190 BELL SVSIEM TECHNICAL JOURNAL 



coils made of stranded wire and with smaller inductance, hut with 

 larger distributed ca|>acil\-. 



In practice, it is not important to go to extremes in order to reduce 

 the distributed capacity by one or two nixf. because the coil will 

 always be shunted 1)>- the tube capacities, which are of the order of 

 10 fjLfif. It may be mentioned that the distributed capacity of the 

 coil used in the above experiment is 3.5 /i/x/. This means that the 

 constructional details of such a coil are not very important, and the 

 coil may be made as a single layer coil or as a coil wound in one or 

 several sections of rectangular or square cross-sections, but in all 

 cases it will be found that coils of the same inductance will have very 

 closely the same resonance frequency pro\idcd that the same tubes 

 and leads are used in all cases. 



Some experiments made with a choke coil (or auto transformer) 

 at about 50,000 cycles show tiiat tlic Inrmulas given al)o\e ma\' be 

 also used here. 



The coil used in these experiments was wound on a core of iron 

 dust and made with square cross-section. The total inductance of 

 the coil was .33 henries and provisions were made so that the plate 

 of the amplifier tube could be tapped across any part of the coil. 



The circuit diagram was the same as that given in Fig. 3 with the 

 exception that the condenser C was omitted. The maximum amplifi- 

 cation curve for this coil, used as a choke coil, is given by Fig. 7, 

 curve A. The step-up ratio necessary to obtain maximum amplifica- 

 tion was 1:16; i.e., the plate was connected across 1 Itl of the total 

 number of turns. 



The resistance of llie coil is obtained from a reson.nice curve as 

 before : 



R' = 27rL(/, -/.) = 27r X .33 X 1300 = 2700 . ,hms, 



and inserting iliis in formula ((>) gi\"es: 



0.1 2rX54800X.33 .. 

 ^ \/22,000X2700 



while liie lApiTlnunl i;a\t- It.."). 



Un I'ig. 7 are also given the amplification curxes B. C and P for a 

 step-up of 1:4, 1:1 and 1:48, resijectiveh'. 



In the two cases B and C, the selectivity of the circuit is determined 

 almost entirely by Rp, the resistance of the circuit itself being negligi- 

 ble, while in case 1) the selecli\il\' is pr.icticalK' delermincd b\' the 

 resistance of the coil itself. 



