[Previous] | [Session 45] | [Next]
M.T. Raley, R.J. Edgar, J.J. Drake (Harvard-Smithsonian CfA), ACIS Team
We present a method for improved calibration of the low energy (less than 0.7 keV) gain for the ACIS-S3 device. We have used multiple off-axis grating observations of the BL Lac object PKS2155-304 with the LETG and ACIS, which put various low-energy portions of the spectrum on the S3 chip. An absorbed power law is used to model the broad-band spectrum. The dispersed spectrum on S3 is sliced into multiple 32-pixel ``tiles" which correspond to the tiles used in the newly released ACIS S3 FEFs (CALDB 2.7) A portion of the broad-band model corresponding to the wavelength range covered by the selected part of the chip is then compared to the data using xspec.
The differences between the model and the data are due to the lack of perfect understanding of the gain at lower energies and can be quantified using the ``gain fit" function in Xspec. Using the outlined method, we can determine the differences in the true gain of the S3 chip and that in the response matrix products.
This research was supported by NASA grant NAS8-39073 to the Chandra X-ray Center and SAO.