US3106700A - Photographic storage system - Google Patents

Description

Oct. 8, 1963 s. P. NEWBERRY PHOTOGRAPHIC STORAGE SYSTEM 5 Sheets-Sheet yl Filed June 27, 1957 INM.

w @y w e r www e. ,5 www fw ww,

OCL 8, 1963 s. nfNEwBERRY PHoToGRAPHIc STORAGE SYSTEM 5 Sheets-Sheet 2 Filed June 27, 1957 Oct. 8, 1963 sQP. NEWBERRY PHOTOGRAPHIC STORAGE SYSTEM 5 Sheets-Sheet 3 Filed June 27, 1957 Oct. 8, 1963 s. P. NEwBr-:RRY

PHoToGRAPmc STORAGE SYSTEM 5 Sheets-Sheet 4 Filed June 27, 1957 f W q fw n 0, ig vd n his W Q lfm d v n e IIIIIII J f g n u lv 21mm IL NRW I. I @I Q11 N MII Wb llllnlllll I III Il' Il! Ivi. .M h .Il dd. w ml f m 5 #V United States Patent O 3,106,700 PHOTOGRAPHIC STORAGE SYSTEM Sterling I. Newberry, Schenectady, N.Y., assignor to General Electric Company, a corporation of New York Filed June 2.7, 1957, Ser. No. 668,490 7 Claims. (Cl. 340-173) This invention relates to a method and apparatus for data storage and, more particularly, a photographic memory system of large capacity, high storage density, high reliability, and rapid access.

There is an urgent and growing need for a memory having very large capacity with rapid access to any part of the memory at high reliability of readout. This is especially important lfor maintaining necessary flexibility in automated production lines, industrial record files, computer applications, as well as the numerous other areas where data storage is desired. The use of a photographic emulsion as the recording medium presents an extremely promising approach since modern emulsions of very high quality and high resolving power are available. In the past, however, the storage densities in systems utilizing photographic emulsions were so low as to make them essentially useless for high capacity and fast access stonage systems.

The limits on storage density in the past have not been from the resolving power of the photographic emulsion but from artifacts caused by dust :and scratches and `from difficulties with tracking errors and data relocation. Because of the low storage density, storage elements of substantial size are necessary in order to store an adequate amount of data and, as a consequence, special handling to eliminate eifects of dust particles, such as hermetic sealing or total oil immersion, is impractical. Furthermore, because of the size of the elements substantial relative movement between the storing and reading heads and the storage element are necessary to store and have access to all of the data. This results in scratches and other marring effects on the film surface further reducing the resolution. In addition, low storage density and large storage elements demand exact mechanical location of blocks of data during readout which requires a formidable level of mechanical precision and, consequently, complex and expensive equipment.

By utilizing an electro-optical system including a reduced and focussed beam of light generated by a stream of electrons striking the fluorescent surface of a cathode ray device, it is possible to produce a storage density of approximately 106 storage sites per square centimeter which is greater by a factor of l()3 than the photographic storage sys-tems hitherto known. With such high storage densities the problem of dust land scratches is eliminated by hermetically sealing the small area of photographic film required for large capacity storage. The problem of registering during readout also is immensely simplified by virtue of the fact that it becomes possible to utilize a rough mechanical register to bring the desired area of stored data into position and then electrically shifting the scanning center of an image tube which reads the data from the magnified image of the data until exact centering and registration of the desired block of data is achieved.

It is an object of this invention, therefore, to provide a photographic data storage system of extremely large capacity :and high storage density.

A further object of this invention is to provide a photographic data storage system which is not limited in reso'- lution by dust particles and scratches.

It is another object of this invention to provide a photographic data storage system having a storage element of sufficiently small dimension to permit sealed handling thereof.

Yet another object of this invention is to provide a photographic data storage system which permits rapid access to any portion thereof by simple mechanical register and readout of information in the selected portion by an electro-optical system.

Yet another object of this invention is to provide a photographic data storage system wherein a simple mechanioal movement provides a rough register and shifting of the scanning pattern of an image device provides the precise registration.

Other objects and purposes of the invention will become apparent as the description thereof proceeds.

In accordance with the invention the foregoing objects are achieved by utilizing -a spot of light from a kinescope flying spot device, reducing Ithe size of the light spot by means of an optical device and focussing it on a photographic element in a predetermined sequence to produce data matrices in binary `form as black and white dots. By virtue of this technique, storage density of the order of 106 storage sites per square centimeter are achieved.

During readout the high capacity and high storage density photographic data element is utilized under hermetically sealed conditions eliminating problems of dust and scratches :and the data stored thereon is projected `onto the face of the image tube device and readout thereby. The image tube sean is shifted electrically in response to -data stored on the photographic element to relocate the information in the image plane so that only approximate mechanical registration is necessary.

The novel features which are believed to be characteristic of this invention are set forth with particularity in the appended claims. The invention itself, however, both as to its organization and method of operation together with further objects and advantages thereof may best be understood by reference to the following description taken in connection with the accompanying drawing in which:

yFIGURE 1 represents an embodiment of an apparatus for storing information on a photographic element;

FIGURE 2 represents la detailed showing of the scanning martrix and current adders of FIGURE l;

FIGURE 3 represents an embodiment of a readout apparatus for the storage system;

FIGURE 4 is a detailed showing of the scanning matrix :and the means for shifting .the image tube can; and

FIGURE 5 is a diagrammatic illustration of the relative position of the stored information and the image tube scan before relocation.

ln the embodiment of the invention, as illustrated in FIGURE 1, a photographic storage element 1 consisting of a high resolution emulsion 2 mounted yon a transparent glass backing plate to provide dimensional stability is rigidly fixed to a metal frame element illustrated at 3 and is henmetically sealed in a casing 4 to eliminate dust particles and scratches on the photographic ernulsion. A rack and pinion element 5, shown diagrammatically, is provided for positioning the storage element 1 both in the horizon-tal and vertical directions, and may be manually operated or driven by a servo-mechanism device. A kinescope 6 on the iiuorescent screen of which is produced a raster utilized as a spot light source is positioned at the conjugate focal position of a projection light microscope 7 which reduces and focuses the light spot from the kinescope onto the photographic emulsion to produce discrete 10 micron diameter light and dark spots, in the form of a matrix, which represent the da-ta to be stored. The individual bits of data represented respectively by the light and dark spots are placed on approximately 10 micron centers in the form of a 32 x 32 binary matrix, covering an area of approximately one-tenth of a square millimeter.

The kinescope 6 has a fluorescent viewing screen 8 and a source of electrons 9 consisting of an electron emitting cathode and an accelerating grid. A focusing coil 11 and an alignment coil 12 are positioned to control the trajectory of the electron beam in the kinescope tuoe in a well gnown manner to focus and align it with res ect to the optical axis of the tube. Vertical and horizontal deflecting coils 13 and 14 constitute the means by which the electron beam is swept `across the face of the fluorescent viewing screen S both in the horizontal and vertical directions to produce the flying spot source of light. A power supply, indicated `generally at 15, provides energization for the cathode of the kinescope element 6 as well as for the focussing and alignment coils 11 and 12.

The horizontal and vertical deflection coils 13 and 14 are connected to a beam deflection coil supply circuit 2f) which produces a staircase deflection current for deflecting the electron beam as a step scan to produce the flying spot of light. The precise construction and manner of operation of the beam deflection supply circuit will be explained in greater detail later with reference to FIGURE 2.

The projection light microscope 7 which reduces the flying light spot on the face of the kinescope 6 and focuses it onto the photographic storage element comprises a color-corrected achromatic amplifying lens systern 16 positioned adjacent to the fluorescent screen 8 of the kinescope, which lens assembly comprises a projection lens element, field flattening lens element, and a field lens. Positioned at the opposite end of the microscope and adjacent to the field photographic element 1 is a color-corrected achromatic objective lens system 18 which, in conjunction with the amplifying `lens 16, reduces the light spot produced by the kinescope and focuses it onto a photographic element 1. A focus control assembly 17 positioned intermediate the lenses 16 and 18 permits manual or, if desired, automatic adjustment of the focus of the `microscope 7.

In order to provide step scanning of the electron beam to produce the flying spot of light, it is necessary that the beam deflection coil currents for the kinescope 6 be in the form of a staircase. This is accomplished by means of the deflection circuit 20 comprising a scanning matrix 21 comprising horizontal and vertical scaling circuits 22 and 23, as seen most clearly in FIGURE 2, and the horizontal and vertical current adder circuits 24 and 25. FIGURE 2 illustrates partially in block diagram form both the horizontal and vertical scalers and current added circuits wherein the vertical scaling circuit 23 is fed from the output of the horizontal Scaler 22.

The horizontal scaling circuit 22 has a pulse input terminal 27 and comprises :a number of bi-stable multivibrator circuits 26 arranged as a binary scaler circuit. Each of the bi-stable multivibrators, one of which is shown in detail, controls one stage of the current adder 24 consisting of a constant current generator circuit '29, such as constant current pentode, for example.

Each of the bi-stable circuits 26 consists of multivibrator connected tube pairs V3 and V4 and a pair of trigger tubes V1 and V2 having a single common input and which are normally non-conducting by virtue of the negative biasing voltage applied to their grids. Whenever a positive pulse arrives on the input, V4 and V2 conduct transmitting a negative pulse to the tubes V3 and V4. Whichever of the tubes is non-conducting is, of course, not affected while the conducting tube is cut off by the negative pulse which, by virtue of the multivibratory connection, reverses the conducting states of the two tubes. The `hi-stable circuit remains in this new state until the arrival of the next pulse which causes it to reverse its state. The output from each lai-stable circuitry 26 drives the next succeeding one by coupling the output pulse from tube V4 to the input lead of the trigger tubes V1 and V2 of the succeeding bi-stable circuit 26.

At the initiation of horizontal Scaler operation each of the V3 tubes are conducting and tubes V4 are non-v conducting and the application of pulses to the terminal 27 of the scaler starts operation thereof. It is clear that each time a tube V4 in a bi-stable circuit changes from a conducting to a non-conducting condition, a positive pulse is introduced to the input of the next bi-stable circuit in the series causing it to change its state. As is well known, scalers of this sort are known `as 2n types and require n bi-stable circuits to scale 2n pulses. For a more detailed description of such scaling circuits reference is made to High Speed Computing Devices, Engineering Research Associates, McGraw-Hill Book Company, lnc., New York (1950), and particularly chapter 3 which contains an excellent discussion of the principles and operation of such scaling circuits.

The output from the last horizontal bi-stable multivibrator is connected to the first vertical bi-stable multivibrator 28 through a lead 3f) to shift the scan by one line at the end of each horizontal scan. That is, at the end of a horizontal line scan when all of the horizontal bistable multivibrators 26 are on, the next clock pulse resets the horizontal Scaler to zero and the output pulse is fed to the first bi-stable multivibrator element 28 of the vertical Scaler 23 and shifts the scan by one line.

Each of the bi-stable devices 26, as has been stated previously, controls one of the constant current generators 29 of the respective current adders 24 and 25 by virtue of triggering pulses aplied thereto by a lead 31 from the anodes of the tubes V3. Thus each time a tube V3 in a bi-stable circuit changes from a conducting to a noncomz'zlctng state, a positive pulse is applied to the respective current generators 29 to initiate operation, while a change from l10n-conduction to conduction produces a negative pulse which terminates operation of the current generator.

The vertical scanning and current adder circuits 24 and 25 operate in the same manner as do the horizontal counterparts described above to produce a vertical step scan with the output from the last horizontal bi-stable multivibrator element, indicating the completion of one horizontal scan, triggering the first vertical bi-stable element 28 to shift the scan. Each of the binary Scaler circuits 22 and 23 as is standard practice in such devices feed back from the last stage thereof to the first stage in order to reset the sealer at the end of the scan. Thus, the horizontal Scaler 22 is reset at the end of every 32 pulses Whereas the vertical Scaler 23 is reset at the end of 1,024 pulses thus producing a scan over a 32X 32 matrix.

The staircase deflection current may be produced not by adding successive equal increments of current, requiring a separate current generator for each increment, but generally by selectively adding current increments of the 2n from (i.e., 201, 211, 221, 21I, 2111) which makes it possible to produce a staircase of 2 steps with n-l-l current generators. Thus, to produce a 32 step staircase deflection current (25 steps) only 6 current generators and their associated bi-stable elements are necessary rather than 32. To illustrate the principle assume that, for simplicity of explanation, a 7 step staircase is desired; if equal increments are to be added then 7 current generators are necessary which are successively turned on to produce the 7 incremental current steps. On the other hand, if current generators producing output currents whose magnitudes are related by a 2n, where factor then only 3 are necessary. That is, 3 current generators having currents of the following magnitudes,

r." .s 201, 211, and 221, may be selectively actuated to produce the staircase as shown in the following table:

As can be seen from the above table, the on-off sequences of the respective current generators follow the counting sequence of a binary scaler such as the horizontal Scaler 24.

The binary scaler illustrated in FIGURE 2 and the constant current generators associated with the individual elements thereof, are so arranged that n-{-lbistaible multivibrator elements are utilized for a storage matrix of 2n and the output currents from each of the constant current generators is adjusted to have a two-to-one ratio of current between any two successive stages so that the selective actuation of the bi-stable element and their associated constant current generators provide the desired staircase deection coil current. However, it is obvious, as pointed out previously, that scaling circuits other than of the type illustrated and described may be utilized to produce the desired deflection coil current wave.

In order to store data on a photographic storage element, such as is illustrated in FIGURE l, in binary form as black and white dots, it is necessary to blank the spot of light produced by the flying spot scanning kinescope element 6, periodically so that discrete areas are selectively exposed to produce the black and white dots. To this end a blanking circuit 32 which provides voltage to blank the electron beam of the kinescope in synchronism with the data to be stored is connected to the accelerating and blanking grid 9 of the kinescope 6. A pair of terminals 33 and 34 provide, respectively, a source of synchronizing clock pulses and the stored data in binary form i.e., pulse or no pulse) which are used to blank the electron beam of the kinescape 6. The terminals 33 and 34 may be connected directly to the output terminals of a computer with the terminal 33 connected to the clock pulse generator of the computer and the terminal 34 being connected to the data output terminal of said computer.

The synchronizing clock pulses appearing at the terminal 33 are connected by means of a suitable conductor to the scanning matrix 21 of the sweep circuit 20 and to the input terminal 27 thereof to provide the triggering pulses for the scalers to produce the proper deflection of the electron beam. The clock pulses from the terminal 33 are also fed, over another conductor, to the input of a gate 35, indicated in block diagram form but which may, for example, be a tetrode, which also has applied thereto the stored data blanking information from the terminal 34. The data representing blanking pulses from the terminal 34 function as the gating signal for the gate 35 and permits passage of clock pulses therethrough only on the occurrence of a pulse from the terminal 34. That is, as has been pointed out, the pulses from the terminal 34 are in binary form; i.e., pulse or no pulse, and thus a clock pulse is permitted to pass through the gate 35 upon the occurrence of a pulse at terminal 34 and none is permitted to pass through during the no pulse condition at this terminal. Thus, clock pulses passing through the gate 35 may then be utilized as blanking pulses for the kinescope.

Connected to the output of the gate element 35 is an amplyfying and pulse shaping and clipping circuit, indicated generally in block diagram form at 36, which functions to amplyfy the clock pulses passed through the gate and to reshape them into steep-fronted square wave pulses in the event the wave shape has degenerated during passage through the computer and the gate 35. The reshaped and amplified clock pulses are then fed by means of a suitable conductor to a delay line 37 in order to delay their application to the accelerating and blanking grid 9 by a suflicient amount to insure that the electron beam has moved to its new scan position prior to the pplication of the blanking signal. There is, since the clock pulses from the terminal 33 are utilized both to trigger the scanning matrix of the kinescope sweep circuit 20 as well as providing the blanking voltage for said kinescope, it is necessary that the application of the blanking pulse to the grid 9 be delayed until the electron beam has been swept to its new position. Under certain circumstances it is possible that the delay inherent in the gate and the amplifying and clipping circuits may be suiciently large to eliminate the need for a delay line; however, it should be kept in mind that the time delay in the scanning circuits 2t) and the blanking `circuits 32 must be equal in order to insure that the blanking of the beam takes place only after the beam has moved to its new scan position. Thus, it can be seen that the blanking circuit illustrated at 32 provides the means by which the light spot produced by the kinescope element 6 is controlled to produce discrete white and blank dots upon the photographic storage element l which dots represent the data to be stored in binary form.

In the data storage system illustrated in FIGURE 1, the pulsed fying spot light source is reduced and projected directly onto a high resolution photographic emulsion which constitutes the ultimate storage element. However, in some cases it may be desirable to utilize a two stage process wherein the data is -irst stored on an intermediate storage element of medium resolution by means of a pulsed fying light spot scanner and optical microscope combination and is then further reduced and stored on a high resolution photographic element which constitutes the ultimate storage element by means of a source of light and another optical microscope of the type illustrated and described with reference to FIG- UREl.

The general system of the present invention which has been described with reference to FIGURES 1 and 2 for storing data on a photographic storage element by imaging a flying light spot on a photographic transparency constitutes a preferred embodiment of a system for storing such data on a photographic plate which information is to be read-out by means of an electro-optical readout system presently to be described and illustrated in FIG- URE 3. However, it is possible to store data in binary form on a photographic storage element with the desired high storage density by means of a focussed, charged particle beam of the type produced by the electron-optics of an electron microscope or an X-ray microscope. Such an apparatus and method is described in the copending application of Sterling P. Newberry, Serial Number 668,489, tiled Iune 27, 1957, now abandoned, and assigned to the General Electric Company.

Assuming that the data has been stored on a photographic storage element by an apparatus of the type described it, consequently becomes necessary to provide means by which such data may be read out. FIGURE 3 illustrates such a readout system which is characterized generally by the fact that an optical projection system enlarges and projects the stored data onto the surface of an image storage tube which translates it into a series of representative electrical pulses. The projected image of the information matrix and the scanning field of the storage tube are aligned by electronically shifting the eld of scan of the storage tube until perfect alignment is achieved at which time the output and readout circuits of the apparatus are activated to produce output signals representative of the data.

Adverting now directly to FIGURE 3, there is illustrated a photographic storage element 40 of the type discussed with reference to FIGURE l consisting of a hermetically sealed high resolution emulsion mounted on a transparent backing plate which contains stored data in the form of 32x32 matrices of discrete light and dark spots. A rack and pinion element, illustrated at 41, is fastened to the storage element and provides a means for positioning the storage element to bring selected matrices into the field of view of a readout means presently to be described, which positioning means may be either manually operated or driven by a servo mechanism device. For the sake of simplicity only the rack and pinion for positioning the element in a single direction, the vertical, is shown; however, it is understood that a complementary horizontal positioning means is also utilized.

Positioned on one side of the photographic storage element is a source of radiant energy such as an incandescent lamp 42 which, in conjunction with a heat absorbing means 43, a filter 44 and a condenser lens assembly 45 projects a beam of light onto the photographic storage element in order to produce a light image of the matrices which, in turn, is amplified by means of a projection light microscope 46 positioned on the other side of the storage element.

The projection light microscope 46 projects the enlarged image onto the face of an image storage tube 47 which, in turn, translates the matrix into a series of electrical output pulses representative of the data. The projection light microscope is of the same type illustrated and discussed with reference to FIG. l and comprises a color-corrected achromatic objection lens system 49 positioned adjacent to the storage element 40 and a colorcorrected achromatic amplyfying lens system 48 positioned adjacent to the face of the image storage tube 47. A focus control assembly 50 is positioned intermediate the lenses 48 and 49 and permits manual or automatic adjustment of the focus of the microscope 46.

The storage tube 47 which has the information matrix applied thereto as a light pattern is of photoconductive target type which has its photoconductivity varied to produce an output pulse train representative of the light pattern. Such a photoconductive image storage tube, commonly known in the art as a vidicon, consists of an electron beam source 5l including an electron emitting filament, an accelerating anode 52, a fine mesh screen 53, positioned adjacent to a photoconductive target assembly. The photoconductive target assembly is positioned to intercept the electron beam and consists of a glass plate 54, a transparent conducting back plate and a photoconductive plate 56. Focussing and alignment coils 57 and 5S are positioned to control the trajectory of the electron beam in a well known manner to focus and align it with respect to the optical axis. A power supply, indicated generally at 59, provides energization for the electron source 51 as well as the coils 57 and S8. First and second vertical deflecting coil pairs 6i) and 61 and corresponding horizontal deflecting coil pairs 62 and 63 constitute the means by which the electron beam is swept across the face of the photoconductive target 56 in a desired sequence to produce the output pulse train representative of the binary information.

Photoconductive storage tubes such as the vidicon are based on the principle that the focussing of a pattern of light on the photoconductive target causes its conductivity to increase at the areas which are illuminated. Since the target conductivity varies with the intensity of the light, the discrete elements or areas shift their potential positive by varying amounts because of leakage currents to the transparent conducting plate 55 and in this manner a pattern of potential variation is established on the target surface 56 corresponding to the input light signal. The electron beam produced by the beam source 51 is scanned across the target and produces capacity current variations which flow to the conductive back plate S5 producing voltage variations corresponding to the input light signal across an output resistor R0 connected to the plate 5S. The output polarity of this device is negative in the sense that an increasing amount of incident light on a target element causes a greater negative variation in voltage across the output resistor R0 when the corresponding element is scanned by the beam. Thus, by projecting the image of a matrix constituted of discrete, exposed and unexposed spots onto the photoconductive target a series of output pulses in binary form (i.e., pulse or no pulse) are produced with the pulse representing an exposed discrete area, and no pulse representing an unexposed area. In this fashion it is quite clear that the output puise train faithfully represents the data stored on the photographic storage element 40.

rfhe respective vertical deflection coil pairs 6ft and 61 as well as the horizontal pairs 62 and 63 are connected to a beam deflection coil current supply circuit, indicated generally at 70, which produces a staircase shaped deflection current for scanning the electron beam of the vidicon 47 in a step-wise fashion across the photoconductive target 56. The deflection circuit 70, broadly speaking, includes a main deflection circuit channel having a first scanning matrix 7l controlling horizontal and vertical current adders 72 and 73 to produce a staircase shaped deflection coil current for scanning the 32x32 matrix projected onto the vidicon tube.

ln addition, there is provided an auxiliary deflection circuit channel for electronically shifting the vidicon scan until there is alignment between the tube scan field and matrix projected thereon and which includes a second scanning matrix 74 controlling a second pair of horizontal and vertical current adder circuits 75 and 76. The actual construction and manner of operation of the beam deflection supply circuit '70 will be explained in detail with reference to FIGURE. 4, at this point, however, suffice it to say that the beam deflection circuit 70 operates in such as fashion that the auxiliary deflection circuit channel 74, 75 and 76) function to shift the scan plaire of the vidicon until a particular information bearing 32X 32 matrix is coincident with the scan field of the vidicon, at which time it is inactivated and the main deflection channel is actuated to scan the matrix.

The pulse train output from the vidicon 47 appearing across the resistor R0 is fed by means of any convenient lead to the input of an amplifying means 66 and in turn to a pulse shaping circuit 67 which sharpens the wave shapes of the amplified vidicon output pulses. The vidicon output pulses appearing on the lead 65 are also applied to the coil deflection circuit 7i) in order to disable the auxiliary scan shifting circuit and actuate the main deflection circuit when the scan field is aligned with the 32x32 matrix in a manner which will be xplained in detail with reference to FIGURE 4.

In order to provide an additional check on the alignment and focus the data matrix and in order to provide a read command signal to the readout circuitry to be described presently, there is provided a monitor circuit 77 comprising an electrostatic deflection cathode ray tube 73, the deflection voltage of which is obtained from the beam deflection circuit through a monitor sweep amplifier 79. The vidicon pulse output is applied to the CRT by means of a lead 89 coupled to the control grid of the CRT (not illustrated) from the output of pulse shaper 67. Since a series of pulses corresponding to the black and white dot pattern comprising the stored information are generated as the electron beam scans the vidicon target, the application of this pulse train to the CRT 78 whose deflection voltage is in synchronism with the vidicon coil deflection voltage makes it possible to observe visually when the vidicon scan field is coincident with the information bearing 32 32 matrix at which time the readout circuitry may be actuated in order to produce output pulses representative of the stored data. In this fashion an additional safeguard is provided to insure that the output of the system contains no ambiguities due to a misalignment of the data matrix and the vidicon scanning system.

`In order to insure that the output pulses from the data storage system are uniform in shape and as precisely timed as possible, the output signal -is composed, not of the amplified and shaped vidicon output pulses themselves, but of clock pulses gated by the vidicon output pulse train. This is accomplished by `applying the vidicon output pulses from the pulse shaping `circuit 67 to a readout .gate means wherein the gating of a source of clock pulses by the vidicon output signal takes place. To this end a clock pulse generator means 81, illustrated in block diagram form, which may be either a blocking oscillator, yfree-running multivibrator or .the like, produces outpu-t pulses having some constant predetermined repetition nate. The clock pulses from the generator 81, in addition to being utilized as the output pulses from the data storage system, provide the triggering pulses for operating the -scanning matrices 71 yand 74 of the main deflection circuit 70 being applied thereto by means of any convenient lead 82.

rPhe lgating of the clock pulses from the generator 81 to produce the output pulses is achieved by a pair of gates 68 and 69 denominated as readout synchronizing gate, and a readout gate respectively. The readout synchronizing gate 68, which may be a pentode or any other sirnilar well known gating arrangement, makes certain that readout is initiated only at the beginning of the vidicon scan. Accordingly, the readout synchronizing gate 68 is opened and closed by means of a readout synchronizing signal generated by a readout control circuit 83 which is actuated by output pulses from the scanning matrix produced at the end of the previous scan `and prior t the beginning of each vidicon scan. The readout control circuit 83, indicated in block diagram form for the sake of simplicity, may be a bistable multivibrator, `for example, which produces a synchronizing gate signal in response to a triggering pulse from the decction circuit 7) applied by means of any convenient lead 84. As will be explained in detail later, with reference to FIGURE 4, the reset pulse from the vertical scaling circuit lof the deflection circuit 7 ti produced at the end of each 32x32, scan is applied to the control circuit `83 to produce the synchronizing signal for the gate 68.

In addition, a readout command circuit 85 is connected to the circuit 83 to insure that a readout synchronizing signal is produced only when readout is actually desired. The circuit 85 maybe a manually operated switch which enables the control circuit 83 and makes it :responsive to the appearance of the trigger pulse on the lea/d 84.

In operation the control circuit `8S is manually operated after the alignment and focus of the data matrix on the vidicon target has been checked by means of the monitor CRT 7S. At this time a read command signal from the control circuit 85 actuates the readout control circuit 83 permitting a pulse from the scanning matrix to open the readout synchronizing gate at the completion of the next scan. The synchronizing si-gnal from the readout control circuit 83 is applied to the gate 68 opening it. The gate `68 stays open d-uring the remainder of the scan allo-wing the shaped vidicon output signals from the pulse shaping circuit 67 to pass to the readout gate 69. At the termination of the scan another pulse from the deilection circuti 70 is transmitted by the lead 84 tothe readout control circuit S3 terminating the synchronizing gate pulse and closing the readout synchronizing gate 68 until the initiation lof another scan. The vidicon output pulses which have been passed through the gate 68 are applied to the readout of gate 69 and act as gating signals to permit the passage yof clock pulses from the clock pulse generator 81, which clock pulses are applied to any desirable utilization circuit and represent the data stored on the photographic data storage element 4t).

Before proceeding with the description of the beam deflection coi-l supply circuit 7 0, which is shown in detail in FIGURE 4, it will be useful to discuss brieily the under-lying reasons and objectives which constitute the basis for the particular approach chosen. As has been pointed out previously in the introductory remarks, one of the objects of the instant invention is to produce a data storage system utilizing a photographic storage elemen-t of high storage density in which selected blocks of data in 32x32 matrix form are located by means of a rough mechanical registering means and then producing exact centering and registration of the desired block of data by electricaly shifting the scanning center of the image tube. Hence, it is no longer necessary to provide high precision mechanical registering means in order to locate the blocks of data.

FIGURE 5 illustrates, schematically, the relative position of the 32x32 matrix and the vidicon scan eld immediately after vthis block of data has been roughly registered by means of the rac-k and pinion. The projected image consists of a 32X 32 matrix of discrete light and dark spots "a representing the data and denominated by the legend matrix Oifset and partially over-lapping the matrix is the vidicon scan field, identified by the legend vidicon scan, which is initially located to the left and slightly above the matrix. Located at the upper left hand corner of the matrix is a relatively large indexing mark B which, as will be pointed out in detail later, provides a registering pulse indicating that exact centering and registration of the desired block of data has been achieved and which inactivates the auxiliary deilection channel and actuates the main deflection channel in order to provide scan of the photoconductive target.

Exact centering `of the vidicon scan may be achieved, generally speaking, by initiating the vidicon scan in the odset position illustrated in FIGURE 5 until the indexing mark B is reached, at which time a registering pulse is produced which disables the auxiliary deflection channel. The auxiliary channel thus produces a biasing deilection current the magnitude of which is related to the deflection necessary to locate the matrix precisely at which time the main dellection circuit takes over and produces the main deflection current which is superimposed on the biasing deflection current and which causes the actual readout scanning of the information bearing matrix.

In order to provide the above described action there is provided electronic circuitry, illustrated in detail in FIGURES 4a and 4b, which can be broken down broadly into four major components: (l) Main Deflection Coil Current Channel, (2) Auxiliary Deection Coil Current Channel for electronically shifting the vidicon scanning until the matrix is located, (3) Switching Means to apply the clock pulses selectively to the main and auxiliary channels, and (4) Resetting Means for the Switching Means.

Main Deflection Coil Current Channel Referring now to FIGURES 4a land 4b, the main deection channel comprises, broadly speaking, a scanning matrix 71 consisting of horizontal and vertical scaling -circuits '71a (FIGURE 4a) `and 71b (FIGURE 4b) with the vertical circuit driven from the output of the horizontal scaler. Horizontal and vertical adder circuits 72 (FIGURE 4a) and 73 (FIGURE 4b) are connected respectively to the horizontal and vertical scaling circuits.

The horizontal scaling circuit 71a and the vertical scaling circuit 7111, in a manner similar to that described with reference to FIGURE 2, comprise a number of bistable multivibrator circuits arranged as binary `Scaler circuits. Each of the bi-stable multivibrators, one of which is shown in detail in each scaler, controls one stage of their respective current adders consisting `of a constant current generator 86 which may, for example, consist of a constant current pentode.

Each of the bi-stable circuits 85 consist of a pair of multivibrator connected tubes V3 and V4 and a pair of normally noneconducting trigger tubes V1 and V2 having a single common input. The arrival of a positive pulse on the common input lead of the trigger tubes V1 and V2 causes them to conduct and pass a negative pulse to the multivibrator connected tubes V3 and V4 thus, causing them to reverse their state and remain thus until the appearance of the next input pulse. The output from each of the bi-stable circuits drives the next succeeding one by coupling the output pulse from the tube V4 to the input lead of the trigger tubes V1 and V2 of the succeeding bi-stable circuit SS. In this fashion a positive pulse is introduced into the input of the next bi-stable circuit in the series each time the tube V4 from a conducting to a non-conducting condition, thus producing a sealer of the sort known as the 2 type.

The constant current generator 86 of the current adder circuit are each connected to the tubes V3 of each of the multivibrator connected circuits. Thus, each time a tube V3 in the circuit changes from a conducting to a non-conducting state, a positive pulse is appliedy to its particular constant current generator S6 to initiate operation while a change from non-conduction to conduction produces a negative pulse which terminates operation of the current generator. As described with reference to FlGURE 7., there will be produced at the output lead of the current adders a dellcction coil current in the form of a staircase which causes the electron `beam of the vidicon tube to scan the photoconductive target in a stepwise fashion.

Each of the binary sealer circuits 71a and 71b, as is standard practice in such devices, feed back from the last stage thereof to the tirst stage in order to reset the sealer at the end of the scan.

Auxiliary Deflection Coil Current Clzannel The auxiliary delieetion channel 74 in a similar fashion comprises a horizontal scaling circuit 74a (FIGURE 4a) and a 'vertical scaling circuit Mb (FIGURE 4b) actuated from the output from the horizontal scaling circuit. A horizontal current adder 7S and a vertical current `adder 75 are connected to the respective horizontal and vertical scaling systems in the manner described above to produce the respective deflection coil currents in a staircase form. The horizontal sealer 74a and the vertical sealer 74h consist of a number of bi-stable multivibrator circuits 37 arranged as a binary sealer circuit with each of the bi-stable elements controlling one stage of its respective current adder, each consisting of a constant current generator S. Since the auxiliary deflection circuit and the components constituting it operate in precisely the same manner as the main deection channel described above, the operational description need not be repeated here. Each of the horizontal and vertical current adders and both of the main and auxiliary deflection channels are connected to their respective vertical and horizontal deflecting coils shown and described with reference to the vidicon of FIGURE 3.

Selective Switching Means In order to apply the clock pulses from the clock pulse generator 8l ot FGURE 3 selectively to the auxiliary and main delleetion channels, a switching means 89 shown in FIGURE 4o at the upper left-hand corner is provided which applies the clock pulses selectively to the auxiliary channel 74- for electronically shifting the vidicon scan until exact registration of the matrix is achieved and then to the main deflection channel to initiate the stepscan of the matrix. To this end an input terminal 90 connected to the clock pulse generator 81 has the clock pulses which actuate the scaling circuits applied thereto. The clock pulses appearing at the terminal 90 are applied through a coupling capacitor 91 and a lead 93 to a normally open gate 92 and a normally closed gate 94 which control the application of the clock pulses to the auxiliary and main channels respectively.

The normally open gate means 92 comprises a tetrode vacuum tube, the anode of which is connected to a source of operating potential through a suitable anode resistor and the cathode to a source of reference potential, such Cir 12 as ground, through a cathode resistor. The lead 93 from the input terminal 9i) is connected to the control electrode of the tetrode 92 while the cathode is connected` by means of a suitable lead to the input of the horizontal scaling circuit 74a.

The normally closed gate means 94 similarly comprises a tetrode vacuum tube, the anode of which is connected to a `source of operating potential by means of an anode resistor and the cathode directly to a source of reference potential, such as ground. The control electrode Ma of the normally closed gate means 94 is connected directly to the anode of the normally open gate 92. Thus, while the normally open gate 92 is in its conducting condition and passing clock pulses to the auxiliary channel, the normally closed gate 94 is maintained non-conducting by virtue of the anode drop of the tube 92 and does not permit the application of clock pulses therethough, thus maintaining its associated del'lection channel 7l in a quiescent state. As soon as the normally open gate 92 is made non-conducting its anode potential rises, and unblocks the normally blocked gate 94 thus applying clock pulses to the main detiection channel, the input terminal of which is connected to the anode of the gate 94. In this fashion the clock pulses which actuate the respective dellection channels are selectively applied either to the auxiliary or main channels.

There is provided, in addition, a triggering means, actuated in response to a positive pulse produced whenever the electron beam of the vidicon scan strikes the registering mark B on the 32x32 matrix, to close the gate 92 and open the normally closed gate 94 to initiate scanning of the now exactly registered matrix. There is provided a gaseous trigger tube 9S, which may be a thyratron or the like, having a cathode, control electrode, and an anode. The anode of the gaseous tube is connected to a source of operating voltage by means of a suitable anode resistor whereas the cathode is connected to a source of reference potential such as ground while the control electrode is connected to a movable tap on a potentiometer connected in shunt with a source of negative biasing voltage 96 such as a battery. As is well Aknown to those skilled in the art, such gas tubes may be maintained in a non-conducting condition by means of such a critical negative bias until the bias is overcome at which time the tube conducts and remains in a conducting condition until the anode potential is reduced suiciently to extinguish conduction.

The anode of the gaseous trigger tube 95 is directly coupled to the screen grid of the normally open gate means 92 and in this fashion provides gating voltage therefor. Also connected to the control electrode of the gaseous tube 95 is the lead 65 from the photoconductive target element of the vidicon 47 illustrated in FIGURE 3. lnitially, the gaseous triggering tube 95 is non-conducting and its anode potential is highly positive maintaining the gate 92 in a conducting condition to permit passage of the clock pulses. Whenever the electron beam of the vidicon tube is deflected sufficiently, by virtue of the deilection current produced by the auxiliary deilectien channel, to strike the registering mark B at the upper left hand corner of each matrix, a positive pulse appears on the lead 65 connnected to the photoconductive target. This positive pulse is of suiicient magnitude to overcome the negative biasing on the control electrode provided by the biasing means 96 causing the tube to conduct heavily. The conduction of the gaseous trigger tube 95 causes a rapid drop of anode potential which is transmitted directly to the screen electrode of the normally open gate 92 causing that gate to close and preventing further clock pulses from being applied to the auxiliary channel. The anode potential of the tetrode 92 rises removing the biasing on the control electrode of the normally closed gate tube 94 permitting the clock pulses to pass to the main deeetion channel to initiate the scan of the now registered and aligned 32 X32 matrix.

Resetting Circuit In order to reset both the switching means 89, as well as the individual bi-stable units of the auxiliary channel 74, there is provided a manually actuated circuit operated from the output of the last bi-stable unit of the main vertical sealer 71b to place the entire circuit in condition to repeat the operation for a new matrix. A reset circuit, indicated at 93, for the auxiliary defiection channel is connected through a manually operated switch 99 and a lead 104 to the last bi-stable element S5 of the vertical scale 71b. The reset circuit 98 consists of a multiplicity of and gates 100, each coupled to individual ones of thebi-stable elements 87. Each of the and gates 100 consist of a pair of crystal diodes 101 and -2 having their respective anodes connected to a common terminal point. A resistor 103 connected between the common terminal point and a source of positive Voltage provides biasing for the igate, while a lead 105 carrying a positive reset pulse is connected between the terminal point and the common input lead of the trigger tubes V1 and V2. The cathode of the diode 101 is connected to the anode of the current generator controlling tu-be V3 of each bi-stable multivibrator while the cathode of the `diode 102, on the other hand, is connected to the lead 104 carrying the positive pulse from the vertical sealer 7i1b.

And gates of the type illustrated at 100 are 'characterized by the fact that if both crystal diodes are simultaneously driven positive the potential at their common junction, to which the lead 105 is connected, is caused to rise thus applying a positive pulse to the trig- Iger tubes V2 and V1. Thus, if in any given bi-stable element the tube V3 is in a non-conducting condition; that is, its anode potential is high and consequently its respective constant current generator is conducting, both of the diodes 101 and 102 will be simultaneously positive thus producing a positive pulse at the output lead 105 which, through the action of the trigger tubes V1 and V2, `reverses the states of the tubes V3 and V4 thus terminating the operation of the current generator. If, on the other hand, the tube V3 is in a conducting condition, its anode potential is low and consequently no positive pulse appears on the lead 10S and the bi-stable element remains in the same condition. Thus, by means of the and gates 100 those bi-s-table elements 87 which are in the state that their associated current generators 88 are conducting are reversed resetting the auxiliary deiiection channel and reducing the deflection coil current produced thereby to Zero.

In addition, a reset circuit 106 returns the gaseous trigger means 95 to a non-cond-ucting condition returning the normally open and normally closed gates to their initial operating conditions. There is provided a triode amplifying element 107 connected to th-e output of the vertical sealer 71b through the manually operated switch 99. The amplifier 107 consists of a cathode, control electrode, and an anode, the latter of which is connected to a source `'of voltage through -a suitable anode resistor and the cathode of which is connected to a source of reference potential such as ground through a suit-able cathode resistor. With the switch 99 in the closed position, the positive pulse produced iby the last bi-stable element 85 at the end of the scan is applied to the amplifier 107 which reverses its polarity. The now negative pulse is coupled by a lead 108 to the diode 97 connected to the anode of the gaseous trigger tube 95. As is Awell known, gaseous tubes of the type illustrated at 95 once iired Iare independent of grid voltage and can only be reset -by reducing their anode potential below the critical extinction voltage.

Thus, the negative pulse appearing on the lead 108 passes through the diode 97 reducing the anode voltage of the tube 95 ybelow the extinction voltage causing conduction to terminate. The cessation of conduction by the tube 95 raises its anode voltage re-opening normally open gate 92 to pass clock pulses 'fonce more to t-he auxiliary channel from the terminal 90. The conduction of the normally open gate 92, of course, reduces its anode voltage which, being connected to control the electrode of the normally closed gate 94, applies a negative bias thereto causing that gate to cease con-ducting and preventing any further clock pulses from lbeing applied to the main defiection channel 71. rIhus, it can be seen that both the auxiliary and main deflection channels 71 and 74 are now in condition to repeat the operation. Clock pulses are again fed to the auxiliary channel until the indexing mark B of the matrix is reached at which time another positive pulse appears yon the lead 65 of the control electrode of the gaseous trigger at which time the gate 92 is once more closed, opening gate 94 and permitting the main deflection channel to scan the new block of data projected onto the face `of the vidicon tube.

A terminal connected to the last bi-stalble element S5 of the vertical scaler 71b of the main deflection channel 71 is connected to the readout control circuit 83, shown and discussed with reference to FIGURE 3, and provides the trigger pulse for that control circuit to produce the synchronizing pulse for operating synchronizing readout gate 68. It is obvious that only at the end of a vertical scan will a positive pulse appear at the terminal 110 in order to trigger the readout control circuit 83. It can be seen that the readout synchronizing pulse thus produced enables the readout synchronizing gate 68 only at the beginning of a vidicon scan.

In discussing the data storage circuit of FIGURES 3 and 4, the storage tube `which translates the information matrix on the photographic dat-a storage element has, for illustrative purposes, been shown `and described as a vidicon which is of the type having a photoconductive target structure. It is obvious to those skilled in the art that many variations and changes may be made in the type of storage tube utilized without going outside the spirit and scope of this invention.

It is clear from the previous discussion that there has been provided a data storage system utilizing a photographic storage element which is capable of large capacity, high storage density, high reliability, and rapid access. Furthermore, a readout system has been disclosed in which the readout and translation of the stored data is achieved by means of a rough mechanical register of the desired block of data and exact centering and registration thereof by ymeans of an electronic shifting of the scanning center of :an image tube.

While a particular embodiment of this invention has been shown it will, of course, -be understood that it is not limited thereto since many modifications both in the circuit arrangement and in the instrumentalities employed may be made. It is contemplated by the -appended claims to cover any such modifications as fall within the true spirit and scope of this invention.

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

l. In a data storage apparatus the combination comprising a photographic storage element having data stored thereon in the form of individual exposure pattern matrices, means to translate said matrices into electrical output pulses including a photoconductive storage tube having a scanning electron beam, optical light microscope projection means including an achromatic lens system having a field fiattening lens to project an enlarged image of entire individual matrices onto said storage tube, said tube operating to produce electrical pulses representative of the stored data as said electron beam scans the image on said tube, means to control the scan field of said beam to align it with said image including a main beam deiiection supply circuit and an auxiliary beam deflection supply circuit, said auxiliary circuit being actuated to shift the beam scan until `a data indexing mark on said matrix image iS located and an output pulse is produced from said storage tube to disable said auxiliary circuit and energize said main circuit in response to said pulse to control the further scanning of said image.

2. In a data storage system the combination comprising a photographic storage element having data stored thereon iu the form oi individual matrices, means to translate the individual data matrices into electrical output pulse trains representative of said data, said translating means including a storage tube means, optical light microscope projection means including an achromatic lens system having a field iiattening lens for projecting an enlarged image of an entire matrix onto said storage tube means, scan control means associated with said storage tube means operative in response to an indexing mark on said matrix to align the tube scan iield with the projected image of one such matrix, pulse generating means connected to said scan control means to initiate and synchronize operation thereof, and means actuated in response to output pulses from said storage tube means to apply pulses from said pulse generating means selectively to a utilization circuit.

3. In a data storage system the combination comprising a photographic storage element having data stored thereon in the form of individual matrices, means to translate said data matrices into output puise trains representative of said data, said translating means including a storage tube readout device having a readout electron beam, beam scan control means for said readout beam, a light microscope optical projection clement including an achromatic lens system having a field liattening lens positioned between said storage element and said storage tube readout device to project an enlarged image of an entire data matrix onto said storage tube device, pulse generating means connected to said beam scan control means to initiate and synchronize said scan, means to apply pulses from said pulse generating means selectively to an output circuit in response to output pulses from said translating means including gating means coupled to said pulse generating and said storage means, said gating means transmitting pulses from said generating means in response tO the output pulses from said storage tube readout device.

4. In a data storage system the combination comprising a photographic storage element having data stored thereon in the form of data matrices, each of said matrices cornprising an exposure pattern of discrete dots, projection light microscope optical means including an achromatic lens system having a eld iiattening lens to produce an enlarged spatial image of said data matrices, electrooptical means having a predetermined field of view to translate the stored data in the matrices from the enlarged spatial image into electrical output pulses, mechanical positioning means for adjusting the position of said storage element in two coordinates relative to said means for producing a spatial image so that selected matrices are brought into substantial alignment with the readout field of view of said electro-optical means, and means to shift the entire scan field of said electro-optical means electronically to further align it with said enlarged spatial image prior to reading out said data.

5. In a data storage apparatus the combination comprising a photographic storage element having data matrices stored thereon in the form of an exposure pattern of discrete dots, projection light microscope optical means including an acnromatic lens system having a field liattening lens to produce an enlarged image of said pattern, cathode ray beam readout means having a predetermined field of scan, said readout means being so positioned that the entire data matrices in the form of said enlarged pattern are projected thereon to be scanned by the cathode ray beam to produce electrical output pulses representative of said data, and means to shift the entire scan iield of said cathode ray means electronically to align it with said enlarged pattern prior to reading out said data.

6. In a data storage system the combination comprising a photographic storage element having data stored thereon in tie form of individual matrices, means to translate said data into output pulses including a photoconductive storage tube having a scanning electron beam, projection light microscope optical means including an achromatic lens system having a iield llattcning lens to form and project enlarged images of entire individual matrices onto the viewing face of said photoconductive storage tube, step scan beam deiiection means for said photoconductive storage tube to scan said projected images and produce said output pulse trains representative of said data, pulse generating means connected to said scan means to initiate and synchronize said step scan, means to gate the output from said pulse generating means in synchronism with the output pulses from said photoconduetive storage tube including a first gate means for passing the photoconductive storage tube output pulses actuated in response to the initiation of said scan, a second gate means for passing pulses from said pulse generating means to a utilization circuit, said second gate being coupled to said tirst gate means and gated by photoconductive storage tube output pulses passed by said first gate means to pass pulses from said generating means.

7. ln a date. storage apparatus the combination comprising a storage element having a pattern of discrete transparent and opaque areas representing stored data, means to project a beam of light through said storage element t0 produce an image of said pattern in space, projection light microscope optical means including an achromatic lenS system having a iield iiattening lens for viewing said image and producing an enlarged spatial image of said entire pattern, electro-optical means having a predetermined field of view of said enlarged image for converting the stored data on said storage element into electrical output pulses representative of said data, and means to align said image and the eld of View of said electro-optical means in response to an indexing mark on said pattern.

References Cited in the tile of this patent UNITED STATES PATENTS 2,295,000 Morse Sept. 8, 1942 2,659,072 Coales Nov. 10, 1953 2,712,898 Knutsen July 12, 1955 2,714,841 Dorner Aug. 9, 1955 2,714,843 Hoover Aug. 9, 1955 2,731,200 Koelsch Jan. 17, 1956 2,738,499 Sprick Mar. 13, 1956 2,795,705 Rabinow June 11, 1957 2,807,728 ilburn Sept. 24, 1957 2,816,246 Bliss Dec. 10, 1957 2,817,041 Urry ec. 17, 1957 2,830,285 Davis Apr. 8, 1958 2,843,841 King July 15, 1958 2,859,427 McNaney Nov. 4, 1958 2,939,632 Demer June 7, 1960 2,950,465 Fox Aug. 23, 1960

Claims (1)

1. 7. IN A DATA STORAGE APPARATUS THE COMBINATION COMPRISING A STORAGE ELEMENT HAVING A PATTERN OF DISCRETE TRANSPARENT AND OPAQUE AREAS REPRESENTING STORED DATA, MEANS TO PROJECT A BEAM LIGHT THROUGH SAID STORAGE ELEMENT TO PRODUCE AN IMAGE OF SAID PATTERN IN SPACE, PROJECTION LIGHT MICROSCOPE OPTICAL MEANS INCLUDING AN ARCHROMATIC LENS SYSTEM HAVING A FIELD FLATTENING LENS FOR VIEWING SAID IMAGE AND PRODUCING AN ENLARGED SPATIAL IMAGE OF SAID ENTIRE PATTERN, ELECTRO-OPTICAL MEANS HAVING A PREDETERMINED FIELD OF VIEW OF SAID ENLARGED IMAGE FOR CONVERTIING THE STORED DATA ON SAID STORAGE ELEMENT INTO ELECTRICAL OUPUT PULSES REPRESENTATIVE OF SAID DATA, AND MEANS TO ALIGN SAID IMAGE AND THE FIELD OF VIEW OF SAID ELECTRO-OPTICAL MEANS IN RESPONSE TO AN INDEXING MARK ON SAID PATTERN.

Images