Previous Article Next Article Table of Contents
IRIM in Action: Performance Information and User...(1Jun93) Evaluation (from KPNO, NOAO Newsletter No. 34, 1 June 1993) As anticipated in NOAO Newsletter No. 31 (September 1992), the infrared camera IRIM has been returned to service at the 1.3-m, 2.1-m, and 4-m telescopes, equipped with a 256 x 256 NICMOS 3 HgCdTe array through a collaboration with Tony Tyson (Bell Labs). I summarize here some performance information for the camera and initial user reactions. For more complete information, including specifics on filters, star and sky signal levels, and detector performance, request the updated IRIM manual from rprobst@noao.edu. The array has a rather nonflat response, with +-25% variations at low spatial frequency across the device. There are about 250 nonresponsive pixels scattered across the array, typically in small clumps. Some users have ultimately rejected up to 700 pixels, but this will depend on the science application. The response linearity falls smoothly with signal level to about 85% of full well capacity, then declines rapidly. Read noise is about 30 electrons. Dark current and well capacity vary with bias, for example 0.5 electron/sec and 140,000 electrons at 400 mV, and 2 electron/sec and 270,000 electrons at 800 mV. Quantum efficiency improves by a factor of two from 60 K to 77 K, and we presently operate at 77 K. There is a "charge retention" effect in which a small portion of incident signal from a given exposure appears to survive many subsequent reset cycles, declining slowly. The practical importance of this phenomenon depends on the original signal level and the general background level on the device. Other users of these devices have reported a variation in sensitivity within a pixel (D. Allen, AAO Newsletter No. 63; and private anecdotal remarks). I have not confirmed this explicitly for the KPNO array, but have noted that point source photometry seems unexpectedly erratic. The instrument was reconfigured with deep detections on the 4-m in mind, and initial user reaction in this application has been very enthusiastic. The first three hours of 4-m science observing produced confirmation of variability in a faint QSO (R. Elston, NOAO) and detection of a high-redshift galaxy cluster around a radio source (M. Dickinson, U. of California, Berkeley). Besides the desirable detector characteristics (high QE pixels and lots of them), the combination of our new f/15 IR secondary for the 4-m and a short-cutoff K' filter (2.0-2.3 um) gives a sky background 1 to 1.5 mag fainter than formerly on this telescope, with corresponding decrease in integration time for a given sensitivity. Source detections at K ~ 22 have been achieved with two hours' integration. Similar results, scaled for aperture, have been obtained on the 2.1-m. Judging from the proposals for fall 1993, word has travelled fast! Ron Probst
Previous Article Next Article Table of Contents