Abstract :
This paper treats of the sources and effect of errors entering into the determination of the power efficiency of rotating electric apparatus. It is shown that where the efficiency is determined by the so-called method of “summation of separate losses” the effect of possible errors in measurement is much less than when determined by the input-output method. In the input-output method of efficiency determination all measurements enter directly into the result, so that errors in them have a maximum effect. Attention is called to the different classes of errors likely to arise, viz., purely instrument errors, observation of reading errors, and errors due to changes in the operating condition of the machine under test. Particular stress is laid on observation errors, these probably being especially large under commercial operating conditions, due to excessive swinging of meter pointers. This swinging of meter pointers may be considerably reduced under special conditions of power supply and loading, and fairly consistent efficiencies then obtained. It is this necessity of operating under special conditions and observing all precautions possible in the layout and conduct of an input-output test that entirely offsets its apparent simplicity. It is next shown that, while the efficiency as determined by the loss method is not the true operating efficiency, it can be so modified by certain factors, depending upon the type of machine, as to approach very closely to that value — much closer, in fact, than the average input-output test. The probable relative accuracies are illustrated in Fig. 1. In consideration of the equivocal results usually obtained, and in view of the laboratory methods of testing required for making accurate input-output tests, the writers do not feel that that method is to be recommended
