Determination of the loss tangent for solid dielectrics at high frequencies

Abstract

Cells with a micrometer screw are still used for testing solid dielectrics at frequencies from units to hundreds of megahertz despite the development of dielectric measurement technology. We present the results regarding elimination of the negative impact of successive stray inductances (L_n) and active resistance (r_n) of such a cell on the accuracy of the dielectric loss tangent (tanδ) determination with allowance for the final value of the Q-factor (Q_k) of the circuit inductor in resonant measurements. Tests with and without a dielectric sample were carried out under the condition of maintaining the cell capacity value. No information is available in the regulatory documents and technical literature about taking the aforementioned parameters into consideration. It is shown that the impact of Ln consists in the overestimation of measured tanδ values. To eliminate this effect, we proposed to introduce an additional factor which depends on L_n, frequency and capacity of the sample into existing formulas used for tanδ calculation. It is also shown that the tanδ error increases significantly with a decrease in the ratio of the dielectric loss characteristics of the measuring circuit with and without a sample which can be attributed to r_n and Q_k. The requirements for r_n and circuit damping without a sample are formulated to ensure the desired accuracy. Conditions providing the possibility of assessing the suitability of a particular equipment for tanδ measurements are specified on basis of the developed requirements. The results experimentally proved in resonance measurements with a Q-meter can be used in high-frequency tanδ measurements to provide the required accuracy.