Effects of H-test Filtering on HRC-I Background

Count Rates

We have evaluated the effects of filtering using the Fine Position criteria (a.k.a. the H-test) on data collected by the HRC-I. We have used the data taken during the instrumental adjustments performed in October and December 1999. During the first adjustment the gain of the detector was lowered by adjusting the High Voltage of the plates, while in the second adjustment the threshold values for selecting the gain sensitivity were raised via a reset of the Range Switch. In each instrumental configuration (three in total), a point source (AR Lac) was used to sample a scan pattern covering the full extent of the microchannel plate.

The unfiltered average background across the entire detector (excluding the source) remained relatively stable at a count rate slightly above 1.1e-5 cts/sec/arcsec**2. There is a decrease in the count rate (from 1.28e-5 to 1.25e-5 and thence to 1.12e-5 cts/s/arcsec**2) correlated to the change in the profile of the instrument. However, the first two data sets, with nearly indistinguishable background rates were collected during October, while the third, which is significantly lower (12%) was collected in December. This may indicate that the variation in the background rate is due to phenomena other than the changes in the instrument.

As can be seen from the accompanying plot, the H-test filter removes approximately 35-40% of the background counts, leaving a residual global (i.e. average) background count rate of just under 0.8e-5 cts/s/arcsec**2.

Global Background	Background Near Source

The "local" (within a radius of 700 pixels) background count rate near each source is consistent with the average/global count rate. The significant increase in the size of the variations is due to the much more localized sampling area of the detector and in fact is most likely attributed to local gain and qe variations.

The mean PHA for the background nicely shows the change in the gain which occurred due to the lowering of the micro-channel plate high voltage in October '99. The bulk of the shift in the mean PHA that occurs due to the H-test filtering is because a large percentage of the events removed by the filtering process have a PHA=255.

Mean PHA	Fraction of Events Least Sensitive Gain Setting (amp_sf=3)

We also show the effect of the instrumental changes and the filtering on the percentage of events received with the least sensitive gain setting. In the event list files, these events have an amplifier scale factor (or amp_sf) value of 3. These events are most affected by the "ringing" in the HRC electronics. The data show that the profile changes in the instrument have substantially reduced this component in the background and that the H-test filtering also reduces this element.

Surface Distribution

For the second component of our analysis we have analyzed the variation in the background count rate across the entire detector for each of the three instrumental configurations, both before and after filtering.

After first removing the sources from the data, the data were blocked 8x8 in degapped pixels and the average determined from the 21 pointings. These results were then further smoothed using a Gaussian with a sigma of 40 blocked pixels, corresponding to 320 raw pixels or approximately 40 arcsecs.

We present three different views of these results: a percentage deviation from the mean, a 3-D projection of the count rate in absolute units and a relative contour plot of the count rate.

Percent Difference from the Mean
Launch Settings	After HV Change	Current Settings
Unfiltered	Unfiltered	Unfiltered
Filtered	Filtered	Filtered

The scale in all of the above frames ranges from -20% to +20% relative to the overall mean within that data set. This was done to allow a direct comparison of all six data sets.

The data shows that the distribution of unfiltered background counts was not appreciably affected by the changes to the instrumental profile in October and December 1999. All three data sets demonstrate a peak in the center of the detector. This is similar to the structure of the gain of the detector.

In contrast the filtered data shows no such peak at center and is better described as a tilted but relatively flat surface (please also see the IPI teams results). The data also indicates that the changes in the instrumental profile have served to significantly improve the overall character of the background.

Contributors:
CXC Calibration Team
HRC IPI Team

Comments to cxccal@cfa.harvard.edu

Chandra Science | Chandra Home | Astronomy links | iCXC (CXC only) | Search

The Chandra X-Ray Center (CXC) is operated for NASA by the Smithsonian Astrophysical Observatory.
60 Garden Street, Cambridge, MA 02138 USA. Email: cxcweb@head.cfa.harvard.edu
Smithsonian Institution, Copyright © 1998-2004. All rights reserved.

Last modified: 09/25/12