THE RING ARMATURE TELEPHONE RECEIVER 117 



their frequency ranges. These solid curves were measured with the receiver 

 on a standard closed coupler^ of six cubic centimeters volume, using a source 

 impedance of 128 ohms, and the ordinate scale is given in terms of the 

 square of the pressure generated in the coupler per unit of electrical power 

 available to a pure resistance of 128 ohms substituted in the electrical cir- 

 cuit in place of the receiver. The new receiver shows 5 decibels improvement 

 in output level and about 500 cycles per second extension in the frequency 

 range. This represents a very substantial increase in transducer efficiency, 

 and the increase in range results in a quite noticeable improvement in the 

 quality of speech sounds. The low frequency cut-off obtained by a hole in 

 the diaphragm of the Ul receiver, mentioned in the preceding section, ap- 

 pears in the response-frequency characteristic below 350 cycles per second. 

 The irregularities in the characteristic of the Ul receiver at 450 and 1200 

 cycles per second are not inherent in the receiver, but are acoustical effects 

 of the passageway molded in the handset handle, which serves as a conduit 

 for the wires connected to the receiver unit. This is indicated by the dashed 

 line curve, which shows the response-frequency characteristic of the receiver 

 when the passageway is plugged at the receiver bowl of the handset. No 

 adverse effect of these irregularities has been discerned. 



For comparison with the closed coupler characteristic, the dotted curve 

 in Fig. 4 shows the pressure generated by the Ul receiver at the entrance to 

 the human ear. This curve is an average of 90 observations on 30 subjects 

 measured by a small diameter search tube inserted into the outer ear cavity 

 through the receiver cap and connected to a microphone external to the 

 handset, so that the ear is used as a passive coupler. Figure 5 shows the 

 manner of using the apparatus, which includes a 640AA condenser trans- 

 mitter mounted on the handset and coupled to the search tube through a 

 very small chamber. It will be noted that the curve of Fig. 4 taken on a 

 human ear shows increased low frequency cutoff because of leakage be- 

 tween the receiver cap and the ear, but that otherwise the 6 cc. closed 

 coupler response is a good representation of the data taken on the ear. Con- 

 siderable deviations from the average curve were observed from person to 

 person, as illustrated in Fig. 6 which shows the maximum and minimum 

 values of all measurements at various frequencies and the standard deviation 

 of the measurements at three frequencies. 



An interesting comparison of the performance of the Ul and HAl re- 

 ceivers is made by holding the receivers slightly away from the ear. It is 

 observed that the degradation in response caused by this condition, which 

 represents a very large amount of acoustical leakage between the ear and 

 the receiver cap, is much greater in the HAl receiver. The effect is illus- 



9T;^e 1 coupler per A.S.A. Standard Z24.9-1949. 



