1. Investigation into the strong residuals seen
near the Oxygen edge in MOS spectra of
bright continuum sources.
Rate Dependent CTI effect ?
Steve Sembay

2. Event 1
Readout direction
Event 2
Trap

3. Emission timescales for filled traps
(Holland et al. 1993, NIMS, A326, 335)
σn electron capture cross section
Xn entropy factor
υth electron thermal velocity
Ncdensity of states in conduction band
E energy level of trap
T temperature

4. Level (eV) te (-90°C) (-100°C) (-120°C) 0.17 0.44
Emission timescales for filled traps
(Holland et al. 1993, NIMS, A326, 335)
Level (eV) te
(-90°C)
(-100°C)
(-120°C)
0.17
0.07 µs
0.19 µs
1.1 µs
0.44
3.8 s
22.9 s
1.4 x 103 s
c.f. row transfer time of MOS = 15 µs

5. Around O edge χ2 -> 3.90 – 2.13 with gain offset = 8.7±1.0 eV

6. Around Si and Au edge: χ2 -> 1.025 – 0.998 Offset ~ 12.5 eV

8. MOS v RGS Line Energy: N line ~ 432 eV J. Carter (Mallorca 06)
Most MOS Large Window Mode
SW Obs.

9. Closed Cal Al line in small window mode

10. Closed Cal Al line in small window mode

11. Is there a “possible” gain shift at O: CTI dependent on rate?
Method 1: Analyse spectra from different regions

12. Sample of bright AGN/BL Lacs: The “Usual Suspects”
MKN
MKN 3C H
PKS
PKS
ARK H
PKS
PKS 3C
MKN

13. MOS1
Count Rate v Offset
In Boxes
Before/After Cooling

14. MOS2
Count Rate v Offset
In Boxes
Before/After Cooling

15. Is there a “possible” gain shift at O: CTI dependent on rate?
Method 2: Flag precursor events from event file

16. Flag Bad
Readout direction
Flag Bad

17. Flagged events

20. |Summary:
Observed offset is not affected by removal of
precursor events in CCD RAWY direction
This suggests offset is not caused by incorrect
application of CTI measurement due to trap
filling in CCD.