CHAP. XII] INDUCTANCE OF WINDINGS 229 



67. The Leakage Reactance in Synchronous Machines. The 

 physical nature of the armature reactance in a synchronous 

 machine is explained in Art. 46; the influence of this reactance 

 upon the performance of a machine is shown in Figs. 37, 38, 40, 

 and 41. The problem here is to calculate the numerical value 

 of this reactance for a given machine, using eq. (150) with 

 empirical coefficients (P'. 1 



It may also be stated here that for standard machines, partic- 

 ularly in preliminary estimates, the ix drop is sometimes taken 

 as a certain percentage of the rated voltage of the machine 

 instead of estimating the inductance from formula (150). In 

 synchronous generators the ix drop at the rated volt-ampere 

 load varies from 5 to 10 per cent of the rated terminal voltage. 

 In synchronous motors, where some inductance is useful, the 

 ix drop ranges from 8 to 15 per cent of the rated voltage. For 

 60-cycle machines, and for machines with a comparatively large 

 number of armature ampere-turns, values must be taken nearer 

 the higher limit. For 25-cycle machines, and for machines 

 with a comparatively small number of armature ampere-turns, 

 values must be taken nearer the lower limit. A considerable 

 error in estimating the value of ix has but little effect upon the 

 calculated performance at unity power factor, because the vector 

 ix is then perpendicular to e (Figs. 37, 38, 40 and 41). However, a 

 considerable error may be introduced at lower values of the 

 power factor if the reactive drop ix has not been estimated 

 with a sufficient accuracy. 



The values of (P f for synchronous machines are different 

 from those given above for induction machines, because of the 

 absence of any secondary current-belts. Parshall and Hobiirt - 

 give the following values for (Pi : 



1 For a theoretical calculation of the coefficient <P', see Arnold, Wechsel- 

 stromtechnik, Vol. 4 (1004), pp. 41-52; Hawkins and Wallis, The Dynamo, 

 Vol. 2 (1009), pp. 901-904. For a comparison between the calculated and 

 actually measured values see an extended series of articles by J. Rezelman 

 in La Lumtire Electriqw, 1909-1911, and in The (London) Electrician. 



1 Electric Machine Design (1906), p. 478. These values are corroborated 

 by those obtained by Pichelmayer; see his Dynnmobau, 1908, pp. 208 and 

 604. Pichelmayer's values for <P/ (which he denotes by ) are somewhat 

 hi nh because the end-connection leakage is not considered separately. 

 Arnold's values, giv.-n in his WechsebtromtechniJc, Vol. 4, p. 280, should bo 

 used with discretion, because they apply to a different formula; namely, 



