Electron Tubes: A Constant Challenge

When the history of our time is analyzed in future centuries, I feel certain thatcommunications will be singled out as the key factor in making possible the"one world" concept-the basis for man's best hopes for the future. RCA'sbusiness is communications, and the people of our Lancaster plant have beena strong link in the development of modern systems through which men ex-change ideas and convey information.

Lancaster contributed to the success of the rapidly growing television indus-try by establishing the first automatic high -volume production facilitiesfor black -and -white picture tubes immediately following World War II.

This was a major milestone in industrial progress. Later, during 1954 to1961, Lancaster built the first automatic high -volume production facilities toproduce shadow -mask color picture tubes. The worldwide industry was

skeptical about the practicality of the shadow -mask precision tolerances and complexity. Lan-caster effort proved to be so successful that theentire industry has adopted the design and fol-lowed RCA's pioneering leadership. Color tele-vision was one of the major growth industries ofthe 1960's. To the engineer goes the credit forthe indispensible design of the tube and the tubeprocessing equipment.

And yet all that has gone before is only a fore-runner of communications systems still waiting tobe developed. Some systems will incorporatecomputers; some will, perhaps, have solid-statedisplay panels; but without doubt, whatever theirnature, the engineers at Lancaster can be ex-pected to play a key development role.

H. R. SeelenDivision Vice President andGeneral ManagerTelevision Picture Tube Div.Electronic Components

The Lancaster operation is an ultra -modern industrial and engineering com-plex producing a wide variety of electron tubes for a myriad of exciting andimportant applications in many fields of modern technology; for example,new medical diagnostic techniques; night vision employing light from thestars as the only source of illumination; power amplification for linear accel-erators used to probe the secrets of the atom; surveillance satellite cam-eras used to better identify the natural resources of the world, just to namea few.The advent of solid-state devices and technologies which was reputed tosound the death knell of tubes has, in fact, accel-erated the growth of the industrial tube business.Combining the older tube technologies with newsolid-state knowledge has advanced the fieldsin which electronics participate and we at Lan-caster look forward to many new advances inthe tube art.At Lancaster, the engineer can pursue his profes-sional career in an amazing number of fieldsranging from mechanical design to new opticaltechniques. The Lancaster Plant is unique in therange of professional skills demanded by theproducts designed and manufactured here. Nowhere is the engineering challenge so varied orthe opportunity for professional recognition sogreat. The engineer is limited only by his ownambition and aspirations.

C. H. LaneDivision Vice President andGeneral ManagerIndustrial Tube DivisionElectronic Components

Our CoverThe three tubes superimposed above the engi-neering scene on our cover are but a small sam-ple of the myriad tube types designed and pro-duced by the RCA Lancaster facility. Yet, thesethree tubes-a color TV tube, an L -band module,and an image orthicon-set the theme for thisissue by representing the three principal areas ofengineering competence at Lancaster: color,power, and conversion tubes. The engineers work-ing at the image isocon test set are E. M. Mussel -man (seated) and 0. Choi, both of RCA Lancaster.Photo credit: Tom Cook, RCA Laboratories.

Contents

Papers

Copyright 1969 RCA CorporationAll Rights Reserved

Milestones in color -picture -tube development H. R. Seelen 2

Development of cathodoluminescent phosphors A. L. Smith 5

Gases and getters in color picture tubesDr. J. C. Turnbull, Dr. J. J. Moscony, A. Month, J. R. Hale 10

Colorimetry and contrast performance of color picture tubes G. M. Ehemann, W. G. Rudy 14

Autotest-automatic production -test and process -control system for color picture tubesW. E. BahIs, F. C. Fryburg, A. C. Grover, J. F. Stewart, N. A. Teixeira, E. D. Wyant

18

Vidicon photoconductors J. F. Heagy 22

The image isocon-an improved image orthicon E. M. Musselman, R. L. VanAsselt 24

Ceramic -metal tetrodes for distributed amplifier service C. E. Doner, W. R. Weyant 29

Klystron for the Stanford two-mile linear accelerator A. C. Grimm, F. G. Hammersand 32

Environmental engineering laboratory J. M. Forman, J. B. Grosh 36

Design of a 915 -MHz power triode for microwave cookingW. P. Bennett, D. R. Carter, I. E. Martin, F. W. Peterson, J. D. Stabley, D. R. Trout 38

RCA Lancaster -25 years of engineering excellence 43

The heat pipe, an unusual thermal device R. A. Freggens, R. C. Turner 44

Ruggedization of camera tubes for space applications J. G. Ziedonis 48

The Coaxitron J. A. Eshleman, B. B. Adams 52

Noble -gas -ion lasers R. J. Buzzard, J. A. Powell, J. T. Mark, H. E. Medsger 56

Design of cermolox tubes for single-sideband A. Bazarian 60

A sterilizable and ruggedized vidicon Dr. S. A. Ochs, F. D. Marschka 64

Broadband, high -gain, L -band, power amplifier module R. L. Bailey, J. R. Jasinski 68

Milestones in color -picture -tube developmentH. R. Seelen

Frequently a company, however heterogeneous its output, becomes identified in themind of the public with a single, best-known product. The RCA Lancaster plant'slife -span has encompassed many developments of high interest to scientists, engi-neers, the military services, and ordinary laymen. But of all the Lancaster products,the public is most aware of television picture tubes, and, in particular, color tubes.The 19 years during which Lancaster engineers have been engaged in developingand producing color tubes has been an eventful period.

Harry R. SeelenDivision Vice President and General ManagerTelevision Picture Tube DivisionElectronic Componentsreceived the BS in Physics from Providence Col-lege, and in June, 1968, Mr. Seelen received anHonorary Doctor of Business Administration de-gree from it. Mr. Seelen was appointed DivisionVice President and General Manager, TelevisionPicture Tube Division, of Electronic Componentson August 20, 1965. Previously, he had been Gen-eral Manager of the division since July 20, 1965.Mr. Seelen joined RCA in 1930 as a tube designengineer at the Harrison, New Jersey, plant. Sub-sequently he became Manager of the Tube Devel-opment Shop. In 1942, he organized and set upthe Engineering organization and laboratories atthe new Lancaster plant. In 1943, with the transferof all non -receiving tube engineering to Lancaster,he assumed charge of Engineering Services.Seven years later, he was appointed Manager of allLaboratory Engineering at Lancaster. In 1954, Mr.Seelen became Manager of the Color KinescopeOperations Department and later returned toHarrison as Manager, Kinescope Operations. In1963, he was appointed Manager, Television Pic -tube Tube Operations Department of the RCATelevision Picture Tube Division, until his promo-tion to General Manager of the division. He is aFellow of the IEEE and a Registered ProfessionalEngineer in the State of New Jersey. He is a mem-ber of the Association for the Advancement of Artsand Sciencies, as well as Sigma Pi Sigma, hon-orary engineering fraternity. In 1955, he receivedthe RCA Victor Award of Merit, the company'shighest tribute.

Final manuscript received October 10, 1968.

'NROM THE BEGINNING, effort hasr been applied principally to theshadow -mask type, which appearedmost suitable for development of theseveral forms of color tubes investi-gated by the RCA Laboratories. Weare often asked, today, why RCA hasadhered steadfastly over the years tothis tube. The answer has good tech-nical and commercial foundations: noother type has shown so desirable acombination of good performancecharacteristics and mass -productionsuitability, brought about by a longseries of engineering developments.Admittedly, the three -beam shadow -mask system is less efficient than othersystems as regards light output, and ithas convergence complexities. But itsadvantages are dominant: high con-trast and resolution, capabilities foruniform color fields and high colorsaturation (which is built into thetube) , fine dot -screen structure, andmoderate circuit requirements, consid-ering both complexity and stability.

Pioneering work in color Tv had beendone by RCA prior to World War II,but the major advances in the com-mercial program occurred after 1945.An important milestone was passed in1949 when field tests began and thefirst demonstrations of the RCA com-patible system were held for the Fed-eral Communications Commission. Forthis early work, the receivers used aseparate picture tube for producingeach of the three primary colors, andthe three pictures were combinedoptically.

The next big step, then, was to providea single, directly viewed tri-color tube.Of prime importance at this time wasthe fabrication of a group of such tubes

to aid continued development of theRCA color TV system as a whole, andto permit additional demonstrations ofthe system at various stages beginningin the Spring of 1950. Ultimately, thesystem developed by RCA served asthe "backbone" of the National Tele-vision Systems Committee recommen-dations, which were finally adoptedby the F.C.C. late in 1953. RCA thusbecame the major contributor to for-mulation of the present national stand-ards for color TV transmission.

Early color tubesIn addition to furnishing tubes for theRCA color program, it was, of course,an important part of Lancaster's as-signment to develop the color -tubedesign, with a view to eventual massproduction. The approach followedduring this early period beginning in1949 was to explore the potentials ofthe flat -screen, flat -mask system. Be-cause accurate positioning of maskapertures with respect to the phosphor -screen dots (register) is extremely im-portant, great emphasis was placed onmounting arrangements for the thin,rather flexible, metal mask. In this con-nection it was felt that tensioning themask in a frame would provide great-est accuracy and stability of apertureposition. As a correlative advantage, aflat screen opened up various possi-bilities for high-speed phosphor appli-cation: initially, silk-screening; later,letter -press or offset methods. Theseapproaches to fabrication of mask as-sembly and screen required that allmask -aperture arrays be similar, andall phosphor -dot arrays be alike, withinthe narrow tolerances permitted by theshadow -mask system's geometry. Inother words, there was a need for

2

"interchangeability"; any mask should"fit" any screen. Another advantage ofusing the flat mask -screen assemblywas the feasibility of adjusting relativepositions of these parts for best reg-ister, as a final step before insertioninto the bulbs. Many acceptable tubesof the flat mask -screen type were made.The 15GP22 was put in production in1953, and a larger 19 -inch tube of thesame type was developed.

Curved screen tubesMeanwhile, various curved -screen,curved -mask proposals had been con-sidered. It was fully realized that aphosphor screen on the face of thebulb would be desirable because ofbetter appearance, a larger picture inthe same size bulb, and several otherfactors; but it was also realized thatinterchangeability might be difficult toattain if curved masks and screenswere used. Furthermore, an inter-changeable system seemed moresuitable for mass production; non -interchangeability would require thateach mask be mated to a particularscreen, through all of the many proc-essing steps. In addition, a problemwould arise if the curved mask didnot remain sufficiently stable throughthe heat treatments to which a vacuumtube must be subjected duringfabrication.

In the fall of 1953, another tube manu-facturer announced and demonstrateda tube of the curved -mask, curved -screen type. Although this tube evi-denced register problems, it gave someindication that a curved mask mightbe reasonably stable through tube proc-essing. RCA decided to investigatethis type more thoroughly.

Earlier work had been done in theRCA Laboratories on a photographicmethod for depositing phosphor dots,a method well suited for screen appli-cation to a curved surface. With thisbackground, the Lancaster programfor design and construction of curved -screen color tubes was initiated late in1953.

Among the first of these tubes wereboth round and rectangular samples.The round tubes used 19 -inch -diameterbulbs that had a final closure of weldedmetal flanges presealed to the face -panel and the funnel. For the all -glassrectangular tubes, standard 21 -inch, 90 -

degree bulbs were cut near the panelseal, and, after screen deposition andmask mounting, were resealed with alow -temperature glass frit.

These first tests were so encouragingthat, within the next few months, anintensive developmental effort wasbeing placed on curved -mask tubes. A21 -inch size was selected because of thepopularity of this size in black -and -white picture tubes; a round shapewas chosen for reasons of stability ofbulb and mask assembly, and a metalenvelope was used for flexibility indesign. From this developmentevolved the 21AXP22, which went intoproduction in the fall of 1954. Thistube incorporated many changes inaddition to the curvature of mask andscreen. It not only provided a largerscreen (260 square inches instead of88 -1/2 -square -inch size of the 15GP22)but did so at a slightly decreased over-all length by widening the deflectionangle from 45 to 70 degrees.

Photographic application of the threephosphors to the curved face plateinvolved development of processingtechniques for production use, as wellas design and development of a specialequipment called a "lighthouse". Thepurpose of this equipment is to provideexposure of the photosensitive phos-phor layers in arrays of "dots" properlylocated so that they are bombarded bythe respective electron beams whenthe tube is completed and in operation.An innovation, called a "radial cor-rection lens", was made in the light-house to provide a correction byoptical simulation of the electronpaths. This compensation is neededbecause of the manner in which thedeflection field acts on an electronbeam, effectively changing its "source"position with increasing deflectionangle.

The photographic method of phosphordeposition required development of amask -frame assembly which could beremoved from its normal position with-in the front end of the tube or "top -cap" and accurately replaced severaltimes. Stability of the curved maskitself was ensured by mounting on arigid frame; repeated, precise position-ing of the mask -frame assembly in thetop -cap was attained by special leaf -springs attached to the frame whichengaged studs on the cap wall.

In addition, a new triple -beam gunstructure, which had originally beendeveloped for the flat -screen tubes, wasintroduced in the 21AXP22. This gunincluded magnetic pole pieces for con-vergence control of individual beams.In this assembly, the three guns weretilted slightly toward the axis to pro-vide static convergence at the screencenter; dynamic fields, in synchronismwith the scanning frequencies, couldthen be applied to the pole pieces toprovide the change in convergenceangle needed as the beams were de-flected. This development permittedaccurate convergence to be obtainedeven at the much wider deflectionangle.

The glass tube

During and following this period, bothround and rectangular glass bulbs wereinvestigated to determine whether glasswas potentially desirable for eithertechnical or commercial reasons, andto assess the problems connected witha rectangular shape. Some of the earlyglass bulbs using metal flanges havealready been mentioned. A major im-provement relating to glass bulbs oc-curred with the introduction of aspecial frit glass developed by CorningGlass Works which devitrifies, duringsealing of the "top -cap" to the funnel,at a temperature low enough not toharm the phosphor or internal parts.

This development was followed by theintroduction, in 1957, of the first colortube in an all -glass bulb, designatedtype 21CYP22. For screening this tube,a "degrouping correction lens" wasintroduced. This lens has an asymmet-ric contour superimposed on the radialcorrection contour to compensate forbeam degrouping incidental to dy-namic convergence. This latter innova-tion permitted enlargement of the maskapertures at the center, and increasedtransmission by about 40%. It also per-mitted a graduated tapering of theaperture walls which effectively in-creased transmission about 20%.

As mask material, cold rolled steel wassubstituted for the expensive copper -nickel alloy previously used. The aper-ture walls were tapered in such a waythat only "knife edges" could be struckby the beams. This modification re-duced electron scattering and conse-quently improved contrast.

3

The glass bulb proved advantageousin several ways. Its insulating proper-ties simplified tube mounting problemsfor the receiver designer. Its mechan-ical stability and uniformity of facecontour provided very significant im-provement in register. Moreover, animproved referencing system for thelighthouse and for the frit-sealing oper-ation also contributed to consistentlygood results. The mask -mountingmethod originally developed for the21AXP22 was well adapted to pro-vide the accuracy needed for thissystem. All of these features whichprovided stability and accuracy con-tributed significantly to improved uni-formity of both color fields and white,and to greater ease of setup in thereceiver. These improvements werethe keys to initial commercial successin the fabrication of shadow -masktubes.

In late 1960, another major improve-ment was introduced in the tube type21FBP22. New green and red (sulfide)phosphors, having higher efficienciesand shorter persistence, supplantedthose formerly used. The advantagesincluded an important enhancement inpicture brightness (about 50%) , free-dom from "smear" in rapid -actionscenes, and better balance between thecurrents required from the three gunsto produce white. Thus, the all -sulfidephosphor screen permitted driving tohigher white -light output without seri-ous red halo. The improvement in cur-rent balance also simplified receiversetup, particularly the adjustment ofvideo drive for black -and -white picturereproduction. In addition, a protectivewindow of 61% transmission filterglass was laminated to the tube facewith a clear resin to improve contrastand color saturation.

Today's tubes

The last few years have seen the de-velopment of the first RCA 90 -degreerectangular tube, the 25AP22, and theproliferation of its brothers and smallercousins. The rectangular bulb shape,and particularly the higher deflectionangle, have introduced additional prob-lems requiring a higher degree ofengineering sophistication for satisfac-tory solutions. Use of the rectangularbulb has required solution to prob-lems of bulb stability and alignment,as well as development of completely

new approaches to design and mount-ing of the mask assembly. Frameweight, replaceability, and thermalstability were among the factors whichhad to be reconsidered. A simple, inex-pensive, lightweight assembly per-mitting good beam -to -screen registercontrol has been developed for theentire tube family. The control of reg-ister and contrast has established astandard which has been or is nowbeing emulated by our competitors.

Important advances have also beenmade in the art and science of light-house lens design, aided by use of acomputer. We have been able to de-posit phosphor dots to a higher degreeof precision than heretofore thoughtpossible. The increase in radial mis-register brought about by the increasein deflection angle and the greater non -uniformity caused by factors of paneldeformation and panel obliquity haverequired greater control. Also, higherbeam degrouping factors broughtabout by the increase in deflectionangle and the greater nonuniformity ofthese factors as a result of yoke char-acteristics also required higher degreesof control. Optimum designs of maskand lens contours have been developedfor such control. We have indeed comea long way from the initial simpleradial lens on a path of progress whichhas been vital to the success of wide-angle tubes.

Another innovation was the introduc-tion of smaller neck size to permit useof smaller -diameter yokes. This changeincreased the efficiency of scanningneeded for the larger deflection angle,and also reduced beam separation. thusminimizing misconvergence. Develop-ment of the small gun to fit the smallneck required further improvementsin beam formation and focus to main-tain the quality performance of thelarger gun used in the round tubes.Low -wattage heaters were developedto reduce thermal problems in the stemarea. As a result, movement of partsbecause of thermal expansion has beenreduced and convergence drift hasbeen minimized. Einzel-lens guns havealso been developed in the small sizewhich permit fixed focus and thussimplify circuitry in portable receivers.

The development of high -efficiency red -emitting phosphors has been extremelyimportant.The red sulfide phosphor wassucceeded by yttrium vanadate in 1964

and by yttrium oxysulfide 2 years later.The yttrium-oxysulfide red -phosphorefficiency has made the "Unity CurrentRatio" mode of operation a reality inthe color picture tube. No longer mustbrightness be limited by red -gun"blooming". Now all guns may beuniformly driven to their resolutioncapability to provide maximum gun -screen efficiency and thus a brighterpicture. Because of this permitted in-crease in current to the screen, themask -frame thermal -compensating sys-tem called "Perma-Chrome" has be-come a necessity. Brightness has beengained without sacrificing resolution,purity, or white uniformity.

The success of these tube develop-ments has literally echoed around theworld. Our licensees in Japan andWestern Europe have been tooling upto make similar tubes, and RCA itselfhas started up new tube plants inScranton, Pa., Midland, Ontario, andSkelmersdale, England; in addition,our RCA plant in Puerto Rico is sup-plying gun assemblies.

Color picture devices in the futureWe continue optimistic about the po-tential of the shadow -mask tube forfurther development, and we expect toexploit ideas not yet fully explored.People frequently ask, "Is the shadow -mask tube likely to be supplanted bysome new device in the near future?"Any new device would have to showat least as good performance, and itwould probably require novelty (suchas a large, thin panel to be hung on thewall) , but far more important thanany physical qualities would be thepotential for cost reduction. A replace-ment for the shadow -mask tube whichwould meet such criteria is not pres-ently within our field of view; we feel,therefore, that advanced versions ofthe shadow -mask tube will be servingus for a number of years to come.

Gen. David Sarnoff's words at thefirst tri-color tube demonstration onMarch 29,1950, still ring true:

"Measured in comparison with everymajor development in radio and tele-vision over the past 50 years, this colortube will take its place in the annals oftelevision as a revolutionary and epoch-making invention. When historians atthe close of the 20th century evaluatethe most important scientific develop-ments, I will predict that this tube willbe among the great inventions of thesecond half of this century."

4

Development ofcathodoluminescentphosphorsDr. A. L. Smith

Phosphors are solid materials that have the ability to convert one or more forms ofenergy into visible or near -visible radiation: luminescence is the generic term for thephenomenon of this conversion. There are a number of excellent books and reviewarticles covering the theoretical aspects of luminescence and the various phosphorsystems that have been studied. Some of these treat cathodoluminescence indetail,' 21° and one is written with practical commercial development problems asits main theme.' None of them, however, deals specifically with the problems asso-ciated with the commercial development of cathodoluminescent materials, the sub-ject of this article.

HOSPHORS have been commerciallyr important for a much longer periodof time than other electronically activesolids." However, despite the longhistory associated with them and thevery extensive literature on the subjectof luminescence, their theoretical basisis not nearly as well understood as thatof the more recently developed semi-conductors, such as transistors. Thingshave not changed much since an intro-ductory paper presented in 1954 statedthat the manufacture of phosphorsis largely an experimental science, ifnot a craft.' Research, development,and technology of phosphors is todaystill a combination of art, intuition,and fundamental knowledge.

RCA's present interest in luminescenceis centered primarily on cathodo-luminescent phosphors, i.e., those ma-terials that have the ability to trans-form the energy of an electron beaminto visible or ultraviolet radiation,because this energy transformation isthe basis of RCA's extensive cathode-ray -tube business. Color -televisiontubes provide by far the greatest dollarvolume in the tube market; for thisreason, phosphor research and devel-opment done by the Materials Groupat Lancaster is slanted heavily towardsupport of this item.

A large number of compounds areknown to be luminescent, but thespecific requirement thay they beeconomically useful in a cathode-raydevice immediately imposes a series of

Final manuscript received October 10, 1968.

design parameters which narrows thefield of investigation to a surprisinglyfew chemical systems. These stringentdesign requirements also make theinvention of a new commercial phos-phor extremely difficult, for not onlymust the proposed material have ex-ceptional luminescent properties, butit must also possess the combinationof chemical and phsyical propertieswhich will make it suitable as a screenmaterial for cathode-ray tubes. Al-though characteristics of a phosphorcan be varied within moderate limits,there is usually an interaction betweenthe variables so that as one is im-proved, another is degraded. Theresult is that any phosphor representsa compromise in which the variableconsidered most important in a givenapplication is optimized.

Design requirementsThe design requirements of commer-cial cathodoluminescent phosphorscan be divided into two broad cate-gories; those dealing with luminescentcharacteristics and those dealing withphysical -chemical properties. The firstconsiders phosphors simply as energyconverters and sets their operationalrequirements in a particular device.The second category views phosphorsas tube components which must re-main stable during the manufacturingprocess.

Luminescent characteristics

A phosphor must convert the energyof an electron beam into emitted ra-diation efficiently. This requirement

Dr. A. L. Smith, Ldr.Phosphor DevelopmentChemical and Physical LaboratoryTelevision Picture Tube DivisionElectronic ComponentsLancaster, Pa.received the BS in chemistry from FordhamUniversity in 1941 mid the MS and PhD from thePolytechnic Institute of Brooklyn in 1943 and1946 respectively. He joined the Chemical andPhysical Laboratory at the RCA Lancaster plartin 1945. The major part of his career has been inthe capacity of Engineering Leader of the PhosphorDevelopment Group. He is the author of a numberof papers on phosphors and has been grantedfourteen patents related to phosphors and colortube screens. Memoerships include the AmericanChemical Society, Phi Lambda Upsilon, Sigma Xiand the Electrochemical Society. He has heldvarious positions in the ECS, being Chairman ofthe Electronics Division in 1953 and a presidentialcandidate in 1955.

5

may seem obvious, but there aremany phosphors that are efficientlyexcited by ultraviolet radiation but notby an electron beam and therefore areuseless in cathode-ray tubes. On theother hand, there are literally thou-sands of chemical combinations thatluminesce under electron bombard-ment but have an efficiency of conver-sion of electron -beam energy intolight that is too low to make themcommercially useful. The most effi-cient and practical cathodoluminescentphosphors have been estimated toconvert, at best, only 20% of beampower into radiant energy". Some-times the tube designer is willing toaccept less than optimum conversionefficiency if some other characteristic,such as a unique decay property, ishis major concern. Under any circum-stance, however, conversion efficiencyalways has a high priority among therequirements. Unfortunately, thereis little theoretical knowledge to guidethe optimization of this parameterbecause the mechanism by whichbeam energy is dissipated in a crystaland transformed by a luminescentcenter into visible emission is butsuperficially understood. The closingof this knowledge gap is, today, thegreatest challenge in the field of phos-phor development. An advance in in-formation in this area could revolu-tionize cathodoluminescence.

Another luminescent characteristicrequired of a phosphor is that it emitin some predetermined portion of thespectrum. Cathode-ray tubes are usedin a wide variety of applications, andthe phosphors used in them must emitin an area of the spectrum suitable toa specific application.

Although most tubes manufactured areused in the black -and -white or colortelevision entertainment area, there are

a number of industrial applications inwhich screens are coupled to a par-ticular photographic film type orphotosensitive surface. In these uses,the output of the phosphor is tailoredto match, as closely as possible, theresponse characteristics of the system.Because the phosphors developed byRCA are used principally for colortelevision, the major interest is inrather narrow regions of the spectrumcomprising the three primaries usedin color TV: red, blue, and green. Col-orimetry of the phosphors is of para-mount importance because the gamutof hues obtainable in a color pictureis defined by the coordinates of thethree primaries" ".

In addition to efficient energy conver-sion and emission, the persistence(the phosphorescence or decay rate)of a phosphor must be of the propermagnitude. The requirements in thiscategory vary widely. In tubes designedfor industrial or military applications,the range may be from 10' second(or less) to several seconds duration,depending on end use. A medium -short persistence is desirable for en-tertainment tube types because thedecay must be fast enough to insurethat there is no smearing of the imagein rapid action, yet slow enough sothat flicker does not become objection-able. In multiple -phosphor screens,such as those used in color televisiontubes, the persistence of the threephosphors must be reasonably matchedor color trailing (image blur) becomesnoticeable.

It is desirable that the light output ofeach phosphor in a multi -color screenvary linearly with power, and eachwith the same slope, so that high andlow brightness areas of a picture willbe properly shaded. Because the totaloutliut of the screen is dependent onbeam power (the product of beamvoltage and beam current) it is im-portant that the voltage and currentcharacteristic of each of the phosphorsbe stable. As the voltage of the tube isusually fixed, the variation in outputcharacteristics of the phosphor withvariable current becomes the majorconcern. In some systems, such as sul-fides, the light output of the phosphordoes not always increase with currentin the linear manner desired, but be-gins to drop off at some intermediateoperating current. This phenomenon

Phosphor slurrybeing dispensed.

is known as current saturation, or"droop -on -drive".

Another equally undesirable charac-teristic is known as color shift, achange of hue of emission with in-creased current. The higher the cur-rent, the more pronounced is thecolor shift. Color shift manifests itselfin the appearance of a new, spuriousemission band (usually at lower wave-lengths) superimposed on the desiredemission band. The phenomenon fre-quently occurs simultaneously withcurrent saturation of the main band.Both main -band saturation and colorshift are detrimental to color picturetube quality because they noticeablydistort the hue in the highlights of thepicture.

Physical -chemical characteristics

Commercially, phosphors are appliedto a tube faceplate to form a screenby a variety of application techniques.Settling, slurrying, and dusting are themethods commonly employed. Themethod used dictates the average par-ticle size and distribution of the phos-phors to be used. Ideally, the phosphorchemist should develop preparationprocesses that are flexible enough toproduce the variety of particle sizesrequired for the different applicationtechniques. This condition is fre-quently difficult to achieve, however,because the high temperatures requiredto develop optimum luminescent char-acteristics simultaneously causescrystal growth. The phosphor develop-mental engineer must therefore makesuitable compromises to obtain thebrightest material that can be appliedin a reasonable manner.

A phosphor must also be chemicallycompatible with the application media.Most application methods involveaqueous systems which are pH sensi-tive. The phosphor must, therefore, bestable in water to the extent that itdoes not decompose to form basic oracidic constituents. Such constituentswould alter the properties of the slurryformulation and its application charac-teristics. Of course, the phosphorshould be inert so that its intrinsicluminescent properties are preserved.If the phosphor is not inert, it willsuffer severe loss of efficiency duringtube processing.

The surface properties of a phosphormust fit the screening techniques so

that the phosphor can be well dis-persed in the medium, yet bond wellto the glass substrate. A variety of"coatings," which might be thought ofas a "cement", is designed to facilitatebonding in the particular applicationtechnique used. Colloidal silica, sili-cates, and phosphates are the mostfrequently used chemical coatings, pre-sumably because they form surfaceswhich bond well to glass, the mostcommon substrate to which phosphorsare applied. The laws governing sur-face properties are at present not com-pletely understood.

Because phosphors are used in avacuum, they must not decompose orevaporate during processing or underelectron bombardment. Organic lumi-nescent materials cannot be usedbecause of their instability underthese conditions. Their decompositionproducts could then poison the elec-tron -emitting cathode or raise the gaspressures that would destroy theusefulness of the tube. Even when amaterial has a vapor pressure lowenough to withstand a vacuum, theenergy of the electron beam itself cancause crystal changes that destroy theluminescence of the materials. Halidephosphors, particularly the fluorides,have notoriously poor tube life be-cause the chemical reducing power ofthe beam causes permanent crystaldamage and ultimately destroys screenefficiency.

A phosphor must be able to maintainits important physical properties dur-ing tube processing; i.e., it mustdemonstrate chemical, thermal, andvacuum stability. During outgassing,a tube is subjected to temperatures ofabout 400 to 450°C. At these tempera-tures, the organic binders used inscreen deposition are decomposed andyield a combination of gaseous re-action products, including CO, COand H20. Although the phosphor maybe unaffected by these individualgases at room temperature and atmos-pheric pressure, the combination ofgases and elevated temperature couldprove disastrous.

Phosphor cost

One of the more important commericalconsiderations not yet mentioned isphosphor cost. It must be possible tomanufacture a phosphor in a repro-ducible manner at a reasonable cost;

a rather high unit cost can be toleratedprovided some outstanding character-istic can be obtained. The red phos-phor now used in RCA's color tubesis a case in point. It costs substantiallymore per tube than any previous phos-phor but its outstanding efficiencymakes the added cost worthwhile intube and set performance.

Group organization andresponsibilities

The RCA engineering group devotedto the achievement of the phosphorcharacteristics described above, theMaterials Group, is located in Lan-caster, Pa. Its activities range fromApplied Research through Develop-ment and into Pilot Plant production.When necessary, factory assistance isgiven and close liaison exists betweenLaboratory and Factory Engineeringpersonnel.

Laboratory engineering

Laboratory Engineering has the abilityto follow an idea from its basic con-ception through factory productionand, in addition, makes its staff avail-able to Marketing and Sales as fieldengineers. Laboratory Engineeringthen is involved in the whole gamutof the business enterprise. The generalphilosophy of the Laboratory Engi-neering group is to concentrate onphosphor technology and to leavetesting and analyses to other groupsmore knowledgeable in those areas.Laboratory Engineering does, how-ever, draw heavily on these supportingspecialists and the well -instrumentedlaboratories available to them.

Analytical group

The Analytical Group performs a widevariety of services. Of particular noteare the spectrographic and X-ray dif-fraction work which is relied uponheavily in purity and compositionalstudies of phosphors. Particle sizeanalysis is another function of con-siderable importance. Because somework involves the interaction of manyvariables, the aid of a statistician isenlisted occasionally in setting andinterpreting statistically designed ex-periments.

Colorimetry group

The Colorimetry Group is essential tothe phosphor development operation,for it is they who make the necessary

Microscopic screen inspection.

efficiency, persistence, and color -coordinate measurements.

Applied research

The worker in the phosphor field isoften accused of alchemy because hesometimes uses rather unorthodoxmethods to achieve rather unusual re-sults. His background should be ineither inorganic or physical chemistryand, ideally, should include a verybroad spectrum of experience; a sur-prising number of disciplines are nec-essary in the successful developmentof a commercial phosphor. (A senseof humor has also been known tohelp.)

The role of Applied Research inphosphor development is twofold.First, its efforts are directed towardthe achievement of a basic understand-ing of those processes which are usedin synthesizing the blue and green -emitting sulfide phosphors. Second, itcarries on a search for new phosphorsystems and new ways of using oldersystems. The approach to each typeof research is quite different.

Efforts in the first area might be calledpurely scientific in that the approach,at least, can be well organized accord-ing to basic principles. Because theblue and green -emitters are sulfidesprecipitated from aqueous solutions,such fundamental properties as thekinetics of the precipitation and therole of the precipitating conditions onthe final characteristics of the phos-

7

Inserting the shadow mask after phosphor application.

phor are being studied. The influenceof the flux and flux systems, includingphase diagram studies, and diffusionas it affects particle growth, are alsounder investigation. Much of the workin this first area is conducted alongclassical lines where chemical theoryis well developed and mathematicalmodels exist. All theory and modelsmust of course be interpreted with aview toward fulfilling present needs.Note that the researcher must be con-versant with a wide variety of chem-ical knowledge, for he deals withsolution chemistry as well as withhigh -temperature and solid-state chem-istry. His life is orderly though com-plex. He seeks new knowledge, butmore or less within the framework ofexisting procedure.

The second area of research, that ofnew systems, is quite different. Here,intuition and art are as important asfundamental knowledge. There is noguiding theory and much of the workis empirical. The chemist's work inthis area is frustrating in large partbecause thousands of samples may beprepared before one of even moderatepromise is found. Many of the chem-ical systems have been worked andreworked, not only within RCA, butin a large number of laboratoriesthroughout the world. Statistically,the chances of finding a completelynew phosphor are exceedingly low.The researcher must not only considera compound as such, but he must beconcerned about the way that com-pound is synthesized, for often successor failure is due to some unique pre-paratory scheme which imparts justthe right amount of crystalline irreg-ularity necessary for an efficient phos-phor. Consideration must also begiven to the "antique" compounds,

8

known to be luminescent, but nevertested under modern cathode-rayconditions.

Development

The job of the chemist who devolepsphosphors is reasonably straight-forward; he develops new phosphorsand their manufacturing processesto a fine degree. His interaction withother groups is much more variedthan the research man and includesan occasional assignment as a tech-nical field representative with Salesand Marketing. (RCA enjoys a verysubstantial share of the market forzinc and zinc -cadmium sulfide phos-phors used in cathode-ray tubesthroughout the industry.) His knowl-edge must be very broad in the fieldof inorganic physical chemistry be-cause he tackles problems as diverseas ultra -purification of materials, high -temperature syntheses, control of crys-tal growth, and surface properties ofmaterials. It is generally he who mustattempt to satisfy all the criteria listedin the earlier part of this article, or atleast determine the optimumcompromise.

Pilot plant

The Pilot Plant has proven to be aninvaluable aid in the phosphor devel-opment operation in that it performsa variety of functions including scale-up of developmental procedures, roughcost analyses, manufacture of phos-phors used in small quantities in eitherlab or factory, and investigationsleading to new or improved processes.The equipment used in the Pilot Plantmore closely approximates factorysize than does the lab equipment,therefore, such items as firing time and

Table I-Commercially used phosphors

temperature are developed best inthis plant. The engineering in the PilotPlant is then true chemical engineeringas distinguished from the pure chem-istry of the other groups. Pilot Plantpersonnel must have a knowledge ofproduction -type equipment and itscapabilities, a talent for process sim-plification, and a healthy respect forcosts. It is up to them to render aphosphor process practical from theviewpoint of production. Frequently,the Pilot Plant has been the initialproduction unit for a new materialand has sold its product to the tubefactory. In the early stages of phosphordevelopment, reasonable cost estimatescan be made, scrap potentials dis-covered and corrected, and processreproducibility assessed. Interactionwith other groups is, of course, great-est with the development engineersand factory personnel, but some liai-son with Applied Research is alsonecessary. Vendor relationshipsthrough Purchasing are also important,not only in equipment areas, but inmaterials as well. Prior to acquisitionof the Pilot Plant, the Phosphor Fac-tory itself had to do its own scale -upand process development. This pro-cedure was undesirable and led tomuch delay because tests had to besqueeezd into existing productionschedules.

Technical progressTable I is a modified, up -dated ver-sion of previously published data"'.The listings within each color arechronological in time of commercialcolor -television usage and show howthe improvement in phosphors hascontributed to the advances in colortube performance.

Types of emitters

Blue

Phosphor notation Powder colorimetric data'x y Y (Lumens/W)

Calcium Magnesium Silicate :Titanium CaO: MgO: 2S102: Ti 0.169 0.134 8.7Zinc Sulfide: Silver: Chloride= ZnS:Ag:CI 0.146 0.052 7.5Zinc Sulfide: Silver: Chloride, ZnS:Ag:C1 0.150 0.059 9.1GreenZinc Silicate : Manganese 2ZnO SiO2 : Mn 0.218 0.712 31.1Zinc Cadmium Sulfide: Silver: Chloride' (ZnCd)S : Ag : CI 0.242 0.529 56.0Zinc Cadmium Sulfide: Silver: Chloride' (ZnCd)S Ag : CI 0.303 0.587 70.3RedCadium Borate : Manganese 2Cd0 :B203: Mn 0.630 0.370 10.7Zinc Phosphate: Manganese fl3ZnO: PAN: Mn 0.674 0.326 7.0Zinc Selenide : Copper ZnSe :Cu 0.652 0.347 17.0Zinc Cadmium Selenide: Copper: Chloride (ZnCd)Se: Cu: CI 0.662 0.338 11.0Zinc Cadmium Sulfide: Silver: Chloride' (ZnCd)S : Ag : CI 0.663 0.337 12.6Yttrium Vanadate : Europium YV04:Eu 0.675 0.325 9.5Yttrium Oxysulfide : Europium Y302S Eu 0.660 0.340 13.8

I. This notation defines color in accordance with that established by the Commission Internationalede I'Eclairage (C.I.E.).

2. The differences between these two phosphors are in silver content, flux composition, and firingtemperature.

3. These phosphors differ primarily in their cadmium content.

Early phosphors

The listing under the color red illus-trates how RCA's research and devel-opment activity operates in seekingto optimize all design specifications.The first phosphor on the list -cad-mium borate: manganese - satisfiedall requirements except one of themost important: quality of emission.The emission of the borate phosphorwas much too orange. For that reasonit was relpaced by zinc orthophos-phate: manganese whose color wasideal. The lower lumens/watt value ofthe phosphate results from its reddercolor and a somewhat lower intrinsicconversion efficiency, a severe handi-cap that led to much research to find areplacement. The phosphate was com-mercially acceptable, however, andwas used as RCA's standard red inthose early years when color was try-ing to get off the ground.

Zinc selenide: copper

The next phosphor to receive consider-able attention was zinc selenide: cop-per. This phosphor is almost a classicexample of one whose initial promisewas very high but which was subse-quently found to lack many of the re-quired characteristics. Zinc selenide:copper has an outstanding lumens/watt value, although its hue is a bittoo orange.

Zinc cadmium selenide: copper

An improved red emission is obtainedby addition of cadmium selenide toform solid solutions of zinc cadmiumselenide. When this is done, however,the lumens/watt value of the com-pound is significantly decreased; thepercentage decrease is greater thanshould be expected on the basis ofcolor change alone. It has been con-cluded, therefore, that the solid solu-tion has an intrinsic lower conversionefficiency than zinc selenide alone. Al-though disappointing in some charac-teristics, the zinc cadmium selenide hasa final lumens/watt value that is stillmuch higher than the phosphate it wasdesigned to replace. A major disad-vantage of zinc -cadmium selenide is itsprocess instability; in water, at roomtemperature, a slow decomposition oc-curs. At the elevated temperatures oftube and screen bakes, the water vaporevolved during decomposition of theorganic binder attacks the phosphor atsuch an accelerated rate that an effi-

ciency loss of approximately fifty percent occurs. Although a coating capa-ble of slowing decomposition wasdeveloped, the close controls requiredin both manufacture and screen proc-essing forced RCA to drop zinc -cadmium selenide as a commercialproduct.

Silver -activated zinc -cadmium sulfide

A silver -activated zinc cadmium sul-fide of high cadmium content eventu-ally replaced the zinc orthophosphate:manganese, primarily because of itshigher conversion efficiency, but alsobecause its persistence matched that ofthe other two sulfides more exactly.All of the brightness gain indicated inTable I was not realized because thesulfide suffered mild degradation dur-ing tube processing and showed aslight color shift on drive, a drawbacknot present in the phosphate. The plusfeatures of the sulfide were consideredto far outweigh its disadvantages, andfor many years it was the standard redcomponent in RCA color tubes.

Europium -activated yttrium vanadate

The europium -activated yttrium vana-date replaced the sulfide for reasonsnot readily apparent from Table I.Again, it was adopted on the basis ofa series of compromises, the sum totalof which made it superior to the sul-fide. One prime disadvantage was itscost, some ten times that of the sulfide.Advantages included no color shiftwith high current and no current satu-ration; persistence was in line with thegreen and blue sulfides. The lumens/watt values shown in Table I appearto put the vanadate at a disadvantagein relation to the sulfide. After process-ing into tubes, however, the vanadateshows a very slight improvement overthe sulfide, because the sulfide effi-ciency is degraded during processingwhile the vanadate is not. This is agood example of why the acceptanceof a phosphor must be based on itsperformance in the tube and not on asimple powder test.

Europium -activated yttrium ozysuifide

Shortly after the yttrium vanadate:europium went into production, RCA'sresearch efforts led to the discoveryof europium -activated yttrium oxysul-fide, a phosphor with a much higherefficiency than the vanadate. The oxy-sulfide is as stable as the vanadate in

all respects. has processing stability,shows no color shift on drive, and nocurrent saturation.

A new process compatible with factoryequipment had to be invented for theproduction of the oxysulfide (here thestaff of the David Sarnoff Laboratorieshelped considerably) . Firing condi-tions capable of yielding proper colorand particle size had to be developed,rigid control procedures established,and impurity levels discovered. ThePilot Plant went into production of thematerial to establish production rou-tine, and actually made many hun-dreds of pounds of product withouta single reject lot, to determine reli-ability, costs, and other commercialconsiderations.

Future improvementsFrom RCA's viewpoint, a phosphor isnot a commercial success until it hasperformed satisfactorily in a market-able cathode-ray tube. To perform sat-isfactorily a phosphor must fulfill twocategories of design parameters, onedealing with intrinsic luminescentproperties, the other with applicationcharacteristics. Future improvementsexclusive to RCA will become increas-ingly difficult because intense competi-tion has increased not only the numberof investigators throughout the world,.but the sophistication of the approachto the discovery of new phosphors.

References1. Garlick, G. F. I., Luminescent Materials (Ox-

ford Univ. Press. New York, 1949).2. Leverenz, H. W.. Introduction to Lumines-

cence of Solids (Wiley, New York, 1950).3. Goldberg, P., Luminescence of Inorganic

Solids (Academic Press, New York. 1966).4. Kaltman. H. P.. and Spruch. G. M.. Lumi-

nescence of Organic and Inorganic Materials(Wiley. New York, 1962).

5. Kroger, F. A.. Some Aspects of the Lumi-nescence of Solids (Elsevier, New York,1948).

6. Curie. D.. Luminescence in Crystals (trans-lated by G. F. ). Garlick, Wiley. New York,1960).

7. Leverenz, H. W.. "Luminescence" Encyclo-pedia Britannica (1966).

8. Palilla, F. C., Elect. Tech., Vol. 6 (1968)p. 39.

9. Ouweltjes, T. L.. Modern Materials (edited byB. W. Gonser, Volume 5, Academics Press.N.Y., 1965) pp. 161-257.

10. Garlick, G. F. J., Brit. I. Applied Physics,Vol. 13 (5962) p. 541.

11. Larach, S., Shrader, R. E.. Yocum. P. N..RCA reprints PE -276, PE -280. and PE -291.

12. Henderson, S. T., "Luminescence," Cam-bridge Symposia, Brit. I. Applied Phys. Vol.6, Supplement 4 (1955) p. 51.

13. Bril. A. and Klasens, H. A.. Philips Res.Rpts. Vol. 7 (1952) P. 401.

14. Bril, A., Klasens, H. A., Philips Res. Rpts.Vol. 10 (1955) p. 305.

15. Bril, A.. Wanmaker, W. L., Philips Tech.Rev. Vol. 27 (1966) p. 22.

16. Hardy, A. E., IEEE Transactions BTR 11,No. 2 (1965) p. 33.

9

Gases and getters incolor picture tubesDr. J. C. Turnbull Dr. J. J. Moscony A. Month and J. R Hale

Picture tubes, like other oxide -cathode electron -tube devices. require a high vacuumfor prolonged operating life. Unfortunately, the inside tube surfaces become gassources and destroy the desired vacuum. This problem results from high-energy elec-tron beams striking the tube screen and the shadow mask, in the case of color picturetubes, and liberating various gases from these surfaces. At the same time. high-energy,back -scattered primary electrons strike and liberate gas from the surface of the con-ductive funnel coating. The use of a barium getter in picture tubes forms a film onthe interior tube surfaces, which reacts chemically with stray gases to form stable.solid, barium compounds and thus maintain the desired vacuum. This paper dDscribesthe use of getters and their function in color picture tubes.

J. C. TurnbullColor TV Chemical and Physical LaboratoryElectronic ComponentsLancaster, Pa.received the BS in Physics from M.I.T. in 1934and the PhD in physics from Brown University in1938. Between 1938 and 1942 he worked on glassdevelopment and research at the Preston Labora-tories. During World War II. he was employed byM I.T., Princeton, and Harvard to work on highpower, high frequency transmitters and modulators.He joined Electronic Components in 1945 and fromthen through 1954 was involved in glass and ce-ramic work in the Chemical and Physical Laboratory.Thereafter, he was involved with the developmentof the metal kinescope and with the development ofboth the metal bulb and the glass bulb'versions ofthe shadow -mask color kinescopes. Since 1958 hehas been leader of the Applied Physics Group ofthe Chemical and Physical Lab investigating ther-mionic cathodes, getters and related phenomenawhich affect performance and life of electron tubes.

Final Manuscript received.

10

J. J. MosconyColor TV Chemical and Physical LaboratoryElectronic ComponentsLancaster, Pa.received the BS in Chemistry from St. Joseph'sCollege in 1951. He worked with the Electric Stor-age Battery Co. as an analytical chemist, servedtwo years in active duty with the U.S. Army andthen worked for Waterman Products Co. as aChemical Engineer in the development of phos-phor screening methods for cathode ray tubes. Hejoined RCA, Defense Electronics Products in 1957,where he assisted in the development of thermo-electric materials for refrigeration purposes. In 1958he received the MS in Chemistry from St. Joseph'sCollege. In 1959 he joined RCA Industrial TubeProducts, Space Components Engineerirg, wherehe worked on a variety of topics related to ther-mionic energy conversion. As a recipient of aDavid Sarnoff Fellowship, he attended the Univer-sity of Pennsylvania from 1962 to 1965 where heearned a PhD in Synthetic Inorganic Chemistry. Hethen joined the Color TV Chemical and PhysicalLaboratory of EC and has been working on massspectrometer analysis of residual gases in vacuumand also on getter and cathode developments. Dr.Moscony has had nine papers published and holdsone patent. He is a member of the AmericanChemical. American Physical and American Vac-uum Societies.

A. MonthColor TV Chemical and Physical LaboratoryElectronic ComponentsLancaster, Pa.received the BS in Chemistry -Physics in 1950 fromthe City College of New York and continued grad-uate studies in Physics at the Polytechnic Instituteof Brooklyn and CCNY. Prior to joining RCA hewas Engineering Leader of the Spectroscopy In-strument Section of the Process and InstrumentsCompany and then the Manager of the InstrumentModel and Glassblowing Shop of Emil GreinerCompany. He joined RCA in 1957 as a Spectro-scopist for the Microwave Chemical and PhysicalLaboratory in Harrison, N.J. In 1959 he assumedresponsibility for Analytical and MetallurgicalAnalysis and Physical Testing for the laboratory.He joined the Electron Physics Group in Lan-caster Color TV Chemical and Physical Labora-tory in 1964. At Lancaster he investigated thefactors affecting operating of color TV tubes andwas responsible for the use of the first high yieldexothermic getters in the color TV tubes. Mr. Monthreceived the EC Engineering Award in 1967 for hiswork in increasing the reliability and life of colorpicture tubes. He is presently engaged in thedesign and development of future color TV picturetubes. He is a member of the American ChemicalSociety and American Physical Society; hasauthored technical papers; and has several patentspending.

J. R. HaleColor TV Chemical and Physical LaboratoryElectronic ComponentsLancaster, Pa.received the BS degree from the New York Collegeof Ceramics at Alfred University in 1961. He wasemployed as a Development Engineer by the FerroCompany and worked on low temperature glassformulations through 1962. Currently, he is en-gaged with various glass technology problems,getter design, heater -cathode reliability problemsand investigations relative to solder glass technol-ogy. Mr. Hale is a member of the American Ce-ramic Society, The National Institute of CeramicEngineers, The Society of Glass Technology, andThe Society of Rheology.

THE GETTERS used in black -and -white picture tubes evolved from

the early loop getters which weremounted and flashed onto the neckarea of the tube above the electron gun.These getters were stable in air andused a barium -aluminum alloy whichcontained approximately 50% of eachmetal.

The alloy was in the form of a wireclad with nickel or stainless steel; thewire was ground down on one side toexpose the alloy for easy flashing. Thegetter wire was welded to a metal loop(see Fig. 1) to facilitate heating by ahigh -frequency source. This heatingcauses the barium metal to be flashedfrom the alloy by evaporation. Thering getter shown in Fig. 1 is animproved version which allows a largerquantity of barium to be flashed witha minimum expenditure of RF energy.This getter has a stainless -steel channelto which the barium -aluminum alloymaterial is pressed.

Ring getters can be exothermic or endo-thermic. The exothermic ring getteris made by mixing nickel powder withthe barium -aluminum alloy powder.When heating of the getter starts, thenickel and aluminum form an alloyand release heat to help to bring thegetter quickly to the flash temperature.The endothermic getter does notinclude the nickel -powder addition.

Getters in shadow -maskpicture tubesGetters used for shadow -mask colorpicture tubes evolved along the samelines as the black -and -white tube ver-sions. The first color picture tubes(type 21AP22. 70° deflection, largeneck tubes) were produced with sixloop getters mounted above the gun ina circle. These getters had a total max-imum yield of 90 mg of barium andwere flashed radially outward with ahigh -frequency coil onto a limited areaof the neck. Later, a number of differ-ent types of getters were used in thistube as part of a transition that led to aring getter 36 mm in diameter, locatedabove the gun mount and flashed up-ward and away from it. This getterprovided a large area of flash on the

2 -inch (diameter) neck of the tube.Barium fill was 150 mg and actualyield was maintained above 100 mg.By observation of the disappearance ofmetallic barium. it was estimated that50 mg of barium were typically con-sumed during the first few thousandhours of tube operation. The good per-formance of this single getter systemhas been demonstrated by the excellentlife of the 70° deflection color picturetubes.

Two -getter systems

A greatly increased flash area wasachieved by using a getter systemmounted in the funnel and directed toflash across the bulb onto the funneland mask areas. Two 130 mg endo-thermic getters were used in thissystem. As shown in Fig. 2, one wasmounted in the tube neck, and theother in tube funnel. The system hadadequate getter -capacity on the oxygencapacity test (described later) , how-ever, experience showed it to be moredifficult to control the flash of twogetters and that the flashing producedhigher methane pressure' than witha single getter. Although the methanedisappeared during tube scanning, itcaused short-term emission instabilityafter cathode aging.

Large exothermic 'unnel gettersRCA's getter -vendors were asked todevelop, to RCA specifications, asingle, large getter for the 90° deflec-tion tubes. An exothermic getter waspreferred because it is easier to flashthan a large endothermic getter. (Largeendothermic getters can present theproblem of liquid aluminum spillingfrom the channel during flash. Thisproblem is not present with endo-thermic getters because a solid nickel -aluminum matrix is formed duringflash.) A goal of a minimum of 135mg of barium -flash yield was set toachieve the same performance as thetwo -getter system. Experience withfactory flash capabilities indicated thatthe getter fill would have to be in therange of 175 to 185 mg, about twicethe amount as in the largest exothermicgetter of 25 -mm size.Based on RCA's specifications, onevendor developed an exothermic getter

ANTENNA MOUNTEDGETTER

GUN -MOUNTEDGETTER

with 185 mg of fill. The antennamounting system needed no changesto position this getter in the funnel.This getter flashed better than the 127 -mg endothermic getter that it replaced.The starting time (time from the appli-cation of RF to initiation of the flash)was 7 ± Is with a total RF applicationtime of 30 seconds, and the averageyields in production exceeded 150 mg.This getter is now being used in thesmaller color picture tubes.

Another vendor produced a large get-ter having 240 mg of barium fill. Thisgetter. also shown in Fig. 3, has anumber of features which departedfrom the conventional ring -type get-ters. First, it used a greater exposedarea, relative to the weight of fill, thanother getters. This was achieved byforming a solid ring of getter materialand pressing the ring into a separateexternal retainer ring. This externalring also aided the RF heating of thegetter material. Second, a reflector wasincorporated which became hot duringflash and re -radiated the barium thatnormally deposited on the glass surfaceas back flash. Third, a ceramic sub-strate was substituted in place of metal-lic supports in the glass funnel aroundthe getter support to minimize local-lized over -heating and cracking. Thissubstrate eliminated glass problemsand reduced the distance that thegetter extended above the surface ofthe glass, resulting in more efficientcoupling to the RF unit. Fourth. thegetter used gas doping with nitrogen.This allowed the getter to releasenitrogen just before flashing and,thereby, to momentarily increase thepressure and confine the flash area.

The performance of the 240 -mg getterhas been favorable. With gas doping,it was found that RF generators withfrequencies above 450 kHz caused ion-ization of the gas released duringflash and that getter yield was belownormal. However, when the frequencywas reduced to 300 kHz, ionizationwas not visible. This getter is used inthe larger color picture tubes, and hasprovided an average yield of 210 mg.

Control criteria for getters andgetter flashGetters delivered to RCA are checkedfor barium fill, barium yield undercontrolled flash conditions, and gascontent. During the inspection forbarium fill, the getter material is

Fig. 1-Endothermic getters:loop getter (left), ring getter(right).

Fig. 2-Two-getter sys-tem used on the 90° de-flection color picturetube.

Fig. 3-Exothermic getters: 185 -mg getter (left), 240 -mg getter(right).

OXYGEN LINE

1.0

o.e

0.6U

10.4

0.2

OXYGENINLET

CAPILLARYLEAN

Fig. 4-Oxygen capacity test system.

broken out of its channel and inspectedfor barium content by chemical anal-ysis. Barium yield is deter mined by theweight -loss method in which longerthan average flashing is used so thatmost of the barium is removed fromthe channel. Gas content is deter-mined by an ASTM test,' in whichthe pressure peak during flash isrecorded in a system of known volume.

In production, barium yield must becontrolled and conveniently evaluated.The weight -loss method is not whollysuitable because the initial weight ofthe getter, in an arbitrarily selectedproduction tube, cannot be accuratelydetermined. However, because themanufacturer maintains control overthe amount of barium fill in each newgetter, the approximate yield can bedetermined. This method of checkingyield is used for endothermic getters.For exothermic getters a less time con-suming approach is used, namely,X-rays are used to perform a residualbarium analysis. This is achieved asfollows: when a primary X-ray beamimpinges on the face of a getter and itsring, the excited elements in the fill(barium, aluminum, and nickel) emitsecondary or fluorescense X-radiationof discrete wavelengths. Quantitativeanalysis by X-ray fluorescence -emis-sion spectroscopy is possible becausethe intensity of the wavelengthsemitted is proportional to the concen-tration of that element in the sample.In this analysis, the barium Ka emis-

'' -=---1--:-- -0

". :

I

RED

BGLRUEEEN -

1

Table I. Efficiencies of Various Getter Types

Getter Tube

70° Tube

Barium fill(mg)

Getter Efficiency(1-11/mg Ba)

Average OxygenCapacity (l-µ)

Endothermic Neck Getter 150 20 240090° Tube

Endothermic Neck Getter 127 10 4500Endothermic Funnel Getter* 127 40Exothermic Funnel Getter 185 40 6200Exothermic Funnel Getter

(nitrogen gas doped)240 35 7350

Used in combination.

sion line is used to determine theamount of remaining barium in aflashed getter. Each determination isthe average of two readings, with thegetter rotated 90° between readings.This X-ray technique gives the amountof barium to within ±3 milligramsand has been very useful in de-termining the best flashing techniquesand in the discovery of faulty or poorflashing stations.

Evaluation of getter performanceGetter performance is best evaluatedin an operating picture tube. The detri-mental effect of oxygen on the oxidecathode and the high degree of activityof barium for oxygen led to a testmethod in which oxygen is introducedinto a finished tube. (Oxygen is knownto be present in picture tubes duringscanning as is carbon dioxide': the ef-fects of these gases can sometimes beseen by oxidation of grid surfaces intubes which are operated with insuffi-cient getter yield.) When oxygen is ad-mitted, cathode current is constant aslong as the barium film is active andabsorbs oxygen. When the film is de-pleted, the oxygen pressure rises andcathode current slumps.

For the purpose of this test method, aprocedure was developed to break intothe tube and admit oxygen gas. In thisprocedure, see Fig. 4, the funnel of atube is partially drilled with a hollow,diamond -impregnated tool. Glass tub-ing from a vacuum station is sealed tothe funnel with epoxy resin and thehole is broken through by impact witha magnetically controlled rod. Oxygenis let into the funnel from the manifoldthrough a fixed leak. The latter is asection of capillary tube 1 mm in diam-eter and 1/4 inch long having a con-ductance (at room temperature) of0.017 liters of oxygen per second. Asecond leak from an oxygen line keepsoxygen pressure in the manifold con-stant, usually at 4 microns -a con-venient value to be checked withMcLeod gauge. The rate of gas admis-sion into the tube is normally 250 liter -

microns per hour. The quantity ofoxygen that causes the cathode currentto decrease to 80% of its initial valueis taken as the oxygen absorption ca-pacity (in liter -microns) of the getter.Fig. 5 shows a plot of cathode currentas a function of oxygen quantity letinto a new RCA-21FBP22 (70° -deflec-tion) color picture tube. The currentfor all cathodes (red, blue, green) isconstant up to 2000 liter -micronswhere it starts to slump. The 80% endpoint is about 2070 liter -microns forall three cathodes. A slower than nor-mal rate of oxygen admission, 34 liter -microns per hour for 60 hours, wasused to provide sharper end -points onthe curves. The amount of bariumflashed into the tube was 109 mg;therefore, the efficiency of the getter inthe neck position of the tube was 19liter -microns per mg of barium.

Oxygen capacity and getter efficiencyIn a single getter system, oxygen capac-ity is a linear function of barium yield.The getter efficiency (liter -microns ofcapacity per mg of barium yield) de-pends on getter location and designand is sensitive to the distribution ofthe flash.For a two -getter system, consisting ofa neck getter plus a funnel getter at-tached to an antenna spring, oxygencapacity can be expressed by the fol-lowing linear relation:

Q =a N-F-bFwhere Q is getter capacity (liter -microns), N and F barium quantities(in mg) , and a and b are constants.

The chemical equivalent of oxygen re-acting with barium to form bariumoxide, BaO, is 67.7 liter -microns permg of barium. This value is the maxi-mum expected for the constants, a andb, for a getter film which reacts com-pletely with oxygen to form BaO(assuming this happens before thecathode emission is affected). Withthis relationship, a least -squares fit ofthe data taken on many new 25 -inch(90° deflection) color picture tubesgave actual values of a equal to 10

0

121000 2000

OXYGEN OUANTITY -LITER MICRON

Fig. 5-Oxygen capacity test on a 70° tube.

0

and b equal to 40 liter -microns permg of barium. Results show that thefunnel getter with its large area of flashcipproaches the theoretical value, whilethe neck getter with its restricted areais less than 1/4 as efficient as the funnelgetter.Getter efficiencies for the various typesof getters used in color picture tubesare listed in Table I. These values area function of flash area. The 70° neckgetter has a large area of flash andgreater efficiency than the 90° neckgetter. The 90° endothermic and exo-thermic funnel getters have the sameflash areas. The 90° exothermic gas -doped getter, also located in the funnel,has a more restricted area of flash andsomewhat lower efficiency in terms ofthis test than the other 90° getters.Getter depletion during tube operationThe oxygen capacity test can also beused for evaluation of getter depletionin tubes that have been operated. Byuse of this procedure, getter depletionis taken as the difference between thegetter capacity of a used tube and thatcalculated for a new tube. Through theuse of this test method, it has beenfound that oxygen absorption capacityof the 2 -getter system in 25 -inch tubeswas depleted at the average rate of 3liter -microns per hour during the first300 hours of normal operation, and atthe average rate of 1.5 liter -micronsper hour during the first 1000 hours.Gas sources within the picture tubecan be evaluated by similar measure-ments on picture tubes which aremodified by omission of tube compo-nents. In one test, the mask, phosphorscreen, and graphite -silicate funnelcoating were omitted from a tube, andan operating tube made by aluminizingthe funnel and face of the bulb, and byproviding small -area patches to allowset up of the picture tube in a receiver.This tube showed no appreciable de-pletion of the neck flashed getterduring hundreds of hours of tube oper-ation, thus demonstrating that theomitted components were the impor-tant sources of gas.Gas measurement in picture tubesMass spectrometer analysis of partialpressures of gas in tubes during theiroperating life can define conditionsthat limit cathode performance. Amass spectrometer developed for usein color picture tubes has the ioniza-tion source incorporated into the

electron -gun mount structure, makingthe partial pressure measurement referto the same gaseous ambient to whichthe cathode is exposed. The analysersector is attached directly to the neckof the tube; therefore, gas measure-ment can be made during scanningoperation since one of the three gunsin the mount is operable.

The picture tube mass spectrometer isused in a normally processed, sealed -off color tube which can then be op-erated and tested for gas pressuresthroughout a life test. These tests haveshown the following: Prior to getterflash, the residual pressure in the tubeis approximately 5 x 10 Torr, andconsists mainly of nitrogen, carbonmonoxide, and hydrogen. Several min-utes after getter flash a significant dropin pressure occurs; the predominantgases are then methane, hydrogen andargon at a total pressure of 5 x 10Torr. Upon subsequent processing,e.g., cathode aging, the total pressurecontinues to fall.

Results from initial scanning of thepicture tube show an increase in nitro-gen and carbon dioxide pressure; witha well flashed getter, however, the com-bined pressure of these gases does notexceed 5 x 10 ° Torr. With continuedoperation, the pressure in the tube isfurther reduced until after severalhundred hours the total and partialpressures do not appear to change ap-preciably. At this point, the total pres-sure, caused chiefly by nitrogen andcarbon monoxide, is 5 x 10' Torr.Thus, the total pressure in the tubedecreased by a factor of about 10' Torrfrom immediately following exhaust(before getter flash) until after severalhundred hours of operation.

The picture -tube mass spectrometercan also define getter performance. Inthis procedure gas is admitted into thefunnel at a constant rate, simulatingthe gas sources from bulb surfaces.The steady-state change in pressure ofthe gettered gas is determined with themass spectrometer.Gettering action is described by theratio between the gas inlet rate and thepressure change at the cathode. Thisratio has the same units as pumpingspeed (liters per second) , but it is notequal to the pumping speed of thegetter for the gas under test, unless thisquantity is very small. The ratio measures directly the ability of the getter

10

EMISSION CURVE

....".:1/ i

/A,..../4

..., PRESSURE OVERGETTERA- .,- 4i,/

SOO

4

zoo arc

00

100 u

0 SOO 1000 MOO 2000OMEN QUANTITY LITER N CRON

Fig. 6-Oxygen pressure over getter andcathode current as a function of quantity ofoxygen absorbed.

to protect the cathode from the testgas and, therefore, its ability to protectthe cathode during normal tubeoperation.

When the mass spectrometer is usedto measure change in oxygen pressureas a function of a constant oxygen -inletrate, and the resulting data is com-pared to the drop in cathode emission,the results are similar to those shownin Fig. 6. The color picture tube usedfor this test was a newly processedRCA-19EYP22 tube in which 50 mg ofbarium fill had been evaporated. FromFig. 6, it can be seen that the cathodeemission current decreases to 80% ofthe initial value at a pressure of about2 x 10-` Torr. At this point the oxygencapacity is 2000 liter -microns. Manysimilar tests carried out on new19EYP22 tubes and tubes which havebeen operated for thousands of hoursindicate that the 80% emission leveloccurs when the pressure rises to therange 5 x 10 ' to 5 x 10' Torr.

Concluding .emarksLarge barium getters have been devel-oped to satisfy gas absorption require-ments in color picture tubes. Testsevolved to assess gettering ability inproduction tubes can show whether agetter system is adequate throughoutlife operation of the tube, and can thusbe used to define the getter system.

References1. Moscony, I. J., Evolution of Methane from

Barium Films (Supplemento at Nuovo CimentoVol. V, No. I. p. 7, 1967).

2. Michon, L. and Giorgi. T. A., Studies of GasComposition Within Finished Television Pic-ture Tubes, (Same p. 104).

3. ASTM Gas Test of Barium Getters (ASTMTentative Proposed Practice for Testing BariumFlashed Getters, June 5. 1964, Rev. 6).

4. Collins, R. H. and Turnbull, J. C., Evolutionand Absorption of Gases in Electron Tubes(Vacuum 10, p. 27, 1960).

5. Todd, B. J., Lineweaver, J. L. and Kerr,J. T., Outgassing Caused by Electron Bombard-ment of Glass (Journal Applied Physics Vol.31. p. 51. 1960).

6. Moscony, J. J. and Turnbull, J. C., A MassSpectrometer for Gas Analysis in Color Televi-sion Picture Tubes. (Supplemento al NuovoCimento Vol. V. No. 1, p. 93, 1967).

13

Colorimetry and contrastperformance of color picturetubesG. M. Ehemann W. G. Rudy

Although manufacturers use many technical terms to describe the quality of colortelevision pictures, the ultimate appraisal is made on a purely subjective basis by theviewer. Therefore, it is more important to produce a picture that is "pleasing" tomost viewers than to strive for an exact reproduction of an original scene which thehome audience does not see. A viewer's subjective appraisal of a color picture de-pends upon the sensations he experiences as a result of various light stimuli. Thesesensations are produced by many properties of the light which can be measured anddescribed in absolute values (e.g., total intensity, spectral energy distribution, hue,purity, brightness, contrast. and resolution). This paper discusses audience responsesto chromatic stimuli of various hues, and relates the physical properties of thesestimuli to the concept of picture "viewability."

VISIBLE LIGHT occupies the wave -length region between 4000 and

7000 angstroms. The spectral energydistribution of a light source is adetailed description of its color com-position, i. e., the relative intensity ofits light components distributed alongthis wavelength region. Each of thethree phosphors used in color tele-vision is a primary emitter of light in aseparate region of the visible spectrum,i.e., the three phosphors produce red,blue, and green light.

Color mixing

The triad of red, blue, and green phos-phors is important in color televisionbecause the proper mixture of theseprimary emitters produces white light.Addition of any primary color to thewhite mixture produces the hue ofthat color; subtraction induces thecomplement of the primary color. Theprimary triad, therefore, can generatea countless number of colors (thetotality of which is called its colorgamut), as well as white. (Althoughthis discussion treats the primary triadas pure, spectral, or "rainbow" colors,it also applies to the phosphors used incommercial color television, whichhave less sharply peaked spectraldistributions.)

The standard color mixing diagramshown in Fig. 1, which was adopted bythe International Commission on Illu-mination in 1931, illustrates the pairs

Final manuscript received October 16, 1968.

of complementary colors generated bythe primary triad. Because the propermixture of complementary colors pro-duces white, the straight line connect-ing each pair passes through the white,or neutral, point. The location of thispoint on the connecting line is an indi-cation of the effectiveness of eachprimary color in neutralizing its com-plementary color. For example, bluelight neutralizes its complement, yel-low light, very effectively; red andgreen are less effective in neutralizingtheir respective complements, cyan(blue-green) and purple. (This prop-erty is described as relative luminanceof each color, or as the luminance ratiobetween the complementary pairs.)

The neutralizing power of a primarycolor over its complement also affectsa viewer's response to a field of thatcolor placed in a luminous neutral(white) background. A neutral back-ground of higher luminance (bright-ness) than the color area induces anapparent grayness into the chromaticfield. The color at which the thresholdof grayness is reached and "fluorence"(the appearance of fluorescence)begins to decrease is called a zero -graycolor.'

Zero -gray colors for a given hue aredescribed by their luminance andpurity. The purity of a zero -gray coloris the ratio of the luminance of thepure or chromatic hue -determiningcomponent to the total luminance ofthe color mixture (chromatic pluswhite additive) which complements

George M. EhemannChemical and Physical LaboratoryTelevision Picture Tube DivisionElectronic ComponentsLancaster, Pa.received the BE in Engineering Physics from Cor-nell University in 1964. That year he joined RCA'sChemical and Physical Laboratory at Lancasterand has been working on color measurement tech-niques and optical property studies of color televi-sion screens. In 1965, while on leave of absence.he earned the ME in Engineering Physics atCornell University.

William G. RudyChemical and Physical LaboratoryTelevision Picture Tube DivisionElectronic ComponentsLancaster, Pa.received the BS at Dickinson College in 1943 andthe MS in Physics at Carnegie-Mellon in 1948 Hehas had additional academic work at the AmericanUniversity at Biarritz, France, Newark College ofEngineering, and Franklin & Marshall College. In1948 he joined the Chemical and Physical Labora-tory at RCA Lancaster and achieved senior engi-neering status in 1955. His major concern hasbeen experimental work embracing conversion,cathode ray, and power tubes. His contributionsresulted in improved performance in camera tubesleading to four patents, enhanced secondary elec-tron emission in conversion tubes, and accurateand meaningful test methods for evaluating thermi-onic emission, residual gas. and life performancein cathode ray and power tubes. Presently, he isworking in the field of colorimetry of color TVpicture tube phosphors and has recently publisheda paper on the chromaticity of narrow -band emit-ting phosphors.

14

320

it in a two -component mix and desat-urates it. Both can be varied inde-pendently of the luminous background.

An experimental technique was usedto determine zero -gray color curves forblue, yellow, green, red, and cyan huesproduced by a P22 television phosphorscreen when the grid voltages of thethree electron guns were adjusted togenerate the proper primary excita-tions for the standard television white.For a desired hue of red, blue, or green,the grid voltage exciting the corres-ponding phosphor primary is increasedeither slightly or noticeably to producelow and medium purities, respectively.Full purity occurs when the hue -pro-ducing primary alone is excited.

For example, yellow and cyan huesare produced when their respectivephosphor primary complements (blueand red) receive less excitation thanthat required to produce white. Whenthe blue or red phosphor is unexcited,full purity results in the yellow andcyan hues, respectively.

Bright light having the color of a3500°K black -body radiator was usedto illuminate a reflecting baffle placedbetween the observer and the televi-sion screen; the light passes throughthe viewing aperture in the bafflewithout striking the screen, as shownin Fig. 2. The observer views thescreen through the baffle illuminatedby a luminous background, adjustsscreen luminance until the zero -gray-ness threshold is reached, and recordstwo measurements: the total lumi-nance of the phosphor screen and theluminance of the hue -producing pri-mary (in the case of yellow hues, theblue -phosphor luminance is recorded) .The first measurement is the zero -graycolor luminance, the second measure-ment allows calculation of colorpurity. Fig. 3 shows the results of suchmeasurements for blue and yellowhues.

Blue colors, which easily neutralizetheir yellow complements, remain flu-orent until a high threshold value ofbackground luminance is reached.while yellows lose fluorence and ap-pear to contain gray at a much lowerthreshold. Television -gamut greens andreds behave as the yellow colors, whilecyans exhibit intermediate behavior.

The uniqueness of the zero -gray colorsfor blue hues lies in the behavior ofrequired blue chromatic luminance asa function of color purity in the pres-ence of a bright background. The bluechromatic luminance curve in Fig. 3is flat at all color purities above a lowthreshold value; therefore, reductionof the white component does notrequire an increase in chromatic illu-mination, and the zero -gray thresholdis maintained all the way to full purityas the white component is graduallyreduced to zero.

Although a decrease in the white com-ponent of luminance in a yellow hueproduces a purer yellow on the colormixing diagram of Fig. 1, for fixedbackground luminance near threshold,the induction of grayness offsets thisphysical gain in purity. As a result,additional chromatic yellow illumi-nation would be required to producethe same total luminance of the orig-inal color to a viewer. At zero -graythreshold, therefore, total luminanceof yellow hues remains constant forpurities ranging from zero (completelywhite) to unity (pure yellow) . Wheth-er the "pure yellow" is spectral orwithin the television gamut of a broad-band green is unimportant in mattersconcerning fluorence of yellow hues.So far as fluorence is concerned, phos-phor efficiency rather than purity ofemitted color is essential in the greenand red primaries.

Picture viewability

Viewability is best studied independ-ently of a television picture so that theviewer can control some of the physi-cal properties. When a television pic-ture is simulated with projected colortransparencies, the viewer can manipu-late independently the color, the lumi-nance (brightness) , the resolution, andthe ambient illumination to "creole"the most desirable picture.

For each picture, the effects of thevarious factors influencing the viewer'sjudgment can be evaluated from apattern produced by substitution of aRETMA resolution -chart transparencyfor the color transparency. The imageprovides a gray scale, broad black andwhite areas for high contrast, and reso-lution wedges. The gray scale, whichhas low contrast in both highlight

00

70

60

So

40

SO

20

0

5,6GR. EN

500

SLOE -60EEP

490 ;.

440

3L

46

0 10 20 50 40 SO 60 70 SO

Fig. 1-Stancard ICI diagram showing spectral colorsand th3ir corr dements (the color of neutral equals thatof a 7 )00` K lack -body radiator).

VIEWER'SETC

P22 PHOSPHOR SCREEN EXCITEDTO DESIRED HUE AND

BAFFLE REFLECTING SURROUNDILLUMINATION TOWARDS VIEWER

IMAGE OF EXCITEDPHOSPHOR SCREEN

Fig. 2-Expe- mental determination o' zero -gray colors.

and shadow areas, proves particularlyuseful.

One important factor is the picturebrightness required for comfortableviewing at various ambient illumina-tion levels. For example, the luminancerequired when pictures are viewedwith 2.5 foot -lamberts reflected am-bient illumination, comparable to thatencountered in a moderately lightedroom, is approximately 17 foot -lam-berts. For prolonged viewing, the pic-ture is judged uncomfortably bright at40 foot -lamberts and too dim at 5 foot -lamberts. Fig. 4 shows the relationbetween the picture luminance fortypical low and high brightness areasand reflected ambient illumination forsatisfactory viewing.

Although the contrast and the resolu-tion of a color picture are decreasedby increases in ambient illumination,viewability can be conserved by in-creased picture luminance. As an ex-ample, a picture was set up with areflected ambient illumination of 0.5foot-lambert. The white luminance of

15

SURROUND

g 10

2

4--,YEL OW WNITE

BLUE COMPONENT

0 02 04 06 0.6COLOR PURITY

Fig. 3-The required luminance of blue andyellow hues and their respective chromaticcomponents at zero -grayness threshold in a500 fL surrounding.

the RETMA slide measured 25 foot -lamberts and the highlight contrastratio was 50. When the ambient illumi-nation was increased to I foot-lambert,the contrast ratio was reduced to 26.The viewer then changed the picturebrightness to "recover" the excellentpicture. However, instead of increas-ing the highlight brightness to 50 foot -lamberts to produce the original con-trast ratio of 50, the viewer settledfor a highlight luminance of about 40foot -lamberts, and a contrast ratio ofonly 40, as shown in Fig. 5.

The conservation of viewability wasstudied by Luckiesh and Moss' usingthe black -and -white Snellen eye chartat various room light levels; their re-sults are shown in Fig. 6. The percentcontrast has been recalculated in termsof contrast ratio, which is the lumi-nance ratio of the highlights to adja-cent areas on the eye chart. Thesecurves should remain valid for com-parison to color television, in which aself -luminous phosphor screen and aprotective gray glass are placed in anambient setting.

The experimental highlight data forconstant viewability shown in Fig. 5correspond to the 11