US2254422A - Electron multiplier - Google Patents

Description

Sept. 2, 1941. D. GABOR 2,254,422

ELECTRON MULTIPLIER Filed May 26, 193'? 3 Sheets-Sheet 2 j INVFNTOR 06777715 6427502" I BY W. Dam mqs'pmz,

ATTORNEYS Patented Sept. 2, 1941 ELECTRON MULTIPIJER Dennis Gabor, Rugby, England Application May 26, 1937, Serial No. 144,787

- In Great Britain May 27, 1936 13 Claims.

This invention relates to electric discharge devices, known under the name of electron multipliers, in which a primary electron stream is amplified by secondary electron emission. The primary electron stream, which may be produced by photoelectric or by thermionic emission, impinges on an electrode to be called a secondary cathode. The secondary electron stream emitted by the secondary cathode impinges in turn on 'a second secondary cathode, maintained at a higher potential than the first secondary cathode, and this process is repeated until the secondary electrons emitted by the final secondary cathode impinge on a collecting electrode or anode. If each secondary cathode emits more than one electron for every impingingelectron, and the number of steps or stages is chosen sufficiently high, the current collected by the anode or output electrode may be a very high multiple of the primary current.

.-One object of the present invention is to provide an improved arrangement of electrodes suitable for a high degree of amplification of photoelectric or thermionic currents by means of electronmultiplication. An important feature of this electrode arrangement is that the secondary cathodes are surfaces of rotation, arrayed along an axis of symmetry. A further feature is that one or more axially symmetrical electrodes, to be called auxiliary anodes, are provided inside the secondary cathodes, in order to produce a substantially radial electric field in addition to the substantially axial electric fields produced by the successively rising potentials of the secondary cathodes. A further feature is, that means are provided; to produce a substantially axial magnetic field in order to prevent the electrons from hitting the said auxiliary anode or anodes.

Another object of the invention is to provide improved means for the amplification of thermionic currents by utilizing electron multiplication. 7 A novel method for controlling the output current is vprovided by means of modifying the gain of said electron multiplication.

The novel features which I desire to protect herein will be pointed out in the appended claims. The invention may be best understood by reference to the following description and the accompanying drawings. In the drawings Figs. 1, 2 and 3 are diagrammatic representations, explaining the principle of the invention. Figs. 4 and 5 show alternative modifications of va device according to the invention. Figs. 6 and 7 are illustrations of certain improved details in the electrode arrangement. Fig. 8 is a suitable arrangement of magnetic coils which may be used in connection with the device according to the invention. Figs. 9 to 19 are diagrammatic representations to explain dimensioning and additional aspects of the device according to the invention. Figs. 20 to 23 relate to a novel method for modifying the gain of the device. Figs. 24 and 25 show methods of controlling the intensity of a thermionically produced. primary beam. Fig. 26 illustrates the circuits of an electron multiplier according to the invention, whereby it may act as a complete superheterodyne set for radio or television reception.

The principle of the new electron multiplier is shown in Fig. 1, which is a diagrammatic longitudinal section of a schematically simplified device. This is understood to be enclosed in an evacuated envelope 3. I and 2 are two cylindrical bodies which have potentials linearly increasing with the distance 2: measured from the top edge of I. In every cross section the potential of 2 shall be greater than of I, and this potential difierence may be called V. V can be independent of z or vary to a certain extent which shall be specified later. In the space between I and 2 the'electric field is therefore composedof a radial field with the intensity er, the intensity of this field varying inversely with the distance r from the axis, and of a longitudinal field of the intensity Ez. To this composite electric field is added according to the invention a longitudinal magnetic field of the intensity H. I

The electrodes in these and subsequent figures are supported within the evacuated glass envelope 3 by suitable means which are not shown in the drawings in the interest of clearness of illustration.

The electric field as described can be realized by making both I and 2 of materials with high resistance, through which currents may be passed in an'upwa-rd direction. Constructions which are more convenient for practical purposes shall be described later. The magnetic field may be produced for example by means of a long coil 4 placed coaxially around the vessel, or by a tubular permanent magnet, made of a suitable material such as cobalt steel.

If by photoelectric'action or by other effects slow electrons are released from the surface of I, they will start moving towards the central cylinder 2. They will be, however, deflected by the magnetic field and forced to return to I.

This is illustrated in Fig. 2, which shows the projection of the electron paths on a plane page pendicular to the axis. In this figure, I is the outer electrode, to be called the cathode, and 2 the inner one, which shall be called auxiliary anode. Fig. 3 shows the paths of the electrons in circular projection on a plane passing through the axis. In these two figures it has been assumed for simplicity that the field components er, Ga, and H are independent of 2. It will be seen that the electrons return to the cathode afterhaving travelled a vertical distance Z, which shall be called the step length. branches of the orbit in Fig. 2 are symmetrical,

The two the electron takes equal times for approaching the inner electrode to a minimum distance and for returning to the cathode surface. As however each electron travels vertically with a uniformly accelerated motion, it will travel one quar ter of the step length in therfirst halfof the time and three-quarters in the secondhalfa .For I this reason the two branches of the orbit as they appear in the longitudinal projectionin Fig. 3' will be strongly, asymmetrical. Assuming that the electron has started with zero velocity, it will-return tangentially to the cathode. If the voltage drop along the step length'is sufficiently high, it will release secondary electrons. These will start again with very low velocities and describe paths of the same shape, until the last electron leaves the space between the cylinders.

If the number of secondary electrons released by one electron is greater than unity, the device separated from one another byfslits Y. The

cathode surfaces preferably comprise materials suitable for secondary electron emission such as The figure shows a longitudinal caesium. Coaxially arranged with the secondary I cathodes, there are provided a number of cylinders,;numbered ll,.l2, l3, Hand is, telescopi cally assembled, The last of these, [5, is connected with the collecting or output anode, the flare I6, which collects the electrons releasedat thelast cathode l0. 7 j V J V 1 The secondary cathodes and auxiliary anodes all have connections with the outside and may be maintained at difierent potentials with respect to one another. In the arrangement illustrated each anode section is connected with the following cathode section, I] with 1, 12 with 8,

l3 with 9 and I4 with In. Co'nsequentlyif. equal potential difierences E are impressed between successive cathodes, there will be a constant-potential difierence between each cathode and the corresponding auxiliary anode. ;It should be understood however that the stages need not. be equal in length or voltage, nor need the radial potential difierences be equal to the longitudinal n t ntiel li eren,. e -v '-.d emete i thwe Q e' c l nders-m ei ryrm a i st ge and flier. ma b .ea on cal nsiead yli r drical. For making; the illustrations simpler {I hav howev as umedqua and i m ic st s. men th eceom enyin d aw I Eisv, sh9ws..-aa. n t ye ccnst e i Jl'h aw lierranodeis her f a z r er.w ch-meme m de o a were: i high spe ic mine e r i may be.e-ja se eie ;insula in t a we r w hs ch am ieria ai iisisfii fid w h n' it; 1815'ii ansil qw iie divi ed? ance into sections. These are connected by means of thin wires 2|, 22, 23 and 24 with the cathode sections 26, 21, 28 and 29. The last ring is connected with the anode flare 25.

In the foregoing constructions the electric field at the ends of the electrode array is necessarily different from that at the central portions. This can be avoided as shown in Fig. 6 by giving the first cathode 30 and the anode flare 3| shapes identical with those of thepotential lines which would exist if the electrode column were infinitely long. Their shape can also be simplified as shown in Fig. 7, where the end plates 32 and 33 are composed of frusto-conical, cylindrical and plane sections.

In some cases it is advantageous to have a coil which produces the magnetic field constructed in such a way that it produces a homogeneous field over the whole volume of the electron multiplier. This has the advantage that in a homogeneous field the relative position of the coil and the device need not be adjusted very accurately, Fig. 8 shows a convenient arrangement. It consists of three single coils 34, 35 and 36 with dimensions which are small as compared with their diameter D, and spaced at the distance D in the axial direction. If. the number of ampere turns of 34 and 36 are equal, and the number of ampere turns of the central coil 35 is 11.6 percent of that of 34 or 36,'the field strength along the axis will vary by less than i 1 percent.

The most advantageous dimensions of the device can be obtained by considering the relation which links up its various factors. This equation can be written in the form:

Here Z is the step length, i. e. the distance in cms. which an electron starting with zero velocity travels in the direction of the axis before returning to the outer cylinder. This must be approximately equal to the stage length X+Y, if the device is to work as a multiplier. H is the magnetic field in gausses. E is the potential difference between two successive cathodes, in volts. R is the radius of the cathodes, in cm.

7 V is the potential difference between the cathode and the opposite auxiliary anode. R1 is the radius ofthe'auxiliary anode. The nature of the function f is represented in Fig. 9. It will be seen that the function f has a maximum. Thismeans, that the step length Z has a maximum for acertain value of V-in an otherwise determined arrangement. Choosing the operating point at or near this maximum affords the advantage, that small variations of V or of R1 are'of no importance. This is'shown more particularly in the following numerical example:

According to Fig. 9 the maximum occurs approximately at We can choose freely 4 quantities, e. g. R=1.5 cms.,' R1'='0.5 cm., .Z 1 cm. and E=' volts. This gives H 22.5"'gausses and 1V=53.5" volts. Because of the maximum this voltage need not bebbservedvery' exactly. Z'will be shortened only by? 1% or 0.1mm. if 'V is 4501 62 volts, or Rile'q'ualj 1:00.42 or 9.6 cm. It will. be seen that this mode I of operation is' particularly advan ta eous for devices ofth kind as shown. in Fig. 4, iasj no correctiohs'j heed to, be introduced for the varying diameter or the auxiliary anodes.

Undr som conditions itLmight be however advantageom ochoose the working conditions so as to remain at the left of the maximum. Near to the origin the function f approaches a straight line, and the equationcan be written approxi- This means that for weak magnetic or very strong electric fields the step length Z depends only on the ratio 'of the radial and longitudinal potential steps. This can be understood from Fig. 10, in which the electron path a corresponds to a weak magnetic field. The length of this path is very nearly equal to the diameter on which the electron would move in the case of a zero magnetic field. This mode of operation has the great advantage, that the step length is determined entirely by the electrical connections. The magnetic field has only the function of hindering the electrons from hitting the auxiliary anode. This mode of operation is particularly advantageous in the case of thin auxiliary anodes like in Fig. 5.

In the same drawing is also shown a path b which corresponds to the maximal step length. The electron approaches the axis to a minimal distance equal to about 60% of the outer radius. The auxiliary anode can have therefore a rather large diameter, without incurring the risk of the electrons hitting it.

As compared with other electron multipliers, axially symmetrical devices according to the invention have the advantage that lateral scattering of the electron paths has no consequence and no cautions need be taken to avoid it. Scattering in the axial direction might however cause inconveniences as it might prevent a part of the electrons from reaching the last cathode. Such an effect is-produced by the inhomogeneity of the longitudinal electric field. Instead of being constant as assumed above, the field is zero along the surfaces of the cathodes and very strong in the slits between them. This produces a focussing effect as shown in Fig. 11. An electron which starts from the lower edge of a cathode 31 comes at the beginning of its path P1 under the influence of a strong field. We assume that it reaches the corresonding edge of the next cathode 38. An electron however which starts from the top edge of 38 will move at the beginning of its path P2 in a weak field and might land at the same spot as P1. Electrons returning to the same cathode are of course wasted. This can be prevented according to the invention by an inhomogeneous magnetic field, which can be produced, as shown in Fig. 12, by putting wedge shaped rings 34, 40 of high magnetic permeability behind the oathodes il, 42. The cathodes themselves can be also wedge shaped and made of a material with considerable permeability like nickel or nickel-iron alloys. The path Pi which starts from the edge of 4| crosses therefore a stronger magnetic field than the path P'z. By choosing a suitable shape and size for the rings 39, 40 it is therefore pos- -to become suddenly homogeneous.

of small diameter in order to avoid obstruction of light.

Tubes of large surfaces areconvenient for certain applications of photo-tubes, especially ifthe light can not well be concentrated by optical means. Figs. 14 and 15 show a construction of the electron multiplier according to the invention, to be used as a self-amplifying photo-tube. The first cathode 46 is longer than the following secondary cathode 41. In order to prevent distortion of the electric field, the photo-cathode 36 is completed to a cylinder by a grid or gauze s8, which covers the window through which the light falls in, but does not cut off too much of the light., The first auxiliary anode 49 is preferably a thin wire, which may be fastened by some insulating member to the top plate 50.

In this case however special measures are necessary in order to prevent the electrons returning to the photo-cathode. This can be done according to the invention by an inhomogeneous magnetic field, the effect of which is shownin Figs. 16 and 1'7. Fig. 1'7 shows in a plane perpendicular to the axis the effect of a magnetic field which increases or decreases along the axis z. An increasing field will produce a stronger curvature in the returning branch of the curve, so that the electron will turn back before reaching the cathode and will go on moving one kind of epicycloidal path. A decreasing field on the contrary will not turn the electron back by a full so that the electron aproaching the cathode under a grazing angle will not be able to reach it and will go on moving onv a kind of hypocycloidal path. The practical result is however in both cases the same. Both increasing and decreasing fields will prevent the electrons from returning to the cathode. Fig. 16 which shows the path in circular projection on a meridian plane corresponds therefore to both cases. Such inhomogeneous fields can be produced easily either by placing or dimensioning the magnetic coils conveniently or by backing the first cathode with wedge shaped rings of ferromagnetic material.

This measure would have, however, the consequence that the electrons which have departed to a considerable distance from the first cathode would not come back to the second cathode, i. e. the first secondary cathode, even if the field were They can be brought back, however, according to the invention, by a decreasing radial electric field, i. e. by making V to decrease with increasing 2. The effect of a decreasing V is that the electron will not have lost all its radial velocity when it reaches again the radius R, and that therefore the path has a tendency to move outwards. This is shown in Figs. 19 and 18-,where also the virtual prolongation of the path is represented by dotted lines.

The decreasing radial field can be produced by choosing the lengths of the anode sections and their connections with the cathode sections'conveniently. The simplest way however is to employ only a single auxiliary anode, such as a solid rod, tube or wire, extending along the whole length of the device. This auxiliary anode may.

be connected with the last or collecting anode,

or with a suitable, potential higher than the 7 advantage that the electrons do not strike the cathodes any more under grazing angles,;"and

their paths are consequently. less strongly modispace charges at-higher current densities. M r .f If the device is to be used as a simple phototube, it is important to make its response independent from the pointof incidence of the light beam, by the means as. explainedin connection with Fig. 14. The newdevice is, however, particularly suitable also for other applications, in

which the output current is deliberately made,

will reach the last circuit only if 20 is larger than a certain minimum, otherwise'they will return to the last cathode Stinsteadof reaching the anode 52. We call the first current, which can be photo-electric, I0, and the current between the last cathode and the anode, I. Fig. 21 showsthe amplification factor or gain I/Io as a function of the starting point of the electrons, 'It is zero below a certain distance and above it jumps suddenly to a finite value, This will be the case, of course, only if the electrons start from one point; in the case of photo-electrons, for example, if the light spot from which they originate is infinitely small. With a spot of finite width the slope be comes finite, as'shown in Fig. 22, the length of the slope S being equal to the spot width. In this range the amplification factor is a linear function of the spot position, if the light density is homogeneous. This phenomenon can" therefore be used, for example, in conjunction with mirror galvanometers, with instruments for precision measurements of small movements, for sound registration and reproduction, etc- The electron emission from the first cathode ring can also be produced by a beam of electrons instead of by a light beam. Thenew electron multiplier is particularlysuitable as a thermionic amplifier because of itsaxial symmetry. In Fig. 23, 53 is an indirectly heatedthermionic cathode with a heating filament 54. 55 and 56 are two cylindrical tubes, forming between themselves an annular slit 51. The potential of 55 and 56 is higher than of the first cathode 58 and the potential of 53'is lower than that of 58 by about an.

equal amount. The beams are therefore ejected from the slit with a rather high velocity and somewhat retarded in the outer space. It can be shown that the magnetic field willnot prevent these primary electrons from reaching the cathode ,58, whereas it will be strong enough for keeping the secondary electrons away from the inner system. H a I Two plane rings 59 and: 60 fittedoutsidethe slit are serving asdeflecting plates. I By impress.- ing a variable potential between them, the beam can be deflected in theagial directionbBy adjusting the dimensions conveniently, deflections of the order of 0.17-1.0 mm. /volt canbeobtained. This can result in variations of the final current of the order of several milliamperesper voltgwhich is of the order of thejfslopes 'ofthe .DIdi-i nary thermionic valves. The amplificationffactor fmu is extremelyhigh, as thegvoltage of .the output anode has hardlyany effect on thecurrent.

This kind of control shall be called deflection .contro ,for differentiating it from intensity control, as shown in Fig. 24. In this second kind of control the intensityof the beam is varied, without changing its position. In Fig. 24, 6| is a thermionic cathode. This is fitted with two grids, 62 and 63, so as to form a small thermionic valve with a perforated anode. 62 is an accelerating grid, 63 is the control grid. This is fitted with side. rings 64 in order to prevent the electrons from flying to the next auxiliary anode section 65.

The electrons fly through the meshes of the "anode grid 62'to the first secondary cathode 66,

from there to 61 and so on. It is a particular advantage of this arrangement that the field between, the cathode 66 and the anode grid 62 is completely separated electrically from the small thermionictube which is serving as electron tained, withoutany perceptible anode reaction,

as the currents flowing in the first stage are extremely small. It is advisable to use self-focussing arrangements; like the one shown in Fig, 12, in order to utilize the whole surface of the secondary cathodes, especially in the later stages, where the currents are considerable.

Intensity control and deflection control can be also combined as shown in Fig. 25. Here 68 is the thermionic cathode with the filament 69; 16 the controlling electrode, which is a small cylinder with a slit; H the accelerating electrode; 12 and 13 are the two deflecting plates. As the position of the beam can be controlled by either of the .defiecting plates, this device contains three possibilities for applying controlling voltages. This arrangement has also the advantage that the controlling and the accelerating electrode can be adjusted in such a way as to concentrate the beam at the secondary cathode 14 ina narrow angular zone.

A thirdmanner of control is control by the longitudinal magnetic field. This can be a very sensitive control if the initial beam is narrow and no self-focussing measures are applied. The sensitivity depends also on the position of the working point on the curve in Fig. 9. Near its maximum Z and H are simply reciprocal, i. e. 1% increase of H causes 1% decrease in step length. If there are e. g. 10 stages, this means a shifting of the position of the last spot of incidence by 10% of the stage length. We see therefore that it .is possible to produce very considerable current variationswith magnetic fields of less than 1 gauss. In order to make highfrequency. control. possible, the cathode'rings must be split.

,A further possibility of control according to the invention is electrostatic control after the last stage. This will. be explained in connection withan example, in which I intend to show, that the new electron multiplier can perform, according to the invention, all'the functions which are performed by valves in superheterodyne receiver sets,

it p ifi requency conversi r c ification (demodulation), and automatic volume control.

Fig. 26 shows the diagram of an electron multiplier with circuit connections and additional circuit elements, in which the multiplier performs all the functions of the several tubes in a superheterodyne receiver, for sound. broadcasting or television reception, without any other tubes.

The signal is applied to the control grid '15 by means of the coupling resistance T6; This point is marked by H. F. which means the high frequency of the carrier wave. The aerial circuit is not shown in the figure. The construction of the thermionic system for the emission of the primary electrons is the same as in Fig. 25 the previous figure. It consists of a thermionic cathode 11, a control electrode 15, an accelerating electrode 18 and deflecting plates 19 and 86. One deflecting plate, 79, is used for heterodyning (frequency conversion) the other, 88, for automatic volume control. The plate 19 is energized by'a local oscillator 96, labeled L. 0., by means of regenerative oscillations with a frequency To. These are produced by picking up the oscillations at one of the intermediary secondary cathodes, 8! in the figure, and feeding them back to the plate 19 across the tuned reactor circuit marked L. O.

The action of the heterodyning plate can be understood from Fig. 22. The voltage of the deflecting plate controls the factor by which the electron current is amplified and produces with the original frequency in the two beat frequencies fh+fo and fir-fotermediate frequency fi. It may be conveniently about 10% of the original frequency fn. This in-- termediate frequency is filtered out and amplified by selective regeneration. It is picked up at one of the intermediate secondary cathodes, 82, and fed back to the control grid through the reactor circuit 91, marked I. R, which is tuned to the intermediate frequency f1. This frequency is now again amplified and also heterodyned, so that the signal received at the final stage is a modulated wave of three carrier frequencies fh, fhfo and fh2fo. This has however no detrimental effect on the selectivity of the set, as all three oscillations are derived from and proportional to the oscillation with the intermediate frequency which has been filtered through the circuit I. F. The regeneration of the intermediate frequency ensures moreover a high degree of modulation throughout the device, and therefore makes it possible to utilize the current capacity of the device to its limit. A further advantage is, that the conversion conductance can be kept rather low and harmonic distortion can be avoided. g

The final stage of the device consists of the last secondary cathode, 83, the anode 8 and three grids 85, 86 and 87. The first grid is placed about one step length below the middle of the last cathode. The electrons have here velocities nearly in axial direction. This has the advantage that as they move through the grids 85-31 they will be proceeding in the direction of the magnetic field, which therefore will not interfere with their movement.

This final system can be used in various ways. It could be used,,e. g., for frequency conversion, if an additional valve were to be used for the loudspeaker or other output device. In the present example I intend, however, to show that the new electron multiplier is a self -contained device The latter may be called in and explain-the use-of the final stage as a rectifier, (current detector or demodulator).

The grid 85, which has the same potential as the guard ring 88, i. e. the potential which the next secondary cathode would have if the system were continued, is acting as an accelerating (space charge) grid. Its potential is derived from a potentiometer 89.. The control grid 86 has a potential only little higher than the last cathode. For simplicitys sake the grid bias is shown in the figure as being derived from the same potentiometer 89. Between anode and control grid is placed the anode screen grid 81.

If the control grid is conveniently biased, its action will be to cut off the anode current beyond a certain current value and make it flow back to the space charge grid 85. The rectified current can be derived either from the anode or from 85. In the figure the loudspeaker 9%) is energized from the anode across a transformer 9|. In the case of television the modulating electrode of the cathode ray tube is to be connected with 34 or through a suitable bias.

Automatic volume control (A. V. C.) is effected in the following way. The reaction with the filtered intermediate frequency is applied not only to the control grid, but also, through the condenser 98, to the second deflecting plate 80 opposite to the plate 19 which is used for heterodyning. The plate 80 is, however, placed so near the electron beam, that at positive potentials it would attract a sufficient number of electrons for neutralizing its charge. It will assume therefore a negative charge so as just to repel nearly all electrons, i. e. it will assume a charge proportional to the high frequency amplitude and produce a corresponding deflection of the beam in the upward direction. The greater, therefore, the intensity of the high frequency signal, the nearer the top of secondary cathode will be the starting point of the electrons, and this will result, according to Fig. 22, in a smaller ampli fication factor.

Fig. 26 shows also the system of potential dividers, formed by resistances 92, 93, 94 by which the suitable potentials are applied to the secondary cathodes.

While I have shown particular embodiments of my invention, it will be understood that many modifications and applications may be made by those skilled in the art without departing from the invention as set forth in this specification and in the appended claims.

What I claim as new and desire to secure by Letters Patent of the United States is:

1. An electron multiplier comprising a plurality of rotationally symmetrical cathodes having inner surfaces capable of emitting secondary electrons, said cathodes being coaxial and axially spaced from one another forming a column, an anode of rotational symmetry inside the column of said cathodes and coaxial with them, an output anode at the end of said column, and means independent of said plurality of cathodes for producing a magnetic field of substantially axial direction.

2. An electron multiplier comprising a plurality of rotationally symmetrical cathodes hav ing inner surfaces capable of emitting secondary electrons, said cathodes being coaxial and axially spaced from one another, a plurality of axially separated anodes of rotational symmetry inside the column of said secondary cathodes and coaxial with them. the axial separation of said anodes being substantially the same as the axial separation of the secondary-emissive cathodes, an output anode at the end of said column, and means for producing a magnetic field of substantially axial direction.

3. An electron multiplier for the amplification of photoelectric currents, comprising a plurality of cathodes capable of emitting secondary electrons, said cathodes having rotational symmetry and being coaxial and axially spaced forming a column, a photoelectric cathode of rotational shape at one end of said column and caxial therewith, at least a portion of said photoelectric cathode being foraminous to admit light to the inner surface thereof, an anode of rotational symmetry inside said column and coaxial with it, an output anode at the end of said column, and means independent of said plurality of cathodes for producing a magnetic field of substantially axial direction.

. 4. An electron multiplier for the amplification of'photoelectric currents, comprising a plurality of .cathodes capable of emitting secondary electrons, said cathodes having rotational symmetry and being coaxial and axially spaced forming a column, a photoelectric cathode of rotational symmetry at one end of said column and coaxial therewith, at least a portion of said photoelectric cathode being foraminous to admit light to the inner surface thereof, a plurality of axially separated auxiliary anodes of rotational symmetry inside said colurrm, an output anode at the end of said column, and means independent of said plurality of cathodes for producing a magnetic field of substantially axial direction.

5. An electron multiplier for the amplification of thermionic currents, comprising a plurality of electrodes capable of emitting secondary electrons, said electrodes having rotational symmetry and being coaxial and axially spaced forming a column, a therminoic cathode of rotational shape in the axis of the column at one end of it for supplying electrons to the secondary-electron emissive electrode at. said one end, said thermionic cathode being surrounded by an electrostatic controlling electrode of rotational symmetry, an anode of rotational symmetry inside said column and coaxial with it, means for producing a magnetic field of substantially axial direction and a collecting output anode at the other end of said column.

6. An electron multiplier for'the amplification of thermionic currents, comprising a plurality of 'electrodes capable of emitting secondary electrons, said electrodes having rotational symmetry and being coaxial and axially spaced from one another forming a column, a thermionic cathode of rotational shape in the axisof the device at one end of it for supplying electrons to the secondary-electron emissive electrode at said one end, said thermionic cathode being surrounded by electrostatic controlling electrodes of rotational symmetry, a plurality of axially separated anodes of rotational symmetry inside said column and coaxial with it, means for producing a magnetic field of substantial axial direction and a collecting output anode at the other end of said column.

7. An electron multiplier including a plurality of coaxial axially spacedsecondary-electron emissive electrodes of rotational symmetry, means for producing an electrostatic field of radial direction, means for producing a magnetic field of substantially axial direction, and means for producing a desired degree of inhomogeneity in said magnetic field, comprising suitably shaped ing a column, means for producing an electrostatic field in radial direction, means for producing a magnetic field of substantially axial direction, a thermionic cathode positioned and adapted to emit electrons to the secondary-electron emissive electrode at one end of said column, and means for deflecting the electrons'emitted by the thermionic cathode to diiferent regions of the secondary-electron emissive electrode at said one end of the column.

9. An electron multiplier including a plurality of coaxial axially spaced electrodes capable of emitting secondary electrons, electrode means for producing an electrostatic field in radial direction, means for producing a magnetic field of substantially axial direction, a collecting output anode of rotational symmetry at one end of the coaxial electrodes, and means for electrostatically controlling the current flowing from the last of the electrodes emitting secondary electrons to the output anode, such means comprising a grid of rotational symmetry before the output anode.

10. An electron multiplier comprising a plurality of substantially cylindrical hollow elements having secondary-electron emissive inner surfaces, said cylindrical elements being mounted coaxially and axially spaced providing a column and the element at the input end of the column being longer in the'axial direction than some of the remaining cylindrical elements, anode means positioned within said column and substantially coaxial with the cylindrical elements to provide a substantially radial electric field, and means surrounding said anode for producmg a substantially axial magnetic field, said magnetic field varying in strength along the axis of the column within said longer element.

11. An electron multiplier for the amplification of thermionic currents comprising a plurality of secondary-electron emissive electrodes of rotational symmetry, said electrodes being coaxial and axially spaced forming a column, a thermionic cathode of rotational symmetry in the ax1s of the column at one end of it for supplying electrons to the secondar -electron emissive electrode at said one end, said thermionic cathode being surrounded by an electrostatic controlling electrode of rotational symmetry, anode means of rotational symmetry inside said column and coaxial therewith for producing an electrostatic field of radial direction, means for producing a magnetic field of substantially axial direction, means for producing a desired degree of inhomogeneity in said magnetic field compris mg. an annular ring of ferromagnetic material surrounding at least a part of one of the secondary-electron emissive electrodes, and a collecting output anode at the end of said column opposite said one end.

12. An electron multiplier comprising a plurality of rotationally symmetrical secondaryelectron emissive electrodes for successively multiplymg a primary beam of electrons, said electrodes being coaxial and axially spaced forming a column, a thermionic cathode of' rotational shape in the axis of said column at one end thereof for supplying an electron beam to the secondary-electron emissive electrode at said one end, ,an electrostatic controlling electrode of ro-.

tational symmetry around said thermionic cathode for controlling the intensity of said beam, and a plurality of deflecting disks mounted substantially parallel to the direction of said beam for varying the axial position of incidence of said beam on the said emissive cathode at said one end.

13. An electron multiplier comprising a plurality of rotationally symmetrical electrodes having secondary-electron emissive inner surfaces for successively multiplying a primary beam of electrons, said electrodes being coaxial and axially spaced forming a column, anode means positioned within said column and substantially coaxial with said emissive electrodes to provide a radial electric field, means for producing a substantially axial magnetic field along said column, a thermionic cathode of rotational shape in the axis of said column at one end thereof for supplying a beam of electrons to the secondary-electron emissive electrode at said one end substantially throughout a rotational area about the axis of the column, an electrostatic controlling electrode of rotational symmetry around said thermionic cathode for controlling the intensity of said beam, and a plurality of deflecting disks mounted substantially parallel to the direction of said beam for varying the axial position of incidence of said beam on the said emissive cathode at said one end.

DENNIS GABoR.

Images