US2037352A - Electronic vacuum tube - Google Patents

Description

prl i4, 1935. A, A THOMAS 2,037,352'

I ELECTRONIC VACUUM TUBE I Original Filed Jan. l2, 1929 2 Sheets-Sheet 1 INVENTOR April 14, 1936. 'A A THOMAS 2,37,352

ELECTRONIC yVACUUM TUBE original Filed Jan. 12,v 1929 2 sheds-sheet 2 INVENTOR @www Pegented Apr. v14, 1936 PATENT oFFicl:

ELECTRONIC VACUUM TUBE Adolph A. Thomas, New York, N. Y.,

assignor to Radio Corporation of America, New York, N. Y., a corporation of Delaware Application January 12, 1929, Serial No. 332,000 Renewed June`24, 1935 My invention relates vacuum tubes, usually called radio tubes. in which a controllable electron 'path allows the passage of current between'the plate and filament electrodes. In prior tubes of this type, the only means employed to control the now of electrons from lament to plate has been an interposed grid on, which a variable potential is impressed. 'I'his grid is veither in the form of a metallic mesh or a' plate having slots through which the electrons must pass to reach the plate'.

The number of electrons that go to the platerapendsv not only on the potential of the grid, but on the size and arrangement of the holes inthe grid. If an electron happens to pass farA enough between the grid wires or the edges of a grid opening, it goes to the plate; if an electron is; near enough-to the metal of the grid, -it either stays there or is repelled. The grid interferes physically with'the ow of electrons to the plate, because it must be placed between plate and lament, and this necessarily lowers the efficiency of the tube. Furthermore. the arrangement of the grid causes the tube to have interelectrode capacity which causes trouble -in radio receivers.

It is the object of the present invention to provide an electron tube free from a grid and its disadvantages. In my vacuum tube, the flow of electrons between filament and plate is controlled by a reflector electrode winch throws the electrons against the plate. 'Ihe number of e1ec' trons reaching the plate depends entirely on the potential charge of said electrode, whereby the electron control is-rendered denite. There is no obstruction in the electron path, as in the old grid tubes, and substantially the `full effect of the electronic output of the filament is utilized. A1- so, the inherent capacity in my tube is reduced to -a negligible minimum.

'I'he novel features and practical advantages ofA my invention will be understood-from a description of the accompanying drawings, ln which-- Flg. 1 illustrates a vacuum tube constructed in accordance with my invention:

Fig. 2 is a diagrammatic plan of the elements in. the tube; l

Fig. 3 shows a modiiiedform embodyins an additional reflector plate;

Fig'. 4 is a diagrammatic plan indicating the 'l Claims.l (01.250-275) to the art of electric.

the parts are shown for clearness in grossly exaggerated proportions. The 'tube as made for commercial use is intended to t standard sockets and is of substantially the same size as the grid tubes heretofore used. Referring to Figs. 1 and 2, the vacuum tube I0 is formed with the usual upstanding stem I2, which in this instance rigidly supports 'live metal rods numbered consecutively from I3 to I1. The rod I3 carries alateral arm I8,` preferably of 10 spring metal, to which the upper end of a filament or wire I9 is secured. The lower end of filament I9 is attached to rod I5. The arm I8 is so mounted that filament I9 is constantly held under tension. The arm I8 and filament I9 are 15 v attached-to their respective supports by means of welding or in any other practical way. The rod I5 is connected by a wire 20 toa contact pin 2i, and a'wire 22 connects rod I3 with acontact pin 23. These pins are mounted or embedded in 20 the insulating base 24 of the tube.

The rod I4 carries an electrode 25, which is a curved metal plate preferably (but not necessarily) in the form of a parabolic cylinder. The inner surface of this plate is` smooth and polished. 25 Among metals suitable for this purpose I may mention tungsten, nickel, uranium and osmium. Platinum may also be used. but it is too expensive for commercial purposes. The member` 26 need not be 'of the same metal throughout, 30 for its surface may be an'electroplating on any suitable material. The upper end of rod I4 may be flattened to facilitate attaching the parabolic 'the lament I9 to the shield 26-are directed to 45 ward vthe parabolic electrode 25. The rods I4 and I0 are connected by conductors 21 and- 20 to a contact pin 29 mounted in base 24. In this way, the electrodes 25 and 20 are always at the same potential. In some instances it may not 50 be necessary to connect the shield 20 to contact 'pin 20, but it may be left as ari electrically inert member, since its functionl is to prevent electrons from travelling directly from filament I0 to the anodeplate 8l arrangedat themouthofthc parabolic electrode 25. The anode 9B,- usually of nickel or nickel-plated metal, is mounted on rod II to which it may be attached by weldingfor otherwise. A wire 3| connects the rod I1 with the fourth contact pin 32.

The filament I9, which is a wire of tungsten or other electronic material, is connected in any suitable circuit for heating to the required temperature to cause the emission of a stream of electrons, asin the old type of vacuum tube. The parabolic reflector 25 is connected to one side of the input circuit and takes the place of the grid used in prior tubes. The anode plate 30 is connected to the positive side of the plate battery or other source of constant potential. These connections will be understood without the necessity of adding a circuit diagram. Assuming the tube to be inserted in a socket and properly connected in a radio circuit or other amplifying system, the operation taking place in the tube is something like this:

When the filament I9 is emitting electrons and there is no electric charge on the electrode 25.

the latter simply acts as a reflector of electrons.

That is to say, as the electrons approach the 'inner surfaceof member 25, they are reiiected to the plate 30. In Fig. 2 the electron path from filament I9 to the parabolic cylinder 25 is diagrammatically indicated by radiating dotted lines 33, and thepath of the electrons as they are reiiected by member 25 to the plate 30 is conveniently indicated by parallel dotted lines 3l. Since the filament I9 in the present arrangement constitutes the focal axis of the parabolic cylinder 25, the reiiection'of the electrons from the latter to plate 30 takes place to some extent in substantially parallel paths, in accordance with a well known law of the parabola. It is not necessary that the electrons shall be regularly reflected fromthe curved surface 25, rbecause at any angle of reflection they will strike the anode plate 30. For example, the crossing line 35 may be the path of a reected electron or stream of electrons after striking the surface 25 from filament I9. There are billions of electrons streaming from the filament at high velocity and their paths are too complicated to follow. Nor is it necessary to speculate upon the precise-course that an electron will follow on reflection from the electrode surface 25, for all that matters is the fact that electrons reected toward the positively charged anode plate 30 will reach that plate. A certain proportion of 'the electrons will doubtless be re-' iiected regularly from surface 25, as that would be the shortest pathy to take, -but this particular point is not important. It will therefore be understood that the dotted lines 33 and 34 in Fig. 2 are not intended to represent accurately the paths traversed by electrons, but are used as a simple convenient way to indicate the travel of V.reflected electrons from electrode y25v to-anode 35. The cylindrical shield '20 isv so arranged that noelectrons can strike the plate 30 directly. kElectrons that strike the cylindrical shield 25Nare repelled toward parabolic cylinder 25, from where they are again reected to plate III. 4T'he path of such doubly reflected electrons is indicated bythe dotelectrode 25, there is a constant stream,l ofelectrons between the filament I9 and the anode lplate 30, so that no current impulses are produced in the plate cli-aut.`

aos'aaoa l Let us now suppose there is a positive charge von cylinder 25. lThe negative electron stream If the cylindrical shield 26 is connected to the contact pin 29 and has the same electric charge as cylinder 25, the action referred to also applies to the shield. The number or percentage of electrons retained by the positively charged members 25 and 25 depends on the voltage of the charge and its strength in relation to thenumber of electrons emitted by the hot iilament I9. The higher the positive charge on reflector 25, the greater will be the proportion of electrons that will stick to the surface of this'electrode, -so that a correspondingly smaller proportion of electrons travel to the anode plate 30. When the potential charge of cylinder 25 isnegative, the electrons striking its surface will be instantly repelled and reflected against the plate 30, so that a correspondingly greater proportion of the electron charge reaches the anode plate. It will be seen from this that the intensity or density of the electronic stream between filament I9 and plate 30 depends upon, yand is proportionate to, the.

of my new gridless tube is that the entire electron charge emitted by the filament I9 is under the control of the electrode 25. In ythe old grid tubes, as explained in the first part of this spec-l iiication, only those electrons that happen to pass through the grid openings far enough away from the metal of the grid will succeed in getting to the plate, irrespective of the potential charge on the grid. This operation is entirely eliminated in my new tube where the entire electron stream iiows toward the charged reflector 25, which either reects the electrons to plate 3,0 or retains them. In the old grid tubes, no plate current ows when the grid is highly negative. In the gridless tube above described, the plate current is a maximum when the reflector 25 carries a high negative charge, and this means that the reflector takes no current (or practically none) when the output of the tube is greatest.

Instead of using a straight filament I9 located in the focal axis of the parabolic cylinder 25, I may use a. reflector in the form of a paraboloid in whose focus the electron-emitting cathode is mounted. The member 25 in Fig. 2 may be considered as a cross-section through a paraboloid in the focus I9 of which is an electrode in the form of a point or small button adapted to emit a stream ofl electrons. In a less eiilcient form of my invention, the shield 26 may be omitted, in which case a certain proportion of the electrons will reach the plate l0 directly without be,- ing reflected 'by the member 25. The precise dimensions and relative arrangement of the electrodes I9, 25 and 30 are a matter of experiment y.for each particular design of vacuum tube, de-

pending largely upon the specific purpose for which the tube is intended.

In the modification of Figs. 35, the mounting and arrangement of the filament are the same as in Fig. 1,and I have therefore used the same reference letters to indicate corresponding parts.

In this case the filament I9 is mounted in the focal axis to a. cylindrical reflector 25' mounted between the cathode I9 and anode plate 30' is controlled by a fiat metal plate 36 rigidly mount-1 ed on one or two rods 31 and electrically connected to contact pin 29 by a wire 38.` The plate 36 is arranged at such an angle to the parabolic reflector R' and the anode plate 30 that the` electrons approaching the surface of plate 36 are reflected to the control electrodet'36, and from there they'are reflected to the positively charged anode'. 'Ihe reflector plate 36 is connected to one side of the input circuit, like the grid in the oldtype vgrid tubes. To prevent any electrons reaching the anode 30' without being under the control of Athe charged reflector plate 36', 1

interpose a shield 39 which may be a cylindrical strip formed integral with (or otherwise attached to) the parabolic reflector 25'. The shield 3,9 is so dimensioned and positioned that it covers the angle 40 in the diagram of Fig. 4 and therefore .blocks the movement of electrons directly to the vanode plate 30'. I do not say that the shield 39 is necessary in all cases, but I believe that it adds to the efilciency of the tube.

The operationof the tube in Figs. 3 andi is the same as that of the tube in Figs. 1 and 2, I need add but little to what has already been said. The only difference in the operation of the two tubes is that in Figs. 3and 4 the electronv stream reaches the charged plate `36 either directly from filament I9, as diagrammatically indicated by ,dotted line 4I, or by reflection from member 25', Vas indicated by dotted lines 42. If

` the reflecting plate 36 has a positive charge,.it

will attractthe electrons and hold them, or at least hold the greater proportionvof them, dei.

pending on the value of the impressed voltage. When the charge on plate 36 is negative,` the electrons are repelled and reflected to the plate 30'. In other words, the operation of plate 36 with respect to the anode 30' is the same as the operation of the lreflector 26 in Fig. l'with respect to anode 30. 'The number of electrons retained by the reflecting plate 36 depends on the character and intensity of the charge impressed on the plate. The reflectors 25' and 36 may be of the same material' as the parabolic electrode 25-of Fig; 1. The parabolic reflector 25' is electrically inert and its function isto reflect electrons. The construction of Fig. 1 is preferable on account of its greater simplicity, because the member 25 combines the functions o'f the members2 and 36 in Fig. 3. What has been said about the electron paths in connection with Fig. 2 applies also to Fig. V4.

Among the practical advantages tube overthe old-fashioned grid tube may be mentioned a greater amplification power, more positive control of the electron stream, and a minimum inherent capacity of the tube. In the old-fashioned grid tube, the control of the elecg trons depends upon the action of the interposed grid asa physical obstruction in thel electron of this new i path. That is to say-the control of the electrons depends entirely on how many succeed in passing through the openings or windows ofthe grid. In the present case, the charged electrode 25 (or 36) does not constitute a physical obstruction in the electronic: path between the filament I9 and the anode 30 (or 36'), but it is a solid metal plate which attracts and holds a definite proportion of the electron emission in accordance withv the potential charge existing on its surface. VThe electron control is therefore definite and positive. As for the' inherent capacity of the tube, that is reduced to a minimum on account of the relative arrangement ci the electrodes 25` and 30 (Fig. 1) or electrodes 36 and 30 (Fig. 3) In the old type grid tube,l the inherent capacity is relatively large on account of the close parallel mounting of grid and plate.

Also, it should be noticed that because of the inherently low capacity eiect in my new tube the tendency of self-oscillation is brought down to a minimum, and may indeed be wholly eliminated in a proper design of tube.

Although I have shown and described certain t specific constructions, I want it understoodthat my invention is not limited to the details set forth. I have no doubt that changes and modi' cations are possible without departing from the scope of the invention as dened in the Iappendedl claims. I claim as my invention:

1. A vacuum tube having an electrode substantially in the form of a parabolic cylinder for receiving and reflecting electrons, an electronemitting filamentv located approximately in the focal axis'of said parabolic cylinder, said filament being constructed to discharge 'electrons against said parabolic cylinder, wherebyy electrons striking said cylinder are reected therefrom, said cylinder being adapted to receive variable electric charges for controlling the electrons that strike saidv cylinder, and a plate electrode mounted closely to said parabolic cylinder for receiving substantially lthe entire electron stream reflected from said cylinder.

-2. A vacuum tube having an electrode in the Aform of a curvedplate provided with a smooth solid inner surface capable of reflecting elec-v trons, an anode inthe form of aplate mounted with its ends closely adjacent the endsl of said curved plate, whereby said plates form a practically enclosed space, and a cathode mounted in said space for emitting electrons to said curved plate which is mounted directly in the path of the electron discharge for reflecting the electronsto said anode in accordancev with the potential impressed on said curved plate.

3. A vacuum tube having an electrode in the form of a curved plate provided with a smooth solid inner surface capableoi.' reecting elec trons, an anode in the form of a plate mounted withl its ends closelyadjacent the ends of said curved plate,` whereby said plates form a practically enclosed space, a cathode mounted in said space, and means in said space for preventing electrons reaching said anode without reection by said curved plate.

4. A vacuum tube having-a. control electrode in the form of a substantially cylindrical solid 'plate adapted to be subjected to variable potential, an electron-.emitting 4filament supported approximately in the axis of said cylindrical electrode which thereby is mounted directly in the path of electrons from said filament for reflecting electrons in accordance with the potential impressedon the control electrode, and an anode 5. A vacuum tube having an' electron-reilecting control electrode in the form ofa substantially cylindrical solid plate, an electron-emitting filament supported approximately in the axis of `said cylindrical electrode, an anode plate mounted closely adjacent said cylindrical elec-- trode for receiving substantially all electrons reiiected by the cylindrical electrode, and means for preventing electrons reaching the anode without reiiection by said control electrode.-

6. A vacuum tube having a iirst electrode for receiving and reiiecting electrons, said electrode being adapted to receive variable electric charges, a second electrode mounted closely adjacent said irst electrode and adapted to operate as an anode, means between said electrodes for emitting electrons to strike the ilrst electrode, and means positioning the lilrst electrode to reflect electrons the opposite terminals to the second electrode in accordance with the electric charge on said ilrst electrode, said anode being mounted to receive practically all electrons reflected by said first electrode.

'1. A vacuum tube having a substantially para bolic electrode for receiving and reflecting electrons, an electron-emitting cathode located approximately in the focus of said parabolic electrode, said cathode being mounted to discharge electrons against said parabolic electrode, whereby electrons striking said parabolic electrode are reflected therefrom, said parabolic electrode being adapted to receive variable electric charges for controlling the electrons that strike said electrode, an anode mounted closely to said parabolic electrode for receiving electrons by reection from said parabolic electrode, and. only four contacts projecting from said tube, one of said contacts being connected to said parabolic electrode, another contact being connected to said anode, and the other two contacts being connected to of said cathode. ADOLPH A. THOMAS.

Images