The ambitious science goals of the James Webb Space Telescope (JWST) have driven spectacular advances in λco ~ 5um detector technology over the past five years. This paper reviews both the UH/RSC team’s Phase A development and evaluation of 2Kx2K arrays exceeding the detector requirements for JWST’s near infrared instruments and also the hardware integration of these into a 4Kx4K (16Mpxl) close packed mosaic focal plane array housed in an Ultra Low Background test facility. Both individual first generation 2Kx2K SCA’s and 4Kx4K mosaic focal planes have been extensively characterized in the laboratory and, since September 2003, a NIR camera utilizing the 4Kx4K mosaic focal plane has been in use for nearly 100 nights at the UH 2.2 m telescope on Mauna Kea. Typical test results for the first generation 2Kx2K arrays and their integration into 4Kx4K mosaic focal planes are reported. Demonstration of the design concepts and both array and mosaic focal plane performance in actual hardware, as described here, has provided the foundation for design iterations leading to later generations of 2Kx2K arrays and 4Kx4K mosaic focal planes. Four major technology developments leading to first generation hardware demonstrations of both 2Kx2K SCA’s and a 4Kx4K mosaic FPA are reviewed. These are: 1) improvement in test equipment and procedures to characterize the detectors against JWST requirements and goals, primarily at 37K but with the capability to test from 30K to 100K; 2) optimization of λc ~ 5 um MBE HgCdTe material on a CZT substrate for low dark current (goal of 0.003 e-/sec at 37K) with high quantum efficiency, low cross-talk and greatly reduced image persistence; 3) development of the 2Kx2K HAWAII-2RG multiplexer designed specifically to take full advantage of these detector characteristics for a wide range of astronomical applications (and fully compatible with an ASIC controller developed under the JWST Instrument Technology Development initiative) and 4) development of molybdenum SCA carriers allowing modules to be close-butted on three sides and easily installed onto a molybdenum plate to form a 4Kx4K mosaic focal plane. We describe both the improvements in the KSPEC test facility and in test procedures for individual 2Kx2K arrays and the Ultra Low Background (ULB) test facility developed specifically to evaluate 4Kx4K mosaic focal plane assemblies required for the NIRCam instrument. The laboratory test configuration of the ULB facility utilizes multiple shields and internal light sources to achieve background fluxes <1 photon/hour per pixel for λc ~ 5um while providing temperature stability <1mK over periods of weeks. An alternate configuration utilizes fore optics to allow the mosaic FPA module of the ULB facility to be mounted at the Cassegrain focus of the UH 2.2 meter telescope, providing an image scale of 0.25”/pixel over a 17’x17’ field. A cold PK 50 lens cuts off around 1.7 um, limiting the background at wavelengths below 1.65 um (where the array can be used with normal filters and where narrow band filters reduce the background to levels comparable to NIRCam on JWST). Observations at the telescope, which provide the best way of verifying certain JWST requirements and allow direct astronomical characterization of the detectors, are reported.

Careful measurements for an engineering grade 2K² 2.5μm cut-off VIRGO detector in a sealed, cold enclosure have yielded dark current twenty five times less than previously reported for these devices, putting Raytheon detectors in contention for low background applications. Global reset followed by Sample Up the Ramp readout was used to allow zero point drifts at the exposure start to be separated from true dark current and mux glow. In a sub-array selected to be far from two photo-emitting defects, mean dark current stabilized 10 hours after power-on, at 0.025 e-/s/pix at 79K (-0.1K, +0.8K). Dark current showed no evidence of the onset of a floor in the 79-104K range, but the temperature dependence was softer than expected, implying a band gap that is 20% below nominal. Shot noise from the dark current will not dominate the 18 e- read noise, unless a substantial noise reduction is achieved through multiple sampling. Hundreds of non-destructive samples will be possible without impact from multiplexor glow, which contributes only 0.04e-/pix/read for a 6μs pixel. Reset Anomaly is dependent on exposure time, settling to -60e- for long exposures, and dominating dark current in exposures less than 2400s. Reference pixels do compensate for these effects, but imperfectly, requiring further study. Attempts to explain Reset Anomaly in terms of self heating were inconclusive.

VISTA awarded a contract to Raytheon Vision Systems in June 2002 to supply 16 SWIR HgCdTe 2Kx2K detectors. Raytheon has delivered VIRGO 2K x 2K readout multiplexers, engineering detectors and the first two science grade detectors. The UKATC has set up a low background test facility to test and characterize the VIRGO detectors and to confirm that the detector performance meets our specifications. The VIRGO 2Kx2K is a new detector being produced for VISTA by Raytheon. In this paper we present the first results and performance of the multiplexer and the science detector. The test facility includes a custom built low background close-cycle cooled cryostat, cryogenic pre-amplifier electronics and uses ESO’s Infrared array controller electronics and detector control and data acquisition software. The detector parameters being measured include trans-impedance conversion gain, quantum efficiency in J, H and K wave bands, read noise, dark generation rate, linearity, well capacity, pixel operability, drift with temperature, persistence and electrical cross-talk.

The NIRCam instrument will fly ten of Rockwell Scientific’s infrared molecular beam epitaxy HgCdTe 2048x2048 element detector arrays, each the largest available with current technology, for a total of 40 Megapixels. The instrument will have two varieties of MBE HgCdTe, a SWIR detector with λ_co = 2.5 μm, for the shortwave channel of NIRCam (0.6-2.3 μm); and a MWIR detector with λ_co = 5.3 μm, for the longwave channel of NIRCam (2.4-5.0 μm). Demonstrated mean detector dark currents less than 0.01 electrons per second per pixel at operating temperatures below 42 K for the MWIR and below 80 K for the SWIR, combined with quantum efficiency in excess of 80 percent and read noise below 6 electrons rms, make these detector arrays by far the most sensitive SWIR and MWIR devices in the world today. The unique advantages of molecular beam epitaxy as well as FPA data on noise, dark current, quantum efficiency, and other performance metrics will be discussed. In addition, the focal plane assembly package designs will be presented and discussed.

VLT instruments increasingly require high sensitivity large format focal planes. Adaptive optics combined with multiple integral field units feeding high resolution spectrographs drive the pixel performance as well as the array format. Three VLT instruments, the wide field imager Hawk-I and the integral field spectrographs SINFONI and KMOS will be equipped with MBE-grown HgCdTe Hawaii-2RG arrays, which have a cut-off wavelength of 2.5 micron. The Hawaii-2RG array was originally developed for the near infrared camera of JWST having a cut-off wavelength of 5 micron. The Hawaii-2RG multiplexer is one of the most advanced readout architectures offering a large variety of operating modes. A special 32 channel package has been developed which allows reading out all 32 output channels of the detector in parallel. Symmetric cryogenic CMOS operational amplifiers are placed next to the focal plane instead of using ASIC’s which are not yet available. The internal bus of the detector is accessed directly, bypassing the on-chip buffer amplifier. Noise performance employing different techniques of using reference pixels is discussed. Basic performance characteristics of the Hawaii-2RG arrays will be presented. Unlike LPE arrays, which lose quantum efficiency at lower temperatures, MBE arrays with λ_c = 2.5 μm do not show this effect. However, the MBE arrays under test still suffer from persistence.

Raytheon Vision Systems is under contract to develop 2K × 2K InSb Focal Plane Arrays (FPA) for the ORION and NEWFIRM projects teaming with NOAO, NASA, and USNO. This paper reviews the progress in the ORION, NEWFIRM, and the JWST projects, showing bare mux readout noise at 30 K of 2.4 e- and InSb dark current as low as 0.01 e-/s. Several FPAs have been fabricated to date and the ongoing improvements for the fabrication of FPAs will be discussed. The FPA and packaging designs are complete, resulting in a design that has self-aligning features for ease in FPA replacement at position of the focal plane assembly with alignment accuracy in the focus direction of ± 12 μm. The ORION/NEWFIRM modules are 2-side buttable to easily form 4K × 4K mosaics while the Phoenix modules, developed under the JWST development program, are 3-side buttable for ease in forming 4K × 2NK mosaics where N can be any integer. This paper will include FPA QE, dark current and noise performance, FPA reliability, and module-to-module flatness capabilities.

The demand for large-format near infrared arrays has grown for both ground-based and space-based applications. These arrays are required for maintaining high resolution over very large fields of view for survey work. We describe results of the development of a new 2048 × 2048 HgCdTe/CdZnTe array with 20-micron pixels that responds with high quantum efficiency over the wavelength range 0.85 to 2.5 microns. With a single-layer anti-reflection coating, the responsive quantum efficiency is greater than 70% from 0.9 micron to 2.4 microns. Dark current is typically less than 4 e-/sec at 80 K. The modular package for this array, dubbed the VIRGO array, allows 3-side butting to form larger mosaic arrays of 4K × 2nK format. The VIRGO ROIC utilizes a PMOS Source Follower per Detector input circuit with a well capacity of about 2 × 10⁵ electrons and with a read noise of less than 20 e- rms with off-chip Correlated Double Sampling. Other features of the VIRGO array include 4 or 16 outputs (programmable), and a frame rate of up to 1.5 Hz in 16-output mode. Power dissipation is about 7 mW at a 1 Hz frame rate. Reset modes include both global reset and reset by row (ripple mode). Reference pixels are built-in to the output data stream. The first major application of the VIRGO array will be for VISTA, the United Kingdom’s Visible and Infrared Survey Telescope for Astronomy. The VISTA focal plane array will operate near 80 K. The cutoff wavelength of the HgCdTe detector can be adjusted for other applications such as SNAP, the Supernova/Acceleration Probe, which requires a shorter detector cutoff wavelength of about 1.7 microns. For applications which require both visible and near infrared response, the detector CdZnTe substrate can be removed after hybridization, allowing the thinned detector to respond to visible wavelengths as short as 0.4 microns.

Large format imaging detectors are required in many modern astronomical optical systems. With the increase in aperture size of large telescopes, the associated cameras for imaging and spectroscopy are much larger than those of a decade ago. Large physical format detectors are required to make full use of these cameras. The detector of choice has been the charge coupled devices (CCDs), although large format 4kx4k CMOS imagers have also been fabricated. We discuss recent developments in 4kx4k pixel imagers, typically with 15 micron pixels, which are over 60 mm per side. Several companies have produced such devices with the characteristics required for astronomy. Backside processing issues are discussed, including results from optimization efforts at the University of Arizona Imaging Technology Laboratory. We also discuss the use of such imagers in the 8kx8k 90Prime prime focus mosaic camera now in operation at the Steward Observatory 2.3 m telescope.

The status of CCD development efforts at Lawrence Berkeley National Laboratory is reviewed. Fabrication technologies for the production of back-illuminated, fully depleted CCD's on 150 mm diameter wafers are described. In addition, preliminary performance results for high-voltage compatible CCD's, including a 3512 x 3512, 10.5 μm pixel CCD for the proposed SuperNova Acceleration Probe project, are presented.

CMOS-based hybrid silicon focal plane array technology is presented as a high-performance CMOS sensor alternative to CCD technology for future space missions and ground-based telescopes. This paper will discuss the unique performance advantages of the hybrid CMOS arrays, including the very high quantum efficiency from UV to near IR, good spatial resolution at moderate voltage bias, readout commonality with IR detector channels in multi-spectral systems, low noise, low power dissipation, high inherent radiation tolerance, and excellent CMOS functionality afforded by the separately optimized readout circuitry. The ability to retain low noise at high video rates and the fact that CMOS sensors do not suffer the charge transfer efficiency (CTE) degradation of CCDs enable an easy scale-up of CMOS-based FPAs to larger formats without compromising sensor performance. The large hybrid CMOS silicon FPAs up to 2048x2048 format in single chip and 4096x4096 format in mosaic configuration that are demonstrated at Rockwell Scientific will be presented.

Electron Multiplying Charge Coupled Devices (EMCCDs) are CCD cameras with potentially single-photon detection ability. Signal amplification is achieved by way of a unique electron-multiplying structure built into the silicon, and the gain can be varied in order to overcome the read-noise floor, which is the usual limiting factor in reading out a conventional CCD at high frame rates. In combination with its high quantum efficiencies, the EMCCD holds great promise for time-resolved photometry. We report here results from two observing campaigns aimed at assessing the suitability of EMCCD technology for detecting short-timescale, low-amplitude variability in blazars. Data were taken on the 2.2m telescope at Calar Alto using both front-illuminated and back-illuminated EMCCD cameras from Andor Technology’s iXon range. Approximately 410,000 science frames were recorded over 10 nights. The results presented here illustrate the photometric stability achieved with the cameras, under typical observing conditions. In general, photometric precision down to the level of a few millimagnitudes is found to be possible. We argue that reliable photometry is best achieved with high data collection rates (typically 4 frames per second) coupled to ultra-low-noise detectors such as the EMCCD.

The Advanced Camera for Surveys (ACS), installed in the Hubble Space Telescope (HST) in March 2002, comprises three cameras: the Wide Field Camera (WFC), designed for deep near-IR survey imaging programs; the High Resolution Camera (HRC), a high angular resolution imager which fully samples the HST full spread function (PSF) in the visible; and the Solar Blind Camera (SBC), a far-UV imager. The WFC and HRC employ CCD detectors. Their performances are affected by the on-going damage due to the space radiation environment where they operate. We present an overview of the performance of the ACS CCD detectors, based on the first two years of flight science operations. We analyze the evolution with time of the basic detector performance that are subjected to degradation due to the on-going radiation damage. Comparison is made with ground testing prediction and with the amount of performance degradation seen in other CCD detectors on board of HST.

The orthogonal-transfer array (OTA) is a new CCD architecture designed to provide wide-field tip-tilt correction of astronomical images. The device consists of an 8x8 array of small (~500x500 pixels) orthogonal-transfer CCDs (OTCCD) with independent addressing and readout of each OTCCD. This approach enables an optimum tip-tilt correction to be applied independently to each OTCCD across the focal plane. The first design of this device has been carried out at MIT Lincoln Laboratory in support of the Pan-STARRS program with a collaborative parallel effort at Semiconductor Technology Associates (STA) for the WIYN Observatory. The two versions of this device are functionally compatible and share a common pinout and package. The first wafer lots are complete at Lincoln and at Dalsa and are undergoing wafer probing.

Recent advances in CMOS read-out integrated circuit (ROIC) design have helped achieve 10e- of read noise with single read and down to 5e- read noise with 32-pair Fowler sampling at slower astronomical frame rates. However, applications like adaptive optics pose even more stringent performance requirements on ROICs for visible and IR focal plane arrays (FPAs). Traditional pixel designs use circuits such as source followers, capacitive trans-impedance amplifiers (CTIA), and buffered-direct injection (BDI) for detector charge integration and readout. Currently these techniques by themselves do not achieve sub 10e- read noise at high readout bandwidths. This paper describes circuit design advances and measured performance that enable ROICs with ultra-low noise readout (3-10e-) at signal bandwidths allowing KHz frame rate on 128x128 and larger arrays. Using deep sub-micron CMOS, high conversion gain has been designed in a small unit-cell area while keeping high bandwidth for reset and readout, and sufficiently low power dissipation to avoid MOSFET self-emission for background-limited sensitivity at ultra-low scene backgrounds. Measured performance of one of the pixel designs reported in detail shows a noise floor of 7e- with HgCdTe detector array, near identical to the design value.

Standard, off-the-shelf, large infrared detector arrays are now available that meet the demanding requirements of the astronomy and civil space communities. This paper describes arrays with more than one million detector elements (1024 × 1024 and 2048 × 2048 formats) developed by Raytheon Vision Systems for these low-background applications. Arrays of this size have been demonstrated with a variety of detector materials: Si PIN, HgCdTe, InSb, and Si:As IBC. All of these detector materials have demonstrated low noise and dark current, high quantum efficiency, and excellent uniformity over wavelengths ranging from visible (optical) to 28 μm. Features of the readout integrated circuits that mate to these detector arrays are also discussed. Summary performance data of each of these materials on arrays with more than one million detector elements are presented. Companion papers in these SPIE proceedings that discuss several of these arrays in more detail are: • "2K × 2K HgCdTe Detector Arrays for VISTA and Other Applications," P. J. Love, A. W. Hoffman, K. J. Ando, D. L. Gulbransen, E. Corrales, N. J. Therrien, J. P. Rosbeck, W. D. Ritchie, R. S. Holcombe. • "1K × 1K Si:As Impurity Band Conduction (IBC) Detector Arrays for JWST MIRI and Other Applications," P. J. Love, A. W. Hoffman, K. J. Ando, A. G. Toth, E. Corrales, N. J. Therrien, W. D. Ritchie, R. S. Holcombe. • "2K × 2K InSb for Astronomy," A. W. Hoffman, E. Corrales, P. J. Love, J. P. Rosbeck, A. M. Fowler, M. Merrill. • "Mosaic Packaging for Large Format Infrared Devices," R. S. Holcombe, A. W. Hoffman, P. J. Love, T. S. Hanson.

Fairchild Imaging has established a unique capability for high volume production of wafer-scale, scientific grade CCD (charge-coupled device) image sensors with active areas ranging from 6 x 6 cm, with 16 million pixels, to 8 x 8 cm devices with 85 million pixels. Large format back-illuminated CCDs are currently in volume production. This paper provides a detailed description of two recent products, and the technologies associated with their development.

This paper discusses the latest technologies for space and ground-based astronomy being pursued by Rockwell Scientific. The discussion covers the latest demonstrated performance of large format NIR (~1.7um cutoff) detectors mated to the HAWAII-2RG readout integrated circuit, our proven readout for large-format arrays for astronomy. Developmental work is presented on the HAWAII-4RG family (consisting of 4k x 4k, 4k x 8k, and 8k x 8k formats), RSC’s newest additions planned to the HAWAII series of astronomy readout integrated circuits. We also present the status of our multifunctional command-and-control ASIC for FPAs, which was first reported at the August 2002 SPIE.

Superconducting tunnel junctions are being developed for application as photon detectors in astronomy. We present the latest results on the development of very high quality, very low critical temperature junctions, fabricated out of pure Al electrodes. The detectors are operated at 50 mK in an adiabatic demagnetisation refrigerator. The contacts to the top and base electrodes of these junctions are fabricated either out of Nb or Ta, which has strong implications on the loss time of the quasiparticles. The Nb contacted junctions show quasiparticle loss times varying between 5 and 80 μsec, depending on the device size. The bias range of the Nb-contacted junctions is limited to the range 0-100 μV, because of the set-in of strong non-equilibrium quasiparticle multiplication currents at higher bias voltages. The Ta-contacted junctions, on the other hand, show quasiparticle loss times in excess of 200 μsec. These long loss times lead to very strong quasiparticle multiplication, which prevents the stable biasing of the junctions even at very low bias voltages. Junction fabrication and characterisation are described, as well as the response of the detectors to monochromatic light with wavelengths varying from 250 to 1000 nm. The energy resolution of the detectors is discussed.

The emergence of curved CCD detectors as individual devices or as contoured mosaics assembled to match the curved focal planes of astronomical telescopes and terrestrial stereo panoramic cameras represents a major optical design advancement that greatly enhances the scientific potential of such instruments. In altering the primary detection surface within the telescope’s optical instrumentation system from flat to curved, and conforming the applied CCD’s shape precisely to the contour of the telescope’s curved focal plane, a major increase in the amount of transmittable light at various wavelengths through the system is achieved. This in turn enables multi-spectral ultra-sensitive imaging with much greater spatial resolution necessary for large and very large telescope applications, including those involving infrared image acquisition and spectroscopy, conducted over very wide fields of view. For earth-based and space-borne optical telescopes, the advent of curved CCD’s as the principle detectors provides a simplification of the telescope’s adjoining optics, reducing the number of optical elements and the occurrence of optical aberrations associated with large corrective optics used to conform to flat detectors. New astronomical experiments may be devised in the presence of curved CCD applications, in conjunction with large format cameras and curved mosaics, including three dimensional imaging spectroscopy conducted over multiple wavelengths simultaneously, wide field real-time stereoscopic tracking of remote objects within the solar system at high resolution, and deep field survey mapping of distant objects such as galaxies with much greater multi-band spatial precision over larger sky regions. Terrestrial stereo panoramic cameras equipped with arrays of curved CCD’s joined with associative wide field optics will require less optical glass and no mechanically moving parts to maintain continuous proper stereo convergence over wider perspective viewing fields than their flat CCD counterparts, lightening the cameras and enabling faster scanning and 3D integration of objects moving within a planetary terrain environment. Preliminary experiments conducted at the Sarnoff Corporation indicate the feasibility of curved CCD imagers with acceptable electro-optic integrity. Currently, we are in the process of evaluating the electro-optic performance of a curved wafer scale CCD imager. Detailed ray trace modeling and experimental electro-optical data performance obtained from the curved imager will be presented at the conference.

Most large telescopes are now implementing sodium laser guide star (LGS) adaptive optics (AO) systems. Most of these systems plan to use the Shack-Hartmann approach for wavefront sensing. In these systems, the laser spots that are imaged in the Shack-Hartmann subapertures suffer spot elongation due to the 10 km extent of the sodium layer. The spot elongation extends radially from the projection point, and increases linearly with the distance the subaperture is separated from the laser. For 8-meter class telescopes with laser projection behind the secondary mirror, the spot elongation is ~1 arc sec at the edge of the pupil, and does not significantly affect the performance of the AO system. However, for the coming generation of extremely large telescopes, sodium LGS spot elongation will significantly degrade the quality of wavefront measurement. Attention should now be given to the development of technologies that can reduce or eliminate the spot elongation problem. The laser spot elongation can be greatly reduced by projecting the sodium laser in a series of short (1-3 μsec) pulses. The Lawrence Livermore National Laboratory (LLNL) has been funded to develop a pulsed fiber laser. In parallel, a new kind of wavefront sensor detector must be developed to properly sense the pulsed laser return. In this paper, we present our project that will develop a novel CCD which is optimized for sensing the return from a pulsed sodium LGS. Our CCD design uses custom pixel morphology that aligns the pixels of each subaperture with the radial extension of the LGS spot. This pixel geometry will allow each subaperture to follow the yellow-orange rabbit (i.e. the 589 nm laser pulse) as it traverses the sodium layer, providing optimal sampling of a limited number of detected photons. This CCD will attain photon-noise limited performance at high frame rates, using MOSFET amplifiers that exist today (2-3 electrons noise). However, we seek even lower noise amplifiers, and as part of our project, we are testing a new generation of JFET amplifiers that may attain sub-electron noise performance. The test CCD will be a standard geometry, 160x160 pixel image area with split frame transfer and a total of 20 readout ports. This test CCD will easily surpass the performance of the CCDs presently in use in astronomical AO systems, and should provide a significant performance improvement in the AO systems of the 8-10 meter telescopes. This project is a collaboration between the Keck Observatory, MIT Lincoln Laboratory, SciMeasure Analytical Systems, Gemini Observatory, Lick Observatory, the University of California, and the Rockwell Scientific Company. Our efforts are being coordinated with the developments at LLNL so that the pulsed laser and novel geometry CCD can be mated together in 3 years when both are fully developed.

We report on the energy resolution of a recently developed superconductor single-photon quantum detector. In a superconducting strip quasiparticles created by a single absorbed photon and a bias current jointly cause a normal domain and, subsequently, a voltage transient that manifests absorption of the photon. Given a constant optical coupling, the mechanism suggests a moderate to good energy resolution in the wavelength range from near-infrared to X-rays. We implemented a meander line from a 80-nm wide and 5-nm thick NbN strip to detect single near-infrared photons with the repetition rate 5•10⁷ sec^-1 and quantum efficiency of few per cent. Although with this detector operated at 2 K we have indeed observed photon-energy dependent detector response, the energy resolving capability appeared smaller than the detector model predicted. We suggest that the inconsistency owes to the influence of the bias current.

Micro-channel plate (MCP) based "Generation 2" image intensifiers are used for initial photon detection and amplification in modern photon counting detectors in astronomy. Input photons initially create photo-electrons in a photocathode. Each photo-electron is then amplified by the MCP or an MCP stack producing a cloud of electrons that are proximity focused onto an output anode. During the electron amplification process charge is stripped from the walls of the associated pore (or pores if an MCP stack is employed) and this needs to be replenished before full gain amplification of following events can take place. Charge depletion then provides the ultimate limitation to the brightness of a point source that can be observed. Experimental work has been undertaken to measure the effect of gain depression in intensified CCD detectors using intensifiers with both P20 and P46 phosphors. The results, along with a discussion on their relevance to detector dynamic range, will be presented.

L3CCDs represent a major step in CCD performance with great potential for astronomical applications because of their ability to work at very high pixel rates with negligible readout noise. This paper describes the results of tests on some of the L3CCDs now available and discusses how the operating conditions may be optimised for a variety of different applications. In particular, at high gain they can be used for photon counting work at photon rates well in excess of one photon per pixel per second. Readout rates which can be as high as 35MHz are entirely practical for a number of astronomical applications. This paper describes some of the compromises and trade-offs that have to be made in designing high-speed controllers to work effectively with these devices. The importance of integrating high-speed controllers for astronomy with significant amount of real-time processing power is also discussed.

Quantum Efficiency (QE) of CCDs decreases at λ >~ 0.7 μm since photons penetrate a depletion layer of CCD. If one makes the layer thicker, the QE will be largely improved. In collaboration with HAMAMATSU Photonics, we have been developing the thicker CCDs which are implemented on the high resistivity n-type silicon wafers. We have made several wafer runs to optimize the basic characteristics of the devices such as charge transfer efficiency (CTE), full-well and node sensitivities of the amplifiers. The results obtained so far mostly satisfied the specifications imposed by astronomical observations. We also attempted to build back-side illuminated devices to realize high QE in wider wavelength. The test devices shows that the QE exceeds 60% at 1 μm, which is roughly 5 ~ 6 times improvement over ordinary CCDs. We will present the current status of the projects.

A GaAs junction field-effect transistor (JFET) is a promising candidate for low noise at low frequency, and low-power cryogenic electronics to read out high-impedance photo-detectors. We report on the spectral noise characteristics, gate leak current, and gate capacitance of a SONY n-type GaAs JFET at a cryogenic temperature of 4.2 K. In our papers reporting performances of FETs, the noise and leak current have been measured separately. However, noise should be measured in the case of the gate terminal surrounded by high impedance devices, so-called in high impedance circumstance. In such high impedance circumstance, the dielectric polarization noise occurs, which is generated in devices and material around the gate terminal. Materials with dielectric loss generate dielectric polarization noise, which cannot be measured in low impedance circumstance because the electric charges compensate the dipoles. Therefore, to estimate the performance of the FETs for cryogenic readout electronics, noise measurement should be done in high impedance circumstance. In our previous work on a GaAs FET, low noise at low-frequency, i.e. ~500 nV/Hz^1/2 at 1 Hz was achieved by the thermal cure (TC) in low-impedance circumstance. By using the thermal curing technique, we have obtained a low noise level of ~500 nV/Hz^1/2 at 1Hz. Moreover, we have demonstrated faint light detection at 4.2 K using an InGaAs pin photodiode and a charge integration amplifier composed of the GaAs JFET. We have succeeded in detecting such ultra low power influx as a few photons per second with a quantum efficiency of ~80% and tolerance of 0.5 photons. The leak current of this detection system is ~500 electrons/hour.

We present the performance of the IR detectors developed for the WFC3 project. These are HgCdTe 1Kx1K devices with cutoff wavelength at 1.7 μm and 150K operating temperature. The two selected flight parts, FPA#64 (prime) and FPA#59 (spare) show quantum efficiency higher than 80% at λ=1.6 μm and greater than 40% at λ>1.1μm, readout noise of ~25 e^- rms with double correlated sampling, and mean dark current of ~0.04 e/s/pix at 150K. We also report the results obtained at NASA GSFC/DCL on these and other similar devices in what concerns the QE long-term stability, intra-pixel response, and dark current variation following illumination or reset.

1K × 1K Si:As Impurity Band Conduction (IBC) arrays have been developed by Raytheon Vision Systems (RVS) for the James Webb Space Telescope (JWST) Mid-Infrared Instrument (MIRI). The devices are also suitable for other low-background applications. The Si:As IBC detectors respond out to ~28 microns, covering an important mid-IR region beyond the 1-5 micron range covered by the JWST NIRCam and NIRSpec instruments. Due to high terrestrial backgrounds at the longer mid-IR wavelengths, it is very difficult to conduct ground-based observations at these wavelengths. Hence, the MIRI instrument on JWST can provide science not obtainable from the ground. A mid-infrared instrument aboard a cryogenic space telescope can have an enormous impact in resolving key questions in astronomy and cosmology. The greatly reduced thermal backgrounds achievable on a space platform (compared to airborne or ground-based platforms) allow for more sensitive observations of dusty young galaxies at high redshifts, star formation of solar-type stars in the local universe, and formation and evolution of planetary disks and systems. We describe results of the development of a new 1024 × 1024 Si:As IBC array with 25-micron pixels that responds with high quantum efficiency over the wavelength range 5 to 28 microns. The previous generation's largest, most sensitive IR detectors at these wavelengths were the 256 × 256/30-micron pitch Si:As IBC devices built by Raytheon for the SIRTF/IRAC instrument. JWST MIRI detector requirements will be reviewed and some model results for IBC device performance will be presented. The IBC detector architecture will be described and the SB305 Readout Integrated Circuit (ROIC), developed specifically for JWST MIRI, will be discussed. The SB305 ROIC utilizes a PMOS Source Follower per Detector (SFD) input circuit with a well capacity of about 2 × 10⁵ electrons. The read noise is expected to be less than 20 e- rms with Fowler-8 sampling at an operating temperature of 7 K. Other features of the IBC array include 4 video outputs and a separate reference output with a frame rate of 0.37 Hz (2.7 sec frame time). Power dissipation is less than 0.5 mW at a 0.37 Hz frame rate. Reset modes include both global reset and reset by row (ripple mode). Reference pixels are built-in to the output data stream. The 1K × 1K IBC is packaged in a robust modular package that consists of a multilayer motherboard, silicon carbide (SiC) pedestal, and cable assembly with 51-pin MDM connectors. All materials of construction were chosen to match the thermal expansion coefficient of silicon to provide excellent module thermal cycle reliability for cycling between room temperature and 7 K.

The detector mounted in the VLTI fringe sensor FINITO is a 256x256 HgCdTe array with a cut-off wavelength of 1.9 micron. The same arrays having cut-off wavelengths of 2.5 micron will be used in the tip tilt sensor IRIS and the PRIMA instrument of the VLT interferometer. The arrays are part of an active control loop with integration times as short as a few hundred microseconds. The fringe tracker FINITO uses only 7 pixels of the array. To take advantage of the four parallel channels of the PICNIC multiplexer, the pixels illuminated in each quadrant are positioned at the same location within the quadrants. A noise analysis of the PICNIC array shows that the main sensitivity limitation of the array is contained in the low frequency part of the noise power spectrum. Similar behaviour has been observed with other infrared arrays. In an effort to optimize the unit cell pixel buffer to achieve high speed and low noise, a prototype multiplexer is being developed at Rockwell for adaptive optics. However, low frequency noise may still be the limiting factor dominating the noise performance of infrared arrays. To overcome this noise barrier, detector architectures have to be envisaged which should allow double correlated sampling on shorter time scales than a full exposure. This might be accomplished by some kind of gate in the IR material which allows charge to be shifted from an integrating well in the infrared pixel to a small sensing node capacitance of the multiplexer unit cell buffer.

We describe the detector subsystem developed at MPIA to operate the Rockwell Hawaii-2 detectors used in the LUCIFER and LINC-NIRVANA instruments for the Large Binocular Telescope (LBT). To fully exploit the capabilities of the LBT, the detector subsystem must meet, especially in the case of the low background applications foreseen for LUCIFER, very stringent requirements in terms of stability and read noise. A read-out electronics has been developed at MPIA, which is able to read the 32 outputs of the Hawaii-2 detector, as well as the 4 reference signals available in this chip. The noise figure associated to the electronics alone is negligible with respect to the intrinsic read noise of the detector, while the cloking patterns and the value of the bias voltages applied to the chip are optimized in order to maximize the signal to noise ratio in the different operating modes. We present the results of the tests performed with the LUCIFER science detector; in particular, we describe the main properties of the detector: read noise, dark current, linearity, and long term stability, and what are the read-out schemes foreseen for different observational modes. We discuss also how the reference outputs can be used in order to correct for thermal drifts, and how effective those outputs are in removing higher frequency noise components.

Limiting the electrical frequency response of the output video prior to analog-to-digital conversion is a common technique used to maximize the signal to noise ratio in imaging systems employing CCD and CMOS based solid-state detectors. Bandwidth limiting effectively reduces the magnitude of the stochastic noise in the video signal prior to digitization; but this filtering action can also change the information content in the video stream if it is applied too liberally. Various "rules of thumb" exist in the detector community for setting the bandwidth in video processing electronics. However, it will be shown that significant signal errors can be induced in many applications if such general design practices are followed implicitly. This paper addresses the general question of how to optimize bandwidth limiting in systems employing either CCD or CMOS based detectors. Induced errors in pixel signal and spatial frequency response (MTF) are analyzed as a function of bandwidth limiting for both CCD and CMOS based detectors and recommended bandwidth values are presented for systems utilizing 8 through 18 bit analog-to-digital converters.

Charge diffusion can vary significantly over the area of a thinned CCD due to local differences in the device thickness. This effect can cause variations in the widths of point spread functions over the field of view. Analyses of on-orbit Hubble Space Telescope Advanced Camera for Surveys (ACS) Wide Field Camera (WFC) star cluster images show a direct relationship between stellar PSF widths and CCD thickness measurements. PSF model fitting was used to separate variations caused by aberrations from charge diffusion effects. The results show that in the ACS/WFC, which undersamples the PSF by a factor of two at 0.5 μm, the core width of a star’s image is determined mostly by charge diffusion and varies by 30% - 40% over the field, with larger differences at shorter wavelengths. This effect should be considered for any instrument that uses thinned CCDs and does not fully sample the optical PSF.

One of the problems found in the design of the electronics for astronomical instruments is the difficulty to find precise digitizers (16 bits) at high speed. In fact, most of the chips which claim to have 16-bit actually have a lower ENOB (Effective Number Of Bits), normally around 14, when considering their noise effects. In this paper, a technique based in auto-adjustable gain amplifiers is proposed as a way to relax the A/D requirements for astronomical CCDs and infrared detectors. The amplifiers will automatically toggle between 2 different gains depending on the pixel value. The technique is based on the fact that, due to the shot (photon) noise of the detectors, the maximum signal to noise ratio achievable in most of these devices is relatively low, allowing the use of A/D converters with an ENOB of only 14 (or even 12) bits when combined with auto-adjustable gain amplifiers. It will be shown that the lower resolution of the A/D converters will not affect the accuracy of the science data, even when many images are averaged out to compensate the effects of the shot noise. Furthermore, given that many real A/D converters do not reach an ENOB of 16, for low level signals the accuracy can be even slightly improved with the technique described in this paper. On the other hand, this relaxing of the A/D requirements can allow the use of off-the-shelf boards for the acquisition systems.

We present a theoretical model of a superconducting kinetic inductance detector which promises high sensitivity and energy resolution from submillimetre to X-ray wavelengths. Cooper-pair breaking photons are absorbed in a superconductor, exciting quasiparticles which change the surface inductance. By arranging the detector in a resonant circuit we can measure the resulting phase-shift of a microwave probe signal. Software has been created to model the superconducting characteristics of the detector and its behaviour when a photon is absorbed. The model predicts the position sensitivity of the detector and calculates how quasiparticles diffuse and recombine to a thermal background level. This temporal evolution of quasiparticle dynamics gives rise to a measurable phase-shift pulse, which will allow the energy and time of a photon to be measured. Pulse shapes have been simulated for photon energies of 1-5 keV being absorbed at the sensitive ground-end of the detector.

At the European Southern Observatory (ESO) in Garching, Germany, several adaptive optics systems using curvature wavefront sensors are being developed for the Very Large Telescope (VLT) and the VLT interferometer (VLTI). Curvature AO-systems have traditionally used avalanche photodiodes (APDs) as detectors due to strict requirements of very short integration times (200 microsec) and very low readout noise. Advances in CCD technology motivated an investigation of the use of a specially designed CCD as the wavefront sensor detector in a 60-element curvature AO system. A CCD has never been used before as the wavefront sensor in a low light level curvature adaptive optics system. This CCD can achieve nearly the same performance as APDs at a fraction of the cost and with reduced complexity for high order wavefront correction. Moreover the CCD has higher quantum efficiency and a greater dynamic range than APDs. A readout noise of less than 1.5 electrons at 4000 frames per second was achieved. Back-illuminated thinned versions of this CCD can replace APDs as a new detector for high order curvature wavefront sensing.

A user-friendly and automatic illuminator with adjustable wavelength and optical power has been developed to obtain precision quantum efficiency (QE) curves of astronomical CCD as well as optical transmission measurements for cryogenic holographic gratings and other optical components. Integrating commercial components with custom mechanical parts and control software, this equipment is able to illuminate a target with light of controlled intensity and wavelength. This facility is primarily intended for testing of Volume Phase Holographic (VPH) gratings at low temperature as well as obtaining CCD quantum efficiencies. A Labview control application runs on a desktop computer allowing full automation of the spectrophotometer. The apparatus includes a Quartz-Tungsten light source, neutral density filters, a monochromator, visible and near-infrared power meters, as well as collimating and focusing optics. Rotation mechanisms allow the characterization of gratings for all angles of diffractions. For CCD testing, network commands allow the facility to get raw images, compute and record QE curves for further detector characterization.

VISTA is a 4-metre survey telescope currently being constructed on the NTT peak of ESO’s Cerro Paranal Observatory. The telescope will be equipped with a dedicated infrared camera providing images of a 1.65 degree field of view. The telescope and camera are of an innovative f/3.26 design with no intermediate focus and no cold stop. The mosaic of 16 IR detectors is located directly at Cassegrain focus and a novel baffle arrangement is used to suppress stray light within the cryostat. The pointing and alignment of the telescope and camera is monitored by wavefront sensing elements within the camera cryostat itself. This paper describes the optical, mechanical, electronic and thermal design of the combined curvature sensor and auto-guider units positioned at the periphery of the camera field of view. Centroid and image aberration data is provided to the telescope control system allowing real time correction of pointing and alignment of the actively positioned M2 unit. Also described are the custom optics, mounted in the camera filter wheel, which are used to perform near on-axis high order curvature sensing. Analysis of the corresponding defocused images allows calibration tables of M1 actuator positions to be constructed for varying telescope declination and temperature.

PMAS is a versatile integral field spectrograph based on the principle of a fiber-coupled lens array type of IFU. The instrument was commissioned at the Calar Alto 3.5m Telescope in May 2001. PMAS is offered as a common user instrument at Calar Alto since 2002. However, it has remained flexible enough to be used as a testbed for new observing techniques. Since the instrument is sensitive in the wavelength range from 0.35 to 1 μm, it is being used to experiment with faint object 3D spectroscopy for a variety of objects in stellar and extragalactic astronomy. Among these experiments, we have implemented a nod-shuffle mode of operation, which is a beam switching technique to achieve a high degree of sky subtraction accuracy. We describe the technical details of the special solution found for PMAS and first results obtained in test observations of faint haloes of planetary nebulae.

CFHT is planning to upgrade its adaptive optics system, PUEO, to a high order system with 104 elements, PUEO NUI. Currently PUEO uses a 19 element deformable mirror with the equivalent 19 avalanche photodiode (APD) detectors as its curvature wavefront sensor. PUEO NUI plans to implement the curvature wavefront sensor using back illuminated CCID-35 detectors developed by J. Beletic et al. instead of 104 APDs, which are prohibitively expensive under the present budget conditions. The CCID-35 detectors, developed at ESO and MIT/LL, were specifically designed to serve as direct replacements for APDs in curvature sensing. The first step in the upgrade is to build and test a system using two CCID-35 detectors, dubbed FlyEyes. These new detectors were successfully tested and integrated in the lab by R. Dorn at ESO but have yet to see sky time. FlyEyes will be their first opportunity. They will directly replace the 19 APDs in PUEO temporarily for a few engineering nights in January of 2005.

SALT, the Southern African Large Telescope, is a 10-m class telescope presently under construction and designed along the lines of the Hobby-Eberly Telescope (HET) at McDonald Observatory in West Texas. The two first light instruments are a simple science imager, SALTICAM, which also doubles as the telescope acquisition camera, and a low resolution spectrometer, the Prime Focus Imaging Spectrograph (PFIS). The detector packages for both instruments, which are being supplied by the South African Astronomical Observatory, will have the capability of readout of a sub-array with a frequency of up to 10 Hz. This is not available on most 8-10 m class telescopes and enables deeper exploration of high-time resolution parameter space. This paper will describe general features of the detector packages, emphasizing the high speed capability, and also touching on the kind of science which is envisaged.

Dynamic range is a very important figure of merit to an imaging system in astronomy since it decides the range of brightness we can observe. This paper describes the design of a CMOS camera with extended dynamic range in which the CMOS sensor achieves high dynamic range by its dual slope response. We first established a model of how the dual slope response works, and gave a method to restore the image from dual slope response to linear response with the extended dynamic range. Then the data needed in the restoration to linear image was obtained in the laboratory by experiments using stable light source based on the model. At last the results of high dynamic range linear response images are shown using these experiment data.

We describe the design of a new IR camera for the Cambridge Optical Aperture Synthesis Telescope that has been designed both to increase our science productivity at COAST and to prototype novel hardware architectures for the Magdalena Ridge Observatory Interferometer IR detector systems. The new camera uses a Rockwell HAWAII sensor in place of the NICMOS device from our previous camera and will be able to sample the temporal fringe patterns from the four outputs of the COAST infrared beam-combiner at frame rates up to 10 kHz. The use of non-destructive multiple reads should allow an effective read noise of 3 electrons to be attained with this chip. The camera controller uses a PulseBlaster FPGA card to generate the timing signals: the advantages of this are flexibility, ease of use, and rapid reconfiguration of the clocking scheme. The new system should improve the IR sensitivity of COAST by around 2 magnitudes. We detail the design of the hardware and the associated software.

Omega2000 is the first near infrared (NIR) wide field camera installed on the 3.5 m telescope at Calar Alto which operates with a 2kx2k HAWAII-2 FPA. Each component of the camera system must suit high requirements to exploit the facilities provided by the imaging sensor. To meet these requirements was a great challenge in design and realization of the optics, the mechanical part and the electronics. The cryogenic optical system with a warm mirror baffle can produce excellent optical quality and high sensitivity over the whole 15.4x15.4 arcmin field of view. The readout electronics together with the camera control software provide multi functional data acquisition and the camera control software can perform the readout and on-line data reduction simultaneously at a high data rate. Different operational and readout modes of the data acquisition of the detector both for engineering and scientific purpose were implemented, tested and optimized and the characteristics of three HAWAII-2 detectors were also determined in their hardware and software environment. Initial astronomical observations were carried out successfully in autumn 2003.

This paper presents results on performance testing of mid-infrared detector arrays for the Faint Object Infrared Camera for the SOFIA Telescope (FORCAST). FORCAST is a two-channel camera that utilizes a Si:As blocked impurity band (BIB) 256 x 256 detector array for imaging through discrete filters at 5 - 25 microns, and a Si:Sb BIB 256 x 256 detector array for imaging at 25 - 40 microns, over a 3.2' x 3.2' field of view, under high thermal background conditions. DRS Technologies has designed and fabricated several Si:As BIB and Si:Sb BIB engineering grade detector arrays which we test as candidate arrays for FORCAST. We present their initial laboratory test performance results.

The Anglo-Australian Observatory has developed a 2nd generation optical CCD controller to replace an earlier controller used now for almost twenty years. The new AAO2 controller builds on the considerable experience gained with the first controller, the new technologies now available and the techniques developed and successfully implemented in AAO's IRIS2 detector controller. The AAO2 controller has been designed to operate a wide variety of detectors and to achieve as near to detector limited performance as possible. It is capable of reading out CCDs with one, two or four output amplifiers, each output having its own video processor and high speed 16-bit ADC. The video processor is a correlated double sampler that may be switched between low noise dual slope integration or high speed clamp and sample modes. Programmable features include low noise DAC biases, horizontal clocks with DAC controllable levels and slopes and vertical clocks with DAC controllable arbitrary waveshapes. The controller uses two DSPs; one for overall control and the other for clock signal generation, which is highly programmable, with downloadable sequences of waveform patterns. The controller incorporates a precision detector temperature controller and provides accurate exposure time control. Telemetry is provided of all DAC generated voltages, many derived voltages, power supply voltages, detector temperature and detector identification. A high speed, full duplex fibre optic interface connects the controller to a host computer. The modular design uses six to ten circuit boards, plugged in to common backplanes. Two backplanes separate noisy digital signals from low noise analog signals.

The scientific instrument ELMER for the GTC 10-m telescope integrates an E2V Technologies CCD44-82 detector and can optionally use the MIT/LL CCID-20 device. Fan-out electronics have been developed that support both detectors reusing the same PCB design and cabling. The main system component, the fan-out board, provides dual-channel input clamp and preamplification of the CCD output signal as well as filtering and active ESD protection of all the CCD lines within a 60x80mm envelope. The preamplifier stage is based on complementary bipolar operational amplifiers with dielectric isolation technology for optimum noise performance and minimum settling time. Matched, discrete JFET buffers are used for improved DC precision. Preliminary tests performed on the system yield 2.0μV_rms preamplifier noise after CDS at 50kHz readout frequency. Slew rate in excess of 230V/μs and settling time well below 75ns have been obtained for a gain 4 preamplifier configuration driving non terminated coaxial cable.

The new UCO/Lick Observatory CCD camera uses a 200 MHz fiber optic cable to transmit image data and an RS232 serial line for low speed bidirectional command and control. Increasingly RS232 is a legacy interface supported on fewer computers. The fiber optic cable requires either a custom interface board that is plugged into the mainboard of the image acquisition computer to accept the fiber directly or an interface converter that translates the fiber data onto a widely used standard interface. We present here a simple USB 2.0 interface for the UCO/Lick camera. A single USB cable connects to the image acquisition computer and the camera's RS232 serial and fiber optic cables plug into the USB interface. Since most computers now support USB 2.0 the Lick interface makes it possible to use the camera on essentially any modern computer that has the supporting software. No hardware modifications or additions to the computer are needed. The necessary device driver software has been written for the Linux operating system which is now widely used at Lick Observatory. The complete data acquisition software for the Lick CCD camera is running on a variety of PC style computers as well as an HP laptop.

The advent of large focal planes requiring many signal channels has exacerbated the problems associated with guaranteeing detector controller performance. The performance specifications for large focal plane controllers includes both 'per channel' requirements such as noise, linearity, dynamic range, etc. and 'system' requirements such as channel cross talk, gain matching, etc. To assess these performance parameters and fully characterize the controller before integration to a focal plane, the MONSOON team has adopted a testing methodology that is based on industry standard practices. This adoption has provided a consistent testing method that produces repeatable results and allows full characterization of the controller performance. This approach will provided a tool to assist in predicting focal plane performance before integration and establishes base line performance values for subsequent detector optimization efforts.

The development of the Orthogonal Transfer Array CCD provides unique control mechanisms that allow a rich set of operating modes necessary to meet the demands of very wide-field imaging programs. The exclusive control modes of the OTA place strong requirements on the CCD controller to support the capabilities of the device while providing detector-limited performance. NOAO and WIYN Observatory have developed a controller based on the MONSOON Image Acquisition concept with the specific application for testing and characterizing the OTA performance and capability. The OTA controller implements control solutions for on-chip cell multiplexing, multiple read modes, high-speed guiding with multiple stars, predictive algorithms for temporal and spatial image motions, and application of electronic tip-tilt corrections. The MONSOON image acquisition system provides the flexibility needed to support the full capabilities of the OTA, while its extensibility can facilitate large mosaics of devices to meet the demands of future very large focal plane instruments.

The PanSTARRS project has undertaken an ambitious effort to develop a completely new array controller architecture that is fundamentally driven by the large 1gigapixel, low noise, high speed OTCCD mosaic requirements as well as the size, power and weight restrictions of the PanSTARRS telescope. The result is a very small form factor next generation controller scalar building block with 1 Gigabit Ethernet interfaces that will be assembled into a system that will readout 512 outputs at ~1 Megapixel sample rates on each output. The paper will also discuss critical technology and fabrication techniques such as greater than 1MHz analog to digital converters (ADCs), multiple fast sampling and digital calculation of multiple correlated samples (DMCS), ball grid array (BGA) packaged circuits, LINUX running on embedded field programmable gate arrays (FPGAs) with hard core microprocessors for the prototype currently being developed.

For the high-resolution IR Echelle Spectrometer CRIRES (1-5 μm range), to be installed at the VLT in 2005, ESO is developing a 512 x 4096 pixels focal plane array mosaic based on Raytheon Aladdin III InSb detectors with a cutoff wavelength of 5.2 microns. To fill the useful field of 135 mm in the dispersion direction and 21 mm in the spatial direction and to maximize simultaneous spectral coverage, a mosaic solution similar to CCD mosaics has been chosen. It allows a minimum spacing between the detectors of 264 pixels. ESO developed a 3-side buttable mosaic package for both the Aladdin II and Aladdin III detectors which are mounted on multilayer co-fired AlN ceramic chip carriers. This paper presents the design of the CRIRES 512 x 4096 pixel Aladdin InSb focal plane array and a new test facility for testing mosaic focal planes under low flux conditions.

The Lawrence Berkeley National Laboratory has been developing fully-depleted high resistivity CCDs. These CCDs exhibit very high red quantum efficiency, no red fringing, and very low lateral charge diffusion, making them good candidates for astronomical applications that require better red response or better point spread function than can typically be achieved with standard thinned CCDs. For the LBNL 2Kx4K CCD we have developed a four-side mosaic package fabricated from aluminum nitride. Our objectives have been to achieve a flatness of less than 10 micrometers peak-to-valley and a consistent final package thickness variation of 10 micrometers or less in a light-weight package. We have achieved the flatness objective, and we are working toward the thickness variation objective.

The desire for larger and larger format arrays for astronomical observatories - both ground and space based - has fueled the development of very large focal plane array (FPA) packaging technology. This has generated new designs and the use of new materials suitable for high reliability and long thermal cycle performance when exposed to operating temperatures from ambient to below 10 Kelvin. We present the design and performance of a series of package designs meeting these requirements evolving from single large mega-pixel arrays through the multiple detector arrays utilizing 4-side butting. This butting arrangement permits future detector arrays of significant size of approaching a meter on a side for infrared astronomy. This packaging technology and the use of thermally compatible materials enable the large format packaging of all detector and Readout Integrated Chip (ROIC) combinations in current production. Current and future applications include the Mid-Infrared Instrument (MIRI) detector for the James Webb Space Telescope (JWST) mission, the 16 VIRGO detector focal planes for the Visible and Infrared Survey Telescope for Astronomy (VISTA) IR survey telescope and future applications such as the Supernova Acceleration Probe (SNAP) mission.

We report on the measurement results for two candidate astrometric CCD designs, the STA700 and the e2v CCD43. Both of these CCDs have been considered as design baseline CCDs for use in a high-precision astrometric instrument in space, similar to the one proposed for the FAME, DIVA, or AMEX missions. We have exposed one sample CCD of each design to a fluence of 5 x 10⁹ p⁺ cm^-2 (@ 63.3 MeV), then measured the relevant readout noise, dark current, charge injection noise, and CTI performance. We compare the two CCDs and assess how well each mitigates radiation damage.

The proton-induced charge transfer efficiency (CTE) behavior for the Lawrence Berkeley National Laboratory (LBNL) p-channel CCD [being developed for the Supernovae Acceleration Probe (SNAP)] is compared with the Hubble Space Telescope’s (HST) Wide Field Camera 3 (WFC3) n-channel CCDs CTE using ⁵⁵Fe x-rays, first pixel edge response (FPR), and extended pixel edge response (EPER) techniques. The pre- and post-proton radiation performance parameters of p-channel CCDs designed by LBNL and fabricated at Dalsa Semiconductor, Inc. are compared with n-channel CCDs from E2V, Inc. LBNL p-channel CCDs both with and without notched parallel registers are compared with the E2V CCD43 [a notched, multi-phase pinned (MPP) device] and the E2V CCD44 (an un-notched, non-MPP device), using the same readout timing and measured over the same range of temperatures. The CTE performance of the p-channel CCD is about an order of magnitude better than similar n-channel CCDs for the conditions measured here after a 63 MeV proton fluence of 2.5 x 10⁹ cm^-2, which is equivalent to 2.5 years in the HST orbit behind shielding comparable to about 2.5 cm Al. Our measurements are compared with previous CTE measurements at 12 MeV by Bebek et al. The ~ 10 x CTE improvements relative to n-channel CCDs is seen at -83°C, a temperature which is optimized for n-channel CCD performance. Advantages from p-channel CCDs should be greater at other temperatures. Dark current measurements and hot pixel issues are also discussed.

Theoretically, L3CCDs are perfect photon counting devices promising high quantum efficiency (~90%) and sub-electron readout noise (σ<0.1 e^-). We discuss how a back-thinned 512x512 frame-transfer L3CCD (CCD97) camera operating in pure photon counting mode would behave based on experimental data. The chip is operated at high electromultiplication gain, high analogic gain and high frame rate. Its performance is compared with a modern photon counting camera (GaAs photocathode, QE ~28%) to see if L3CCD technology, in its current state, could supersede photocathode-based devices.