The Joint Milliarcsecond Pathfinder Survey (JMAPS) is a small, space-based, all-sky, visible wavelength astrometric
and photometric survey mission for 0th through 14th I-band magnitude stars with a planned 2013
launch. The primary objective of the JMAPS mission is the generation of an astrometric star catalog with 1
milliarcsecond (mas) positional accuracy or better, and photometry to the 1% accuracy level or better at 1st
to 12th mag. Achieving this level of accuracy in the final catalog requires a demanding attention to reducing
We present our findings on distortion, signal to noise, and the astrometric bandpass necessary to obtain the
desired accuracy for JMAPS.
We present both laboratory and telescope testing results describing the performance of the H4RG-10 CMOS-Hybrid
detector. The H4RG-10 is the largest visible hybrid array currently in existence and shows great potential for use in
future space missions. We report read noise, dark current, pixel connectivity, persistence, and inter-pixel
capacitance measurements for the temperature range 110-240 K. We report on quantitative astrometric and
qualitative photometric performance of the instrument based on observations made at USNO's Flagstaff Station
observatory and establish an upper limit to the astrometric performance of the detector. We discuss additional
testing and future work associated with improving detector performance.
CMOS-based focal planes have many potential advantages over CCDs for use in space for star mapping/star
tracking applications. These include more flexible readout circuitry and improved radiation tolerance. There are
also weaknesses, including noise performance, quantum efficiency, and potential systematics introduced by the
presence of circuitry on the photosensitive side of the detector. In this paper, we measure the effects of these sources
of error on centroiding and photometry for the HDTV (1k x 2k) SI-1920HD camera by observing stellar reference
fields using USNO's 8-inch Twin Astrograph telescope in Flagstaff, AZ. This camera serves as an archetype for the
entire family of related TIS detectors, including the 3k x 4k V12M and the 7.6k x 7.6k V59M. We determine an
upper limit for the astrometric centroiding performance for this class of detector to be ~1/30<sup>th</sup> of a pixel. There are
indications that better performance may be possible if improvements are made to the temperature control system
used for this first set of observations.
This paper presents the radiometric and noise characteristics of 12-bit SI-1920HD cameras built from the AltaSens ProCamHD 3560 FPA as a function of integration time and temperature. Our measurements are for two integration time regions: 1 to 50 millisecond, which is standard for video operation; and 1 to 240 seconds, of possible use for stellar observations. For 1 to 50 millisecond integration times, the cameras are extremely linear with a Gaussian-like dark frame. As we increased to seconds-long integration times, the camera initially remains radiometrically linear, but develops a dark frame with the vast majority of pixels at dn=5. Further increases in integration time eventually result in a saturated dark frame with all pixels at dn=4095. Reducing the operating temperature to -7.2°C increased the integration times at which the camera's two transitions occur by a factor of 20. The calibration parameters determined from our measurements were applied to the image data collected by Dorland <i>et al</i>. (these proceedings).