Search

This paper is aimed at expressing the expanded combined measuring uncertainty of capacitive divider with concentrated capacitance on a high-voltage scale. The tested capacitive divider is constructed to have a minimal measuring uncertainty. Measuring uncertainty of type A is expressed by means of the statistical processing of an experimentally determined random variable of pulse voltage amplitude and random variable of step voltage amplitude uncertainty budget of type B was derived from: 1) the influence of uncertainty values of a high-voltage and low-voltage capacitances on the transmission ratio; 2) the influence of frequency on the transmission ratio; and 3) measuring uncertainty type B of the used instrument. The experiments were performed under well-controlled conditions. Measuring uncertainty of type B for the first budget component is obtained theoretically using the Monte Carlo method, for the second budget component by using experimental-analytical method, while for the third budget component it is taken from the instructions manual of the manufacturer. The Gaussian distribution is attributed to the influential effects of the instruments measuring uncertainty type B. The results obtained show that most impact on measuring uncertainty of type B has the measuring uncertainty type B of the used instrument and the measuring uncertainty type B of divider by the influence of frequency. Values of expanded combined measuring uncertainty for 95% of coverage probability are 5.2% for frequencies equal to 50 MHz, 6.1% for frequencies equal to 1 GHz, and 7.1% for frequencies equal to 4 GHz. Comparing to our results for relative error of capacitive divider for measuring fast pulse voltages, it is shown that the tested capacitive divider has very good characteristics.