1494 THE BELL SYSTEM TECHNICAL JOURNAL, NOVEMBER 1957 



to 



Q 



z 

 o 

 u 



to 



z 



a. 

 o 

 a. 

 tr 



25' 



50' 

 SHAFT ANGLE, ©rn 



I'lS' 



Fig. 5 — Error of a 27th order vernier resolver over | vernier interval. 



27th order vernier resolver this common factor is 1 and consequently the 

 slot harmonics are of order: 9, 11, 19, 21, etc. 



The effect of the slot harmonics can be reduced by the following 

 means : 



a) Selecting the dimensions as well as the number of rotor and stator 

 teeth such as to keep the higher order flux components low. 



b) Using a ''skewed" rotor or stator, in which successive laminations 

 are progressively displaced with respect to their angular orientation. 



III. PERFORMANCE 



Clifton Precision Products Co. built experimental resolver models of 

 order 26, 27, 32 and 33 using the laminations shown in Figs. 3 and 4. 

 The best results were obtained with 27th order resolvers. Their per- 

 formance is described in the following sections. 



3.1 Repeatability and Accuracy 



The repeatability is better than d=3 seconds of shaft angle. 



Figs. 5, 6 and 7 show the error curves taken on a 27th order vernier 

 resolver after compensating with trimming resistors for the fundamental 

 and second harmonic error with respect to the vernier interval. (In es- 

 sence, the effect of these trinnning resistors is either to add or to sub- 

 tract a small voltage to one or both of the resolver signals.) Fig. 8 shows 

 an error curve before trimming.* 



3.2 Temperature Sensitivity 



The error introduced by a temperature change of 70°C is less than 25 

 seconds of shaft rotation. 



* The error curves really represent the combined error of the tested resolver 

 itself plus that of the testing apparatus. 



