MEASUREMENT OF PHASE DISTORTION 543 



tion is obtained in the galvanometer. Then with the system to be 

 measured inserted between the terminals of the measuring set, the 

 brushes are again adjusted until the galvanometer shows no deflection. 

 The .setting in both cases can be noted by means of a suitable scale and 

 the difference between the two settings for calibration and measure- 

 ment is, of course, the phase shift of the envelope of the modulated 

 wave in the system used for measurement. This scale can be cali- 

 brated in terms of seconds so that it measures the envelope delay 

 directly. It is evident from the above description that a one-degree 

 shift of the commutator brushes corresponds to an envelope delay of 

 .00222 second. 



This set has been found useful in measuring the phase distortion in 

 circuits below 300 cycles per second, especially recently when con- 

 siderable importance has been attached to the low-frequency distortion 

 on circuits which have been developed for program transmission. 



(3) When small amounts of distortion are to be measured and the 

 frequency range will permit, a higher frequency may be used advan- 

 tageously for modulation. Such a circuit, adapted for straightaway 

 measurements, was used for checking up the phase correction of certain 

 circuits used for television demonstrations.'' The circuit arrangement 

 has been described in the reference given. 



In this particular case, it was not expected that the distortion of the 

 overall system including the phase correction would be very great, so 

 that the chief point of interest in these measurements was the detection 

 of small changes in delay over the relatively large frequency range 

 concerned. The modulating frequency used was 250 cycles per second, 

 this larger value being used to obtain the desired accuracy. The 

 frequency required for reference at the receiving end was provided 

 by sending the modulating frequency over another circuit in the same 

 manner as that used on the circuit being measured, except that 

 a constant frequency for the carrier was used in the reference circuit. 

 The purpose of this was to introduce approximately the same delay 

 in the reference circuit as in the measured circuit because of the 

 fact that the phase shifter used at the receiving end was capable of 

 measuring only small differences in phase. 



d. Direct Measurement of Delay of Envelope 



Instead of measuring the envelope delay, which is d^fdo} by defini- 

 tion, it may sometimes be desirable to measure the delay of the en- 

 velope, say, the interval between the instant of application of a sin- 

 usoidal wave and the instant of the received wave reaching a prede- 

 termined value. One suitable arrangement which may be utilized to 



