Research and experimental results on Hg_1-xCdxTe and Pb_1-ySn_yTe are reported.4,8,16 I-Y and 24-element HOT infrared detector arrays have been prepared and found applications in thermal imagingeLTT/LT heterostructure detectors with D+(500K,1000Hz,1Hz)greater than 1x1010cm Hz1/2have been fabricated and will be developed into hybrid CCD devices.Some problems on the fabrication of CCD focal plane arrays are discussed.

This paper describes essentially three technologies that are involved in making self-scanning pyroelectric arrays. The first technology is that of the pyroelectric detector itself. The second is the technology of the multiplexing electronics used to read out the signal from the pyroelectric detector. The third technology is the hybrid integrated circuit technique of mating the two components together.

In analyzing the performance of direct-injected hybrid focal plane arrays, many factors must be consiaered in determining the minimum detector resistance-area product R_DA neces-sary to obtain background-limited performance (BLIP) and good array uniformity. In photo-aioue arrays, a necessary but not sufficient condition is that noise due to the diode generation-recombination and diffusion currents are less than the background photon shot noise. Tnis places a minimum requirement on the magnitude of RoA, the zero bias resistance area product. In addition, there are generally much more stringent requirements on R_DA uue to input MOSFET l/f noise and threshold variations which exceed the single detector R_oA requirement for BLIP operation at a given background. In general, the input thresnolu variations require that the photodiodes be somewhat back-biased. This produces a substantially higner average R_DA at the expense of higher detector l/f noise due to surface leakage. In tnis study we have investigated the detector impedance requirements in terms of tne injection efficiency, threshold nonuniformities, the input MOSFET excess (l/f) noise, and tne detector excess noise. For state-of-the-art parameters, it was determined tnat tne input MOSFET l/f noise always dominates the other elements in determining the required detector impeaance:k

Thermal imaging in the 3-5 micrometer band has been reported using Pt-Si detectors in staring mosaic arrays of more than 2000 elements. Excellent imagery has been obtained directly from the arrays without any need for signal processing or image enhancement. The only processing required for display is data reformatting to make the images compatible with standard TV frame rates. The absence of any need for image enhancement is a direct consequence of the excellent uniformity which can be obtained over large areas of the silicon wafer using Pt-Si detectors. Detector uniformities of better than 1/2% over a one centimeter area and extremely low fixed pattern noise in the readout CCD's have allowed image details to be displayed with 1/10°C accuracy.

Useful charge integration has been achieved in high quality photovoltaic InSb diodes when combined with a field effect transistor (FET) multiplexing (MUX) scheme. An experimental detector assembly with a linear array of 128 InSb diodes coupled to a FET MUX has been developed. The first stage J-FET preamp, pixel reset switch, MUX, and the InSb array are contained in a hybrid microcircuit of compact dimensions (1-3/4" by 1" by 1/8"). A low noise preamplifier topology which capitalizes on the low video line capacitance inherent with hybrid fabrication techniques is also utilized. Measurements indicate a signal-to-noise of over 1000:1 and response uniformity of +2 percent. Sample images show subtle detail which supports the estimated radiometric sensitivity of 0.01°K NEAt.

The Teal Ruby Experiment (TRE) , employing a space-shuttle-launched infrared telescope, will mark a major milestone in the application of mosaic infrared detectors in space. However, it is generally recognized that supplemental overview imagery acquired in the visible portion of the spectrum is necessary to assist in accurately interpreting the data generated by the infrared focal plane. For this purpose, the TRE will be equipped with a Charge Coupled Device (CCD) Visible Light Sensor (VLS) that will yield the cloud and ground truth data needed to assist in pointing and scene verification. The optical format of the VLS is designed to overlay the TRE footprint on the earth from a distance of 1000 nautical miles. The VLS is designed, built and will be qualified to meet the environmental and reliability requirements for an on-orbit mission in excess of one year. The TRE/VLS is tentatively scheduled for launch in mid 1983.

The Jet Propulsion Laboratory (JPL) is currently developing two imaging sensors using 800 x 800 charge coupled devices (CCD's) developed by Texas Instruments (TI). The Space Telescope Wide Field and Planetary Camera uses a three phase device. The original intent was to use the same device for the Project Galileo imaging system, however the radiation environment near Jupiter would have caused problems of both radiation damage and radiation induced noise with the three phase device. The availability of virtual phase technology provided a solution to the radiation problem. This paper compares the performance of the two different imagers and describes plans for further development of large scale virtual phase imagers.*

The use of AlGaAs/GaAs Schottky gate, heterojunction CCDs for visible/NIR imagers offers a number of distinct advantages over the use of silicon MOS homojunction CCDs, several of which are of particular importance for space applications. The use of the heteroepitaxial technology possible with the GaAs/AlAs alloy system allows the design of devices each of whose layers is optimized to a specific functions. A CCD imager, for example, has a wide bandgap electrically inactive AlGaAs layer for optical access to the device, a narrow bandgap p-GaAs layer as an optical absorber for good response to the visible-NIR spectrum, end an intermediate bandgap n-AlGaAs CCD channel for charge transfer and low dark current. adapting glass bonding technology originally developed for III-V photocathodes, a backside illuminated CCD can be realized which eliminates optical obscuration from the CCD gates. Finally, the fundamental properties of the AlGaAs/GaAs system together with the use of Schottky barrier gates eliminates cross-talk and blooming and results in a device which is intrinsically radiation hard. During this presentation, the above considerations will be discussed in detail and the status of the development program at Rockwell International will be summarized. Important demonstrations including glass bonded, backside illuminated imaging and the observation of room temperature dark currents <100 pA/cm² will be descirbed.

A signal processing system has been designed and constructed for a pyroelectric infrared area detector which uses a matrix-addressable JFET array for readout and for on-focal plane preamplification. The system compensates for all offset and gain nonuniformities in and after the array. Both compensations are performed in real time at standard television rates, so that changes in the response characteristics of the array are automatically corrected for. Two-point compensation is achieved without the need for two separate temperature references. This paper describes the focal plane circuitry used to read out the array, the offset and gain compensation algorithms, the architecture of the signal processor, and the system hardware.

Several IFPA programs having large quantities of deliverable devices provided Rockwell with the necessity of solving the many problems associated with high volume production. The facility contains processing, assembly, and cryogenic testing equipment.

The new technologies embodied in the Teal Ruby Experiment are surveyed to illustrate how the sensor represents much more than a new detector art. It is indicated that, as in all such multidimensional advances, much accommodation is inevitable among the competing design elements, requiring continuing systems engineering attention. The Teal Ruby Experiment is emerging as a significant benchmark in sensor design (see Figure 1).

Mosaic sensors have been advanced as superior sensors for observing moving targets in cluttered background, because they are potentially more sensitive than scanning sensors. Processing data from mosaic sensors, however, differs greatly from processing data from scanning sensors. Mosaic sensors produce much more data and the staring array introduces problems in rejecting background and assembling tracks. This paper surveys the mosaic data processing sequence showing how the nature of mosaic sensors introduces processing challenges in most of its stages. Also identified are design variables, such as detector size or integration time, that affect how well these challenges can be met. As an example of a design variable's influence on processing performance, the effect of blur circle size on the performance of a specific tracking and trajectory estimation algorithm is presented.

The problem of developing an on-board signal processor which meets stringent power and performance constraints applicable to space-based mosaic sensors is addressed. A large sensor system concept is used to derive an example power budget. Speed power product requirements as derived from the example power budget are used to evaluate applicable LSI technologies in terms of their present and projected performance. The success of a given signal processor design in meeting the requirement of low power will depend on LSI technology, algorithms, their implementation, and the overall processing strategy. A signal processing concept that makes use of reconfigurable pipeline elements is introduced and discussed in terms of its applicability to low-power signal processor implementations. Particular attention is given to operational strategies that include segmented field-of-view processing for target classes, low duty cycles achieved through multiple sampling intervals, dynamic range compression, quick/long-look stare periods, and dual resolution.

Infrared focal plane array (FPA) technology has advanced rapidly in the past two years. This technology promises advantages for future PLIR systems as well as advanced missile concepts. RCA has been in the forefront of this technology and has pioneered an IR focal plane array based upon Schottky Barrier detectors. This paper describes the technology and its status. Following a short description of the technology, the background of its development will be presented, and then a more detailed description of the device will be provided, including some performance measurements of the device.

The following is a short review of pyroelectric detectors manufactured at SPIRICON and how they are made.