Recent progress in infrared astronomy has been very rapid because of the introduction of two dimensional detector arrays. Practical experience in using these arrays has changed the techniques and outcomes of infrared astronomical research profoundly. In this paper those changes are illustrated by an assortment of images and spectra.

Meeting the stringent requirements of astronomy provides some of the most difficult and interesting challenges to the infrared industry. This paper evaluates InSb array performance, compares this performance with astronomy requirements, and predicts what could be available before the end of this century. Technology trends in infrared arrays for astronomical applications are analyzed to stimulate development in the industry and to assist astronomers in planning future instruments.

SIRTF, and other infrared space astronomy projects, require detector arrays with extremely high sensitivity. It is a goal of SIRTF to achieve background limited performance at all wavelengths. Especially critical is the spectral region around 3 micrometers , where the detector current due to the zodiacal background radiation is a minimum. For background limited operation, at a spectral resolution of 100, the dark current must be less than 0.1 e^-/s/pixel. The detector noise must be less than the noise given by fluctuations in the number of zodiacal background photons (< 9 e^-/pixel). Other detector array goals include: high quantum efficiency (> 90%), radiation hardness, minimal image latency, and excellent photometric accuracy and stability. Many of the performance goals have been met with Santa Barbara Research Center's 256 X 256 InSb arrays.

The evolution of Santa Barbara Research Center (SBRC) InSb focal plane arrays (FPA) for astronomy has been significant since the introduction of the SBRC 58 X 62 array in the mid 1980's. A 256 X 256 array, whose performance is evaluated herein, is currently available from SBRC, and a 1024 X 1024 InSb FPA is under development by the U.S. Naval Observatory (USNO). National Optical Astronomy Observations (NOAO), and SBRC. This represents an increase by a factor of approximately 300 in the number of pixels in less than 10 years, with attendant increase in performance as well. The availability of high performance infrared arrays has opened up the field of infrared astronomy. As this paper shows, the 256 X 256 InSb FPA is a great improvement over its predecessor in all areas of performance. Its most striking improvement is the complete absence of a latent or residual image after exposure to very high fluxes. Because of its low node capacitance, which results in very low read noise, the full well capacity is somewhat compromised.

We are investigating trapping/recombination centers in near-infrared (1 - 5 micrometers ) InSb imaging arrays via experimentation and theoretical modeling. The presence of impurities, lattice defects and/or surface states can compromise the operational qualities of an imaging array by introducing latent images, signal rate/quantum efficiency loss at low signal levels, and by increasing noise and dark current. Identification of these trapping centers should enable a reduction in their number density by appropriate changes in the material processing and fabrication steps. We have performed experiments and analyses on both gate-controlled arrays (SiO_x surface passivation) and recently received gateless arrays (Si₃N₄ surface passivation). All of the gated arrays showed latent images at temperatures 6 - 26 K for signal fluxes as low as 1500 e^-/s/pixel, while neither of the two gateless arrays examined has shown latent images in the same temperature range; no latent image was detected (to a level of < 50 e^-) in a 5 second integration after exposure to a 2 X 10⁶ e^-/sec/pixel signal flux. We interpret this as evidence for surface state charge trapping in the region of the gate oxide, which is largely eliminated by the new passivation. The physics of surface states is investigated theoretically in order to gain an understanding of the surface contribution to the observed behavior, and a model is presented to explain the experimental results.

As part of our instrument development work for the U.S. Air Force Advanced Electron Optical System (AEOS) telescope to be built on Haleakala, Maui, the Institute for Astronomy is contracting with the Santa Barbara Research Center (SBRC) for the development of a 1024 X 1024 InSb detector array with 30 micrometers pixels optimized for groundbased astronomical applications. The device design is based on the successful 256 X 256 InSb devices currently produced by SBRC.

This array was developed for use in future NASA space infrared instrumentation and was funded by Craig McCreight at NASA Ames Research Center. The multiplexer was designed at Valley Oakes Semiconductor and fabricated using TRW's radiation hard 1.2 micron CMOS process as a baseline. Several processing variants were explored as this effort was directed toward the development of a low temperature (< 10 K) multiplexer. The 8 to 256 CMOS address decoders allow for random access and subarray readouts. Cincinnati Electronics developed the 2-D array of photovoltaic InSb mesa diodes. Much of this paper presents an evaluation of the bare multiplexer to determine the optimal operating point of the hybrid array. At 10 K, the detector reset node must be operated at a minimum of 2.2 V with respect to the reset on control signal (multiplexer ground) to overcome the threshold drop of the PMOS reset transistor. A linear regression of the multiplexer response at 10 K indicates that for linear gain response the reset voltage must be between 2.4 to 2.8 V. The standard deviation of the pedestal values gives an indication of multiplexer uniformity and is 32.0 mV. Charge pumping through the reset transistor adds bias to the detector. Within the operating range of the multiplexer, this charge pumping was measured to be 70 to 100 mV. The multiplexer operates continuously from 77 to 10 K with no anomalies due to threshold crossover in the CMOS gates.

A 256 X 256 element InSb (indium antimonide) focal plane array has been specifically developed for use in ground-based astronomy. The array is an indium bump hybrid of a high- quality InSb detector array fabricated with an improved process, mated to a new, specially- designed low-background multiplexer. The performance parameters have been tuned to best reflect the requirements of ground based astronomy. The circuit is a direct readout detector integrator. It has a well size typically around 1,000,000 electrons, a readout rate of about 400 kHz, and has an expected noise level of about 200 electrons.

We report on progress in the construction of highly sensitive arrays of germanium photoconductors for operation at wavelengths between 40 to 120 micrometers . This technology is being developed for future astronomical missions in the far infrared such as SIRTF. We are evaluating construction techniques that will allow us to build 32 X 32 format arrays that are reliable and suitable to be qualified for space flight, while retaining the excellent performance of a prototype we have built and tested previously. We review the technical issues in constructing large format far infrared arrays, along with a description of the approaches that have been adopted in a prototype array to deal with these issues. We report on the development of two advanced methods to interconnect the detectors and readouts, that allow the extension of the Z-plane concept to larger formats: (1) Flex cable and (2) tape automated bonding. Both methods are being evaluated in a detailed series of laboratory experiments, including the electrical performance of the integrated readout/interconnect systems, thermal performance, ease of assembly of completed arrays, and mechanical ruggedness. These approaches will be discussed in terms of their use both with readouts that need to operate at temperatures above those acceptable for the detectors (approximately 20 K) and with readouts that can operate close to the detector temperature (approximately 2 K).

As part of a program to extend the wavelength response of infrared array technology beyond the cutoff of extrinsic silicon photoconductivity, we have calculated the optimum parameters for a front-illuminated two dimensional array of Ge:Be detectors. Detector arrays to this prescription have been produced by indium bump bonding onto sapphire fanouts. We report a preliminary performance evaluation of these arrays.

Four 58 X 62-element Si:As impurity-band-conduction (IBC) detector arrays produced by the Hughes Technology Center were tested to evaluate their usefulness for space- and ground- based astronomical observations. PMOS circuitry was used in the multiplexers to improve low-temperature noise performance. Laboratory tests at background levels simulating those expected on space-based observing platforms were combined with ground-based telescope IR observations. The devices have shown read noise levels below 120 rms e-, dark currents below 10 e-/s, and detective quantum efficiencies of 20%.

Cryogenic space telescopes such as the Space Infrared Telescope Facility (SIRTF) require large-area focal plane arrays (FPAs) with high sensitivity. Such applications set requirements for the readout arrays to simultaneously provide low noise and low power dissipation at very low temperatures. The Hughes Technology Center (HTC) has developed a low-noise 256 X 256-pixel hybrid FPA composed of a PMOS readout array hybridized to an arsenic- doped silicon (Si:As) impurity-band conduction (IBC) detector which is designed to operate below 10 K. The readout unit cell employs a switched source-follower-per-detector (SFD) design where in signals are multiplexed onto four outputs. The detector was processed using high-purity, multilayered epitaxial processing. The readout was processed using the p-channel subset of HTC's CryoCMOS process.

The design and construction of 100 mK composite bolometers for low background submillimeter and millimeter-wave astronomy are discussed. The bolometers are cooled to 100 mK using an adiabatic demagnetization refrigerator. The bolometers consist of a silicon substrate suspended by nylon fibers, a bismuth film absorber, a neutron transmutation doped germanium thermometer with graphite fiber electrical leads, and a brass wire thermal strap. Heated JFET amplifiers located on the 1.5 K cold plate are used to read out the bolometer signals. Electrically measured noise equivalent powers as low as 2 X 10^-17 W/(root)Hz have been achieved.

The Long Wavelength Spectrometer (LWS), part of the Infrared Space Observatory (ISO), uses an array of ten doped-germanium photoconductors to cover the wavelength range 45 - 200 micrometers . The array comprises a Ge:Be detector for the 45 - 50 micrometers range, five unstressed Ge:Ga detectors for wavelengths between 50 - 110 micrometers and four stressed Ge:Ga for the long wavelength range 110 - 200 micrometers . We have calibrated the performance of the detectors, both individually and in the LWS. Optimum values of detector bias and operating temperature that maximize the sensitivity of the LWS have been established, resulting in detector NEPs of 10^-17 WHz^-1/2 for Ge:Be, 5 X 10^-18 WHz^-1/2 for unstressed Ge:Ga and 5 X 10^-19 WHz^-1/2 for stressed Ge:Ga. Under the low flux conditions expected in the LWS (< 10^-14 W at the detector), doped-Ge photoconductors exhibit a number of other non-linearities and memory effects. Additionally, the performance of the detectors will be strongly influenced by ionizing radiation encountered during orbital perigee. The characteristics of these effects are described and the implications for operation and calibration of the LWS considered.

Rockwell's short wavelength infrared (SWIR) focal plane arrays (FPA) were originally designed for use in Orbital Replacement Instrument for the Hubble Space Telescope, but the 256 X 256 FPA version subsequently has found a home in many observatories. Developed for the University of Arizona under a NASA-Goddard prime contract to the University, the device is designated NICMOS3 due to its original relationship with the Hubble's Near Infrared Camera Multi-Object Spectrometer. Typical NICMOS3 FPAs have read noise < 35 e- with < 1 e-/sec detector dark current at 77 K and broadband quantum efficiency > 50% from 0.8 to 2.5 micrometers . These devices are in use all over the world by many researchers for SWIR astronomy. Based on long-term interaction with these scientists and on our own tests, the consensus is that the NICMOS3 is an extremely useful device. We are working to facilitate several paths for the subsequent low risk development of significantly upgraded astronomical FPAs. These include an even higher performance 256 X 256 FPA consisting of an upgraded readout mated to either standard or improved PACE-1 HgCdTe detector arrays, the near-term development of a 512 X 512 FPA via a proposed astronomical research consortium, and the longer term development of a 1024 X 1024 via several possible paths.

Features of the flight hardware version of the NICMOS 256 X 256 mercury-cadmium- telluride (MCT) detector array for the Hubble Space Telescope (HST) are presented and described. Detector flowdown requirements for flight production are reviewed and discussed. Detector cross section and array architecture features are analyzed in relation to quantum efficiency and crosstalk behavior. Features of the charge integration scheme employed are analyzed in assessing dynamic range.

The NICMOS3 infrared focal plane array (FPA), which was designed as a Hubble Telescope upgrade device, provides excellent low-noise images in the 1 - 2.5 micrometers (SWIR) band. Both the detector array and the readout multiplexer of this hybrid FPA are optimized for low- noise operation. The NICMOS detector array is fabricated in HgCdTe grown on a sapphire substrate (PACE-I material). The sapphire substrate is very rugged and provides a good thermal contraction match to the silicon multiplexer, producing excellent reliability. The composition of the HgCdTe is adjusted to yield a response cutoff at 2.5 micrometers which limits the detector response to thermal background from the atmosphere and telescope. The quantum efficiency of the detectors is %GRT 50% over the 1 - 2.4 micrometers range. The dark current of the NICMOS detector is < 1 e^MIN/s at 77 K, which is unprecedented for an IR detector. The multiplexer is a switched-FET CMOS design with a single source-follower per unit cell. The photocurrent is integrated on each detector diode, and the diode voltage level can be read nondestructively, or reset after each readout. This flexibility in the FPA operation makes it possible to generate images at a 12 Hz data frame rate or to optimize for low-noise exposures of many thousands of seconds. With a readout before and after each reset, off-chip correlated double sampling can be implemented to reduce the read noise to < 30 e^-.

Several cameras employing NICMOS2 (128 X 128) and NICMOS3 (256 X 256) arrays have been used in diverse applications on ground-based telescopes. Such use of arrays provides many insights to their performance that are not usually obtained in laboratory tests, and provides a baseline of performance over repeated thermal cycles. Astronomical use of these arrays also involves extracting accurate photometric information and detection of sources in spite of high background levels. The NICMOS arrays have proven very capable as astronomical imagers.

A comprehensive program has been developed for the production of focal plane assemblies (FPA) for use on the University of Arizona Near Infrared Camera and Multi-Object Spectrometer (NICMOS) instrument which is to be installed in the Hubble Space Telescope (HST). This paper describes the current schedule, tests to be performed, test conditions and unique test facilities of the flight FPA qualification test program. This test series is intended to validate design, assembly, performance and reliability of flight qualified FPAs. Also described are the design features, performance characteristics and test results obtained with prototype FPAs used as engineering evaluation units prior to committing the flight qualification units to manufacture. The qualification tests will demonstrate performance margins over and above requirements under operating environmental conditions. Included in the qualification tests are electrical, mechanical and thermal tests. Random vibration tests and mechanical shock tests will be performed at 1.5 times the load level specified for acceptance requirements. The random vibration tests simulate launch conditions and will induce stresses to uncover any potential structural deficiencies that might exist. The mechanical shock tests will simulate potential impacts incurred as a result of handling or transport. The qualification test program is intended to maximize confidence in the quality and integrity of the flight FPAs.

Exploiting hybrid focal plane array methodology and a flexible multiplexing readout, 128 X 128 FPAs were made and directly compared using several short wavelength infrared (SWIR) and long wavelength (LWIR) detector technologies. The detector types include two GaAs/AlGaAs quantum well infrared photodetectors (QWIP), 1.7 micrometers InGaAs/InP, and 2.5 micrometers PV HgCdTe. The tests were performed at operating temperatures ranging from 35 K for the LWIR devices to as high as 175 K for the SWIR FPAs. Highlights include the first FPA demonstrations (to the best of our knowledge) of BLIP-limited detectivity (D*) for both LWIR GaAs/AlGaAs QWIP and 1.7 micrometers PV InGaAs/Inp. The 9 micrometers QWIP peak detectivity is near the theoretical background limit at 1.2 X 10¹⁰ photons/cm²-s background and 35 K operating temperature. The mean D* of 4.5 X 10¹³ Jones at 8.3 micrometers peak wavelength is 75% of BLIP. A maximum peak D* of 5.7 X 10¹⁴ Jones was achieved with the PV InGaAs/InP device at 200 K. This is also believed to be the highest reported FPA-level D* for a staring mosaic array operated at TV-type frame rate and integration time.

The NICMOS camera to be used on the Hubble Space Telescope will acquire near-infrared images with extremely high spatial resolution. To extract scientifically useful data from these images will require a complete understanding of the arrays used to produce the images. The NICMOS team has developed a program for characterizing arrays which will lead to this understanding and which will also allow selection of optimum devices for each section of the NICMOS instrument. The overall plan will be described and related to the scientific goals of NICMOS. The characterization plan includes standard infrared array testing such as electrical properties, read noise, dark current, and quantum efficiency, and will be expanded to include testing such as crosstalk measurements, hysteresis testing, and radiation testing.

A program is currently in process at Rockwell for the production of focal plane assemblies (FPA) for use on the University of Arizona Near Infrared Camera, Multi-Object Spectrometer (NICMOS) instrument which is to be installed in the Hubble Space Telescope (HST) during a Space Shuttle mission in 1997. This paper describes the testing approach for the production of the FPA's, the tests to be performed, the test equipment, and facilities used in the built of two prototype FPA's used as engineering evaluation units (EEU). A sample of the data output from the various tests will be discussed for one of the EEU FPA's, both in the intermediate screen tests which will be described, and for the final performance tests.

KWIC (Kuiper Widefield Infrared Camera) is a wide field imaging spectrometer/spectrophotometer designed for use between 18 and 44 micrometers on NASA's Kuiper Airborne Observatory (KAO). KWIC achieves the highest possible spatial resolution (approximately 5 to 10') over this band, by using a Rockwell 128 X 128 pixel Si:Sb BIB array. Even with spatially oversampled, 3' X 3' pixels, the large array gives KWIC a very wide field of view: 6 X 6 arc minutes. KWIC delivers high spatial resolution images both in imaging spectrophotometry mode and in imaging spectrometer mode. KWIC is thereby optimized for detailed investigations of dust and discerning lines in galactic and extragalactic sources. these modes are quickly interchangeable during a KAO flight. We plan to have the entire instrument complete by mid August 1993 and hope to have our first flights on the KAO in early 1994. KWIC will address a wide variety of astrophysical issues. The fine-structure lines available to KWIC are excellent probes of the density and ionization structure of interstellar gas clouds. The mid-IR continuum arises from warm dust heated by nearly starlight. We will use KWIC's large scale diffraction limited imaging capabilities to: (1) explore molecular cloud structure, probing the physical conditions of the gas near embedded sources and condensations, (2) examine the relationship between the interstellar medium and star formation activity on galactic scales through complete imaging of nearby galaxies, (3) examine the complex kinematics and density structure of the gas clouds associated with the center of the Galaxy, and (4) conduct high spatial resolution, large scale imaging of young stellar objects in the dust continuum thereby probing the dusty disks predicted to accompany low mass star formation.

The Mid-Infrared Spectrometer (MIRS) is one of four instruments that will fly aboard the orbiting Infrared Telescope in Space (IRTS). This telescope is a joint NASA/Japanese Space Agency (ISAS) project that is scheduled for a Spring, 1995, launch aboard a Japanese expendable launch vehicle and subsequent retrieval by the space shuttle. The telescope itself is liquid helium-cooled with a 15 cm aperture and will survey approximately 10% of the sky before its cryogen runs out and it begins to warm up. The MIRS was developed jointly by NASA, the University of Tokyo, and ISAS and operates over a wavelength range of 4.5 to 11.7 microns with a resolution of 0.23 and 0.36 microns. The MIRS has a conventional entrance aperture, so that spectral studies can be made of extended as well as point-sources. A cold shutter and an internal calibrator allow accurate absolute flux determinations. Calibration and sensitivity tests in the laboratory have shown that the instrument sensitivity will be limited by the fluctuations due to the zodiacal dust emission over the wavelength range of the spectrometer. The large A-omega of the spectrometer, the cryogenic optics, and the survey nature of the telescope will allow very sensitive studies of the spectral characteristics of diffuse extended emission. These observations will help in determining the composition of the galactic dust responsible for the warm component of the infrared cirrus. In secondary observing programs, the MIRS will also take spectra of the zodiacal dust emission as well as measure the infrared spectra of an estimated 9,800 point-source objects.

The flight model of ISOPHOT has been completed and delivered to ESA. The pre-flight calibration verified the high performance specifications for the far-infrared detector channels. At times, limiting factors in orbit may be high-energy-radiation induced noise and responsivity drifts after passage through the radiation belts, and signal drifts following large incident flux changes.

The Infrared Space Observatory (ISO) has a complement of four focal plane instruments for making a range of astronomical observations at infrared wavelengths. The telescope and instruments are operated at cryogenic temperature. Spectroscopy is shared between two of these instruments, with the Long Wavelength Spectrometer (LWS) providing for spectroscopic observation over the wavelength range 43 micrometers to 198 micrometers at two resolving powers. The flight model of the LWS has been completed, following an extensive program of performance testing and calibration at the Rutherford Appleton Laboratory. For this, a test facility has been developed to provide the necessary operating and environmental conditions, including a very low thermal background. The design and operational details of the test facility are given, followed by examples of the LWS performance values obtained. The data from these measurements will provide the initial calibration of the LWS in-orbit.

ISOCAM, the ISO camera is designed to map selected regions of the sky in the spectral regions 2.5 to 17 microns. It will make images, within the 3 arcmin field of view of ISO with two 32 X 32 infrared array detectors, one for the short wavelength range, below 5 microns, the second for the long wavelength range, above 4.5 microns. Filter wheels and lens wheels allow to change the spectral resolution and the pixel field of view. Circular variable filters are also mounted on the filter wheel. The instrument is ready for delivery to ESA, after a thorough testing and calibration phase. Test has been conducted in a facility that simulate the ISO environment. A particular care has been taken to ensure the appropriate level of IR background inside the calibration cryostat, to check the detectors in the actual IR flux range that they will experience in flight. This paper presents the results on this test and calibration campaign, with a particular emphasis on the optical performances and on the behavior of the detectors. Photometric performances have been obtained for all the observing modes of ISOCAM. Some of these results, like stabilization of the detectors, have strong impacts on the observing strategy with ISOCAM.

The Short Wavelength Spectrometer is one of the four instruments for the Infrared Space Observatory. The instrument operates at about 3 K. Employing diffraction gratings, if offers a resolving power between 1000 and 2000 in the wavelength range 2.45 to 45 micrometers . An additional Fabry-Perot interferometer offers resolutions between 25,000 and 30,000 in the range 12 to 45 micrometers . The instrument employs arrays of discrete detectors: InSb photo- diodes and Si:Ga, Si:Sb and Ge:Be photo-conductors. In the course of 1992, the flight unit was tested and characterized, with interruptions for minor modifications. The paper discusses the SWS design and its performance.

The design of the cryogenic, scanning Fabry-Perot interferometers for the Long Wavelength Spectrometer on the Infrared Space Observatory is presented. The spectroscopic performance is compared with the model developed for the interferometer design. Although the measured efficiency is in reasonable agreement with calculations, the spectral resolution is lower than expected. This is ascribed to non-flatness of the reflectors.

A 1 - 5.4 micrometers Cryogenic Echelle Spectrograph (CSHELL) for the NASA Infrared Telescope Facility is described. It achieves a resolving power of 5,000 to 40,000 using slits ranging from 4.0' to 0.5' in width and 30' long. It operates in a single-order long-slit mode, and a circular variable filter is used as an order sorter. Two infrared arrays are employed to achieve spectral coverage from 1 - 5.4 micrometers : a 256 X 256 HgCdTe NICMOS-3 array for 1 - 2.5 micrometers and a SBRC 58 X 62 InSb array for 2.8 - 5.4 micrometers . A closed- cycle cooler is employed to keep the optics and supporting structure at 73 K and to maintain the detectors at their proper operating temperatures. The entire spectrograph fits within an envelope of 64 cm X 35 cm X 27 cm. The instrument is controlled by a microcomputer mounted on the telescope, but the observer commands the instrument from a UNIX X Windows workstation on the Internet. This use of the Internet for communication between instrument control and user interface computers facilitates remote observing. A limiting magnitude of 12.3 mag is achieved for S/N equals 10 in 1 hour integration time, at resolving power of 20,000 at 2.2 micrometers wavelength.

Infrared (IR) observations have traditionally been limited to a relatively small number of specialized telescopes since: (1) the cost of detectors and detector system development is large; and (2) there are a number of significant technical differences associated with an 'infrared-capable' telescope when compared to a traditional optical telescope. With the advent of lower cost infrared detectors in recent years, IR instrumentation now becomes accessible to observatories with budgets unable to support the traditional high costs. We have assembled a complete observing system making use of a Hughes 256 X 256 pixel PtSi 1 - 5 micron array detector. This particular PtSi detector was chosen because it has several characteristics conducive to precise photometric observations. The detector has 100% fill factor, a large dynamic range of 10⁴, low dark current and the potential for extremely good stability. The system we describe was designed with an emphasis on simplicity and the use of commercially available hardware and software, while retaining full performance of the detector. The system proved to integrate easily into the CCD observing environment used on our telescopes.

SpectroCam-10 is a 10 micrometers spectrograph and camera built at Cornell University as a facility instrument for the 200 inch Hale telescope. The instrument is optimized for operation from (lambda) equals 8 to 13 micrometers in three modes: a medium-resolution spectrography (R equals (lambda) /(Delta) (lambda) approximately equals 2000), a low-resolution spectrography (R approximately equals 100), and a camera with diffraction limited (0.5 arcsec) spatial resolution. An optical flat and two reflection gratings mounted on a cryogenic rotating mechanism allow rapid switching between modes. The detector is a Rockwell 128 X 128 Si:As Back Illuminated Blocked Impurity Band array. We discuss the design and operation of the instrument, and present some scientific results from our early observing runs at Palomar.

Low-frequency, low-noise, low-power cryogenic electronics to read out photodetectors is being investigated for the star-tracking telescope of the Gravity Probe B spacecraft. We report additional results in evaluating low-frequency '1/f' noise of commercial and non-commercial GaAs field-effect transistors (FETs) at room and liquid-helium temperatures. No correlation was found between noise at these two temperatures. For our dc biasing conditions, the lowest- noise non-commercial GaAs FETs give a typical value of K_f (equals A_f x gate area) approximately equals 2 X 10^-22 V² (DOT) m²; this corresponds to a noise voltage of approximately equals 80 nV/Hz^1/2 at 1 Hz for a gate area of 3 X 10⁴ micrometers ², only a factor of approximately equals 3 higher than the best Si JFETs of comparable gate area operated at their optimum temperature. RTSs (random telegraph signals) were observed for many GaAs MESFETs at 4 K, for gate areas up to approximately equals 5000 micrometers ². We also examined low-frequency '1/f' noise in relation to FET materials, processing, and pinch- off voltage but the results were inconclusive.

On-focal-plane signal processing circuits for enhancement of IR imager performance are presented. To enable the detection of high background IR images, an in-pixel current-mode background suppression scheme is presented. The background suppression circuit consists of a current memory placed in the feedback loop of a CTIA and is designed for a thousand-fold suppression of the background flux, thereby easing circuit design constraints, and assuring BLIP operation even with detectors having large response non-uniformities. For improving the performance of low-background IR imagers, an on-chip column-parallel analog-to-digital converter (ADC) is presented. The design of a 10-bit ADC with 50 micrometers pitch and based on sigma-delta ((Sigma) -(Delta) ) modulation is presented. A novel IR imager readout technique featuring photoelectron counting in the unit cell is presented for ultra-low background applications. The output of the unit cell is a digital word corresponding to the incident flux density and the readout is noise free. The design of low-power (< 5 (mu) W), sub-electron input-referred noise, high-gain (> 100,000), small real estate (60 micrometers pitch) self-biased CMOS amplifiers required for photon counting are presented.

An optical link can provide an interface channel for the focal-plane array that is immune to electromagnetic interference (EMI) and can lower the heat load on the dewar. Our approach involves the use of fiber-optics and an on-focal-plane optical modulator to provide an interface to the focal-plane array (FPA). The FPA drives the modulator with an electrical signal. We evaluated specially fabricated AlGaAs/GaAs multiple quantum well (MQW) optical modulators, operating near 840 nm, for analog modulation, and we have used the results to calculate the performance of an optical interface link using experimentally determined device parameters. Link noise and dynamic range for an analog link were estimated from a separate experiment using pigtailed fiber components. The performance of the MQW modulator system is compared to alternative strategies. Significant improvement in performance in comparison to conventional electronic interfaces appears to be possible.

This paper discusses the latest results of a continuing study of the properties of the complementary heterojunction field-effect transistor (CHFET) at 4 K. The electrical characteristics, including the gate leakage current and the subthreshold transconductance, and the input-referred noise voltage for a new lot of discrete CHFETs is presented and discussed. It is shown that the inclusion of a sidewall spacer on the gate substantially reduced the gate leakage current, as compared to a previous lot without the sidewall spacer. The input-referred noise is approximately the same order of magnitude as previous devices, on the order of 1 (mu) V/(root)Hz at 10 Hz for subthreshold operation. The noise is relatively unaffected by changes in the bias current and drain voltage, but decreases with increasing device size, and is increased by the inclusion of dopants in the channel region. Several simple multiplexer circuits using CHFETs are presented, and the open-loop transfer curve of a multiplexed single gain stage operational amplifier at 4 K is shown.

The newest 256 X 256 InSb arrays (CRC-590 and CRC-463 multiplexers) from Santa Barbara Research Center have reached a milestone in low noise performance. Using the Fowler-sampling technique to acquire data, we have achieved 10 - 13 e^- multiply sample read-out (MSR) noise with the new arrays. With this remarkably low noise performance, background limited performance occurs at relatively small signal levels, viz. a few percent of full-well depth. The signal-to-noise capability of infrared detectors operating in both read noise and background limited performance is an important parameter for evaluating the efficacy of various data sampling methods. We conclude that both line-fitting and Fowler- sampling are the best sampling methods covering the whole performance regime, providing large improvement in the read-noise dominated regime and approaching CDS capability in the photon-noise dominated regime.

The primary purpose of this paper is to present the theory of the pixon, its role in Bayesian image reconstruction, and a new method for selecting a pixon basis using fractal dimension concepts: the Fractal Pixon Basis (FPB). In addition, we review our recent work involving the Maximum Residual Likelihood (MRL) criterion for the goodness of fit (GOF) and the Uniform Pixon Basis (UPB). The MRL GOF statistic is based on the autocorrelation of the residuals and eliminates spatially correlated residuals which commonly occur in image reconstruction methods. The presence of spatially correlated residuals causes photometric inaccuracy. Consequently, photometry is greatly enhanced by using the MRL statistic. We also show that through the use of the UPB image representation, a 'Super-Maximum Entropy' solution can be obtained in which entropy is maximized exactly. We present reconstructions obtained with the above methods and compare them to those obtained with the best Maximum Entropy algorithms and demonstrate that use of UPB/MRL concepts provides consistently superior solutions. Finally, we present reconstructions which demonstrate that the FPB obtains the best results so far, significantly superior to both UPB/MRL and ME-based reconstructions.

NASA Ames Research Center and Lick Observatory have jointly developed two prototype infrared camera for astronomical imaging. The new cameras have been built using Amber Engineering Inc. 128 X 128 SiGa and InSb focal plane arrays. We report observations of the BN-KL source complex in M42 and HD44179 which are used as a demonstration of the cameras' performance when used as narrow band spectral images. The images have a dynamic range in excess of 10³. The high signal to noise ratio and linearity of the images are used to demonstrate the power of spectral differencing as a tool for astronomical image enhancement at mid infrared wavelengths. The observations were made over a range of wavelengths 3.1 - 3.3 micrometers and 8.4 - 11.5 microns in and out of the silicate absorption band, and in the PAH emission bands at 3.3 micrometers and 11.3 micrometers using approximately 2% spectral resolution. Bell and Crisp, 1991, have reported spatial resolution enhancement of images for the surface of Mars by division of multispectral images in and out of the H₂O absorption feature at 3 micrometers . We find that a similar image processing technique using normalized image differencing can be used in the 10 micrometers silicate absorption feature and PAH emission bands to improve the contrast in mid-infrared astronomical images. This technique is expected to be applicable to imaging of galactic nuclei and other astronomical objects as well as embedded stellar infrared sources.

A room temperature camera based on a 128 X 128 pixel InGaAs focal plane array (FPA) is described. The photodiode array (PDA) was a backside-illuminated device with an In_.53Ga_.47As epitaxial active layer deposited on an InP substrate. The PDA was bump- bonded to a silicon CMOS readout multiplexer. The FPA was sensitive from below 0.9 micrometers (the cutoff of the substrate) to 1.7 micrometers (the bandgap of the active layer). At room temperature, the mean detectivity of the FPA, D^*_(lambda pk), exceeded 10¹³ cm-(root)Hz/W and increased to 3.5 X 10¹⁴ cm-(root)Hz/W when cooled to 230 K. Potential applications for this wavelength band are discussed.

The Instituto de Astrofisica de Canarias (IAC) is undertaking the construction of an IR camera for astronomical use at the 1.5 meter (f/13,8) Carlos Sanchez IR Telescope (CST), sited at the Observatorio del Teide (Tenerife). The camera will employ a 256 X 256 InSb focal plane array, and will be used in the 1 - 5 micron atmospheric windows. The Camera uses an optical reimaging system which maps 0.5 square arcseconds of sky per pixel. The optical system will be diamond turned in aluminum and mounted in such a way that the optical alignment is facilitated. Two filter wheels will accommodate 14 broad and narrow band filters. A SUN SPARCstation will control the camera and allow data handling and displaying of the images. With this configuration we expect to achieve sensitivities of 17 and 12.5 magnitude (3 (sigma) in 10 sec) at the K and L band respectively.

MIRAC is a Mid InfraRed Array Camera built for ground-based astronomy by Steward Observatory at the University of Arizona, Harvard Smithsonian Center for Astrophysics, and the Center for Advanced Space Sensing at the Naval Research Laboratory. It utilizes a Hughes Aircraft Co. 20 X 64 pixel arsenic-doped silicon impurity band conduction hybrid array with a capacitance transimpedance amplifier readout (CRC 444A) operating at 10 K in a liquid helium-cooled cryostat. It has an operating wavelength range of 2 to 26 micrometers . Using 20 parallel readout lines and frame rate of 10 KHz, the array exhibits both low noise and good linearity at high background flux, which is essential for 10 and 20 micrometers ground-based observing conditions. It has a peak quantum efficiency of 0.42 at 22 micrometers , and a well size of 120,000 electrons. MIRAC has been operated on the Steward Observatory 2.3-meter and NASA 3-meter IRTF telescopes a number of times for observing a variety of objects including infrared-luminous galaxies, planetary nebulae, star forming regions, and young stellar objects. The NEFD of MIRAC on the IRTF at 11.7 micrometers is 0.1 Jy/square-arcsec in one second of integration on-source, four seconds total time, including nodding and chopping off-source.

The ESO Very Large Telescope will be equipped with a complementary suite of instruments for infrared imaging and spectroscopy between 1 and 20 micrometers . ISAAC (Infrared Spectrometer and Array Camera) is being developed by ESO and is scheduled to be commissioned during 1997 on the first of the 8 m unit telescopes where it will provide for 1 - 5 micrometers imaging and long slit, low and medium resolution spectroscopy. Its design provides for separate 1 - 2.5 micrometers and 2.5 - 5 micrometers cameras equipped with large format array detectors which can be used either to re-image the telescope focal plane or the intermediate spectrum formed by a grating spectrometer. HgCdTe and InSb arrays of 256 X 256 pixels are currently baselined but the possibility of accommodating larger arrays of up to 1024 X 1024 pixels is foreseen if and when these become available. The complete instrument will be cryogenically cooled using two stage closed cycle coolers and housed in a approximately 1 m diameter vacuum vessel attached to the adapter/rotator at one of the Nasmyth foci. Subject to satisfactory results of prototype tests now in progress it is planned to use diamond turned metal mirrors for the reflecting optics and cryogenic stepper motor drive systems for all moving functions.

In March, 1988 the Science Steering Committee of the California Association for Research in Astronomy selected four instruments which were to be built by the time the Keck Telescope received first light. A long wavelength infrared camera was one of the instruments chosen and is described here. The camera was divided into two parts, (1) the detector, the optics and the dewar; (2) the electronics and the computers.

An infrared, cryogenically-cooled, grating spectrometer has been designed for the Columbus Project (2 X 8.4-m telescopes) and MMT Conversion (6.5-m). On one barrel of the Columbus Telescope and using a NICMOS3 array of 256 X 256 40 micrometers HgCdTe detectors, the instrument will project each pixel to 0.33 arcsec. With a slit of 0.66 arcsec width (2 pixels), the available spectral resolutions will range from (lambda) /(Delta) (lambda) equals 670 to 19,000. The optics are achromatic from 1.4 to 5 micrometers , allowing use of a variety of array types. The first version of this instrument has been built and fitted with optics that allow its use with the Steward Observatory 1.5-m and 2.3-m telescopes. It is relatively inexpensive (< $DOL400 K) and compact (approximately 0.3 m³). The high spectral resolution in such a compact instrument will be achieved through an echelle grating immersed in silicon. We discuss the processing for producing such gratings, including demonstrations that we have conducted on test blanks. We report on the preliminary performance of the prototype instrument and on unique design features that may be useful for other spectrometers.

The design of a multipurpose 1 - 5.5 micrometers infrared camera (NSFCAM) for the NASA Infrared Telescope Facility (IRTF) on Mauna Kea, Hawaii, is described. The camera is built around the new 256 X 256 InSb array manufactured by Santa Barbara Research Center (SBRC) and incorporates a variety of observing modes to fulfill its role as a major facility instrument. These include three remotely-selectable image scales, a selection of fixed bandpass filters, R equals 50 - 100 spectral resolution circularly variable filters, a grism, coronographic masks, and a polarization imaging capability. Through the use of flexible array clocking schemes, driven by programmable digital signal processors (DSPs), we plan to implement several new operating modes, including real-time shift and add for image stabilization, and fast subarray readouts for occultations. Simultaneous optical and infrared imaging of the same field will be possible through the use of a cold dichroic beamsplitter. This feature is primarily intended for use with the IRTF tip-tilt image stabilization system currently being built. Given a suitable guide star, the camera should achieve near-diffraction limited imaging at 2 - 5 micrometers . In this paper we discuss the design of the optics, cryogenic, electronics and software needed to provide the camera with these capabilities.

As part of the Canada-France-Hawaii Telescope's (CFHT) New Imaging Program we recently fabricated a pair of 1.0 - 2.5 micrometers facility cameras, known as 'Redeye'. Each camera uses a Rockwell NICMOS3 Hg:CdTe array with 256 X 256 pixels. The two cameras are virtually identical in all respects except one houses 1.7:1.0 reimaging optics, while the other houses 0.7:1.0 reimaging optics. The cameras were commissioned in January 1993 and we anticipate developing several new observing modes in the near future.

This paper describes a new infrared imaging system being developed at UCLA for use on both the Lick Observatory 3-m telescope and the W.M. Keck 10-m telescope. The instrument has a relatively wide field of view on each telescope and is intended for infrared surveys and deep imaging. To enhance efficiency, the new instrument incorporates a dichroic beam splitter to provide two simultaneous imaging systems, one short-wave (SW) from 1 - 2.5 micrometers and one long-wave (LW) from 2 - 5 micrometers . Each wavelength channel is independently optimized. The SW channel contains a Rockwell NICMOS3 256 X 256 HgCdTe array and the LW channel has an SBRC 256 X 256 InSb array. The thermal design employs a closed cycle cooler. A control and data acquisition system based on transputers and high speed analog electronics is being developed to handle the high data rates.

In this paper we describe the main subsystems that constitute the Mt. Palomar Prime Focus InfraRed Camera (PFIRCAM), together with some of the characterization data obtained for the focal plane array. This camera is currently a facility instrument at the 200-inch Mt. Palomar Observatory. It helps to satisfy the observational needs of astronomers in the spectral range of 1 micrometers to 2.5 micrometers by utilizing a HgCdTe NICMOS3 array. The camera has a plate scale of 0.54 arcsec/pixel for an overall FOV of 138 X 138 arcsec.

First light with the advanced cooled grating spectrometer (CGS4) was achieved at the United Kingdom Infrared Telescope on February 4, 1991 following successful delivery of the instrument from the Royal Observatory, Edinburgh. We discuss the performance of CGS4 and summarize our experience in maintaining optimum array sensitivity. CGS4 is unique in that both the data acquisition and reduction can be almost completely automated, and the key elements of the software and their impact on observing are described. We discuss how various aspects of CGS4 such as the reproducibility of flat fields relate to the ability to provide users with flat-fielded, sky-subtracted spectra almost in real-time. We also discuss the problems of the variability of OH line emission and atmospheric transmission and describe the sky subtraction techniques which we have been using both at the telescope and in post observing analysis.

We describe ALICE (Array Limited Infrared Control Environment), the new array control and data acquisition system being developed at the Royal Observatory, Edinburgh, for use with instruments at the UK Infrared Telescope (UKIRT) on Mauna Kea, Hawaii. The first ALICE systems are described to control Santa Barbara Research Center (SBRC) 256 X 256 indium antimonide arrays to be installed in a near-IR camera (IRCAM) and the IR spectrometer (CGS4) at UKIRT, ALICE, however, may be reconfigured for other sizes and formats of array, its use of parallel processing techniques helping to make it a powerful, flexible and extensible system.

This paper describes an infrared imaging system developed to match the physical pixel sizes of the near infrared arrays used in the UCLA IR camera to the typical seeing disk at the f/15 focus of the W. M. Keck telescope. A field of view of 64 arcseconds (') at 0.25' per pixel is required, no internal focussing mechanism is possible and the system must perform under vacuum at 77 K. In our design, an achromatic triplet lens of calcium and barium fluoride provides an image of the entrance pupil and collimates the beam before it passes through a dichroic beam-splitter which divides the system into two independent channels; the wavelength split occurs at approximately 2.5 micrometers . Each beam is re-imaged onto IR arrays with 30 and 40 micrometers pixels to yield 0.25'/pixel and each imager is independently optimized to be achromatic and aberration free. Spot diagrams and aberration plots will be given. We also describe the steps required to compensate for environmental changes, since these lenses are used at LN₂ temperature, and we discuss AR coatings and throughput.

The design of an OH airglow suppressor spectrograph for use on the University of Hawaii 2.2 m telescope is presented. The unique feature of the pre-optics system for low resolution spectroscopy in the 1.1 to 1.8 micrometers range is the capability of removing most of the intense OH emission lines by a specially designed spectroscopic mask. With the OH suppressor spectrography, the background flux is reduced to about 1/30 the natural background on the average. The sensitivity gain in terms of limiting magnitude is expected to be approximately 1.5 mag, compared with the conventional method.

An infrared liquid nitrogen (LN2) cooled focal reducer camera has been designed to be used at the 1.5 m (f/13.8) Carlos Sanchez IR Telescope (CST) at the Observatorio del Teide. The image quality is better than 30 micrometers (typical pixel size of the detector) with an image scale of 0.5 arcsec/pixel. The final design is a very compact system which consists of a centered reflective system with a CaF₂ lens which will be used as window of the dewar. The system is free of chromatic aberration, reduces the amount of coma presents in the telescope and allows us to get a homogeneous image quality over the detector. Vignetting effects on the transmission are studied and found below the 20%. The use of aluminum diamond turned optics lets us obtain the tight tolerances needed to get the maximum optical performance with a system which is practically insensitive to temperature changes. Spot diagrams at different wavelength from 1 to 5 micrometers using a real simulation of the complete system (CST + camera) are presented. The baffling of the system is also analyzed.14

ARNICA (ARcetri Near Infrared CAmera) is the imaging camera for the near infrared bands between 1.0 and 2.5 micrometers that Arcetri Observatory has designed and built as a general facility for the TIRGO telescope (1.5 m diameter, f/20) located at Gornergrat (Switzerland). The scale is 1' per pixel, with sky coverage of more than 4' X 4' on the NICMOS 3 (256 X 256 pixels, 40 micrometers side) detector array. The optical path is compact enough to be enclosed in a 25.4 cm diameter dewar; the working temperature is 76 K. The camera is remotely controlled by a 486 PC, connected to the array control electronics via a fiber-optics link. A C-language package, running under MS-DOS on the 486 PC, acquires and stores the frames, and controls the timing of the array. We give an estimate of performance, in terms of sensitivity with an assigned observing time, along with some details on the main parameters of the NICMOS 3 detector.

A low-cost solution to the problem of cryogenic motors for infrared instrument is presented. Stepper motors have been modified to work in a cryogenic environment (T approximately equals 80 K) and laboratory tests have been performed to estimate both wear and torque. The modifications to the motors and the results of the measurements on torque of two different motors are reported.

This paper describes the basic design, operation, and initial performance of MAGIC, the new MPI fur Astronomie General-purpose Infrared Camera. MAGIC uses a 256 X 256 NICMOS3 HgCdTe detector array and has flexible optics and drive electronics that permit a variety of observing configurations. The camera was designed and built to MPIA specifications by Infrared Laboratories of Tucson, Arizona. MAGIC is based at the 3.5 meter telescope on Calar Alto, although it may be used at a number of other sites, including the 2.2 and 1.2 meter telescopes on Calar Alto and the 2.2 meter MPIA/ESO telescope at La Silla. The design of MAGIC places particular emphasis on wide field, deep imaging at the f/10 focus of the 3.5 m telescope and on providing some spectroscopic and speckle interferometric capability.

UCSD's IR astronomy group is building an imaging mid-IR spectrometer for the Keck Telescope. This instrument, the Long-Wavelength Spectrometer (LWS), is built around a 96 X 96 element, Si:As impurity band conduction array built by GenCorp Aerojet Electronics Systems Division. The LWS has low and moderate spectroscopy modes with nominal spectral resolutions of R (equals (lambda) /(Delta) (lambda) ) equals 100 and 1400 respectively, operating in the 10 micrometers (second order) and 20 micrometers (first order) ground- based atmospheric spectral windows. The LWS is also capable of direct imaging from 5 micrometers to 27 micrometers through a selection of 16 filters. For each of the spectroscopic modes and the direct imaging mode, the plate scale is 0.12 arcsec/pixel, which Nyquist samples the telescope's diffraction pattern at 10 micrometers . Because of the large light gathering power of the Keck Telescope and it's small diffraction pattern, the LWS will have unparalleled point source sensitivity, making it the premiere instrument for extragalactic and general faint-source mid-IR spectroscopy.

An immersion grating with a high refractive index, n, increases the spectral resolution by a factor n over that of a reflective surface grating of equal length. A silicon immersion grating has been developed and tested; initial results are reported here.

We present the optical design for Michelle, the UK Infrared Telescope's next major spectrometer which is now under construction at the Royal Observatory Edinburgh. Operating in the 10 and 20 micrometers atmospheric windows, Michelle will be capable of taking spectra at resolving powers ranging from 40 to 20,000 over a region of sky an arcminute long and varying in width from 0.75 arcseconds upwards. This high degree of versatility reflects Michelle's role as the U.K. astronomy communities sole long-slit mid-infrared spectrometer and as a primary tool for following up on the discoveries of the ISO satellites.

The Coude focus of the first VLT 8 m-telescope unit will be equipped with a high resolution near-infrared camera (CONICA). This multimode instrument will be specialized in making use of the high spatial resolution offered by the Adaptive Optics System of the VLT. It is being developed by the Max-Planck-Institut fur Astronomie (Heidelberg), the Max-Planck-Institut fur Extraterrestrische Physik (Garching) and the Osservatorio Astronomico de Torino (Italy). CONICA will provide the operation of two detector arrays, one for low background application between 1 and 2.6 micrometers and a second one which is sensitive in the whole 1 - 5 micrometers region but will be used mainly in the thermal near infrared. The pixel size is adapted to diffraction limited resolution (between 10 mas and 100 mas per pixel) by variable camera magnifications. The various observational modes of polarimetry (Wollaston prisms, wiregrid analyzer), low resolution spectroscopy (grisms) and coronographic masking techniques can be combined independently with speckle application or the use of the Adaptive Optics System.

We have implemented an infrared array camera for high-background, ground-based astronomical imaging in the mid-infrared spectral region from 5 to 27 microns. The camera is built around a 20 X 64 element Si:As Impurity Band Conduction (IBC) device manufactured by GenCorp Aerojet Electronic Systems Division. Observations with the camera on a 1.5 meter telescope (Mt. Lemmon Observing Facility, Tucson, Arizona) yield a noise- equivalent flux density (NEFD) of 23.5 mJy min^-1/2 arcsec^-2 at (lambda) equals 10 micrometers , (Delta) (lambda) equals 1 micrometers , and a readout frame rate of 366 Hz. We discuss the design and implementation of the camera and operational procedures for observation in a high-background environment (e.g. background subtraction and flat-fielding). We also present sample images obtained with our camera.

TIMMI is the ESO Infrared camera dedicated to 10 micrometers high angular resolution imaging of the austral sky. This camera, built for the European Southern Observatory (ESO) by the Service d'Astrophysique at Saclay (SAp), has been successfully commissioned during 2 observing runs at the ESO 3.6-m telescope: one in July 1992 and the other in January 1993. Based on a LIR 64*64 pixel Si:Ga/DVR detector array optimized for ground-based broad-band 10-micrometers astronomical observations, the camera, operated at a frame rate of 120 Hz, has achieved a noise equivalent flux density of 0.01 Jy min^-1/2 pixel^-1 (1 (sigma) ) during N₂-band observations with a pixel field of view of 0.48 arcsec. Two other fields of view are available: 0.32 and 0.65 arcsec. A filter wheel allows to select between 14 broad-band and narrow-band filters covering the atmospheric window (8 - 13 micrometers ); a M filter allows observations in the 4 micrometers atmospheric window with substantial sensitivity. The instrument is now available to visiting astronomers.

As part of the NASA Space Infrared Telescope Facility (SIRTF), a low noise multiplexer has been developed. The hybridization of this multiplexer to a high quality indium antimonide (InSb) photodiode array has resulted in a MWIR detector of outstanding performance. The multiplexer is made of a 256 X 256 array of source follower amplifiers on a 30 micron square pitch. Random access binary decoders are used to access each pixel of the array, allowing any read-out scheme to be implemented. Dark current has been measured at temperatures ranging from 4 K to 77 K. Generation recombination currents dominate above 45 K. With 100 mV of reverse bias applied, less than 3 X 10^-17 A is typical below 50 K with 8 X 10^-19 A (5 electrons/second) at 4 K. Under the same conditions 0.25 pA was measured at 77 K. Read noise has been measured as low as 186 electrons using non-correlated techniques. Detector quantum efficiency is 50 to 80 percent through the entire wavelength band of 1 to 5 micrometers .

Fabry-Perot interferometer (FPI) is an important instrument to study the composition and dynamics of the mesosphere and thermosphere. Despite the inherent large throughput of FPI, its sensitivity and applications has been largely limited by the use of low quantum efficiency and limited spectral range photomultiplier tubes. The high quantum efficiency CCD has gained wide acceptance in recent years. CCD is almost an ideal detector, it has high quantum efficiency and wide spectral range from UV to NIR, but its square format is not compatible with the circular fringe pattern generated by the FPI. In this paper we will show a new focal plane detection technique called the circle to line interferometer optical (CLIO) system. CLIO converts the circular FPI fringes into linear fringes and makes the CCD more applicable to Fabry-Perot interferometer. The design of an advanced FPI based on CLIO and modern back- thinned, AR-coated high quantum efficiency CCD operating in the MPP mode will also be presented.

This paper provides a brief description of the development of adiabatic demagnetization refrigerators (ADRs) at Berkeley for cooling bolometric infrared detectors. This development was stimulated by the needs of the Multi-Band Imaging Photometer (MIPS) for the NASA Space Infrared Telescope Facility (SIRTF). A description will be given of the ways in which classical ADR technology has been developed to provide a compact refrigerator to meet the requirements of this space experiment. Similar refrigerators will be useful in a number of forthcoming space missions for cooling both infrared and X-ray bolometers.

ESO has recently installed a new infrared array camera IRAC2 which is equipped with a large format NICMOS3 256*256 Hg_1-xCd_zTe array detector. The performance of the instrument and the detector array will be discussed briefly. A new nondestructive readout scheme of the array will be presented which allows first order wavefront corrections of images degraded by atmospheric seeing. During the stare time the integration ramp of the detector signal is sampled every 100 msec. A regressional fit of the sample data points yields the slope of the integration ramp which is proportional to the flux received by a detector pixel. To this readout mode which is commonly used for IR arrays a small software module can be added to compensate the image motion of the observed object by shifting the nondestructively sampled images. This has the same effect as a tip tilt correction by an active optical element--but without the extra complexity of such a device. Lab tests and first results obtained at the telescope are presented.

The Ohio State Infrared Imager/Spectrometer (OSIRIS) is a general purpose near infrared (0.9 to 2.5 micrometers ) instrument that can be used at a wide variety of telescope focal planes. OSIRIS currently uses a 256 X 256 HgCdTe array detector and will accommodate larger arrays when available. OSIRIS has two modes of operation: imaging and spectroscopic. This paper describes the general instrument design and sample scientific results.

New linear multiplexed focal plane arrays have been developed for the AVIRIS (Airborne Visible and Infrared Imaging Spectrometer) instrument. The new arrays are designed to improve the AVIRIS mission placing emphasis on greater sensitivity and lower noise. An order of magnitude improvement in sensitivity is achieved primarily by use of a higher performance input circuit, and to a lesser degree, by maximizing detector photoresponse in region where the terrestrial signal is weakest. The silicon detectors were fabricated using a process that enhances their blue response, and the indium antimonide (InSb) detectors were fabricated with carefully selected spectral anti-reflection coatings. The multiplexer also incorporates a 'snapshot' mode eliminating the 'spectral smear' inherent with the previous design.