Discussion and Summary

For small fluxes, the pileup models seems to work well. For larger fluxes, systematic differences can be seen between the MEG+1 and MEG-1 orders. This appears to be associated with differences between the frontside and backside CCDs.

The following figure shows a plot of the number of counts per angstrom per frame vs wavelength in the MEG plus and minus first orders for a number of bright HETGS observations:

As the figure shows, the MEG+1 count-rate "saturates" at about 12 counts per angstrom per frame on the ACIS-S3 (backside) chip, whereas the MEG-1 saturates has a higher value on the S3 (frontside) chip. The red-dashed line in the figure shows the location of saturation point of the HETGS pile-up model.

The conclusion that one can draw from the figure is that the pile-up model underestimates the saturation point on the front-side chips (for wavelengths above the silicon edge) and over estimates the saturation point on the backside chips:

Work is currently underway to include frontside/backside differences into the model.

Previous: Cygnus X-1