84 BELL SYSTEM TECHNICAL JOURNAL 



as the upper curve of Fig. 9. As may be seen here, it was found to be 

 nearly linear — with a maximum deviation of about 10°. When the 

 experiment was repeated with a string that was resistance damped, 

 considerable phase distortion was found as shown by the middle curve 

 of the plot. 



This method of measurement cannot be used for much higher fre- 

 quencies because of the difficulty of stretching a string to appreciably 

 more than 6000 cycles. The amount of the phase distortion in an 

 instrument equalized to higher frequencies may be judged, however, 

 by taking oscillograms of square-front flat-topped waves and noting 

 the irregularities produced. The phase correction required was de- 

 termined by making such oscillograms with a resistance-capacity phase 

 corrector in the circuit, and adjusting the phase corrector to bring 

 about a minimum amount of distortion. An electrical equivalent of 

 this experimental network, giving the same phase correction but with 

 negligible attenuation, is included as part of the equalizing circuit of 

 the new oscillograph. Although it is realized that the resulting phase 

 characteristics are not perfect up to 10,000 cycles, the oscillogram of 

 Fig. 10 shows that there is no great amount of phase distortion present. 

 It has been found that the amount of distortion indicated here is not 

 usually of practical importance. 



Fig. 10 — Oscillogram of square front, flat top wave from an instrument tuned to 4500 

 cycles and equalized to 10,000. Abscissa divisions are .001 second. 



A few years ago a recording oscillograph, of the string type, was 

 developed by Bell Telephone Laboratories, which would satisfactorily 

 record frequencies over the part of the voice range important in tele- 

 phone work. It represented a distinct advance over the oscillograph 

 of similar type developed during the war for locating enemy guns by 

 sound ranging and improved subsequently for studying circuit phe- 

 nomena. This earlier oscillograph "* would record frequencies up to 

 200 cycles per second and had facilities for developing and fixing the 

 paper record at the rate at which it was exposed, while the improved 

 oscillograph increased the frequency range to 3000 cycles. Although 



^Bell Laboratories Record, March 1927, p. 225. 



