Uros Kovacevic

Measurement uncertainty of fast pulse voltages measurements with capacitive divider

Causes of measurement uncertainties might be numerous and as the rule, not all of them can be considered. Two basic types of measurement uncertainty are the type A and the type B. Measurement uncertainty of type A is determined only by method of statistical data processing. A result of this is that measurement uncertainty of type A exists only in case of a measurement repeated several times. Measurement uncertainty of type B is determined by all other methods, except by the statistical method. Measurement uncertainty type B is used regardless if individual or repeated measurements are the issue. The combined measurement uncertainty represents a resultant value in case when a measuring uncertainty originates from two or more different components. The influence of measurement uncertainty of type A and B as well as combined uncertainty during the measuring of fast pulses by using capacitive divider have been considered in the paper. A budget of measurement uncertainty has been formed for that purpose, the one that is characteristic for measuring fast pulses with a capacitive divider. Several types of capacitive dividers have been designed and used (with different types of high-voltage capacitor, low-voltage capacitor and with adjustable resistance) that provide minimization of certain measurement uncertainty budget components. Measuring of pulse voltages under well-controlled laboratory conditions has been performed for each type of capacitive divider. The parameter of these experiments will be pulse rising time (in the ranges of ms, μs, ns). On basis of the obtained results, it is possible to conclude which components of the voltage divider (construction, materials, capacitance, inductivity etc.) influence the measurement uncertainty. The paper provides recommendations for choosing the optimal type of capacitive divider in accordance with minimal measurement uncertainty.

The construction of capacitive voltage divide for measuring ultrafast pulse voltage

The problem of pulse voltage measurements in the nanoseconds range has become of great importance with ongoing problems related to the construction of pulse fusion generators as well as with simulations of electromagnetic pulses related to the atmospheric nuclear explosion. Ultrafast pulses in the nanoseconds range can be measured by the capacitive voltage divider. The problem that arises with ultrafast voltage occurrence is inductivity. Under conditions of the nanoseconds pulse increase rate, conductors of about few centimeters should be regarded as inductivity. Therefore, capacitive divider can be treated as an equivalent RLC circuit. The ohmic resistance R in that equivalent circuit originates from the characteristic impedance (either 50 Ω or 75 Ω) and, depending on its design, it can introduce an additional parasite inductivity. The aim of this paper is to consider and analyze the constructive solutions of capacitive voltage divider for measuring ultrafast voltage occurrence. Three types of capacitive divider will be constructed for that purpose. The high-voltage capacitor of each divider will be gas capacitor (in order to avoid both electrostatic and electrodynamic influence on measurement accuracy). The variations between divider types will be made for low-voltage capacitor. The designed low-voltage capacitors will be: 1 - in the shape of 10 mica capacitors connected in a parallel with a discrete ending resistor of 50 Ω; 2 - in the form of gas trimmer capacitor with a discrete ending resistor of 50 Ω, and 3 - the continual mica capacitor with waveguide resistor of 50 Ω (L=0). Ultrafast pulse response will be measured with dividers constructed in such manner, and will be compared with the corresponding response obtained from the numerical simulations.