Radio Broadcast (May 1928-Apr 1929)

were supplied from an oscillator as shown in Fig. 4, and the performance of the tube and output transformer circuit were examined for undistorted output at 500 cycles. The input voltage to the stage, as determined by the reading of the meter, Mi, and the equal resistors, Ri, was increased in small steps and the output power at each step computed as the product of the 8000-ohm resistor, R2, and the square of the current through it. In this way the output in watts per volt input squared could be computed, and plotted as shown in Fig. 5. This curve was taken with two uxITIa tubes operating with 180 volts on the plates, and the choke of Fig. 4. It will be noted that the output available before the gain falls appreciably is about two watts. The manufacturer's tables give an output of 0.7 watt for this tube which would permit of 1.4 watts output with the two tubes operated in simple parallel arrangement. The difference, 0.6 watt, in output is due to the push-pull arrangement. As a matter of fact, the ear does not notice the distortion until the output per volt squared has dropped about 2 db, allowing the operation of this stage to 2.8 watts output. This output is sufficient to operate a dynamic loud speaker with sufficient undistorted volume for the home and for small halls and has the advantage over a 250 tube of employing low voltages that are available easily at low cost. Where larger outputs are required the 250 tubes will and can be used with the same choke. The performance of this single stage with 250-type tubes was measured in the same way as the previous combination of 171's and the results are plotted in the curve of Fig. 6-a. It will be seen that this arrangement is suitable for operation where about 12 watts are required for coverage of considerable area. The impossibility of its operation for home use at full volume is indicated in the fact that none of the smaller dynamics on the market 0L L 1 1 1 1 0 1 2 3 4 5 OUTPUT WATTS Fig. 5 will stand this output without rattling badly on the high frequencies. Attention is called to the fact that both of the curves discussed were made with an 8000ohm resistor in the output circuit. It has been shown many times that maximum undistorted volume (not maximum volume) is realized from a tube circuit when it is operating into a resistance of double its own plate resistance. The condition for maximum volume, neglecting distortion, is satisfied when the output resistance is equal to the tube's resistance. In order to illustrate the latter condition, the curve of Fig. 6-b was prepared using a 4000ohm resistor in the output circuit. It will be seen that while the gain is somewhat greater than in curve a, the bend indicating distortion occurs at a much lower level. This illustrates a common malady in power amplifiers. In order to secure high volume the temptation is to operate the tubes into a low-impedance output circuit when properly the amplification should be increased in the preceding stages. As seen in the curves of Fig. 6, the volume is increased by this means, but the 0 2 4 6 8 10 12 14 16 18 OUTPUT <WATTS> Fig. 6 l.Oi 1 1 .0.7 25 100 400 1000 5000 FREQUENCY JN CYCLES Fig. 7 -12 c B fit? _ "T f 25 100 400 1000 5000 FREQUENCY IN CYCLES Fig. 8 undistorted output available is actually reduced. In view of the fact that many loud speakers do not have sufficiently high impedances to permit operation under conditions similar to those of Curve a if the loud speakers are connected to the plate terminals of the power tubes (terminals A and A1, Fig. 3), two other sets of taps have been directly provided on the choke for the loud speaker connection. These provide two available step-down ratios the use of which permit a low-impedance speaker to present the proper impedance to the tubes for attainment of the greatest possible undistorted power. The actual ratios available are 1:1 , 1:0.73, and 1:0.48. The first of these will be used when the low-frequency impedance of the speaker is 8000 ohms or greater; • february, 1929 . . . page 242 • 3.0 5 oil 1 1 | 1 I 25 100 400 1000 4000 8000 FREQUENCY IN CYCLES Fig. 9 3.0 5 2.0 Type Small Type 25 100 400 1000 4000 8000 FREQUENCY IN CYCLES Fig. 10 6000 25 50 100 400 1000 5000 13.000 FREQUENCY IN CYCLES Fig. 11 the second for speakers of 4000 ohms or less, while the third tap is useful where several speakers are to be operated in parallel. Where the impedance of the speaker is not known, the best connection can be determined by trial by connecting a milliammeter in the plate supply lead and operating with that ratio which allows the greatest volume with a minimum of variation in the plate current. A Further Advantage A FURTHER advantage of the pushpull circuit was noted during this work. This was the possibility of obtaining the plate voltage from the output of the rectifier tubes as shown in Fig. 12 without the use of a filter other than the 4-6-mfd. condenser from the center of the choke to ground. This is of particular advantage in the operation of 250type tubes for the size of the smoothing chokes is reduced to that necessary for handling the small currents of the preceding stages. In addition, the voltage divider can be inserted at the input, the filter as shown allowing the safe use of lower voltage condensers for the remainder of the filter, which is a distinct economy. It was found that with this arrangement the best of dynamic speakers with large baffle could be operated without objectionable hum from the power supply. In the same way the hum from thefilament circuit was observed to be negligible. Changing the plate supply to batteries, temporarily, it was found that there was no increase in hum when the lead to the center tap of the filament resistor was