Instrumentation packages for the new generation of 8 m. class Telescopes aim at delivering global observational capabilities in the most effective way: (1) technically, with 'perfect' individual instruments, (2) observationally with extensive covering of the parameter space and most of the instruments tailored to specific classes of astronomical targets, (3) operationally with reliable, automatic, fast systems linked with an 'adapted-observing' scheduler, and finally, (4) scientifically with flexible observing strategies and target finding capabilities. The ESO VLT with four Unit Telescopes and twelve foci, all but one equipped with a (quasi) permanently mounted instrument, offers a good ground for this challenging task.

We present the current design of the VIsible Multi-Object Spectrograph (VIMOS) and the Near InfraRed Multi-Object Spectrograph (NIRMOS) for the European Southern Observatory Very Large Telescope. The basic scientific requirement is to conduct very deep redshift surveys of large quantities of objects, in a minimum number of nights. The technical specifications are to allow for a large multiplex gain over a wide field, and a high efficiency of the optical train, over the 0.37 to 1.8 micrometer domain. The baseline technical concept is built around 4 channels, covering 4 X 7 X 8 arcmin² for VIMOS and 4 X 7 X 7 arcmin² for NIRMOS. Each channel is an imaging spectrograph with a large field adaptation lens, a collimator, grisms or filters, and a F/1.8 camera, coupled to a 2048 X 4096 pixels CCD for VIMOS, and a 2048² HgCdTe Rockwell array for NIRMOS. The unique multiplex gain allows to obtain spectra of up to 840 object simultaneously with VIMOS, and up to 170 with NIRMOS (10 arcsec slits). An integral field spectroscopy mode with more than 6400 fibers coupled to micro-lenses will be available for VIMOS, covering a 1 X 1 arcmin² field. The VLT-VIRMOS survey of more than 150,000 galaxies is planned down to magnitudes I_AB equals 24, coupled to an ultra deep probe to I_AB equals 26.

FORS1 (FOcal Reducer/low-dispersion Spectrograph) is an all dioptric focal reducer designed for direct imaging, low- dispersion multi-object spectroscopy, imaging polarimetry and spectro-polarimetry of faint objects. Two identical copies of the instrument (FORS 1 and 2) are being built by a consortium of three astronomical institutes (Landessternwarte Heidelberg and the University Observatories of Gottingen and Munich) under contract and in cooperation with ESO. FORS 1 and 2 will be installed, respectively, in 1998 and 2000 at the Cassegrain foci of the ESO VLT unit telescopes nos. 1 and 2. For the tests of FORS in Europe, a telescope and star simulator was built, which allows to incline and rotate the whole instrument and to simulate stars in the field of view at various seeing conditions. FORS 1 was integrated at the telescope simulator and saw its 'first light' in the integration facility in November 1996. Since then the electro-mechanical functions, the image motion due to flexure, the calibration units, the optical performance and the instrument software were tested and optimized. This paper presents a summary of the procedure and the results of the tests.

An Australian consortium of astronomers and engineers (based at AAO, MSSSO and UNSW) were contracted by the European Southern Observatory to carry out a one-year concept design study for a near-infrared multi-object spectrograph for the VLT. The primary scientific motivation for this instrument was the detection and study of high-redshift galaxies. The scope of the study included the elucidation of the main science drivers and the development of a VLT instrumentation strategy best suited to those goals. The underlying instrumental philosophy was to supply a significant object multiplex at a high enough spectral resolution to resolve the internal kinematics of galaxies. This science-driven goal also permitted digital OH sky-suppression, yielding better S/N and spectral coverage than at lower resolutions. A full contiguous wavelength coverage from 0.9 micrometer to 1.8 micrometer is achieved through the use of multiple HgCdTe-based spectrograph cameras. A preliminary optical design for the spectrographs has been achieved as has a detailed concept design for the 400-fiber positioner and multiple integral field units. With these capabilities, the proposed instrument is highly effective both for statistical studies of large numbers of object and detailed studies of individual objects. In addition, the development of a novel fiber optic switching facility permits simultaneous wavelength coverage over the entire optical and near-infrared windows (from 0.45 micrometer to 1.8 micrometer) by the use of dichroics and additional CCD- based spectrographs. The very broad range of science that can be carried out with AUSTRALIS spans fields as diverse as galaxy evolution and large-scale structure, the detection of primeval galaxies, the spatially-resolved kinematics of nearby AGN and star-forming regions, globular cluster dynamics and follow-ups to all-sky surveys.

The Echellette Spectrograph and Imager (ESI) for Keck 2 is a versatile Cassegrain instrument which will take data in 3 independent modes. In the echellette mode, it is a medium- dispersion prism cross-dispersed spectrometer featuring a 20- arcsec slit height, 11.5 km/sec/pixel average resolution and full coverage of the (0.39 to 1.09)-micron spectral range in a single exposure. In the prism-only mode, it is a low- dispersion multi-slit spectrometer which covers a 2.0-arcmin- wide field area with an 8.0-arcmin height perpendicular to dispersion. Prismatic resolution is roughly linear with wavelength, ranging from about 62 km/sec/pixel at 0.39 microns to 285 km/sec/pixel at 0.80 microns. In direct-imaging mode, the aforementioned 16.0 sq arcmin field area is reimaged directly unto the CCD detector at a resolution of 0.153 arcsec/pixel. ESI contains an on-axis reflecting collimator which accommodates an off-axis field of view. Cross dispersion is provided by an Ohara BSL7Y prism used in double-pass, followed by a second prism of the same material used in single-pass. The camera is a 10-element all-spherical Epps lens which services a single flat (2048 by 4096 by 15-micron) CCD. The same camera and detector are used for all 3 operating modes without modification. The ESI mechanical design is based upon the 'space-frame' concept which was used successfully for the Keck telescope(s) mechanical structure(s). This results in large weight reduction relative to more typical Cassegrain spectrographs, with the added expectations of very high stiffness and sub-pixel image stability during long exposures. ESI is funded by a grant from CARA and the project has been under way for about 27 months. Most of the mechanical design work is finished and construction is in progress. Electronics, data reduction and user-interface software are nearing completion. All of the optics (including coatings) have been completed and delivered. A thinned science-grade MIT/Lincoln Laboratory CCD for ESI has also been delivered. It is anticipated the ESI may be operational toward the end of 1998.

The Echellette Spectrograph and Imager (ESI), currently being developed for use at the Cassegrain focus of the Keck II 10-m telescope, employs two large (25 kg) prisms for cross dispersion. In order to maintain optical stability in the spectroscopic modes, these prisms must maintain a fixed angle relative to the nominal spectrograph optical axis under a variety of flexural and thermal loads. In this paper, we describe a novel concept for mounting large prisms that has been developed to address this issue. Analytical and finite element analyses (FEA) of the mounts are presented. Optical and mechanical tests are also described.

DEIMOS is a dual beam, off axis, multi object spectrograph of medium resolution being designed for the Keck II Telescope on Mauna Kea in Hawaii. The difficult an advanced scientific goals of the DEIMOS project have generated many challenging design requirements. The DEIMOS team at Lick Observatory has been responding to these challenges with new and unique concepts in instrument design and fabrication. This paper is an update to the paper presented at the SPIE conference in Landskrona, Sweden in 1996.

This paper summarizes the optical, mechanical, electrical, and software design of LRIS-B, the blue channel of the Keck Low Resolution and Imaging Spectrograph. The LRIS-B project will shortly be completing the existing LRIS instrument through the addition of dichroic beamsplitters, grisms to disperse light on the blue channel, broad-band u, B, and V photometric filters, a blue and near-UV transmitting camera lens, and a large format blue-sensitive CCD detector. LRIS-B will also introduce piezoelectric xy-actuation of the CCD detector inside its Dewar, in order to compensate for flexure in the existing instrument; ultimately the red-side CCD detector will be similarly equipped, its PZT xy-stage being independently programmed. The optical design of the LRIS-B camera uses only fused silica and calcium fluoride elements, and includes a decentered meniscus element to compensate for coma introduced by the LRIS off-axis paraboloid collimator. The design of the blue channel grisms have been optimized for maximum blaze efficiency, the highest dispersion grism having a groove density of 1200 gr/mm. Optical elements not in use at any given time will be stowed in carousels externally mounted to the instrument sidewalls. The entire instrument is designed to permit remote operation.

As the only two optical instruments appearing in its first fleet of instrumentation, the GEMINI MultiObject Spectrograph (GMOS) are indeed being developed as workhorse instruments. One GMOS will be located at each of the GEMINI telescopes to perform: (1) exquisite direct imaging, (2) 5.5 arcminute longslit spectroscopy, (3) up to 600 object multislit spectroscopy, and (4) about 2000 element integral field spectroscopy. The GMOSs are the only GEMINI instrumentation duplicated at both telescopes. The UK and Canadian GMOS team successfully completed their critical design review in February 1997. They are now well into the fabrication phase, and will soon approach integration of the first instrument. The first GMOS is scheduled to be delivered to Mauna Kea in the fall of '99 and the second to Cerro Pachon one year later. In this paper, we will look at how a few of the more interesting details of the final GMOS design help meet its demanding scientific requirements. These include its transmissive optical design and mask handling mechanisms. We will also discuss our plans for the mask handling process in GEMINI's queue scheduled environment, from the taking of direct images through to the use of masks on the telescope. Finally, we present the status of fabrication and integration work to date.

The field of integral field spectroscopy is approaching maturity with a large range of devices built or under construction. The integral field unit (IFU) of the GEMINI Multiobject Spectrographs (GMOS) is currently being developed by the Durham Astronomical Instrumentation Group (AIG). This gives 0.2 arcsec sampling over a field exceeding 60 square arcseconds. A prototype has been built for the near-infrared which is now in operation at UKIRT and another instrument using fiber-lenslet technology is under construction for use with Adaptive Optics systems at the Herschel telescope in La Palma. The next generation of integral field spectrographs will address yet more ambitious targets which arise from opportunities presented by the new ground-based and space- borne telescopes. Foremost among these is the desire to (1) obtain finer spatial resolution so as to exploit the excellent images available from AO and space telescopes; (2) exploit the thermal infrared, in order to understand the evolution of galaxies at very high redshift; and (3) to provide a multiplexed integral field capability. Realization of these requirements will, most likely, involve a combination of fiber-lenslet and image-slicer technology.

An international partnership of the USA, UK, Canada, Chile, Argentina and Brazil are projecting two high performance 8- meter Aperture Optical/Infrared Telescopes to be commissioned between the years 1998 and 2000 at Mauna Kea, Hawaii, and Cerro Pachon, Chile. The aim is to exploit the best natural observing conditions to undertake a broad range of astronomical research programs within the national communities of other countries. Zeiss was awarded two mayor contracts within this project, to design, manufacture and test the secondary mirrors and the Gemini Acquisition and Guiding system (GAG) for the 8m telescopes. The current paper refers to the GAG. The Gemini Acquisition and Guiding Unit are complex integrated systems, including optics, mechanical assemblies, detectors and control and signal processing software. The project is to be accomplished under subcontract and in cooperation with the Royal Greenwich Observatory in the United Kingdom. We briefly review the technological and astronomical background of the project and the technical specification of the GAG. The paper describes the design of the Acquisition and Guidance Unit, highlights some design criteria and discusses the expected performance. A report about the actual status of work is given.

The Inamori Magellan Areal Camera and Spectrograph (IMACS) will be one of three first-generation instruments for the Magellan 6.5 m telescopes. It will be installed at the f/11 (Gregorian) Nasmyth focus. This instrument drove the specification and design of the f/11 configuration, which it uses to feed an all-spherical, wide-field collimator. The combination of the Gregorian secondary and refracting collimator lead to 0.2 arc-sec images over a 17 arc-min field with an f/2.66 camera, and 0.4 arc-sec images over a 27 arc- min field with an f/1.49 camera. This paper describes the preliminary specifications for the multiple spectrographic and imaging modes, the optical layout of the instrument and Epps cameras, and strategies for the design and fabrication of the instrument.

Binospec is a wide-field, multi-object optical spectrograph to be used at the f/5 focus of the converted 6.5 m Multiple Mirror Telescope. Its dual beams will address adjacent 8' X 15' fields of view, yielding a total slit length of 30'. Binospec will offer approximately 1 - 6 Angstrom resolution at wavelengths between 0.39 and 1.0 micrometer with a 200 mm collimated beam diameter. Although it is difficult to design an f/5 wide-field collimator, f/5 optics are compact, allowing a small and stiff instrument structure. Binospec uses refractive optics throughout; the collimator contains three lens groups and the camera contains four lens groups. Three aspheric surfaces are used: two in the collimator and one in the camera. A pair of 2048 by 4608 pixel CCD detectors are used for each beam, yielding a sampling of 0.22' per pixel. Binospec's innovative optical design allows excellent image quality. Including the contribution of the MMT optics with the f/5 wide-field corrector, the RMS image diameter at Binospec's focal plane is 18 micrometer (1.3 pixels) averaged over field angles and colors.

The Hectochelle will be a fiber-fed, multi-object spectrograph for the post-conversion MMT which will take 255 simultaneous spectra at a resolution of 32,000 - 40,000. The absolute efficiency, including optical fiber losses, is predicted to be 6% - 10%, depending on the position of a line within a diffractive order. In one hour, features with 60 mangstrom should be resolved in m_R equals 18 stars with a signal to noise of 10.

We report on first astronomical results with a cross-dispersed optical echelle spectrograph fed by a near single-mode fiber. We also present on a novel design of a new adaptive optics (AO) optimized fiber-fed cross-dispersed echelle spectrograph. The spectrograph is designed to match with AO corrected images in the optical bands provided by such as the Mt. Wilson 100 inch, Starfire Optical Range 3.5 m AO telescopes. Ultimately, it will be installed at the 6.5 m MMT, when this has high resolution AO correcting the optical spectrum. The spectrograph, fed by a 10 micron fused silica fiber, is unique in that the entire spectrum from 0.4 micron to 1.0 micron will be almost completely covered at resolution 200,000 in one exposure. The detector is a 2k X 4k AR coated back illuminated CCD with 15 micron pixel size. The close order spacing allowed by the sharp AO image makes the full cover possible. A 250 X 125 mm² Milton Roy R2 echelle grating with 23.2 grooves mm^-1 and a blaze angle of 63.5 deg provides main dispersion. A double pass BK7 prism with 21 deg wedge angle provides cross dispersion, covering the spectrum from order 193 to 77. The spectrograph is used in the quasi- Littrow configuration with an off-axis Maksutov collimator/camera. The fiber feeds the AO corrected beams from the telescope Cassegrain focus to the spectrograph, which is set up on an optical bench. The spectrograph will be used mainly to study line profiles of solar type stars, to explore problems of indirect detection of planets and also study interstellar medium, circumstellar medium and metal abundance and isotopic ratios of extremely metal-poor stars.

This short contribution presents the demanding requirements of scientific cases for ultra-high stability spectrography, from the study of subtle radial velocity changes induced by asteroseismology to the ones produced by extra-solar planetary companions. The analysis of physical conditions in cool interstellar clouds is presented as a typical application of ultra-high resolution spectrography. The main technical challenges associated with such instruments are outlined, including focal stations and light feeding methods. As a possible way to combine both modes in a single instrument, we describe a design case for the ultra-stable high resolution spectrograph that we are proposing for the Gemini south telescope, specified to detect radial velocity variations down to one meter per second and also to achieve near diffraction- limited spectral resolution approaching one million. This versatile instrument would occupy an insulated room in the pier of the telescope and would be fed initially by an optical fiber coming from the Cassegrain focal station.

The atmospheric refraction affects the position of objects in the sky in two ways: a chromatic effect and a field differential effect. The former is about the same in the field and can be, in principle, corrected using an Atmospheric Dispersion Compensator (ADC). The latter is dependent of the field size and cannot be corrected. For spectroscopy in wide fields, both effects have to be carefully considered because they affect the spectrophotometry in terms of signal to noise ratio and spectral distortions. The present study aims to evaluate the atmospheric effects in the case of the VIMOS instrument for the ESO VLT. It is shown that provided a careful operational mode the distortion of spectra can be kept at a level less than 15% with reasonable constraints.

The basic concept of a new type of versatile light feed using a microlens array and optical fibers configured as a fiber pupil slicer is described in comparison with the existing image slicers. The fiber pupil slicers will open up a wide variety of applications in feeding the images of point sources produced at telescope focus into the entrance slit of the spectrographs by transforming the beam shape as well as the focal ratio. We describe two concrete examples for applying the fiber pupil slicers on the instrumentation of the 8.2 m Subaru Telescope. (1) A new feature provided with these pupil slicers is the realization of multi-object spectroscopic mode for the High Dispersion Spectrograph with cross dispersed echelle format by squeezing the seeing disk in the narrow inter-order spacing at the expense of spectral resolution. (2) Another more obvious application is to improve the throughput and the resolving power of a given spectrograph.

This paper describes a high dispersion spectrograph (HDS) now under construction towards the first-light phase of Subaru Telescope. It is located at a Nasmyth focus. HDS is an echelle spectrograph with grating as a post-disperser. The collimated beam size is 272 mm, and the echelle is 300 mm by 840 mm in total size (31.6 gr/mm, R-2.8). The overall throughput (resolution X slit width product) achieved is 38,000 arcseconds. HDS has two cross-dispersers (400 gr/mm, 250 gr/mm) which are optimized for blue- and red-wavelength regions, respectively. The camera is of catadioptric type system, consisting of three correctors and a mirror. It spans the entire chromatic range from 0.30 to 2.0 micrometer. It delivers 10-micrometer images on average within a flat 60 mm- diameter focal plane area in the wavelength range from 0.3 to 1.1 micrometer, without refocusing. This image quality corresponds to typical limiting spectral resolutions well above 300,000 though the resolution will generally be limited to less than this by the entrance slit width and finite CCD pixel sizes. The detector will be a 2 X 1 mosaic of 2k X 4k CCD's with 15-micrometer pixels. Typical limiting spectral resolution will be 100,000.

The Hobby-Eberly Telescope (HET) is a revolutionary large telescope of 9.2 meter aperture, located in West Texas at McDonald Observatory. First light was obtained on December 11, 1996. The start of scientific operations is expected in the late summer of 1998. The Low Resolution Spectrograph [LRS, an international collaboration between the University of Texas at Austin (UT), the Instituto de Astronomia de la Universidad Nacional Autonoma de Mexico (IAUNAM), Stanford University, Ludwig-Maximillians-Universitat, Munich (USM), and Georg- August-Universitat, Gottingen (USG)] is a high throughput, imaging spectrograph which rides on the HET tracker at prime focus. The LRS will be the first HET facility instrument. The remote location and the tight space and weight constraints make the LRS a challenging instrument, built on a limited budget. The optics were partially constructed in Mexico at IAUNAM, the mechanics in Germany, and the camera and CCD system in Texas. The LRS is a grism spectrograph with three modes of operation: imaging, longslit, and multi-object. The field of view of the HET is 4 arcmin in diameter, and the LRS will have a 13-slitlet Multi Object Spectroscopy (MOS) unit covering this field. The MOS unit is based on miniature components and is remotely configurable under computer control. Resolving powers between R equals (lambda) /(Delta) (lambda) approximately 600 and 3000 with a 1 arcsecond wide slit will be achieved with a variety of grisms, of which two can be carried by the instrument at any one time. The CCD is a Ford Aerospace 3072 X 1024 device with 15 micrometer pixels, and the image scale is approximately 0.25 arcsec per pixel. Here we present a detailed description of the LRS, and provide an overview of the optical and mechanical aspects of its design (which are discussed in detail elsewhere in these proceedings). Fabrication, assembly, and testing of the LRS will be completed by mid 1998. First light for the LRS on the HET is expected in the summer of 1998.

A high-resolution spectrograph (HRS) is taking shape for the Hobby-Eberly Telescope (HET). HRS will be mounted on a Newport bench in an insulated chamber in the 'basement' of the HET building, internal to the pier, and will be linked to the corrected prime focus of the HET through its Fiber Instrument Feed. Spectra will be recorded on a 4096 X 4096 Orbit CCD mosaic at resolving powers of 30,000 to 120,000, using a mosaic of two R-4 echelle gratings replicated on a single blank at Spectronic Instruments and a selection of two cross- dispersing gratings. This paper will describe the design and report on the current construction status, and will describe innovative aspects, in particular a new all-refractive camera designed by Epps. The HRS design incorporates Barranne's white-pupil concept as adapted for ESO's VLT by Delabre and Dekker. Predicted performance from ray-trace analysis will be demonstrated. With the 316-gr/mm cross-dispenser the full 420 - 1,100 nm spectral range can be recorded in two exposures at any resolving power up to 120,000. Greater order separation can be had with the 600-gr/mm grating, requiring four exposures if the full range is needed. Designed for remote, queue-scheduled operation, HRS will be capable of rapid response to discoveries of transient phenomena. It will be directed toward discoveries of extra-solar planetary systems and studies of long-period variable stars and absorption line systems in QSOs as well as stellar nucleosynthesis and chemical compositions.

The Medium Resolution Spectrograph (MRS) is a high throughput, versatile, fiber-fed echelle spectrograph for the Hobby-Eberly Telescope (HET). This instrument is designed for a wide range of scientific investigations; it includes single-fiber inputs for the study of point-like sources, synthetic slits of fibers for long slit spectroscopy, multi-fiber inputs for multi- object spectroscopy, and an optical fiber integral field unit. The MRS will have resolution settings between 3500 less than (lambda) /(Delta) (lambda) less than 21000 and will consist of two beams. The initial, visible wavelength beam will have wavelength coverage from 450 - 900 nm in a single exposure. This beam will also have capability in the ranges 390 - 450 and 900 - 950 nm by altering the angles of the echelle and/or cross-disperser gratings. Later, a second beam operating in the near-infrared will be added which will have coverage of 950 - 1300 nm in a single exposure and capability out to 1800 nm. The HET Fiber Instrument Feed (FIF) is mounted at the focal plane of the telescope and positions the fibers feeding the MRS and the High Resolution Spectrograph (HRS). The unique and economical design of the FIF enables the HET's versatility in performing a wide range of scientific investigations with the telescope operating in a queue-scheduled mode.

Back illuminated CCDs have been the detectors of choice for most astronomical imagers and spectrographs during the past decade. In recent years, we have developed processes to improve the performance of these detectors. Recent work has resulted in improved absolute quantum efficiency (QE), QE stability with temperature, QE stability with environmental contamination, and enhanced near-IR response. We demonstrate that QE near 100% can be achieved which is stable against hydrogen, dewar outgassing, and water contamination. We show that QE decrease with temperature can be eliminated for blue/visible optimized CCDs using a backside passivation layer, and significantly reduced for UV optimized CCDs which require very thin backside films. We also show that a 20% QE increase at 900 nm can be obtained by coating the frontside of back illuminated CCDs with a reflective metal film, without increasing interference fringing.

Within the last few years several manufacturers have been producing the 'next generation' of scientific CCDs. These devices have small pixels (approximately 15 micrometers), high UV and broad-band spectral response (greater than 80%), very low readout noise (less than 4 e^- rms), large format (2048*4096) and close butting capability. We present examples of recent data taken on the WHT (at the Roque de los Muchachos Observatory, La Palma) obtained from one such device -- the EEV CCD42 array. The detector has been used for spectroscopy and direct imaging with excellent results. Design and performance details, as well as various special operational modes will be discussed. This device has been adopted for scientific imaging on the Gemini telescopes, as well as several other major observatories -- and so these first operational results should demonstrate the power of these new sensors. Variants of the CCD42 design are now being made to yield slightly different architectures and packaging options. We will compare predicted with actual performance, and discuss characteristics and applications of this new sensor.

We have entered an era of large CCD mosaic camera construction with many observatories developing large mosaic focal planes for wide field cameras and spectrographs. In this review, we outline the history of CCD mosaic development, describe the current state of the art while illustrating the many projects underway, and attempt to peer into the future.

Megacam is a 36 CCD mosaic camera that will cover a 24' X 24' field of view at the f/5 wide-field focus of the converted 6.5 m Multiple Mirror Telescope. The mosaic is a 9 X 4 array of thinned 2048 X 4068 pixel CCDs with 13.5 micrometer pixels. The CCDs are dual-output EEV devices in a custom package to allow the devices to be closely butted on all four sides. The dewar will be mounted to a 2 m diameter assembly that contains the filter wheels (for 30 X 30 cm filters) and the shutter. Telescope guiding will be accomplished with two additional CCDs mounted at the edges of the focal plane. The guider CCDs will be operated slightly defocused, one on either side of focus, to allow simultaneous focusing and guiding. Guide stars will be selected by reading out the full guider frame, after which only a small area surrounding the guide star will be read out. Our simulations show that the defocused guide star images will also be useful for low order wavefront sensing, allowing corrections to the telescope collimation. We are developing a new CCD controller capable of reading the full Megacam in 24 seconds. This controller will also be used to operate the guide chips.

NOAO is testing Scientific Imaging Technologies, Inc. (SITe) thinned backside 2k by 4k charge coupled devices (CCDs) to be used at KPNO and CTIO. NOAO's Mosaics will use as the basic 'tile' the SITe ST-002A CCD Imager. These CCDs will be used to upgrade the wide field Mosaic imager now in use at KPNO's 4- Meter Mayall telescope and 0.9-Meter telescope. SITe 2k by 4k CCDs will also be used in MOSAIC II now under construction for CTIO's 4-Meter Blanco telescope. Additionally NOAO will implement 2 Mini-Mosaics in the two device 4K by 4K configuration. The first of the ST-002As arrived in mid August 97 and this paper will discuss test results of all devices received and tested prior to publication. Additionally this paper will discuss the mounting methods of Mosaics. This will include; geometric stability, techniques used for measuring CCD surface flatness and co-planer requirements for NOAO's Mosaic Instruments.

The NOAO Mosaic CCD Camera consists of 8 CCDs producing an 8K X 8K format. The Mosaic Data Handling System (MDHS) receives data packets from the detector system and broadcasts them on a message bus. A data capture agent (DCA) receives the data and formats it into a set of distributed shared images (DSI). The images are displayed as they are received by a real-time display (RTD) and are recorded to disk as multiextension FITS (MEF) files by the DCA. The DCA triggers a data reduction agent (DRA) to do standard pipeline processing and archiving with a graphical user interface (GUI) for user interaction. The MDHS provides a suite of IRAF data reduction tools for the user and DRA to apply to the MEF mosaic data. The tools include quick-look analysis of the data in the RTD, basic CCD calibration, mosaic reconstruction, and combining of dithered observations into a fully populated (gaps removed) image.

New generation astronomical CCD controllers are being required to operate a variety of CCDs in a range of ground-based applications. These include simultaneous readout from two or four corners of the same CCD (multiple readout), operation of several CCDs in the same focal plane (mosaics), fast readout of small devices for wavefront sensing in adaptive optics systems, readout of only a small region or number of regions of a single CCD (sub-image or region of interest readout), merging the charge from neighboring pixels before readout (binning), continuous readout of devices for drift scan observations, and low contrast polarimetric or spectroscopic differential imaging. Most astronomical applications require that the controller electronics not contribute significantly to the readout noise of the CCD, that the dynamic range of the CCD be fully sampled, that the CCD be read out as quickly as possible from one or more readout channels, and that some flexibility in readout modes and device format exist. A further requirement imposed by some institutions is that a single controller design be used for all their CCD instruments to minimize maintenance and development efforts. The Generation II controller design recently completed at San Diego State University to address these requirements is reviewed. A user-programmable digital signal processor (DSP) operating as a sequencer and communications processor is combined with 12-bit digital-to-analog converters for setting all CCD voltages, a video processor chain with speeds of up to 1 MHz, 16-bit analog-to-digital converters, and a bussed backplane architecture for incorporating the control and readout of multiple CCDs by replicating the clock driver and video processing elements.

This paper describes two important concepts employed by ESO's high-speed low-noise FIERA CCD controller: (1) a distributed sequencer that allows for an unlimited number of high-speed control signals, and (2) a video preprocessing chain which is based on the clamp and sample method. The reasoning for both concepts is described in detail, and the clamp-sample approach is quantitatively compared to the more common dual-slope video processing technique. Furthermore, we describe aspects of the clamp and sample approach we implemented to avoid image artifacts. First results with FIERA are presented.

We present the characteristics of the new CCD imager, SUSI2, installed at the ESO 3.5 m NTT. The instrument shares the Nasmyth focus A with the new infrared imager-spectrograph SOFI. The focal plane array of USSI2 is a mosaic of 2 EEV44- 82, 2k X 4k, 15 micrometer pixels, thinned, anti-reflection coated CCDs, which are placed at the direct focus of the telescope (scale 0.08 arcsec/pixel, field of view 5.5 X 5.5 arcmin). The average QE for the two devices is 76, 90, 85, 80, 68, 49, 23% at 350, 400, 500, 600, 700, 800, 900 nm respectively. The overall instrument efficiency, including the three mirrors of the telescope and the detector but without filters, is computed to be 46, 55, 51, and 48% at the central wavelengths of the U, B, V and R bands. The CCDs are driven by the new ESO CCD controller FIERA. The system performance was measured during the commissioning of the instrument at the telescope in February 98. The mosaic is read in 16 seconds in the standard operating mode (2 X 2 binning of the CCDs) with a read-out-noise of 4.7 e^-/pixel. The other CCD parameters such as CTE, dark current and linearity, were also found to comply with the requirements. The FWHM of stellar sources in images obtained in good seeing conditions were measured to be 0.49 arcsec, with no significant variation over the field of view.

MEGACAM is the next generation wide-field camera to be used at the prime focus of the 3.6 m CFHT telescope. This instrument, which will cover a full 1 square degree, is designed around a mosaic of 36 to 40 2K X 4K CCDs. such a large detector requires new approaches for the hardware as well as the software, and will have major impacts on the telescope structure, optics and operations. We present in this paper several novel ideas and techniques that will be implemented as part of this challenging project.

The Big Throughput Camera (BTC) recently celebrated its first anniversary as a user instrument on the Blanco 4-meter telescope at Cerro Tololo Interamerican Observatory (CTIO), where it collects more photons per second than any other nighttime astronomical camera in the world. We offer a look at the successes and lessons learned during the first year of operation. After an overview of the hardware, we describe the software from the user's point of view, and then offer examples of the observing targets and strategies used. BTC has become very popular among CTIO observers -- more than one- third of dark time is now assigned to BTC -- but the large field of view leads to some new data reduction challenges which we discuss in the final section.

The Quasar Equatorial Survey Team (QUEST) collaboration has built and commissioned a large CCD-based astronomical camera covering most of the focal pale of the 1 meter Schmidt telescope at the Llano del Hato National Observatory (ONLH) in Venezuela. The camera consists of a 4 X 4 mosaic of CCDs. The CCDs are a standard astronomical design with front- illuminated (coated) 2048 X 2048 15 micrometer pixels. Using a novel position adjustment technique and precision CCD alignment, we are able to driftscan within 6 degrees of the celestial equator. Using a combination of UBV filters and an objective prism, we expect to discover quasars in unprecedented numbers and with well understood, quantifiable systematic effects. Such a quasar sample can be used to study cosmological questions -- large-scale structure of the universe, distribution of dark matter via gravitational lensing and Hubble's constant via gravitational lensing time delays -- as well as many other topics. Preliminary commissioning results will be presented along with preliminary quasar identification results. A second-generation camera (403 Mpixels, back-illuminated 4 X 24 mosaic with 1024 X 4096 7.5 micrometer pixel CCDs) is being designed.

We describe the design and performance of a dichroic-mirror camera (DMC) which can take 15 narrow-band images simultaneously. We separate the wavelength range of 390 - 950 nm into 15 narrow bands with 14 dichroic mirrors. The detector of DMC is a mosaic CCD camera which has 15 CCDs (TI TC-215). When we put DMC to the MAGNUM 2-m (F/9) telescope being built at Haleakala, Hawaii, the field of view becomes about 4.5 arcmin in diameter. The design of optics shows that we can get an image size of about 0.13 arcsec r.m.s. or better (without atmosphere), though we use only two different kind of lenses (the camera lens and the collimator lens). The system throughput of DMC as a function of wavelength is quantitatively estimated. Simulations using spectra of galaxies ad Qnd QSOs show that DMC can get a signal-to-noise (S/N) of approximately greater than 5/band/object for galaxies (I_AB equals 22) and QSOs (I_AB equals 23) in the images of 30 min - 1 hour exposure taken with a 2-m telescope. Future prospects for possible enhancements and applications of dichroic-mirror system are also discussed. Having a DMC with resolution of about 30 would be very adequate for high redshift supernovae search. To get higher resolution, DMC combined with Fabry-Perot's is an interesting possibility.

LOTIS is a rapidly slewing wide-field-of-viewtelescope which was designed and constructed to search for simultaneous gamma- ray burst (GRB) optical counterparts. This experiment requires a rapidly slewing (less than 10 sec), wide-field-of-view (greater than 15 degrees celsius), automatic and dedicated telescope. LOTIS utilizes commercial tele-photo lenses and custom 2048 X 2048 CCD cameras to view a 17.6 X 17.6 degree field of view. It can point to any part of the sky within 5 sec and is fully automated. It is connected via Internet socket to the GRB coordinate distribution network which analyzes telemetry from the satellite and delivers GRB coordinate information in real-time. LOTIS started routine operation in Oct. 1996. In the idle time between GRB triggers, LOTIS systematically surveys the entire available sky every night for new optical transients. This paper will describe the system design and performance.

A description is given of the fiber based Thousand Element Integral Field Unit (TEIFU) that is being built by Durham University for use with the WYFFOS fiber spectrograph and the ELECTRA Adaptive Optics system at the William Herschel Telescope. With TEIFU there will be at least one thousand spatial elements with selectable sampling scales of 0.125, 0.25 and 0.5 arcsec corresponding to field areas of 14, 56 and 222 square arcsec, the field can be divided into 2 to facilitate background subtraction. The first two scales of 0.125 and 0.25 arcsec are designed to take advantage of the improved image quality provided by the ELECTRA AO system. The 0.5 arcsec sampling is designed for use without ELECTRA and samples the uncorrected image quality provided by the WHT. In this mode there is a larger field that is not constrained by the guide star requirement of the ELECTRA AO system. The operating wavelength of the system will be approximately 500 - 1000 nm.

PMAS, the Potsdam Multiaperture Spectrophotometer, is a new integral field spectrograph currently under development at the Astrophysical Institute Potsdam (AIP). The design is optimized for linear and stable behavior in order to allow for 2D spectrophotometry which is expected to become an important new observing technique at 8 - 10 m class telescopes.

One of the most exciting operational modes of the new generation of 8-m class telescopes will be integral field spectroscopy. Many of the current designs of integral field units use microlens arrays as a way of sampling the focal plane of the telescope. Integral field spectroscopy places demanding requirements on the optical quality of the microlens arrays. In this paper we describe these requirements and report on an extensive series of optical tests to characterize microlens arrays in the laboratory. The tests were carried out on prototype microlens arrays intended for use in various Durham instrument projects. The lenses tested had excellent positional accuracy and consistency of focal length. However detailed measurements of the point spread function showed a core with good image quality but with a background level of diffuse scattered light. This scattered light causes a throughput loss in the instrument of around approximately 20%. This result was verified for the Spectroscopic Multimode InfraRed Fiber System (SMIRFS) integral field unit by absolute throughput tests carried out during commissioning of the instrument at the United Kingdom Infrared Telescope.

INTEGRAL is an optical fiber unit for performing 2D spectroscopy of extended objects at the 4.2 m. William Herschel Telescope (WHT). It is mounted at the GHRIL Nasmyth focus together with newly built acquisition, guiding, and calibration units. It makes use of the specially designed fiber spectrograph WYFFOS. This system allows up to six bundles to be mounted simultaneously. It currently contains three science oriented fiber bundles, any one of which can be easily and quickly placed in the telescope beam. Their spatial resolution elements (fiber core diameters) are 0'.45, 0'.9, and 2'.7, respectively. Hence, depending on the prevailing seeing conditions the instrument can be easily optimized for the scientific program. INTEGRAL was successfully commissioned at the WHT during a six night period in July 1997. Here we will discuss its main characteristics.

The 2dF project aims to provide the Anglo-Australian Telescope with the ability to perform multi-object fiber spectroscopy of 400 objects within a 2 degree field. The 2dF project has now reached a stage where telescope time is being allocated to use this new facility on a regular basis for science observations. We will review the current status of the 2dF project in terms of the performance and operation of the instrument hardware and software. We will describe some of the wide variety of scientific projects attempted during the first year of shared risks observing and commissioning.

Multi-object spectroscopy at the Anglo-Australian Observatory's 1.2-m UK Schmidt Telescope (UKST) is carried out with the FLAIR multi-fiber system. The FLAIR front-end feeds an optically-efficient, all-Schmidt spectrograph mounted on the dome floor. However, positioning of the 92 available fibers within the 40 sq. deg. field of the telescope is essentially a manual operation, and can take from four to six hours. Typical observations of sufficient signal-to-noise usually take much less than this (e.g. about an hour for galaxy redshifts to B approximately 17). Clearly, therefore, the system is working at well under its potential efficiency for survey-type observations where repeated reconfigurations of fibers are required. To address the imbalance between reconfiguration time and observing time, a fully-automated, off telescope, pick-place fiber-positioning system known as 6 dF has been proposed. It will allow 150 fibers to be reconfigured across a 6-degree circular field in under an hour. Three field plates will be available with a 10 - 15 minute field-plate changeover anticipated. The resulting factor of 10 improvement in observing efficiency will deliver, for the first time, an effective means of tackling major, full-hemisphere, spectroscopic surveys. An all southern sky near-infrared-selected galaxy redshift survey is one high- priority example. The estimated cost of 6 dF is $A450k. A design study has been completed and substantial funding is already in place to build the instrument over a two-year timescale.

FEROS is a new fiber-fed bench-mounted prism crossdispersed echelle spectrograph for the ESO 1.52-m telescope at the European Southern Observatory (ESO) in La Silla, Chile. It works with a 79 lines/mm R2 echelle grating in quasi-Littrow mode and in white pupil configuration. With two fibers of 100 micrometer core diameter for the object and the nearby sky, the complete optical spectrum from 370 - 860 nm is recorded in one single exposure on a 2k X 4k thinned CCD with 15 micrometer pixels. Therefore, the instrument can work in a fixed configuration on the optical bench without movable parts besides the CCD shutter mechanics. For the highest-possible opto-mechanical stability. FEROS will be housed in a temperature and humidity controlled room in the former Coude room of the telescope. The resolving power of R equals 48,000 is reached by the use of a newly designed two-slice image slicer which is fed by the two fibers. Alternatively, the sky fiber can be illuminated with a calibration lamp during the whole object exposure to monitor the spectrograph's residual motions for high-precision radial-velocity work. FEROS is built for ESO by a consortium of four European astronomical institutes under the leadership of the Landessternwarte Heidelberg, Germany. Further members of the consortium are the Astronomical Observatory Copenhagen, Denmark, the Institut d'Astrophysique de Paris, and the Observatoire de Paris/Meudon, France. It is planned that FEROS will be fully operational at the ESO 1.52-m telescope in December 1998 and will be available to the community in early 1999.

Time-resolved spectroscopic studies of relatively faint stars that vary on many time scales, such as interacting binary stars, benefit from a particular combination of wide wavelength coverage, medium spectral resolution, and high efficiency. We describe two similar fiber-fed echelle spectrograph designs that address this need, with an emphasis on eventual use for automated, unattended spectroscopy. The coverage is approximately 385 - 850 nm in about 25 echelle orders or approximately 350 - 500 nm in about 16 orders, each with a resolution of about 5000. Cross dispersion is supplied by two prisms in series and the collimator is an off-axis paraboloid. One of the designs incorporates white pupil optics to minimize camera vignetting in the red echelle orders.

We are building the second version of the Kyoto Tridimensional Spectrograph (Ohtani et al., this symposium). This will be mounted on the MAGNUM, a 2-m telescope under construction at Haleakala, and also on the SUBARU. 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 new spectrograph is significantly better than the first version in several ways. The IFS has as many as 37 X 37 microlenses, each of which subtends 0' .39 at the MAGNUM. The optics is designed to be used in wide wavelength ranges from 360 nm to 900 nm. The transmission at any wavelength between 370 and 900 nm is designed to exceed 50% for the collimator plus camera system, and to reach almost 40% even at 360 nm. In order to achieve high efficiency at short wavelengths, we use an anti- reflection coated backside-illuminated 2K X 2K CCD. We are also planning a further improvement by using multi-layer anti- reflection coatings for lenses, in collaboration with National Astronomical Observatory, Japan. In order to assure good image quality under a severe weight limit of 150 kg for this instrument, we have carried out mechanical design by calculating the flexure of the instrument for all telescope attitudes with finite element analysis, and succeeded in limiting the maximum flexure to 30 micrometer. This does not degrade image quality. The movements on the CCD of the light from the center of the focal plane have also been simulated, depending on the telescope attitudes. This is important to obtain not only a good image, but also a correct flat field and wavelength calibration in the IFS mode. The movements are expected to be confined almost within one pixel for an attitude, which is considered to be small enough.

This paper describes the operating parameters and initial design of a new spectrograph proposed for the 200-inch Hale Telescope at Palomar Observatory. The instrument, whose working name is THuCIDIDES (Two Hundred-inch Cassegrain Image- Deblurred Interchangeable-Disperser/Echelle Spectrograph), will feature high system efficiency and multiple modes of operation, including low- and intermediate-resolution long slit and multi-slit capability over 12.5 X 3 arcmin fields, and a cross-dispersed echelle mode covering 3800 - 8500 angstrom at R equals 20,000 (with a 1.2 arcsecond slit) up to R equals 60,000 (with an image slicer). A 4096 X 4096 pixel CCD will serve as the detector. The quasi-Littrow echelle configuration and use of a prism cross-disperser will result in high system efficiency, estimated at approximately equals 14%. The compact design will permit mounting in the Cassegrain ring plane, to reduce susceptibility to flexure. An optional fast-guiding tilt mirror provides modest improvement to seeing FWHM and slit throughput.

The technique of spectroastrometry (measuring the wavelength dependence of the position of an object) provides a means of studying the spatial structure of astronomical sources on scales much smaller than the seeing disk size or the diffraction limit. Despite successful demonstrations in the past, the method does not seem to be widely known. This paper describes techniques used at the Anglo-Australian Telescope to obtain such observations with standard instrumentation, and presents some examples of astronomical results on sources including binary stars and active galactic nuclei. The potential for combining spectroastrometry with interferometry to observe structure on microarcsecond scales is described.

Progress in Active Optics Methods has led to the invention and production of blazed aspherical gratings. These developments use jointly 'vase form' submasters and a two-stage replication technique. It has been shown that the use of aspherized gratings greatly minimizes the number of optical surfaces. This improves the optical throughput of astronomical spectrographs and has a capability of correcting camera mirror aberrations up to f/1.2. With respect to refractive designs, the full achromaticity in correcting mirror aberrations by constant line spacing reflective gratings allows much broader spectral coverages -- hereafter [(lambda) (lambda) ] approximately equals 2 octaves. In addition, and also due to a full reflective design, such instruments provide quasi- constant spectral dispersions and are distortion free. These latter features increase the accuracy in the data reduction process (sky substraction, etc. ...), and are particularly convenient in the multi-aperture mode. Recent developments in this field are presented with imager-spectrograph ISARD, dedicated to the Cassegrain focus of the 2m Bernard Lyot Telescope at Pic-du-Midi Observatory for faint object studies in the optical domain [320 - 1200 nm], and with spectrograph OSIRIS, to be launched in a ODIN orbital mission in 1998 and built by the Canadian Space Agency for studies in the spectral range [295 - 800 nm].

The Taurus Tunable Filter (TTF) is a pair of tunable narrowband interference filters covering 3700 - 6500 Angstrom (blue 'arm') and 6500 - 9600 Angstrom (red 'arm'). The TTF offers monochromatic imaging at the Cassegrain foci of both the Anglo-Australian and William Herschel Telescopes, with an adjustable passband of between 6 and 60 Angstrom. In addition, frequency switching with the TTF can be synchronized with movement of charge (charge shuffling) on the CCD which has important applications to many astrophysical problems. Here we review the different modes of TTF and suggest their use for narrowband imaging.

A diffraction grating technology based upon volume-phase holograms shows promise of enhanced performance for many applications in astronomical spectroscopy over classical surface-relief grating technology. We present a discussion of the underlying physics of a volume-phase grating, give some theoretical performance characteristics, present performance data for a real volume-phase grating, and discuss some potential applications for this grating technology.

Birefringent, nematic liquid crystals (LC) have been laminated between the substrates of several Fabry-Perot etalons. The application of an electric field allows the effective index of refraction of the LC to be varied. A polymer alignment layer is used to align the crystals perpendicular to the optical axis of the Fabry-Perot etalon. An oscillating electric field is used to rotate the crystal around the optical axis of the etalon, effectively changing the index of refraction. This change in index is used to tune the Fabry-Perot etalon in a manner similar to traditional pressure and mechanical tuning systems. However, the approach described here has the advantage of producing a solid-state etalon that is tunable without use of a bulky pressure system or environmentally sensitive piezo-electric stacks. Several etalons have been constructed to further develop this technology. Clear apertures greater than 2' have been achieved, and a hybrid spacer technique has been developed to allow for etalons with spacings of up to 1 cm. Fabry-Perot partial reflective coatings capable of operation from the visible (400 nm) to the near infrared NIR (2.2 micrometer) are also in use.

Megacam is a wide-field optical imager for the converted MMT that uses thirty-six 2048 by 4608 pixel CCDs to cover a 24' X 24' format. We describe a computer architecture designed to accommodate the expected data volume and show benchmark results from prototype implementations that demonstrate the performance attained by each of the design decisions. We show that our time budget allotments can be met using a modular, scalable architecture design that exploits the natural parallelization of multiple, identical detector components.

We briefly describe the design and the characteristics of the Italian Panoramic Monochromator installed at the focal plane of the THEMIS telescope built in Izana by a joint venture of the French and Italian National Research Councils. The Panoramic Monochromator substantially is a narrow band filter (approximately equals 22 mAngstrom bandwidth) tunable on the visible spectrum for quasi simultaneous bidimensional spectrometry of the solar atmosphere. The narrow bandwidth is obtained by using a non standard birefringent filter and a Fabry Perot interferometer mounted in series. This assembly has the advantage of the spectral purity of one channel of the Fabry Perot interferometer and a very large free spectral range. Moreover the spectral stability depends on the interferometer, the environment of which may be carefully controlled. The design of this instrument is not really new, but, only now it has been possible to build it thanks to the development of servo controlled Fabry Perot interferometers, which are stable in time and may easily be tuned. The system seems to perform well. It is stable in wavelength and the spectral pass band and stray light are within the expected values, as it may be deduced by very preliminary tests performed at the THEMIS Telescope and in Arcetri (Firenze) at the 'G. B. Donati' solar tower.

One of the most critical issues in designing a spectrograph is the motion of opto-mechanical components due to flexure especially when it will be mounted to the Cassegrain focus of a telescope. Image motion on the detector has to be kept small in order not to affect the value of the scientific data. The FORS spectrographs fulfil those requirements by a proper design and by a passive compensation of the instrumental flexure. Image motion of the 2 metric tons instrument could be reduced in this way to a tiny fraction of one pixel's size thus not affecting the data gathered with those spectrographs. It is tested and approved at a telescope simulator that all specifications regarding those motions are fully met. A fine tuning flexure compensation is built into the spectrograph's design and is tested on its tuning range which allows to adapt the compensation to effects eventually caused by the Cassegrain flange of the telescope.

We present the preliminary study of a low cost, high performance spectrograph for the VLT, for observations in the V, R and I bands. This spectrograph is meant for intermediate (R equals 16,000) resolution spectroscopy of faint (sky and/or detector limited) sources, with particular emphasis on the study of solar-type (F-G) stars belonging to the nearest galaxies and to distant (or highly reddened) galactic clusters. The spectrograph is designed to use the adaptive optics (AO) systems at the VLT Telescope. Even if these AO systems will not provide diffraction limited images in the V, R and I bands, the photon concentration will still be above approximately 60% of the flux in an 0.3 arcsecond aperture for typical Paranal conditions. This makes the construction of a compact, cheap and efficient echelle spectrograph possible. AVES will outperform comparable non adaptive optic instruments by more than one magnitude for sky- and/or detector-limited observations, and it will be very suitable for observations in crowded fields.

We describe the techniques that we have used to mount large optics in three wide-field instruments for the converted MMT: the wide-field corrector uses to provide a 1 degree diameter field at the f/5 focus of the converted MMT, the Hectospec bench spectrograph fed by 300 optical fibers and the wide- field dual-beam Binospec spectrograph. These optics are primarily refractive elements with diameters between 0.2 and 0.8 m that must be mounted from their edges, although we also describe mounts for two large mirrors in the Hectospec bench spectrograph. Both the wide-field corrector and Binospec mounts must perform under varying gravity loads: the corrector is fixed to the converted MMT's primary mirror cell and is tilted from zenith to horizon while Binospec is mounted at the converted MMT's Cassegrain focus. Furthermore, the optics mounts for both instruments must fit within tight space constraints. The Hectospec spectrograph is mounted in the MMT's rotating building and experiences a constant gravity vector. In all cases, the mounts must perform over a wide temperature range, -20 to 20 degrees Celsius, so the issue of differential thermal expansion between the mounts and optics must be carefully considered. As a result, the mounts we discuss include either RTV elastomeric or flexural elements.

The Hectospec consists of a robotic positioner that will position 300 optical fibers at the f/5 focus of the converted MMT and a bench mounted moderate-dispersion spectrograph. Hectospec will be the first wide-field instrument to be used at the converted MMT and is now under construction at the Smithsonian Astrophysical Observatory. Commissioning at the converted MMT is scheduled for mid 1999, shortly after first light at the f/5 focus. The innovative features of the instrument are described, emphasizing recent developments.

Sometimes Fabry-Perot, and other narrow-band filters, are used for astronomical imaging in the so-called telecentric mode. In it the pupil is collimated through the filter, resulting in different incidence angles on the filter for rays coming from different parts of the objective. This results in variations of the central transmission wavelength, which broaden the effective filter bandpath. In addition each wavelength within this filter bandpath sees a different illumiatnon of the pupil when viewed from behind the fitler. This causes the diffraction limited point-spread-function to vary with wavelength. With the advent of diffraction limited imaging using adaptive optics, this can cause complications. In this paper, which elaborates further on research published elsewhere, I examine the magnitude of this effect.

Sol gel coatings are an important alternative to conventional broadband dielectric antireflection (AR) coatings. Originally developed in the early 1940s for AR coating of lens elements, interest in this technology has been rekindled with the advent of high energy laser systems. Fusion laser designs push the damage threshold limits of conventional dielectric films. In addition to these laser applications, sol gel AR coatings are important for broadband applications in the ultraviolet, visible or near infrared. Repeatable and predictable broadband coatings of this type are frequently required astronomical instruments. Lens size does not impose severe limitations for this technology, an important aspect when considering conventional dielectric AR coatings. Sol gel AR coatings typically have a minimum reflectivity per surface on the order of 0.1% and behave as a near perfect quarterwave coating for substrates with refractive indices between 1.38 and 1.60. Sol gel coatings can be removed from glass elements without damage or repolishing of substrates thereby eliminating the risk of a spoiled coating run. These coatings simplify changes for different wavelength requirements. Optical glass, fused silica and many types of crystalline elements are appropriate substrates for sol gel AR coatings. Recent technical development has improved mechanical durability of standard sol gel coatings on many substrates, allowing them to be more easily cleaned. Sol gel coatings have been used successfully in astronomical instruments, and it appears that they may find additional application as their properties become more widely understood.

The advent of new large telescopes and the promise of ever larger CCDs and CCD arrays has spawned the need for spectrograph cameras with unprecedented high-performance characteristics. During the last decade, familiar catadioptric cameras gave way to lenses which offered less obstruction, better transmission, wider spectral coverage and larger field of view. These developments culminated in a series of successful multi-aspheric camera lenses with elements as large as 225 mm in clear diameter such as the LRIS lens for the Keck 1 Low Resolution Imaging Spectrograph. Advances in optical glass technology and in the production of CaF₂ ingots some 400 mm in diameter and 140 mm thick have brought even larger, faster lenses within reach. Optical designs now exist and in some cases construction is under way of a generation of camera and collimator lenses with elements as large as 360 mm in clear diameter and focal lengths as short as (305 to 381) mm. These optics will service a variety of spectrographs with collimated beam diameters in the (150 to 205)-mm range and anamorphic factors in the (1.0 to 1.5) range. Specific optical design examples of aspheric lenses are presented including the DEIMOS camera (for Keck 2), the IMACS short camera (for Magellan 1), the BINOSPEC camera and the BINOSPEC collimator (for converted MMT). All-spherical lenses are presented including the ESI camera (for Keck 2), the HRS camera (for Hobby Eberly) and the IMACS long camera (for Magellan 1).

We show that a crystalline lithium niobate (LiNbO₃ : LN) grism and a hybrid grism made of a LN transmission grating and a zinc sulfide (ZnS) prism, in spite of their birefringent properties, can be used as new and powerful dispersing elements with high refractive indices to realize high spectral resolution for optical to near infrared astronomical spectrograph with transmission optics. We also show new fabricating methods for grisms and immersion gratings with high index material and deep grooves.

Once the optical design for a spectrograph is finalized, a number of tasks remain for the optical designer which largely simplify the engineering, fabrication, and assembly of the instrument. Such tasks include sensitivity analysis for alignment tolerances, flexure tolerances and flexure compensation, distributions of radiation incidence angles for coating design, and thermal analysis for thermal compensation. For the Keck ESI instrument, the entire spectrograph and guiding system optical designs were directly translated into a 3-dimensional AutoCAD^r file, complete with clear apertures, actual traced rays, beam envelopes, stray light, and footprints of the beam paths at the optical surfaces. The mechanical engineers could then design the spectrograph structure in 3-dimensions around the existing optical layout.

Instruments for large telescopes often require cameras with large, deeply-curved, and temperature-sensitive lenses. The instrument error budgets require each lens to be supported so that excellent performance is maintained in the face of gravitational and thermal perturbations. We describe here elastomeric mounts that address these requirements. We first describe the general design principles, the effects of errors in design and fabrication, and the performance under static and dynamic loads. We describe specific examples; the elastomer RTV560 and the lens supports for the camera of the W. M. Keck Observatory DEIMOS spectrograph.

Ground based observations beyond 0.68 microns are severely affected by intense and variable atmospheric OH bands that dominate the night sky spectrum. A number of OH suppression techniques have been proposed in the past using either dispersive or filter based techniques. Dispersive techniques are promising but potentially expensive and complex to implement. Filter based solutions have the advantage of being relatively cheap and easy to incorporate into existing instruments. However, the inflexibility of simple filter designs has made significant improvements in sensitivity difficult to achieve. Rugate filters are thin film filters with refractive indices that very continually through the coating. They enable the production of transmission profiles comprising a series of irregular and sharply defined band passes. In this paper we investigate the potential of rugate filters as OH suppression devices. We demonstrate through numerical simulation that it is possible to achieve almost complete suppression of the OH features in the J photometric band while retaining roughly half the spectral coverage. Such filters can potentially increase the signal to noise ratio for faint continuum sources by a factor of 2 in the J band and allow for considerably longer exposures before the detector saturates on the sky background.

The Hobby Eberly Telescope (HET) is a revolutionary large telescope of 9.2 meter aperture, which is currently undergoing commissioning at McDonald Observatory. First light was obtained on December 11, 1996. Scientific operations are expected in 1998. The Low Resolution Spectrograph (LRS, a collaboration between the University of Texas at Austin, the Instituto de Astronomia de la Universidad Nacional Autonoma de Mexico, Stanford University, Ludwig-Maximillians-Universitat, Munich and Georg-August-Universitat, Gottingen) is a high throughput, imaging spectrograph which rides on the HET tracker at prime focus. The LRS will be the first HET facility instrument. The unique nature of the HET has led to interesting optical design solutions for the LRS, aimed at high performance and simplicity. The LRS is a grism spectrograph with a refractive collimator and a catadioptric f/1.4 camera. The beam size is 140 mm, resulting in resolving powers between (lambda) /(Delta) (lambda) approximately 600 and 3000 with a 1 arcsec wide slit. The LRS optics were designed and partially fabricated at the IAUNAM. We present a description of the LRS specifications and optical design, and describe the manufacturing process.

Each of the two Gemini telescopes will be instrumented with the Gemini Multi-Object Spectrograph (GMOS), a general purpose optical spectrograph mounted at one of the Cassegrain foci. Two GMOS are currently being designed and built by a team of scientists and engineers in Canada and in the UK. A stringent flexure specification is imposed on these instruments by the scientific requirement to measure velocity to high precision, 2 km/s at R equals 5,000 with 0.5 arcsec slits. This implies a basic stability specification of 3.125 micrometer/hour at the detector focal plane. The GMOS design has met this specification by using a combination of stiff structure (where flexure is minimized); Serrurier trusses (where the flexure is controlled); precision mechanisms (where mechanical hysteresis and error are minimized) and, finally, an open-loop active correction system at the detector focal plane (where the CCD is translated to counteract any residual flexure). Once the GMOS design was conceptualized and its component groups were identified, the design team divided the basic stability specification into allowable contribution from each group. The final division was weighted according to the degree of design difficulty, based on inputs from the engineers. An error budget was developed and maintained to ensure that GMOS would meet its overall flexure specification by controlling the contribution from each component. The error budget approach will be described and discussed in the paper. We will also look at examples from the GMOS design with reference to calculations, analyses, FEA and actual measurements from prototype components.

The Hobby-Eberly Telescope (HET) is a revolutionary large telescope of 9.2 meter aperture, located in West Texas at McDonald Observatory. The Low Resolution Spectrograph [LRS, an international collaboration between the University of Texas at Austin (UT), the Instituto de Astronomia de la Universidad Nacional Autonoma de Mexico (IAUNAM), Stanford University, Ludwig-Maximillians-Universitat, Munich (USM), and Georg- August-Universitat, Gottingen (USG)] is a high throughput, imaging grism spectrograph which rides on the HET tracker at prime focus. The remote location and tight space and weight constraints make the LRS a challenging instrument, built on a limited budget. The mechanical design and fabrication were done in Germany, and the camera and CCD system in Texas. The LRS is a grism spectrograph with three modes of operation: imaging, longslit, and multi-object. Here we present a detailed description of the mechanical design of the LRS. Fabrication, assembly and testing of the LRS will be completed by mid 1998. First light for the LRS on the HET is expected in the summer of 1998.

The Echellette Spectrograph and Imager (ESI) is being built at UCO/Lick Observatory for the Cassegrain focus of the Keck II telescope. The collimator mirror is optimally constrained by a space-frame structure. It will be actively moved to provide the focus and flexure (tip and tilt) control for the instrument. Careful attention to space-frame geometry has simplified the mechanical design. Analytical and Finite Element Analysis (FEA) are presented to demonstrate how a simple but very stiff structure is used to provide support, flexure control, and focus.

HROS is the high resolution optical spectrograph which is being constructed by University College London for use on the Gemini South Telescope in Chile. We report here the structural analysis which was performed by CLRC on the proposed design as part of the conceptual design review. The UCL design uses an immersed echelle grating and has a number of very large and hence heavy optical components. Supporting these elements while providing a stable spectrum on the detector is particularly challenging as the instrument is designed to be mounted at the Cassegrain focus of the telescope and will therefore be subject to changing gravity fields during integrations. Our analysis demonstrates that it is possible to design a structure which will limit the bulk of the flexure in the spectrograph to rotation of the major elements with relatively little inherent decentering. This structure lends itself to active flexure compensation. Our analysis suggests that by careful design of the support structure, active control can provide compensation for quite large flexures with little or no degradation to the quality of the spectrum delivered at the detector.

The High Resolution Optical Spectrograph (HROS) will operate at the Cassegrain focus of the Gemini South telescope, at a resolving power of R equals 50,000. It will use an immersed echelle for dispersion, and fused-silica prisms for cross- dispersion. The instrument will weigh two tonnes and will need to meet stringent engineering performance requirements. This paper describes these requirements and the engineering approach used in the HROS design.

The Echellette Spectrograph and Imager (ESI) is one of several second-generation instruments for the Keck telescopes. The motivation for the f/15 Cassegrain-mounted instrument has been to provide a versatile, extremely efficient, and stable system for faint object spectroscopy and imaging, on a comparatively limited schedule and budget. In keeping with these goals, a space-frame instrument structure has been designed, analyzed, and fabricated. The mainframe structure provides the mechanical interface between the telescope and instrument, support points for all the optical, mechanical, and electronic sub-systems, and provides a rigid base for the active- collimator flexure control system. The fundamental concepts and motivation for using a space-frame are discussed, and their application to the design, analysis, and fabrication of the ESI structure is presented.

The S.A.O. Megacam focal plane will consist of 36 large CCDs (2 K X 4.5 K format), each with 2 output ports running at pixel rates up to 200 kHz. The unbinned data field is thus as large as 324 megawords and is presented at a sustained data rate during readout approaching 28 megabytes/sec. We have developed a simple camera controller to deal with the problem of handling so many channels and their digitized outputs at fairly high speeds. We also present some results from ongoing efforts to optimize the 200 kHz signal processing chain for low noise and low crosstalk.

During the ground calibration of the Space Telescope Imaging Spectrograph (STIS) large scattered light haloes were identified in images of point sources and long slit spectral images at long wavelengths (greater than 750 nm). The long wavelength scattering was traced to the SITe 1024 X 1024 CCD and its header package, raising concerns for the performance of the Advanced Camera for Surveys (ACS) CCD detectors. ACS is a third generation axial instrument for the Hubble Space Telescope (HST) and will be installed during the 1999 Servicing Mission. Two of the ACS imaging channels employ SITe CCDs, so the ACS team have conducted a study of the long- wavelength scattering, in collaboration with SITe, to assess the impact to the ACS science program and develop a solution. In this paper we discuss our solution, its implementation on ACS CCDs, and describe the results of initial tests.

A new generation wide-field imager is being developed and will be put into service at sites in North and South America. Driven by the requirement for larger imaging areas and more pixels but limited by manufacturing process constraints, manufacturers are developing 2, 3, and 4-side buttable CCDs that can be tiled to achieve large imaging areas as opposed to developing a single large CCD. NOAO has designed, fabricated, and tested a wide-field imager called Mosaic that tiles 8 CCDs to produce an imaging area slightly greater than 123 mm X 123 mm. Several successful science observation runs have been completed using Mosaic at the KPNO Mayall 4 m and .9 m telescopes. A second Mosaic Wide-Field Imager is presently being manufactured and will be deployed at the CTIO Blanco 4 m telescope early next year. This report will focus on the mechanical design aspects of the Mosaic Imager and the upgrade path to achieve the scientific requirements will be discussed.

Precision CCD photometry over large fields-of-view and detection of faint and diffuse objects demand great care in the acquisition of flat-field frames. Variations in the illuminated field will introduce spatial variability in the flat field. Small angular displacements of the CCD, telescope, and instrument, with respect to the external light source yields a corresponding linear translation of the image on the CCD with respect to the true flat-field function, so that a functional equation relates the flat field to these shifted data sets. In two-dimensions an elliptic partial differential equation relating the offset CCD measurements to the flat- field function can be derived enabling relaxation solution techniques to be used. A simulation is presented and discussed.

There are a number of application areas of CCDs in ground and space based astronomy and earth observation, which could benefit from compact but versatile control electronics. These include visible imaging and spectroscopy, auto-guiding, star tracking and wavefront sensors. We describe here our design of an Application Specific Integrated Circuit (ASIC) which integrates the complete functionality of a dedicated programmable waveform generator and waveform sequencer to provide a CCD controller on a single chip. The ASIC can sustain waveform state changes at the clock rate of up to 40 MHz, and can implement complete sequence changes on a frame by frame basis. The ASIC is designed specifically to provide programmable CCD waveform generation and sequencing, and thus yields significant shrinkage in component count, circuit complexity, PCB circuit size, and power dissipation compared to previous DSP or general purpose microprocessor-based design solutions. More compact and light weight cameras are thus realized without compromising the ability to program any complexity of CCD waveform patterns and waveform sequences such as multiple window and/or pixel binning readout format. A second generation ASIC is to be fabricated on a radiation hard silicon process for use in Space-borne CCD camera systems. Applications of our prototype ASIC will be presented along with future development plans.

This paper discusses a magnetically focused electron bombarded CCD type ultraviolet image sensor, employing an opaque substrate photocathode. The novel feature, reported in this paper, is the unique light weight permanent magnet structure employed in the electron optics to generate a uniform focus field. The approximately 200 Gauss magnetic field of this 'PXL' prototype is relatively uniform over a considerable region. The weight is only 5 kg, compared to 14 kg for permanent magnet structures of conventional design having comparable field strength and uniformity over equivalent volume. The light weight and low volume of this type of magnet assembly makes it especially attractive for space applications where size and weight are critical design factors. The paper will present experimental data showing the spatial resolution achieved with this new magnet assembly when employed in imaging photoelectrons from an opaque substrate photocathode onto an electron bombarded CCD.

The Hectospec is a moderate dispersion spectrograph fed by 300 optical fibers. Hectospec's pair of five-axis robots will position fibers at the 1 degree diameter f/5 focus of the converted MMT, allowing efficient multi-object spectroscopy. We discuss algorithms that we have developed to match the optical fibers to celestial objects and then to compute the appropriate sequence of robotic positioner moves to reconfigure the fibers between successive observations. Both algorithms require essentially no user interaction, consume only modest computer resources and allow effective deployment of the Hectospec's 300 fibers. The target-to-fiber matching algorithm is a recursive procedure which allows simultaneous optimization of the multiple observations that are required to complete a large survey. The robotic motion sequence algorithm allows the two Hectospec robots to work together efficiently to move fibers directly between observing configurations.

Not all realms of observation require an 8-meter telescope. Some, such as space surveillance of asteroids and man-made satellites, are too important to ignore, yet obviously inappropriate to consign to the new generation of large telescopes. The United States Air Force Research Laboratory (AFRL), with Boeing, Rocketdyne Technical Services (RTS), has developed a low-cost, rapidly deployable surveillance telescope concept called Raven which takes advantage of commercial off-the-shelf (COTS) telescopes, detectors and software. The development of the Raven concept was originally a response to a recognized need to support the timely follow- up of asteroid discoveries. Early astrometric tests using Raven telescope in the 12- to 16-inch diameter range proved to be comparable in accuracy to the much larger telescopes of the existing space surveillance network. Observations of man-made satellites have also produced quality results. A high level of productivity is achieved by automating all of the observing functions and much of the data reduction and analysis. Performance data in both the areas of asteroid and satellite metrics will be presented, and performance parameters discussed.

Instrumental problem/Task: Adaptation of large Ritchey- Chretien Telescopes to CCD-detector fields of the 90-s. A solution could be found as a result of a common development of Research Institutes and optical industry. Focal reducer systems for 2.2 m and 3.5 m telescope at Calar Alto, Almeria, Spain. A focal reducer system is a lens system, which is highly corrected and optimized for a wide spectral range (350 nm . . . 1000 nm) and reduces the telescope image field to the size of the CCD-chips. The system complies to the following requirements: Improved polychromatic imaging quality. Well- accessible position of the pupil, with enough space for filter turret, dispersing elements such as grating prisms, Fabry- Perot etalons, etc. Good accessibility of the focus area, with the user having the possibility to insert a detector of his preference. The poster describes the optical design of the focal reducer systems and shows practical application results with the 2.2 m and 3.5 m MPI telescopes. Furthermore the poster reviews the utilization of the same know how for the development of modern imaging spectrographs with high photometric stability. As an example the design for the Potsdam Multiaperture Spectrophotometer PMAS, which is currently under manufacturing in strong co-operation with Astronomical Institute Potsdam/Germany, is described in the poster.

Development of a spectrograph for area spectroscopic observations of faint extended objects in optical spectral region (Ohtani et al., 1994) was completed and the instrument is now in commission at the 188 cm telescope of the Okayama Astrophysical Observatory. With this instrument can be made four different kinds of spectroscopy which are switched to one another by remote operation. The four modes are filter imaging at several narrow (or wide) bands, slit scan in the long-slit spectrograph mode, imaging Fabry-Perot interferometer observation, and integral field spectroscopy (IFS). The first two modes of these are conventional types. Two Fabry-Perot etalons of Queensgate Instruments, resolving power R equals 300 (tunable narrow band filter) and 7000, are available. Prior to development of the spectrograph, characteristics of these etalon were examined in detail by laboratory experiments. Effective finess and nonuniformity were measured. Further, behaviors of drift of transmitting wavelength with variations of ambient temperature was carefully examined. Based on the results of the experiment, temperature of the etalon in the spectrograph at the telescope is stabilized within plus or minus 0.5 degrees Celsius to attain high performance that drift of the transmitting wavelength does not exceed one tenth of the width of the Airy profile during observations. For the IFS mode, the type of the TIGER spectrograph is employed. Our spectrograph has a dual-channel enlarging optics, one channel acquiring a target object and the other an 'uncontaminated' sky field well apart from the target. The microlenses array is shared by the target and the sky. With the 188 cm telescope a 9' X 15' (7 X 11 lenslets corresponding 1'.3 square) target field and a smaller sky field 3'.7 apart are observed simultaneously. Spectra on the CCD partially superpose each other with adjacent spectra because appreciable deviations in the microlenses array format exist from an exact square array. However, energy loss due to truncation of superposed part is not serious. Total quantum efficiency of the system is 2 percent at peak. Some results of astronomical applications for active galaxies are presented. Results of super-wide-field observations with an objective of a short focal length is also demonstrated.

We discuss the design and construction techniques of the Integral Field Units (IFU's) for SPIRAL and COHSI (Cambridge OH Suppression Instrument). The design for both is similar and we explain our reasons for adopting our particular approach. Both IFU's have been used on telescopes and found to perform very well. Finally, our plans for future instruments which will use the same techniques are briefly discussed.

At the Canada-France-Hawaii Telescope (CFHT) recent gains in Spectrographic capability has pointed out the need for a telescope-mounted Spectral Calibration Source that has three features: (1) A precise simulated pupil image, (2) Emission lines from .36 microns to 2.5 microns, (3) High throughput in the infrared. This paper describes the CFHT f/8 focus and then describes the Spectrographs and the Spectral Calibration Source (SCS) that use this focus.

The Kitt Peak Mayall 4-m telescope required a new prime focus corrector having a flat focal plane covering 36 arcmin on a side (51 arcmin diagonal) to accommodate the Mosaic 8K X 8K CCD system. The scientific requirements for the new corrector included atmospheric dispersion compensation (ADC), excellent near-UV efficiency, excellent image quality, and extremely low scattered light and ghosting. The optical system designed to meet these demands exhibits excellent and stable performance through its first year of operation. This paper describes the innovative design and engineering aspects of the corrector. Science verification data are presented to demonstrate some of the attributes of the new corrector.

The Ohio State Instrument Control and IMage ACquisition System, ICIMACS, is the computer hardware and software used by all instruments developed by the Imaging Science Laboratory (ISL) to control the detector, pre-process data, record image data on a separate computer system for data reduction and analysis, generate real time data display, control the mechanisms within an instrument, interface with the telescope controller, connect to a user interface, and perform engineering functions such as temperature or pressure logging. ICIMACS has now been used on 12 different instruments and is herein described as applied to 'MOSAIC' the near IR imager/spectrometer in use on the Kitt Peak 2.1 and 4 meter telescopes and on the MDM 2.4 and 1.3 meter telescopes.

A fiber optic array, called DensePak, has been built for the WIYN 3.5 meter telescope located on Kitt Peak, Arizona. DensePak incorporated the use of existing instruments at WIYN to provide a low-cost option for integral field spectroscopy. The array consists of 91, 310 micrometer diameter fibers bonded into a 7 X 13 rectangle with fibers spaced at 400 micrometer center-to-center. The array provides spatial coverage on the sky of approximately 30 X 45 arc-seconds at the WIYN F/6.4 Nasmyth focus. The fiber bundle feeds the same bench spectrograph as that used for the Hydra multi-object spectrograph which provides many configuration options. DensePak has been used successfully in several scientific programs where mapping of spectral characteristics of extended objects such as galaxies, planetary nebulae, and comets is desired. The options of low to high resolution spectroscopy with DensePak allow two-dimensional spatial measurements of abundances, velocity kinematics, and line ratios to be obtained. Proven success of DensePak at WIYN has led to consideration of future upgrades to expand scientific capability through improved spatial resolution.

The kinematics of the interstellar medium may be studied by means of a scanning Fabry-Perot interferometer (SFPI). This allows the coverage of a wider field of view with higher spatial and spectral resolution than when a high-dispersion classical spectrograph is used. The system called PUMA consists of a focal reducer and a SFPI installed in the 2.1 m telescope of the San Pedro Martir National Astronomical Observatory (SPM), Mexico, in its f/7.5 configuration. It covers a field of view of 10 arcmin providing direct images as well as interferograms which are focused on a 1024 X 1024 Tektronix CCD, covering a wide spectral range. It is considered the integration of other optical elements for further developments. The optomechanical system and the developed software allow exact, remote positioning of all movable parts and control the FPI scanning and data acquisition. The parallelism of the interferometer plates is automatically achieved by a custom method. The PUMA provides spectral resolutions of 0.414 Angstrom and a free spectral range of 19.8 Angstrom. Results of high quality that compete with those obtained by similar systems in bigger telescopes, are presented.

In order to increase the efficiency, reliability, and cut quality of the mask generating YAG laser system, several improvements have been made to the hardware and software. High quality cuts with a low number of passes have been the goals in producing unique multi slit aperture masks. The description of these improvements and their results is presented.

Fiber optics scramblers are use to improve the stability of the shape of the spectral lines along the scientific exposures. The optical efficiency of two prototypes is revised in this paper. Transmission and focal ratio degradation on two new fibers are also presented.

We are building an 8192 X 10240 CCD mosaic camera for 8 m Japanese Telescope (Subaru). The mosaic will consist of 2 X 5 arrays of 3-edge buttable 4096 X 2048 15 micrometer pixel imagers. Although several vendors have just started supplying the type of large format CCD, it is still in the development phase. Therefore, careful characterization and optimizations of individual CCD are critical. We describe the system used to evaluate several kinds of the CCDs. In addition to the CCD characterization, we also present the mechanical design of the mosaic focal plane which is an another issue to realize the large mosaic.

This paper presents the design of an assembly to support the Hectochelle multi-object spectrograph grating. The grating consists of two mosaicked, 420 mm by 320 mm by 74 mm thick, 110 line/mm echelle grating segments, nominally operated at a 72 degree angle of incidence. The fact that the grating consists of 2 segments that must be co-aligned to within 0.27 arcsecs creates a challenging design problem. After much preliminary examination, two support deigns were studied, one utilizing Invar, and the other, selected for development, using a Zerodur support structure. The trade study to select the final design is discussed. The Zerodur design was selected because it maintained higher grating stability throughout the operating temperature range than the Invar design. The Zerodur design concept is discussed in detail, including the results of finite element modeling that shows system feasibility.

The DEIMOS spectrograph, being designed and built by Lick Observatory, for the Keck-II telescope, requires a dewar system that can support a mosaic of eight 2K X 8K science CCDs and two 600 X 1200 CCDs for flexure control. The science CCD must be mounted on the mosaic backplane with a height difference of less than 10 (mu) . Inside the dewar the mosaic must translate in the focus direction plus or minus 1 millimeter and in the direction of the slit by 90 (mu) for flexure control. This system has now completed the design stage, with many of the subsystems having been prototyped, and is now in the later part of the fabrication stage.

2dF is a multi-object instrument mounted at prime focus at the AAT capable of spectroscopic analysis of 400 objects in a single 2 degree field. It also prepares a second 2 degree 400 object field while the first field is being observed. At its heart is a high precision robotic positioner that places individual fiber end magnetic buttons on one of two field plates. The button gripper is carried on orthogonal gantries powered by linear synchronous motors and contains a TV camera which precisely locates backlit buttons to allow placement in user defined locations to 10 (mu) accuracy. Fiducial points on both plates can also be observed by the camera to allow repeated checks on positioning accuracy. Field plates rotate to follow apparent sky rotation. The spectrographs both analyze light from the 200 observing fibers each and back- illuminate the 400 fibers being re-positioned during the observing run. The 2dF fiber position and spectrograph system is a large and complex instrument located at the prime focus of the Anglo Australian Telescope. The mechanical design has departed somewhat from the earlier concepts of Gray et al, but still reflects the audacity of those first ideas. The positioner is capable of positioning 400 fibers on a field plate while another 400 fibers on another plate are observing at the focus of the telescope and feeding the twin spectrographs. When first proposed it must have seemed like ingenuity unfettered by caution. Yet now it works, and works wonderfully well. 2dF is a system which functions as the result of the combined and coordinated efforts of the astronomers, the mechanical designers and tradespeople, the electronic designers, the programmers, the support staff at the telescope, and the manufacturing subcontractors. The mechanical design of the 2dF positioner and spectrographs was carried out by the mechanical engineering staff of the AAO and the majority of the manufacture was carried out in the AAO workshops.

An updated overview of the 7 scientific instruments and 3 baseline instruments under construction for the 8.2-m Subaru telescope is presented with reference to papers addressed during this series of conferences. Somewhat detailed descriptions are given for Faint Object Camera And Spectrograph (FOCAS) and for COoled Mid Infrared Camera and Spectrograph (COMICS) that are not specifically reported in the conferences. The infrastructures for developing these instruments and the scope of scheduling the first light of these instruments are also mentioned.

Computer controlled polishing of large optics has required complex and expensive tooling, which is impractical at the smaller scale. This paper describes the simulation of a simple aspheric polishing regime using a large active lap, and its embodiment in a new generalized aspheric polishing machine for instrumentation optics.

CCID20 CCDs are designed and produced by personnel at MIT/Lincoln Laboratory. The CCDs are thinned, back illuminated, 4096 X 2048 15-micrometer square pixel, three- side buttable devices. Some CCDs have been made on high- resistivity bulk silicon and others have been made on standard resistivity epitaxial silicon. Recently many devices from the first round of production of these CCDs have been tested at the UCO/Lick Observatory Detector Development Laboratory. In this paper we present the results of the measurements of horizontal and vertical charge transfer efficiency, low- temperature dark current, localized charge traps, full well, responsive quantum efficiency, and fringing. We present performance measurements of the on-chip amplifier including measurements of read-out noise, gain and linearity with different bias voltages. Cross-talk between the two on-chip amplifiers is discussed. High resistivity CCDs made by MIT/Lincoln show higher QE and less QE variation at long wavelengths than regular thin CCDs. However, they are subject to additional lateral charge diffusion and cosmic-ray effects. We will give a comparison between the two kinds of CCID20 CCDs. CCID20 CCDs are not MPP devices. It is much more difficult to get high full well, low spurious charge, low dark current and low residual image, simultaneously. We present optimized parallel clocks and a special erasing procedure to help solve these problems. Devices from this first round of CCID20 CCDs exhibit a rectangular pattern of QE variations caused by backside surface treatment problems.

WYFFOS is a bench-mounted spectrograph based on a Baranne white pupil design, located on one of the Nasmyth platforms of the William Herschel Telescope. It is fed by 126 optical fibers which are accurately positioned in the corrected one degree focal plane by the robotic positioner. AUTOFIB-2 at the prime focus of the WHT. We describe two enhancements to the normal use of WYFFOS; the provision of an 'echelle' mode of operation, and the design for a new longer focal length camera.

Image intensification systems have been used for photon counting applications in ground based and space based astronomy for many years. Various charge amplification and image readout techniques have been used, but some of the most successful to date utilize the microchannel plate (MCP) intensifier in a 'Gen II' configuration with an electronic image sensor. The two major competitive solutions for image sensor are the CCD, which reads the optical image from a phosphor, and the electronic image readout, which collects the charge from the MCP directly, replacing the phosphor. We describe the advantages of the electronic charge division image readout and compare its performance with other imaging techniques. For example, intensifiers using charge division readouts have distinct advantages over intensified CCDs for applications where localized count rate capability (point source count rate) and temporal resolution (event time tagging accuracy) are important. We discuss existing and potential astronomical applications for these detector systems. We discuss recent improvements in imaging performance obtained with charge division readouts, comparing the performance of traditional resistive anode readouts with the latest generation of pattern designs including the Vernier anode. We present results showing the imaging performance of the Vernier anode. We describe a photon counting imaging intensifier system for ground and space based applications in astronomy. The intensifier, which is manufactured by Photek Ltd., uses a proximity focused photocathode, microchannel plate intensifier and conductive charge division electronic readout. A variety of pattern designs offer performance ranging up to 4000 X 4000 pixel format, with position resolution down to 10 microns FWHM and sub-microsecond timing accuracy.

A design for a new-concept speckle imaging system is presented. The instrument, now under construction at Rochester Institute of Technology (RIT), will be able to use any large- format scientific-grade CCD as the imager. The high quantum efficiency, low noise, and linear response of current CCDs are attractive characteristics when compared with traditional photon-counting speckle imaging systems. The RIT system consists of an optics package, placed between the telescope and the imager, that contains a piezoelectric tip-tilt mirror capable of executing a timed sequence of movements to place many speckle patterns over the active area of the CCD. The system will either (a) place a series of speckle images in a row and then use the CCD electronics to periodically shift charge toward the serial register or, if the CCD electronics do not allow, (b) have the mirror perform a serpentine step- and-scan motion over the entire CCD. When the entire CCD is full of speckle images, the chip is read out as normal. This kind of 'burst mode' speckle data collection effectively uses the large area of the CCD as a memory cache of speckle data frames, allowing large format scientific-grade CCDs that already exist at many observatories to be used efficiently and inexpensively in speckle imaging. The expected performance of the system, which is dependent on the CCD imager, is discussed. CCD speckle observations at the WIYN* 3.5-m telescope and simulation results indicate that, when used with a very low noise CCD, this system could obtain speckle data that are superior to those of even the best photon-counting cameras at the fainter magnitudes where such cameras are currently used.

Optical fibers with broadband transmission from the UV through the IR have not been available because the silica core material either has OH absorption bands in the IR or UV absorption due to intrinsic structural defects or chlorine. We have developed a new silica core material which can be fabricated into an optical fiber with very good transmission characteristics from 350 nm to 2000 nm. The transmission performance is stable with time because the fiber is not doped with hydrogen.

The Gemini telescopes implement active optics and atmospheric tip-tilt-focus correction through the use of a number of Shack-Hartmann wavefront sensors at several locations and providing various orders of wavefront data. All are based on a common Charge Coupled Device (CCD) housing and controller. This paper provides an overview of the electronics in the CCD head and controller. The two frame-transfer CCDs that are used (80 X 80 and 1024 X 1024), are described. A parameter based multiple subaperture readout algorithm is presented. The system is capable of operating at 2000 frames per second in a 2 X 2 subaperture mode with 36 pixels per subaperture and a read-noise floor of approximately 4.5 erms.

The high resolution spectrograph of the TNG (SARG) was projected to cover a spectral range from (lambda) equals 0.37 up to 0.9 micrometer, with resolution ranging from R equals 19,000 up to R equals 144,000. The dispersing element of the spectrograph is an R4 echelle grating in Quasi-Littrow mode; the beam size is 100 mm giving an RS product of RS equals 46,000 at order center. Both single object and long slit (up to 30 arcsec) observing modes are possible: in the first case cross-dispersion is provided by means of a selection of four grisms; interference filters are used for the long slit mode. A dioptric camera images the cross dispersed spectra onto a mosaic of two 2048 X 4096 EEV CCDs (pixel size: 13.5 micrometer) allowing complete spectral coverage at all resolving power for (lambda) less than 0.8 micrometer. Confocal image slicers are foreseen for observations at R greater than or equal to 76,000; an absorbing cell for accurate radial velocities is also considered. SARG will be rigidly fixed to one of the arms of the TNG fork by means of an optical table and a special thermally insulating enclosure (temperature of all spectrograph components will be kept constant at a preset value by a distributed active thermal control system). All functions are motorized in order to allow very stable performances and full remote control. The architecture of SARG controls will be constructed around a VME crate linked to the TNG LAN and the instrument Workstation B by a fiber optic link.

An overview of the scientific goals and specifications of the instruments to be installed on the two Nasmyth foci of the Unit Telescope 2 of the VLT is presented. The combination of a fiber position with three independent spectrographs provides a powerful utility for multi-object spectroscopy of large samples. Two spectrographs will cover the optical and infrared regime at medium resolution and a fiber link to the echelle spectrograph mounted on the second Nasmyth focus will permit multiplexing observations at high spectral resolution.

In this paper we describe the solutions adopted for the design and the realization of an astronomical CCD imaging system, the results achieved on a Schmidt telescope (1 degree X 1 degree Field Of View, FOV) and, as well, the images obtained on a 30 cm prototype of a really innovative wide field telescope (2 degree X 2 degree FOV), a two-mirrors Three- Reflection Telescope (TRT), adopting aspherical reflecting- only surfaces. This solution allows the correction of every aberration, removing completely the vignetting and the field- curvature on very large fields of view (FOV). The CCD camera (equipped with a Loral 2k X 2k chip), operating under 'inverted mode' (Multi-Phase-Pinned mode, MPP), shows a dark current less than 0.1 e^-/min at a temperature of only 200 degrees K. This particular operational mode enables the camera to provide accurate photometry even when the CCD is not at 'conventional' cryogenic temperatures. To cool-down the CCD chip it has been designed and realized a sophisticated Thermo- Electric-Cooler (TEC), which uses a three-stage Peltier module and glycol circulating in closed circuit at minus 10 degrees Celsius as heat exchanger. This TEC can reach, in this arrangement, an operative temperature of minus 80 degrees Celsius, approaching the typical performance of usual cryogenic systems. For its small encumber and for its reliability it is especially suitable for applications in which the room available is small, such as in internal-focus telescopes (e.g. Schmidt telescopes). Finally we briefly report new CCD and telescope projects.