Automatic Correction of Centering and Convergence Errors in CRT Displays

The present application is related to application Ser. No. 423,906 filed Sept. 27, 1983, (now U.S. Pat. No. 4,485,394) by Sanjar Ghaem-Maghami and Howard Eugene Holshouser entitled "Automatic Convergence and Gray Scale Correction for Television Receivers and Projection Television Systems" and assigned to the assignee of this application.

The invention generally relates to the correction of centering and convergence errors in cathode ray tube (CRT) displays, and more particularly to a method and apparatus for the automatic correction of such errors during normal operation of the display.

CRT displays, whether they be monitors, television receivers or projection systems, periodically require adjustments to be made to maintain proper centering of the displayed image. Color CRT displays of the type having three cathode ray beams and a screen with a mosaic of phosphor dots or stripes of recurring groups of three colors must be adjusted to maintain the convergence of the three beams over the visible surface of the screen. An analogous adjustment must be made for projection displays employing three projection CRTs. These adjustments are initially made at the factory, but with age, temperature and other environmental conditions, it is necessary to readjust centering and convergence in order to maintain the quality of the displayed image. Ordinarily this is accomplished by a skilled technician with test instruments. The test instruments used to measure convergence often resort to the use of an appliance that is placed over the CRT screen to facilitate detection of the landing point of the cathode ray beam. Such an appliance obscures the screen, and therefore these instruments are not intended to be used simultaneously with the viewing of the display. Examples of such instruments are U.S. Pat. No. 4,001,877 issued to Theodore Frederick Simpson and U.S. Pat. No. 4,035,834 issued to Anthony M. Drury.

The Simpson patent describes a test instrument that employs a photosensitive array comprising a plurality of individual photo cells, this array being placed over the CRT screen. Further, a special post-deflection coil is required to introduce magnetic fields in the region just forward of the deflection yoke to displace the scanned beams in a controlled pattern from their normal landing points on the screen. The displaced beam causes the emission of an error color, the intensity of which is measured by those photo cells which are sensitive to the error color emitted. The intensity of a reference color emitted by the phosphor deposits stimulated by the undisplace beam is then measured, and the ratio of the error color to the reference is calculated for each measurement location on the screen. The largest ratio is displayed as an indication of the color purity tolerance of the CRT. The Simpson test instrument is used primarily as a quality control device in the manufacture of color CRTs.

The patent to Drury describes a beam landing indicator for a color CRT which also employs a holder for positioning a plurality of photo cells over the screen of the CRT. While the Drury instrument does not require a special deflection coil, it does employ a special deflection generator in order to produce a clockwise rotation of the beam landing shift of the beam. This rotation is stepped in increments which occur once each vertical field of the television raster. Light variations sensed by the photo cells are combined with a reference signal to control the dot location on an oscilloscope display of the vector beam landing error. The technician can then make purity adjustments and yoke adjustments of the CRT by observing the oscilloscope display.

Automating the adjustment of color television receivers is also known. An example is described in U.S. Pat. No. 3,962,722 issued to Walter S. Ciciora. More specifically, the Ciciora patent describes a color television setup apparatus for use in a factory. Once again, a holder positions a plurality of photo cells over the CRT screen in such a manner as to obscure the view of the screen. Patterns indicative of the characteristics of contrast, brightness, color and tint are displayed on the CRT. The photo cells develop corresponding electrical signals which are supplied to circuitry that energizes a plurality of bi-directional motors that are engageable with the receiver contrast, tint, brightness and color level adjustment elements.

While the systems described by Simpson, Drury and Ciciora are useful in a factory or shop environment, what is needed is an automatic means for adjustment of convergence which is part of the CRT display. In this way, the display would be continuously maintained in proper adjustment for optimum viewing. Such a system has been provided in the above-referenced application Ser. No. 423,906 (now U.S. Pat. No. 4,485,394) filed by Ghaem-Maghami and Holshauser. According to that invention, a system is provided for the automatic correction of convergence and gray scale which employs light sensors, either singly or in an array, on or adjacent to the beam landing surface of a CRT or on or adjacent the screen of a projection receiver. The sensors can be placed proximate the overscanned area of the raster such that they are outside the normal viewing area, or in the viewing area if the sensors are made sufficiently small. In the vicinity of a sensor, two of the three cathode ray tubes or electron guns are blanked. As the light beam, in the case of a projection system, crosses the sensor, an output is produced. This output is processed to obtain accurate timing characteristics. Since the position of the sensor is known in terms of counts in both the vertical and horizontal directions, error signals can be developed by comparing the timing of the sensor output with the proper count. These error signals are used to develop vertical and horizontal correction signals to correct the convergence of the one cathode ray tube or gun. The process is then repeated for the remaining two cathode ray tubes or guns. The output of the sensor is also amplified by gated amplifiers for each of the cathode ray tubes or guns in sequential order, and the outputs of these amplifiers are compared to a preset value to develop error signals. These error signals are used to set gun drives to correct the gray scale.

The system described in application Ser. No. 423,906 (now U.S. Pat. No. 4,485,394) filed by Ghaem-Maghami and Holshauser generally operates well under most conditions; however, that system is affected by a change in picture size due to a change from a dark scene to a light scene, a change in line voltage, or a change in the blanking pulse transmitted by the television station, for example. The system is sensitive to both the amplitude and speed (slope) of the incoming light pulse. So while the system according to Ghaem-Maghami and Holshauser operates satisfactorily, it is nevertheless desirable to improve on the performance of that system.

It is therefore an object of the present invention to provide a system for the automatic correction of centering errors in cathode ray tube displays which is substantially insensitive to the amplitude and slope of the outputs from the light sensors.

It is another object of the invention to provide an improved system for the automatic correction of convergence errors in color cathode ray tube displays which accomplishes the convergence correction within a high degree of accuracy.

According to the present invention, two or more light pulses are generated in the vertical and horizontal blanking intervals. A corresponding number of optical sensors are positioned about the periphery of the viewable display to receive the light pulses. The outputs of these optical sensors are connected to a feedback-type controller. The controller corrects for centering errors by producing an error signal which alters the centering of the displayed image until the light pulses fall on the sensors. When the process is applied to all three colors sequentially, static convergence errors are corrected. If, in addition, information about the width and height of the actual scan is known, a device which can correct the instantaneous position of the beam is used, and a multiplicity of light pulses and sensors are used, dynamic convergence correction can also be accomplished. The present invention accomplishes a similar objective to that of the invention disclosed in Ser. No. 423,906 (now U.S. Pat. No. 4,485,394) filed by Ghaem-Maghami and Holshauser; however, this invention accomplishes that objective in a different way.

The invention will be better understood from the following detailed description of the invention with reference to the drawings, in which:

FIG. 1 is a plan view of a CRT display showing the relationship between the displayed area and the actual area scanned;

FIG. 2 is a block diagram of a basic embodiment of the invention employing two optical sensors and a feedback-type controller;

FIG. 3 is a modification of the system shown in FIG. 2 for accomplishing static convergence corrections;

FIG. 4 is a further modification of the basic embodiment employing a plurality of optical sensors with the feedback-type controller for accomplishing dynamic convergence corrections;

FIG. 5 is a block diagram of a microprocessor based feedback-type controller for the embodiment shown in FIG. 3;

FIG. 6 is a schematic diagram of a blip generator which is used with the controller shown in FIG. 5;

FIG. 7 is a schematic diagram of an optical sensor circuit which is used with the controller shown in FIG. 5; and

FIG. 8 is a block and schematic diagram of the 1 to 8 analog switch, the switched current source circuit and the sample and hold circuits shown in FIG. 5;

In a raster-scan CRT display, the electron beam actually scans past the edges of the displayed image, but it is normally blanked during this time. This is illustrated in FIG. 1 where the arrows VB indicate the vertical blanking intervals and the arrows HB indicate the horizontal blanking intervals. Notice also that the display area is bisected horizontally by a major axis and vertically by a minor axis. It is possible to generate one or more pulses which are timed to occur during the vertical and horizontal blanking intervals. These pulses are used in the present invention to unblank the display and cause the emission of a blip of light. Since this blip occurs during the blanking interval and the displayed image is normally set to fill the entire area which may be seen by the viewer, these light blips would not be seen.

If the centering of the displayed image is changed, the light blips will move along with the image. Thus, if optical sensors are positioned where the blips should fall physically, a feedback-type controller can be implemented which will correct for centering errors by producing error signals which alters the centering of the displayed image until the light blips fall on the optical sensors. A block diagram of this system is shown in FIG. 2 wherein the controller 10 receives as inputs the outputs of masked optical sensors 12 and 14 placed along the major and minor axes, respectively. Horizontal and vertical timing pulses are provided to both the controller 10 and the blip generator 16. In response to these timing pulses, the blip generator 16 provides an output which unblanks the video circuits causing light pulses to be generated at known timing intervals. The rectangles 17 and 19 surrounding the optical sensors 12 and 14, respectively, indicate blips of light produced when the displayed image is properly centered. Since the optical sensors are masked, they provide an output only if the blips of light are physically coincident with the sensors as indicated by the drawing. If the light blips 17 and 19 do not fall on the optical sensors, the controller 10 begins an iterative routine that produces an error signal which is supplied to the centering circuits.

If the process is applied to all three colors sequentially, using the same sensors 12 and 14, static convergence errors will also be corrected. This is accomplished by the system shown in FIG. 3 wherein the output of the blip generator 16 is supplied to a 1 of 3 analog switch 18 that generates in sequence unblanking pulses for the red, green and blue video signals. The output of the controller 10 is supplied, in corresponding sequence, to the red, green and blue centering circuits represented by block 20.

If, in addition, information about the width and height of the actual scan is known, a device which can correct the instantaneous position of the beam is used, and a multiplicity of blips and sensors are used, dynamic convergence correction can also be accomplished. Such a system is shown in FIG. 4 wherein a first plurality of vertical optical sensors 12A, 12B, 12C, 12D, and 12E are shown along the left edge of the display, a second plurality of vertical optical sensors 13A, 13B, 13C, 13D, and 13E are shown along the right edge of the display, a first plurality of horizontal sensors 14A, 14B, 14C, 14D, and 14E are shown along the top edge of the display, and a second plurality of horizontal sensors 15A, 15B, 15C, 15D, and 15E are shown along the bottom edge of the display. The outputs of these optical sensors are supplied to the input of the controller 10 which generates an error output to the red, green and blue positioners represented by block 22.

Referring now to FIG. 5, there is shown a microprocessor based feedback controller for the system shown in FIG. 3. Vertical and horizontal sync pulses are supplied to vertical and horizontal pulse shapers 30 and 32, respectively. The outputs of these pulse shapers are provided to the blip generator 16 which is described in more detail hereinafter with reference to FIG. 6. The output of the vertical pulse generator 30 is also supplied to the microprocessor 34 to supply a timing reference. The output of the blip generator 16 is supplied to the video circuits as an unblanking signal and also to the video control 38. The video signal 38 may be the analog switch shown in FIG. 6. Note that a two-bit binary input is supplied to the video control from the microprocessor 34. This two-bit input is used to make the one-of-three switch selection. The vertical and horizontal light sensors 40 and 42 provide inputs to the microprocessor 34 when a light pulse is produced at their respective physical locations on the periphery of the display screen as shown in FIG. 3. If there are no light sensor inputs, the microprocessor 34 begins an iterative process that generates error signals to correct for red, green or blue centering errors as may be appropriate. More specifically, if the light sensors 40 and/or 42 do not receive a pulse of light when they are supposed to, the microprocessor begins a process of ramping up and down to achieve centering and static convergence. This is accomplished by means of the charge pump 44, the one-of-eight analog switch 46 and the sample and hold circuits 48 to 58. These are shown in more detail in FIG. 8.

Turning now to FIG. 6, there is shown the blip generator used in the controller system of FIG. 5. The vertical sync pulse is applied to the bases of NPN transistors T1 and T2. A capacitor C1 is connected across the emitter-collector circuit of transistor T1 so that when a positive going vertical sync pulse is applied to the base of transistor T1, the transistor conducts discharging capacitor C1. At the same time, transistor T2 conducts shorting the output of comparator A1 to ground. When the vertical sync pulse ends and transistors T1 and T2 are again nonconducting, capacitor C1 charges from the +5 V supply through load resistor R1. The ramping voltage across capacitor C1 is supplied to the negative input of comparator A1 as an analog timing signal. The positive or reference input to the comparator A1 is connected to the wiper of a potentiometer P1, the winding of which is connected between the +5 V supply and ground. The purpose of potentiometer P1 is to establishe the vertical blip height. Thus, initially during a vertical sync pulse, the output of comparator A1 is forced low by the conduction of transistor T2, but immediately thereafter, the output of the comparator A1 is high. The output of the comparator A1 remains high until the ramping voltage across capacitor C1 exceeds the voltage set by potentiometer P1. In effect, a number of horizontal scans at the beginning of the field is selected by adjusting the potentiometer P1.

The output of comparator A1 is connected to pin 2 of a monostable multivibrator M1 which may be one half of a standard commercial 74123 integrated circuit (IC). A positive voltage on pin 2 enables the monostable M1. The horizontal flyback pulses are supplied to the base of an NPN transistor T3 connected in grounded emitter fashion with a diode D1 connected between the base-emitter circuit. Thus, transistor T3 functions as a clipping inverter. The negative going pulses from the collector of transistor T3 are supplied to pin 1 of the monostable M1. Connected between pin 15 of the monostable M1 and the +5 V supply is a variable resistor P2, and connected between pins 14 and 15 is a capacitor C2. The monostable M1 is triggered by the negative going horizontal timing pulses and times out at a time determined by the RC time constant of P2 and C2. The monostable output goes low immediately and then returns high at the end of the time period determined by P2 and C2. Thus, the monostable M1 produces a negative going pulse having a width determined by the capacitance of capacitor C2. In this way, the monostable M1 provides a series of negative going pulses, one each horizontal scan, as long as it is enabled by a positive output from comparator A1. These pulses are produced at some point along the horizontal scan as determined by the RC time constant P2C2. In other words, adjustment of the potentiometer P2 is made to locate the blip on the minor axis (see FIG. 1) at the top of the display. This blip occurs outside the field of view as determined by the setting of potentiometer P1. When the output of monostable M1 returns to the high state, a pulse is coupled to the base of transistor T8 causing its collector to go low. The width of this pulse is determined by the value of the capacitor in the base of transistor T8 and thereby determines the width of the pulse on the minor axis.

The comparator A2 is similar to comparator A1 receiving as it does the ramping voltage across capacitor C1 at its negative input and a voltage from potentiometer P3 at its positive or reference input. At the beginning of a vertical scan, the output of comparator A2 is positive, but at some point as determined by the setting of potentiometer P3, the output of comparator A2 goes low. The negative going output of comparator A2 triggers monostable M2 which times out after a period of between one and two horizontal scans or about 75 .mu.sec. During this time, the output of monostable M2 is positive and enables monostable M3. This provides a window during which only one negative going horizontal timing pulse from transistor T3 can trigger the monostable M3. Like monostable M1, the monstable M3 is provided with a variable resistor P4 connected between pin 15 and the +5 V supply and a capacitor C3 connected between pins 14 and 15. The RC time constant of P4C3 determines the width of the output pulse. Thus, by adjusting potentiometer P3, the pulse at the left edge of the display can be positioned to fall on the major axis (see FIG. 1), and by adjusting potentiometer P4, the pulse width can be adjusted so as to be outside the field of view.

The negative going pulses from monostables M1 and M3 are capacitively coupled to the bases of NPN transistors T8 and T4, respectively. These transistors serve as drivers and their collectors are connected to pin 13 of an analog switch SW1. The switch SW1 may be one half of a standard commercial CD4052 IC which functions as a one-of-three switch controlled by a two-bit binary code supplied to pins 9 and 10 from microprocessor 34 (see FIG. 5). Switch SW1 connects pin 13 to pins 12, 14 or 15 as determined by the binary code. These pins are, in turn, connected to the bases of NPN transistors T5, T6 and T7 which supply unblanking pulses to the video circuits.

The blip that is generated on the vertical axis requires applying a narrow pulse to the cathode of the electron gun. The width of this pulse, and thus the bandwidth needed in the video circuits, is dependent upon the viewable diagonal measurement (DM), the allowable convergence error (CE), the aspect ratio (AR), and the active scan time (AST). The aspect ratio is defined as the ratio of the width of the displayed image to its height. The actual size of the image is specified as the viewable diagonal. Thus, AR=h/v, where h and v are the horizontal and vertical dimensions, respectively, of the displayed image, and DM, h and v forms a right triangle so that

But since

then ##EQU1## Substituting Equation 3 into Equation 1, ##EQU2## Factoring, ##EQU3## Taking the square root of both sides ##EQU4## If the sweep is linear, the ratio of the width of the blip to the active scan time is the same as the ratio of the allowable convergence error to h. That is, ##EQU5## Solving for the pulse width

Substituting Equation 7 into Equation 9 and simplifying ##EQU6##

Consider the following specific example where the values of the terms are DM=40, AR=1.33, and AST=50 .mu.sec. The allowable convergence error is .+-.1/32. Thus, ##EQU7## Assuming a Gaussian rolloff, 10%-90%-10% drive leads to ##EQU8## In contrast, consider the example of the system disclosed by Ghaem-Maghami and Holshouser in application Ser. No. 423,906 (now U.S. Pat. No. 4,485,394). The 3.58 MHz fed to the horizontal counter in that system limits the horizontal resolution:

The active scan time is 50 .mu.s; therefore, each horizontal scan can be divided into ##EQU9## A 40" projection receiver has a horizontal screen dimension of 0.8(40)=32 inches. Thus, the finest resolution of which the Ghaem-Maghami and Holshouser system is capable is

In contrast, the present invention is capable of converging to within 1/32" (0.03125). This represents an improvement in the convergence error of almost six times. Obviously, the digital clocking scheme of Ghaem-Maghami and Holshouser can be improved by increasing the number of bits in the horizontal counter and clocking it at a faster rate. The present invention offers a simpler solution.

FIG. 7 shows the light sensor circuit used in the system of FIG. 5. A sensitive photodiode D2 is provided with an aperture (not shown) so that it will respond to the blip of light only if it occurs at the correct physical location. The diode D2 may be a Siemens BPW34 or equivalent and is connected in the base-collector circuit of an NPN transistor T9 having a gain of 200. The collector and emitter leads of transistor T9 are respectively connected by twisted pair to the +5 V supply and a load resistor R2. Thus, the transistor is connected as an emitter follower. The load resistor R2 is part of an RC bandpass filter F1 which is connected to the positive input of operational amplifier A3. Negative feedback for amplifier A3 is derived from an voltage divider to provide a gain of 100. The output of amplifier A3 is connected to a diode D3 followed by a low pass RC filter F2. The diode D3 and the filter F2 function as a peak detector to provide an output pulse to the microprocessor 34 (see FIG. 5) when a blip of light falls on the photodiode D2.

In FIG. 8, the switched current source circuit comprises a transistor T10 which acts as a gated current source. Transistor T11 is used to reset the sample and hold circuits before a charging current is supplied by transistor T10. A 3-bit binary input to the switch 46 from the microprocessor 34 selects one of six output lines to which the junction of the collectors of transistors T10 and T11 is connected. The switch 46 may be implemented with a standard commercial CD4051 IC and functions as a one-of-six analog switch controlled by the three-bit binary code supplied by the microprocessor 34. The six output lines of switch 46 are connected to the inputs of the sample and hold circuits each of which comprises a capacitor and an operational amplifier. Thus, when the microprocessor selects one of the sample and hold circuits, the capacitor for that circuit is first discharged by a reset pulse to the base of transistor T11. Then, the current supplied by transistor T10 charges the selected capacitor to a voltage determined by the duration of the pump pulse from the microprocessor 34 so that a correcting voltage is supplied at the output of the appropriate operational amplifier.

Returning now to FIG. 5, the microprocessor 34 is provided with a nonvolatile random access memory (NOVRAM) 36. The NOVRAM is preferably used to store factory setup data, e.g. nominal values for convergence are stored there during production. The microprocessor then uses this data as a starting point for convergence correction. The following listing is a specific example of a program which may be used in the practice of the invention. This listing is in assembly language for a Z80 microprocessor. ##SPC1##