January 15, 1920] 



NATURE 



501 



TRIODE VALVES AS ELECTRIC 

 AMPLIFIERS. 



AMONG the most exquisite tools that modern 

 wireless telegraphy now proffers to investi- 

 gators working in the fields of pure science, that 

 known as the amplifier stands out as being of 

 the most obvious promise in various directions. 

 The amplifier offers the means of magnifying vary- 

 ing electro-motive forces and currents, otherwise 

 imperceptible, so that they come within the range 

 of ordinary laboratory measuring and record- 

 ing instruments. It was developed during the 

 war to a high pitch of excellence, not only for 

 the improvement of wireless telegraph signals, 

 but also for other kinds of signalling and for 

 listening under water and under the ground — that 

 is to say, it has been fully developed for the mag- 

 nification of the high-frequency currents used in 

 wireless telegraphy and for currents of telephonic 

 frequency. Descriptions of the apparatus have 

 now been published in many places, and the tool 

 as thus developed will in due course take its place 

 in the laboratory. For many purposes, however, 

 an amplifier that will faithfully magnify slow varia- 

 tions of a current or electro-motive force is de- 

 manded, and since little has been published about 

 such apparatus, the following notes of methods 

 used in the writer's laboratory during the past few- 

 years are now presented. 



As the term is usually under- 

 stood nowadays, an amplifier 

 consists of one or more of the 

 three-electrode thermionic 

 vacuum valves of wireless tele- 

 graphy associated with auxiliary 

 transformers or analogous ap- 

 paratus. This particular kind 

 of valve may for brevity 

 be called a triode valve, or 

 even a triode. It comprises a hot filament for 

 supplying electrons, which serves as cathode, a 

 plate, or cylinder, which serves as anode, and an 

 intervening grid, all contained in a highly evacu- 

 ated bulb. The bulbs generally used in amplifiers 

 are about the same size as the common incan- 

 descent filament lamp, but the filament cathode of 

 the triode is proportioned so as to become white 

 hot when a battery of about 5 volts is joined to its 

 terminals to supply about three-quarters of an 

 ampere of current. A battery of, say, 50 volts, 

 connected with its positive pole to the anode and 

 its negative pole to the cathode, causes a current 

 of order one milliampere to flow when the grid is 

 at the same electric potential as the mid-point of 

 the filament, and of perhaps twice this value when 

 the grid is given a potential one volt higher. 



The- reason for this influence of the grid may 

 be briefly explained as follows : When, in obedi- 

 ence to the electro-motive force applied between 

 anode and filament, an electron current flows 

 from the filament, the distributed electric charge 

 in the space creates an electric field that tends to 

 repel electrons back to the filament, or, in other 

 words, gives rise to a back electro-motive force 



NO. 2620, VOL. 104] 



acting against the battery. But making the grid 

 positive relative to the filament partially neutral- 

 ises the field of the space charge, and therefore 

 reduces the back electro-motive force. This influ- 

 ence is greater the closer the mesh of the grid ; 

 in some commercial triodes one volt on the grid 

 will cancel ten volts of the back electro-motive 

 force, or, in other words, one volt applied to the 

 grid is worth ten volts applied in the anode circuit. 

 At the same time, the current flowing on to the 

 grid when one volt is applied between grid and 

 filament is, perhaps, only a microampere ; the 

 multiplication of current performed by the triode 

 is thus a thousandfold. Moreover, since the 

 energy input to the grid is, in the assumed 

 circumstances, i x 10-* watt, and the consequent 

 additional energy output of the high-voltage bat- 

 tery in the anode circuit 50 x iq-^ watt, the 

 energy ratio is 50,000. Not all this output is 

 available for use, however; we may, in fact, 

 scarcely hope to use half of it. 



It is worth while emphasising here a difference 

 between an electro-magnetic transformer and a 

 triode regarded as a transformer. The trans- 

 former may be arranged to give in its secondary 

 circuit a voltage many times that applied to the 

 primary, but the current is correspondingly dim- 

 inished to keep the output of the energy equal to 

 the input (losses being neglected). But in the case 



Fig. 



of the triode valve the current, as well as the 

 electro-motive force, is multiplied, and the con- 

 sequently multiplied energy output takes place at 

 the expense of the high-voltage battery. 



The most highly developed type of amplifier is 

 that intended for the magnification of currents 

 alternating more than 100 times per second, and 

 consists of a number of triode valves linked in 

 tandem by means of the mutual inductance of 

 transformers. The earliest instruments were prob- 

 ably constructed by de Forest. Excellent instru- 

 ments can now be made for any frequency between 

 100 and 1^000,000. It is stated that Mr. H. J. 

 Round, of the Marconi Co., has used up to tvventy- 

 two triodes in tandem, and obtained magnifica- 

 tions of potential difference of about half-a-million- 

 fold. As already stated, amplifiers for rapidly 

 alternating current have been described elsewhere, 

 and are not the subject of this article. 



The type of amplifier described in Fig. i may 

 be used for magnifying currents that vary slowly. 

 It appears to have been conceived first in the 

 French Military Radio-telegraphic Laboratory in 

 Paris. In this apparatus the linkage between suc- 

 cessive triode valves is accomplished by means 



