On the dead time problem of a GM counter.

Characteristics of the dead time of a Geiger-Müller (GM) counter was studied by using a decaying source method and tests were performed for two-parameter dead time models including the general models. Activated (56)Mn was used as the decaying source and the counting rates were observed in one minute steps for 20 half-lives using a GM probe and a multichannel scaler. Measured data were compared with predictions of a generalized dead time model, a dead time model of nonextending-extending (NE-E) types, and a model of extending-nonextending (E-NE) types in series. The NE-E series model described the measured counting rates better than the other models. The model rates were consistent with the data within 10% up to a true rate of approximately 10 kcps, however, its application in the region of higher counting rates was limited. Possible causes of this limitation are discussed. For practical purposes, a phenomenological model is suggested to describe measured counting rates above approximately 10 kcps.