Calibration methods for charge integrating detectors

Since the introduction of the extremely intense X-ray free electron lasers, the need for low noise, high dynamic range and potentially fast charge integrating detectors has increased significantly. Among all the problems that research and development groups have to face in the development of such detectors, their calibration represents one of the most challenging and the collaboration between the detector development and user groups is of fundamental importance. The main challenge is to develop a calibration suite that is capable to test the detector over a wide dynamic range, with a high granularity and a very high linearity, together with a certain radiation tolerance and the possibility to well define the timings and the synchronization with the detector. Practical considerations have also to be made like the possibility to calibrate the detector in a reasonable time, the availability of the calibration source at the experimental place and so on. Such a calibration test suite is often not represented by a single source but by several sources that can cover different parts of the dynamic range and that need to be cross calibrated to have a final calibration curve. In this respect an essential part of the calibration is also to develop a mathematical model that allows calibrating the entire dynamic range, taking into account features that are calibration source and/or detector specific. The aim of this contribution is to compare the calibration for the AGIPD detector using several calibration sources such as internal current source, backside pulsing, IR pulsed laser, LED light and mono-energetic protons. The mathematical procedure used to calibrate the different sources will be discussed in great detail showing how to take into account a few shortcomings (like pixel coupling) that are common for many charge integrating detectors. This work has been carried out in the frame of the AGIPD project for the European X-ray Free Electron Laser.