Radio broadcast .. (1922-30)

A, C, As The Problems of Filament Heating With Alternating Current — Determining the Cause of and Remedying Undesirable Hum — The Best Tubes to Use By B. F. MIESSNER ADIO receivers and tubes have, since the beginning, been designed and developed for operation on absolutely steady direct current, such as that delivered by batteries. To operate a receiver from machine-generated direct current presents some problems, and to operate it from alternating current presents problems of a much higher order. Because operation from direct-current mains was easier, direct-current receivers of this type were the first to make their appearance in spite of the fact that fully 90 per cent, of electrically wired homes are provided with alternating current, and only between 5 and 10 per cent, with direct current suitable for use with radio receivers, that is, I ic-volt current. With battery operation, it is customary to connect filaments in parallel with a heavy-current A-battery source. When, however, the direct current lighting mains are used it has been found necessary to connect the filaments in series and to add resistance to this series circuit which, when connected across the i lo-volt circuit, would provide a current of proper value for the fila- ments of the tubes. A simplified circuit arrange- J.110 Volts D.C. Input 80V. FIG. I ment of the general type used for this purpose is shown in Fig. i. The next step in electric power receiver de- velopment was the substitution, in a circuit of this type, of an a. c. line with a suitable rectifier for supplying the direct current. In Fig. 2, it will be noted that the vacuum- tube supply circuits are preceded by a filter de- vice, that the vacuum tube filaments are con- nected in series with a resistance across the rec- tified current line, and that the grids are biased by voltage drops in filaments more negative than the filament of the tube whose grid requires a negative bias. The first receivers of this type were designed for 201 -A tubes requiring 250 milliamperes in the filament and an additional plate load of perhaps 25 milliamperes. To pro- vide this rectified power output of 250 milliam- peres at approximately 100 volts, a full-wave rectifier consisting of two Tungar gas type recti- fier tubes was used together with a heavy-duty filter, consisting of very large inductances and very large capacities. With the introduction of 6o-milliampere fila- ment tubes and also 6o-milliampere rectifier tubes of the Kenotron type, by the Radio Cor- poration of America, came the possibility of using the same general receiver scheme with such rectifier and radio tubes. This is shown in Chief Engineer, Garod Corporation Fig. 3. A simplification in the power conversion system was made possible with such a scheme be- cause the output load of the converter was re- duced to approximately 30 per cent, of that required by 2OI-A tubes. Even then, two such rectifier tubes were required to take care of the filament and plate current loads unless a single rectifier tube were to be considerably overloaded. The ideal scheme for eliminating batteries in receivers is one which will use standard tubes of the larger types, without the necessity of develop- ing rectified current specially for the heating of their filaments, and one which will also provide the large power so necessary for the development of high quality and plenty of volume. A receiver which will operate satisfactorily with raw a. c. filament supply will require less than half the rectified current required when rectified current is applied throughout. For ex- ample, if the small dry-cell tubes are used with filaments in series across the rectified current output, their filament supply will be 60 milliam- peres, which is more than the necessary plate supply. If 2OI-A type tubes are used, the filament consumption is 250 milliamperes in excess of the plate current load. If still larger tubes, such as the 112, are used, the rectified power must neces- sarily be still further increased to provide the additional filament heating current. The cost of current-supply devices of this type varies nearly in proportion with the rectified output power for which they are designed; their size, weight, com- plexity and upkeep cost vary in like proportion. It will be understood, therefore, that a receiver so designed as to permit the use of standard tubes of proved design with a. c. current excita- tion of their filaments represents probably the ultimate and ideal type of design for operation from home lighting circuits. The only rectified power required in such a receiver is that used for the plate circuits of the receiver, and the alter- 5V. 5V. 5V. 5V. 5V. Tungar Tubes FIG. 2 nating current required for the filament lighting is obtained from a single winding of a few turns of wire on the power transformer used with the B power rectifier. The A power, therefore, re- quires none of the complex, costly, and bulky elimination apparatus, and the B power require- ments are such that the rectifier and filter appar- atus is small and inexpensive. In the author's receiver plan, the filaments of the amplifier tubes are heated by a. c., and the plate circuits are energized by rectified a. c. The filament of the detector tube is lighted by the B current of all the other tubes which is regulated to the 60 milliamperes required by this filament. By the use of this scheme, therefore, a single 2i6-B rectifier tube provides ample plate power for all of the tubes, including the powerful 210 second audio tube. The filament power for the detector tube, and the C voltage for all the tubes requiring a grid bias, are also supplied from the 216-8 rectifier tube. AN INTERESTING EXPERIMENT LET us now consider a two-element vacuum tube connected as shown in Fig. 4. The filament of the tube is excited by a. c. The plate is connected through a telephone or other indi- cating device to one leg of the filament without any external source of potential included in its path. If we listen at the telephone in this cir- cuit, we will hear a humming noise in which a trained hear can discern a mixture of tone fre- quencies including 60 cycles, 120 cycles, and FIG. 3 some other higher harmonic frequencies. This may appear strange, inasmuch as the plate cir- cuit is not provided with any source of potential for attracting the electrons emitted by the fila- ment. We realize, however, on examination of the diagram, that the plate is connected to the fila- ment at a point of potential variation. The plate itself has at all times the same potential with respect to the rest of the filament as the leg to which it is connected. It is clear that, when the plate and its leg of the filament is positive with respect to the other leg, the plate and this leg both may, by virtue of this positive potential, attract electrons emitted from the negative leg of the filament. In fact, we may conceive of the negative leg as repelling electrons from it while the positive leg is not only holding within it the electrons attempting to escape, by virtue of the releasing effect of the filament tem- perature, but in addition, this positive leg, along with the plate electrode connected to it, is attract- ing some of the electrons liberated from the negative leg. Thus we have impressed upon the plate electrode a 6o-cycle voltage variation by its connection to the filament circuit, and it func- tions as a single-wave rectifier under these con- ditions. If now we change the connection of the plate