Next: Relative Quantum Efficiency Uncertainty Up: Measurement Method and Analysis Previous: Spatial Variations in CCD
The average count rate for each CCD is used to estimate pileup correction factors, which are then used to obtain corrected nominal values of the QE ratio and the sigma value. The pileup correction is typically very close to unity when both the flight and reference CCD have the same exposure time. Larger corrections were found in several cases where the reference CCD used older (and slower) electronics than the flight CCD and therefore recorded higher counts/frame and pileup values. Also, data comparing FI CCDs to BI CCDs can be affected by slight differences in pileup behavior exhibited by the two types of detector.
The pileup model and parameters are explained in section 4.4; flux scaling studies were used to fit typical FI and BI pileup model parameters at energies above 1.5 keV. These studies consisted of recording data from fluorescent targets at different current values on the commercial X-ray tube (at constant voltage); analysis of total charge collected indicates that the X-ray flux is linear with respect to this current. For energies below 1.5 keV (O and F data), parameters measured with the Al source were employed. A more refined treatment for these energies is planned later. Note, however, that empirically fit pileup parameters were typically independent of energy below 3 keV. For a given flux level, pileup is greatest for high energy photons because their charge clouds typically spread to more pixels, especially if the X-ray is energetic enough to penetrate the depletion layer of the CCD before initiating a photoelectric interaction.
Pileup effects may be expressed for one measurement in terms of a correction factor (f), defined as the ratio of counting rate (corrected) to counting rate (raw). Typical correction factors (f) are shown in Table 4.25 for the relative QE calibration of flight CCD I1 (w193c2) vs. reference CCD w103c4. This is an example in which the reference CCD used the older electronics and was limited to a 7.15 second exposure time, whereas the flight CCD using engineering versions of the flight electronics was capable of running with 3.28 second exposures if the flux was high enough to warrant it. As mentioned above, the effect of pileup correction is larger when the two CCDs employ different integration (frame exposure) times.
|Energy||I1 f||I1 exp||Ref f||Ref exp||f ratio||+/-|
|.525||1.061||7 sec||1.071||7 sec||0.991||0.013|
|.677||1.031||7 sec||1.041||7 sec||0.990||0.001|
|1.74||1.046||3 sec||1.109||7 sec||0.943||0.016|
|2.01||1.022||3 sec||1.051||7 sec||0.972||0.009|
|4.51||1.027||3 sec||1.059||7 sec||0.969||0.006|
|5.89||1.065||3 sec||1.146||7 sec||0.930||0.014|
|8.03||1.068||3 sec||1.144||7 sec||0.934||0.012|