US3229213A - Bistable electron device comprising axially spaced dynodes - Google Patents

Description

Jan. 11, 1966 J. w. SCHWARTZ 3, 9,

BISTABLE ELECTRON DEVICE COMPRISING AXIALLY SPACED DYNODES Filed July 20, 1962 2 Sheets-Sheet 1 UHF FIG 2 SOURCE 29 FIELD FORMER) L i 40/fifi /26 a OUT UT SECONDARY P B ELECTRONS 1 I ACTUATING g 3y, 54 BEAM m 2 56 32 \MESH REFLECTOR r 36 24A COLLECTING ELECTRODE /DYTLODE RING INVENTOR JAMES W. SCH WARTZ Jan. 11, 1966 J, w. SCHWARTZ 3,22

BISTABLE ELECTRON DEVICE COMPRISING AXIALLY SPACED DYNQDES Filed July 20, 1962 I 2 Sheets-Sheet 2 29 FOCUSING UHF SlGNAL SOURCE A UTiLiZATlON DEVICE -\1T --{TlE-- s4 DELAY AND AND AND GATE GATE GATE START EXTINGLHSH VQLTAGE VOLTAGE.

72 p58 66 RESET sET SET INVENTOR JAMES w. SCHWARTZ F IG. 3. B

Y a A ATTORNEY United States Patent 3,223,213 BISTABLE ELECTRGN DEVICE COMPRISENG AXHALLY SPACED DYNQDES James W. Schwartz, Phoenix, Ariz., assignor, by mesne assignments, to the United States of America as represented by the Secretary of the Navy Filed July 20, 1962, Ser. No. 211,463 8 Claims. (Cl. 328255) This invention relates to flip-flops for use in computers and more particularly relates to the use of an electron multiplier as a flip-flop.

The prior art discloses many devices for use as flipflops in computers, such as relays, bistable vacuum-tube rnultivibrators, and bistable magnetic cores. The various states of these devices represent bits or parts of a code character for use in the computer. The speed at which these flip-flops operate eventually determines the speed of the computer. However, the speed at which these devices can be switched from one state to another is limited by their reactance.

One object of this invention is to provide a faster flipflop, and more particularly one which will operate in the millimicrosecond range.

Another object of this invention is to provide a flipflop in which the speed of operation is dependent upon the transit time of electron beams.

A further object of this invention is to provide a memory element which consists of a secondary-emission electron multiplier in which a multiplication factor of unity or greater will indicate one state and a multiplication factor of less than unity another state.

In one illustrative embodiment of this invention the electron multiplier comprises a primary cathode, control electrodes, and accelerating electrodes for producing an electron beam; two dynodes containing central holes through which the electron stream passes; and a collector electrode between the dynodes which indicates the space charge density therein. The inner surfaces of the dynodes are coated with a secondary-electron emissive surface. A reflector may be used at the exit of the holes to reflect the electron stream between the two dynodes. The electrons which are between the dynodes, bombard the secondaryelectron emissive surfaces in response to the ultra high frequency voltage applied to these dynodes. This bombardment creates secondary emission which increases the space charge.

The electron multiplication is obtained by secondary emission. In this process an electron is accelerated toward a dynode by the alternating potential applied to the dynode. When it strikes the dynode, it imparts suflicient energy to enable several electrons to escape. These electrons are, in turn, accelerated toward the other dynode on the second half cycle of the alternating potential to repeat the process. in this way the number of electrons are increased at each half cycle of the alternating voltage that is applied to the dynodes.

The low multiplication-factor state is maintained by a focusing field which causes the secondary electrons to move toward the central holes in the dynodes. They pass through the holes and then are lost so far as electron multiplication is concerned.

The high multiplication-factor state is maintained at a level of multiplication in which electrons are lost through the center holes of the dynodes at the same rate that secondary electrons are provided by bombardment of the dynodes. When the electron multiplier is operating at a high current, the effects of the focusing field are reduced by the high space charge so as to effectively increase the multiplication factor of the device.

In accordance with one feature of this invention, a high electron multiplication is triggered by a low actuating cur- 3,2292% Patented Jan. 11, 1966 rent from the primary cathode, and a low multiplication factor is caused by an increase in this actuating current, which drives the core of electrons far enough so that they leave the multiplication region.

In accordance with another feature of this invention, the memory is switched from one state to the other by producing bunching in the actuating beam which is in such a phase relationship with the dynode voltages so as to spoil or to aid multiplication.

The invention and the above noted and other features thereof will be understood more clearly and fully from the following detailed description with reference to the accompanying drawings in which:

FIG. 1 is a diagrammatic view illustrative of one embodiment of this invention showing an actuating beam, a reflector for directing the electrons into the multiplying region, two dynodes for secondary emission purposes, a field former for urging the electrons towards the central holes in the dynodes, and electrodes for read out purposes;

FIG. 2 is a perspective view illustrative of another embodiment of this invention, in which additional holes are placed in one dynode and the electrodes are placed opposite these holes, so that electrons spilling out of the dynode will be collected by the electrodes for read out purposes; and

FIG. 3 is a diagrammatic view of the memory element showing the output terminal, the set and reset terminals.

Referring now to the drawings an electron multiplier is shown with an evacuated enclosing vessel 10 as shown in FIG. 2 which may be made for example of glass. A cathode 12 is mounted adjacent to one end of the enclosing vessel. The cathode may be of the equal-potential, indirectly-heated type, having an electron emissive surface, normal to the longitudinal axis of the enclosing vessel It], as shown in FIG. 2. The cathode heating element is not shown. A control electrode 16 is mounted opposite the emissive surface 14 of the cathode 12. This control electrode contains an aperture 18 for focusing the electron beam.

Mounted opposite to the control electrode 16, in line with it and the cathode 12, is an accelerating electrode 20. This accelerating electrode contains an aperture 22 for accelerating the beam of electrons. A field former 28 is mounted opposite to the electrodes 16 and 20. This field former may be a cylinder with an axis normal to the aforementioned electrode. At each end of the field former is a dynode 24 and 26. These dynodes are insulated from field former 28 and are connected to a source of ultra high frequency voltage 29. The dynode 24 contains a central aperture 3%! and a secondary-electron emissive surface 34; similarly, dynode 26 contains central aperture 32 and secondary-electron emissive surface 36.

In the embodiment of FIG. 1, a mesh reflector 38 is mounted opposite the aperture 32 of the dynode 26. The collector electrode ring 40 is mounted between the dynodes.

In the embodiment of FIG. 2 additional apertures along the edges of the dynode 26 are shown, for example 41, 43, 44 and 46. Opposite these apertures and out side of field former 28 are collector electrodes 48 and St The embodiments of FIGS. 1 and 2 differ in the position of the collecting electrodes and in the structure of dynode 26. In the embodiment of FIG. 1 the collecting electrode is placed between the dynodes 24 and 26; in the embodiment of FIG. 2 the collecting electrodes are placed ice on the opposite side of dynode 26 from dynode 24. In

the latter embodiment, holes 41, 43, 44 and 46 must be made in dynodes 26 so that the electrons may reach the collecting electrodes 48 and 50. FIG. 1 shows the read out terminal 52, which develops a voltage in the absence of an electron flow through from the collecting electrode ring 40, through resistor 54 to voltage source 56.

The memory element may be best understood with reference to FIG. 1. When the memory element is in the state, there is no output voltage on terminal 52. This is because the space between the dynodes 24 and 26 contains an electron cloud, which provides a source of current from the electron collector 40 through resistor 54 to the positive terminal of voltage source 54. This drops the voltage from 56 across the resistor 54 and drives the voltage of terminal 52 toward 0.

When the memory element is in the 1 state a voltage appears at terminal 52. This is because there are no electrons between the dynodes 24 and 26. The circuit containing voltage source 56, resistor 52 and electron collector 40 is effectively open-circuited. The voltage at the positive terminal of voltage source 56 appears at terminal 52 since the current through resistor 54 is negligible.

Assume first that the memory element is in the 1 state, that is, there are no secondary electrons between the dynodes and therefore a constant output voltage appears at terminal 52. To switch the memory element from the 1 state to the 0 state an actuating beam must be introduced at aperture 30. This beam, which may originate in a beam selection section such as selectron geometry will supply the space charge forces and initiating electrons to go from this vacant space to the multiplying state in which secondary electrons are present between the dynodes.

The electrons which have been supplied by the actuating beam are accelerated first towards dynode A and then towards dynode B by the ultra high frequency voltage source connected to these dynodes. When these electrons strike the secondary-electron emissive coating 34 and 36 on the dynodes 24 and 26 more electrons are driven off. This field former 28 focuses the electron towards the center hole in the dynodes. Then the secondary electrons which are driven off of the dynodes are equal to the electrons that are lost through the center holes in the dynodes due to the focusing field, an equilibrium state is reached. This equilibrium is reached because, when a large cur-rent exists in the device, the space charge forces spoil the 0- cusing for the electrons near the edge of the core in the memory element so that they do not pass through the hole, thus increasing the number of electrons between the dynodes. The electrons between the dynodes are attached to the collecting electrode ring 40. These electrons cause a current to flow through resistor 54 to the source of voltage 56. The voltage drop across resistor 54 drives the terminal 52 to ground to indicate the 0 state of the memory device.

Reversion to the vacant state is also produced by the actuating beam. One way of doing this is to drive the multiplying core of electrons far enough that they completely pass outside of the multiplying region. The electric field created by the increased actuating beam drives the secondary electrons beyond the edges of the dynodes and away from the central holes so that they are lost for electron multiplication purposes.

. Another means of reverting to the vacant state is to produce bunching in the actuating beam which is in such a phase relationship with the dynode voltages as to spoil multiplication. The electrons are bunched in such a way as to create the greatest electric field due to the bunched electron at the cross-over point of the ultra high frequency voltage applied to the dynodes. This will provide the maximum displacement away from the central holes towards the edges of the dynodes to the secondary electrons since at that point they have the greatest travel time under the influence of the electric field from the actuating current. This reduces the number of secondary electrons available to strike the next dynode and create further secondary electron emission. Also, the electric field will counteract the UHF field during one half of each cycle and thus slow down the speed of electrons.

When the memory element has reverted to its vacant state and no electrons are made available to the collecting electrode ring, the current through resistor 54 ceases and the voltage at terminal 52 is driven to the 1 state.

The embodiment of FIG. 2 operates in substantially the same manner as the embodiment of FIG. 1. However, the dynode 26 contains further holes 41, 43, 44 and 46, which are used for read out purposes. When the memory element is in its multilpying state electrons will spill out of these holes. The collecting elements 48 and 59 for read out purposes are opposite these holes and will attract electrons in the same manner as the collecting electrode ring 40 collected the electrons in the embodiment of FIG. 1.

The control and accelerating grids 16 and 20 shown in FIG. 2 may be used to produce the bunching and actuating beam which is necessary to return the memory element to the vacant or 0 state. To accomplish this, the accelerating electrode must be modulated by the ultra high frequency voltage source of the same frequency as that used to modulate the dynodes 24 and 26. This modulation will create a velocity modulated beam, in that the electrons will pass through the electrodes during the positive cycle will be accelerated and those that pass through during the negative cycle will be retarded. Consequently, the faster moving electrons of the beam will catch up with the slower moving electrons thus causing an in creased electron density in the beam. This variation in the density of the beam of the electrons is called bunching. To spoil the multiplication of the memory element, the bunches of electrons should pass across the aperture 30 at the time that the ultra high frequency voltage begins to accelerate the electrons which have been emitted from the secondary-electron emission surface of dynode 24 towards the opposite dynode. This will cause the maximum deflection of these electrons away from the multiplication region of the dynode. The phase of this bunching is controlled by the distance from the modulated electrodes to the aperture 30. If the electrodes 16 and 20 are used for the purpose of bunching electrons, the beam must be provided by another electron accelerator for the electron gun.

The number of secondary electrons emitted per primary electron depends upon the velocity of the primary bombarding electrons, the nature of the material, the condition of its surface and the potential conditions surrounding the bombarded surface. For typical surfaces of barium oxide with a primary voltage of 400 volts, the

ratio of secondary electrons to primary electrons will be approximately 4.5.

A common secondary-electron emission coating consists of a thin film of cesium on an intermediate film of cesium oxide covering a silver electrode. The secondary emission from this surface approaches milliamperes per watt of incident primary-electron energy. As the energy of the primary electrons is increased the secondaryernission ratio increases to a maximum and then gradually decreases.

FIG. 3 shows a noncomplementary embodiment of the invention. When the set terminal 58 is pulsed AND gate 60 is opened to pass an extinguishing voltage from voltage source 62. This voltage is applied to accelerating electrode 20 which increases the electron stream from electron gun 12 through aperture 30 of dynode 24.

If the memory element is in its 0 state, there is an electron cloud between the dynodes 24 and 26. This cloud results from the multiplication of electrons by secondary emission. The electrons are collected by collector electrodes 48 and 50 and cause a current to fiow through resistor 54, to drive the voltage at 52 to zero.

The extinguishing voltage 62 which appears at electrode 20 accelerates the electrons from electron gun 12 causing a high current to flow through the apertures 30 and 32 of the dynodes. This current drives the secondary electrons to the edges of the dynodes and beyond the secondary-electron emissive surface, where they are lost for electron multiplying purposes.

Since no electrons can be collected by electrodes 48 and 50, no current flows through resistor 54 and therefore terminal 52 rises to the positive voltage of source 64 to indicate the 1 state of the memory element.

This 1 state can also be obtained by pulsing terminal 66. This will open AND gate 68 to the UHF voltage from source 29. This voltage is applied through the AND gate to electrode 20 to modulate the electron beam from electron gun 12. The delay 76 is adjusted to provide the proper phase in the modulating volt-age such that the electron beam is bunched so as to spoil multiplication. This will interrupt the current from battery 64 and cause terminal 52 to register a l.

A pulse at reset terminal 72 opens AND gate 74 to voltage source 76, which appears on grid 20. The voltage from source 7s is less than that from voltage source 62. The electron beam, controlled by this voltage, provides the initiating electrons for the secondary emission process. The electric field created by the electron current through the apertures 39 and 32 in the dynodes must not be so strong as to tangentially accelerate the secondary electrons to a sufficient degree as to force them beyond the edges of the dynodes.

This initiating voltage will start the secondary-electron emission to switch the memory element into its electron multiplying state. The initiating electrons from electron gun 12 and newly released secondary electrons will be forced back and forth to bombard dynodes 24 and 26 under the influence of the UHF source 29. This bom bardment will, in turn, cause further secondary-electron emission until an equilibrium state is reached in which the newly created secondary electrons equal the elec trons lost through the apertures 3i? and 32 in the dynodes.

In this state an electron cloud is present between the dynodes. The electrons are collected by electrodes 48 and St) to provide an electron flow through resistor 54 to voltage source 64. This current causes a voltage drop across resistor 54 and drives the terminal 52 to the 0 state.

The embodiment of FIG. 3 is that of a noncomplementing ffip-flop. However, it is clear that conventional steering gates can be used to convert this embodiment to that of a complementing flip-flop.

The above embodiments can be used as memory elements which wi l operate in the millimicrosecond range. These elements may be used for fast acting random access or cyclic access computers.

It is important to realize that the beams used in this device are not required to alter the potential of any electrode but only to alter the free space potential. Hence, there are no capacitive lags involved and the only time limit to the speed of the device is that of the transit time of the electrons. For spacings of a few thousandths of an inch and dynode voltages in the order of a hundred volts, a transit time of less then one tenth of a millimicrosecond are involved. The secondary emission process itself takes less then one one-hundredth of a millimicrosecond. The basic element is capable of either essentially continuous readout or an interrogated type of readout requiring a period appreciably less than one millimicrosecond.

Control of the state of the binary element is readily achieved by means of a control electron beam which enters the multiplication region and alters the electrostatic fields therein. Beam deflection techniques such as used in traveling wave oscilloscopes may be useful in directing or steering control and/ or interrogating beams. Similarly klystron type beam bunching techniques may also be applied. The bandwidth and operating frequency of either of these devices is high enough to provide subnanosecond operation.

It is apparent that the basic element is capable of power gain. This immediately suggests the direct coupling of elements to perform control and logic functions. The coupling would, of course, be by means of electron beams emanating from individual multiplier cells and going to one or more other cells. Each cell can control not only the presence or absence of a beam of secondary electron origination but one or more beams from a static or DC. flood source.

A complete computer consisting of appropriately intercoupled layers of many cells could be contained in a small volume. Obviously multiple traversals of a logic sequence through the layers of the computer can be effected by beam reflection or bending procedure beyond the outside surfaces of the multilayered array.

While there has been described what is at present considered the preferred embodiment of the invention it will be obvious to those skilled in the art that various changes and modifications may be made therein Without departing from the invention and it is therefore aimed in the appended claims to cover all such changes in modifica tion that may fall within the scope of the invention.

What is claimed is:

1. A flip-flop, having a first state with an electron multiplication factor of less than unity and a second state with an electron multiplication factor of more than unity, comprising:

a source of a beam of electrons;

a plurality of axially spaced dynodes positioned in the path of said beam of electrons containing holes through which electrons may pass;

an alternating voltage source connected to said dynodes, whereby electron in the space between said dynodes are alternately accelerated first toward one dynode and then toward another dynode to cause increased secondary emission such that the increase in elec trons due to secondary emission from said dynodes will be greater than the loss of electrons through said holes when said flip-flop is in said second state and less than the loss of electrons through said holes when said flip-flop i in said first state; and

means for initiating and terminating the secondary emission action of said dynodes whereby said flipfiop may be switched from said first state having an electron multiplication factor of less than unity to said second state having an electron multiplication factor of more than unity.

2. A flip-flop, in accordance with claim 1, including mean providing a focusing field which causes said secondary electrons to move toward at least one of said holes in said dynodes.

3. A flip-flop, in accordance with claim 2, in which at least one of said dynodes has a secondary-electron emissive surface.

4. A flip-flop, in accordance with claim 3, in which said alternating voltage source is characterized by an UHF signal.

5. A flip-flop, in accordance with claim 2, in which an electron collector is mounted between said dynodes whereby said collector is nonconductive when said flip-flop is in said first state and said collector is conductive when said flip-flop is in said second state.

6. A flip-flop, in accordance with claim 2, in which one of said dynodes contains at least one additional hole through which electrons may pass for subsequent computer operation purposes.

7. A flip-fiop, in accordance with claim 4, in which said source of electrons comprises:

(a) a beam selection section; and

(b) a reflector mounted in the path of the beam.

8. A millimicrosecond flip-flop, having a first state with an electron multiplication factor of less than unity and a second state with an electron multiplication factor of more than unity, comprising:

(a) a plurality of axially spaced dynodes;

7 8 (b) at least one of said dynodes having a secondary- (g) means responsive to the electron density between electron ernissive surface; said dynodes for controlling further electronic corn- (c) an UHF voltage source connected to said dynodes; ponents. (d) a source of primary electrons; (e) means for forming said primary electrons into a 5 Refemmes fitted y the Examiner (S i l f 1 t l d f d b UNITED STATES PATENTS con ro means or a terna e y irec ing sai earn of electrons between said dynodes with such a veloc- 2026892 1/1936 Hemtz 328 64 2,071,515 2/1937 Farnsworth 32824 3 X ity and phase as to cause them to be multlplled by 2 639 12/1941 Honmqnn 328 277 secondary emission when dnven against said second- 10 2,726,328 12/1955 clogston 328 255 X ary-electron emissive surface by said UHF voltage source and for directing said beam of electrons be- GEORGE N WESTBY Primary Examiner tween said dynodes with such a velocity and a phase as to cause them to spoil secondary emission multi- ROBERT SEGAL, Examinerplication; and v 15

Claims (1)

1. A FLIP-FLOP, HAVING A FIRST STATE WITH AN ELECTRON MULTIPLICATION FACTOR OF LESS THAN UNITY AND A SECOND STATE WITH AN ELECTRON MULTIPLICATION FACTOR OF MORE THAN UNITY, COMPRISING: A SOURCE OF A BEAM OF ELECTRONS; A PLURALITY OF AXIALLY SPACED DYNODES POSITIONED IN THE PATH OF SAID BEAM OF ELECTRONS CONTAINING HOLES THROUGH WHICH ELECTRONS MAY PASS; AN ALTERNATING VOLTAGE SOURCE CONNECTED TO SAID DYNODES, WHEREBY ELECTRONS IN THE SPACE BETWEEN SAID DYNODES ARE ALTERNATELY ACCELERATED FIRST TOWARD ONE DYNODE AND THEN TOWARD ANOTHER DYNODE TO CAUSE INCREASED SECONDARY EMISSION SUCH THAT THE INCREASE IN ELECTRONS DUE TO SECONDARY EMISSION FROM SAID DYNODES WILL BE GREATER THAN THE LOSS OF ELECTRONS THROUGH SAID HOLES WHEN SAID FLIP-FLOP IS IN SAID SECOND STATE AND LESS THAN THE LOSS OF ELECTRONS THROUGH SAID HOLES WHEN SAID FLIP-FLOP IS IN SAID FIRST STATE; AND MEANS FOR INITIATING AND TERMINATING THE SECONDARY EMISSION ACTION OF SAID DYNODES WHEREBY SAID FLIPFLOP MAY BE SWITCHED FROM SAID FIRST STATE HAVING AN ELECTRON MULTIPLICATION FACTOR OF LESS THAN UNITY TO SAID SECOND STATE HAVING AN ELECTRON MULTIPLICATION FACTOR OF MORE THAN UNITY.

Images