XSPEC FITS Response Matrix: w140c4r_norm.rmf

Release notes for preliminary, normalized spectral redistribution matrix for ACIS back-illuminated flight detector S1 (w140c4r).

F. Baganoff/M. Bautz
30 October 1997




xspec FITS response matrix format. (See Goddard Office of Guest Investigator Programs Calibration Memos CAL/GEN/92-002 & CAL/GEN/92-002a available in Postscript format from cal_gen_92_002.ps and cal_gen_92_002a.ps.


This matrix includes only the CCD spectral redistribution for the back-illuminated detector S1, and does not incorporate detector quantum efficiency, UV/optical blocking filter transmission, or AXAF telescope effective area. The matrix is based on subassembly calibration measurements made at MIT CSR with flight detectors and engineering model detector electronics. It is intended to represent the spectral response of the back-illuminated flight detector S1 when 1-, 2- and 3-pixel events are selected with a split-event threshold of 15 electrons. The FITS header comment fields provide additional details about the matrix generation process.

Intended use:

When combined with ancillary response functions (arf) representing the HRMA effective area, ACIS optical blocking filter transmission, and back-illuminated CCD detection efficiency as functions of energy, the matrix should be useful for simulating ACIS observations with detector S1 for purposes of proposal preparation.

  1. Header keyword entry HDUCLAS3=FULL is incorrect, since telescope, filter and detector efficiencies are NOT included in this matrix.
  2. The spectral response functions herein should be accurate to better than 15% below 2 keV and better than 10% above 2 keV. A more accurate spectral response function for this device will be released by MIT CSR as calibration data analysis continues. Generally, this matrix provides slightly poorer spectral resolution than is observed with the flight detector and flight electronics.
  3. The particular energy scale (mapping from input energy to output pulseheight) incorporated in this matrix is arbitrary; the flight energy scale may differ in slope from that represented in this matrix by about 10%.
  4. The energy scale of this detector is measurably non-linear. This non-linearity is accounted for in the Response Matrix itself (i.e., in extension 1 of the FITS file). However the EBOUNDS Extension (extension 2) assumes a linear energy scale. This bug does not affect the accuracy of spectral simulations, since the simulator does not require the EBOUNDS extension.
  5. This spectral redistribution matrix was produced by analyzing the summed spectrum of X-ray events from all pixels in quadrant B of detector S1. Spectral resolution in the other quadrants may differ by up to 15%. Calibration measurements performed with the HRMA at XRCF indicate that the spectral resolution was about 20-30% better than in the subassembly calibration measurements made at MIT CSR. This effect probably reflects the charge transfer inefficiencies of this detector. Observations of astrophysical objects covering only a small region of the detector may exhibit better spectral resolution than predicted by this spectral redistribution matrix.

Fred Baganoff
Thu Dec 4 18:33:45 EST 1997 Last modified: Thu Dec 4 19:14:34 EST