c.iKRii H ri.i.i.i'iioxy ox men roi.i.iar. i.ixi.s i7,i 



of ,1 III) K.\'. pnwrr liiir (liri'C(l>- to oili- of llic line trriiiiii.ils of llii- 

 protectivf circuit. In every case tlu- circuit lias operated satis- 

 factorily. Ill no case has any of the standard apparatus lieeii <lani- 

 aj;e<l nor has there heen any evidence tliat the elements of protection 

 beyond the third, th.it is, tlie slunil coil with the mid-point grounded, 

 h.i\e lieen c.dled upon to fmiction. 



Transmission I.kvki, CiiARACTF.RtsTrcs 



Fij;. 22 shows the attenuation (expressed in transmission units) of 

 the hi^h frecjuency line versus the carrier frec|iienc>- of K.C. It will 

 Ih- noted that o\er the rani;e from .'lO K.C". to l-")0 K.C the vari.uion 



HIGH F;?E0UENCY line (T.U.) 

 JO iO 50 bO 



Kig. 13 — \'ariation of Overall ('.a!ii with llu- .\ltemiation of the \Ug,h f-'requeiicy Line 



in attenuation is less than o T.l'. This curve was made with a constant 

 audio frequency input of 3.35 mils and an output of 3.35 mils from 

 the carrier circuits, the audio frcfjuency being 1,000 cycles. The 

 \ariation of audio frequency level with the attenuation of the high 

 frequency line is shown in Fig. 23. The observations given in Fig. 24 

 were made on an artificial transmission line in which the line constants, 

 and therefore the attenuation, could be readily changed without 

 changing the carrier frequency. The shape of this curve is a function 

 of the receiving circuit since the audio input, carrier frequency- and 

 the mtxlulated output of the transmitting circuit are maintained 

 constant. It shows that for audio frequency levels lying between 

 — 10 and +10 T.U. the equivalent is appro.ximately a straight line 

 fimction of the attenuation of the high freciuencN' line, and ihat 

 therefore the receiving circuit is not overloaded. 



Fig. 24 shows the audio frequency load characteristic. This curve 



