A solid-state camera system designed for use as a multichannel detection system for spectroscopic applications is evaluated. The camera is based on a charge-injection-device sensor that permits very high dynamic range operation by virtue of its unique nondestructive readout capability. Methods of determining gain and readout noise are discussed. Performance of the camera system is evaluated with respect to sensitivity, linearity, readout noise, and dynamic range. The use of nondestructive readout to reduce effective readout noise and increase dynamic range is described. Spectral responses in the visible and ultraviolet of coronene-coated and uncoated devices are reported. Various time-dependent anomalies caused by interface trapped charge are discussed.

The need for a single-element charge-transfer device as a spectroscopic detector is discussed. Such a detector promises to offer superior performance compared to current photomultiplier tubes over a wide range of illumination levels. As a detector to address this need, the prototype CID75 manufactured by General Electric Co. is described and characterized. The CID75 is a single-element charge-injection-device sensor with a 1 mm by 1 mm photoactive area and a readout rate adjustable from 0 to 20 kHz. The electro-optical characteristics reported in this study include linearity, read noise, full-well capacity, dark count rate, and quantum efficiency. The sensors have good photometric linearity with a full-well capacity in excess of 1.2 x 108 e -. The read noise of the detector can be lowered to 80 e - when its nondestructive readout mode is employed. The quantum efficiency of the CID75 is reported for the wavelength range from 200 to 1000 nm. Combining a simple dynamic range of 106 with the ability to vary integration times over four orders of magnitude allows this detector to quantify photon fluxes varying over 10 orders of magnitude. The conclusions of this study are that the CID75 sensor is a useful detector for a variety of applications.

The Texas Instruments TC-104 linear CCD array with 3456 elements per line has been investigated for applications in spaceborne imaging systems for multispectral and stereoscopic earth observation. Of special importance are parameters such as linearity, pixel nonuniformity, dynamic range, and calibratability. The TC-104 was found to be suitable for imaging systems with a radiometric resolution of up to 12 bits.

Charge-coupled-device imagers have great potential to replace photographic film for recording cathode ray tube (CRT) displays. The basic problem, detection of electrons impacting with variable densities on a sensing screen, implies high sensitivity for single-electron detection. Few data have been published on CCD performance in this regime, especially as one component of a complete readout system. We report on the basic sensitivity and resolution of one possible system, the electron beam and phosphor coupled with fiber optics to a thermoelectrically cooled CCD camera. Observed performance indicates that present sensitivity is several CRT beam electrons per pixel and that resolutions of 25 µm trace width present no problem. Improved experiments are expected to yield single-electron detection capability in the near future. The outlook for broad use of this readout in oscilloscopes, streak tubes, and imagers is discussed with regard to this performance.

We describe the acquisition, reduction, and analysis of long slit spectra of active and normal galaxies using the Steward Observatory Texas Instruments TI 3PCCD system on the University of Arizona Observatory 2.3 m telescope. We also present fully reduced data for one Seyfert and one normal galaxy to illustrate the scientific motivations for using such a detector system.

This paper describes the design and performance of a laboratory instrument for imaging and nondispersive spectroscopy of soft x rays (300 eV to 10 keV) utilizing the Texas Instruments TI-4849 virtual-phase CCD. This instrument has achieved a spatial resolution of 22 p.m (limited by pixel size) with an overall array area of 584x390 pixels. It has achieved an energy resolution of ~140 eV FWHM for single-pixel 55Fe x-ray events (5.9 keV) with the CCD operated at -30°C. We have operated the CCD in photon-counting mode at room temperature and have obtained x-ray spectra with an energy resolution of ~450 eV at 5.9 keV. The low energy x-ray sensitivity of the CCD also has been demonstrated by detecting carbon Ka x rays (277 eV).

In general, attention to a multitude of details is essential when designing low noise CCD cameras. This paper describes some of the critical details of the MIT x-ray CCD camera design. Special attention is given to those portions of the system that deviate from conventional practice in the design of CCD cameras. Furthermore, an attempt has been made to generalize the design so that both optical and x-ray requirements can be satisfied whenever possible. Currently, noise levels of <8 rms are routinely achieved with this design, and even lower levels (<5 e) should be realizable.

We have operated two types of CCD detectors over the past several years: the three-phase thinned Texas Instruments 800 x 800 and the thinned RCA 320 x 512. Both detectors are operated in a slow-scan, cooled camera behind a variety of optical instruments at ground-based observatories, often involving low light levels. Emphasis is placed on optimizing the camera's operation with respect to quantum efficiency, readout noise, charge-transfer efficiency, dark current, full-well satura-tion, bias image flatness, residual image, and on-chip binning. Techniques for optimizing these properties are discussed. Our camera system is described, and information about the performance of our five operational CCDs is given.

This special issue of Optical Engineering is the third of three consecutive issues on charge-coupled devices. Papers contributed to this issue present various topics related to theoretical aspects of the CCD and include three papers discussing a close cousin of the CCD, the charge-injection device (CID).

This paper is concerned with the use of thermoelectric coolers (TECs) to cool charge-coupled devices(CCDs). Heat inputs to the CCD from thewarmer environment are identified, and generalized graphs are used to approximate the major heat inputs. A method of choosing and estimating the power consumption of the TEC is discussed. This method includes the use of TEC performance information supplied by the manufacturer and equations derived from this information. Parameters of the equations are tabulated to enable the reader to use the TEC performance equations for choosing and estimating the power needed for specific TEC applications.

A two-camera system has been developed for photographing and analyzing images consisting of particle tracks from high energy collisions in a streamer chamber track detector. The charge-coupled-device (CCD) cameras view the collision region, with typical images consisting of more than 100 tracks. Each camera consists of a 1024 X1024 pixel CCD array. Up to three cameras can be read out simultaneously, at a rate of 5 X105 pixels/s/camera digitized to 9 bits, through a custom-designed multiplexing interface into an integer array processor, computer, and color display station. A software system has been developed that allows multiple asynchronous computer processes to record, analyze, and display the streamer chamber images. The system has been used in a high energy nuclear physics experiment for on-line analysis, digital data acquisition, and storage of streamer chamber events, with subsequent off-line image and physics analyses.

This book deals with a broad range of subjects related to acoustics but will appeal to many readers of Optical Engineering because there are so many points of contact made to optics and to optical processing. Not only is there a great similarity in many of the concepts of optics and acoustics, but the two disciplines often converge in the same device.

The charge-coupled device has shown unprecedented performance as a photon detector in the areas of spectral response, charge transfer, and readout noise. Recent experience indicates, however, that the full potential for the CCD's charge-collection efficiency (CCE) lies well beyond that realized in currently available devices. In this paper we present a definition of CCE performance and introduce a standard test tool (the photon-transfer technique) for measuring and optimizing this important CCD parameter. We compare CCE characteristics for different types of CCDs, discuss the primary limitations in achieving high CCE performance, and outline the prospects for future improvement.

The maximum entropy principle (MEP) is applied to the problem of reconstructing an image from knowledge of a finite set of its moments. This new approach is compared to the existing method of moments approach and is shown to have a clear edge in performance in all of the applications attempted. Compression ratios more than twice as high as those previously achieved are possible with the new MEP method.

Infrared CCD arrays generally require a compensation for the effect of gain and offset variation among the individual detectors of the array. Linear compensation techniques do not suffice for focal planes that exhibit a large nonlinearity of response combined with order-of-magnitude variations in threshold and saturation flux levels. This situation is common among hybrid architecture CCDs, particularly when HgCdTe is the detector material. This paper reports on a multipoint piecewise-linear correction scheme employed on a HgCdTe infrared CCD focal plane. This technique allows a compensated response to be obtained in a computationally efficient manner. An experimental relationship between the number of calibration points and the amount of residual fixed-pattern noise is presented and compared to previous analytical models.

In my editorial published in the May 1987 issue of Optical Engineering, I addressed a question that had been posed by Adolf Lohmann regarding the appropriate sequence for listing authors' names on multiple-author technical publications. Adolf indicated that authors' names are listed in alphabetical order at his institute but wondered how the sequence was determined elsewhere. There have been several responses to that editorial, most of which have tended to agree with the opinion I expressed therein: if the contributions of one of the authors are substantially greater than the contributions of the others, that author's name should be listed first, followed by the names of the remaining authors in alphabetical order.

Noise performance in imaging arrays is often specified simply by the variance of the pixel levels. In this paper a more complete characterization technique is presented, based on the spatial-frequency power spectrum of the noise data on the detectors. This is seen to provide additional information for cases in which the noise spectrum is nonwhite. Experimental data demonstrate the nonwhite nature of the spectrum under certain conditions, especially resulting from spatial correlation of detector nonuniformity and from sampling-lattice-related artifacts in the data.

In this paper we describe and analyze the internal noise sources of a signal charge partitioning circuit and indicate techniques for minimizing the internal noise of the circuit. Two noise sources are identified and described as a function of the circuit architecture and operating speeds: noise due to the trapping of thermal charge-density fluctuations (Johnson noise) and noise due to electrons scattered by the action of the partitioning gate. The noise due to the trapping of thermal charge-density fluctuations is dependent on the time it takes to divide the charge; at most, it is equal to the thermal Johnson noise of the integration well times the partition ratio. Noise due to the closing of the partitioning gate is dependent on the division ratio, the length of the septum gate, and its closure speed; it is, however, independent of the size of the charge packet being divided. With small division ratios (<10) and optimal operating speeds, the partition noise can be limited to fewer than 100 e-.

An optical multichannel detector system based on a charge-injection device (CID) sensor has been designed and constructed. This system is intended for use in a variety of spectroscopic studies, including atomic emission spectrometry (AES), in the ultraviolet to near-infrared spectral region. A General Electric Co. CID11 B was selected as the sensor because preliminary research indicated that this device has excellent characteristics for detection of the very wide dynamic range signals encountered in AES. During initial characterization of this detection system, a phenomenon was observed in which the signal obtained from reading a pixel of the array was affected by the quantity of charge in orthogonally located pixels. Investigations indicate that this spatial crosstalk effect is not caused by charge migration among pixels ("blooming") as in many other types of imaging detectors. Instead, the source of crosstalk for the CID11B is capacitive coupling of charge-collection and charge-sensing electrodes during the readout process. Because no charge migration is involved with this process, crosstalk in a CID can be corrected by simple algorithms, which are described.