Click for next page ( 142


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 141
SESSION IV: NEW TECHNIQUES UNDER DEVELOPMENT CHAIRMEN Cal. Charles S. Gersoni MSC, and Col. Charles W. Hill, MS' THE PRESENTATION OF ALPHANUMERIC INFORMATION S. H. Boyd and C. W. Johnson, Stromberg-Carlson Company, San Diego TARGET INTENSITY ON A RADAR INDICATOR AS A FUNCTION OF THE RADAR PARAMETERS J. W. Ogland, Air Arm Division, Westinghouse Electric Corporation BRIGHT DISPLAYS VIA SCAN CONVERSION Thomas S. Wonnell and William E. Miller, Technical Development Center, Civil Aeronautics Administration THE ARMY-NAVY AIRCRAFT INSTRUMENTATION PROGRAM George W. Hoover, Office of Naval Research TRANSPARENT PHOSPHORS FOR CATHODE-RAY TUBES Charles Feldman, U. S. Naval Research Laboratory DIRECT-VIEW STORAGE TUBES George F. Smith, Hughes Research Laboratories BRIGHT-STEADY ELECTROLUMINESCENT DISPLAYS UNDER DEVELOPMENT W. L. Gardner, Lincoln Laboratory, Massachusetts Institute of Technology 141

OCR for page 141
The Presentation of Alphanumeric Information S. H. BOYD and C. W. JOHNSON, Stromberg-Carlson, San DIego Summary-Presentation o f letters numbers and symbols by means of cathode-ray tubes imposes problems whether characters are generated by raster scan, Lissajous, or shaped beam. In the Charactron shaped-beam tube, the electron beam imposes a char- acter from a matrix on the phosphor screen. Several important parameters interact with the light output from the tube and the number of characters presented. Operation of these factors in relation to room illumination is illustrated by surveillance systems. Not too long ago in surveillance systems, it was sufficient to present raw radar and supplementary data to observers. The systems performed satisfactorily, since the observer had sufficient time to calculate the necessary information for his task manually from the raw data. It is becoming increasingly important to provide the observer with sufficient quantity and quality of information to make decisions more accurately and rapidly. The problem is being continually mag- nified because of the increased speed and number of objects under surveillance. As a result, raw radar data and manual calculation are no longer adequate. Therefore' electronic data-processing equipment and character-generating cathode-ray tubes are now being incorporated. The data-processing equipment converts the raw data into digital form, after which it is presented to the ob- server as letters, numbers, and symbols by the cathode-ray tube. Presentation of alphanumeric data by means of cathode-ray tubes has im- posed new conditions and problems upon the display designer. It is the intent of this paper to discuss methods of character generation and some of the problems involved. Character generation Character-type information on cathode-ray tubes is generally generated by one of three methods: the raster scan' Lissajous, or the shaped beam. In the raster- scan technique, the intensity of the electron beam during the sweep is controlled. The characters are generated in segments in much the same manner employed in facsimile recording. In the Lissajous method, the electron beam is used as one would use a pencil; a como~na~on of intensification and deflection is used to obtain the desired share. In the sha~ed-beam approach' a series of stencil-like ~. ~.~ ~. 1 . 1 1 . 1 . 1 1 ~1_ ~1 ~ _ ~ ~ ~ openings within the tube are used to shape the electron beam into Ine prop~- configuration. The following will be confined to the shaped-beam method of character display. Figure 1 is a sketch of a Charactron shaped-beam tube and a typical matrix format. The tube has a 19-in. diameter screen and is used for visual displays. The operation of the Charactron shaped-beam tube is essentially the same as 142

OCR for page 141
that of any cathode-ray tube. An electron beam is generated by a conventional, high-resolution, electron gun. This beam is accelerated to the phosphor screen, where its energy is transformed into light. The major difference in the shaped- beam approach is that, instead of a spot striking the screen, a beam having a prescribed cross section is used. Formation of this shaped beam and the need to retain the shape at the screen make this tube different from standard cathode- ray tubes. The heart of the tube is the beam-forming matrix. As indicated in Fig. 1, there are 64 characters arranged in an 8 x 8 format.) The choice of char- acters is essentially unlimited. MU-METAL SH I ELD COI | ~ ADJUSTERS HELICAL ACCELERATOR7 ELEcTRON/ S ~ act _~ ////~/~ ~ CONVERGENCE / / ~ \ COIL / DEFLECTION \ REFERENCE YOKE \ PLATES DISPLAYED\ \\ CHARACTER ~ {DOTTED LINE INDICATES PATH \ OF ELECTRON BEAM) \ Fig. 1. Charactron shaped-beam tube: C1 9K tube and coil assembly; typical matrix format. ., In operation, an electron beam is generated by the electron gun. The size of the beam is adjusted to cover only one of the available characters in the matrix. Voltages applied to the horizontal and vertical selection plates direct the electron beam to the desired location on the matrix. The matrix intercepts this beam in much the same way that a stencil intercepts a light beam. The beam on passing through the aperture, takes on the cross- section of a character. The electron beam is usually deflected off-axis in the process of character selection. Since this would give poor registration at the screen, a convergence coil and set of reference plates are used to return the beam to the tube axis. The deflection yoke is used to deflect the beam to the desired location on the tube face. After deflection, the beam is accelerated to a high energy by means of a helical-type accelerator. The shaped electron beam then strikes the phosphor screen where the Here is converted to light. 7 - - - ~- -- - - _ _ _ _ Only one character is generated at a time. Hence, any character may be positioned at any point on the screen. In most applications a display rate of approximately 10,000 characters per see is used. Depending upon the system HA greater or lesser number of characters may be used. [Iowever, 64 have been found adequate for most applications. 143

OCR for page 141
parameters, a rate up to 20~000 characters per see is attainable. At higher speeds, component limitations destroy the display usefulness.2 At present, there are two types of Charactron shaped-beam tubes in quan- tity production: the C19K shown in Fig. 1, and the C7C which has a 7-in. di- ameter screen and presents characters 0.035 in. high for use in photographic recording and with projection equipment. In addition, a new tube type, the C19Q, has been developed to serve as a combined display. This tube, which has the capability of presenting high-resolution data in typical radar form ir1 con- junction with characters which can be varied in size, will be used in future air-traffic-control installations. Display design considerations A problem encountered in any CRT display is obtaining sufficient light output. In a character-generating system, the maximum beam-current density is less than that obtainable when conventional radar data are presented because of "space charge" effects. The influence of "space charge" becomes more noticeable ir1 character presentations since identification has been added to the problem of detection. A degree of blooming which is not considered harmful in radar displays may become serious when presenting characters such as a B and an 8. Blooming in character displays may be considered analogous to noise in radar displays. The total light output obtainable when using the shaped-beam principle cannot be compared to other types of displays by merely comparing the ratios of beam-current densities employed. A proper evaluation requires a comparison 2This statement holds true only for magnetically deflected displays. With electostatic deflec- tion, display rates of 50,000 characters per see or more are possible. This method is not practical in large screen displays at present because of deflection distortions. _ n) C;Pf1T WRITING >< - - - - - - \~ _b) SHAPED BEAM WRITING - I Tl ME Fig. 2. Comparative light output curves for spot writing and shaped-beam writing of an elemental area. Only one pulse is shown. - 144_

OCR for page 141
of coulombs per unit area per unit time. This can best be recognized by reference to Fig. 2, which gives curares of light output as function of time for an elemental area. In radar or other types of displays employing a sweep, the time a given point on the screen is excited depends upon the sweep speed and spot size This time is generally of the order of 0.01 to 1 parsec. Hence, curve a represents the light output for spot writing. In the shapecl-beam display, a given elemental area is generally excited for 10 to 100 Alec, yielding a curve of the type shown in b. Total light output is obtained from the area under the curve. Therefore, the excitation time becomes an important parameter when analyzing a system em- ploying cathode-ray tubes. A second important condition that must be considered in designing character-generating displays is the number of characters to be presented. Let us suppose that the generation of a character requires, on the average, 100 ,usec. The consequent presentation rate of 10,000 characters per sec is a fundamental restriction. As the quantity of information increases the number of times per second that it can be presented decreases. If it is desired to present 1,000 char- acters, then the maximum frame rate is 10 per sec. If 2,000 characters are to be presented, then the maximum frame rate is 5 per sec. In addition, if several different displays are being generated from the same source (for instance, the storage drum of a computer), and if it is desired to display different information serially on different consoles then the frame rate is determined by the total data to be presented and not merely by the data which happen to be utilized. This relation can be expressed as 1 Fn~a.r ~ X to where Flax is the maximum frame rate in see -1, to is the average time in see necessary to present a character, and N is the total number of characters being considered. This fundamental relation influences many of the parameters in- volved in the display, such as integrated light output from a given character, flicker, and legibility, as well as the choice of phosphor and ambient illumination. In most character-generating displays only a fraction of to is used for actual character generation. The remainder of to allows the components to stabilize. The actual print time per character varies from 10 to 100 ,usec, depending upon various factors such as the component stabilization period and the output speed of the data-processing equipment. Systems applications The Charactron shaped-beam tube has been used in several applications having different operating parameters. In addition, systems being developed introduce new conditions which must be considered in system design. As examples of different applications, a Navy surveillance system and the SAGE displays will be compared. The display portion of the Navy surveillance system is cur- rently being manufactured at Stromberg-Carlson, and many of the problems concerning its ambient lighting are still unsolved. In the Navy surveillance 145

OCR for page 141
- D lo. . I ~ >.ol tool \ \ a! HI 1 ~ AMBIENT ILLUMINATION 0 3 6 10 IS 20 2S 3S 4$ S5 65 75 am_ ~I~3 Cam) _ w ~ 0.1 HI ! . I ~ _~ .~_ ooo S - ~ {ma) _ . * ~ooo ! : I 30~ ~ 2~0 _ m 10= / w ... . ~ / / HI a. / ~ / ~^~aiENT lLLu~'NATION ~ _ ~ 0 ~ 6 uNGLANK[NG :~PLlTuOE (VOLTS) LOW SCREEN CURRENT HIGH SCREEN CURRENT ! lava / 1 _... O. ~5 (ma) _ h fig. 3. Measurements taken on the screen of a C1 9K tube. 146

OCR for page 141
system, it is required to display as many as 1356 characters. This requirement, coupled with the output speed of the computer storage drum, results in a frame rate of 15 per see, or a total available time per character of approximately 50 ,usec. However, the actual printing time per character is 10 Basic, giving a total of 150 user per see per character. Figure 3 shows experimental data taken on a P14 phosphor screen of a C19K tube. The data were recorded under operating conditions comparable to those which will be employed with the Navy surveillance display. Included in the figure are curves of mean character brightness character-background con ~ . . . . .. ... . ~ ~ . ~ ~ , trast, apparent flicker, and 1eglblllty plotted as function of screen current yu~l- blanking amplitude). Since it is generally desirable to maximize light output without destroying character legibility, the tube is operated just below the blooming point. Analysis of the figure shows that at this point. the pulsating brightness flicker is high, ~ ~ ~ - ~ ~ ~ ~ ~ 1 . . . 1 1 ~ ~ _ _ ~ ~ ~ the apparent legibility is only fair, and the brightness contrast has been reclucea. It would appear that the overall character presentation could be improved by reducing the mean character brightness. It can be seen that a reduction in mean brightness can reduce the apparent flicker considerably and thereby increase apparent legibility. This can be accomplished by reducing the screen current, using a filter over the screen to remove a portion of the character brightness' or 0 1 ~ 1 ~ ~ ~1 1 ~ ~ . 1 . ~ 1 ~ ~ 1_ _ _ t _ ~ by a combination ot the two. lne choice is lalrly stralgntIorwara 1I Ine OL)~-V~1 is located in an isolated room and if his only task is that of viewing the character presentation. In this case, the screen current can be reduced to a point at which .1 1 . 1 1 1 1 1 . tne cnarac~er-oac~rouncr Mess is high, the flicker is just noticeable, and O O lo, the legibility is good. Unfortunately, this situation rarely presents itself. Usually, a compromise must be reached such that adequate illumination is available for the observer to carry out a secondary task without destroying character legibility. To achieve the need for adequate illumination, two techniques appear to be feasible: selective spectrum and polarization. Each technique employs a filter over the display screen to reduce the mean character brightness to a level of just noticeable flicker, high background contrast, and good legibility. Appropriate lighting sources, when coupled with these filters, apparently provide satisfactory levels of illumination in the working area, but do not wash out the character- background brightness. This simplified analysis cannot be expected to solve all the problems associ- ated with the design of the Navy surveillance (or any other) lighting nor does it choose between those schemes that appear to be feasible. However, this type of analysis, as a minimum, should be used before the display parameters, such as frame rate and print time' have been fixed, in order that satisfactory illumi- nation can be provided. In contrast to this example, let us consider the SAGE display. Because of the quantity of data to be presented, the frame rate is limited to a 2~/2-sec cycle. The actual printing time per character is 25 ,usec. Under these conditions, an on-off type flicker is present, and the total light output is low. Because of the 147

OCR for page 141
operating conditions imposed on the tube, a low-level ambient lighting is required. These problems, coupled with others described in a previous paper, have appar- ently limited the choice to a selective spectrum lighting system, the broad-band blue. In conclusion, character displays are playing a more important part in surveillance situations, and the outlook for the future is that the applications will continue to grow. Because of this growth it is important that systems designers become familiar with the variables involved in this type of display. Only in this manner can they hope to improve the position of the observer who, in the end, is faced with making accurate decisions and making them quickly it is this task that determines the worth of any display system. -148

OCR for page 141
Target Intensity on a Raclar Inclicator as a Function of the Raclar Parameters J. W. OGLAND, Westinghouse Electric Corporation Summary With a fighter aircraft radar, detection is determined by the intensity impulse. For a typical radar CRT, maximum brightness is reached long before the grid voltage reaches zero, and at about ]/4 maximum beam current. Brightness saturation zs not serious for target ctetectzon because the intensity continues to increase uniformly up to zero grid voltage. When the numbers of overlapping excitations is increased, the intensity also is in- creased, but not proportionately. A maximum value is reached asymptoticaZZy, and more sZowZy the longer is the persistence of the phosphor. Although the intensity as a function of the num- ber of excitations saturates, the intensity impulse increases Zine- arZy. The spot intensity increases with increasing prf. It is essen- tiaZZy a negative exponential relationship to the interval between pulses. The intensity increases with increasing puZse-width, but less than proportionately. The intensity increases essentiaZZy ZinearZy with with screen voltage. The spot brightness increases nearly pro- portionateZy to the square of the screen voltage. In the direct-view storage tube, the brightness or intensity attozned zs not higher than in a conventional CRT, but the persistence, which is con- troZZed eZectricaZZy, is much longer. The intensity-time integral therefore is much larger, which gives the impression of much higher brightness, target detectability is better. Transparent phos- phors reduce the background brightness and improve the contrast with a given signal strength. Presently avaiZabZe transparent phosphors are efficient for low beam currents. If research in progress leads to good efficiency also at high beam currents, good contrast and detectability wiZZ be attained. At high altitudes, the radar indicator of fighter airplanes is subject to in- tense environmental illumination which tends to wash out the radar information presented on the screen. To attain good detectability, the radar designer should choose the electrical parameters in such a way that the target spots have the best light output possible. Very little information is available on the behavior of the cathode-ray tube pertaining to this situation. To provide some data which may assist the designer in making the best choice of parameters, some measurements have been made of the CRT under simulated radar operating conditions. Vision scientists have already provided data by means of which the visual requirements for detection of a small bright spot on a background of varying brightness can be predicted. These investigations were made by varying the spo brightness' sizes duration and the background brightness. and registering the ,_, , ~ ~ 7 ~ response of a crew of observers. . ~, ~- -~ ~ ~ --- - ~ At, -- For investigating the performance of radar indicators, the same technique may be employed by varying the electrical parameters. It appears, however, that -149-

OCR for page 141
with the above information on human vision available, this method is unneces- sarily time consuming for a study of this kind. Furthermore, since the radar indicator tube does not have an instantaneous respone and is nonlinear, valu- able information on its behavior will be lost if this technique is used. For these reasons, CRT operated under simulated radar conditions has been investigated by recording the output in photometric units when varying the parameters. These investigations will, thus, utilize existing knowledge on human . . vision. Human requirements for target detection Before discussing some of the measurements made on the CRT, it appears desirable to draw attention to some requirements for detection. Figure 1 gives Am m a He o Cut C) - a to .01 = :: ~ 10 100 1000 1 09000 BACKGROUND BRIGHTNESS (FT-L) Fig. 1. Contrast required for detection as a function of background brightness, for various spot sizes. The spot sizes are given in mm diameter at a viewing dis- tance of 20 in. Exposure 30 msec {Graham and Bartlettl). the results of an investigation by Graham and Bartlett. Similar investigations have been made by Blackwell.~ The individual curves for different sizes of spot show the contrast ratio needed for detection as a function of background bright- ness. The larger spots need less contrast ratio than the smaller. Since the intensity of a bright spot is equal to the product of its area and brightness (as long as spot radius is small compared with the viewing distance), Fig. 1 may be replotted showing spot intensity required for detection as a function of background bright- ness. This is shown in Fig. 2. The ordinate is called incremental intensity because it is the product of spot area and ~ B. the difference between spot brightness and background brightness. This is justified since any brightness ~ B produced by the CRT electron beam will add to, or float on top of, and background brightness produced. by ambient light. At high brightness, the curves for medium-sized spots overlap. These are the spots that occur in a radar presentation' where de- tection is determined by the spot intensity rather than the brightness alone. The brightness may be reduced if the spot area is increased in proportion, or vice -150

OCR for page 141
loo 10 L J Cl z C' - __ Z .10 zl~ clot 10 100 1000 BACKGROUND BRIGHTNESS (FT-L) 10,000 Fig. 2. Incremental spot intensity required for detection as a function of back- ground brightness. ~ Derived from Fig.1.) versa. For low' background brightness, these curves do not overlap' but this is without importance since our problem of detection is at high brightness. The measurements of Graham and Bartlett were made with a duration of the bright spots of 30 msec. For shorter durations a higher contrast is needed for detection. It is known that up to about 0.1 sec intensity and duration are reciprocal, i.e., the product is constant.3 The vertical scale on Fig. 2 could therefore be multiplied by 30 and expressed ire candle-milliseconds. This product may be called intensity impulse. The significance of this term will be seen below. iGraham, C. H. and Bartlett, N. R. The relation of size of stimulus and intensity in th human eye: III. The influence of area on foveal intensity discrimination. J. Exp. Psychol., 1940 27, 149-1 59. 2Blackwell, H. R. Contrast thresholds of the human eye. J. Opt. Soc. Amer., 1946, 36, 624-643. 3Stevens, S. S. (Ed.) Handbook on Experimental Psychology, Wiley, 1951, 964. 151

OCR for page 141
systems. A 21-in. halftone tuber is under development for such applications. This tube makes use of curved' self-supporting screens and an envelope with a curved faceplate. Because of mechanical considerations, neither the meshes nor the faceplate can be flat when the tube diameter becomes this large. The thick, self-supporting target material, which cannot be fabricated with as many holes per in. as thin, flat material, still provides several times as many holes in the entire target and a substantially increased total resolution. Current develop- mental tubes provide resolutions in excess of 750 distinguishable white lines. The spot-writing version of the 21-in. tube can write at least 40,000 in. per see, and the light output is in excess of 50 ft-l, an output quite adequate for use is normally lighted areas. A letter-writing version of the 21-in. tube can write 40~000 letters per sec with a displayed letter size of 0.15 in. Halftone uniform- ity comparable to that of the 5-in. storage tube has been achieved. fig. 6. One field of a television picture stored on a 5-in. Tonotron tube. Also in the advanced development stage is a multicolor storage tube'89 which incorporates shadow-mask color selection together with halftone storage in a 10-in. envelope, as shown in Fig. 7. A storage screen is inserted between the shadow mask and the color-dot plate of a conventional three-gun color tube. A ring-type flood gun has been added to provide the low-energy viewing beam. 7Koda, N. I;, Lehrer, N. H., and Ketchpel, R. D. Twenty-one-inch direct-view-storage tube. IRE WESCON Convention Record, 19577 1 (Part 7), 78. SBeintema, C. D., Smith, S. T., and Vant-lIull, L. L. Multicolor storage tube. IRE Trans- actions PGED, 1957, ED-4, 303 309. ~Vant-Hull, L. L., and Beintema, C. D. Multicolor storage tube. Wright Air Development Center, 1957, Technical Report 57-400. 186

OCR for page 141
As shown in Fig. 8' the storage target has one hole for every phosphor color dot. The writing beam from one gun, say the red gun' can strike the dielectric around only those storage-surface holes that are aligned with red phosphor dots. The flood beam coming through the shadow mask from the extended- source flood gun illuminates the entire storage surface but penetrates it only where the charge pattern is positive. Excellent color purity has been achieved over the entire 6-in. useful screen diameter of the tube and at least seven distinct colors can be distinguished. A color storage tube could be used for any applica- tion in which another display dimension is required, e.g., for the display of identification, speed, or altitude, in addition to the usual radar parameters of range and azimuth. _ VIEW PLATE STORAGE M ASK \ SHADOW MASK \ \ SHIELD GRID\\ \ \. 0~ . li ~1 l l , I ll ll ~1 1 I:,,= ~ THREE- BEAM WRITING GUN ool l: ~ Rl NO- TYPE FLOOD GUNK 1~ 1-\ . ~ ~ 7~ ~ COLL IMATOR -18 INCHES Fig. 7. Schematic of the multicolor storage tube. 187 , ~ 1 9 1/4 I NCHES

OCR for page 141
SHADOW MASK APERTURE STORAGE MASK APERTURE PHOSPHOR COLOR DOT r ~ ~ ~ I ~ ~ V ~Y I ~ ,_ _~/ / Fig. 8. Alignment of shadow mask, storage mask, and phosphor dots in the multicolor store ge tube. 188

OCR for page 141
BrighI-Steady Electroluminescent Displays Uncler Development* W. L. GARDNER, MIT Lincoln Laboratory Summary-Displays for presenting the output of computers can be improved by utilizing the solid-state processes of electro- luminescence and photoconductivity. Persistent-emission panels and storage panels are of particular interest. Persistent-emission panels, the construction of which is described, have adequate resolution and are superior in their light output and persistence to the phosphors of conventional cathode-ray tubes. The con- struction of storage panels is similar to that of persistent-emission panels, and methods for triggering the pane, are of particular interest. Our group at the Laboratory, working on display techniques, has in the past been associated with display components such as the Charactron tube, the Typotron direct-view storage tube, and high-speed film-processing systems. For some time now' we have been investigating improvements to computer output displays that might be made possible by utilizing the solid-state processes of electroluminescence and photoconductivity. The work can be divided into two areas: (1) the application of electro- luminescent and photoconductive processes to display panels with persistent light-emission characteristics; (2) the application of these processes to storage or memory panels. Persistent-emission panels These panels are distinctly different from the class of memory-display panels (bistable storage cells) in that they exhibit a characteristic steady emis- sion pulse for a controlled period, with a sharp fall-off of emitted light at the end of this period. For applications where repetitive information in short-pulse form is displayed, the overall advantages potentially available are a steady and high light output and automatic cut-off at the cessation of input pulses. Figure 1 shows a comparison of this light-emission characteristic with that of conven- tional phosphors. It can be seen that the light output is higher and holds rela- tively steady for a period of 5 to 7 sec. This period can be varied by panel con- struction and applied-voltage variation. Comparison emission values for PI and P19 phosphors currently used are also shown. The difference in the area under each curve is a measure of the integrated useful light, and this difference shows a possible improvement factor of 50 to 100. :Figure 2 shows the construction of a persistent-emission panel. The ma- terials utilized are capable of withstanding high temperatures and can be fabricated in large, flat-panel construction. The first layer is transparent glass, e.g., the faceplate glass of a cathode-ray tube. The second layer is a transparent electrode of tin oxide. The next layer is an electroluminescent layer, 0.002-in. *The research in this document was supported jointly by the Army, Navy, and Air Force under contract with the Massachusetts Institute of Technology. 189

OCR for page 141
In a fired glass medium. The next thick, consisting of the phosphor supported in OA~ layer is a 0.001-in. webbing of optically-opaque and electrically-insulating black glass. On this is deposited a 0.005-in. layer of sintered powder cadmium sul- phide. This is baked at 535 C for 15 min. Finally, a transparent top electrode is applied. This is indicated by the metallic strip electrode. 10 us 4 - 4 a 0 1 . 1 \ 200ClS 600 v 7 0 1 2 3 4 5 6 \ Time ( seconds ) Fig. 1. Light emission from a persistent panel. 8 9 With application of 500 c at 300 ~ and with pulse radiation incident on the metallic electrode side, the light-emission characteristic from the corresponding electroluminescent pillar is of the general form shown in Fig. 1. Pulse radiation can be in the form of either an incident electron beam or light emission from an intermediate phosphor. At present' the phosphor method is the most efficient in terms of utilization of beam energy. Figure 3 is a photograph of light emission from an operating panel with a specific light pattern incident. This is a 4-in. X 4-in. panel with cell spacing of 25 cells per in. As Fig. 3 indicates, with this resolution the delineation is still sharp, and the information pattern does not diffuse into adjacent cells. Panels with 40 cells per in. have been made, and it is estimated that 80 cells per in. can be attained without changing present panel-fabrication methods. Photometer traces of the light emission from a persistent panel have been re- corded when a series of four input pulses were spaced over approximately 2~/2 sec and then terminated. Because of the logarithmic response of the eye the fall-off in light during the 2~/2 sec cycle was not significant, and the emission 190-

OCR for page 141
appeared steady. However, at the cessation of the input pulses the light emission fell quite rapidly, and within 1 see was at the. background level. - O I. ~ Hi. - ~... .~. Fig. 2. Panel construction for a persistent-emission display. Fig. 3. Photograph of operating panel. 191

OCR for page 141
Figure 4 is a plot of light emission as a function of time for various applied voltages at the fixed frequency. The family of curves demonstrates that light output is increased, and that the shape of the emission curve changes, as the voltage is increased. a 0 4 600 v. us 6 so Q 50~ ~ 1 i, 2 - 500 c/s 40~\\ \\ \N ~ _ ~ ~ '- --I- - - . 0 3 6 9 Time ~ seconds ~ Fig. 4. Insight emission vs. time for various voltages. Figure 5 shows a similar family of curves for various frequencies at a fixed applied voltage. It is of interest here that the shape of the plateau, i.e., its height and length, can be significantly changed by varying the frequency. On the other hand, the integrated light output, i.e., the total area under each curve, is approximately constant. In general, the voltage and frequency stability re- quired of the power supplies is not at all critical. However, an ability to vary these over a range would be useful for some applications. Storage panels Large, flat displays utilizing electroluminescent and photoconductive components have also been under investigation at the laboratory. These con- sist of an optical storage panel (a flat array of optically bistable cells) with a flat X-Y trigger panel. There are other combinations of light-emitting and voltage-control elements which might constitute a display, e.g., electrolumines- cent cells with ferro-electric or ferro-magnetic control elements. However, we will discuss here only the type of work our group has been concerned with. 192

OCR for page 141
8e, u] 4 ~ - 4 o o _ 4 4~ o .= 2 :~2000c/s o 500 v.A.C. 1000 c/s \ \\ 500 c/s Time ( seconds ) Fig. 5. Light emission vs. time for various frequencies. ~200 c/s 9 The optical storage panel resembles the Persistron panel (cf. Fig. 2~. The various layers are as follows: First, there is a supporting glass substrate, and then a transparent tin oxide conductor continuous over the whole panel. The . ~. . ~ . next layer consists of isolated tin oxide conductors; this is produced by a photo- etch technique. Next is a layer of thin black glass which is optically and elec- trically insulating. ~. The next layer IS 01 sistered powder But, VVI11~11 Ib ~11 :^ of ^;~^Y.^A ~r`~7`rA^~ CA~ `xrh;~h ;~ rare J _ _ , . . ~ , veniently formed by a mechanical mask and air abrasive techniques. Finally, a top electrode is applied which is continuous and does not go down into the hole, or cavity. The inside wall of the hole represents a photoconductive element. The square cell of electroluminescent material at the bottom of the hole is a capaci- tive component. The electrical circuit, then, is a resistive component in series with a capacitive component. In the dark, most of the voltage appears across the resistor. When trigger light is incident, the resistance drops and the electro- luminescent cells light up. A portion of the emitted light feeds back on the inside wall of the hole and maintains the resistance in the low state so that the cell remains on. To extinguish the cell, the applied voltage is cut, allowing the dark resistance to be restored, and' when the voltage is reapplied, the cell 193

OCR for page 141
remains off until triggered again by incident light. Present photoconductive material and available light processes are such that a 1-msec trigger time per cell is apparently feasible. Figure 6 shows one method of triggering an optical memory panel. Trans- parent electrodes' consisting of parallel strips, are crossed with respect to each other on the top and bottom of a thin (0.002-in.) electroluminescent panel. When a voltage wave is applied to the one electrode on each axis, the inter- secting point lights up. Light levels of 100 to 1000 ft-l are achievable with 500 pulses to 50,000 c. A 1-msec envelope of these pulses is sufficient to trigger a corresponding storage cell when the storage panel is adjacent to the trigger panel. . ~ . ::~: Fig. 6. Trigger panel. Unfortunately, with this crossed electrode structure a i/2-voltage wave form appears along the line, and the whole line also lights up with decreased lumi- nance. The contrast between the intersecting point and the line is an important consideration, and' with certain modes of driving, this contrast can be as high as 100 to 1. The driving circuitry, shown in Fig. 67 consists of a saturable core which performs the double function of a voltage transformer and switch. There is one core per line. The core is of the order of 3/8-in. diameter. The transistor driver network is so arranged that only one core at any one time is in the unsaturated state and, therefore, can act as a transformer of the single high- frequency driver. Our present developmental goal is an 8-in. X 8-in. array with 16 cells per in. The array should have 128 cores for each axis. However' because of the binary logic in the transistor drivers considerably fewer transistors than this will be required for each axis. The complete assembly will work from low-level binary computer output signals. The light emission should fall in the 5- to 10 194

OCR for page 141
ft-l range with good contrast. At this emission level' components have yielded 4000 hr of operation at 100 per cent duty cycle. There are several applications for a display panel with this resolution and writing speed. The combination of trigger panel and storage panel represents a complete display unit. It is an initial step away from the standard CRT display. ., 195

OCR for page 141