US2797321A - Bias generating matrix - Google Patents

Description

June 25, 1957 .1.A. RAJCHMAN Erm. 2,797,321

BIAS GENERATING MATRIXi original Filed nec. 1, 1949 :s sheds-sheet 1v (Vb: i A Y (yf, @w21 gw@ [gr/min ma INvEN'roRS eanA.

June 25, 1957 J. A. RAJCHMAN E TAL 2,797,321

BIAS GENERATING MATRIX Original Filed Dec. l. 1949 3 'Sheets-Sheet 2 i :inventors JMA. Ilma,

Jne 25, 1957 J. A. RAJCHMAN ETAL BIAS GENERATING MATRIX Original Filed Dec. l, 1949 3 Sheets-Sheet 3 (lttorneg States BIAS GENERATLJG MATRlX Jan A. Rachman and Max H. Mesner, Princeton, N. l., and Milton Rosenberg, Santa Monica, Calif., assignors to Radio Corporation of America, a corporation of Delaware 4 Claims. (Cl. 25d-27) This invention relates to vacuum tube matrices and more particularly to vacuum tube matrices for generating bias voltages.

This application is a division of application Serial No. 130,412, filed December 1, 1949, now Patent No. 2,666,161, for a Bias Generating Matrix.

In an application for an Electronic Discharge Device by J an A. Rajchman, tiled September 30, 1949, Serial No. 118,758, now Patent No. 2,635,201, there is described and claimed a target area selection type of tube. This is an improvement over the similar type of tube which is described and claimed in a copending application for an Electronic Discharge Device, Serial No. 665,031, tiled April 26, 1946, by Jan A. Rajchman, which is now Patent No. 2,494,670. The type of tube described is a memory tube in which any discrete area of a storage target may be selected, as desired, for bombardment by electrons from the electron source for the purpose ot either storing information in that area or reading the information stored therein.

One of the novel features of both of these types of tubes is the selecting grid structure which provides a facile means for selecting any desired area `of the storage target for the purpose of being bombarded by electrons from the cathode. This grid structure essentially consists of a first network of parallel, spaced, separately insulated conductors and a second network of parallel, spaced, separately insulated conductors. The conductors of the first network are angularly disposed with reference to the conductors of the second network. Preferably, the angle made by the conduct-ors of the two networks is a right angle and the conductors are accordingly known as horizontal selecting wires and Vertical selecting wires. When viewed from the cathode looking toward the target, these horizontal and vertical selecting wires deine a plurality of Windows through which electrons from the cathode pass on the way to the storage target.

Whether or not electrons pass through any 'one window depends upon the bias existing on the four selecting bars defining that window. A detailed explanation of the theory involved is found in the application for an Electron Discharge Device, tiled April 26, 1946, by Jan A. Raich- Inan and bearing Serial No. 665,031, now Patent No. 2,494,670. In the application Serial No. 118,758, filed September 30, 1949, now Patent No. 2,635,201, for an Electronic Discharge Device, it is shown and explained that, when the four selecting wires defining a window are at cathode potential, electrons may pass through that Window, but, when any one of the four selecting wires is suiciently negative with reference to the cathode, then the window is closed to the passage of electrons. The target is constructed with storage areas opposite these windows. Through the application of a proper bias to the selecting wires, one window may be left open and all the others closed. It is thus possible to select a single storage area for electron bombardment.

One system for applying a proper bias to the selecting latented .inne 25, 1957 wires is to bring a lead from each one of the selecting wires external to the tube envelope and to apply a proper bias to each of these leads so that either a desired window is left open and the remainder are closed, or all the windows are `open depending upon the condition desired. It should be obvious that, with a tube having a large number of windows and therefore selecting wires, a cornplex and cumbersome arrangement is required-for applying a bias to these wires. However, systems have been developed for interconnecting the vertical selecting wires and the horizontal selecting wires in combinatorial combinations inside the tube envelope so that, although the number of leads required to be brought external to the tube is considerably reduced, complete control in selecting a desired Window is provided. Systems for interconnecting the selecting wires in a combinatorial combination are described and claimed in both the above noted application for an Electron Discharge Device and in an application Serial No. 694,041, filed August 30, 1946, by George W. Brown for Control of Electron Discharge Device of Area Selection Type, and now Patent No. 2,519,172.

Although the number of leads, brought external to a tube for applying a bias, is considerably reduced where the selecting wires are internally connected in a combinatorial combination, the problem of providing a system which is simple to construct and operate and which permits the application of bias voltages selectively to these external leads is still present. A number of systems, for applying bias to a selecting grid for the purpose of opening a desired window, are described and claimed in an application Serial No. 702,775 for Scanning Circuits for Area Selecting Tubes and the Like, by I an A. Rajchrnan, tiled October 11, 1946, now Patent No. 2,558,460, issued June 26, 1951. These systems, however, are `desirable for scanning the target of an area selecting tube in a desired sequence and do not readily lend themselves to a selection of a desired window without going through a sequence. These systems are also complex in operation and construction.

It is an object of the present invention to provide a bias voltage generating matrix which permits selective bias control.

It is a further object of the present invention to provide a bias voltage generating matrix which is simple to construct and operate.

These and further obje-cts of our invention are achieved by providing a bias generating tube for each biasing lead brought external to the tube from the combinatorial interconnection of the selecting wires in the tube. A plurality of first gate and second gate tubes are also provided which are equal in number to the number of bias generating tubes. Each of the rst gate and second gate tubes have a common plate load to which the control grid of an associated bias generating tube is coupled. The rst gate tubes are normally conducting and thus the bias generating tubes are normally cut olf and the bias lead coupled to its plate receives its most positive voltage. The second gate tubes are multi-grid tubes. Similar grids of groups of second gate-tubes are connected in parallel and then coupled to a plurality of multivibrators. In one of the groups of second gate tubes, each tube also shares a common cathode bias resistor with an associated single gate tube. The conductive condition of these single gate tubes is determined by a multivibrator to which the single gate tubes control grids are coupled.

The multivibrators used herein are of the two stable condition type. By the judicious application of a signal to each of the multivibrators, each multivibrator is placed in one of its two conditions of stability thus in turn applying signals to the associated second gate tubes to cause 'desired ones of the second gate tubes to be in a conducting condition and the others of the second gate tubes to be in a non-conducting condition. A signal may then be applied to all the first gate tubes to simultaneously render them non-conducting. In View of the common plate loads for the first gate and second gate tubes, when all the first gate tubes are rendered non-conducting, a positive pulse is app'lied to the control grids of all the bias generating tubes except those associated with the second gate tubes which are selected to be maintained in a conducting condition. The positive pulse applied to the bias generating tubes causes them to draw current and their anodes accordingly go negative and bias the associated bias leads negative. The bias generating tubes which are associated with the conducting second gate tubes are unaffected and maintain a positive bias on the associated bias leads. The bias leads which are maintained at the positive bias are connected to the selected wires which define the window which is desired to be kept open to the passage of electrons from the cathode to the target. The remaining windows are closed by virtue of the negative bias applied to the remaining bias leads.

The novel features of the invention, both as to its organization and method of operation will best be understood from the following description, when read in connection with the accompanying drawings, in which,

Figure 1 is a diametral section of an electronic discharge device, an understanding of which is necessary for an understanding of our invention,

Figure 2 is a schematic diagram of a system of interconnection of the vertical and horizontal selecting wires and their connection to the external bias leads, an understanding of which is necessary for an understanding of our invention,

Figure 3 is a circuit diagram of a selective bias generating matrix constituting an embodiment of our invention, and

Figure 4 is a diagram showing all possible combinations of signals supplied by multivibrators in the matrix and the leads which are biased positive as a result of each signal combination.

The diametral section of the target area selecting type of tube, shown in Figure 1, as well als the schematic diagram of a combinatorial interconnection of its vertical and horizontal selecting wires are both shown, described and claimed in the copending application of I an A. Rajchman, Serial No. 118,758, now Patent No. 2,635,201 filed September 30, 1949. They are both reproduced and explained here to facilitate the explanation of the selective bias generating matrix which constiutes an embodiment of our invention.

Considering Figure 1, it may be seen that the tube is in a glass envelope and is constructed symmetrically about a plane formed by the cathodes 12. The catho'des 12 are preferably of a rectangular cross section. The cathodes 12 are alternate with, between and parallel to a set of nine vertical selecting bars or wires 14 of square cross section. They are also substantially co-extensive with the vertical selecting wires 14. On either side of the plane made by the cathodes 12 and the vertical selecting bars 14 is a set of 18 parallel horizontal selecting bars 16 of square cross section. These two sets of horizontal selecting bars are parallel to, and sandwich the cathodes and vertical bars, as do all the subsequent electrodes of the tube.

It will be readily appreciated that, when viewed perpendicularly to the parallel planes formed by the vertical and horizontal selecting bars, a grid mesh is seen having square openings or windows in which the horizontal sides are defined by two adjacent horizontal selecting bars and the vertical sides are defined by two adjacent vertical selecting bans. These windows are perpendicular to the path of the electrons from the electron source to the target and electrons -may pass through them.

Spaced on either outer side of the horizontal selecting Vbars 16, and parallel to the plane thereof, is positioned a plate.

first target assembly 24. This first target assembly consistis of a collector electrode 18, a storage target and a writing electrode 32. The collector electrode 18 is made of two fiat metal plates 20, 22 perforated with round holes, the centers of which are aligned with the centers of the windows formed by the vertical and horizontal selecting bars. The iirst plate 20, which is nearest the horizontal selecting bars, is known as the collector mask and has the lsmaller holes. The second plate, or collector spacer 22, is in intimate contact with the collector mask and has the larger holes.

On the outer side of each of the collector electrodes 18 is positioned the storage target assembly. This consists of two perforated sheets 26, 28 of an insulating materiail, such as mica, holding between them, by means of the perforations, metallic eyelets 30. Next cornes another metallic plate with aligned perforations which is known as the writing plate 32. The eyelets 30 are generally cylindrical and have shoulder offset portions to be insulatingly retained thereby by the perforated mica sheets. The perforations in the insulating sheets 26, 28 are so spaced as to position the eyelet openings opposite the center of the respective grid windows. An eyelet comprises a conical head, a center hole, a collar and a tail. The writing plate 32 is separated from the eyelets 30 by the insulating material sheet 28 and serves as a common capacity plate for all the eyelets 30 with which it is associated. The two collector plates 20, 22, the two insulating sheets 26, 28 supporting the eyelets 30 and the writing plate 32 form a tight assembly which is riveted together at the upper and lower ends and in the center. 'This target assembly 24 is more fully shown and described in Patent No. 2,604,606, issued July 22, 1952, entitled Target for Storage Tubes, to Ian A. Rajchman.

On the outer side of either target assembly 24 and spaced therefrom is a second target assembly 25 consisting of a reading plate 34, which is another metallic plate, having perforations substantially aligned with the centers of the windows formed by the horizontal and vertical selecting bars.

Beyond each of the reading plates is a Faraday cage 36. This comprises a rectangular metallic box in which two walls are parallel to the reading plate and have perforations aligned with the reading plate perforations, A glass plate 38 coated with a fluorescent and secondary electron emitting material 40, such as willemite, is placed against the outer perforated wall of the Faraday cage. In the central plate of the cage there are nine reading wires 42' which are positioned so that they are between the perforations in the perforated walls and are thus shielded from any electrons which may be coming directly from the target. The reading wires are also substantially shielded from electrostatic field leakage from the reading These reading wires are connected together and the corresponding lead to the stem of the tube is shielded. This second target assembly 25 is also more fully shown and described in the above identified Patent No. 2,604,606, entitled Target for Storage Tubes.

In Patent No. 2,494,670 for Electronic Discharge Devices, there has been explained at length the method by which the selection of an area of a target is made by applying the proper bias to the selecting bars defining the window which opposes the target area selected. Reference should be made thereto for detailed consideration of the subject. Briefly, however, it is explained therein that,

when all four selecting bars defining a window are at cathode potential or higher, electrons can pass through that window. Should any one of the selecting bars defining the window be at a potential which is sufficiently lower than cathode potential, then electrons do not pass through that window but are deflected therefrom. This assumes that there is a sufiicient accelerating potential exerted on the electrons to enable them to pass through the window when the defining selecting bars are at cathode potential. In order to secure the required accelerating amasar potential in the area selecting type of tube described in the previous above noted application, in one form, the selecting bars are all positively biased to permit the passage of electrons therethrough, and have their bias lowered to cathode potential or slightly negative to block electron passage. In another form, positively biased, accelerating electrodes are interposed both between the horizontal and Vertical selecting bars and between the cathode and the selecting bars and the selecting bars are then either biased to cathode potential to permit passage therethrough of electrons or are biased highly negative to block such passage. The provision of a sufficient accelerating potential is thus provided either by the selecting bars themselves or the accelerating electrodes.

In the tube described above, because of its structure, the collector plate 18, which is positively biased, acts to provide the required accelerating potential, In order to permit passage of electrons from the cathode' through a window, the selecting bars defining vthat window are left at cathode potential. The biasing of any of the selecting bars dening a window suiciently negative with respect to the cathode prevents further passage of electrons through that window. Since the two potential values applied to the selecting bars are either cathode potential or a negative potential with respect to the cathode, it will be appreciatexl that the power requirements for the selecting bars is minimal, since these bars never draw any current. Furthermore, the cathode potential is easily attained with accuracy in external circuits, while the negative voltage is not critical.

For biasing purposes, each of the vertical 14 and horizontal selecting bars 16 may be individually insulated and brought out through the envelope of the tube and separately biased so that the window opposing the desired target area is opened. Methods for effecting complete `control of the windows defined by the vertical and hon'- zontal selecting bars utilizing a number of external leads which is less than the number of horizontal `and vertical selecting bars have been described and claimed in an application, Serial No. 702,775, filed October 11, 1946, by I an A, Rajchman, now Patent No. 2,558,460, and in the application of George W. Brown, Serial No. 694,041, filed August 30, 1946, now Patent No. 2,519,172 and -assigned to this assignee.

The principle of the combinatorial connections, by means of which the number of external leads can be greatly reduced, is the fact that, the electron current through a gate, formed by two metal bars, can be controlled by either bar. In the case of the window, any of the four defining bars can block the electron current. ln the present system the 'stoppage of current is `actually `affected by suppressing almost totally the emission from the particular area of the cathode corresponding to a Window by biasing any one of the horizontal or Vertical bars forming it. The small remaining part, perhaps l percent, is so badly deflected off the direction of the axis of the hole that it strikes one face of the collector electrode 18 without reaching the eyelet 30.

Referring to Figure 2, wherein is shown a preferred system of connection of the selecting bars, the nine vertical selecting bars 14 are connected to six separate leads which are brought external to the tube. These' leads are in two groups and are designated as V1, V2, V3, V4 and V1 and V2. The thirty-six horizontal selecting bars are connected to twelve separate leads which are brought external to the tube. These leads are also in .two groups and are designated H1, H2, Hs, H4, and H1', Hz', H3', H4', H5', Hs', H7', and Hs. The nine vertical selecting bars '14 are employed to operate as eight gates since there ,are only eight combinations of V1, V2, V3, V4 and V1 and Vz, when any one of the Vs and any one of the Vs is taken (4X2). The 36 horizontal selecting bars are employed to operate as 32 gates since there are only thirty-two combinationsof H1, H2, H3, H4 and H1' through Ha', when one of the Hs and one of the Hs is taken (4X8). The excess number of bars are used to-take care of the end elects. The eight vertical gates and 32 horizontal gates separately control 256 windows. For operation of the tube as a two channel device leads H1 and H5', H2' and Hs', H3 and H7', and H4 and Hs' should be connected together.

Because of the positioning of the vertical selecting bars adjacent each cathode and also the horizontal selecting bars adjacent the cathode, all these bars being at cathode potential, and because of the positive collector plate, with its holes in register with the windows formed by the selecting bars, an almost perfect electron optical system is formed, Emission from the cathode is sharply focussed through the collector hole. No current lgoes to the vertical and horizontal selecting bars because of this focussing and because they are at cathode potential. Furthermore, because of the sharp focussing action, Veryfew electrons strike the collector plate but-most of them are directed through the'perforations and atthe storing eyelets 30.

In the quiescent state of the tube, the vertical and horizontal selecting bars are all at cathode potential and the collector plate 18 is positively biased with reference thereto. The electrons emitted by the cathodes will therefore be focussed into 256 beams by the combined action of the Vertical andhorizoutal selecting bars which form 256 windows. These 256 beams are focussed through the center of the collector holes and are directed at the heads of the eyelets 30.

The act of writing or reading requires the selection of one eyelet 30 or target element (two eyelets or target elements if the two halves ofthe tubes are run in parallel). This selectionis obtained by applying a negative pulse to all the selecting leads except to the one in each of the four groups V, V, H and H which connect to the selecting bars defining the window associated with the desired eyelet. These leads are left at cathode potential.

For positive writing, a window to a desired eyelet is left open and the remaining windows are closed. A highly positive pulse is then applied to the writing plate and is then allowed to slowly subside.

This pulse should have a sharp rise time, this rise time being sufficiently short in order to overcome the locking electron current to the eyelet. The amplitude of the pulse should be sufficient to drive the eyelet potential above the collector potential. In other words, the displacement current due to the capacity plate pulse should be greater than the locking electron current to the eyelet from the cathodes. During the plateau time or" the pulse the real current to the eyelet will then charge the eyelet down to anode voltage. r)The writing plate pulse then is permitted to decay slowly. The real current to the eyelet during the decay period is now greater than the displacement current and the eyelet as a result is locked at collector potential. The eyelet then remains at a positive or approximately collector potential.

If it is desired to write negatively on an eyelet, the above outlined procedure is repeated except that before the pulse applied to the writing plate is allowed to` decay, the window to the eyelet is closed by applying a negative pulse to one or more of the leads in the groups V, V', H, H', which, during selection were left at cathode potential to keep open the window associated with the selected eyelet. After the end of the writing pulse, all other pulses are ended and current is re-established to all the eyelets again.

The reading or interrogating of the tube is done one eyelet at a time (or two if the tube is used as a two channel device). First, a selecting pulse is applied to all leads to the selecting bars except the one in each of the groups V, V', H and H' which is connected to the selecting bars defining the window associated with the eyelet desired to be read. Some arbitrary short safety period thereafter, a positive reading pulse is applied to the reading plate which was previously negatively biased. A pulse of electron current flows to the reading wires as a ,result Yof this reading pulse if the selected eyelet isat ,collector potentialv but no electron current Viiows ifthe ing plate holes, through the Faraday cage 34 until they strike the uorescent screen 40. The area of the screen defined by the holes in the Faraday cage 36 fluoresces and secondary electrons are emitted and are attracted to the reading wires 42. Thus a visual, as well as an electrical indication, is given as to whether an eyelet is at collector t or cathode potential. The reading has no effect on the eyelet potential.

Figure 3 is a circuit diagram of the bias generating matrix by means of which bias voltages for proper operation of the selecting grid are secured. In the combinatorial connections shown in Figure 2 it may be seen that there are four groups of bias leads brought external to the tube. A first group has eight leads designated as H1', Hz',-H3', H4', Hs', H6', H7' and Hs', a second group has four leads designated as H1, H2, H3 and H4, a third group has four leads designated as V1, V2, Va, and V4, and a last group has two leads designated as V1 and V2'. One lead in each of these four groups must be at the most positive potential (in this instancecathode potential) when the other leads are'at a negative potential in order that a single Vwindow be open. Y

A bias generating tube 50 is provided for each of the bias leads in the four groups H', H, V, V. Each of the bias generating tubes 50 has its individual load resistor .52 connected to its anode 54. The leads are connected to the anodes 54 of the bias generating tubes 50 and are shown as stub leads having the same designations as the leads in Figure 2. It is to be understood that the leads of FigureV 2 and Figure 3, which are similarly designated, are connected together. A first gate tube 60 and a second gate tube 70 are provided for each bias generating tube 50. The first gate tubes 60 and the second gate tubes 70 each have a common load resistor 62 connected to their anodes 64, 72. The cathode 56 of each bias generating tube is connected to the B| supply for the first and second gate tubes. The control grid 58 of each of the bias generating tubes is connected to the common anode connection of the first and second gate tubes to derive a control signal therefrom.

All the first gate tubes 60 are normally in a conducting condition. This results in a more negative signal being applied to the control grids 58 of the bias generating tubes 50. As a result, the bias generating tubes are nonconducting and the biasA applied to all the bias leads is the more positive one. Therefore, since all the selecting wires connected to the bias leads are at their most positive condition, all the windows formed by the selecting wires are open. All the cathodes 66 of all first gate tubes 60 are connected to ground. The control grids 68 of the first gate tubes associated with the first bias lead group H are all connected to a first, two-position switch 61 which normally is in the position where it is connected to a low impedance negative pulse source 69. The control grids 68 of the first gate tubes 60 associated with the second bias lead group H, the third bias lead group V and the fourth bias lead group V are also respectively connected to second, third and fourth two-position switches 63, 65, 67 which are also normally connected to the low impedance negative pulse source 69. The second position of all the switches 61, 63, 65, 67 connects all the control grids to ground to isolate them from the effect of negative pulses from the source 69. Upon the arrival of a negative pulse from the negative pulse source 69 all the first gate tubes, in any. group whose switch Vis thrown to its first position are. rendered non-conducting.

All of the second gate tubes 70 are multi-grid tubes.

All their screen grids are connected together and to the B| source through a resistor 77. Considering first, the Veight second gate tubes 70 associated with the first group of bias generating tubes andbias leads, the suppressor grids 74of alternate second gate tubes 70 are connected together. The two sets of four suppressor grids 74 connected together are then respectively coupled to the two grids 92, 94 of a first multivibrator 90. The control grids 76 of alternate second gate tubes are also connected together and the two sets of four control grids 76 are respectively coupled to the two grids 102, 104 of a second multivibrator 100. The reason for connecting to the grids instead' of the anodes of the multivibrators is that the grids are at the proper D. C. level.

A single gate tube 80 is provided for each second gate tube 70 in the first group. Each of the second gate and associated single gate tubes have their cathodes 78, 82 connected together and have a common cathoderbias resistor 84. Therefore, in addition to the signals applied to the suppressor 74 and control grids 76 of the second gate tubes 70, another signal is applied to the cathodes 78 by means of the single gate tubes 80,

In order for any second gate tube 70 in the first group H to be rendered conducting, a positive signal must be applied to its suppressor and control grids 74, 76 and no positive signal should be applied to its cathode 70. A positive cathode signal is applied when the single gate tube is conducting. If a single gate tube is non-conducting, then the conductivity of the second gate tube is determined by the signals on its grids. The control grids 86 of the rst, last and middle two of the single gate tubes 80 are connected together. The control grids of the remaining single gate tubes 80 `are also connected together. These two sets of interconnected control grids are then coupled to the two grids 108, `of a third multivibrator 106 through a pair of D.C. coupled butter amplifiers 96, 98. The anodes 80 of the single gate tubes 80 `are connected to the same B-fsource as are the first and second gate tubes. The D.C. coupled buffer amplifiers 96, 98 are required to raise the D.C. level of the signal applied to the single gate tubes in order that the positive signal, which is applied to their grids, have sufficient amplitude to maintain the single gate tubes conducting and thus bias off the second gate tubes 70 regardless of any other signals which may be applied to the second gate tubes control grids.

For the second bias lead group H only four bias gen,- erating tubes 50 and associated four first gate 60 and four second gate tubes 70 are required. The connections between these tubes are the same as the connections in the first group H. It will be noted, however, that no single gate tubes are provided for the second gate tubes. Instead the cathodes 78 of the second gate tubes are returned to ground. The control grids 76 of alternate second gate tubes are connected together and then connected to the grids 114, 116 of a fourth multivibrator 112. The suppressor grids 74 in the alternate second gate tubes 70 of the second group H are likewise connected together and coupled to the grids 120, 122 of a fifth multivibrator 118.

The third bias lead group V also contains four bias generating tubes 50, four first gate 60 and four second gate tubes 70 which are interconnected in a fashion similar to the interconnections of the second bias lead group H. A sixth 124 and seventh multivibrator 130 are provided which are coupled to the suppressor and control grids. However, the control grids 76 are coupled to the sixth multivibrator 124 through buffer amplifiers 142 paralleled by triodes 146 for `reasons which are subsequently explained herein.

The fourth bias lead group V' requires only two bias generating tubes 50 with two associated first 60 and second gate tubes 7 0. The suppressor grids 74 of the second gate tubes 70 are connected to ground through a resistor 71 and also to a negative pulse source 73 for reasons which will be subsequently explained herein. Each con- 9 trol grid 76 is coupled to oneof the grids 138, 140 of an eighth multivibrator 136. i

Each one of the eight multivibrators are of the type, well known to the art, having two tubes `including cross connected grids and Yanodes, two conditions of stability, and are capable of being triggered from one condition oi stability to the iother condition of stability by pulses of the proper polarity applied to the control grids. The theory of operation of the multivibrator is well known and may be found explained in a book, Theory and Applications of Electron Tubes by H. J. Reich, published by McGraw-HillV Book Company, pages '362-365. A typical multivibrator circuit is shown for the lirst multivibrator; the remaining seven multivibrators are represented vestigially in dotted line boxes.

From'the foregoing explanation it will be appreciated that, when all the Vfirst gate tubes 60 are conducting, current is drawn through all their common load resistors 62 'thus causing the bias generating tubes S0 to apply a more positive bias to the bias leads. When the 'rst gate tubes 60 are rendered non-conducting by throwing the fourswitohes 61, 63, 65, 67 to the second position, the bias voltages applied to the leads by the various bias generating tubes 50 are determined by the conducting or non-conducting condition of the second gate tubes 70. If a second gate tube is in a conducting condition, current is still dnawn through the common load resistor 62 when the associated rst gate tube is rendered non-conducting and the associated bias generating tube will still apply a more positive voltage to its bias lead. However, should a second gate tube 70 be in a non-conducting condition, when a first gate tube 60 is rendered non-conducting, the anodes of both of these tubes rise up to the value of their B+ supply. Therefore, la positive pulse is applied to the bias generating tube grid 68 causing the tube to draw current and thus apply a negative pulse to the associated bias lead and the selecting bars connected thereto. This lin turn closes the windows controlled by these selecting bars to the passage of electrons.

The condition `of conduction of each of the second gate tubes is determined by the signals impressed on these tubes by the eight mul-tivibnators. In an actual electronic computing machine, the setting of these eight multivibrators is determined by a coded set of voltages, which is obtained from the coded input circuit S of the computer and is impressed upon the grids of the multivibrators. The signal from the coding circuit is a short sharp pulse whose duration is just long enough to trip the multivibrator and not alt'ect the signal on the multivibrator grid from the anode to which the grid is cross connected.

Referring to Figure 4, there is illustrated the selection effect of the signals derived from the grids of the eight multivibrators. Four rectangles are shown, one for each group H', H, V, V. ln each `rectangle are two vertical columns representing the grid signals supplied by each multivibrator in the group. The numeral at the heading of each vertical column corresponds to the identifying numeral of the multivibrator grid, the signal being impressed on the associated second gate tube by the multivibrator grid is shown thereunder. At the right side oi the horizontal column is shown, for the signal combination shown, the bias lead i-n each group which has a positive bias applied to it. All other leads of the group have a negative bias applied. All possible combinations of signals applied by the multivibrators to the selecting bars are shown. By applying the proper coded signals, a single bias lead in each gro-up is selected to be maintained positive and all others negative. Accordingly, a single selected window, detined by the positive (cathode potential) selecting bars attached to these bias leads, is open to the passage of electrons while the remaining windows are closed.

For writing negatively on any target storage eyelet of the tube, it was previously explained that only a single window opposite the selected eyelet is kept open, a positive pulse is applied to the writing plate and before this pulse can decay, the window is quickly closed to the flow of current. Some time after the positive -pulse has decayed, all windowsare again opened. Several methods for closing the window are available and some of these are shown in Figure 3. One method is to couple a source, which provides a negative or cutoff pulse when required, to the suppressor grids of any group of second gate tubes. This is shown for the fourth group in Figure 3. The negative pulsek from the source '73 causes any connected conducting second gate tube 70 to be cut off. The resulting negative bias voltage generated by the associated bias generating tube is applied to one of the selecting wires which is also one of the wires defining an open Window. The window is therefore closed.

Other apparatus, by means of which an open window may be closed, is shown in Figure 3 as serving to couple the control grids of the third group of second gate tubes to the sixth multivibrator 124. This apparatus consists of a pair of D.C. coupled ampliiiers 142, each having an associated triode tube 146 whose anode 14S and cathode 150 Vare coupled to the associated D.C. amplilier anode 152 and cathode 154. The grids 156 of the triode tubes 146 are connected together and through an Yisolating resistor to a negative bias source so Athat they are normally not conducting. One of the D.C. coupled ampliers 142 is conducting and the other is not conducting depending on the signal applied from the multivibrator grids. A positive pulse is applied from a positive pulse source at the proper time to the grids of the triode tubes. This causes them to conduct and draw current through the load resistors common to both triodes and D.C. ampliers. The voltage drop, caused thereby at the anode of the D.C. amplifier, causes a negative pulse to be applied to the control grid of Whichever second gate tube has been conducting causing it to cease conducting. The end result is that the associated bias generating tube applies a negative bias to one of the selecting wires deining an open window, thus closing the Window.

The B+ supply for the bias generating tubes may have its positive side connected to the cathode of the area selecting tube and its negative side below cathode potential so that the more positive bias applied to the leads when the bias generating tubes are not conducting is at cathode potential and the less positive bias applied when the bias generating tubes are conducting, is well below cathode potential.

There has been shown and described herein a bias generating matrix wherein the response of the bias generating tubes to the change in the conductive condition of rst gate tubes is determined by the conductive condition of second gate tubes which in turn is determined by a coded signal input.

From the foregoing description it will be readily apparent that a selective bias generating matrix has been described which readily permits facile application of bias to a selecting type of grid to permit the opening and closing of a desired window, defined by any four of the selecting wires, to the passage of electrons. Although a single embodiment of the bias generating matrix has been herein shown and described in connection with one type of selecting grid having four groups of interconnecting bias leads, it should be apparent that the bias generating matrix may be altered to accommodate other selecting grid combinatorial interconnections or serve as an apparatus for providing signals in response to coded input signals. It should therefore be apparent that many changes may be made in the particular embodiment herein disclosed and that many other embodiments are possible, all within the spirit and scope of the invention. Therefore, it is desired that the foregoing description shall be taken as illustrative and not as limiting.

What is claimed is: Y v

1. A matrix for the selectiveV generation ofrbias, voltages comprising a plurality of rst gate' tubes, Aa plurality of second gate tubes, each of said rst gate tubes together with separate ones of said second gate tubes having a common load impedance, and a plurality of bias generating vacuum tubes having at least a cathode,'an anode and a control grid, the control grid of each of said bias generating tubes being coupled to a separate one of said load impedances to derive therefrom a bias voltage control signal, and means to control said second gate tubes to cause selected ones of said second gate tubes to assume one condition of conductivity and simultaneously to cause the remainder of said second gate tubes Vto assume a second condition of conductivity, thereby selectively to determine the bias voltage generated by each of said bias generating tubes.

2. A matrix for the selective generation of bias voltages comprising a plurality of normally conducting rst gate tubes, a plurality of second gate tubes, each of said first gate tubes together with separate ones of said second gate tubes having a common load impedance, a plurality of bias generating tubes, each of said bias generating tubes having at least a cathode, a control grid and an anode, the control grid of each of said bias generating tubes being coupled to a separate one of said load impedances to derive a bias control signal therefrom, means to render simultaneously non-conducting said plurality of normally conducting first gate tubes, and means to render desired ones of said second gate tubes conducting and the remaining ones of said second gate tubes nonand said generated bias voltages to diter accordingly.

j 3. A matrix for the selective generation of bias voltages as recited in claim 2 wherein each of said second gate tubes has a cathode, an anode and a plurality of control grids and said means to render desired ones of said second gate tubes conducting and the remaining ones of said second gate tubes non-conducting includes a plurality of multivibrators, each of the control grids in each one of said second gate tubes being coupled to a separate multivibrator.

4. A matrix as recited in claim 2 wherein said means to render desired ones of said second gate tubes conducting and the remaining ones non-conducting includes in addition several vacuum tubes each having at least cathode, control grid and anode electrodes, each also having a common cathode bias resistor with a corresponding one of said second gate tubes, the control grids of one half of said several vacuum tubes being connected in parallel and t0 one output from one of said plurality of multivibrators, the remaining half of said several vacnum tubes being connected in parallel and to the other output of said one of said multivibrators.

References Cited in the tile of this patent UNITED STATES PATENTS 2,541,932 Melhose Feb. 13, 1951 2,559,499 Gillette et al. July 3, 1951 2,611,813 Sharpless et al. Sept. 23, 1952 2,615,127 Edwards Oct. 2l, 1952 2,666,848 Goodwin Ian. 19, 1954

Images