Patent Application
20230136688
This invention is directed generally to displays that use light emissive devices such as OLEDs and, more particularly, to image correction using stress history modelling in such displays to compensate for aging of the light emissive devices.
Active matrix organic light emitting device (“AMOLED”) displays offer the advantages of lower power consumption, manufacturing flexibility, and faster refresh rate over conventional liquid crystal displays. In contrast to conventional liquid crystal displays, there is no backlighting in an AMOLED display as each pixel consists of different colored OLEDs emitting light independently. The OLEDs emit light based on current supplied through a drive transistor. The drive transistor is typically a thin film transistor (TFT). The power consumed in each pixel has a direct relation with the magnitude of the generated light in that pixel.
During operation of an organic light emitting diode device, it undergoes degradation, which causes light output at a constant current to decrease over time. The OLED device also undergoes an electrical degradation, which causes the current to drop at a constant bias voltage over time. These degradations are caused primarily by stress related to the magnitude and duration of the applied voltage on the OLED and the resulting current passing through the device. Such degradations are compounded by contributions from the environmental factors such as temperature, humidity, or presence of oxidants over time. The aging rate of the thin film transistor devices is also environmental and stress (bias) dependent. The aging of the drive transistor and the OLED may be properly determined via calibrating the pixel against stored historical data from the pixel at previous times to determine the aging effects on the pixel. Accurate aging data is therefore necessary throughout the lifetime of the display device.
In one compensation technique for OLED displays, the aging (and/or uniformity) of a panel of pixels is extracted and stored in lookup tables as raw or processed data. Then a compensation module uses the stored data to compensate for any shift in electrical and optical parameters of the OLED (e.g., the shift in the OLED operating voltage and the optical efficiency) and the backplane (e.g., the threshold voltage shift of the TFT), hence the programming voltage of each pixel is modified according to the stored data and the video content. The compensation module modifies the bias of the driving TFT in a way that the OLED passes enough current to maintain the same luminance level for each gray-scale level. In other words, a correct programming voltage properly offsets the electrical and optical aging of the OLED as well as the electrical degradation of the TFT.
The electrical parameters of the backplane TFTs and OLED devices are continuously monitored and extracted throughout the lifetime of the display by electrical feedback-based measurement circuits. Further, the optical aging parameters of the OLED devices are estimated from the OLED's electrical degradation data. However, the optical aging effect of the OLED is dependent on the stress conditions placed on individual pixels as well, and since the stresses vary from pixel to pixel, accurate compensation is not assured unless the compensation tailored for a specific stress level is determined.
There is therefore a need for efficient extraction of characterization correlation curves of the optical and electrical parameters that are accurate for stress conditions on active pixels for compensation for aging and other effects. There is also a need for having a variety of characterization correlation curves for a variety of stress conditions that the active pixels may be subjected to during operation of the display. There is a further need for accurate compensation systems for pixels in an organic light emitting device based display.
According to a first aspect, there is provided a method of compensating for degradation of pixels of a display panel including organic light emitting devices (OLED), said method comprising: determining a stress history model for the pixels generating model parameters including a drive level power factor which linearizes the correction drive level versus accumulated stress in the model; storing the correction parameters in a memory of the display; maintaining a stress history of the pixels of the display according to the stress history model with use of the model parameters; generating stress history corrected drive levels for the pixels with use of said parameters generating corrected drive levels for the pixels; and displaying images on the display with use of the corrected drive levels.
In some embodiments, the stress history model defines stress as a summation of drive levels raised to the power factor, up to some constant factor.
In some embodiments, determining the stress history model comprises stressing each population of pixels while periodically testing the pixels at one or more testing drive levels to determine correction drive level versus accumulated stress.
In some embodiments, stressing each population of pixels comprises continuously driving the pixels at predetermined stress drive levels.
In some embodiments, the model parameters include parameters defining a pixel gain and an offset in terms of the accumulated stress.
In some embodiments, the model parameters include parameters which define a linear relationship between pixel gain and accumulated stress and a linear relationship between offset and accumulated stress.
According to a second aspect, there is provided a system for compensating for degradation of pixels of a display panel including organic light emitting devices (OLED), said system comprising: the display panel; optical correction processing and a controller adapted to: determine a stress history model for the pixels generating model parameters including a drive level power factor which linearizes the correction drive level versus accumulated stress in the model; store the correction parameters in a memory of the display; and a controller of display panel adapted to: maintain a stress history of the pixels of the display according to the stress history model with use of the model parameters; generate stress history corrected drive levels for the pixels with use of said parameters generating corrected drive levels for the pixels; and display images on the display with use of the corrected drive levels.
In some embodiments, the optical correction processing and controller are adapted to determine the stress history model including testing populations of pixels of each color for the type of display panel and determining a drive level power factor for each population of pixels.
In some embodiments, the optical correction processing and controller are adapted to determine the stress history model including stressing each population of pixels while periodically testing the pixels at one or more testing drive levels to determine correction drive level versus accumulated stress.
Additional aspects of the invention will be apparent to those of ordinary skill in the art in view of the detailed description of various embodiments, which is made with reference to the drawings, a brief description of which is provided below.
The invention may best be understood by reference to the following description taken in conjunction with the accompanying drawings.
and 15B are flow charts of procedures for identifying the stress condition of a device based on the rate of change or absolute value of a parameter of the device and the rate of change or absolute value of a parameter of another device.
While the invention is susceptible to various modifications and alternative forms, specific embodiments have been shown by way of example in the drawings and will be described in detail herein. It should be understood, however, that the invention is not intended to be limited to the particular forms disclosed. Rather, the invention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the invention as defined by the appended claims.
The display system 100 may also include a current source circuit, which supplies a fixed current on current bias lines. In some configurations, a reference current can be supplied to the current source circuit. In such configurations, a current source control controls the timing of the application of a bias current on the current bias lines. In configurations in which the reference current is not supplied to the current source circuit, a current source address driver controls the timing of the application of a bias current on the current bias lines.
As is known, each pixel 104 in the display system 100 needs to be programmed with information indicating the brightness of the light emitting device in the pixel 104. A frame defines the time period that includes a programming cycle or phase during which each and every pixel in the display system 100 is programmed with a programming voltage indicative of a brightness and a driving or emission cycle or phase during which each light emitting device in each pixel is turned on to emit light at a brightness commensurate with the programming voltage stored in a storage element. A frame is thus one of many still images that compose a complete moving picture displayed on the display system 100. There are at least two schemes for programming and driving the pixels: row-by-row, or frame-by-frame. In row-by-row programming, a row of pixels is programmed and then driven before the next row of pixels is programmed and driven. In frame-by-frame programming, all rows of pixels in the display system 100 are programmed first, and all of the frames are driven row-by-row. Either scheme can employ a brief vertical blanking time at the beginning or end of each period during which the pixels are neither programmed nor driven.
The components located outside of the pixel array 102 may be disposed in a peripheral area 106 around the pixel array 102 on the same physical substrate on which the pixel array 102 is disposed. These components include the gate driver 108, the source driver 110, and the optional supply voltage control 114. Alternately, some of the components in the peripheral area can be disposed on the same substrate as the pixel array 102 while other components are disposed on a different substrate, or all of the components in the peripheral area can be disposed on a substrate different from the substrate on which the pixel array 102 is disposed. Together, the gate driver 108, the source driver 110, and the supply voltage control 114 make up a display driver circuit. The display driver circuit in some configurations may include the gate driver 108 and the source driver 110 but not the supply voltage control 114.
The display system 100 further includes a current supply and readout circuit 120, which reads output data from data output lines, VD [k], VD [k+1], and so forth, one for each column of active pixels 104 in the pixel array 102. A set of optional reference devices such as reference pixels 130 is fabricated on the edge of the pixel array 102 outside the active pixels 104 in the peripheral area 106. The reference pixels 130 also may receive input signals from the controller 112 and may output data signals to the current supply and readout circuit 120. The reference pixels 130 include the drive transistor and an OLED but are not part of the pixel array 102 that displays images. As will be explained below, different groups of reference pixels 130 are placed under different stress conditions via different current levels from the current supply circuit 120. Because the reference pixels 130 are not part of the pixel array 102 and thus do not display images, the reference pixels 130 may provide data indicating the effects of aging at different stress conditions. Although only one row and column of reference pixels 130 is shown in
The reference pixel 130 may be stressed at a certain current level by applying a constant voltage to the programming voltage input line 216. As will be explained below, the voltage output measured from the monitoring line 218 based on a reference voltage applied to the programming voltage input line 216 allows the determination of electrical characterization data for the applied stress conditions over the time of operation of the reference pixel 130. Alternatively, the monitor line 218 and the programming voltage input line 216 may be merged into one line (i.e., Data/Mon) to carry out both the programming and monitoring functions through that single line. The output of the photo-sensor 132 allows the determination of optical characterization data for stress conditions over the time of operation for the reference pixel 130.
The display system 100 in
As the OLED material of an active pixel 104 ages, the voltage required to maintain a constant current for a given level through the OLED increases. To compensate for electrical aging of the OLEDs, the memory 118 stores the required compensation voltage of each active pixel to maintain a constant current. It also stores data in the form of characterization correlation curves for different stress conditions that is utilized by the controller 112 to determine compensation voltages to modify the programming voltages to drive each OLED of the active pixels 104 to correctly display a desired output level of luminance by increasing the OLED's current to compensate for the optical aging of the OLED. In particular, the memory 118 stores a plurality of predefined characterization correlation curves or functions, which represent the degradation in luminance efficiency for OLEDs operating under different predetermined stress conditions. The different predetermined stress conditions generally represent different types of stress or operating conditions that an active pixel 104 may undergo during the lifetime of the pixel. Different stress conditions may include constant current requirements at different levels from low to high, constant luminance requirements from low to high, or a mix of two or more stress levels. For example, the stress levels may be at a certain current for some percentage of the time and another current level for another percentage of the time. Other stress levels may be specialized such as a level representing an average streaming video displayed on the display system 100. Initially, the base line electrical and optical characteristics of the reference devices such as the reference pixels 130 at different stress conditions are stored in the memory 118. In this example, the baseline optical characteristic and the baseline electrical characteristic of the reference device are measured from the reference device immediately after fabrication of the reference device.
Each such stress condition may be applied to a group of reference pixels such as the reference pixels 130 by maintaining a constant current through the reference pixel 130 over a period of time, maintaining a constant luminance of the reference pixel 130 over a period of time, and/or varying the current through or luminance of the reference pixel at different predetermined levels and predetermined intervals over a period of time. The current or luminance level(s) generated in the reference pixel 130 can be, for example, high values, low values, and/or average values expected for the particular application for which the display system 100 is intended. For example, applications such as a computer monitor require high values. Similarly, the period(s) of time for which the current or luminance level(s) are generated in the reference pixel may depend on the particular application for which the display system 100 is intended.
It is contemplated that the different predetermined stress conditions are applied to different reference pixels 130 during the operation of the display system 100 in order to replicate aging effects under each of the predetermined stress conditions. In other words, a first predetermined stress condition is applied to a first set of reference pixels, a second predetermined stress condition is applied to a second set of reference pixels, and so on. In this example, the display system 100 has groups of reference pixels 130 that are stressed under 16 different stress conditions that range from a low current value to a high current value for the pixels. Thus, there are 16 different groups of reference pixels 130 in this example. Of course, greater or lesser numbers of stress conditions may be applied depending on factors such as the desired accuracy of the compensation, the physical space in the peripheral area 106, the amount of processing power available, and the amount of memory for storing the characterization correlation curve data.
By continually subjecting a reference pixel or group of reference pixels to a stress condition, the components of the reference pixel are aged according to the operating conditions of the stress condition. As the stress condition is applied to the reference pixel during the operation of the system 100, the electrical and optical characteristics of the reference pixel are measured and evaluated to determine data for determining correction curves for the compensation of aging in the active pixels 104 in the array 102. In this example, the optical characteristics and electrical characteristics are measured once an hour for each group of reference pixels 130. The corresponding characteristic correlation curves are therefore updated for the measured characteristics of the reference pixels 130. Of course, these measurements may be made in shorter periods of time or for longer periods of time depending on the accuracy desired for aging compensation.
Generally, the luminance of the OLED 204 has a direct linear relationship with the current applied to the OLED 204. The optical characteristic of an OLED may be expressed as:
L=O*I
In this equation, luminance, L, is a result of a coefficient, O, based on the properties of the OLED multiplied by the current I. As the OLED 204 ages, the coefficient O decreases and therefore the luminance decreases for a constant current value. The measured luminance at a given current may therefore be used to determine the characteristic change in the coefficient, O, due to aging for a particular OLED 204 at a particular time for a predetermined stress condition.
The measured electrical characteristic represents the relationship between the voltage provided to the drive transistor 202 and the resulting current through the OLED 204. For example, the change in voltage required to achieve a constant current level through the OLED of the reference pixel may be measured with a voltage sensor or thin film transistor such as the monitoring transistor 210 in
I=k*(V−e)a
In this equation, the current is determined by a constant, k, multiplied by the input voltage, V, minus a coefficient, e, which represents the electrical characteristics of the drive transistor 202. The voltage therefore has a power law relation by the variable, a, to the current, I. As the transistor 202 ages, the coefficient, e, increases thereby requiring greater voltage to produce the same current. The measured current from the reference pixel may therefore be used to determine the value of the coefficient, e, for a particular reference pixel at a certain time for the stress condition applied to the reference pixel.
As explained above, the optical characteristic, O, represents the relationship between the luminance generated by the OLED 204 of the reference pixel 130 as measured by the photo sensor 132 and the current through the OLED 204 in
Then by using the electrical and optical characteristics measured from the reference pixel, a characterization correlation curve (or function) is determined for the predetermined stress condition over time. The characterization correlation curve provides a quantifiable relationship between the optical degradation and the electrical aging expected for a given pixel operating under the stress condition. More particularly, each point on the characterization correlation curve determines the correlation between the electrical and optical characteristics of an OLED of a given pixel under the stress condition at a given time where measurements are taken from the reference pixel 130. The characteristics may then be used by the controller 112 to determine appropriate compensation voltages for active pixels 104 that have been aged under the same stress conditions as applied to the reference pixels 130. In another example, the baseline optical characteristic may be periodically measured from a base OLED device at the same time as the optical characteristic of the OLED of the reference pixel is being measured. The base OLED device either is not being stressed or being stressed on a known and controlled rate. This will eliminate any environmental effect on the reference OLED characterization.
Due to manufacturing processes and other factors known to those skilled in the art, each reference pixel 130 of the display system 100 may not have uniform characteristics, resulting in different emitting performances. One technique is to average the values for the electrical characteristics and the values of the luminance characteristics obtained by a set of reference pixels under a predetermined stress condition. A better representation of the effect of the stress condition on an average pixel is obtained by applying the stress condition to a set of the reference pixels 130 and applying a polling-averaging technique to avoid defects, measurement noise, and other issues that can arise during application of the stress condition to the reference pixels. For example, faulty values such as those determined due to noise or a dead reference pixel may be removed from the averaging. Such a technique may have predetermined levels of luminance and electrical characteristics that must be met before inclusion of those values in the averaging. Additional statistical regression techniques may also be utilized to provide less weight to electrical and optical characteristic values that are significantly different from the other measured values for the reference pixels under a given stress condition.
In this example, each of the stress conditions is applied to a different set of reference pixels. The optical and electrical characteristics of the reference pixels are measured, and a polling-averaging technique and/or a statistical regression technique are applied to determine different characterization correlation curves corresponding to each of the stress conditions. The different characterization correlation curves are stored in the memory 118. Although this example uses reference devices to determine the correlation curves, the correlation curves may be determined in other ways such as from historical data or predetermined by a manufacturer.
During the operation of the display system 100, each group of the reference pixels 130 may be subjected to the respective stress conditions and the characterization correlation curves initially stored in the memory 118 may be updated by the controller 112 to reflect data taken from the reference pixels 130 that are subject to the same external conditions as the active pixels 104. The characterization correlation curves may thus be tuned for each of the active pixels 104 based on measurements made for the electrical and luminance characteristics of the reference pixels 130 during operation of the display system 100. The electrical and luminance characteristics for each stress condition are therefore stored in the memory 118 and updated during the operation of the display system 100. The storage of the data may be in a piecewise linear model. In this example, such a piecewise linear model has 16 coefficients that are updated as the reference pixels 130 are measured for voltage and luminance characteristics. Alternatively, a curve may be determined and updated using linear regression or by storing data in a look up table in the memory 118.
To generate and store a characterization correlation curve for every possible stress condition would be impractical due to the large amount of resources (e.g., memory storage, processing power, etc.) that would be required. The disclosed display system 100 overcomes such limitations by determining and storing a discrete number of characterization correlation curves at predetermined stress conditions and subsequently combining those predefined characterization correlation curves using linear or nonlinear algorithm(s) to synthesize a compensation factor for each pixel 104 of the display system 100 depending on the particular operating condition of each pixel. As explained above, in this example there are a range of 16 different predetermined stress conditions and therefore 16 different characterization correlation curves stored in the memory 118.
For each pixel 104, the display system 100 analyzes the stress condition being applied to the pixel 104, and determines a compensation factor using an algorithm based on the predefined characterization correlation curves and the measured electrical aging of the panel pixels. The display system 100 then provides a voltage to the pixel based on the compensation factor. The controller 112 therefore determines the stress of a particular pixel 104 and determines the closest two predetermined stress conditions and attendant characteristic data obtained from the reference pixels 130 at those predetermined stress conditions for the stress condition of the particular pixel 104. The stress condition of the active pixel 104 therefore falls between a low predetermined stress condition and a high predetermined stress condition.
The following examples of linear and nonlinear equations for combining characterization correlation curves are described in terms of two such predefined characterization correlation curves for ease of disclosure; however, it is to be understood that any other number of predefined characterization correlation curves can be utilized in the exemplary techniques for combining the characterization correlation curves. The two exemplary characterization correlation curves include a first characterization correlation curve determined for a high stress condition and a second characterization correlation curve determined for a low stress condition.
The ability to use different characterization correlation curves over different levels provides accurate compensation for active pixels 104 that are subjected to different stress conditions than the predetermined stress conditions applied to the reference pixels 130.
The limited number of reference pixels 130 and corresponding limited numbers of stress conditions may require the use of averaging or continuous (moving) averaging for the specific stress condition of each active pixel 104. The specific stress conditions may be mapped for each pixel as a linear combination of characteristic correlation curves from several reference pixels 130. The combinations of two characteristic curves at predetermined stress conditions allow accurate compensation for all stress conditions occurring between such stress conditions. For example, the two reference characterization correlation curves for high and low stress conditions allow a close characterization correlation curve for an active pixel having a stress condition between the two reference curves to be determined. The first and second reference characterization correlation curves stored in the memory 118 are combined by the controller 112 using a weighted moving average algorithm. A stress condition at a certain time St (ti) for an active pixel may be represented by:
St(ti)=(St(ti−1)*kavg+L(ti))/(kavg+1)
In this equation, St(ti−1) is the stress condition at a previous time, kavg is a moving average constant. L(ti) is the measured luminance of the active pixel at the certain time, which may be determined by:
In this equation, Lpeak is the highest luminance permitted by the design of the display system 100. The variable, g(ti) is the grayscale at the time of measurement, gpeak is the highest grayscale value of use (e.g. 255) and is a gamma constant. A weighted moving average algorithm using the characterization correlation curves of the predetermined high and low stress conditions may determine the compensation factor, Kcomp, via the following equation:
K
comp
=K
high
f
high(ΔI)+Klowflow(ΔI)
In this equation, fhigh is the first function corresponding to the characterization correlation curve for a high predetermined stress condition and flow is the second function corresponding to the characterization correlation curve for a low predetermined stress condition. ΔI is the change in the current in the OLED for a fixed voltage input, which shows the change (electrical degradation) due to aging effects measured at a particular time. It is to be understood that the change in current may be replaced by a change in voltage, ΔV, for a fixed current. Khigh is the weighted variable assigned to the characterization correlation curve for the high stress condition and Klow is the weight assigned to the characterization correlation curve for the low stress condition. The weighted variables Khigh and Klow may be determined from the following equations:
K
high
=St(ti)/Lhigh
K
low=1−Khigh
Where Lhigh is the luminance that was associated with the high stress condition.
The change in voltage or current in the active pixel at any time during operation represents the electrical characteristic while the change in current as part of the function for the high or low stress condition represents the optical characteristic. In this example, the luminance at the high stress condition, the peak luminance, and the average compensation factor (function of difference between the two characterization correlation curves), Kavg, are stored in the memory 118 for determining the compensation factors for each of the active pixels. Additional variables are stored in the memory 118 including, but not limited to, the grayscale value for the maximum luminance permitted for the display system 100 (e.g., grayscale value of 255). Additionally, the average compensation factor, Kavg, may be empirically determined from the data obtained during the application of stress conditions to the reference pixels.
As such, the relationship between the optical degradation and the electrical aging of any pixel 104 in the display system 100 may be tuned to avoid errors associated with divergence in the characterization correlation curves due to different stress conditions. The number of characterization correlation curves stored may also be minimized to a number providing confidence that the averaging technique will be sufficiently accurate for required compensation levels.
The compensation factor, Kcomp can be used for compensation of the OLED optical efficiency aging for adjusting programming voltages for the active pixel. Another technique for determining the appropriate compensation factor for a stress condition on an active pixel may be termed dynamic moving averaging. The dynamic moving averaging technique involves changing the moving average coefficient, Kavg, during the lifetime of the display system 100 to compensate between the divergence in two characterization correlation curves at different predetermined stress conditions in order to prevent distortions in the display output. As the OLEDs of the active pixels age, the divergence between two characterization correlation curves at different stress conditions increases. Thus, Kavg may be increased during the lifetime of the display system 100 to avoid a sharp transition between the two curves for an active pixel having a stress condition falling between the two predetermined stress conditions. The measured change in current, may be used to adjust the Kavg value to improve the performance of the algorithm to determine the compensation factor.
Another technique to improve performance of the compensation process termed event-based moving averaging is to reset the system after each aging step. This technique further improves the extraction of the characterization correlation curves for the OLEDs of each of the active pixels 104. The display system 100 is reset after every aging step (or after a user turns on or off the display system 100). In this example, the compensation factor, Kcomp is determined by
K
comp
=K
comp_evt
+K
high(fhigh(ΔI)−fhigh(ΔIevt))+Klow(flow(ΔI)−flow(ΔIevt))
In this equation, Kcomp_evt is the compensation factor calculated at a previous time, and evt is the change in the OLED current during the previous time at a fixed voltage. As with the other compensation determination technique, the change in current may be replaced with the change in an OLED voltage change under a fixed current.
As explained above, an electrical characteristic of a first set of sample pixels is measured. For example, the electrical characteristic of each of the first set of sample pixels can be measured by a thin film transistor (TFT) connected to each pixel. Alternatively, for example, an optical characteristic (e.g., luminance) can be measured by a photo sensor provided to each of the first set of sample pixels. The amount of change required in the brightness of each pixel can be extracted from the shift in voltage of one or more of the pixels. This may be implemented by a series of calculations to determine the correlation between shifts in the voltage or current supplied to a pixel and/or the brightness of the light-emitting material in that pixel.
The above described methods of extracting characteristic correlation curves for compensating aging of the pixels in the array may be performed by a processing device such as the controller 112 in
In addition, two or more computing systems or devices may be substituted for any one of the controllers described herein. Accordingly, principles and advantages of distributed processing, such as redundancy, replication, and the like, also can be implemented, as desired, to increase the robustness and performance of controllers described herein.
The operation of the example characteristic correlation curves for compensating aging methods may be performed by machine readable instructions. In these examples, the machine readable instructions comprise an algorithm for execution by: (a) a processor, (b) a controller, and/or (c) one or more other suitable processing device(s). The algorithm may be embodied in software stored on tangible media such as, for example, a flash memory, a CD-ROM, a floppy disk, a hard drive, a digital video (versatile) disk (DVD), or other memory devices, but persons of ordinary skill in the art will readily appreciate that the entire algorithm and/or parts thereof could alternatively be executed by a device other than a processor and/or embodied in firmware or dedicated hardware in a well-known manner (e.g., it may be implemented by an application specific integrated circuit (ASIC), a programmable logic device (PLD), a field programmable logic device (FPLD), a field programmable gate array (FPGA), discrete logic, etc.). For example, any or all of the components of the characteristic correlation curves for compensating aging methods could be implemented by software, hardware, and/or firmware. Also, some or all of the machine readable instructions represented may be implemented manually.
Referring to
OLED efficiency degradation can be calculated based on an interdependency curve based on OLED electrical changes versus efficiency degradation, such as the interdependency curve in
First, one can create a stress history for each pixel (or group of pixels). The stress history can be simply a moving average of the stress conditions. To improve the calculation accuracy, a weighted stress history can be used. Here, the effect of each stress can have a different weight based on stress intensity or period, as in the example depicted in
In another case, one can assume the stress history is low frequency in space. In this case, there is no need to sample all the pixels. Here, a sub-set of pixels are used to calculate the stress history, and then an interpolation technique can be used to calculate the stress history for all the pixels.
In another case, one can combine both low sampling rates in time and space.
In some cases, including the memory and calculation block required for stress history may not be possible. Here, the rate of change in the OLED electrical parameter can be used to extract the stress conditions, as depicted in
As illustrated in
One can compensate for OLED efficiency degradation using interdependency curves relating OLED electrical change (current or voltage) and efficiency degradation, as depicted in
In
The following are some examples of procedures for finding a proper curve from a library:
Semiconductor devices (e.g., OLEDs) may age differently under different ambient conditions (e.g., temperature, illumination, etc.) in addition to stress conditions. Moreover, some rare stress conditions may push the devices into aging conditions that are different from normal conditions. For example, an extremely high stress condition may damage the device physically (e.g., affecting contacts or other layers). In this case, identifying a compensation curve may require additional information, which can be obtained from the other devices in the pixel (e.g., transistors or sensors), from rates of change in the device characteristics (e.g., threshold voltage shift or mobility change), or by using the change in a multiple-device parameter to identify the stress conditions. In the case of using other devices, the rate of change in the other device parameters and/or the rate (or the absolute value) of change in the other-device parameter compared with the rate (or the absolute value) of change in the device parameter can be used to identify the aging condition. For example, at higher temperature, the TFT and the OLED become faster and so the rate of change can be an indicator of the temperature variation at which a TFT or an OLED is aged.
In
In
In another embodiment, one can look at the rates of change in different parameters in one device to identify the stress condition. For example, in the case of an OLED, the shift in voltage (or current) at different current levels (or voltage levels) can identify the stress conditions.
In the procedure illustrated in
The stress pattern can include duration and stress level. In one embodiment of the invention, the pixel parameters are monitored in-line during the stress to assure the parameters of the pixels do not pass the specified range. In another embodiment of the invention, the parameters of selected pixels or some reference pixels are monitored in-line during stress. In another embodiment of the invention, the pixels are stressed for a period of time and then the pixel parameters are extracted. After that the pixel parameters are updated and the stress pattern and timing can be updated with new data including new pixel parameters and the rate of change. For example, if the rate of change is fast, the stress intervals can be smaller to avoid passing the specified ranges for pixel parameters.
The setting for the parameters of the pixels can be variation between the parameters across the panel. In another embodiment it can be specific value.
In one example, the pixel information (or parameter) can be the threshold voltage of the drive TFT. Here, the stress condition of each pixel is defined based on its threshold voltage. In another example, the pixel parameter can be the voltage of the emissive devices (or the brightness uniformity).
The pixel information can be extracted through different means. One method can be through a power supply. In another case, the pixel parameters can be extracted through a monitor line.
In
Based on the extracted pixel parameters and the stress history, stress patterns are generated at step 1803. Then the pixels are stressed at step 1804, in accordance with the generated stress pattern. The parameters of the stressed pixels are monitored and updated at step 1805 by extracting the pixel parameter from the stressed pixels. Step 1806 determines whether the pixel parameters extracted from the stressed pixels is within a preselected range, and if the answer is negative, step 1807 updates the stress history of the pixels, and then steps 1803-1806 are repeated. This process continues until step 1806 produces a positive answer, which means that the pixel parameters extracted from the stressed pixels are within the preselected range, and thus the pixels are returned to normal operation.
In one example, the pixels are assigned to different categories based on the stress history, and then the pixels are stressed with all the other categories that they are not assigned to. At the same time, the pixel parameters are monitored similar to the previous case to assure they do not pass the specified ranges.
In another example, the stress history has no timing information, and the change in pixel parameters can be used to identify the stress level and timing. For example, in one case, shift in the electrical characteristics of the emissive device can be used to extract the stress condition of each pixel for the stress pattern.
In yet another embodiment, the interdependency curves between pixel parameters and its optical performance can be used to extract the stress condition for each pixel. In the case of electrical characteristics of the emissive device, the interdependency curves can be used to find the worst case of efficiency degradation. Then, the delta efficiency between each pixel and the worst case can be determined. After that, the corresponding change in electrical characteristics of the emissive device of each pixel can be calculated to minimize the difference in efficiency between the pixel and the worst case. Then the pixels are stressed, and their pixel parameters (e.g., electrical characteristics of the emissive device) are monitored to reach the calculated shift. Similar operations can be used for other pixel parameters as well.
Efficiency degradation of electro-luminance devices can affect the performance of devices such as displays. This degradation is due to stress and other conditions such as temperature. Interdependency curves are the relation between an OLED's characteristics and its luminance degradation, therefore, interdependency curves are what connect the measurement data (electrical characteristics) to the characteristic (luminance degradation) that needs to be compensated for. For example, in the case of an emissive device, the electrical characteristics of the device can be measured easily. In one example, the OLED characteristic can be OLED voltage shift for a given current as a result of stress. However, the final characteristic that is required to be compensated for are its optical characteristics. In this case, the change in electrical characteristics due to aging (or other conditions) is measured and based on the interdependency curve one can determine how much the optical performance of the device is affected.
A correction algorithm fixes the drive circuit issues by extracting parameters related to the driver circuit and also fixes the optoelectronic device issues such as burn-in by extracting parameters from the device (or other related parameters) and with use of the interdependency curves. Interdependency curves thus show the relation between the extracted parameters (or stress history) for the optoelectronic device and its optical performance degradation.
One method of calculation of the correction factor involves extracting the relationship of the optical degradation and the given value of extracted parameter(s) as a function of stress level. The stress history of a pixel (or a group of pixels) is calculated, and based on the stress level, one or more interdependency curves are selected from different interdependency curves representing different stress levels. From the selected curves and the extracted parameters a correction factor is calculated as a function of the stress level. One simple function can be a linear approximation.
Using interdependency curves to solve the aging issues in optoelectronic devices can eliminate the need for optical sensors. However, some devices may experience different aging behavior as a function of temperature.
Referring now to
In some optoelectronic devices, the temperature may affect the interdependency curves or as described below, an effective stress. As a result, the system needs to accommodate for the temperature effect as well as the stress levels as described hereinabove. Both the stress levels and the temperature are operating conditions which affect the interdependency curve. To accommodate for the temperature effect as well, the temperature profile of the panel is either measured or estimated and taken into account in the compensation of the display.
In one embodiment depicted in
Next, from the selected set of the interdependency curves (or the calculated new interdependency curves or the points of interest) and stress information 1904 (and with use of the OLED characteristic parameter(s) 1902 if not used already to restrict calculation to points of interest) one or more pixel correction factors 1905 are calculated 1920. The one or more correction factors 1905 are used in the correction algorithm 1930 to fix for optical degradation of the optoelectronic device as described hereinabove, so that for example a video signal 1906 is displayed on the display 1940 accurately.
It is to be understood, that since the interdependency curves are stored for various stress conditions and various temperatures, the order of selection and/or calculation based on temperature and stress history 19101920 may be changed. For example, as an alternative to the above, given a set of interdependency curves for N stress conditions, and for each stress condition M temperatures, when analyzing stress conditions first, for every temperature within a threshold, interpolation curves of the closest higher and lower stress conditions are utilized to interpolate a curves corresponding the stress condition of the pixel for each close temperature condition. To reduce calculation and storage requirements the OLED characteristic of interest (the measure of OLED voltage shift for example) may be used to extract or generate only the points of interest on the new interdependency curves. Furthermore, a single selection and/or calculation taking into account both temperature and stress history may be utilized to generate appropriate at least one correction factors 1905. In such an algorithm, for example, the interdependency curves for various temperature and stress conditions could be interpolated in terms of both the temperature and stress information of the pixel to extract the correction factor corresponding to the OLED characteristic parameter 1902.
In the case of calculating a new interdependency curve for a given temperature based on a few of the stored interdependency curves 1901, the optoelectronic device characteristic parameters may be used to calculate required output for just those parameters to reduce the calculation load, i.e. generating only points of interest rather than generating entire interdependency curves. In some embodiments utilizing functional curve fitting, in calculating interdependency curves 19101920 the between value for each corresponding curve in the sets is extracted for the parameters and then a function is generated for the extracted values and temperature. Here, the value for the given temperature then is calculated based on that function. This is repeated for all the curves in the set.
In another embodiment depicted in
Here, since effective stress takes into account both temperature and standard stress conditions, one can change the order of incorporation of temperature and stress history into the calculations or mix them in one selection function.
For calculating an effective stress condition based on temperature, one can either use models or lookup tables. In some embodiments, the same model or lookup tables utilized to calculate the effective stress 2010 are used to generate and/or index the interdependency curves 2001.
One can mix the two methods described here to improve the correction factor calculation. In addition, if the temperature difference between a pixel (or a group of pixels) temperature and a reference temperature is larger than a threshold, calculation of the correction factor can be performed more often to reduce the effect of higher order conditions. For example, if there is a large temperature change for a short time, its effect might otherwise be ignored if the periodic update time for the OLED correction factor is too long.
In another case, illustrated by
After the degradation factor 2120 (or degradation factor as calculated from the correction factor) is updated with use of curves in calculations similar to as outlined above, either the degradation-time curve 2112, 2114, 2116 or the electrical-optical curves (not shown) corresponding to different stress conditions, the start-point of the curves can be reset for the next update. One method is finding the related x-index (e.g. stress-time) of the curve for the degradation value for each curve and using that as the new start point for those curves. For example in
For embodiments which utilize degradation-time curves, the stress time can represent an actual time in which case a temporary stress history tracking actual stress on the pixel for a short time may be recorded. In other embodiments an effective stress time may be tracked which combines the actual stress level and time between each update for example as described hereinbelow.
Another method is to calculate the effective x-index from the stress (or temperature) level for each curve. This can be empirical or modeled for each curve, or it can be measured from different reference devices being stressed at different levels.
The new effective x-index can be used as the new start point for each curve.
The x-index could be time as shown in
In one aspect, the stress history and temperature history of pixels (or group of pixels) are stored. During a status update period of the optoelectronic device, one or more interdependency curves are chosen based on temperature. Then from the stress history and selected interdependency curves a correction factor is calculated. Here, an electrical measurement from the optoelectronic device or a representative device can be used to fetch proper points from the interdependency curves.
In another aspect, the temperature is used in adjusting the stress history generating an effective stress. Here, based on the temperature and the luminance value (it can be also current, voltage or ON time) of the pixel, the effective stress is calculated. For example, if the pixel is program to offer L1, at higher temperature the “effective stress” of L1 can be similar to a “higher” stress case according to a standard of stress which does not take temperature into account.
In another aspect, if the temperature of a pixel (or a group of pixels) is significantly different from a reference temperature, the stress history calculation for said pixel (or the group of pixel) gets updated more often. In addition, the calculation for the correction factor based on the interdependency curves can also be performed more often.
In another aspect, the interdependency curves are the relation between stress time and luminance degradation of the OLED.
In another aspect, the interdependency curves are the relationship between OLED electrical characteristic and the luminance degradation of the OLED.
In another aspect, the stress history is reset to a default value after the correction factor is updated. Here, some other parameter is stored (in addition to retaining the degradation value or correction factor), to track the new origin point in the interdependency curves. For example, correction factor, time or extracted OLED parameter can be used, with the previous degradation or correction factor.
In some applications, the device performance may vary due to process variations. This can also affect the interdependency curve that a device will actually exhibit and hence affect the accuracy of calculations relying on interdependency curves which do not correspond to the device in question. It follows that the interdependency curves are a function of the initial status of the device. For example, in the case of printed OLEDs, the initial device characteristics of the OLED at different pixels or in different displays can vary due to process variation. This can also affect the aging behavior of the OLED and so influences the interdependency curve, i.e. the change in OLED electrical characteristics versus OLED efficiency degradation, exhibited by each pixel.
In the embodiment depicted in
As described above the interdependency curves include curves for various stress conditions and various initial device characteristics. With reference also to
Referring now to
In another aspect, test units go under different test conditions to extract interdependency curves directly or indirectly. In the case of indirect measurement, some parameters are extracted from the test units pointing to interdependency curves from the library. In one embodiment, test units from the same or similar batch are utilized to produce initial curves which are then utilized to select more complete curves (subjected to longer testing time) from the library.
The interdependency curves then can be updated at different stages: at product manufacturing or at a consumer site. In addition, the new data extracted may be used to update the interdependency curve library. In some embodiments updates are performed remotely, i.e. even when the device is remote from the origin of the interdependency curve library or the aging of the test devices and the preparation of the interdependency curves.
Referring specifically to the steps of the method 2400, once the device fabrication is complete 2410, test devices on a substrate are aged 2420 continually, interdependency curves are prepared. The device is shipped to the product manufacturer, for example a display with an array of OLEDs 2430. In one case aging 2420 is performed on test devices of the device itself also, in which case the prepared interdependency curves measured from that display are shipped with the device 2430. At the point in time of shipping the prepared interdependency curves may be provided to the manufacturer. In either case, the aging of the test devices continues 2420 and further interdependency curves are prepared 2442 so that by the time there is integration of the devices into the products 2440 there is another opportunity to update the shipped device with calculated interdependency curves. The aging of the test devices continues 2420 and yet further interdependency curves are prepared 2452 so that by the time the device in the product is at the consumer site 2450 there is another opportunity to update the shipped device with calculated interdependency curves. In some embodiments updates are provided over the internet. In some embodiments, preparing the interdependency curves 2432, 2442, 2452 and updating those of the shipped device at various points in time utilizes data from testing devices 2420 from the same or similar batch of devices as those that went into the product.
Optionally the process can include updating a central library with interdependency curves 2460 stored in an interdependency curve library 2480, which can collect data from multiple devices and batches of devices and serve as a comprehensive repository for similar devices and which can be used to update the interdependency curves of the shipped device at various points in time from fabrication to operation at a consumer site. In some embodiments, interdependency curves of the library 2480, each of which may for example contain data representing a many hours of stress testing, are only chosen to augment those of the shipped device when they are close a enough match to those curves already associated with the shipped device, such as for example initial interdependency curves which contain data representing fewer hours of stress testing. Although
Modelling can be one approach to fix the burn-in effects caused by pixel stress. However, keeping long stress histories for every pixel and also other parameters requires significant memory. Another issue is that proper modelling is very complicated due to the multi-input system with long input dynamic range. Moreover, process variations cause divergence in the real performance of the device from that predicted by the model.
The following embodiments illustrated in
First, a total effective stress time is initialized 2510. Here, the total effective stress time for each pixel or group of pixels are set to a known value (for example zero). Alternatively, after calculating the degradation value during a previous update, the remaining or residual value which otherwise would have been rounded off and lost due to the data resolution in degradation factor is used to calculate the initial value for the effective stress time.
After the total effective stress time is initialized, video brightness data is sampled 2520. In one case, after a fixed time the pixel value is sampled. The sampling time should be less than the frequency of change in the pixel data. In another case, if there is a significant change in the pixel value, the previous value and its time on the panel is used as the sampled video brightness data and the new value is used for calculating the new stress time. One can also use a combination of both.
In another case, temperature is sampled in addition to sampling the video data and time. In this case, temperature change can also be used as a trigger value for sampling the video data. For example, once the temperature change exceeds a threshold new video data is sampled.
Once the video brightness data has been sampled 2520, the effective stress time for at least one given reference stress level is calculated. Here, if one or two reference stress conditions are used, then the stress time of the pixel under sampled stress is translated to said reference conditions. For this translation, also one can use temperature as one of the translation factors. For example, the sampled video data, stress time, and temperature of the pixel are used to calculate the effective stress time for a given reference stress value, at a given temperature level 2530.
In one case, several degradation curves based on different stress and different temperature are stored. For a sampled temperature level, corresponding curves are selected. From the selected curves the conversion factor of the stress time for the sampled stress to the effective stress time of a given reference stress level is calculated. If there is no direct curve for the sampled temperature, the curves are extracted from the existing curves first. The calculation can be performed in reverse order. In this case, the curves for given sampled stress are extracted first and then the conversion factor for the temperature is calculated. Once the effective stress time for the pixel has been calculated the total effective stress for the pixel is updated 2540. The total effective stress replaces the stress history normally utilized in the process of determining from the interdependency curves the degradation factor as described hereinabove. The effective stress time therefor acts to effectively calculate the change in the total effective stress of a pixel from the various conditions contributing to effective stress since the last degradation factor update. In some embodiments, degradation-time curves are stored and utilized in the calculations. In other embodiments, a single degradation-time curve, having the single reference conditions is stored.
To simplify the calculation, one can linearize the curves around the degradation factor to calculate the change in the degradation factor for a given video data and stress time.
After some conditions are satisfied 2550 the degradation factor is updated 2560 otherwise another sample is taken 2520. These conditions can be a threshold for total effective stress time or the change in degradation factor. Here, the threshold value can be dynamic. For example, when the degradation factor changes faster, the threshold predetermined time value can be smaller to accommodate the faster degradation. The threshold parameters' value for this decision can be different for each pixel. In some embodiments, the threshold is set to ensure that only once the total effective stress time has accumulated by an amount having a magnitude of sufficient significance, is the degradation factor updated. As mentioned above any residual which would be rounded off can be used as the value to initialize the total effective stress time during the next update.
In updating the degradation factor 2560, from the effective stress time and the previous degradation factor, the change in degradation is calculated. After updating the change in degradation, the degradation factor itself is updated. In one case, after the degradation factor is calculated, the error due to quantization and other factors is calculated to be used as part of the calculation of the new initial value for the total effective stress time.
First, the total change in degradation factor is initialized 2610. Here, the change in the degradation factor for each pixel or group of pixels are set to a known value (for example zero). Alternatively, after calculating the degradation value of a previous update, the remaining or residual value due to the resolution in the degradation factor which otherwise would have been rounded off during the last update is used to initialize the total change in degradation factor.
After the change in degradation factor is initialized, video brightness is sampled 2620. In one case, after a fixed time the pixel value is sampled. The sampling time should be less than the frequency of change in the pixel data. In another case, if there is a significant change in the pixel value, the previous value and its time on the panel is used as the sampled video brightness data and the new value is used. One can also use a combination of both. In another case, temperature is sampled in addition to sampling the video data and time. In this case, temperature change can also be used as a trigger value for sampling the video data. For example, once the temperature change exceeds a threshold new video data is sampled.
Once the video brightness data has been sampled 2620, a resulting change in degradation factor is calculated 2630. For example, the sampled video data, stress time, degradation factor, and temperature are used to calculate the change in the degradation factor.
In one case, several degradation curves based on different stress and different temperature are stored. For a sampled temperature level, corresponding curves are selected. From the selected curves, the change in degradation factor can be calculated based on the degradation factor, the sampled stress, and stress time. If there is no direct curve for the sampled temperature, the curves are extracted from the existing curves first. The calculation can be performed in reverse order. In this case, the curves for given sampled stress are extracted first and then the change in the degradation factor for the temperature is calculated. In a similar manner to embodiments described hereinabove, histories of the pixel are discarded by adopting new starting points for the degradation-time or interdependency curves. As such a degradation factor is stored for each pixel i.e. OLED, and updated.
To simplify the calculation, one can linearize the curves around the degradation factor to calculate the change in the degradation factor for a given video data and stress time.
After some conditions are satisfied 2650 the degradation factor is updated 2560 otherwise another sample is taken 2620. These conditions can be a threshold for the change in degradation factor. Here, the threshold value can be dynamic. For example, when the degradation factor changes faster, the degradation threshold value can be smaller to accommodate the faster degradation. The threshold parameters' value for this decision can be different for each pixel.
In updating the degradation factor 2660, the change in degradation factor is added to the degradation factor. In one case, after the new degradation factor is calculated, the error due to quantization and other factors is calculated to be used as the initial value for change in the degradation factor. In some embodiments, the threshold is set to ensure that only once the total change in device degradation has accumulated by an amount having a magnitude of sufficient significance, is the degradation factor updated. As mentioned above any residual which would be rounded off can be used as the value to initialize the total change in device degradation during the next update.
Compensation for OLED efficiency degradation based on electrical characteristics of the OLED devices is prone to error due to different aging conditions. One solution is to keep history of the aging, for example stress and temperature histories, of each pixel (or a group of the pixel). This may require significant memory size. To address that, event driven stress history was developed which reduces the memory size significantly. Further, to reduce the system complexity and eliminate the need for memory, the new embodiment uses the rate of change in the OLED characteristic as an indicator for correcting the aging of the OLED.
OLED correction=f(VOLED or IOLED,dVOLED/dt or dIOLED/dt)
here, different interdependency curves can be used for correcting the OLED efficiency degradation. To select the curve, one can use the rate of change. The higher the aging rate at a certain aging point can be an indicator of the stress status.
Although the above shows the function specifically with respect to voltage or current and the change in voltage or current other parameters of an interdependency curve may be used.
Compensation According to Stress History Modelling
As described by the foregoing, stress history can be used to compensate for OLED panel degradation by measuring and tracking the aging effect on each of the red, green and blue pixels and creating a correction model to eliminate or reduce the effects of aging by compensation of the pixel though adjusted, usually increased, drive levels.
In some embodiments stress history may be used to compensate for OLED panel degradation in the form of a pixel drive level correction model whose parameters are a function of the stress history. Stress history compensation is done by measuring the aging effect of each of red, green, and blue pixels in any of the aforementioned devices or embodiments, and creating a correction model which utilizes that stress history to eliminate the aging by compensation of the pixel though increased drive levels. In methods for compensating for stress history the individual points of aging are determined and then compensation is modeled using:
O=B*X+C (1)
Where O is the output (corrected) drive level for the pixel, X is the input drive level for the pixel, B is the gain correction, and C is the offset correction.
It should be noted, that although the correction model presented in equation (1) is based on a 1st order linear equation, in some embodiments the model can utilize higher order polynomial equations along with the requisite higher order coefficients (higher order “gains” e.g. BO1, BO2, . . . ) as well as the appropriate offset correction C. Moreover, in some embodiments a linear piecewise model is utilized, for which there are multiple segments each with independent slopes (gains for each segment e.g. Bi, Bii, . . . ) as well as the multiple appropriate pixel offsets (Ci, Cii, . . . ). It should be noted that in some embodiments the correction model has independent B and C parameters for each different color of pixel (e.g. red, green, blue) as different types of pixels generally age differently.
If there were no stress history to inform the correction model of equation (1), B and C would be constants which would need to be determined for each pixel from time-to-time, based on periodic measurements (whether optical or electrical) as each pixel ages. The drawbacks of that kind of solution include the high bit resolution required for B & C and the sheer number of values to be stored which translate into larger memory and higher power for completing the compensation. In addition, as pixels age more and more B & C parameters for various segments may be required, for example in a piecewise linear model, to complete the compensation at all gray levels putting further increases in both power and memory. The actual aging factors for pixels can change very slowly and as the materials used in the fabrication of the panel become more stable there can be very small changes in aging over steps of 10 or more hours. This translates into potential execution of wasteful and unnecessary periodic pixel aging measurement cycles which use resources (e.g. battery power) at a much higher frequency than warranted.
Modelling compensation with B and C being functions of stress history mitigates one or more of the aforementioned problems. In order to minimize the number of bits stored, the calculation of the correction coefficients B & C (or more for polynomial solutions) are instead moved into compensation flow. This solution uses a quantized stress where 8-bits are used to represent total accumulated stress levels up to the end of life for the panel, representing typically 20% luminance reductions for any color. Tracking for each pixel (either R, G, or B etc.) only one 8-bit total accumulated stress value, and utilizing model parameters defined for all pixels of each color, helps to minimize the storage requirements and power consumption of the compensation. This is most applicable to compensation using GPU processor of smart phones but is equally applicable to other solutions in other contexts.
In embodiments for which compensation is performed according to the compensation model corresponding to equation (1):
and
C=P
1
−BP
m1 (3)
where B is the pixel gain, C is the offset, and Pn is the corrected value at a gray scale, at some correction point represented by n, which in this case is a correction from the stress history model for that gray scale or drive level point, and Pm1 is the uncorrected gray scale of the correction point. In some embodiments, B and C are respectively the slope and intercept of a line representing a linear model mapping the input drive level X, to the desired i.e. new corrected output drive level O. As such, P2 and Pm2 correspond to any output-input drive level point (X2, O2) along the line of the linear model. Similarly, P1 and Pm1 correspond to any other output-input drive level point (X1, O1) along the line of the linear model. In some embodiments, the model is piecewise linear in which case Bs and Cs are defined for each segment, and may be expressed in terms of points along the respective segments.
To base a compensation model relating the input and output drive levels on a stress history model, the parameters of the compensation model, namely, gain B and offset C in the present example of equation (1), are defined as functions of the total accumulated stress (5) on the pixel. Those functions are determined and defined by data collected on the aging of pixels according to a stress history model. As will be described further below, the functional form of B and C are chosen to help minimize the number of variables used to compensate for all of the pixels of the display. Since, the parameters B and C of the compensation model are each functions of the stress history model, they are referred to herein as being part of the stress history model.
Part of generating the stress history model involves the testing of pixels of each color of similar displays. The data collected in that testing is used to determine the parameters for the stress history model, which defines a quantitative measure of stress, and applies to pixels of the same color, which are present in similar displays, e.g. same kinds families of displays. As such, although utilizing the same type of model, each color of each type of display panel will have its own set of parameters for that stress history model.
During operation of a panel, a stress history of each pixel is kept according to the stress history model, as the panel ages. The stress history model tallies samples of the actual grayscale drive level applied to a pixel raised to an aging power factor according to the pixel's color (e.g. a power factor for each of R, G, B), and uses that as the accumulated or total stress for the pixel. This accumulation can be achieved for example, by sampling every second, processing, and then saving to non-volatile storage in one-hour increments, and in some embodiment the accumulation is normalized to a desired scale. In some embodiments, each hour of stress is normalized to a number between 0 and 1000 where 1000 is corresponds to aging from stress caused by driving the pixel at max gray scale, e.g. 255. The aging equation for each color is as follows:
In equation (4), Sh is the accumulated stress for the pixel over a particular hour h. The aging power factor gamma γ is different for each color found empirically and used to linearize the model equations as described further below. In an example corresponding to the sample graphs illustrated in
The total accumulated stress is:
In equation (5), y is the total number of hours of data collected. The stress logging file contains a 32-bit value for each red, green, and blue pixel on the display panel. In some embodiments, the total accumulated stress for each pixel is replaced every hour, i.e. the hourly stress Sh is added to the previous total accumulated stress S.
In order to determine the power factors of the stress history model for each color, a population of pixels of the same color (e.g. R, G, B) and from one or more displays of the same family or type of display are driven and measured during an initial testing phase carried out by a panel testing system to determine the stress model parameters.
Referring to
The panel testing system 2700 includes test display system 2750 which is to be tested, optical sensing 2730, a controller 2702 for overall control of the process, which in the embodiment of
The optical sensing 2730 may include a digital camera based on a digital photography system with lenses, and may be a monochromatic digital camera or a standard digital camera, such as a monochromatic or RGB, CCD, CMOS or other sensor array based camera, or any other suitable optical measurement technology capable of taking optical images of the luminance of the pixels of the test display panel 2720. In some embodiments, optical sensing 2730 includes an optical scanning system which works on similar principles to a flatbed scanner, but without relying on reflection from a light source. In other embodiments, optical sensing 2730 includes photo sensors 132 distributed throughout the test display panel 2720 itself, and hence forms part of the test display system 2750.
Luminance measurement image data refers to any matrix containing optical luminance data corresponding to the output of the test display panel 2720, and may comprise multiple channels such as red (R), green (G), blue (B) etc. and in some cases may be monochromatic as in the case where the optical sensing 2730 is monochromatic. Advantages of using monochromatic optical sensing versus RGB optical sensing may include faster exposure times, avoidance of display and sensor R,G,B frequency mismatch, aliasing, and/or crosstalk, avoidance of mismatching numbers or arrangements of the R,G,B sub-pixels of the display and the R,G,B elements of an optical sensor array, and ease of handling yellow or white subpixels of display panel 2720. In some embodiments utilizing either a monochromatic or an RGB optical sensing, measurements occur while the display only displays a single channel, primary color, or subpixel color (R, G, B, Y, or W etc.) at any one time.
The pixels are driven with predetermined stress inducing drive signals which progressively ages the pixels. Periodically, this is interrupted and the pixels are driven with one or more predetermined testing drive levels. The optical sensing 2730 is arranged to measure the luminance of the pixels 104 of the display panel 2720 while pixels are driven with the predetermined testing drive levels. The optical sensing 2730 may be operated manually or automatically controlled by one or both of the controller 2702 and optical correction processing 2740. The optical sensing 2730 generates a luminance measurement image representative of the optical output of the pixels of the test display panel 2720, and the optical correction processing 2740 receives the luminance measurement image data from the optical sensing 2730. Optical correction processing 2740 then processes the measurement image data while controlling the test display panel 2720 in the process of generating the test data for generating the stress history data, and may store it in memory storage 2706. The data generated from testing test display system 2750 may be stored and processed in the optical correction processing 2740 or an external tool of which it may be a part. In some embodiments multiple test display systems 2750 are utilized to generate the parameters for the stress history model.
With reference also to
Pixel populations of one or more test display panels are driven and tested during an initial testing phase to determine the stress model parameters 2810 using pixels of the same color in panels of the same kind or display family. Sometime prior to use, the stress model parameters appropriate for a color and kind of display are stored 2820 in the display, for example in a memory of that display.
In determining the stress history model parameters 2810, the population of pixels (all of the same color) are stressed to induce aging. In some embodiments, the pixels are driven continuously at a particular stress drive level, for example at 300, 600, or 800 out of 1023 (in the case of a 10-bit drive level). In some embodiments, the pixels are driven at various different stress drive levels and tracked, however this is more memory intensive.
By driving the pixel continuously at the same stress drive level, multiple samples and or time periods are effectively combined into one value since x1γ+x2γ+ . . . +xnγ simplifies to n*xγ which can simply be stored as n, it being explicitly known that it represents n samples of xγ. This also enables convenience in fitting various power factors. Using the same identity, in embodiments where the stress drive levels vary, the number of samples of all at the same stress drive level can be similarly reduced, amounting to a total of 255 different counts for 8-bit, and 1023 different counts for 10-bit, each count simply representing a total (up to that point) number of samples of xγ for each possible x. In any case, a history of the actual drive levels being used to drive the test panel are tracked and possibly reduced or compressed. In some embodiments, the counts of each stress drive level for each pixel is tracked and indexed by time. In some embodiments, the counts of each stress drive level for each pixel is tracked and updated with no indexing by time, but is associated with drive correction values as they are determined as discussed below.
After data tracking drive levels x over the entire lifetime of a test panel, the system generates candidate stress histories with equations (2) and (3) using a candidate power factor γ with the actual tracked drive levels. As mentioned above, periodically, the test panel is driven at one or more predetermined test drive levels, to measure aging in the form of the necessary drive correction values which need to be applied to the drive level which corrects the luminance. In some embodiments, correction values for multiple drive levels are recorded. Generally speaking, correction levels are measured by varying the input drive level until the pixel displays the desired luminance.
In some embodiments, the drive correction values, when determined, are indexed by time, which indexes may be used to determine the counts of the stress drive levels that had been applied by that time, enabling easy generation of a stress drive level history value using a candidate power factor γ and its association with that drive correction value. In other embodiments, the correction values are not indexed by time, however correction values are stored or otherwise associated with the counts of the stress drive levels at the time the correction values are determined.
With respect to periodically determining the drive correction value, in some embodiments, while the optical sensing 2730 takes luminance measurements of the test display panel 2720, at least one of the optical correction processing 2740 and the controller 2702 varies the driving of the pixel circuit 104 until a desired output from the pixel is achieved. For example, in an embodiment, a predetermined initial drive level is used to drive all of the pixels, and a luminance measurement image is compared to an expected luminance corresponding to that drive level. In a loop of one or more iterations, a pixel circuit whose luminance measurement is darker than expected is driven with greater drive levels, while pixel circuits whose luminance measurement is brighter than expected is driven at lower drive levels, while the system 2700 captures and generates further luminance measurement images. Once the luminance measurement of each pixel matches the expected level, the final drive levels used to drive each pixel circuit 104 of the test display panel are stored in memory 2706. A difference between this final drive level and the predetermined initial drive level constitutes the drive correction value for the pixel at that time. In some embodiments, this drive correction value is stored indexed by time in others it is stored along with or associated with the current accumulation counts of drive levels x, which will be converted into stress according to equations (4) and (5).
Referring also now to
Referring also now to
Referring also now to
In some embodiments, once the maximum life of the display panel is determined, this life value, in hours, is converted into 255 steps via a scaling factor used on the stress history equation. This allows an 8-bit value to represent the accumulated stress history of each pixel, which can be used as an eight-bit input based on the panel stress to generate the B and C parameters as described hereinbelow. In one example embodiment, a factor of 4 is utilized to convert an 8-bit stored stress value for use in a 10-bit color context.
Referring once again to the compensation model represented in equations (1), (2), and (3), and with a linearized stress history model as defined in equations (4) and (5), the functional form for B and C in terms of the total accumulated stress history of a pixel will now be discussed.
In the linearized stress history model, the correction drive values are linear with respect to the stress history, as shown in
Cor=kS+c (6)
where Cor is the correction drive value, S is the total accumulated stress, k is the slope and c is the intercept. Each of k and c may be determined empirically for any particular drive level point. It turns out that for each drive level point, the same power factor linearizes the correction drive values. It can be shown that the correction model parameters B and C of equation (1) may be expressed as linear functions of S, as follows:
B=ΔS+δ (7)
C=ΛS+λ (8)
where S is the total accumulated stress, Δ and Λ are slopes respectively of the linear functions B and C of S and δ and λ are the respective intercepts.
In order to solve for the four unknowns, multiple correction values for the same stress S but for two different drive level points (P1, Pm1) and (P2, Pm2) are empirically determined. In embodiments in which multiple correction levels are stored for multiple drive levels points (one or more predetermined test drive levels) while creating the stress history model, the correction for drive level points for each stress level are already available. In such a case one may use two different linear relationships between correction levels and stress history:
Cor1=k1S+c1 (9)
Cor2=k2S+c2 (10)
In equations (9) and (10), the subscripts are representative of drive level points (P1, Pm1) and (P2, Pm2) corresponding to each linear relationship between correction and stress. It should be clear from the above that the correction levels by definition are the difference between the input and the adjusted drive level: Corn=Pn−Pmn. From equation (2) and the definition of the correction levels:
Letting Pdiff=Pm2−Pm1, and using equations (9) and (10), B reduces to:
This is of the form of equation (7), and Δ and δ are each constants which are empirically measurable, and moreover, in some embodiments are directly derivable from part of the data gathered during generation of the stress history model.
From equation (3) and the definition of the correction levels:
C=P
m1+Cor1−BPm1 (13)
Using equations (12) and (9), C reduces to:
This is of the form of equation (8), and A and A are each constants which are empirically measurable, and moreover, in some embodiments are directly derivable from part of the data gathered during generation of the stress history model.
The stress model parameters which are stored 2810 for each color pixel of any given type of display therefore include γ, Δ, δ, Λ, and λ. During operation a stress history is kept according to (4) and (5) 2830. During regular use of the display panel, it is common for the drive levels x to vary over most, if not all of the possible values. While the stress history is maintained, it is also used to generate stress history corrected drive levels for each pixel according to equations (1), (12) and (14) using the total accumulated stress stored for the pixel at that point in time. The stress history corrected drive levels are used to actually drive the display to display images 2850 which are corrected for the effects of stress.
While particular embodiments, aspects, and applications of the present invention have been illustrated and described, it is to be understood that the invention is not limited to the precise construction and compositions disclosed herein and that various modifications, changes, and variations may be apparent from the foregoing descriptions without departing from the spirit and scope of the invention as defined in the appended claims.
This application This application claims the benefit of U.S. Provisional Patent Application No. 63/273,773, filed Oct. 29, 2021, which is hereby incorporated by reference in its entirety.
| Number | Date | Country | |
|---|---|---|---|
| 63273773 | Oct 2021 | US |
