Discussion at annual convention: Separate leakage reactance of transformer windings: Transformer harmonics and their distribution: Resolution of transformer reactances into primary and secondary reactances

Sabatoga Springs, N. Y., June 25, 1925 J. F. Peters: In Mr. Dahl's paper, as far as I can see, the assumption is made that the triple-frequency component of magnetizing current follows Ohms' law, that is, there is inherently in the transformer a triple-frequency voltage and the triple-frequency component of current that will flow is that voltage divided by a triple-frequency impedance. If this is the ease, then by decreasing the triple-frequency impedance to a small value, the corresponding current could be made quite large. Obviously this cannot be the case because when the triple-frequency current reaches a certain value, which is approximately 40 to 45 per cent of the fundamental-frequency current, the voltage wave takes on a true sine shape in which case the triple-frequency voltage disappears. It may not be possible to decrease this impedance to a very small value within the transformer, but if it is a true impedance, it can be counteracted to any desired degree externally. Also, if no triple-frequency current is permitted to flow, there will be a large triple-frequency voltage appear across each of the phases. In a transformer of commercial proportions and flux density, this voltage would amount to approximately 75 per cent of the fundamental-frequency voltage, which, in the transformer analyzed by Dahl, would amount to 100 volts triple-frequency. Then the triple-frequency current that should flow in any winding-should be that voltage divided by this triple frequency impedance. He finds in winding one a triple-frequency impedance of approximately one-half ohm. This should give a triple-frequency current in the order of 200 amperes. Actual measurements show approximately two amperes.