Radio broadcast .. (1922-30)

394 RADIO BROADCAST FEBRUARY, 1927 FIG. 4 circuit to the filament circuit as indicated in Fig. 5, so as to reach a point which is neither positive nor negative with respect to the two ends of the filament, the plate will never be posi- tive or negative with respect to the filament as a whole, and it will, therefore, not have the positive potentials applied to it as in the preceding case. If we listen, however, with such an arrangement, we will still hear a humming signal in the tele- phone. This signal is of a I2o-cycle frequency, or in general terms, double that of the exciting fre- quency. If we include a battery in the plate circuit so connected as to make the plate nega- tive with respect to the filament, we will find that a potential of several volts is required to stop the hum signal. With the aoi-A type tube, a negative plate voltage of about 9 volts is neces- sary to stop this signal. With the 199 type tube, a negative voltage of about 3 volts on the plate will accomplish the same result. One might ask many questions concerning the cause of this phenomenon. It might be due to a bicyclic thermo electromotive force set up between plate and filament by a bicyclic temperature variation of the latter; it might be a bicyclic contact electromotive force; it might be photo- electromotive force, or a magnet electromotive force emanating from the filament current. Possibly the best explanation is that there is a bicyclic variation in initial emission velocity. We know that when a cathode is heated, it allows a freer swing to the natural vibration of the electrons within it and we know that the higher the temperature of the cathode, the greater the velocity of emergence of the electrons liberated by the heating. If then the temperature of the cathode is varying under the varying heat- producing electric current, the velocity of emer- gence will vary. Consequently the plate elec- trode, with no attractive force of its own for these electrons, will receive a mild bombardment of them which varies (in number of electrons striking it) with their emission velocity. We see, therefore, according to this explanation, that electrons reach the plate through no attractive force of its own and with a bicyclic variation following the temperature variations of the fila- ment itself. This temperature effect, along with the effects of the voltage on the plate due to the latter's connection with one side of the filament, as shown in Fig. 4, occur simultaneously. We should not forget in this connection that the positive voltage of one leg of the filament is attempting to equalize the emission reaching the plate by stealing from the negative leg a portion of the excess electrons liberated at the periods of higher temperature. That is, while the negative leg tends to emit more electrons, due both to its rising tempera- ture and to its rising repulsive negative po- tential, the po- tential of the positive leg is rising also and attract- ing to it an increasing number of freed electrons. Thus, the effect of the positive leg which we shall call "voltage effect," is in direct opposition to the "temperature effect" and tends therefore to stabilize the electron flow to the plate electrode. These effects are shown in Fig. 6. In this dia- gram, curve A represents the exciting voltage applied to the filament of the tube. Curve B indicates the temperature of the filament and shows that the temperature variation is bicyclic with reference to the exciting current. Curve C indicates that the plate current also is bicyclic although the definite relation between tempera- ture and emission is not indicated in this curve. Curve D indicates the voltage variation on the positive leg of the filament insofar as its action as a plate electrode is concerned. The numerals i and 2 indicate that during the first cycle one leg of the filament is the positive leg and that during the other half cycle the second leg acts as the positive plate electrode, so that, irrespec- tive of the fact that a given leg of the filament is alternately positive and negative, one or the other of the legs is positive during all periods except when the exciting voltage passes through the zero point, and therefore one or the other leg is constantly acting as a plate electrode of varia- ble potential. The effect upon an otherwise steady emission to the plate electrode of the vacuum tube is shown in curve E, which indi- cates that the voltage effect of the filament causes a periodic decrease in the electron flow to the plate. When the two effects shown in curve C and E are present simultaneously in the same tube, one tending to increase the emis- sion to the plate and the other tending todecrease it, both of these effects are constantly opposed and a neutralization results which has the effect of stabilizing the plate current, as shown in curve F. While these curves indicate the tendencies toward plate current stabilization, they are not meant to represent exactly the effects found experimentally. Todetermineexactly theprecise amount of hum signal developed in the plate circuit of various types of vacuum tubes under more normal operating conditions, a series of measurements have been made upon different types of tubes under different operating condi- tions. In order to obtain this information, a FIG. vacuum tube with direct current plate potential and with steady grid bias was set up for alter- nating current filament excitation. FURTHER EXPERIMENTS IN FIG. 7, the circuit arrangement employed for making these measurements is shown. It will be noticed that the filament of the vacuum tube is energized from a i lo-volt 6o-cycle light- ing circuit through a step-down transformer and controlling resistance. A voltmeter across the terminals of the filament indicates the voltage impressed thereon. A milliameter in series with the plate circuit indicates the plate current therein, while a telephone in the same circuit serves as an aural indicator of hum signals. An output transformer primary is also connected in this plate circuit and its secondary is con- nected to the terminals of a vacuum tube volt- meter, whose function it is to measure the peak voltage of the alternating currents produced by hum causes within the vacuum tube. There is no input to the grid circuit other than the grid biasing C battery. The grid- and plate-circuit filament returns are made to the central point of the potentiometer shown connected across the filament terminals. The plan of measurement here is to fix the grid and plate voltages at some definite values and then to vary the filament voltage through definite steps and to measure the hum signal as well as the plate current for each such filament voltage. Curves are then drawn with the filament volt- age as abcissae and the hum signals as ordinates for one curve and the plate current as ordinates for another curve. These two curves are plotted together and various sets of this type are ob- tained under varying plate and grid voltage conditions. In Fig. 8 are shown two such curves obtained with an ux-i 12 type tube with a plate voltage of 135 and a grid voltage of 4.5 volts. The filament voltage was varied from approximately two to six volts and the plate current and hum voltage curves were obtained as indicated. We are impressed at once with the unexpected fact that the hum does not increase uniformly with the filament voltage as would a grid impressed signal voltage under the same conditions. There is, strangely, a rather pro- FIG. 8 FIG. 9 FIG. IO FIG. I I