US2869111A - Electron beam switch tube operation of a ferroelectric matrix - Google Patents

Description

Jan. 13, 1959 D. R. YOUNG 2,859,111

ELECTRON BEAM SWITCH TUBE OPERATION OF A FERROELECTRIC MATRIX Filed Nov. 17, 1953 2 Sheets-Sheet 1 OUTPUT 7 Y SWITCH TUBE X SWITCH 'TUBE 3 INVENTOR. DONALD R.YOUNG FLIP FLOP. 26

.oos;n m 6 4; AGE

35 SIGNAL SOURCE Jan. 13, 1959 D. R. YOUNG ELECTRON BEAM SWITCH TUBE OPERATION OF A FERROELECTRIC MATRIX 2 Sheets-Sheet. 2

Filed Nov. 17. 1953 T Iii T Y SWITCH TUBES STORAGE MATRIX X SWITCH TUBES m N R U m X T O W M 00 T U T l Y a T O PA HTT N N w T T C CNN E E D W m WO VD G m M N w E NL E TNH llll l. I s R IA UEC N PKA O AE D T Y 0 B 5 2 p 8 2 v I 2 3 4 .IX 1- y I .4 .w .w s ilx R T N O C E T W v. R w v D 2 3 4 Z Z T Z 2 m e w 2 L F P L F I 1 2 3 m" Y X I a a s 4 5 Y r v 3 T w 2w w 5 3 03F w 3 02m fi ice ELECTRON BEAM SWETQH TUBE OPERATION OF A FERRUELECTRIC MATRIX Donald R. Young, ionghkeepsie, N. Y., assignor to international Business Machines Corporation, New York, N. Y., a corporation of New York Application November 17, 1953, Serial No. 392,615

11 Claims. (Cl. 34il173) This invention is directed to memory systems and principally to a system for storing binary information in a ferroelectric matrix and subsequently determining which one of two representations has been stored therein.

Ferroelectric capacitors are readily adapted for use as binary storage elements as they exhibit two stable states of polarization somewhat similar to the remanence states of magnetic cores. Coincidence currents may be employed for selection of particular elements of a matrix for storage, however, the problem of switching associated with the selection and driving of coordinate channels has remained complex.

My invention relates more specifically to the use of cathode ray tubes having a plurality of discrete target electrodes arranged to be connected to terminals of a ferroelectrix matrix and which provide both a constant voltage source and switching means for selectively driving the ferroelectric elements to one or the other stable polarization states. Two such electron beam switch tubes are employed, one for pulsing the coordinate row channels and the other the coordinate column channels of a two dimensional coincident current ferroelectric capacitor memory array.

An object of the invention is to provide a cathode ray type switch tube of the character described, wherein the beam is caused to advance step by step over each of the target rows cyclically and which is selectively operable to provide potentials to individual ones of a plurality of targets.

Another object is to provide a ferroelectric memory system using cathode ray tube switching means for a matrix array.

A further object is to provide a switching tube of the cathode ray type using plural secondary emissive target electrodes, the emissive current of which may be selectively controlled.

Another object of the invention is to provide a readily controlled high speed switching system for coincident current operated memory matrices.

Other objects of the invention will be pointed out in the following description and claims and illustrated in the accompanying drawings, which disclose, by way of example, the principle of the invention and the best mode, which has been contemplated, of applying that principle.

In the drawings:

Figure 1 is a diagrammatic representation of the hysteresis curve for a ferroelectric capacitor such as that employed in the system described.

Figure 2 is a schematic diagram of the circuit arrangement and apparatus for operation of a two dimensional ferroelectric array.

Figure 3 illustrates an alternative arrangement of a ferroelectric matrix and signal detecting means.

Figure 4 is a diagram of the circuit for control of a number of matrices such as a cubical array.

Ferroelectric capacitor elements employed in memory systems preferably employ dielectrics having somewhat rectangular hysteresis loops and low coercive force. Figure 1 illustrates the characteristic for a barium titanate crystal of this type and it will be observed that the vertical axis is designated P, representing the degree of electrical displacement or polarization, and the horizontal axis is designated E, representing the applied electrix field strength or voltage presented to the capacitor terminals. Points on the curve labeled a, b, c and d represent stable and saturation states of polarization. When polarized in one direction by an electric field plus E applied to one terminal, the capacitor exists at point c, for example, which is a saturation point, and upon termination of the field will return to point a where it will remain in a stable condition. Application of an electric field of reverse polarity to the same terminal will cause the capacitor to traverse the curve from point a to point 0' and, on relaxation of the field, will return to point 12 which is a second stable state of polarization of reversed polarity to the first mentioned state a. With the capacitor in an initial state b and a negative potential applied to the same terminal, the curve is traversed from point b to point d and subsequently returns to b when the negative field is removed.

At residual points a and b, there is no net electric field within the dielectric or external to it and the polarization charge is equal and opposite to the surface charge.

out pulses are conventionally applied to cause or fail to cause a change in its polarization. For example, with point a arbitrarily selected as representing a binary one and b a binary zero, a negative pulse applied causes a shift from a to d and finally to b if a one is represented, and a shift from b to d and back to b if a zero is represented. The slope of the curve traversed in going from these two storage points to point d is diiferent and, as the slope is proportional to the capacitance presented by the ferroelectric, the two states can be distinguished for example by comparison with a fixed value standard capacitor or by the magnitude of current flow as a result of the read-out pulse.

The ferroelectric matrix employed in the present invention may be made up of a plurality of separate individual capacitors or as illustrated in Figure 2 where the dielectric for each element is integral and consists of a large single crystal F. The crystal may have a plurality of parallel conductive lines X etched or deposited by other methods on one surface and a second group of parallel lines Y on the other surface, with those on each surface arranged perpendicular to those on the opposite surface. Only nine lines labeled X and Y with corresponding sub scripts are illustrated on each surface, however, it is known that they may be closely spaced to provide a great number of storage cells in a single crystal and furnish a memory unit of small physical proportions.

A separate capacitor unit is formed at each point where the X and Y lines or channels are in juxtaposition and a bit of binary information may be stored in each one individually by energizing the lines forming its two terminals with a potential individually less but coincidentally greater than the coercive force. Referring to Figure 1, if each of the capacitors formed by intersection of X and Y lines are assumed to be in a zero representing polarization state b and it is desired to store a one, for example, in capacitor XY9, a positive force ofE applied to one terminal will cause the capacitor to go from b to c and return to a A pulse of such magnitude may not be applied to line X, as the row of capacitors having line X as one terminal would each go to point a. The coincident system provides for application of a field of plus E/2 or half the voltage required on line X and the other half as minus E/ 2 applied on line Y The negative voltage applied to the other terminal is equivalent to applying a positive voltage on the first terminal and, as a result, only capacitor X Y is changed from state b to state a while the remaining condensers in row X and in column Y have had afield applied which is insufficient to change their final state of polarization as it is less than the coercive force.

one of 2 output lines.

The present invention employs a cathode ray type tube both for switching and providing the potentials for the X and Y lines in a novel manner now to'be'described As shown in Figure 2, each of the X and Y coordinate lines are connected individually to discrete targets of a pair of tubes Sx and Sy. The targets are indicated collectively by the reference character 10 and'are positioned at one end of each cathode ray tube. Other elements are provided within the tubes and include an indirectly heated cathode 11, control grid 12, accelerat ing anode 13, shield electrodes 14, horizontal deflection plates 15, vertical deflection plates 16 and a collector electrode 17 arranged within an evacuated envelope 18. Elements 1.1, 12 and 13 comprise the electron gun structure provided for producing an electron beam which is directed by means of the deflection plates 15 and 16 toirnpinge upon a selected one of the targets 10. The beam is turned on under control of the grid 12 and is directed to particular elemental targets 10 by voltages applied to the deflection plates. As is'well known in the art, the pairs of deflection plates are physically arranged to produce electrostatic fields at right angles to the beam and to one another. It is to be understood that these plates will have varying voltages applied to them from a saw-tooth waveform generator, for example, in order to produce line and frame scanning as in a television raster or staircase type scanning system, however, appropriate voltages may be applied for producing single line or row scansion or to position the beam selectively at any one target. The means for producing different types of scansion are well known in the art and, as they are not necessary for an understanding of the present invention, need not be described in detail here.

As the primary electron beam bombards an elemental target 10, secondary electrons are emitted and are attracted to the collector electrode or grid 17. The target is caused to adjust its potential to that of the collector 17 since it is essentially floating because of its connection through the diode 19, capacitor 20 and resistance 21 for tube Sx, and diode 22 and resistance 23 for tube Sy.

Momentarily, the only source of electrons to the target is from the primary beam and the only output for electrons is through those emitted as secondary electrons to the collector grid 17. If the collector is at a positive potential with respect to the target, a greater number of secondary electrons will be attracted to it than are received by the target in the primary beam and, as a result, the target reaches a positive potential with respect to ground. If the collector is at a negative potential with respect to the target, the secondary electrons are not attracted to it but as the primary beam electrons 4 accumulate on the target, it becomes increasingly negative with respect to ground until it becomes more negative than the collector 17-at which time the latter starts to attract secondary electrons and the target and collector potentials are again in equilibrium. The potential of the target then may be made positive or negative depending upon the potential of the collector 17. A resistor r of high ohmic value is connected between each target lead and ground to bleed off spurious charges 10 which are accumulated in a progressive manner when adjacent targets are bombarded.

It will be recalled that the potential applied to a particular X lead of the matrix in a coincidence selection system should be one-half or at least less than that which produces an electric field equal to the coercive force while the additional field is supplied to the Y lead as a potential of opposite polarity. For this reason, the collector grid 17 of the Sx and Sy tubes are connected by leads 24 and 25 respectively to a flip-flop unit 26. The

latter is a conventional trigger device and is shown in block diagram form as it constitutes no part of the present invention per se and will be but briefly described. Triggers or flip-flops conventionally include two vacuum tubes so interconnected that when one is conductive the other is cut off. The network will remain in either of ..these states stably until a controlling pulse is applied to'reverse the conductive status of the tubes. Voltage values at points in the trigger circuit differ according to whether the tubes are in one or the other relative conducting states and are used for various circuit controlling purposes as is well known. With the plate potential supply for the tubes positive with respect to ground and the cathodes connected to a source negative with respect to ground, the two conventional output terminals of trigger 26 may be made to alternately assume positive and negative potentials in response to input pulses successively applied to their commonly connected grids through an input lead 28. Assuming flip-flop 26 to be in a first state of equilibrium, lead 24 will be positive and lead 25 negative, however, an input pulse applied to lead 28 causes a reversal of the potentials appearing on the output terminals and applied to leads 24 and 25 and the trigger remains in this state of equilibrium. A succeeding input pulse then returns the output terminals and leads 24 and 25 to their original polarities.

The grids 12 of each of the cathode ray switch tubes S are commonly connected by a lead 30 and are biased negatively so as to normally cut off the electron beam. This biasing voltage is applied to the grids by a source of potential 32 coupled to lead 30 by a paralleled resistor 33 and diode 34. Input signals of positive polarity 'and of sufficient magnitude to overcome the negative control grid bias are applied to the grids 12 through a terminal 35 which is connected to lead 39 by a coupling capacitor 36. The heater elements for the cathodes 11 of each tube are also commonly energized from a source 37 and the anode 13, cathode 11 and focusing discs 14 have suitable potentials applied from a source 38.

Considering storage of a binary one in a ferroelectric capacitor element located at the intersection of leads X and Y the cathode beams of each of the tubes Sx and 8 2 are turned on by a positive pulse applied to terminal 35. This pulse traverses condenser 36 and overcomes the bias of battery 32 so that the grid 1.2 becomes less negative. The potentialsapplied to the deflection plates 15 and 16 of both tubes are adjusted so as to direct the electron beam to the targets 1;!) connected with leads X and Y respectively. With the trigger 26 in a first state of equilibrium, lead 24 is positive and lead 25 negative so that the collector grid 17 of tube Sx is positive and that of tube Sy negative. Under these conditions, the targets and leads connected thereto assume similar potentials during the period that the beam is turned on. A coincidenceiof two electric fields, each of magnitude E/2, is applied only to the capacitor element between the X and Y leads and this capacitor is caused to traverse its hysteresis loop from point 17 to point 0. As the pulse applied to terminal 35 ends, the cathode beams in both tubes cut off as the bias from source 32 takes over and the tmgets return to ground potential. Capacitor element X Y then shifts from point 0 to point a Where it remains stable regardless of the fact that both leads X and Y and both condenser terminals are at ground potential.

The deflection system of tube Sy is regulated to scan each of its targets in sequence during the period that the beam in tube Sx is positioned at one target. The frequency of the saw tooth wave generator controlling the deflection plates of tube Sy is, therefore, made to be n times that of the generator controlling the deflection plates of tube Sx, where n is the number of columns of the storage matrix. Each of the rows of the matrix are thus conditioned for reading or writing in sequence.

To read out the binary one stored in address position X Y a pulse is applied to the lead 23 and flip-flop 26 reverses its output polarity so that lead 24 is now negative and lead 25 is positive. At the time during scansion of the targets that the beam would be directed to the proper target of both tubes, a pulse is again applied to terminal 35 to turn on the cathode beams of both tubes. Collector 17 of tube 'Sx is now negative and a voltage of minus E/Z is applied to lead X collector 17 of tube Sy is now positive so that a voltage plus E/2 is applied to lead Y This reversal of polarity is equivalent to a minus E field applied and capacitor X Y now reverses its polarization, shifting from point a to point d and, when the pulse to terminal 35 ceases and the cathode beams are turned off, returns to point b. The slope of the hysteresis curve in going from point a to point b is great and a large capacitance is presented by the ferroelectric, however, as it cannot charge instantaneously the voltage appears principally across the capacitor 20 and an output signal is obtained.

With the flip-flop 26 in this state of equilibrium, that is with lead 24 negative and lead 25 positive for readout operations, turning on the cathode beam of the tubes S when directed to a ferroelectric element in a zero representating state or at point 12, causes a change in polarization from b to d and a low capacitance is presented by the ferroelectric capacitor with the drop appearing thereacross and no output signal obtained at capacitor 20. It is, therefore, seen that binary ones and .zeros are distinguished by the presence or absence, respectively, of a voltage at the output terminals. Detection of the stored binary representation may also be determined by sensing the current flow in the collector electrode circuit as in electrostatic storage systems.

As previously mentioned, the state at which the several ferroelectric storage elements exist in representing binary information may further be determined by means of a standard capacitor element and also by means of current sensing devices. Apparatus of the latter type is illustrated in Figure 3 and in this embodiment the storage matrix is illustrated as comprising a plurality of separate ierroelectric capacitors F rather than a single crystal.

In order to avoid second order effects from non-switching cores on read-out, the row coordinate channels X are double leads and, at one end, are wound oppositely on a read-out magnietic core 40. The column coordinate channels Y are single leads as heretofore. In alternate columns, the capacitor elements are connected alternately to one and the other of the double leads X. The double leads are commonly connected at their input terminals to targets of a cathode beam switch tube and at their opposite terminals to ground through diodes ll and a resistor 42.. The Y leads are likewise connected to ta;- gets of a second cathode beam switch tube at one end and the opposite terminals connected to ground through diodes 43 and a resistor 44. An output circuit comprising a secondary winding 45 is positioned about the core 40 and is thereby inductively related to each of the X channel double row windings. To interrogate a particular ferroelectric capacitor, the cathode ray switch tubes have their beams directed to the corresponding corrdinate lines X and Y and are pulsed for read-out as heretofore mentioned. If a binary one has been stored and the particular ferroelectric element changes from stable polarization state a to states d and b, a large capacitance is presented, however, as the ferroelectric storage element cannot charge instantaneously, a current pulse appears on the X lead and traverses the coil on core 40 associated therewith to ground and a voltage is induced in the output winding 45. With a binary zero stored and capacitor element changing between 11 and d and return, the ferroelectric capacitor blocks flow of current through the associated windings on core 40 and no output signal is obtained. In reading out the binary one stored in a particular terroelectric capacitor as described, each condenser in the associated coordinate X row is subjected to a voltage of minus E/2 and each condenser in the associated coordinate Y column is subjected to a voltage of plus E/Z. These condensers may be in one or the other stable polarization states represented as points a and b on the hystersis curve of Figure 1 since other binary bits are stored at these addresses. The field to which these elements are subjected is therefore plus or minus an amount less than the coercive force of the dielectric and less, therefore, than the field necessary to cause a change in polarization, however, since particular ones are driven away from their stable states at a and b some capacitance is presented. cumulatively, the effective capacitance may be sufiicient to produce an erroneous read-out signal and employment of the double leads and oppositely wound coils on the core 40 tends to cancel out these second order effects of read-out.

Cathode ray switch tube S may be employed to control and drive the coordinate channels of a plurality of matrices such as that illustrated in Figure 2 or Figure 3. For example, the collector electrodes 17 of each of the X and Y switch tubes of the plurality of matrices can be connected together for common operation in a serial type system such as that illustrated in Figure 4. Here the read-write control lead 28 and flip-flop 26 control the leads 24 and 25 as in Figure 2, however, the collector electrode of a plurality of switch tubes associated with matrices Z to Z, are coupled to the leads 2 and 25. Particular ones of the matrices are selected for writing or storage by turning on the beam of any of those tubes S associated with the selected matrix by pulsing the Z leads 35-4, 352, 353, 354, etc.

For parallel type operation of a plurality of matrices Z, a separate flip-flop 26 is required for each matrix as the entries to each are made simultaneously and the control grids 12 of the associated switch tubes S must be simultaneously pulsed.

While there have been shown and described and pointed out the fundamental novel features of the invention as applied to a preferred embodiment, it will be understood that various omissions and substitutions and changes in the form and details of the device illustrated and in its operation may be made by those skilled in the art without departing from the spirit of the invention. it is the intention, therefore, to be limited only as indicated by the scope of the following claims.

What is claimed is:

l. A switch tube system adapted to selectively apply a positive or a negative potential to selected ones of a plurality of channels and comprising an evacuated envelope having mounted therein means for producing an electron beam, a control grid, for deflecting said beam along a path, a collector grid, a plurality of separate mutually spaced conductive secondary electron emissive target electrodes directly connected to a corresponding 7 one of said channels, means for biasing said control grid to cut oif, means for selectively varying the polarity of said collector grid with respect to said target electrodes, and means for overcoming said control grid bias for intervals when said beam is deflected to a selected one of said target electrodes to cause said target electrode and the associated channel connected thereto to assume the potential selectively maintained on said collector grid. 2. A switch tube system adapted to selectively apply a positive or a negative potential to selected ones of a plurality of channels and comprising an evacuated envelope having mounted therein means for producing an electron beam, a control grid, means for deflecting said beam in a predetermined path, a collector grid, a plurality of spaced conductive secondary electron emissive target electrodes arranged in said predetermined path and directly connected to a corresponding one of said channels, means for varying the potential of said control grid, means for varying the polarity of said collector grid with respect to said target electrodes whereby selected ones of said target electrodes assume the potential of. said collector grid when impinged by said beam, and elec trostatic storage means connected to said channels and responsive to the potentials attained thereon.

3. In a memory stystem comprising an array of storage elements having coordinate channels arranged in rows and columns; means for applying potentials simultaneously to particular ones of said channels comprising a pair of beam switch tubes each comprising an evacuated envelope having mounted therein means for producing an electron beam, a control grid, means for deflecting said beam, a collector electrode and a plurality of spaced conductive secondary electron emissive target electrodes; circuit means connecting said target electrodes with individual ones of said coordinate channels, one switch tube being provided for said row channels and the other switch tube for said column channels; means for varying the potential of said collector grids whereby selected ones of said targets and said channels assume varied potentials when said control grids are turned on and said selected targets are bombarded.

4. A memory system in accordance with claim 3 wherein said latter means maintains said collector electrodes of each of said pair of switch tubes at potentials of opposite polarity in storing information and of converse polarity in determining the information stored.

5. A memory system as set forth in claim 4 wherein said means for deflecting said beam causes the beam of each said pair of switch tubes to be deflected in a predetermined pattern along a path and said target electrodes are arranged in said pattern so as to be bombarded successively at intervals when said control grids are energized.

6. A memory system in accordance with claim 5 wherein said beam deflection means causes said beams to be deflected in a predetermined pattern in a cyclic manner and at dissimilar cyclic rates.

7. A memory system comprising an array of ferroelectric capacitor storage elements each capable of assuming either one of two stable states of polarization in representing binary information and having terminals thereof coupled to coordinate channels arranged in rows and columns, means for applying coincident potentials of opposite polarity to terminals of particular ones of said elements through selective simultaneous-energization of one of said row channels and one of said column channels, said means including a pair of beam switch tubes each comprising an evacuated envelope having mounted therein means for producing and controlling an electron beam and a plurality of spaced conductive secondary electron emissive target electrodes, circuit means directly connecting said target electrodes with individual ones of said coordinate channels, one switch tube being provided for said row channels and the other for said column channels, and means for selectively varying the polarity of said targets to cause said ferroelectric elements to assume one or the other of said stable states of polarization in writing and reading binary information.

.8. A memory system according .to claim 7 wherein read-out means is provided which means is coupled to one group of said coordinate channels.

9. A memory system according to claim 8 wherein said read-out means comprises a standard capacitor of fixed value connected in series with one group of said coordinate channels and across which voltage signals of distinguishable character are developed in response to read-out of binary representations.

10. A memory system according to claim 8 wherein double leads are provided for one group of coordinate channels and alternate ones of the associated ferroelectric elements are connected at one terminal to one of said double leads and the remaining ferroelectric elements associated with that particular channel are connected at one terminal to the other of said double leads, said read-out means comprising a magnetic core having said channels inductively related thereto by said double leads wound in opposite directions thereon, and output signal means comprising a winding on said core.

11. A memory system comprising plural arrays of ferroelectric capacitor storage elements each element of which is capable of assuming either one of two stable states of polarization in representing binary information and having terminals coupled to coordinate channels arranged in rows and columns in each array, means for applying coincident potentials of opposite polarity to terminals of particular ones of said elements in each array through selective simultaneous energization of one of said row channels and one of said column channels, said means including a pair of beam switch tubes each com.- prising an evacuated envelope having mounted therein means for producing and controlling an electron beam and a plurality of spaced conductive secondary electron emissive target electrodes, circuit means directly connecting said target electrodes with individual ones of said coordinate channels of its associated memory array, one switch tube being provided for said row channels and the other for said column channels of each array, means comprising flip-flop devices for varying the polarity of said targets to cause said ferroelectric elements to assume one or the other stable polarization state, and means for selectively controlling the electron beam in particular pairs of said switch tubes in writing into and interrogating storage elements in a particular one of said arrays.

References Cited in the file of this patent UNITED STATES PATENTS 843,746 Hall Feb. 12, 1907 1,595,735 Schmierer Aug. 10, 1926 2,355,212 Farnsworth Aug. 8, 1944 2,547,386 Gray Apr. 3, 1951 2,565,486 Feinstein Aug. 28,1951 2,577,283 Stamper Dec. 4, 1951 2,658,670 Morton Nov. 10, 1953 2,717,373 I Anderson Sept. 6, 1955 FOREIGN PATENTS 143,105 Australia Aug. 29, 1951 OTHER REFERENCES The Snapping Dipoles of Ferroelectrics as a Memory Element for Digital Computers (Pulvari), Proceedings of the Western Computer Conference (Feb. 4-6, 1953) (page 152 and Fig. 16, page 159 relied upon, Fig. 14, page 158 of interest).

Images