In this work the reproducibility of the efficiency of different HPGe detectors was analyzed under a series of different conditions. The detectors studied were plugged either to a regular analogical amplifier or to a digital signal processing (DSP) device, in order to evaluate the possible differences between either setup. Detectors were inspected by performing a long series of sequential measurements with standard calibration sources, and comparing the standard deviation of the number of counts per second in each series to the uncertainty of the individual measurements. Detectors were also subjected to distinct count rates, in order to verify the possible experimental issues associated with this parameter. The results allow a discussion on the stability of the detectors’ efficiencies over a few days, the possible dependence with the count rate, and the estimation of the uncertainty related to the efficiency variation.