi-object spectrograph slits. The MARS design provides 99 individually positionable variable-width slits over a 180 mm square telescope focal plane. It offers fast, remote configurable, suitable for optical and IR ground and space based multi- object spectrographs. The slit apertures are arranged in three columns, to optimize the placement of the spectra on the detector. Two distinct MARS designs are presented, mechanically actuated strips and mechanically actuated rolls. The methods configure the slits by translating thin strips of material across the telescope focal pane. Slits are formed either by an aperture cut in the strip or by bringing two separate strips together head-to-head.

ISAAC is an IR imager and spectrometer installed at the first Unit Telescope of the ESO-VLT. ISAAC started science operations in April 99. We present the instrument characteristics, its performance and operation.

This paper describes the performance of NIRSPEC, the cryogenic cross-dispersed IR echelle spectrograph for the Keck II telescope on Mauna Kea. NIRSPEC employs a 1024 by 1024 InSb array, diamond-machined metal optics and closed- cycle refrigeration on achieve high throughput and low backgrounds. The instrument operates directly at the f/15 Nasmyth focus, but can also be used in conjunction with the Keck adaptive optics system. First Light was obtained on April 25, 1999. As expected, the performance is detector- limited at short wavelengths and background-limited at longer wavelengths. All of the design goals have been met and result illustrating the optical performance and sensitivity are reported.

We report current status of the IR Camera and Spectrograph (IRCS) for the Subaru Telescope. IRCS is a Subaru facility instrument optimized for high-resolution images with adaptive optics (AO) and tip-tilt at 1-5 micrometers . IRCS consists of two parts: one is a cross-dispersed spectrograph providing mid to high spectral resolution, the other is a near-IR camera with two pixel scales, which also serves as an IR slit-viewer for the echelle spectrograph. The camera also has grisms for low to medium resolution spectroscopy. We have just completed the first engineering run about one month before this SPIE conference. It was an initial performance evaluation without AO or tip-tilt to check IRCS and its interface to the telescope. We confirmed the basic imaging and spectroscopic capability we had estimated.

We present results on the integration and testing of an imaging spectrometer for the 1-5micrometers wavelength band. UIST offers high angular resolution imaging and spectroscopy and has been designed to exploit the best performance of the UK IR Telescope. In addition to imaging with 0.24arcsec and 0.12arcsec resolution, long-slit and cross-dispersed spectroscopy, UIST has an integral field mode using a reflective image slicer. An image rotator allows the slits and the rectangular field of view of the IFU to be oriented at any position angle on the sky. The UIST optical design relies on refractive optics with the spectroscopy provided by both replicated and direct-ruled grisms. The lenses are mounted in mechanical modules which also contain the mechanisms, such as the filter and slit wheels. The integration of the opto-mechanical system will be discussed. The high tolerances on positioning the optical components to be used under cryogenic conditions are achieved by mechanical alignment on an optical bench which is an integral part of the instrument. Initial tests of the cryogenic performance of the optics will be presented. The UIST detector is an 1024 by 1024 InSb 'ALADDIN' array from the Raytheon IR Center of Excellence. The array controller is modified from 'EDICT', a VME processor based system which was developed at the UK ATC to control the mid-IR arrays used in the MICHELLE spectrometer. Progress on the integration of the UIST detector and controller will be discussed.

The 10⁵ resolution spectrograph 'CRIRES' for the 1-5 micrometers wavelength range is under construction for ESO's Very Large Telescope. The Nasmyth-mounted instrument sues an adaptive optics front-end for light concentration and 0.2 arcsec spatial resolution in the main long-slit spectroscopy mode. Three detector arrays will be used for large single- order wavelength coverage. Many components for this large cryogenic spectrograph are based on developments for earlier ESO IR instruments.

IRIS2 will provide direct imaging and low dispersion spectroscopy on the Anglo-Australian Telescope (AAT) using a 1K by 1K HgCdTe array and be capable of a future upgrade - by installing a new camera - to use a 2K by 2K array. It will be used primarily with the f/8 telescope configuration but can also be used with f/15 and f/36 configurations. The optics will be entirely transmissive with all spherical surfaces. The collimator will have four elements and the camera, five. With a 50mm collimated beam diameter and an f/2.2 camera, the 1K array will provide a field 7.6 arcmin square with pixels equivalent to 0.45 arcsec. If a 2K array is fitted later, an f/4 camera will give an 8.0 arcmin square field with pixels 0.24 arcsec square. Zemax was used to optimize the design as a multi-configuration system, so that a balance was maintained between direct imaging and spectroscopic performance and between the three atmospheric windows. For direct imaging with the f/2.2 camera, a diffraction based calculation indicates the energy inside a circle inscribed within one pixel is always better than 80 percent of that for a diffraction limited system. Particular care was taken to provide good imaging of the telescope pupil onto the cold stop in K.

LUCIFER-MOS is a liquid nitrogen cooled near IR multi object spectrograph imaging 20 freely selectable sub-fields of about 2.2 inch by 1.8 inch and 6 by 4 image elements each on the entrance slit of the LUCIFER spectrograph. The image elements are re-arranged by 480 fused silica fibers of 50 micrometers core diameter and 100 micrometers total diameter with integrated, hexagonal lenslets of 0.6 mm width corresponding to a 0.3 inch field. The pre-optics magnifies the telescope image by a factor 3.3, thus adapting the telescope plate scale to the lenslet scale, and additionally providing a cold stop. The post-optics converts the f/3 fiber output beam to the f/15 beam accepted by the spectrograph. Each of the 20 6 by 4 fiber arrays together with its pre-optics is mounted in a spider arm which can be freely positioned within the 200 mm diameter field of view by a cryogenic robot. The robot performs three rotational movements to position the spider arms and is driven by cold stepper motors. The spider arms are locked in their positions by two permanent magnets each. Their magnetic field can be compensated by coils to unlock the arms and move them across the field of view.

We present the preliminary conceptual design of a Mosaic IR Camera and Multi-Object Spectrography (MIRCAMOS) for the Canada-France-Hawaii Telescope. The instrument houses 4 Hawaii-2 2048 by 2048 HgCdTe detectors sensitive between 0.8 and 2.5 micrometers . The optics is all reflective, featuring a warm corrector with fast tip/tilt capability and 4 cryogenic optical trains. The pixel scale is 0.20 inch/pixel yielding a field of view of 13.7 feet by 13.7 feet. Z, J, H or K band spectroscopy at R approximately 1500 is obtained with a single grating setting. A cryogenic slit wheel unit featuring several positions for multi-object custom masks is mounted within a separate cryostat designed to be thermally cycled within 8 hours for rapid exchange of MOS masks. Each mask can hold up to approximately 300 slitlets distributed over a FOV of 7 feet by 13.7 feet. MIRCAMOS is very competitive compared with similar instruments planned for 8- 10 m telescopes.

Design concept of the fiber multi-object spectrograph (FMOS) for Subaru Telescope together with innovative ideas of optical and structural components is presented. Main features are; i) wide field coverage of 30 arcmin in diameter, ii) 400 target multiplicity, iii) 0.9 to 1.8 micrometers near-IR wavelengths, and iv) OH-airglow suppression capability. The instrument is proposed to be built under the Japan-UK-Australia international collaboration scheme.

In this paper, we present the status of VISIR, the mid-IR instrument to be installed in 2001 at the Cassegrain focus of YEPUN, the telescope unit number 4 of the European Very Large Telescope program. This cryogenic instrument, optimized for both mid-IR atmospheric windows, combines imaging capabilities over a field up to about 1 arcmin at the diffraction limit of the telescope, and long-slit grating spectroscopy capabilities with various spectral resolution up to 25000 at 10 micrometers at 20 micrometers . The contrast to design and build VISIR was signed in November 1996 between the European Southern Observatory and French-Dutch consortium of institutes led by Service d'Astrophysique of Commissariat a l'Energie Atomique. A key step in the project has been passed in 1999: the final design review. The instrument is now in the manufacture phase. Several subsystems have already been built and tested. The integration of the whole instrument is scheduled to start in December 2000.

TIMMI2 is a focal reducer with variable magnification using a reflective collimator and various camera lenses from Silicon, CdTe, Germanium and KRS-5. The primary operating wavelength is 8-24 microns with limited access also to the 3-5 micron region. Longslit and Echelle spectroscopy up to a resolving power of 1000 are implemented with grisms. A cryogenic wire grid polarizer allows for imaging polarimetry. TIMMI2 uses a 240 X 320 As:Si detector array and is cooled by a 2 stage Gifford-McMahon cooler. Maximum field is 72 X 96 arcsec. TIMMI2 has 5 internal cryogenic functions and one external wheel holding calibration targets. TIMMI2 will be interfaced to the 3.6m telescope with a special IR adapter allowing wheel holding registration of the IR images with respect to astrometric reference frames. The instrument design, electronics and results from laboratory test will be presented. Sensitivity estimates as well as an outlook on possible astronomical programs will be given.

In this paper, we present the design and test performance of the COMICS, the mid-IR instrument for the 8.2 m Subaru Telescope at Mauna Kea. The instrument has both imaging and long slit grating spectroscopy capabilities in the 8-26 micrometers wavelength range. In the camera section, there are selectable three sets of lens assembly, one for the 10 micrometers imaging, another for the 20 micrometers imaging, and the other for the 10 micrometers pupil imaging. This camera section has an SBRC 320 X 240 Si:As IBC array and serves as a slit viewer and as a camera pixel scale of 0.130 arcsec. The spectrograph section is designed to have fiber SBRC 320 X 240 Si:As IBC arrays. Five arrays will cover 8-13 micrometers wavelength range in two positions of the grating with resolving power around 2500. So far, two arrays are installed for the spectrograph section and full spectral region is covered with tilting the grating. Selectable four sets of gratings provide spectral resolution ranging from 250 to 10000 in the N band and around 2500 in the Q band.

We describe a near-IR coronagraphic camera built for use with the Subaru 8.2-m telescope and its adaptive optics system. The purpose of this instrument CIAO is to obtain high-resolution images of faint objects in close vicinity of bright objects at IR wavelengths. Such a desire is strong in astronomy, especially in the study of companion brown dwarfs and extra-solar planets, circumstellar disks around both young stellar objects and main-sequence stars, jets and outflows form both young stellar objects and main-sequence stars, jets and outflows from both young stars and evolved stars, circumnuclear regions around AGNs, and host galaxies of QSOs. CIAO is a 1-5 micron camera with tow focal plate scales: 22 milli-arcsec/pixel and 11 milli-arcsec/pixel. The camera is equipped with the standard broad-band filters as well as a number of narrow-band filters. Choice of masks, filters and camera lenses and optical alignment with collimator and detector are made with cryogenic motors. CIAO utilizes one ALLADIN II 'science'-grade detector array manufactured by SBRC. Occulting masks whose diameter ranges from 0.1 to 3 arcsec and several types of pupil masks are selectable, all cooled down to about 60 K and the detector is cooled to about 30 K. Also available are a R <EQ 1000 grism with coronagraphic slits and a polarimetric module. We also present preliminary results from the first commissioning run at the Subaru telescope.

LIRIS is a near-IR intermediate resolution spectrograph with added capabilities for multi-object, imaging, coronography, and polarimetry. This instrument is now being constructed at the IAC, and upon complexion will be installed on the 4.2m William Herschel Telescope at the Observatorio del Roque de Los Muchachos. The optical system uses lenses and is based on a classical collimator/camera design. Grisms are used as the dispersion elements. The plate scale matches the median seeing at the ORM. The detector is a Hawaii 1024 X 1024 HgCdTe array operating at 60K.

The GEMINI Multiobject Spectrograph (GMOS), due for delivery in late 2000, will include a powerful integral field spectroscopic capability. The instrument scan switch to this mode by the remote insertion of an integral field unit (IFU) into the focal plane in place of multiobject masks. The initial implementation of the GMOS IFU will cover a field in excess of 50 square arcsec with a sampling of 0.2 arcsec via 1500 spatial elements with spectra covering up to 3000 pixels. The spectrum length may also be doubled by halving the field. A separate field is provided at fixed offset to facilitate accurate background subtraction. The system employs a fiber-lenslet technique that provides significant benefits over unlensed fiber reformatters and fiberless lenslet arrays. The specific advantages are unit filling factor, high throughput and long spectra. The IFU has been designed in the light of our experience with two other successful devices of this type. We summarize the design of the device and discuss how the IFU will be operated within the context of GMOS and the GEMINI telescopes. Finally, we present options for implementing IFUs with finer spatial resolution on GMOS.

The Gemini Near IR Spectrograph (GNIRS) currently under development at NOAO and scheduled for delivery in the summer of 2002, will include a powerful and innovative Integral Field Spectroscopy (IFS) capability. The design, integration and test of the GNIRS Integral Field Unit (IFU) are the responsibility of the University of Durham's Astronomical Instrumentation Group. The Critical Design Review is scheduled during the second quarter of the year 2000. Its design is based on the Advanced Image Slicer concept developed as a result of research conducted under the auspices of the Durham Instrumentation R and D Program. A slicer-based system has many advantages over fiber-based designs, especially for cryogenic instruments. The GNIRS IFU consists of two self-contained modules mounted inside the GNIRS slit slide mechanism. This slide mechanism is employed to select the required spectroscopy mode by sliding the respective module into the instrument's optical path. The low resolution option provides a field of view of 3.2 inch X 4.4 inch with a sampling resolution of 0.15 inch over 625 spatial elements and a spectrum length of 1024 pixels, whereas the high resolution optic provides a field of view of 1.0 inch by 1.5 inch with a sampling resolution of 0.04 inch over 972 spatial elements and a spectrum length of 1024 pixels. This paper gives an overview of the IFUs optical design, which has been optimized to take full advantage of the excellent image quality provided by the Gemini telescopes, and the mechanical design.

We describe CIRPASS which is currently being completed at the Institute of Astronomy in Cambridge. The optical design is presented and contrasted with more conventional software- suppression designs. The limiting magnitudes on Gemini are expected to be J 22.6 and H 21.7 in a 3 hour exposure.

Recently integral field spectroscopy has become a band-wagon among the optical astronomical community and most of the 8m class telescopes plan to offer this as part of their instrumentation package. The possibility of expanding integral field spectrsopcy into the near IR opens exciting new possibilities for the study of a variety of astronomical objects.

We present her the opto-mechanical design of an image slicer with reference to designs for a deployable IFU for UIST and two mid-IR IFUs for NGST. Particular attention is paid to how the design achieves a number of goals required of an IFU working in an astronomical IR instrument.

The current state-of-the-art in large format near-IR detector array fabrication is discussed. Near-IR detector arrays of 1024 by 1024 pixels format, both in HgCdTe and InSb, are now in use in many astronomical instruments and are performing very well. The next step in detector array size has just been taken and Rockwell has produced the first science-grade 2048 by 2048 HgCdTe array. Further advances in performance are required for medium and high resolution spectroscopy and for future space missions, in particular the next generation space telescope. Technologies are being developed that will allow us to overcome most of the performance limitations of today's near-IR detector arrays.

Rockwell Space Center is developing low-noise visible and IR imaging sensors and systems for astronomy, high-end commercial, NASA, and advanced military applications. The first science grade 2048 by 2048 HAWAII-2 focal plane array (FPA) for astronomy was recently demonstrated for the SWIR waveband. Science-grade deliveries to the University of Hawaii's Institute for Astronomy, the European Southern Observatory and the Subaru Telescope, among others, will soon start. MWIR/visible 2048 by 2048 HAWAII-2 arrays are also being developed for the NGST program using our process for removing the CdZnTe substrate from the back-side illuminated HgCdTe FPAs to detect visible radiation in addition to IR. Previously, more than 25 science grade 2.5micrometers 1024 by 1024 HAWAII FPAs were delivered for use in many observatories; these typically exhibit < 0.1 e-/s dark current and < 10 e- read noise after correlated double sampling at temperatures above 60K. 1024 by 1024 FPAs development is also continuing; dark current < 1 e-/s has been measured at 140K for a NIR 1024 by 1024 HAWAII array. In a related effort, development of high frame rate, low noise FPAs has begun for wavefront sensing including adaptive optical systems for both visible and NIR/SWIR bands. Hybrid Visible Silicon Imager development is also continuing, expanding the success achieved with prior 640 by 480 FPAs. We are now demonstrating 1024 by 1024 arrays with 0.3-1.05 micrometers response. The silicon detectors in HyViSI FPAs are independently processed on silicon wafers and mated to the same multiplexers fabricated originally for interface to HgCdTe detectors. HyViSI FPA quantum efficiency is > 90 percent with near-100 percent fill factor, and the dark current is negligible with minimum cooling. Our near-term plan to develop 4096 by 4096 visible and IR FPAs will also be discussed.

Hg:Cd:Te grown by Molecular Beam Epitaxy onto a lattice matched Cd:Zn:Te substrate with Double Layer Planar Heterostructure architecture holds promise of extremely low, near theoretical dark current out to wavelengths beyond 5 micrometers while eliminating the persistent images and reduced short-wave quantum efficiency disadvantages of the liquid phase epitaxial (LPE) material now in widespread astronomical use. We report on the characterization of two Rockwell FPAs consisting of (lambda) c approximately 4.75 micrometers MBE material hybridized to 1K by 1K HAWAII multiplexers within the context of establishing their performance relative to the stringent focal plane goals for the Next Generation Space Telescope (NGST). The effort has concentrated primarily on characterizing total FPA noise at selected temperatures from 30 to 90 K although short wave quantum efficiency and image persistence have also been measured. The test procedures are based on the use of both DC and variance nosie techniques and, at T approximately 60K, have allowed characterization of the total noise as the sum of separate contributions due to dark current, read noise and read charge. The test facility and characterization techniques are described and results, which verify the remarkable potential of this material to exceed the NGST performance goals at temperatures both at, and also significantly higher than, the approximate 30K now anticipated for the NGST NIR focal pane, are presented.

The first VLT IR instrument, ISAAC, was installed at the 8 meter Antu telescope in 1998. Experience and results with both InSb and HgCdTe large format arrays will be reported. Effects limiting the performance and strategies to partially overcome these limitations will be discussed.

Quantum Well IR Photodetectors (QWIPs) afford greater flexibility than the usual extrinsically doped semiconductor IR detectors. The wavelength of the peak response and cutoff can be continuously tailored over a range wide enough to enable light detection at any wavelength range between 6-20 micrometers . The spectral band width of these detectors can be tuned from narrow to wide allowing various applications. Also, QWIP device parameters can be optimized to achieve extremely high performance sat lower operating temperatures due to exponential suppression of dark current. Furthermore, QWIPs offer low cost per pixel and highly uniform large format focal plane arrays (FPAs) mainly due to mature GaAs/AlGaAs growth and processing technologies. The other advantages of GaAs/AlGaAs based QWIPs are higher yield, lower 1/f noise and radiation hardness. Recently, we operated an IR camera with a 256 by 256 QWIP array sensitive at 8.5 micrometers at the prime focus of the 5-m Hale telescope, obtaining the images. The remarkable noise stability - and low 1/f noise - of QWIP focal plane arrays enable camera to operate by modulating the optical signal with a nod period up to100 s. A 500 s observation on dark sky renders a flat image with little indication of the low spatial frequency structures associated with imperfect sky subtraction or detector drifts.

The ongoing development of large IR array detectors has enabled wide field, deep surveys to be undertaken. There are, however, a number of challenges in building an IR instrument which has both excellent optical quality and high sensitivity over a wide field. We discuss these problems in the context of building a wide field imaging camera for the 3.5m telescope at Calar Alto with the new 2K by 2K HgCdTe HAWAII-2 focal plane array. Our final design is a prime focus camera with a 15 feet field-of-view, called Omega 2000. To achieve excellent optical quality over the whole field, we have had to dispense with the reimaging optics and cold Lyot stop. We show that creative baffling schemes, including the use of undersized baffles, can compensate for the lost K band sensitivity. A moving baffle will be employed in Ogema 2000 to allow full transmission in the non-thermal J and H bands.

The search for galaxies at redshifts > becomes increasingly difficult in the visible since most of the light emitted by these objects is redshifted into the near IR. The recent development of high-performance near IR arrays has made it practical to built a wide field survey instrument for operation in the near IR part of the spectrum. CIRSI, the Cambridge IR Survey Instrument, uses four of the Hawaii-1 MCT arrays each of which has 1024 by 1024 pixels. This paper describes a number of the novel feature of CIRSI and summarizes the present performance achieved by CIRSI and the scientific programs it is principally engaged in.

This paper describes an ambitious new wide field IR camera for the 3.8m UK IR Telescope (UKIRT), located on Mauna Kea, Hawaii. The camera, currently under design at the UK Astronomy Technology Center, will include 4 2048 by 2048 pixel focal plane array IR detectors operating over a wavelength range of 1-2.5 micrometers . The optics provide a 1 degree diameter corrected field of view and a pixel scale of 0.4 arcsec per pixel. A novel Schmnidt type optical design allows the large field to be imaged with excellent image quality. The optical design includes a cold stop to maximize rejection of background radiation and stray light. Precise microstepping will be used to improve sampling. Four parallel data acquisition and processing channels will be used to cope with the large data rates expected. It is envisaged that a substantial fraction of UKIRT time will be devoted to large area sky surveys once WFCAM is operational, resulting in a unique IR catalogue containing hundreds of millions of objects.

The NIRI for the Gemini North telescope is now undergoing acceptance testing. NIRI is the main near-IR facility camera on the Gemini North telescope and is designed to fully exploit the excellent characteristics of the s