128 THE BELL SYSTEM TECHNICAL JOURNAL, JANUARY 1957 



90 

 85 

 80 

 75 



70 



a 

 y 65 



>- 



5 60 



uj 



u 



(r 55 



LU 

 Q. 



50 

 45 

 40 



35 



30 



JFMAMJJASONO JFMAMJJASONO 

 1954 1955 



JFMAMJJASONO 

 1956 



Fig. 7 — Resistor yield. 



wire on a shuttle, wind by hand the turns required to produce an in- 

 ductor. These varied from a very small inductor, smaller in diameter 

 than a pencil, to a fairly large "figure eight" inductor with turns having 

 a major diameter of about Ij inches. Each layer of a winding was in- 

 spected with a microscope to insure that the wire had not been twisted 

 or kinked, or that the insulation was damaged or uneven. Some of the 

 shuttles became fairly long so that they could hold the amount of wire 

 required to make a continuous winding. The operator's handling of this 

 shuttle, as she moved it down around the openings in the methacrylate 

 part, or placed it on a bench to proceed with the interleaving tape, de- 

 manded considerable dexterity and concentration to insure that the 

 shuttle was not turned over — which in effect would put a twist in the 

 wire. Although best known means were used to sort cores for their mag- 

 netic properties prior to the time a winding was made, the limits on the 

 inductors themselves were so close that subsequently a large number of 

 windings were lost. The best cores that could be selected, plus the best 

 winding practice, could not produce 100 per cent of the inductors within 

 the required limits. Crazing of the insulation on the wire; cementing 

 together of two methacrylate parts or of permalloy cores into pockets of 



