Radio broadcast .. (1922-30)

.RADIO BROADCAST. :\ TYPE OF TUBE NUMBER OF TUBES CIRCUIT Ig (mA) THIRD HARMONIC Po (WATTS) INPUT (VOLTS R-M 5) CX-112A .002 5.0% .45 19.7 B CX-112A .46 10.2% 1.16 36.5 CX-U2A 4.1 24.2% 2.62 82.4 CX-371A PUSH-PULL PUSH-PULL PUSH-PULL PUSH-PULL PUSH-PULL PUSH-PULL .002 5.6% 1.37 59.5 CX-371A 0.6 10.5% 2.89 86.5 CX-371A 6.7 21.0% 6.38 175.0 Fig. 5 out a small fraction of one milliampere d.c. which flowed through it. In some earlier listening tests we used resistors bridging the input and output to overcome a slight tendency to sing at some high frequency. While no difficulty was ex- perienced with singing in the set-up shown in Fig. 6, we continued the use of bridging input and output resistors. The percentage of harmonics registered by the oscillograph may be somewhat high because of loss of fundamental in the output in- ductances, but it is not believed that this discrepancy is very great. In Fig. 5 A is shown the wave form of the output current when 19.7 volts (r.m.s.) is indicated on the meter, E e , in Fig. 6. The load resistance R p was 18,100 ohms. The peak value of the 19.7-volt (r.m.s.) signal is 27.9 peak volts. This is slightly over twice the bias voltage which was 13.5 volts. Each tube was drawing about 2 microamperes of grid current. On increasing the signal voltage the grid current increased rapidly and both the upper and lower peaks of the output current were seen to flatten somewhat. When the signal reached 36.5 volts (r.m.s.) each tube had an average d.c. grid current flow of 460 microamperes. (Fig. SB.) The r.m.s. plate current in Fig. SA was 5.0 milliamperes and the power output 450 milliwatts, or 225 milliwatts per tube. In Fig. SB the r.m.s. plate current was 8.0 milliamperes and the power output 1160 milliwatts or 5.15 times the out- put per tube obtained in Fig. SA. Fig. 5c shows the result of increas- ing the signal to an extreme value of 82.4 volts (r.m.s.). The power output is 2620. milliwatts or 11.6 times the output per tube obtained in Fig. SA. The data and the results of an analysis of Fig. 5 (A, B, and c) are shown in Table 1 The second har- monics in Fig. 5 (A, B, and c) in order of increasing signal, are, re- spectively, 2.6 per cent., 6.8 per cent., and 4.8 per cent. The per- centage was calculated from the ratio of harmonic to fundamental. Considering second harmonics alone as a measure of the distortion, the quality is good for any signal amplitude. The corresponding amounts of third har- monic were, respectively, 5.0 per cent., 10.2 per cent., and 24.2 per cent. The percentage of higher harmonics was rel- atively small. Evidently the third har- monic is the principal distortion com- ponent from a push-pull amplifier, the OSCILLOGRAPH VIBRATOR Fig. percentage increasing (roughly) in direct proportion to the signal voltage. Similarly it will be found that with a single-tube amplifier the percentage second harmonic (principal distortion component) increases in direct proportion with the signal volt- age. As the operation of the single tube extends into the grid-current region, or into the region of plate-current satura- tion, and the second derivative of the plate current with respect to plate voltage is negative, third harmonic distortion is produced and the second harmonic dis- tortion will not increase in direct propor- tion to the signal voltage. These tests were repeated with two CX-371A tubes. Fig. 5 (D, E, and F) shows the wave form of the output current. The corresponding data are given in Table I. The results indicate, as in the tests with the CX-112A tubes, that for highest quality output the signal voltage amplitude should not exceed twice the normal bias voltage for the tube. The load resistance required for max- imum undistorted power output from a single tube, considering 5 per cent, of second harmonic as a criterion of distortion, is equal to approximately twice the plate resistance of the tube. In the push-pull stage the load resistance has less effect upon the percentage of harmonics. Since the maximum power output from a tube (or any source of power) is obtained when the resistance of the load equals the re- sistance of the tube (source), the max- imum power output will be obtained from the push-pull stage when the load resist- ance equals the sum of the plate resistances of the tubes. \\ ith the load resistance equal to twice the plate resistance of one tube and normal bias voltage, the distortion is usually negligible until the signal is large enough to start a flow of grid current. The flow of grid current lowers the grid-filament re- sistance of the tubes putting a load on the input transformer. In Fig. 6 the input im- pedance was shunted with two 10,000-ohm resistors. The effect of loading the input transformer was then eliminated. The data, Fig. 5 (A to F) in Table 1, show an in- crease in the bias voltage due to the flow of grid current through the d.c. resistance of the input. Figs. 2, 3, and 4 and the data of the table show an inter- esting result due to the flow of the grid current through the impedance of the input circuit. Fig. 2 shows the wave-form of the output current when the resistances II in Fig. 6 arc 100 ohms. The signal was 39.0 volts (r.m.s.), the d.c. grid current 1.05 milliamperes, and the output alter- nating current 10 milliamperes. The resistances R were then changed to 10,000 ohms. For the same signal voltage, 39 volts, the grid cur- rent was reduced to 0.45 milli- ampere and the output alternat- ing current was reduced to 9.15 milliamperes (see Fig. 3). Increas- ing the signal until the d.c. grid cur- rent rises to 1.05 milliamperes, as it was initially, the required signal voltage is found to be 52 volts (r.m.s.) The output is 10.4 mA. a.c. The output is 8.3 per cent, higher than was obtained with the low- impedance input, though the signal voltage was increased 33 per cent. Fig. 1 shows the wave-form of the output current. It is apparent that the distortion also is greater. In other words, a low-impedance input circuit is to be preferred. Table I — Data and Results of Analysis Figure 5A* 5B* 5C* 5D** 5E** 5F** 2t 3t 4t D.C. Plate Cur. (If) Mil- liamperes 26.4 34.0 46.0 60. 81. 33.5 30.5 33.7 D.C. Grid D.C. Grid Bias, (E t ) Cur. (I g ) Volts Milliamp. 13.5 14.7 24.4 40.5 40.8 56.7 14.5 16.8 19.4 .002 .460 4.1 .002 .60 6.7 Grid Signal (Eg) R.M.S. mils 19.7 36.5 82 I 59.5 86.5 175.0 1.05 .45 1 05 *2-cx-112\ tubes in push pull. E p = + 180 volts. EJ = - 13.5 volts. Rp **2-cx-37lA tubes in push pull. Ep = + 180 volts. Ec = - 40.5 volts. Rp 39.0 39.0 52.0 180 volts. EJ A.C. Plate Cur. (If) R.M.S. Mil. 50 8.0 12 0 12.9 18.7 27.9 10.0 9.15 10.4 Power Out- put (2f Rf.) Milliwatts 450. 1160. 2620. 1370. 2890. 6380. Per cent. 2nd 3rd Harmon ia fith 5th 2.6 6.8 48 3.1 .4 1.2 5.0 10.2 24.2 5.6 10.5 21 0 1.7 1.8 1.3 .3 1.1 2.3 2.0 0.9 2.0 7.5 1.3 3.7 1612. 1350. 171.V Input resistance 100 ohms 10000 ohms 10000 ohms 18.100 ohms. 8260 ohms. t 2 - CX-112A tubes in push pull. E p = + 180 volts. E c = 13.8 volts. R p = 16.120 ohms. 220 • FEBRUARY 1930