Radio Broadcast (May 1928-Apr 1929)

-2L RADIO BROADCAST tifier circuit causes an undue drop in output voltage, because of the reactive drop in the choke, to compensate which the transformer voltage would have to be greatly increased. The filter action was also impaired. The favorable factor was that this connection did result in a much lower ratio of peak tube current to average load current, reducing the maximum required emission to a low value. (See Table II) Determining Efficiency IN DETERMINING the overall efficiency, the transformer, the tube, and filter losses were included, but the power supplied to the filament was omitted because it remained constant and if included would have affected the readings taken with low load current, 4c and 4e, dispropor \ tionately. Tbe reduced efficiency in j these two cases was caused by the j fact that the internal resistance of the tube rises rather rapidly at small plate current values. Had the transformer voltage been increased to maintain the load current at a higher value, much higher efficiency woidd have been obtained. It should be noted that in Figs. 4a to 4c, the power delivered to the load could not be determined from the average reading of output current and voltage because of the performance of the full-wave rectifier's irregular wave form. Fig. 5, shows the more favorable results obtained with a full-wave rectifier tube in the same circuit combinations covered in Fig. 4. The rectifier tube was a cx-380, operating at 600 volts, or 300 volts per anode. In Fig. 5a, with a resistance load, the load current flows during nearly the entire cycle since current is flowing through one or the other of the two tubes except at the time the transformer voltage passes through zero. The ratio of peak to average current is, therefore, reduced 50 per cent, as compared with the half-wave rectifier, being 150 and 103 mA., respectively, a ratio of 1.4:1. In Fig. 5b, the 4-mfd. condenser is added, the increase in load current being much less marked than with the half -wave circuit, rising 19 mA., to 122 mA. The peak current through the tubes rises to 310 mA., ratio 2.4:1. The next, Fig. 5c, shows the effect of the choke placed in series with the load. In this case it is interesting to note that the choke has quite a different effect from the condenser, since it reduces the peak current carried by the tube while it also keeps current flowing through one anode or the other during the entire cycle. The ratio of peak to load current is very low, 105 mA., to 97 mA., or 1.1:1. Fig. 5d, shows the performance of the fullwave rectifier with the usual filter. The ratio of peak to average current is again high, 290 mA., to 118 mA., or 2.5:1. Fig. 5e shows the performance with the first filter condenser of Fig. 4 removed. The peak current is now only 110 mA., or only 1.15 times larger than the average load current, which is 96 mA. The output voltage was only 45 volts, or 18 per cent, lower than that obtained with the usual filter, Fig. 5d, the readings being, Fig. 5d, 250 volts, Fig. 5e, 205 volts. The various readings including, power output and efficiency, are tabulated in Table III The ripple voltages present at the outputs of the filter arrangements shown in Figs. 4 and 5 have not been measured, but a few tests rndicated that there was not a marked difference, especially if the condenser omitted at the input was added across the output of the filter. The only disadvantage noted in using the arrangement shown in Fig. 5 was the fact that the transformer voltage had to be increased 22 per cent, to obtain a load voltage equal to that obtained with the usual filter. The very greatly reduced peak current demand on the rectifier tubes makes the use of this circuit arrangement highly desirable. Furthermore, the efficiency improves rapidly as the load current is increased, and with equal current outputs it was found that the energy dissipated in the tube was lower and the efficiency slightly higher with the filter 100 125 150 ANODE VOLTAGE (E) Fig. 7 arrangement shown in 5e than with that of 5d. The possibilities of this arrangement were called to our attention by Mr. J. C. Warner of the General Electric Research Laboratory, and these tests have shown clearly the advantages obtained in the full-wave rectifier circuit, Fig. 4, indicating that the arrangement is not suitable for half-wave rectifiers. Effect of Revised Filter Circuit THE remarks made in discussing the lower emission requirements with the revised filter arrangements shown in Fig. 4e, might lead the experimenter to believe that this change would permit much higher load currents to be obtained from the rectifier. However, the amount of energy dissipated in the tube must be considered, as that is one of the important factors limiting the output obtainable from a rectifier. Fig. 6, shows a condition which may occur if an excessive load is placed on the tube, the tube used being a cx-381. Vibrator 1 on this figure reads the full voltage across the tube both in the conducting and in the nonconducting direction. The transformer voltage was 750 volts, and the peak approximately 1050 volts. It will be noted that the peak voltage across the tube reached a much higher value, 1400 volts, due to the fact that the voltage across the first filter condenser is added to the voltage across the tube. In the conducting direction the voltage was quite low, as would be expected, 185 volts. Vibrator V3, shows the current through the tube, which reached a peak value of 450 milliamperes. The load current is not shown on the film, but it was 140 mA., an overload of 37 per cent, on this type tube, a value which resulted in excessive heating of the plate. As a result of the heat developed by the excessive current, the plate reached a temperature at which a slight amount of electron emission occurred which is shown on an exaggerated scale by vibrator Vi. The average value of this reversed current was 4.5 mA., and the peak value 6.8 mA. The fact that during a portion of the cycle this emission was drawn across to the filament at an instantaneous voltage of 1400 volts means that considerable energy was dissipated on the filament and the overheating of the filament became evident in a visible increase in brightness in the center. The circuit arrangement and placing of the vibrators in obtaining this record are shown in the circuit diagram, the double rectifier arranged in series with the tube under test being necessary to separate the two components of the current through the rectifier tube. A small biasing voltage, not shown in the diagram was added to prevent a circulating current between these two rectifier tubes. In taking the voltage across the Fig. 5 april, 1929 page 395 •