Detail Descriptions of Data Selections and Reductions

The Goal is to extract information on the HRC background when the detector is in the "stowed" position.

For our purposes the stowed position will be anytime that the SIM translation motor step position is greater than zero (TSCPOS > 0).

The detector MCP HV must be on and at the operational level

For the HRC-I, this means:

For the HRC-S, this means:

The configuration of the HRC-I (or HRC-S) matters. Data from the SI_MODE = DEFAULT configuration are the data of interest. For more information on the HRC SI_MODEs (HRCMODE) see:

http://cxc.harvard.edu/contrib/juda/hrc_flight/macros/hrcmode.html

The DEFAULT configuration for each instrument has changed over time. Some of these changes are not expected to have an impact on the background (the WIDTH_THRESHOLD and RANGE_SWITCH_LEVEL parameters) but do have an impact on how the event data are processed.

The HRC-I has been operated in a non-DEFAULT configuration a few times, most notably when we are performing the ACIS undercover background. There can be times after these observations when the stowed background data would be collected in this configuration. these time intervals should not be included.

For the HRC-I the important configuration settings to satisfy are:

For the HRC-S the important configuration settings to satisfy are:

The third HRC "stowed" background to do is for the HRC-S high-precision timing configuration: For the HRC-S in the high-precision timing mode (HRCMODE S_TIMING) the important configuration settings to satisfy are:

SCIDPREN has a C-style bit numbering:

BitMSIDDescription
12-15 SPARE  
11 2CLMDAST CALIBRATION MODE ENABLE
10 2FIFOAVR DATA FIFO ENABLE
9 2OBNLASL OBSERVING/NEXT-IN-LINE MODE SELECT
8 2SPMDASL SPECT DETECTOR SPECT/IMG MODE SELECT
5-7 SPARE  
4 2EBLKAVR EBLK VALIDITY ENABLE
3 2CBLKAVR CBLK VALIDITY ENABLE
2 2ULDIAVR ULD VALIDITY ENABLE
1 2WDTHAVR WIDTH VALIDITY ENABLE
0 2SHLDAVR SHIELD VALIDITY ENABLE

The bit-mask filter SCIDPREN=0000xxxx000xxxxx requires the spare bits to be 0.

Given the time intervals when the SIM translation table has the HRC in the "stowed" position, the subset of times for which the HRC-I has the HV on and at the operational level and which satisfies the configuration settings can be found. Similar time intervals can be found for the HRC-S.

A complication in dealing with the HRC serial-digital housekeeping telemetry (the hrcf*_hk0.fits files) comes from "glitches". These can mark nominally good times as bad. There are three types of glitches:

In principle, the QUALITY bit-array flags the first type of glitch. The second can be at least partially identified by non-zero values to the spare bits in SCIDPREN. A bit-mask check of

SCIDPREN=0000xxxx000xxxxx

should remove some of these second type of glitch. The third type are harder to spot but I have had some success with using MSID 2SMTRATM (the selected motor temperature). Normally, there is no motor selected and the value floats to some arbitrary value. The corruption of the data stems from bytes in the secondary science FIFO getting duplicated and the two bytes preceding 2SMTRATM are spares that stay at a "high" value. A reading of 55-56 for 2SMTRATM indicates corrupted data. For more information on this data corruption see:

http://cxc.harvard.edu/contrib/juda/memos/anomaly/sec_sci/index.html

and

http://cxc.harvard.edu/contrib/juda/memos/anomaly/sec_sci/byte_shift.html

For each interval the secondary science rate data should be extracted and merged into a single file for the interval. From each of these interval files the mean, sigma, min, and max of the total (TLEVART) and valid (VLEVART) MCP rates should be determined and incorporated into a table. This will permit investigation of the time history of the background and any potential changes in the valid/total ratio. It will probably be necessary to iteratively clip the rates when calculating the statistics to avoid outliers.

Processing the event data from each interval will require specifying the values for the keyword RANGELEV to do the AMP_SF correction. While earlier I described using time to divide between when it is appropriate to use 90 or 115 the automatic way to do it is to set the value based on MSID 2RSRFALV, which should be essentially constant over the interval. The prescription for the value of RANGELEV is:

RANGELEV = (int)((mean(2RSRFALV) - 127) * 2 + 0.5)

It is also necessary to specify the value of the keyword WIDTHRES to obtain the best tap-ringing correction. Its value is determined by the value of MSID 2WDTHAST, which should be constant during the interval. Its prescription is:

WIDTHRES = (int)(mean(2WDTHAST) + 0.5)

To run hrc_process_events, we also set following parameters.

ParametersHRC IHRC S
ampsatfile hrcsD1999-07-22sattestN0002.fits hrciD1999-07-22sattestN0002.fits
ampsfcorfile hrciD1999-07-22amp_sf_corN0001.fits hrcsD1999-07-22amp_sf_corN0001.fits
badpixfile hrcf10702_000N001_bpix1.fits hrcf10600_000N001_bpix1.fits
degapfile hrciD1999-07-22gaplookupN0003.fits hrcsD1999-07-22gaplookupN0002.fits
evtflatfile hrciD1999-07-22eftestN0001.fits hrcsD1999-07-22eftestN0001.fits
hypfile hrciD1999-07-22fptestN0003.fits hrcsD1999-07-22fptestN0004.fits
obsfile obs.par obs_hrc_s.par
tapfile hrciD1999-07-22tapringN0002.fits hrcsD1999-07-22tapringN0002.fits

There are several gainfile for HRC I.

Before Oct 4, 1999 hrciD1998-10-30sampgainN9999.fits
After Oct 4, 1999 Before Dec 12, 2000 hrciD1999-10-04sampgainN9999.fits
After Dec 12, 2000 Before Jan 26, 2002 hrciD2000-12-12sampgainN9999.fits
After Jan 26, 2002 Before Feb 22, 2003 hrciD2002-01-26sampgainN9999.fits
After Feb 22, 2003 Before Nov 25, 2004 hrciD2003-02-22sampgainN9999.fits
After Nve 25, 2004 Before Oct 17, 2005 hrciD2004-11-25sampgainN9999.fits
After Oct 17, 2005 Before Sep 20, 2006 hrciD2005-10-17sampgainN9999.fits
After Sep 20, 2006 Before Sep 17, 2007 hrciD2006-09-20sampgainN9999.fits
After Sep 17, 2007 Before Sep 07, 2008 hrciD2007-09-17sampgainN9999.fits
After Sep 07, 2008 hrciD2008-09-07sampgainN9999.fits
 
For HRC S, there is only one gainfile file
  hrcsD1999-07-22gainN0001.fits

A separate level 1 type event file should be made for each of the stowed-background intervals. These can later be merged into larger data sets based on time or other discriminators.

Last modified: Mon Jan 12, 2010