ECONOMICS OF TKLKIMIOXK HKLAV APPLICATIONS 247 



millisecond in workino- time will reduce the average quantity of equip- 

 ment needed by an amount e(iui\'alent to S35, or e\'en more. For such 

 cases a considerable sum of mone}' can be spent on each I'elay in order 

 to make it faster. The following approach helps guide the circuit and re- 

 lay designer in the choice of relay they should make in order to gain the 

 greatest o\'er-all advantage. 



It is well known to relay designers that one of the most important 

 design parameters controlling operating time of a relay is the spacing 

 between the contacts when they are at rest. This space cannot be less 

 than a certain small distance of about 0.005" to a^'oid contact failures 

 due to ^'ibration, buildups, sparkover, etc. However, because it is costly 

 to hold adjustments accurately to a few thousandths of an inch, the 

 maximum spacing is often quite large, sometimes as much as 0.050". 

 The spread in this distance is primarih' a matter of economics; if it were 

 clear that more money could be spent on the relay to narrow down this 

 spacing while gaining material reductions in operating time as a result, 

 then some closer adjustment would be specified. Such a problem is read- 

 il}^ treated bj' summing (a) the cost penalties for differing values of this 

 spacing and (b) the cost penalties of the corresponding functioning time, 

 and finding the optimum condition that results. One such case is shown 

 below^ 



The cost penalt}' due to changing spacing may be estimated by con- 

 sidering various actual values, and determining what procedures would 

 be used in the factory in each case. Such a cost study can be made in the 

 factor}'. In illustration, assume a particular relay type deemed to suffer 

 no cost penalty if its armature stroke is allowed to take on any value up 

 to 0.050", but which cannot be less than 0.010" without impairing per- 

 formance. The actual cost per unit of various kinds of manufacturing 

 procedures might be found according to the hj^pothetical conditions given 

 in Table I. Thus a curve can be plotted of the cost penalties for each 

 setting of gap. 



The design may also be studied for the influence of the gap settings 

 on speed. The operating time of the magnet as a function of spacing is 

 readily determined by experiment, and may also be checked against 

 known operating principles, to be sure the figures are reasonable. For a 

 typical case, operating times corresponding to various spacings might be 

 as given in Table II. At the same time, the value of this functioning time 

 is given by equation (20) to be 



n _ 9^' 



Or — • 



