When a HgCdTe IR detector is used for low flux astronomy mission, persistence is one of the main characteristic to calibrate. However, it is not yet a true figure of merit since no generic persistence characterization protocol exist. Indeed, a protocol is usually developed for each mission in order to calibrate this parameter according to the specific needs and specific operating point of the detectors (Integration time, temperature, flux dynamics ...). It is consequently not possible to compare persistence performance of two detectors if the characterization protocol differs and no definitive calibration scheme exist.

In the frame of ALFA and ASTEROID programs, we started characterizing persistence at CEA LETI, CEA IRFU and Lynred a few years ago. The objective was to improve the performance of our built in house detectors in terms of persistence. The goal of this paper is to present our approach to identify which protocol is best suited to compare persistence between two detectors manufactured with different technological steps. With a dedicated instrument, we compare systematically persistence protocols in a controlled environment and on a single detector. Electrical and optical stimulation are compared as well as several stimulation parameters (amplitude and stimulation duration, ie soak time).