We describe how the James Webb Space Telescope (JWST) Near-Infrared Spectrograph's (NIRSpec's) detectors will be read out, and present a model of how noise scales with the number of multiple non-destructive reads samplingup-the-ramp. We believe that this noise model, which is validated using real and simulated test data, is applicable to most astronomical near-infrared instruments. We describe some non-ideal behaviors that have been observed in engineering grade NIRSpec detectors, and demonstrate that they are unlikely to affect NIRSpec sensitivity, operations, or calibration. These include a HAWAII-2RG reset anomaly and random telegraph noise (RTN). Using real test data, we show that the reset anomaly is: (1) very nearly noiseless and (2) can be easily calibrated out. Likewise, we show that large-amplitude RTN affects only a small and fixed population of pixels. It can therefore be tracked using standard pixel operability maps.

ESA and NASA recently selected two 5 μm cutoff Teledyne H2RG sensor chip assemblies (SCA) for flight on the James Webb Space Telescope (JWST) Near Infrared Spectrograph (NIRSpec). These HgCdTe SCAs incorporate Teledyne's "improved barrier layer" design that eliminates the degradation that affected earlier JWST H2RGs. The better indium barrier, together with other design changes that Teledyne phased in from other programs over the years, has improved the performance and reliability of JWST's SCAs. In this article, we describe the measured performance characteristics that most directly affect scientific observations including read noise, total noise, dark current, quantum efficiency (QE), and image persistence. As part of measuring QE, we inferred the quantum yield over the full NIRSpec pass band of λ ¼ 0:6-5 μm and found that it exceeds unity for photon energies E γ > ð2:65 AE :2ÞE g , where E g is the HgCdTe bandgap energy. This corresponds to λ ≲ 2 μm for NIRSpec's 5 μm cutoff HgCdTe.

No abstract

We examine proton-damaged charge-coupled devices (CCDs) and compare the charge transfer efficiency (CTE) degradation using extended pixel edge response, first pixel response, and 55 Fe X-ray measurements. CTEs measured on Marconi and Fairchild imaging sensors CCDs degrade similarly at all signal levels, though some of the Fairchild CCDs had a supplementary buried channel.

No abstract