Badanie nieliniowości w precyzyjnym pomiarze przedziału czasu dla wybranych typów częstościomierzy-czasomierzy

Celem niniejszego artykułu jest zaprezentowanie metody i wyników badań nad nieliniowością częstościomierzyczasomierzy w pomiarach precyzyjnych przedziałów czasu, prowadzonych w GUM przy współpracy z AGH. Dokonano porównań wyników pomiarów dla częstościomierzy-czasomierzy SR620 i HP 53132A zarówno stałych i zmiennych przedziałów czasu. Potwierdzono przydatność dwukanałowego generatora/dzielnika typu Dual Channel 1/100 PPS Generator oraz komparatora częstotliwości typu A7-MX.

In this paper we present the method and the last results of verification of nonlinearities in time interval measurements with the usage of Time Interval Counters (TIC), performed at Central Office of Measures (GUM) with cooperation of AGH University of Science and Technology. We consider the non-ideal operation of TIC manifested by fluctuations of differences between the current indication of TIC and the real value of the measured time interval. We observed instabilities of indications of TIC, even if the measured time interval is keeping constant. For verification of TIC operation at continuously changed time intervals, it was set up a special measurement system that performs simultaneous measurements of the phasetime changes between two standard frequency signals (with the usage of A7-MX standard frequency comparator) and time intervals between two pulse signals (with the usage of TIC under test). Due to applying a special, developed by AGH, dual channel 1/100 pps generator, the measured 1 pps (1 pulse per second) and 10 MHz signals are standardized and the phasetime between 1 pps and 10 MHz signals are kept in pairs constant. Differences between the indications of TIC and the proper measurement results for frequency allow to verify internal nonlinearities of TIC in time interval measurements. Here, we show and shortly discuss the results obtained for SR620 and 53132A universal counters. The obtained range of fluctuations of indications of SR6200 with relation to the measured time interval approaches about 190 ps, and for 53132A – about 450 ps, whereas the observed short-term noise of measurement for SR620 is about 10 ps, and for 53132A – about 100 ps typically. It was confirmed the complex and random influence of many factors on operation of TIC. Our investigation will be continued with the usage of 100 pps (100 pulses per second) signals and with automatic switching of the measured pulse signals in pairs.