Chandra Aspect Operations

Flickering Pixels Follow-Up

Summary

From September 2008 to February 2009, 49 perigee passes were used to gather flickering pixel data as a follow-up to the 2002 study. Analyzed data show that the behavior of warm pixels remains largely unchanged.

Data Gathering

The Star Selection & Acquisition Working Group submitted a special activity request to gather data similar to the data gathered during the 2002 data collection. Fixed-position monitor windows were commanded during perigee pass observations. The positions of these monitor windows were selected to match a subset of the positions used in the 2002 study.

In the coverage figure, the positions used for the 2008/2009 data gathering are marked in blue; 2002 locations are marked in red. Thus, fewer locations on the CCD were used in the new study. However, in the 2008/2009 study, an attempt was made to command the same monitor window over multiple observations during the same perigee pass. This allowed the collection of much more data. Data was generally collected in only one obsid per perigee pass during the 2002 study.

Analysis

As in the 2002 analysis, we define a pixel as 'warm' if the 50th percentile dark current exceeds 100 e-/sec. However, instead of using by-eye light curves and an IDL routine to mark pixel 'jumps', a Bayesian Blocks algorithm was applied to the pixel data. This algorithm results in similar 'jump' detections to the previous method, but the new method may be totally automated. The following figures show 2002 data analyzed using the old and new methods.
Old AnalysisNew Analysis
For consistency when comparing the 2002 data to the 2008/2009 data, the 2002 data was re-analyzed using this new Bayesian Blocks method.

More Data Plots

Warm pixels with raw (median filtered) data and Bayesian blocks intervals

Warm Pixels Plotted Simultaneously Per Perigee Pass

Concatenated Data from Warm Pixels seen in 2002 and 2008/9.

Results

(In these plots, all 2002 data is red/salmon, 2008/2009 data is blue, overlapping sets display as ~ plum)

Histograms of 'jump' fraction. Here "A" is the pixel count rate in e-/sec. The shape of the histogram of the 'jumps' as a fraction of the preceding count rate is slightly more broad for 2008/2009 data:

The dip in the middle of the above plot is due to the fact that the Bayes analysis is not sensitive to small jumps.

The following plot is a histogram of the durations of the intervals determined by the Bayesian Blocks analysis.

The differing slopes of the older and newer data in this duration histogram show that the 2008/2009 data appears to have more intervals of longer durations than the 2002 data. However, as the 2008/2009 data was collected over longer intervals (more full perigee pass data instead of the first obsid of a pass as used in 2002) it is not unexpected that there is a larger fraction of longer intervals. This does not rule out other possible contributing factors, such as operating temperature and accumulated radiation exposure.

Gotchas

As the CCD is exposed to incoming photons from the observed field and the fixed monitor window passes over areas of the sky, the light from dithering faint stars may contaminate these data. Ionizing radiation may also cause rate spikes. In both cases, the 'flickering' behavior is not inherent to the 'aberrant' pixel. Intervals which appear to be contaminated by stars have been manually marked and excluded from this analysis. See Pass 2008:357 for an example.

The Bayesian Blocks analysis cannot distinguish small magnitude flickering from noise, and thus this analysis is not sensitive to any such behavior.

Conclusions

The behavior of flickering pixels in the ACA CCD appears largely unchanged from 2002 to 2008/2009. No new behaviors have emerged and the CCD appears to be very stable. As mentioned, the appearance of more intervals of longer duration in the new data set is likely a result of the data collection strategy, but as other causes cannot be ruled out with this analysis, it may be valuable to repeat the data collection strategy used for this study in a future study.

As the collected data from 2008/2009 will be used to update the flickering pixel simulation library, it may also be valuable to gather data to update that library again in the future.

Analysis data

The scripts and data used in this analysis are in /proj/sot/ska/analysis/flickering_pixels_2008


Last modified:01/13/11



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