Abstract
Debye's theory of polar molecules has been extended to give simple expressions for conditions of maximum loss per cycle in terms of equivalent circuit bridge or substitution measurements. The dielectric loss in a good grade of viscous mineral insulating oil has been separated into two components; one resulting from conduction, and the other showing characteristics with frequency and temperature qualitatively explained on the basis of the presence of polar molecules in the oil. A quantitative check shows the order of magnitude of this second loss to be the same as that which would be predicted by Debye's theory. The experimental curve shows a wider frequency response than the theoretical curve, indicating that the simple theory is not sufficient to account for the observed data. This may result from the presence of polar molecules of many different sizes giving an average response rather than molecules of one size only, as assumed in the theory. The size of the polar molecules necessary to give the observed effects was calculated from Deybe's theory and found to be of the right order of magnitude. Data taken over a wide range of frequency and temperature indicate the danger in drawing conclusions as to the characteristics of a material from data taken for limited ranges of experimental conditions. The curves show that either in-increasing or decreasing power factor can be obtained as a function of either frequency or temperature for certain narrow limits of experimental conditions. The dielectric loss in good commercial mineral insulating oils at power frequencies and operating temperatures results from conduction only, the contribution of polar molecules to the loss being negligible. Before a definite statement is made regarding the mechanism of the observed loss, studies should be made to determine whether a satisfactory theory can be developed which is based only on the motion of charged particles in a viscous medium.

This publication has 3 references indexed in Scilit: