SLOW im;lkask uklay dksign 215 



of these is small if v'o/<p" is as small as 10 per vvni, as in I he AG relay. 

 Hence the principle variations in the release time ar(! those associated 

 with the variations in the sleeve dimensions and the core cross section, 

 which determine the values of G and (p" respectively. They cause varia- 

 tions in the AG relay of from 20 to 30 per cent. Combined with the varia- 

 tion in \I resulting from the adjustment tolerance, they result in release 

 time variations of 40 or 50 per cent. 



For the great majority of slow n^h^asc^ relay applications, much larger 

 variations have no effect on circuit performance, and for only a small 

 minority is it desirable to employ requirements assuring this measure of 

 control. It is apparent from the preceding discussion that changes in 

 construction and adjustment procedure might be made which would 

 result in closer time control: as they would add to the costs of both 

 manufacture and use, they are not economically justified. 



Engineering Data 



A specific relay design is termed a code: all relays of the same code 

 are nominally identical. A code is distinguished from other codes of the 

 same relay type by the contacts and coil provided, and by the sleeve or 

 other special features used. To design slow release relay codes, and to 

 specify their adjustment reciuirements and the performance proA'ided, 

 rec[uires means for determining from the general pull and timing char- 

 acteristics the values applying to specific cases. For this purpose the pull 

 and timing data are prepared in the form illustrated in Fig. 15. 



The pull information appears in this chart in the form of the two 

 curves marked ''hold pull" and ''release pull." These show the relation 

 between operated load and release ampere turns for the two limiting 

 cases of minimum and maximum pull, and reflect the variations in the 

 terms of (16). No individual relay releases on ampere turns above the 

 hold values, all relays release on ampere turns below the release values. 



The timing information appears in the form of a pair of curves for 

 each sleeve size that is used: one giving the minimum time, and one the 

 maximum. Only one such pair is included in Fig. lo. The difference be- 

 tween the two curves represents the 20 to 30 per cent variation associated 

 A\ith the variations in G and (p" — (po . 



The information on a chart of this type is used in conjunction with 

 engineering data for the opeiated loads of the various spring combina- 

 tions, and the changes in these loads that (an be pi'()\i(led l)y back ten- 

 sion and buffer spring adjustment. 



For each spring combination, there is a lower limit to the adjusted 



