1292 THE BELL SYSTEM TECHNICAL JOURNAL, NOVEMBER 1954 



nors averaged 0.24 pulses per second while dials with fly-bar governors 

 decreased 0.33 pulses per second. The slight additional loss in speed 

 noted experimentally was probably due to friction in the gear mechanism 

 which is not considered by the theoi'etical analysis. However, since the 

 differences recorded are quite small, one can conclude that the theoretical 

 and experimental results are in good agreement even when concerned 

 with small changes in torque experienced during normal rundown of a 

 dial. 



CHATTER IN GOVERNORS 



It is not uncommon for governors with fine regulating ability to produce 

 an objectionable chattering noise when operated near or at the vertical 

 position. This chattering, while in most cases not particularly adverse 

 from a regulation or wear point of view creates in the mind of the lis- 

 tener grave doubts as to the correctness of the design. In severe cases a 

 sharp noise is heard during every half revolution of the governor shaft 

 as each weight alternately leaves the case and strikes against the end of 

 the other governor weight. During every revolution of the governor shaft, 

 each weight is alternately supported as show^n in the schematic. Fig. 4. 

 At this instant, the gravity moment about 5 is a maximum, and along 

 with the spring moment, opposes the centrifugal moment. If the gravity 

 component, or effective mass of the weight, is sufficiently large, a new 

 system is produced which has a critical velocity in excess of the regulated 

 speed. Since the governor speed is continually regulated by the bottom 

 weight at a speed lower than the new critical velocity, the top weight 

 falls from contact with the case. 



The magnitude of the gravity component is a function of the angle at 

 which the governor operates and is a maximum when the dial is in the 

 vertical position. As the operating plane of the dial decreases to the 

 horizontal, the gravity effect decreases to zero. Chatter will not occur 

 when the operating angle produces a gravity component smaller than 

 the difference between the centrifugal force and the spring force. 



Since the chattering effect is the result of a balance of forces on the 

 governor, it is apparent that a relationship can be derived which ^\^ll 

 express the effect in terms of governor constants. This derivation is given 

 in Appendix III and shows the chatter equation for a conventional fly-bar 

 governor to be 



.m S fW-^ (19) 



2rn(, sui /? 



