It has been ten years since solid state devices, such as CCDs, have been discussed for use in astronomy. There is by now enough information, in the form of laboratory data on a variety of devices and published scientific results, to confirm the original assessment that solid state arrays will become a permanent entry in the repertory of astronomical instruments.

The design and performance of the Texas Instruments 800 x 800 CCD imager are described. This device is fabricated utilizing a three phase, three level polysilicon gate process. The chip is thinned to ~ 8 μm and is employed in the rear illumination mode. Detailed measurements of device performance including dark current as a function of temperature, linearity, and noise are presented. The device is coated with a UV downconverting phosphor which allows imaging with the same device over an extremely wide optical bandwidth.

Both TI and RCA CCDs have been applied to two-dimensional photometry and spectroscopy at Palomar. The optical, mechanical, and electronic properties of these chips are described along with a discussion of drive, processing, and recording techniques. A focal conversion optical system, the "Prime Focus Universal Extra-galactic Instrument" (PFUEI) is used to more optimally match the Hale 5 meter plate scale. Calibration and data analysis are discussed with illustrations of results.

Astronomical photometry with CCDs will often involve relatively low signal levels--less than 1000 charge carriers per pixel--at which non-linear effects sometimes become significant. These effects, which have already been troublesome to us in attempting accurate photometry of faint objects with ground-based telescopes, will doubtless play a role in the processing of Space Telescope data. The problem evidently arises because signal charges are not read out equally completely from all columns and all pixels. Although a first-order correction can be made by subtracting a low-level flat field from all frames before conventional processing, higher accuracy can be achieved by modeling the response with a non-linear function.

When CCD imaging arrays are used for astronomy the demands placed upon them are generally considerably different than those placed upon them in general TV applications. In both cases the parameters used to characterize a given array are generally the same but the target values for the parameters may vary significantly. The emphasis placed upon certain characteristics may also vary significantly depending on the particular application. Since, for some of the parameters at least, different methodologies provide different values it is important to understand how each value is obtained. It is the purpose of this paper to describe some of the methods used by one group at KPNO for determining the values of the primary parameters.

Because the astronomical performance capabilities of solid state imaging devices are generally beyond the manufacturer's testing interest, Kitt Peak National Observatory has undertaken a program to do this testing. This paper deals with the evaluation of the RCA 512 x 320 bulk (buried) channel CCD (SID 53601) and, to a more limited extent, the thinned version of the same device (SID 53612). The emphasis of the evaluation involved optimization of performance and performance parameters such as noise, linearity, quantum efficiency, and cosmic ray response. The results indicate that the devices can be useful in many astronomical applications, especially those requiring good blue response, but noise and low level transfer efficiency may limit its use in some spectroscopic uses.

An RCA buried channel thinned 320x 512 CCD has recently been put into operation as the detector on the blue side of a new double spectrograph used at the Cassegrain focus of the Hale 5.08-meter telescope. The chip is used in a vacuum dewer at a temperature of -132°C. The electronics are essentially the same as those used by Westphal and Gunn for the Texas Instruments 800x 800 CCD's. The readout noise over most of the chip is 46 electrons per pixel, but on one side of the device there are columns with readout noise three times as high as this. The dark current is approximately 85 electrons per pixel per hour and for most purposes can be neglected. When used in a spectroscopic mode with an f/1.5 beam the CCD shows interference patterns with an amplitude of 5 to 20 percent. This pattern occurs most clearly above 5000 Å. It appears to be stable and can be removed by using standard flat-field exposures. At signal levels of a few hundred electrons there are some charge transfer problems. The quantum efficiency between 4000 and 6500 Å is very high and the efficiency at 3260 Å is about 20 percent of that at 4200 Å. Above 6500 A the sensitivity begins to drop steadily.

We present a number of new and revised measurements of the imaging properties of the buried-channel RCA CCD at cryogenic temperatures, based on our experience with this device over the past 18 months in both a laboratory setting and at the telescope. Included will be a discussion of the effect of some of these characteristics on the problems of astronomical imaging as well as a description of some of our ongoing efforts to both apply photometric corrections to such images and remove defects.

We have built and tested a soft X-ray/XUV imaging camera that uses a thinned, back-illuminated, all-buried channel RCA CCD for radiation sensing. The camera is a slow-scan device which allows frame integration if necessary. The detection characterics of the device have been tested over the 15-1500 eV range. The response was linear with exposure up to ~0.2-0.4 erg/cm²; saturation occurred for greater exposures. Attempts to resolve single photons with energies of 1.5 keV are described.

A brief description is given of a CCD camera facility which has been designed and built for testing and characterization of CCDs in response to single X-ray photons of energy ~ 1 - 10 keV. The camera system facilitates ease of change from one type of CCD to another. Results are described from the frontside illuminated Fairchild 211 and 221 area imagers, and the disadvantages of commercially available devices are discussed. Preliminary results are presented. from (100 x 200) CCD arrays manufactured under contract by Westinghouse on high-resistivity silicon for deep depletion.

This paper describes the design and performance of two unintensified Reticon photodiode array detectors for high resolution spectroscopy with the Shane 3-meter telescope at Lick Observatory. Reticon arrays of 2 x 936 and 1 x 1872 formats were chosen for the initial installation. The remaining optical and mechanical components of the system were designed to allow easy addition of charge coupled devices (CCD's) as they become available. Detailed descriptions are given of cooling techniques, temperature control, and signal processing electronics which represent major improvements in performance and simplicity of operation over many previous Reticon systems. Both Reticon arrays are in routine operation for high resolution spectroscopy at Lick, and their performance is illustrated by astronomical observations obtained for several on-going research programs.

The Faint Object Spectrograph (FOS), one of five scientific instruments being built for the Space Telescope, has recently completed a significant milestone in its progression toward final instrument assembly, integration, and test. Seven flight Digicon phototubes have been successfully manufactured and tested and a development unit of the integrated detector assembly has successfully completed a set of difficult environmental and functional tests. This paper presents the results of these tests and details the results of the Digicon tube tests including photocathode quantum efficiency, dark count rate, uniformity, distortion, resolution, and diode array performance.

Two spectrophotometric detectors based on self-scanned RETICON arrays are in operation at the ASIAGO Observatory; one directly illuminated and one electron bombarded (a so called self-scanned-Digicon). A brief review of their performances is presented here.

A photon counting detector system, using a 4-stage EMI image intensifier tube followed by a lens coupled linear diode array (200 elements) with parallel outputs (each diode has its own pulse counting electronics), is described. The fast component of the phosphor decay (about 1 msec) limits the maximum linear range count rate to about 100 counts per channel and per second. This together with a dark count rate of about 10^-2 counts per channel and per second leads to the rather high dynamic range of 10⁴, which is required for the observing of emission line objects. The data handling and control functions are done by a PDP 11/34 minicomputer. In addition to the fundamental functions several supervision and test functions are integrated to monitor and compensate hardware situations which otherwise could falsify data. Some laboratory measurements are presented to demonstrate the system properties.

Many types of image tube intensified self-scanned array (ISSA) detectors are presently being used for x-ray, uv, and visible light astronomy. Notable types of ISSA detectors have been discussed in the literature. Several of these ISSA detector systems made use of lens coupling between the image tube and the SSA.

An imaging photometer employing a Bell-Northern Research (BNR) CCD has been developed at UBC and used on the Canada-France-Hawaii telescope (CFHT) in both a direct and an intensified mode. The intensified system is photon noise limited and linear from a 15th magnitude stellar source to the night sky limit. Intensified skylight provides a "bias" illumination that improves the low light level performance of the CCD.

A camera system has been constructed using the 188 x 244 element Charge Injection Device (CID) and a micro-programmable controller. The CID is used in a cooled direct imaging mode and controlled by a bit-slice sequencer. An LSI-ll computer with FORTH language is used. Results obtained with direct imaging, polarimetry and spectrograph work are presented.

This paper describes the CCD camera system that is currently operating at ESO, the performance characteristics of this camera, and presents some test results. The system has been developed with the philosophy that it will be important to be able to utilize a wide variety of CCD chips since chip development is proceeding rapidly. Thus, the system is based around a μ-processor, programmable in FORTH, which allows a areat deal of flexibility in driving the CCD chip. One feature of interest is the ability to do on chip summing of adjacent pixels. Test results in the laboratory used a thinned, special version of the RCA SID 52501 chip are encouraging, but the chip itself is far from optimized, having about 80 electrons/pixel of noise. However, this is considered sufficient for many direct imaging applications where sky noise will dominate.

A flexible microprocessor-based controller for charge coupled device two-dimensional detectors is described. The controller can operate under manual control or as a slave to a remote computer linked by coaxial cable. The system is discussed along with data taken at the telescope and in the laboratory.

We review some of the recent CCD detector development projects now in progress at KPNO. An overview is given of our CCD controller systems, which are designed to accept a wide variety of CCD detectors with a minimum amount of modification. The performance characteristics of two currently operational CCD detector systems are summarized -- the cryogenic camera system and the prime-focus CCD system. Finally we give a progress report of the development work now under way in evaluating the performance of two Fairchild CCD-221 detectors.

A prototype of low noise camera using a cooled Fairchild 202 CCD chip was developped for far red astronomical observations. A short description of the system is given. with a report on some of the first astronomical observations : edge-on Saturn configuration, galaxies and enveloppes of geant star.

A camera was designed for the use of large size CCD, such as the 380 x 488 pixels from Fairchilds or the 320x512 pixels from R.C.A. The aim of this camera is laboratory experiments however it can easily be settled up on telescope through a CAMAC interface. A description of this camera and preliminary performance obtained with a Fairchild CCD are given

A low light level CCD imaging system is described in which the polarization as well as the intensity for each pixel in the image is measured and displayed on a color display. Overall performance is good, and the availability of quick-reduced Stokes images makes CCD polarimagery attractive for astronomy. We show polarization images of two objects recently studied.

The Princeton observing and picture-processing equipment is a portable system for using charge-coupled detectors. It was designed to meet these goals: 1) Data may be reduced during an observing run. 2) Different CCD's may be used without modifying the hardware. 3) Reliability is stressed. The equipment is described.

A cooled slow-scan microprocessor-controlled CCD camera system is briefly described. The potential application of this CCD camera to astronomical measurements requiring relatively rapid chopping between two states, such as polarization, is discussed and the results of recent tests are described.

The General Electric Company in the UK (unrelated to US-GE) already have in production their MA357 device which has 576 x 385 pixels each 22 microns square giving a sensitive area 12.7 x 8.5 mm. These are thick devices so have no blue or ultraviolet response. They are commercially available in two versions; a normal thick device and one built out of epitaxial silicon that has the cosmic ray event rate of a thinned device. The performance of this device for astronomy is described elsewhere in these proceedings in some detail by the same author. It can be summarised as: 20 electrons rms system read-out noise, no charge transfer problems at any charge level and excellent linearity.

This paper gives details of the Cambridge CCD system hardware and of its performance at the telescope. A new technique has been developed (and called the Drift Scan technique) that eliminates most of the problems associated with flat field calibration and permits operation under photon shot-noise limited conditions. Some examples of recent astronomical results are given.

In 1985 NASA will launch a Galileo spacecraft using a new generation slow-scan planetary imaging system based on a recently developed single-phase charge-coupled device (CCD) sensor. This paper presents the fundamentals of operation of this new technology called "Virtual Phase" as used in Galileo. The design and operation of an 800 x 800 Virtual Phase CCD is described, and performance parameters as well as advantages and limitations are discussed.

The RCA 512 x 320 pixel buried channel CCD has been employed in a number of astronomical observations ii the past year and the astronomical results are presented in several companion papers. This paper reports on recent laboratory evaluation of an experimental version of this CCD which has a lower input capacitance on-chip amplifier. The data presented includes readout noise measured in several ways, charge transfer efficiency as a function of signal level, and dark current as a function of temperature. The response of the RCA buried channel CCD to single energetic electrons and soft X-rays is also presented.

The plasma torus of Jupiter is a diffuse, faintly emitting ionized gas surrounding Jupiter in the vicinity of Io's orbit. We have observed this region in the [SII] lines at 6716 Å and 6731 Å and in the [SIII] line at 9531 Å with an imaging spectrometer consisting of a narrow band tunable Fabry-Perot interferometer coupled to a CCD camera. Because of the extreme faintness of the emissions and the requirement of short exposure times imposed by the motion of the torus, which co-rotates with Jupiter, a prebinning mode of operation of the CCD was developed. This mode partially overcomes the rather large readout noise of the available CCD's when maximum spatial resolution is not required. While [SIT] images have been obtained previously using conventional image intensifiers, the [SIII] images at 9531 Å, beyond the wavelength limit of sufficiently sensitive image intensifiers, are the first ones obtained in this line. The Fabry-Perot, designed to cover the region of high CCD sensitivity eyond the limit of sensitive intensifiers, was also used to evaluate the system at the HeI wavelength 10830 Å. with short exposures of the Orion Nebula. The characteristics of the system and the method of producing the final quantitative images are described in the text.

We present details of a filter set which we have found useful for cosmological investigations with the CFA-CCD camera system. We discuss transformations to standard BVRI and gr I photometric systems, as well as systems of photographic photometry. We present colors and magnitudes of stars in an open cluster useful as standards.

The first observing run with the MASCOT (MIT Astronomical Spectrometer/Camera for Optical Telescopes) on the 1.3 m telescope at the McGraw-Hill Observatory in March 1981 demonstrated that the spectrophotometric sensitivity and stability of the instrument is consistent with the estimates made by Meyer and Ricker (1981) and Dewey and Ricker (1990), based on their calculations and laboratory calibrations. A pair of Texas Instruments virtual phse CCD's were used as the MASCOT optical detectors. In the W band (4000Å-7000Å), the MASCOT achieved a sky-limited sensitivity of +24.4 mag arc sec ^-2 in an 1800s integration (1 arc sec pixel size; 5σ level of significance). The ability to "flatten" pictures to a level consistent with (sky + source) photon statistics and readout noise was demonstrated. As an example of the capabilities which the MASCOT has demon-strated, we discuss our preliminary results from optical observations of four high galactic latitude Einstein x-ray sources without optical counterparts (i.e. "empty field" sources). For the four sources we observed, we established an optical counterpart for one source (1413+13), based on positional coincidence (better than 1.8 arc sec); detected four possible candidates in the error bode of anotper source (1009+35), and established sensi-tive upper limits (>22.3 mag; 4000 Å - 7000 Å, 5 upper limits) for optical counterparts in the error boxes for the other two sources (0920+39 and 0931-11).

Observations made with an RCA thinned back illuminated CCD on the echelle spectrograph of the Kitt Peak 4-meter Mayall telescope are discussed. With emphasis on interstellar line measurements in bright and faint background sources, the system efficiency, spectrum format, achieved signal-to-noise, interference fringes, and overall performance are discussed. A signal-to-noise of > 100/1 has been achieved at 0.2 Å resolution in two hours on a 13.2 mag star. Projected CCD improvements should yield signal-to-noise ratios of 20/1 for 18th magnitude objects (V) in 2 hours. With appropriate gratings, 1500 Å of continuous spectrum can be obtained.

A parallax program using a Charge Coupled Device (CCD) at the prime focus of the KPNO 4-m telescope has been underway since July 1980. The goal of the program is the determination of the distances to a sample of intrinsically faint normal and degenerate main-sequence stars. Program stars were selected on the basis of their large proper motions, extremely red colors, and faint apparent magnitudes. The CCD appears to be an excellent astrometric detector. Preliminary analysis of several standard fields indicates that image centers accurate to better than 0.25 micron (5 milli-arcseconds) can be obtained from 120 second integrations.

A program of precision radial velocity measurements using a RL1872 F/30 Reticon as detector and absorption lines of HF for reference (see Campbell and Walker 1979) is being pursued at both CFHT and at DAO, Victoria. While the long term accuracy of the technique will depend on the ability to duplicate and control the state of the HF gas in the absorption cell, oe can achieve a velocity precision in a single observation of a strong-lined star at 10 Å/mm dispersion, of ~15 ms^-1 at a s/n ~1200. This corresponds to a displacement ~2.10^-2 of a 15μ diode spacing. The critical instrumental and calibration techniques are discussed, in particular the temperature stability of the array, photometric calibration of the individual diodes and the measurement of individual diode to diode spacings.