HIE socxns or sriiECU 



597 



wiTc ilt'p.irtures from thi-se roiulitions in tlu' fri-tiuoncy iiitiTwiI from 

 /(.•ro to ")(M) ryclfs for whicli allowance had to In- made. 



In the amphtier the effect of capacity reactance was nearly ehnii- 

 nated. Owing to the small remaining capacity reactance there was a 

 phase leail of amplifier current with respect to driving force which wag- 

 appliixl to offset the excessive phase lag in the condenser transmitter 

 at the low frequencies. The particular adjustment of amplifier finally 

 arrived at represented the best compromise, considering the difficulty 



Fig. 5 — Overall frequency char.ictcristics of amplitude and phase of the recording 



system. Curve .'1; Oscillographic amplitude per unit of pressure on transmitter 



diaphragm. Curve B: Phase lag of oscillographic amplitude behind pressure on 



diaphragm 



encountered with the transmitter characteristics. With this com- 

 promise made there was an unavoidable phase lead in the whole appa- 

 ratus for frequencies below 125 cycles, but this was not serious as 

 most of the speech energy is in higher frequencies. After all final 

 adjustments were made the overall frequency characteristics of ampli- 

 tude and phase were as shown in Fig. 6. Thus ultimately there was 

 obtained a system with practically uniform amplitude characteristic 

 from o(X) to 5,000 cycles, without serious departure from this level for 

 frequencies from 50 to 500 cycles; and with phase lag nearly a linear 

 function of frequency from 125 to 5,000 cycles, after passing through a 

 perirxl of lead in the narrow interval from .50 to 215 cycles. 



