Goal:
We discuss the evolution of CTI and its impact on spectral resolution for X-ray CCDs in different orbital environments. We specifically prepare
models of the energy scale and resolution as a function of the background, CCD type (FI vs. BI), and application of charge injection, considering
energies spanning the CCD range.
Outline:
- CTI evolution (energy scale, FWHM)
- effects of different radiation environments
- effects of SCI
- calibration sources from corners (Mn K alpha, akin to LaMarr 2008)
- compare ACIS and XIS
- calibration sources (all energies?) across FOV
- ACIS only
- celestial calibration sources (1 keV, 6 keV) across FOV
- Cygnus Loop, Perseus, E0102, N132D, other clusters
- separate CTI from electronics gain
- verify gain change negligible
- checker-flag CI (Ozawa 2009)
- Cygnus Loop/Perseus at different roll angles (XIS)
- (are we interested in electronics performance?)
- CTI of FI vs. BI
- soft proton damage
- effects of BG sac charge (Grant 2005 SPIE, LaMarr 2008 SPIE)
- about the instruments
- first describe each separately (label CCDIDs)
- then similarities/differences between ACIS/XIS that impact CTI (just the instruments and operation, no environment yet) in a table and summarize in text
- initial pre-launch CTI, ACIS BI > XIS BI, XIS FI > ACIS FI
- Transfer speeds, fast transfer (image-to-framestore) serial transfer not the same
- Frame time, 3.2s vs 8s
- Focal plane temperature, -90C vs -120C
- about the orbits and backgrounds
- steal from Bev's paper
- steal from O'Dell, Markevitch papers
- about the calibration sources
- ACIS, Fe55 with Al&Ti (Fe-L), uniform illumination, getting pretty wimpy, only sampled twice per orbit
- XIS, Fe55 in corners, getting wimpy, continuously sampled (except SAA)
- measure Fe55 half-life extremely well
- Measuring CTI, ACIS vs XIS (methodology)
- ACIS, fit center pixel pulseheight vs ccdy/ccdx (binning/fitting details too)
- XIS, fit good grades summed pulseheight from top cal source corners
- (should we process ACIS the same way as XIS for comparison? only use center pixels?)
- only use Mn K alpha
- CTI evolution, plots of measured CTI vs time
- for ACIS, apply corrections for temperature and sacrificial charge
- not done for XIS; temperature is stable, background is integrated over 1 day = 16 orbits
- compare differences in rate of CTI increase (and shape?)
- (no parallel vs serial)
- FI vs BI
- low vs high orbit
- with and without CI (for XIS, when possible)
- charge trailing vs time
- trailing fraction shows how initial ACIS from low energy protons is different from ongoing, higher energy particle damage
- metric is average lost charge of all events divided by average trailed charge of all events
*FWHM evolution, plots of measured FWHM vs time
-
- ACIS and XIS can pretty much measure this one the same way
- (G02346, summed pulseheights, fit Gaussians, etc.)
- discussion related to all the above stuff
- somewhat more complicated to link to physical causes w/ charge trailing, multi-pixel events
- relate CTI and FWHM increases
- depend on BI/FI; are ACIS/XIS different?
- relate CTI/FWHM increase to measures of particle fluence, type
- maybe beyond scope of this paper)