US3476025A

US3476025A - Cathode ray tube and method of manufacture

Info

Publication number: US3476025A
Application number: US557663A
Authority: US
Inventors: Fred Herzfeld; Frans Van Hekken
Original assignee: RCA Corp
Current assignee: RCA Corp
Priority date: 1966-06-15
Filing date: 1966-06-15
Publication date: 1969-11-04
Anticipated expiration: 1986-11-04
Also published as: NL168363C; GB1194023A; DE1614369A1; NL168363B; MY7300269A; BE699947A; JPS5428711B1; SE338111B; ES341726A1; NL6708246A; AT279683B

Description

NOV. 4, 1969 F HERZFELD ETAL 3,476,025

4CATHODE RAY TUBE AND METHOD 0F MANUFACTURE 3 Sheets-Sheet 1 Filed June l5, 195

Nov. 4, 1969 F. HERzFELD ETAL 3,476,025

CATHODE RAY TUBE AND METHOD OF MANUFACTURE Filed June l5, 1966 3 Sheets-Sheet 2 Oohm.

cou@

OOK.

ogm.

oo .N

vob SQ NOV. 4, 1969 F, HERZFELD ETAL 3,476,025

CATHODE RAY TUBE AND METHOD OF MANUFACTURE Filed June l5. 1966 3 Sheets-Shen 5 United States Patent O 3,476,025 CATHODE RAY TUBE AND METHOD F MANUFACTURE Fred Herzfeld, Princeton, NJ., and Frans van Hekken,

Lancaster, Pa., assignors to RCA Corporation, a corporation of Delaware Filed June 15, 1966, Ser. No. 557,663 Int. Cl. G03c 7/00; G03f 5/00 U.S. Cl. 95-1 16 Claims ABSTRACT OF THE DISCLGSURE A shadow mask type color picture tube mosaic screen pattern is laid down by means of a lighthouse including a light refracting element or lens having an analytical surface approximating the theoretical lens surface required to provide misregister compensation at each of -a multiplicity of points distributed over the entire surface of the screen as determined from misregister measurements `at those points on the screen of an uncompensated color tube. The lens may be formed by sagging a plate of glass onto a mold fabricated by numerical control milling techniques. The lens surface has no symmetry.

This invention relates to cathode ray tubes, and to the manufacture thereof, and particularly to shadow mask type color cathode ray tubes comprising a plurality of electron guns, a multi-apertured shadow mask, and a mosaic screen of systematically arrayed color phosphor deposits, or dots.

The phosphor dots of the screen of such a tube may be laid down in trios (groups of three dots of different coloremitting phosphors) by a direct photographic printing technique wherein a photosensitive coating on the faceplate panel of the tube is exposed through the apertures of the mask by a point source of light, and the coating is then developed, as by washing off the unhardened unexposed portions, leaving the desired pattern of exposed hardened dots. This process is repeated for each color. The shadow mask is preferably detachably mounted on the faceplate panel so that it can be easily removed and replaced in exactly the same position every time. Phosphor powder may, e.g., be mixed directly with each photosensitive coating before application to the faceplate, or else applied to the coating after the latter has been eX- posed.

In the operation of the tube, the electron beams are subjected to forces such as scanning (i.e. horizontal and vertical deflection) and dynamic convergence (to maintain convergence of the beams near the screen at various angles of deflection), which affect the electron beam paths (and hence, the landing points or spots of the beams on the screen) in ways that the screen-printing light rays are not affected. Thus, unless compensation is made for the differences between the beam paths and the light ray paths, serious misregister of the beam spots with the phosphor dots will result, i.e., the spot `and dot centers will not coincide. Hereinafter, the electron spots and phosphor dots will usually be referred to simply as spots and dots, respectively.

Misregister of the type wherein a trio of beam spots is shifted as a unit radially outward from its associated dot trio, caused by an axial shift of the deflection centers of the beams toward the screen with increasing angles of deection, is termed radial misregister.

Misregister of the type wherein the individual spots of a spot trio are all three moved away from each other, caused primarily by dynamic convergence forces, is termed degrouping misregister.

Misregister of the type wherein the beam spots of a spot trio are distorted away from equilateralism because of nonuniformities of the raster scanning deflection fields and/ or the phosphor dots of atrio are distorted away from equilateralism because of inherent characteristics of the dot printing system, is termed astigmatic misregister.

Misregister of the type wherein a phosphor dot and its associated beam spot are individually moved different distances relative to the centers of their respective trios due to an apparent decrease of the spacing of the dot printing light sources and of the electron beams from the central axis at increasing angles of deflection, is termed foreshortening misregister.

Other sources of misregister of the spots and dots include the effect of the earths magnetic field upon the paths of the electron beams, and the effect of distortion of the faceplate when the tube is evacuated.

In the commercial manufacture of prior art tubes of the type in question, compensation for radial, degrouping and foreshortening misregisters has, to a degree, been provided in the form of selected tube geometry and in the use of unique screen printing methods and apparatus. One example of such compensation is taught in U.S. Patent 2,885,935, issued to D. W. Epstein et al. That patent teaches that best correction for degrouping and radial misregister can be obtained by (a) using a mask-to-screen assembly such that the spacing between the mask and screen varies in a particular manner from a maximum at the central axis to a minimum at a given radius near the edge of the assembly, and (b) interposing, in the path of light from the light source to the mask in the screen printing apparatus, or lighthouse, a special light refracting device, or lens, having a single central plane of symmetry. The contour of the lens along the diametrical section perpendicular to this plane is determined by measuring the average radial misregister of the spots and dots on the screen of an evacuated tube, calculating the rate of change in thickness of a lens along this section necessary to correct or compensate for the measured misregister and fitting an odd power polynomial to the rate of change values. The contour of the lens at all other points (not on the diametrical section) is calculated by multiplying the odd powers of the polynomial by the cosine of the azimuthal angle with respect to the plane of symmetry and adding the even powers of the polynomial thereto. The result is a lens that is relatively simple to fabricate by known grinding methods. However, such a lens provides acceptable compensation for beam path distortion, and hence acceptable register of the beam spots and phosphor dots in color tubes having a maximum beam defiection angle of 70, only along said line of symmetry. Particularly in the case of tubes with relatively large maximum deflection angles, e.g. and non-circular shape, e.g. rectangular, the amount of misregister produced in regions not on the line of symmetry in tubes made with such lenses exceeds desirable tolerances.

Accordingly, an object of the present invention is to provide an improved cathode ray tube of the shadow mask type wherein the phosphor dots are laid down by a light printing process in which acceptable register of the beam spots and phosphor dots is obtained over the entire screen of the tube.

Another object of the invention is to provide an improved method of photographically printing the phosphor dot screen of a shadow mask type cathode ray tube which will exhibit acceptable register of the beam spots and phosphor dots over the entire surface of the screen.

Still another object is to provide an improved method of making an improved corrective lens to be used in the manufacture of phosphor dot screens of shadow mask type cathode ray tubes.

These and other objects are achieved by printing the phosphor dot screen of the tube by means of a light house including one or more special lenses designed to produce acceptable correction or compensation for all of the causes of misregister over the entire surface of the screen. In the case of a three-color kinescope with three beams, a different lens is used to print the dot pattern for each color. However, the invention could be used to make a dot screen for a tube having only one beam and one kind of phosphor dot.

The improved lens for use in printing one color dot pattern on the screenplate of a tube of given kind as, for example, a 19" rectangular color kinescope with 90 deflection, may be made by the following steps:

(l) A faceplate of given curvature, e.g. spherical, is fabricated;

(2) A multi-apertured shadow mask of desired curvature is fabricated;

(3) A phosphor dot pattern is printed on the faceplate according to prior art techniques, by applying a photosensitive layer to the faceplate, assembling the screenplate and mask in predetermined spaced relation on a lighthouse, exposing the photosensitive layer to light rays passing through the mask aperatures, and developing the layer to remove the unexposed portions;

(4) The mask and exposed faceplate are incorporated into a complete, operative cathode ray tube;

(5) The tube is operated, with an electron beam scanning the printed dot pattern, and the direction and amount of misregister of the beam spots and corresponding phosphor dots is measured at each of a multiplicity of points distributed over the entire surface of the screen;

(6) The total number of data points may be multiplied, e.g. to 250-300 points, by interpolating additional data points located between the measured points;

(7) For each of the data points in steps (5) or (6) the elemental three-dimensional slope required at the corresponding point on the surface of a theoretical lens, which, if used in a subsequent printing operation, would compensate for the conditions which cause said misregister at that data point, is determined, and these slopes are preferably converted to vectors normal to the lens surface;

(8) The theoretical lens surface of step (7 is approximated by fitting a bivariant polynomial to the calculated slopes or normal vectors to obtain a description of a continuous surface of the lens to be made;

(9) Tme bi-variant polynomial surface description of step 8) is converted to one or more tapes for controlling a numerical-control (NC) milling machine;

l) A lens mold is fabricated by the NC machine and tape to the desired lens surface; and

(11) The desired lens is sagged on the mold.

The printing step (3) could be carried out with no lens used in the lighthouse, in which case the lens produced could be used alone in a lighthouse to print the phosphor screen (one color) for a commercial tube. However, the method is preferably practiced by using at least one lens of known configuration in step (3), in which case the description of this lens is incorporated in step (7), and the resulting new lens is used in combination with the first lens to make commercial tubes. Alternatively, one or more lenses could be used in step (3) Vand a new lens could be fabricated to be substituted for one or more of the first lenses in manufacturing tubes, which new lens would incorporate the refractive properties of the original lens of step (3).

In the process of fabricating corrective lenses for use in making three-color tubes, step (3) is repeated to print each color dot patem prior to step (4), and the misregister measurements of step are preferably made simultaneously for all three colors. However, the subsequent steps are carried out independently for each color, to produce three different lenses, one for each color.

Each final lens resulting from the process described briefly above is tested, before being used commercially, by using it to print a dot pattern on the faceplate of an experimental tube, operating the tube, and again measuring the misregister at the same or different points. In most cases, the amount of misregister is within predetermined limits or tolerances, and hence, commercially acceptable. In the event that the residual misregister in this test is not acceptable, the entire process is repeated, using in step (3) the last lens fabricated, to produce a new lens capable of producing tubes with acceptable register over the screen surface.

The invention will be described in greater detail in conection with the accompanying drawings, wherein:

FIG. l is a schematic illustration of a three-beam shadow mask cathode ray tube showing the causes of radial and degrouping spot-dot misregister;

FIG. 2 is a partial axial section of a lighthouse apparatus that may be used in practicing the method of the invention;

FIG. 3 is a plan view of cathode ray tube faceplate showing a distribution of points at which misregister may be measured;

FIGS. 4 and 5 are schematic views, partly in section in the plane of FIG. 2, to be used in explaining the dot printing operation;

FIG. 6 is a transverse section through a lens mold made according to the invention and a glass plate to be sagged onto the mold;

FIG. 7 is a view similar to FIG. 6 after the sagging operation; and

FIG. 8 is a transverse section of the finished lens.

FIG. 1 illustrates a cathode ray tube of the type described which comprises an envelope 10 containing therein three electron guns 11, 12 and 13, which may, for example, be disposed co-planar or in a triangular array, for projecting three electron beams towards a faceplate panel 14. A delta triangular array symmetrically disposed about the central axis A-A of the tube is preferred.

For a delta gun array, a shadow mask 15 is usually formed with a multiplicity of apertures 15a systematically hexagonally arrayed thereover; and a mosaic screen 14a is provided on the faceplate 14 comprising a multiplicity of phosphor dots similarly arrayed, with a trio of three dots, each of a different color emitting phosphor, being provided for each aperture 15a in the mask 15.

In operation of the tube, three separate electron beams are projected from the three guns 11, 12 and 13 and are directed to be converged to a cross-over point near the screen 14a by virtue of the mechanical disposition of the guns and/or convergence forces generated by convergence means 16. The three beams approach the mask 15 and portions thereof pass through the mask apertures 15a and excite the different color-emitting dots of the same dot trio.

In FIG. 1, the

numerals

18, 20 and 22 indicate the paths of three beams passing through a central aperture of the mask. The centers of

deflection

24, 25 and 26 of the

beams

18, 20 and 22 lie in a plane P-P perpendicular to the central tube axis A-A, which is referred to as the plane of deflection of the tube.

When the beams are deflected away from the

paths

18, 20 and 22 to e.g., the

beam paths

28, 29 and 30, the beams would, in the absence of dynamic convergence forces being applied thereto, converge to a cross-over before they reach the screen 14a, because of the greater distances between the centers of deflections and the screen. To prevent such a premature cross-over, the beams are spread apart, by dynamic convergence forces established by convergence means 16. For example, the dashed lines 2S and 30', by comparison with the

paths

18 and 22. illustrate the spreading of the beams from guns 11 and 13 by convergence means 16 when the beams are deflected to follow paths 28 and 3).

For a deflected beam, the center of deflection is defined as the point of intersection of the beam path through the central aperture and the rearward extension of the beam path in the region beyond the influence of the defleeting field. Thus, the deflection centers of the deflected

beams following paths

28, 29 and 30 are respectively indicated by

points

32, 33 and 34 in FIG. 1. When the beams are deflected from the center of the screen, and proper convergence beam-spread is applied, the centers of deflection of the beams move both forward and outward to the

new points

32, 33 and 34, and the plane of deflection moves forward to plane PP.

As a result of the forward axial shift of the centers of deflection, the landing spots of the three beams are shifted radially outward on the screen 14a, the radial shift being substantially equal for each spot of a given trio. This forward shift of the deflection centers with outward defiection of the beams will result in lmisregister of the beam spots with respect to phosphor dots printed in a lighthouse having the light source at the zero deflection center for each color, with no corrective lens.

As a result of the outward shift of the defiection centers, the trio of beam spots will be spread apart, or degrouped, at the outer portions of the screen 14a. Unless compensated for, the trios of beam spots will be larger than the printed phosphor dots at the outer edge of the screen, causing degrouping misregister. Both radial and degrouping misregister increase as a function of increasing distance from the center of the screen.

In order to photographically print phosphor dot screens with acceptableD register of the beam spots and phosphor dots, it is necessary to incorporate one or more corrective lenses in the screen printing lighthouse to compensate for all of the various causse of misregister. As pointed out above, substantial compensation and register have been produced in the prior art only for the average degrouping along a single section of the lens used, with only partial compensation over the remainder of the lens (and screen).

FIG. 2 illustrates a typical lighthouse apparatus of the type normally used for printing a phosphor dot screen of a cathode ray tube. The lighthouse 110 comprises an open-top cabinet 111 having a shoulder 112 on which the bowl-shaped faceplate panel 14 of the cathode ray tube is disposed. The panel 14 is adapted to be subsequently sealed at its open end 115 to another member (not shown in FIG. 2) to form a completed cathode ray tube bulb. The panel 14 includes a surface 14b for supporting the phosphor screen 14a of FIG. 1 and a plurality of studs 118 on which the apertured shadow lrnask 1,5 (FTG. 1) can be removably mounted. Prior to mounting the panel 14 on the cabinet 111, the surface 14h is coated with a conventional photoresist material, such as polyvinyl alcohol sensitized with ammonium dichromate. On the external surface of the panel 14 are a plurality of bosses 119 which cooperate with mating recesses in the cabinet 111, whereby the panel 14 may be positioned in a prescribed orientation on the lighthouse 110.

At the base of the cabinet 111 is a housing 120 which contains a lamp 121 and a tapered light conductor or collimator 122. The lamp 21 may, for example, be an ultraviolet light emitting device such as a General Electric Company 1 kilowatt, high pressure, mercury arc lamp, Type BH6. The collimator 122 is positioned above the UV lamp 121 and tapers away therefrom to a small area point 124. The point 124 is positioned in a selected plane of deflection at a predetermined distance d from the central axis A-A of the panel 14. The section of FIG. 2 is taken in the plane passing through the central axis A-A and the point source 124. The housing 120 may, as shown, be mounted on a rotatable table 125 which can be indexed at a plurality of predetermined positions. Means 126 including a spring-loaded plunger is provided which cooperates with mating depressions in the rim of the table 125 to fix the table in these positions. Such indexing of the table 125 and housing 1201 is designed to selectively position the light source 124 at a different location for each of the dot-printing exposures required.

A support bracket 127 disposed between the light source 124 and the panel 14 is supported on a plurality of legs 128 from the table 125. The bracket 127 has an opening 129 therein, opposite the light source 124, across which a light refracting element or lens 130 may be disposed. The lens 130 is thus maintained in a fixed angular relationship to the light source 124 when the latter is moved.

Alternatively, the table may be omitted, the housing .120 mounted directly on the base of cabinet 111, and the bracket 127 mounted directly on the wall of the cabinet. In such case the light source 124 and lens are fixed. A plurality of such light-houses 110` are then provided for making the plurality of required dot printing exposures on a given faceplate panel 14.

According to usual screen printing methods the light source for each phosphor dot exposure is located substantially at a position termed the first order color center of the beam. The lirst order color center may be defined as the intersection of a line extending through the center of a given phosphor dot and its associated mask aperture with the plane of deflection associated with the given phosphor dot. The term associated aperture refers to the same aperture that the beam passes through to excite the given phosphor dot. A copending application of Albert M. Morrell et al. Ser. No. 208,044, tiled July 6, 1962, now Patent No. 3,282,691, granted Nov. l, 1966, discloses and claims an improved method of dot screen printing in which the light source is disposed at a second (or higher) order color center, displaced laterally in the plane of deflection a predetermined distance, from which the light rays pass through a mask aperture different from the electron beam aperture associated with the dot being printed. This arrangement inherently compensates for most of the degrouping effects and improves the register of individual spots and dots at the screen.

The present invention may be used with either first or higher order color center printing. In the example shown and described, first order color center printing will be used for simplicity. Thus, in FIG. 2, the light source 124 is located substantially at the first order color center. This center is displaced a distance d from the central axis A-A, as determined from the formula q=La/ 3d, where q is the mask to screen spacing along the central axis, L is the distance between the color center and the screen surface 14b is on the central axis, and a is the spacing between mask apertures.

The lens 130 in FIG. 2 could be any known lens which is usable for partial correction or compensation of the various beam or printing errors involved. Preferably, a lens is selected which is known to produce as much correction as possible, thereby minimizing the additional correction required. In the present example, it will be assumed that lens 130 is a known lens fabricated according to the Epstein et al. Patent 2,885,935 to provide substantial correction for both radial and degrouping errors along its line of symmetry.

The first step in the process of developing a new lens for a new tube type, such as a 19" rectangular 90 color tube, is the selection and/or fabrication of a faceplate panel 14 of given contour. The faceplates for the 19 and V15" tubes are nearly spherical, as compared to the complex contour of the earlier 25" tube faceplates. For example, the internal surface 14b may have a spherical radius of 26.812 inches. Next, a shadow mask 15 for use with the chosen faceplate 14 is fabricated. Preferably, the contour of the mask 15 is made such that when mounted on the panel the spacing q therebetween, as measured along the beam path, varies from a predetermined maximum value, e.g., about .4 inch, at the center to a predetermined smaller value at a given radius near the edge of the screen. The variation in q' is preferably such that, after correction for all errors, the size of the beam spot trios and phosphor dot trios is substantially the same over the entire screen surface. A mask having a spherical radius of 27.27 inches may be used with the panel radius given above.

The selected panel 14 and mask 15 are mounted in the lighthouse with the lens 130, as in FIG. 2, with a photosensitive coating on the faceplate surface 14b, and a dot pattern is printed and developed in conventional manner on the surface 14b. This may, for example, be the green-emitting dot pattern. Then, the printing process may be repeated to add the red and blue dot patterns. The same lens 130 may be used for each pattern in this step, if desired.

The assembly of the printed panel 14 and mask 15 is then combined with the bulb funnel, stem and other parts, and processed to produce an operative, evacuated, shadow mask color tube. The tube is operated in normal manner, scanning the screen with the beam (or beams, if all three colors are screened), and the misregister between the electron spots and phosphor dots is measured at a multiplicity of points or locations distributed over the entire face of the tube. FIG. 3 shows an example of a distribution of points P1, P2, P3, etc., on the faceplate 14 at which the misregister may be measured. The points shown are systematically arranged on circles concentric with the axis A-A of FIGS. 1 and 2, and on radial lines at the various clock positions. At each point, the measurement includes the three-dimensional location of the point in space, and the magnitude and three dimensional direction of the misregister.

One method that can be used to measure and record the misregister is as follows. At each of 63 points distributed over the screen, a photomicrograph is taken of a small group of phosphor dots during operation of the tube. The points are recorded by clock positions, which are later converted to azimuthal angles in radians `from a given axis. Each photograph cover an area of about 6 by 7 rows of dots. First, the centers of three pairs of phosphor dots and electron spots for each color of dots in the central area of each photograph are deter mined. Then, the magnitude and direction of misregister are measured and recorded for each pair of spots and dots, for each color of dots. The misregister measurements are converted to vectors, and the three misregister vectors for each color are averaged (by averaging vector components) Preferably, the mask to screen spacing q parallel to the beam path is measured at each point to determine the variation, if any, from the desired, or bogie, q'. Then, if necessary, the measured misregister vectors for each color are corrected for the variation `from bogie q'.

Misregister measurements are usually made on live substantially identical tubes printed by the same lens 130, and the corrected misregister vectors for each color at each coresponding point on the live tubes are averaged. In the above example, this results in 63 average misregister vectors for each color. Since the subsequent surface fitting calculations require a larger number of data points, additional data points are obtained by interpolating additional data points located between the measured data points up to a total of about 250 vectors for each color.

The next step in the process is the determination of the location in space of the mask aperture associated with the dots and spots at each data point. This is done, as shown in FIG. 4, by first determining the path of a light ray from the position B of the light source 124, through the known lens 130, to the position D of each data point. In FIG. 4, ray BEFCD enters the lens at point E, at an angle 01 with the normal EG to the lens surface at that point, and is retracted to path EF within the lens, at an angle 02 with the same normal extended (EH). In leaving the lens 130, the ray is again refracted from path EF to new path FD. Since the lens 130- is wedge-shaped at the ray path shown, the two paths BE and FD are not parallel. The point of intersection C of the ray path FD with the mask 15 is the location of the corresponding mask aperture. In determining the ray path BEFCD, an approximation is made by drawing a straight line BD between points B and D (shown dashed). From the point of intersection E' of this line BD with the lens 130, the path of an actual light ray BEFD through the lens and to the screen is determined, using Snells law, i.e., N1 Sin 02=N2 Sin 02, where N1 is the index of refraction of the first medium (air) and N2 is the index of refraction of the second medium (glass).

The vector DD', which represents the error due to the straight line approximation, is then subtracted vectorially from the straight line BD, and a second straight line approximation is drawn from B to D and a new corrected ray path through the lens is determined. This series of approximations is continued until the error is not more than 50 microinches tolerance.

Having determined the location C of the mask aperture for each data point, the next step in the process is to determine the surface contour of a new lens for each color, to be used with the lens to print cathode ray tube screens with acceptable registry of spots and dots over the entire screen surface. FIG. 5 shows the lens 130 in the same position relative to the mask 15 and screen 14a as in FIGS. 2 and 4. For example, C and D represent the same aperture and dot, respectively, as in FIG. 4. Point S is the measured center of the electron spot corresponding to dot D, hence DS is the measured misregister therebetween. The object is to determine the elemental slope at a corresponding point on a theoretical lens 132 which, if used in combination with lens 130, would print a dot at point S, instead of D. A straight line is drawn from spot S through aperture C to lens 130, at point I. The path JK of a refracted ray passing through lens 130 is then -retermined, using Snells law. Then the path KL of this ray through lens 132 is determined. If the two lenses are in optical contact and made of the same index of refraction material, path KL will merely be a straight line continuation of path J K. Having determined the location of point L in space, the elemental slope of the surface of lens 132 at that point necessary to refract the ray KL to a ray path passing through the light source at B is determined. Point B may be, but is not necessarily, at the same point as in FIGS. 2 and 4. The elemental slope of the lens 132 at each point x, y, z thereon may be expressed in terms of z/x and z/y, in the coordinate system shown in FIGURE 4, where the Z axis is parallel to the axis A-A, and X and Y are the rectangular coordinates in a plane perpendicular to the Z axis. Preferably, the slopes of the lens surface are converted to vectors normal to the elemental plane of the surface, with one normal vector at the lens data point corresponding to each screen data point.

The next step is to prepare a description of a continuous lens vsurface which approximates as nearly as possible the contour of the surface defined by the multiplicity of normal vectors obtained from the measured misregister data. This continuous lens surface may be expressed as a bi-variant polynomial 2:2'1/1. jxiyj (1) where z is the height of the lens surface at each point (x, y) thereon from the X-Y plane z' and j are the exponents of x and y, respectively, in each term of the polynominal, and 7i, j represents all of the coeicients of the (xiy) terms in the desired polynominal. Equation l may be expressed 2: i.-Pi.i 21,

where ai, j are the expansion coetiicients, and P1J 3 (x, y) are bi-variant polynomials in x and y.

In order to simplify the equations involved, a mapping of j) into integers is introduced as follows (i+j)(i|j+l) Thus, the values of I(1',j) for the integral values of and j shown in Eq. 2 are:

i 1 0 j 0 0 l 1(1', )=0 1 2 Let N--i-l-j. Then the surface ZN is of degree N if al1 (i, j) pairs in Eq. 2 `satisfy the condition I (i, j)I(0,N), and if )I(0, N-l) for at least one (i, j) pair. For example, for N=10, I(0, N)=65 and [(2, 8)=63. Now Eq. 2 is written for a surface of degree N, as

follows where z' and iN. Then the partial derivatives of ZN with respect to x, y and z are z waarnemer/ef,

where pX represents the P/x, the P1l represents the '6P/3y. The vector having ax, ay, and rz as direction numbers represents the vector normal to the surface z. In the least square sense, one would like to minimize the dierences between the above vector and the set of normal vectors given by the data. Thus, one would like to ind the a i, J such that m oz )2 (oz xl a by (5) where l represents a particular data point, m is the total number of lens data points, and 'z/'x and x/y represent the elemental slopes determined from each lens data point. Substitution from Eq. 5 to Eq. 7 yields Forming the normal equations for the expansion coefficients (N) aid' one gets where I(u, v)=0, 1 [(0, N). After rearrangement of terms, Equation 8 becomes m z z y E nhe/LPM)- m 12ga (Pi, je, a1-Pf; a]

The system of equation defined by Eq. 9 will decouple if If conditions (10 are -met by proper determination of the Pi, js then one can solve Eq. 9 for the individual expansion coeflicients (N ai, j)

Under conditions (10) all of the ai, J

terms in the summation of (9) vanish except those for which 1(1', j) :[(p, u) and Eq. 9 can be written Shown below is a way in which the P1, js may be determined so that conditions (1) will indeed hold.

am) l adding the resulting equations, and summing over the data, one gets (PrjPt. +PriPi.. v)

'1)-1 2, tarwe) a+ By virtue of conditions (10), the left side of Eq. 14 is zero. Rearranging the right side gives:

m ZKxPi-Ls'iPi-LPWF (UPiy-1.i)PiL.vi=- 1=1 1 1 Since the inside sum on the right side of Eq. 15 s zero -by conditions (10) unless k=1(p, u), one can solve for the required for the condition eO by substituting k==l(;t, v).

Equations corresponding to

Equations

13, 14 and 15 can -be similarly derived from Equation 12b for the condition =0, and the equation corresponding to Eq. 15 can be solved yfor where [(u, 10:0, 2, 5, 9

Equation 4 for the desired lens surface can now be evaluated to any desired degree, using the key Equations 11, 12, 13a, 13b, and 16, for

am): PM: Pix-i: Pili: and sill) ZN=0, oPo, o-l-al. oPl, oi-Oo, 1Po,1+2, oP2, o (4A) Since P0, :1, the rst term is a0, 0, which is a Iconstant which determines the z location of .the central point on the lens surface. Now take P1, 0=x -and P0, 1=y, noticing that condition is satisfactory by this initial choice; thus FX1, 0:1, Pxo. 1:0, P571, 0:0, and Py, 1:1. From Eq. 11,

y lpg' 1) mPwwPwl 12 and m z y (570 1PM) NPMHPM] From Eq. 12a, where 70,

from Eq, 16a, where 1%0, and

111 l? [(Pa. 02+ (P3. u2]

Now all of the terms of the ZN polynomial of Equation 4a up to and including the P2' o term 'are known and one can IWrite Assume, for example, that the data points and slopes on the assumed lens surface are as follows:

Substituting these values in Eq. [2a] one gets zN=a-0+g(0-.2+.1+.2.2+.1+.2+o-.1)

When zN is evaluated to degree N=10, for which I (0, N )=65, the total number of terms is 66 minus the ones which drop out as zero. Obviously, it would be impractical if not impossible to evaluate such an equation and utilize it to fabricate the desired lens without the use of a computer.

To evaluate zN to the desired degree, an RCA-604 computer, for example, is programmed to perform the mathematical operations involved in Equations 4, 11, 12, 13a, 13b and 16 from the input data z/x and x/y at the various points (x, y) on the ideal lens determined from the misregister data, up to the desired degree N. The output from the computer is a description of the surface of the desired lens surface on magnetic tape in APT language. APT (automatically programmed tools) is the designation given to a known system or series of computer programs that will: (a) read-in APT language statements; (lb) perform the calculations ordered or implied by these statements; calculate cut vectors to machine a part described by the APT statements; and output a punched tape to machine the part on the NC (numerical control) machine specified.

The APT language tape output from the computer is then fed into a computer which has been programmed in the APT system, for example, an IBM 7094 cornputer, to convert the surface description to an NC tape for controlling the particular NC milling machine to be used to fabricate the part, which machine may, for example, be a Pratt & Whitney machine with a Bendix controller.

The NC tape from the computer is then used to control the specified machine to cut the desired surface on a metal plate of stainless steel, for example, which will be used as a mold for fabricating the desired lens. FIG. 6 shows a transverse section of such a mold 140 with an upper surface 142 cut by the NC machine to the surface contour of the desired lens.

The desired lens, as well as the first lens 130 if used, may be made of any suitable transparent optical material, such as glass or clear plastic.

There are several ways in which the mold 140 can be used to fabricate the desired lens. One way is to use a sagging technique in which the mold surface is suitably treated to prevent glass from sticking to it, and a fiat plate 150 of optical glass is placed on the mold and heated in an oven to the softening temperature of the glass to cause the glass to sag by gravity into intimate contact with the mold surface 142, as shown in FIG. 7. In this sagging operation, the upper surface of the plate 150 is given a surface contour 162 substantially identical with the surface contour 142 of the mold. The plate 150 is then removed from the mold 140 and formed into the desired lens 160 (shown in FIG. 8) by grinding flat or otherwise removing the bottom side, which was in contact with the mold, along the plane indicated by the dashed line 164 in FIG. 7.

Since the surface 162 of lens 160 is substantially the surface described by Eq. 4, it is a sufliciently close approximation to the theoretical lens surface defined by the normal vectors (z/x, z/'yL as determined in 14 connection with FIG. 5, that the lens 160 can be used in combination with lens in a lighthouse to print a phosphor dot screen of one color with acceptable spotdot register over the entire surface of the screen.

Alternatively, the refractive properties of the two

lenses

130 and 160 can be combined in a single lens for use in the lighthouse.

As pointed out above, three separate lenses, one for each color, are designed and fabricated from the misregister data, and the three different color dot patterns. are sequentially printed with the three lenses, each alone or in combination with a lens 130.

What is claimed is:

1. In the manufacture of a cathode ray tube having a mosaic phosphor screen comprising an array of discrete phosphor elements, a multi-apertured shadow mask spaced from said screen, and means for projecting an electron beam through the apertures of said mask and onto said screen; the method of laying down said array of elements onto a screen support by a direct photographic process, said method comprising the steps of:

(a) applying a photosensitive layer to said screen support;

(lb) projecting light rays from a substantially point source along paths through said apertures and onto said layer;

(c) optically refracting said rays between said source and said support in such manner as to provide acceptable compensation at each of a multiplicity of predetermined points distributed over the entire screen area for all conditions that would otherwise cause misregister between said elements and the impingement spots thereon of the electrons of said beam at said predetermined points; and

(d) developing said layer to produce said array on said support.

2. The method of claim 1, wherein said rays are refracted by passing them through a light refracting member having a continuous curved surface with no symmetry.

3. The method of claim 2, wherein said rays are refracted by passing them successively through said light refracting member and another light refracting member of given surface contour located adjacent to the first named member.

4. In the manufacture of a color picture tube having a mosaic phosphor screen made up of a plurality of arrays of discrete phosphor elements, each array adapted to emit light of a different color, a multi-apertured shadow mask spaced from said screen, and means for projecting a plurality of electron beams, one for each array, through the apertures of said mask and onto said screen; the method of laying down said arrays of elements onto a screen support by a direct photographic process, said method comprising the steps of:

(a) applying a photosensitive layer to said screen sup- Port,

(b) projecting light rays from a substantially point source along paths through said apertures and onto said layer;

(c) optically refracting said rays between said source and said support in such manner as to provide acceptable compensation at each of a multiplicity of predetermined points distributed over the entire 'area of said screen for all conditions that would otherwise cause misregister between said phosphor elements and the impingement spots thereon of the electrons of the corresponding one of said beams at said predetermined points;

(d) developing said layer to produce one of said arrays on said support; and

(e) repeating steps (a) through (d) to lay down each of the others of said arrays on said screen support.

5. In the manufacture of a color cathode ray tube having a mosaic phosphor screen made up of a plurality of arrays of discrete phosphor elements, each array adapted when energized by an electron beam to emit light of a different color, a mu-lti-apertured shadow mask spaced from said screen, and means for projecting a plurality of electron beams, one for each array, through the apertures of said mask and onto said screen, and wherein the several electron beams in their transit from separate sources to said screen are subjected to various forces including (1) horizontal and vertical scanning forces, and (2) dynamic convergence forces, all of said forces operating to jointly shift the landing spots of said beams on said screen, and hence, cause misregister of said beam spots 'with their corresponding phosphor elements in the absence of compensation therefor; the method of laying down each of said arrays of phosphor elements onto a screen support by a direct photographic lprocess, said method comprising the steps of (a) applying a photosensitive layer to said screen support; (b) projecting light rays from a substantially point source along paths passing through said mask apertures and onto said layer; (c) optically refracting said rays between said source and said screen support in such manner as to provide acceptable compensation for said forces at a multiplicity of predetermined points distributed over the entire area of said screen support, and thereby to produce acceptable register of said beam spots with the corresponding phosphor elements over the entire screen area during operation of said tube; and (d) then developing said layer to produce said array on said screen support. 6. A shadow mask type color cathode ray tube having a mosaic screen made by the method of claim 5.

7. In the manufacture of a cathode ray tube having a mosaic phosphor screen comprising an array of discrete phosphor elements, a multi-apertured shadow mask spaced from said screen, and means for projecting an electron beam through the apertures of said mask and onto said screen; the method of making a corrective light refracting device for use in laying down said array of elements on a screen support by a direct photographic process; said method comprising the steps of (a) fabricating and mounting an apertured shadow mask adjacent to said screen support in predeter mined spaced relation therewith;

(b) applying a photosensitive layer to said screen support;

(c) projecting light rays from a substantially point source along paths through said apertures and onto said layer;

(d) developing said layer to produce said array on said support;

(e) fabricating a complete cathode ray tube using the screen resulting from step (d) with the mask of step ta);

(f) operating said tube by scanning said electron beam over the surface of said screen, and measuring the direction and amount of misregister of the beam spot and phosphor element at each of a multiplicity of predetermined points distributed over the entire surface of said screen;

(g) fabricating la light refracting device for interposition between said source and said screen, said light refracting device having a continuous curved surface closely approximating an ideal surface having a multiplicity of slopes, at points thereon corresponding respectively to said multiplicity of misregister points on said screen, required to alter the paths of light rays through said device to compensate for said misregister at said screen points.

8. The method of claim 7, wherein the last named step comprises:

(a) determining for each of said points, the elemental slope required at the corresponding point of the sur- L3 face of an ideal light refracting device interposed between said source and said mask to alter the paths of said rays and provide complete compensation for the conditions which caused said misregister and thereby eliminate said misregister at said screen point; and

(b) fabricating a light refracting device having a continuous curved surface closely approximately said elemental slopes at said corresponding points.

9. The method of claim 8, wherein the last named step comprises:

(a) fitting a bi-variant polynomial to said elemental slopes; and

(b) fabricating a light refracting device having a continuous curved surface described by said polynomial.

10. The method of claim 9, wherein the last named step comprises:

(a) fabricating a sagging mold having a surface conforming to said continuous curved surface;

(b) sagging a flat plate of transparent optical material onto said mold; and

(c) grinding flat the side of said plate that was sagged into contact with said mold.

11. The method of claim 10, wherein said mold is fabricated in step (a) by a numerical-control machining process.

12. The method of claim 7, wherein step (c) comprises the use of a first light refracting element designed to cornpensate for at least one of the conditions which tend to produce misregister of the beam spot and phosphor element, and step (g) involves the design and fabrication of a second light refracting element to be used in combination with said rst element in said subsequent printing operation to substantially compensate for all other conditions tending to cause misregister.

13. In the manufacture of a cathode ray tube having a faceplate, a mosaic phosphor screen comprising an array of discrete phosphor elements on said faceplate, a multi-apertured shadow mask spaced from said screen, and means for projecting an electron` beam through the apertures of said mask and onto said screen: an apparatus for laying down said array of elements onto said faceplate by a direct photographic process, comprising:

(a) a support for holding said faceplate and mask in the same relative position that the faceplate and mask are to occupy in said tube;

(b) means for projecting light rays from a predetermined point toward said mask and faceplate; and

(c) a light refracting device interposed in the path of said light rays for refracting said rays in such manner as to provide acceptable compensation over the entire screen area for all conditions that would otherwise cause misregister between said elements and the impingement spots thereon of the electrons of said beam in the operation of said tube, said light refracting device having an asymmetrical continuous surface described by the bi-variant polynomial:

where z is the distance of said surface at any point (x, y) thereon from a given X-Y plane, P1, (x, y) are orthogonal bi-variant polynomials in x and y, am `are the expansion coefficients of the terms P1,j(x,y), and the values of am and Pi,j(x,y) are determined from measurements of misregister at a multiplicity of predetermined points distributed over the screen surface of an uncompensated operating cathode ray tube.

14. The apparatus of claim 13, wherein said continuous lSurface of said light refracting device is a close approximation, at the points (x, y) thereon corresponding respectively to said predetermined screen points, to the elemental slopes on the surface of a theoretical light refracting device required to compensate for said conditions at said screen points.

15. In the manufacture of a cathode ray tube having a faceplate, a mosaic phosphor 4screen comprising an array of discrete phosphor elements on said faceplate, a multi-apertured shadow mask spaced from said screen, and means for projecting an electron beam through the apertures of said mask and onto said screen: the method of laying down said array of elements onto said faceplate by a direct photographic process, comprising the steps of: (a) supporting said faceplate and mask in the same relative position that the faceplate and mask are to occupy in said tube; (b) projecting light rays from a predetermined point toward said mask and faceplate; and (c) interposing a light refracting device in the path of said light rays for refracting said rays in such manner as to provide acceptable compensation over the entire screen area for all conditions that would otherwise cause misregister between said elements and the impingement spots thereon of the electrons of said beam in the operation of said tube, said light refracting device having an asymmetrical continuous surface described by the bi-variant polynomial:

where z is the distance of said surface at any point (x, y) thereon from a given X-Y plane, P1,j(x, y) are orthogonal bi-variant polynomials in x and y, am are the expansion coefficients of the terms of Pi, (x, y), and the values of am and Pi' j(x, y) are determined from measurements of misregister at a multiplicity of predetermined points distributed over the screen surface of an uncompensated operating cathode ray tube.

16. The method of claim 15, wherein said continuous surface of said light refracting device is a close approximation, at the points (x, y) thereon corresponding respectively to said predetermined screen points, to the elemental slopes on the surface of a theoretical light refracting device required to compensate for said conditions at said screen points.

References Cited UNITED STATES PATENTS 2,885,935 5/ 1959 Epstein et `al 95-1 2,986,080 5/ 1961 Burdick 95-1 3,279,340 10/ 1966 Ramberg et al. 95-1 NORTON ANSHER, Primary Examiner DAVID B. WEBSTER, Assistant Examiner UNITED STATES PATENT OFFICE CERTIFICATE 0F CORRECTION Patent No 3,476,025 November 4 1969 Fred Herzfeld et al.

It is certified that error appears in the above identified patent and that said Letters Patent are hereby corrected as shown below:

Column 6, line 48 cancel "is". Column 8 line 8 "e2" first occurrence, should read 91 Column l0, line 30, after "0" insert and line 3l, after "0" insert a semicolon; line 35, cancel "Then,"; line 36, "determined" should read determined. Then, line 37 after "O" insert and line 69, "Py", first occurrence, should read PX Column ll line 9, "i=0" should read O same line 9, in the denomina tor, insert an opening parenthesis after the opening bracket; line 25,

Y PY

u ,v should read u ,v

Column l2, line l0 "B2 (3 should read B2 0 line ll "from 9 ,l Y O ,l m

should read From line l2 cancel "and"; line 24 shoul l .Y read nil after line 26 insert and line 40 "P2 0" second occurrence should read x2 line 60 at the end of the equation, insert (.4b) line 70, the fourth and fift horizontal rows of the table should appear as shown below:

al 0 .2.1.2 .2 .1.2 0 .1 3X

Column 13, line l (2.3)?" should read (2) line 3, before 2" insert l line 8 l 2" should read l X2 line 24, "3X/ay" should read Bz/By Column l6, line 66 Signed and sealed this 3rd day of November 1970.

(SEAL) Attest:

EDWARD M.PLETCHER,JR. WILLIAM E. SCHUYLER, JR. Attesting Officer Commissioner of Patents