Calibration of the AXAF High Resolution Camera (HRC) proceeds in three stages: X-Ray Calibration Facility (XRCF), subassembly calibration (SAC), and On-Orbit Verification (OOV).
The XRCF data were collected at Marshall Space Flight Center in Huntsville, Alabama during the spring of 1997.
In mid-May, the HRC was returned to the HRC lab at SAO in Cambridge, Massachusetts. The HRC was attached to the "pipe 1" beam line on 15 May 1997. Over the next month, SAC measurements were made in order to determine the flat-field response of the HRC detectors. The HRC was removed from pipe 1 on 15 June 1997 for shipment to Ball Aerospace.
At XRCF, a 91 kHz noise signal was discovered when the microchannel plate (MCP) high voltage was turned on, resulting in a large number of false events. The problem was ultimately traced to the RC filter network between the high voltage power supply (HVPS) and the detector. Due to time constraints, the actual repair of the HVPS was delayed until the HRC was returned to the HRC lab in Cambridge. For more detail on this, refer to S. Murray's 1997 SPIE paper
In order to mitigate the effects of the noise signal at XRCF, the threshold voltage level for event triggers was raised from the nominal flight level; the MCP plate voltages were also lowered to correct for pulse height saturation. As a consequence, both the threshold and MCP high voltage levels used at XRCF are different from the nominal in-flight settings. In order to understand both the behavior of the MCP's at XRCF and in-flight, some of the flat-field measurements were made before the HVPS was repaired ("pre-fix") and some after ("post-fix").
Five separate intervals of flat-field measurements were made with the HRC components.
| pre-fix | |
|---|---|
| May 21-23 | HRC-I flat fields |
| May 26-June 1 | HRC-S flat fields |
| post-fix | |
| June 10-12 | HRC-S flat fields |
| June 13-14 | HRC-I flat fields |
| June 15 | HRC-S segment 0 high energy flat fields |
The HRC-I flat field measurements have been summarized in a table. Note that the pre-fix measurements were performed twice: once with the higher flight MCP high voltages and once with the lower XRCF MCP high voltages. But both were made with the higher threshold voltage that was used at XRCF. The flight HV measurements were made at eight energies: 183, 277, 525, 852, 1487, 2984, 4511 and 6404 eV, while the XRCF HV measurements were made at only four: 183, 525, 1487 and 4511 eV. The post-fix flat fields were made at all eight energies and all at the flight setting MCP HV and threshold levels.
The flat fields were all taken at X-ray fluxes far higher than any that might be expected in flight in order to collect statistically significant numbers of counts in a reasonable amount of time.
All of the flat fields, as well as the background data, are stored in raw data files (in "fast telemetry" format) with the measurement date and sequence number embedded in the file name. The names all begin with "p1" in reference to the beam line pipe 1. The data are all stored (compressed with the gzip algorithm) on the hrc data archive on hrc4:/hrcdata1.
The data files were moved from the data archive via FTP. It was noted that several data files which were identified in the data sheets were not in the archive.
Within the ASCDS environment, the data were processed from "raw data" format to QPOE ("Quasi-Position-Ordered Event" format) by use of the shell script "do_fastrd2lev1_v2". This script makes use of S. Murray's "fasttm2fftm" and M. Juda's "fftm210_fits" to convert the fast "telemetry" type raw data to a level 0 FITS file (unpacked event data). The "fits2evt" utility converts FITS to QPOE to permit processing by the ASCDS tool "hrc_process_events", which stops at the minimal instrument coordinates (i.e. not World Coordinate System (WCS) coordinates). The images are blocked with a blocking factor of 128.
The raw data files were also processed using the "do_fastrd2epr" shell script which results in an epr format file. This script makes use of the "fasttm2fftm", followed by "fftm2prd" (extracts the HRC primary science events and unpacks) and "nprd2epr" (handles degapping and hotspot filtering). The epr files were then processed using "do_epr2img_gmi" script which generates a binary counts image and gain (mean pulse height) image for each epr file, using the "nepr2img" and "nepr2gmi" code of J. Chappell and S. Murray. Both images are blocked by 128 into a 512x512 array.
During the collection of the flat field and background data, a Manson flow proportional counter (the "monitor detector") was used to measure the true X-ray intensity in the beam. The monitor detector was positioned on a short arm perpendicular to the main beam. In order to calibrate the rates measured in the monitor to the rate in the main beam (and incident on the MCP), several cross calibration measurements were made, using a second flow proportional counter placed in the main beam line (the "beam normalization detector" or "BND"). These calibration measurements are summarized in several tables.
Unfortunately, the exact X-ray flux in the beams is highly sensitive to the exact position of the filters placed in front of the Manson source. Multiple calibration measurements made on different days and with intervening changes in Manson source anodes and filters display very poor repeatability. This can be seen in several graphs which plot the simultaneous rates measured in the monitor and BND. Consequently, only those calibration measurements which are made immediately before or after a flat field measurement, with no intervening changes in Manson source anodes or filters, are useful. Since all of the calibration measurements were made after the HRC was removed for repair, many of the pre-fix flat fields, including all of the HRC-I pre-fix flats, cannot be calibrated. All of the HRC-I post-fix flats were calibrated using the MD/BND calibration measurements taken on June 12 (for B, C, O and Al) and June 17 (for Ni, Ag, Ti and Fe).
The Quantum Efficiency (QE) of the MCP is the product of four terms:
The QE of the BND is itself the product of three terms: the gas absorption coefficient (which is a function of the temperature, pressure, and composition of the gas), the window transmission (which depends on the composition of the window), and the mesh transmission (which depends on the mesh open area). All of these parameters are specified for each measurement in the cross calibration tables.
The MCP rate is calculated using the total number of events recorded in the detector region of interest, the duration of the observation, and the dead time factor. The total number of events was determined by using the FTOOL "imcnts" tool on the QPOE format file of the flat field data. This was verified using IDL utilities on binary format flat field images. For QE calculations on the HRC-I, three different regions were used: the total open area of the MCP (86.5 cm2), the portion of the MCP which is coated with CsI (81.0 cm2), and a square central region, to provide a QE for the region most likely to be used for imaging astronomical sources (10.5 cm2). Since the flat field data were collected at artificially high rates, the correction for dead time is important. The dead time factor of 68.5 µsec per event is multiplied by the total number of events in the entire detector to determine the dead time. When subtracted from the observation duration, this yields the live time which is used for calculating the rates for the different detector regions. A background rate, determined from images collected with the beam blocked, was subtracted from the MCP rate.
The rates on the proportional counters were determined from the total counts and the live time, found using the M. Juda's "chn_summary" program. Background rates, collected with the beam blocked, were subtracted from the fpc rates.
The solid angles are calculated from the
tabulated areas and distances, as
recorded in the HRC team's logs and data sheets.