Method of driving bistable electro-optic displayThis application is a divisional application filed on application No. 200910163444.6 entitled "method of driving a bistable electro-optic display", filed on 20/11/2002.
The present invention relates to a method of driving a bistable electro-optic display and to an apparatus for use with such a method. More particularly, the present invention relates to a driving method and device controller for more accurately controlling the gray state of pixels of an electro-optic display. The invention further relates to a method of maintaining a long-term Direct Current (DC) balance of driving pulses provided to an electrophoretic display. The invention is particularly, but not exclusively, intended for use with particle-based electrophoretic displays in which one or more types of charged particles are suspended in a liquid and moved in the liquid under the influence of an electric field to change the display of the display.
In one aspect, the invention relates to a device capable of driving an electro-optic medium that is sensitive to the polarity of a supplied electric field using circuitry for driving a liquid crystal display, wherein the liquid crystal material is insensitive to polarity.
The term "electro-optic" as used herein, as applied to materials or displays, is used in its conventional sense in imaging technology to refer to a material having first and second display states which differ in at least one optical characteristic, the material changing from its first display state to its second display state by application of an electric field to the material. Although this optical characteristic is typically a recognizable color to the human eye, it may also be other optical characteristics such as light transmission, reflectance, brightness or, in the case of machine-readable displays, pseudo-color in the sense of variations in reflectance of electromagnetic wavelengths outside the visible range.
The term "grey state" as used herein in the conventional sense of imaging technology relates to a state intermediate the two extreme (extreme) optical states of a pixel and does not necessarily imply a black-to-white transition between the extreme states. For example, several of the patents and published applications referred to below describe electrophoretic displays whose extreme states are white and deep blue, so that the intermediate "grey state" is effectively pale blue. In fact, as already indicated, the transition between two extreme states may not be a color transformation at all.
The terms "bistable" and "bistability" as used herein in their conventional meaning in the art relate to a display comprising display elements having different first and second display states at least in one optical characteristic such that after driving of any given element to assume its first or second display state is completed by means of an addressing pulse of limited time, the state will last at least several times (temporal times) after the end of the addressing pulse, for example at least four times the minimum time of the addressing pulse required to change the state of the display element. As shown in co-pending application No. 10/063236 filed on 2.4.2002 (see also corresponding international application publication No. wo02/079869), some particle-based, greyscale electrophoretic displays are stable not only in their extreme black and white states, but also in their intermediate greyscale states, which is equally suitable for some other types of electro-optic displays. This type of display is better suited to be referred to as "multi-stable" relative to bi-stable, although in general the term "bi-stable" may be used herein to cover both bi-stable and multi-stable displays.
The term "gamma voltage" as used herein refers to an external voltage reference that a driver uses to determine the voltage supplied to a pixel of the display. It will be appreciated that bistable electro-optic media do not exhibit a type corresponding to a one-to-one relationship between supplied voltage and liquid crystal optical state characteristics, and that the use of the term "gamma voltage" herein is not as accurate as in conventional liquid crystal displays, where gamma voltage defines an inflection point in the voltage level/output voltage curve.
The term "pulse" as used herein in its conventional sense refers to the integral of voltage with respect to time. However, some bistable electro-optic media act as charge sensors, and for such media an alternative definition of a pulse, i.e. the integral of the current over time (which is equal to the total amount of charge provided) may be used. The proper definition of the pulse depends on whether the medium is acting as a voltage-time pulse sensor or a charge pulse sensor.
Several types of bistable electro-optic displays are known. One type of electro-optic display is the rotating bichromal type, such as described in U.S. patent nos. 5808783; 5777782, respectively; 5760761, respectively; 6054071; 6055091, respectively; 6097531, respectively; 6128124, respectively; 6137467, respectively; and 6147791 (although this type of display is often referred to as a "spinning bicolor ball" display, the term "spinning bicolor cell" is more accurate because the cell is not spherical in some of the above patents). Such a display uses a large number of bodies (usually spherical or cylindrical) having two or more sections (sections) with different optical properties and an internal dipole. The bodies are suspended in liquid-filled vacuoles located in a matrix, and the bodies are free to rotate as the vacuoles are filled with liquid. The appearance of such a display is changed by applying an electric field thereto, thereby rotating the corpuscles into various positions and changing the portion of the corpuscles viewed from a viewing surface.
Another type of electro-optic medium uses an electrochromic medium, such as one in the form of a nanochromic film, the film including an electrode at least a portion of which is comprised of a semiconducting metal oxide and a plurality of reversibly color-changeable dye molecules attached to the electrode; see, e.g., O' Regan, B., et al, Nature 1991, 353, 737; and information display, 18(3), 24 (3 months 2002) of Wood, d. See also Bach, u., et al, adv.mater., 2002, 14(11), 845. Nanochromic films of this type are also described, for example, in U.S. patent No. 6301038, international application publication No. wo01/27690 and its co-pending serial application No.60/365368, both filed 3/18/2002; 60/365369 and 60/365385; serial application No.60/319279, both filed on 31/5/2002; 60/319280 and 60/319281 and serial application No.60/319438 filed on 31/7/2002.
Another type of electroluminescent display that has been studied earnestly and developed for many years is a particle-based electrophoretic display in which a plurality of charged particles move through a suspension under the influence of an electric field. Electrophoretic displays have superior brightness and contrast, wide viewing angles, bistable states, and low power consumption characteristics compared to liquid crystal displays. However, problems with the long-term image quality of these displays have prevented their widespread use. For example, the particles that make up electrophoretic displays are prone to settling, resulting in insufficient lifetime of these displays.
A number of patents and applications describing encapsulated electrophoretic media have been published recently, under the name of or assigned to the institute of technology and technology (MIT) and E Ink companies. Such an encapsulation medium comprises a plurality of small internal bodies (capsules), each of which itself comprises an internal phase comprising electrophoretically mobile particles suspended in a liquid suspension medium, with an internal body wall surrounding the internal phase. Typically, the endosome holds itself in a polymeric binder to form a coherent layer between two electrodes. Packaging media of this type are described in the following documents, such as U.S. patent numbers 5930026; 5961804, respectively; 6017584, respectively; 6067185, respectively; 6118426, respectively; 6120588, respectively; 6120839, respectively; 6124851, respectively; 6130773, respectively; 6130774, respectively; 6172798, respectively; 6177921, respectively; 6232950, respectively; 6249721, respectively; 6252564, respectively; 6262706, respectively; 6262833, respectively; 6300932, respectively; 6312304, respectively; 6312971, respectively; 6323989, respectively; 6327072, respectively; 6376828, respectively; 6377387; 6392785, respectively; 6392786, respectively; 6413790, respectively; 6422687, respectively; 6445374, respectively; 6445489 and 6459418; and U.S. patent application publication No. 2001/0045934; 2002/0019081, respectively; 2002/0021270, respectively; 2002/0053900, respectively; 2002/0060321, respectively; 2002/0063661, respectively; 2002/0063677, respectively; 2002/0090980, respectively; 2002/106847, respectively; 2002/0113770, respectively; 2002/0130832, respectively; 2002/0131147 and 2002/0154382, and international application publication No. WO 99/53373; WO 99/59101; WO 99/67678; WO 00/05704; WO 00/20922; WO 00/38000; WO 00/38001; WO 00/36560; WO 00/20922; WO 00/36666; WO 00/67110; WO 00/67327; WO 01/07961; WO 01/08241; WO01/17029 and WO 01/17041.
A number of the above-mentioned patents and applications recognise that the walls surrounding discrete micro-bodies (microcapsules) in an encapsulated electrophoretic medium can be replaced by a continuous phase, thereby producing a so-called dispersed polymer electrophoretic display in which the electrophoretic medium comprises a plurality of discrete droplets of an electrophoretic fluid and a continuous phase of a polymeric material, and the discrete droplets of the electrophoretic fluid in such a dispersed polymer display can be considered as endosomes or microbodies, even if a discrete endosomal membrane is not associated with each individual droplet; see, for example, WO01/02899, page 10, lines 6-19. See also co-pending serial application No.09/683903 filed on 28.2.2002, and corresponding international application PCT/US 02/06393. Thus, for the purposes of this application, such dispersed polymer electrophoretic media are considered to be a subspecies of encapsulated electrophoretic media.
Encapsulated electrophoretic displays generally do not suffer from the clumping and settling failure modes of conventional electrophoretic devices and provide further advantages such as the ability to print or coat the display on a variety of flexible and rigid substrates. (the use of the word "printing" is intended to include all forms of printing and coating including, but not limited to, premetered coating such as die coating, slot or extrusion coating, slide or step coating, curtain coating, roll coating such as roll-on-knife, roll coating in forward or reverse direction, gravure coating, dip coating, spray coating, meniscus coating, spin coating, brush coating, air knife coating, screen printing processes, electrostatic printing processes, thermal printing processes, ink jet printing processes, and other similar techniques.) the displays thus obtained may be flexible. Furthermore, since the display medium can be printed (using various methods), the display itself can become inexpensive.
A corresponding type of electrophoretic display is the so-called "microcell electrophoretic display". In microcell electrophoretic displays, the charged particles and the suspension are not encapsulated within microcapsules but are held within a plurality of cavities formed within a carrier medium, typically a polymer film. See, for example, International application publication No. WO02/01281, and published U.S. application No.2002-0075556, both assigned to Sipix Imaging, Inc.
The bistable or multistable nature of particle-based electrophoretic displays and the like that other electro-optic displays exhibit, are in sharp contrast to that of conventional liquid crystal ("LC") displays. Twisted nematic liquid crystals are not bistable or multistable but act as voltage sensors, so that a pixel of such a display is supplied with a set electric field, regardless of the gray level originally present on the pixel, to produce a specified gray level at the pixel. In addition, liquid crystal displays are driven in only one direction (from non-transmissive or "black" to transmissive or "bright"), and the reverse transition from the bright state to the black state is accomplished by reducing or eliminating the electric field. Finally, the grey scale of a liquid crystal display pixel is not sensitive to the polarity of the electric field, but only to its magnitude, whereas in practice commercial liquid crystal displays often reverse the polarity of the driving electric field at frequent intervals for technical reasons.
In contrast, bistable electro-optic displays act, to a first approximation, as pulse sensors, so that the final state of a pixel depends not only on the applied electric field and the time of the applied electric field, but also on the state of the pixel before the electric field is applied. Furthermore, it has now been found that, at least in many particle-based electro-optic displays, the pulses required to change a given pixel by equal changes in gray scale (as judged by the eye or standard optical instrumentation) need not be constant, nor need they be interchangeable. For example, consider a display in which each pixel can display a 0 (white), 1, 2, 3 (black) gray scale, preferably at regular intervals. (the spacing between gray levels may be linear in percent reflectance as measured by the eye or instrument, but other distributions may be used; for example, the distribution may be linear in L, or alternatively a specific gamma value may be provided; a gamma value of 2.2 is often used for monitors where the display is used as a replacement for a monitor and a similar gamma value may be used as desired.) it has been found that the pulses required to change a pixel from 0 to 1 (hereinafter referred to as a "0-1 transition" for convenience) are often different than those required for a 1-2 or 2-3 transition. Moreover, the pulses required for a 1-0 transition need not be the same as for the opposite 0-1 transition. In addition, some system representations exhibit a "memory" effect, such that (for example) the pulses required for a 0-1 transition vary slightly depending on whether a particular pixel undergoes a 0-0-1, 1-0-1 or 3-0-1 transition. (where the symbol "x-y-z" represents a sequence of chronologically addressed optical states, where x, y, z are all optical states 0, 1, 2 or 3.) although these problems can be alleviated or overcome by driving all pixels of the display to one of the extreme states for a fundamental period before driving the desired pixel to the other state, the resulting solid color "flicker" is often unacceptable; for example, a reader of an electronic book may need the text of the book to scroll down the screen, and if the display needs to be solid black or white to flash at frequent intervals, the reader may be confused or lose his position. Furthermore, such flickering of the display increases power consumption and may reduce the lifetime of the display. Finally, it has been found that, in at least some cases, the pulses required for a particular transition are influenced by temperature and the total operating time of the display, as well as the time a particular pixel remains in a particular optical state before a given transition, which factors need to be compensated for in order to ensure accurate grey scale reproduction.
In one aspect, the present invention seeks to provide a method and controller that can provide accurate gray scale levels to an electro-optic display without requiring pure color flashing on the display at frequent intervals.
Furthermore, as is readily apparent from the above description, the driving requirements of a bistable electro-optic medium are such that drivers designed for driving Active Matrix Liquid Crystal Displays (AMLCDs) are not adapted for use with bistable electro-optic medium-based displays. However, such AMLCD drivers are readily available commercially, have large allowable voltage ranges and high pin count packages, have an off-the-shelf basis, and are inexpensive, and thus such AMLCD drivers are attractive for driving bistable electro-optic displays, while customizing similar drivers for bistable electro-optic medium-based displays can be substantially more expensive and take up substantial design and manufacturing time. Thus, there are cost and development cycle advantages to modifying AMLCD drivers for bistable electro-optic displays, and the present invention seeks to provide a method and modified drivers which can meet this.
Also, as already mentioned, the present invention relates to a method of driving an electrophoretic display, keeping the driving pulses supplied to the electrophoretic display in long-term Direct Current (DC) balance. It has been found that encapsulated and other electrophoretic displays need to be driven with a precisely dc-balanced waveform (i.e., the integral of current over time for any particular pixel of the display is to be held to zero throughout an extended period of display operation) to keep the image stable, maintain symmetric switching characteristics, and provide maximum useful life of the display. Conventional methods for maintaining accurate dc balance require a precisely controlled power supply, precise voltage modulation drivers for gray scale, and crystal oscillators for timing, the provision of upper edges and the like add significant expense to the display.
Moreover, even with the addition of such expensive components, true dc balance is not achieved. It has been empirically found that many electrophoretic media have asymmetric current/voltage (I/V curves), and although the present invention is not limited in any way by this knowledge, it is believed that such asymmetric curves are attributable to the electrochemical voltage source in the medium. This asymmetric curve means that even when the voltages are carefully controlled to be exactly the same in both cases, the current when the medium is addressed to one extreme optical state (say black) is not the same as when the medium is addressed to the opposite extreme optical state (say white).
It has now been found that the extent of dc imbalance in an electrophoretic medium used in a display can be ascertained by measuring the open circuit electrochemical potential (hereinafter conveniently referred to as the "residual voltage" of the medium). When the residual voltage of the pixel is zero, it is already well balanced for dc. If the residual voltage is positive, it is a direct current imbalance in the positive direction. If the residual voltage is negative, it is DC imbalance in the negative direction. The present invention uses the residual voltage data to maintain long-term dc balance of the display.
Accordingly, in one aspect, the invention provides a method of driving a bistable electro-optic display having a plurality of pixels, each of which can display at least three gray levels (as in conventional display technology, the extreme black and white states are considered as two gray levels for purposes of calculating gray levels). The method comprises the following steps:
storing a look-up table holding data representing pulses required to convert an initial gray level to a final gray level;
storing data representing at least one initial state of each pixel of the display;
receiving an input signal representing a desired final state of at least one pixel of the display; and
an output signal is generated representing the pulses required to switch the initial state of the pixel to its desired final state, as determined from the look-up table.
This method is hereinafter referred to as the "look-up table method" of the present invention for convenience.
The invention also provides a device controller using the method. The controller includes:
storage means for storing a look-up table holding data representing the impulse data required to convert an initial grey level to a final grey level and data representing at least one initial state of each pixel of the display;
input means for receiving input signals indicative of a desired final state of at least one pixel of the display;
computing means for determining from the input signal, the stored data representing the initial state of said pixel and a look-up table the pulses required to change the initial state of said one pixel to the desired final state; and
output means for generating an output signal representative of said pulses.
The invention also provides a method of driving a bistable electro-optic display having a plurality of pixels, each of which can display at least three gray levels. The method comprises the following steps:
storing a look-up table holding data representing pulses required to convert an initial gray level to a final gray level;
storing data representing at least one initial state of each pixel of the display;
receiving an input signal representing a desired final state of at least one pixel of the display; and
an output signal is generated representing the pulse required to switch the initial state of said pixel to its desired final state, as determined from said look-up table, the output signal representing the time period of the substantially constant drive voltage to be supplied to said pixel.
The invention also provides a device controller using the method. The controller includes:
storage means for storing a look-up table holding data representing the impulse data required to convert an initial grey level to a final grey level and data representing at least one initial state of each pixel of the display;
input means for receiving input signals indicative of a desired final state of at least one pixel of the display;
computing means for determining from the input signal, the stored data representing the initial state of said pixel and a look-up table the pulses required to change the initial state of said one pixel to the desired final state; and
output means for generating an output signal representative of said pulse, the output signal being representative of a time period of a substantially constant drive voltage to be supplied to said pixel.
In another aspect, the invention provides a device controller using the method of the invention. The controller includes:
a storage means for storing a look-up table holding pulse data indicative of a desired gray level to be converted to a final gray level and a look-up table holding at least one initial state data indicative of each pixel of the display;
input means for receiving input signals indicative of a desired final state of at least one pixel of the display;
computing means for determining from the input signal, the stored data representing the initial state of said pixel and a look-up table the pulses required to change the initial state of said one pixel to the desired final state; and
output means for generating an output signal representative of said pulses, the output signal representing a plurality of pulses varying in at least one of voltage and duration, the output signal representing zero voltage after expiration of a predetermined time period.
In another aspect, the invention provides a driver circuit having output lines for connection to drive electrodes of an electro-optic display. The driver circuit includes a first input means for receiving a plurality of (n +1) -bit numbers representing a signal voltage and a polarity to be placed on the drive electrodes; and a second input means for receiving a clock signal. Upon receiving the clock signal, the driver circuit displays the selected voltage on its output line. In a preferred form of the driver circuit, the selected voltage may be between R and R + V by 2nAny one of discrete voltages or between R and R-V2nAny one of a number of discrete voltages, where R is a predetermined reference voltage (typically the voltage of the common front electrode of an active matrix display, as described in detail below) and V is the maximum voltage difference from the reference voltage that the driver circuit determines (assert). These selected voltages may be linearly distributed within the R + -V range or may be distributed in a non-linear manner, the non-linearity being controllable by two or more gamma voltages within a specific range, each gamma voltage defining a linear condition (region) between the gamma voltage and an adjacent gamma value or reference voltage.
In another aspect, the invention provides a driver circuit having output lines for connection to drive electrodes of an electro-optic display. The driver circuit comprises first input means for receiving a plurality of 2-bit numbers (2-bit numbers) indicative of the signal voltage and polarity to be placed on the drive electrodes; and a second input means for receiving a clock signal. Upon receiving the clock signal, the driver circuit displays on its output lines the voltages selected from R + V, R and R-V (where R and V are defined as above).
In another aspect, the invention provides a method of driving a bistable electro-optic display, in particular an electrophoretic display, which displays a residual voltage. The method comprises the following steps:
(a) providing a first drive pulse to a pixel of the display;
(b) measuring a residual voltage of the pixel after the first drive pulse; and
(c) a second drive pulse is supplied to the pixel after the measurement of the residual voltage, the magnitude of the second drive pulse being controlled in dependence on the measured residual voltage to reduce the residual voltage of the pixel.
The present method will hereinafter be referred to as the "residual voltage" method of the present invention for convenience.
FIG. 1 is a schematic diagram showing the apparatus of the present invention, a display driven by the apparatus and associated apparatus, designed to represent the overall system configuration;
FIG. 2 is a schematic block diagram of the controller unit shown in FIG. 1 and illustrating the output signals produced by the unit;
FIG. 3 is a schematic block diagram illustrating the manner in which the controller unit shown in FIGS. 1 and 2 generates certain of the output signals shown in FIG. 2;
FIGS. 4 and 5 illustrate two different reference voltage settings that may be used in the display of FIG. 1;
FIG. 6 is a schematic diagram showing the trade-off between pulse width modulation and voltage modulation methods in the look-up table method of the present invention;
FIG. 7 is a block diagram of a custom (custom) drive for use in the look-up table method of the present invention;
FIG. 8 is a flow chart illustrating a program executable by the controller unit shown in FIGS. 1 and 2;
figures 9 and 10 illustrate two drive configurations of the present invention;
fig. 11A and 11B show two parts of a third drive arrangement of the invention.
As has been noted above, the look-up table portion of the present invention provides a method and controller for driving an electro-optic display having a plurality of pixels, each of which can display at least three gray levels. The invention can of course be used for electro-optical displays having a larger number of grey levels, for example 4, 8, 16 or more.
Also as described above, driving bistable electro-optic displays requires a method that is significantly different from that typically used to drive Liquid Crystal Displays (LCDs). In a conventional (non-cholesteric) LCD, a particular voltage is applied to a pixel for a sufficient period to cause the pixel to attain a particular gray level. Furthermore, liquid crystal materials are only sensitive to the magnitude of the electric field, but not to its polarity. In contrast, bistable electro-optic displays act as pulse sensors, so there is no one-to-one mapping between applied voltages and acquired grey states; the pulse (and hence the voltage) that must be applied to a pixel to achieve a given grey state varies with the "initial" grey state of the corresponding pixel. Furthermore, since bistable electro-optic displays need to be driven in two directions (white to black, and black to white), the polarity and magnitude of the required pulses need to be specified.
It is contemplated that some of the terms used herein will be defined according to their ordinary meaning in the art of presentation. Much of the discussion below will focus on one or more pixels undergoing a single gray scale transition (i.e., a change from one gray scale to another) from an "initial" state to a "final" state. Obviously, the initial and final states are specified to take into account only a single transition of interest, whereas in most cases a pixel has undergone a transition before the "initial" state and has undergone a transition after the "final" state. As explained below, some embodiments of the present invention may not only consider the initial and final states of a pixel, but also the "previous" state that the pixel existed before the initial state was reached. Where a distinction is to be made between a plurality of preceding states, the term "first preceding state" is used to refer to a state in which the corresponding pixel has a (non-zero) transition prior to the initial state, the term "second preceding state" is used to refer to a state in which the corresponding pixel has a (non-zero) transition prior to the first preceding state, and so on. The term "non-zero transition" is used to refer to a "transition" that effects a change in at least one unit of gray scale; the term "zero transition" is used to refer to a "transition" that does not produce any change in the gray scale of the selected pixel (although other pixels of the display may undergo non-zero transitions at the same time).
As will be readily apparent to those skilled in the art, a simple embodiment of the method of the present invention may consider only the initial and final states of each pixel, in which case the look-up table is two-dimensional. However, as already indicated, some electro-optical media exhibit memory effects, and a medium is needed for which it is necessary to consider not only the initial state of each pixel, but also (at least) the first previous state of the pixel when generating the output signal, in which case the look-up table is three-dimensional. In some cases it may be necessary to consider more than one preceding state per pixel, thereby resulting in four (if only the first and second preceding states are considered) or more dimensions in the look-up table.
From a formal mathematical point of view, the invention can be seen to comprise an algorithm giving information about the initial and final and (optionally) previous states of an electro-optical pixel, and (optionally-see more detailed discussion below) about the physical state of the display, which will produce a function v (t) that can be applied to the pixel to effect a transition to the desired final state. From this form of perspective, the controller of the present invention can be viewed essentially as one physical embodiment of the algorithm, the controller serving as an interface between the device desiring to display information and the electro-optic display.
Ignoring the physical state information for the moment, the algorithm is encoded as a look-up table or transformation matrix according to the invention. The matrix will have one dimension for each desired final state and each dimension for the other states (initial and any previous states) used in the calculation. The elements of the matrix will include the function v (t) to be used for the electro-optic medium.
The elements of the look-up table or conversion matrix may take a variety of forms. In some cases, each element may contain a single number. For example, electro-optic displays may use highly accurate voltage modulation driver circuits capable of outputting a variety of different voltages above and below a reference voltage and simply provide the required voltages to the pixels at a standard predetermined period. In this case, each entry in the look-up table may simply be in the form of a single integer that specifies which voltage is to be supplied to a given pixel. In another case, each element would include a series of numbers relating to different locations of the waveform. For example, the embodiments of the invention described below use single or double pre-pulse waveforms, and specifying such a desired pulse requires specifying several numbers of different positions of the waveform. Also described below is an embodiment of the invention which effectively applies pulse length modulation by supplying a predetermined voltage to the pixel during the selection of several of a plurality of sub-scanning periods in a full scan. In such an embodiment, the elements of the conversion matrix may be in the form of several bits indicating whether a predetermined voltage is applied in each sub-scanning period of the corresponding conversion. Finally, as described in more detail below, in some cases, such as temperature compensated displays, it may be more convenient for the elements of the look-up table to be in the form of functions (or, indeed, more precisely, the coefficients of the various terms in such functions).
It is obvious in the first aspect of the present inventionThe look-up table used in some embodiments can become very large. To take an extreme example, consider an algorithm that considers initial, final, and two previous states for 256 (2)8) Operation of the invention for gray scale display. The required four-dimensional look-up table would be 232An entry. If each entry requires (say) 64 bits (8 bytes), the total amount of the look-up table would be about 32 gigabytes. While storing such large amounts of data on a desktop computer is not problematic, it may be problematic in portable devices. However, in practice the size of such a large look-up table can be reduced considerably. In many cases it has been found that only a small number of waveform types require a large number of different transitions, for example the length of the individual pulses of a common waveform varies between different transitions. Thus, the length of the individual entries in the look-up table can be reduced by having each entry include the following: (a) a pointer to an entry in the second table specifying one of a small number of waveform types to be used; and (b) a small number of parameters that specify how the normal waveform varies for the associated transition.
The values of the entries in the look-up table may be predetermined by an empirically optimized process. Basically, the pixels are set to the respective initial states, a pulse is provided which estimates that the desired final state to be achieved is approximately the same, and the final state of the pixel is measured to determine the deviation, if any, between the actual and desired final states. The process will then be repeated with modulated pulses until the deviation is less than a predetermined value, which can be determined by the function of the instrument used to measure the final state. In the case of this method which takes into account the previous states of one or more pixels, it is generally convenient, in addition to the initial state, to first determine the pulse required for a particular transition and then "fine tune" the pulse taking into account the different previous states when the pixel state is constant in the initial state used to determine the pulse and all the previous states.
The present invention seeks to provide compensation for pulsing when taking into account changes in temperature and/or overall operating time of the display, and operating time that may be required due to "aging" of some electro-optic media and changes in their state after prolonged operation. Such modulation may be achieved in one or two ways. First, the look-up table may be extended by additional dimensions for each variable considered in computing the output signal. Clearly, when dealing with continuous variables such as temperature and operation, the continuous variables need to be quantized in order to keep the look-up table within a certain limited size. To find the waveform to be applied to a pixel, the computing device may simply select the look-up table entry as the table closest to the measured temperature. Alternatively, to provide more accurate temperature compensation, the computing device may look for two contiguous look-up table entries on either side of the measured continuous variable and apply a suitable interpolation algorithm to obtain the desired entry at the middle of the measured variable. For example, assume that the matrix includes temperature entries that are incremented by 10 ℃. If the actual display temperature is 25 deg.C, the calculation looks for entries of 20 deg.C and 30 deg.C and uses the middle of the two. Note that since the change in the characteristics of the electro-optic medium with temperature is often not linear, the entries stored in the look-up table for the temperature settings may not be distributed linearly; for example, many electro-optic media change in temperature more rapidly at higher temperatures, so a pitch of 20 ℃ in the look-up table at low temperatures is sufficient, and a pitch of 5 ℃ at high temperatures is satisfactory.
An alternative method for temperature/operating time compensation is to use look-up table entries in the form of functions of physical variables or perhaps more accurate coefficients of standard terms in such functions. For the case of a display that simply considers the use of a time modulated drive scheme, where each transition is controlled by supplying a constant voltage (of either polarity) to each pixel for a variable length of time, thus, omitting any correction of environmental variables, each entry in the look-up table may contain only a single signed number representing the duration of the constant voltage to be applied and its polarity. If it is desired to correct such a display for temperature variations, the time T of the constant voltage applied is required for a particular transition at temperature TtIs represented by the formulaThe following are given:
Tt＝T0+AΔt+B(Δt)2
wherein T is0Is the time required at some standard temperature, typically the midpoint of the expected operating temperature range of the display, and Δ T is at T and at T0The difference between the measured temperatures, the entry in the look-up table may comprise T0And the values of a and B for a particular transformation involving a given entry, and the computing means may use these coefficients to calculate T at the measured temperaturet. Put it more generally, the computing means finds a look-up table entry suitable for the respective initial and final states and then uses the function defined by this entry to compute a suitable output signal that already takes into account the other variables that need to be considered.
The relevant temperature for the temperature compensation calculation is the temperature of the electro-optical medium at the corresponding pixel, which may differ significantly from the ambient temperature, especially in the case of displays intended for outdoor use, where, for example, sunlight acts through the front protective panel causing the temperature of the electro-optical medium layer to be substantially higher than the ambient temperature. In fact, in the case of a large bulletin board type outdoor sign, for example, if part of the display falls under the shadow of a nearby building while the other part is in the sun, the temperature will be different between different pixels on the same display. Thus, it may be desirable to embed one or more thermocouples or other temperature sensors in or adjacent to the electro-optic layer to detect the actual temperature of that layer. In the case of large displays, it may also be necessary to provide interpolation between the temperatures measured by the plurality of temperature sensors to estimate the temperature of each particular pixel. Finally, in the case of a large display constructed of many individually replaceable modules, the method and controller of the present invention can specify different operating times for the pixels in different modules.
The method and controller of the present invention may also take into account the dwell time of a particular pixel to be driven (i.e., the period during which the pixel maintains a non-zero transition). It has been found that, in at least some cases, the pulse required for a given transition varies with the dwell time of the pixel in its optical state. Thus it is desirable or necessary to vary the pulse for a given transition as a function of the dwell time of the pixel in its initial optical state. To accomplish this, the look-up table may include an additional dimension indexed by a counter to indicate the dwell time of the pixel in its initial optical state. In addition, the controller requires an additional memory area containing a counter for each pixel in the display. This also requires a display clock which is incremented at set intervals by the count value stored at each pixel. The length of the interval must be an integer multiple of the display frame period and must therefore not be less than one frame period. The size and clock frequency of the counter will be determined by the length of time that the varying pulses are used over and the necessary time resolution. For example, storing a 4-bit counter for each pixel may allow pulses to be varied at 0.25 second intervals over a 4 second period (4 seconds × 4 counts/second 16 counts 4 bits). The counter may be cleared upon the occurrence of a particular event, such as the transition of a pixel to a new state. Once its maximum value is reached, the counter may be set to "roll over" to a zero count, or to maintain its maximum value until it is cleared.
The look-up table method of the present invention may of course be varied taking into account any other physical parameter having a detectable effect depending on the pulse required to produce any one or more specific transitions of the electro-optic medium. For example, if the electro-optic medium is found to be humidity sensitive, the method may be modified to incorporate a modification of the ambient humidity.
For bistable electro-optic media, the look-up table will have the characteristic that for zero transitions where the initial and final states of any pixel are the same, the entry is zero, or in other words, no voltage is applied to the pixel. As a corollary, if no pixel change occurs on the display in a given interval, no pulse need be applied. This allows for ultra-low power operation and also ensures that the electro-optic medium is not overdriven when displaying a static image. Typically, the look-up table will only retain information about non-zero transitions. In other words, for both images, I and I +1, if a given pixel is in the same state in I and I +1, then state I +1 is not stored in the previous state table and no information is stored until after the pixel has undergone a transition.
The controller of the present invention may have various physical forms, as will be apparent to those skilled in the modern electronics art. And any conventional data processing components may be used. For example, the method may be implemented using a general purpose digital computer in combination with appropriate devices (e.g., one or more digital-to-analog converters, "DACs") for converting the digital output from the computer to appropriate pixel voltages. Alternatively, the method of the present invention may be implemented using an Application Specific Integrated Circuit (ASIC). In particular, the controller of the present invention may be in the form of a video card that can be inserted into a personal computer so that the image produced by the computer is displayed on an electro-optic screen that replaces or is a supplement to an existing display screen, such as an LCD. Since the structure of the controller of the present invention is well at the level of technology in image processing technology, it is not necessary here to describe its circuit details in detail.
The preferred physical embodiment of the controller of the present invention is a timing controller Integrated Circuit (IC). The IC receives input image data and outputs control signals for the data collection and selection driver ICs to generate appropriate voltages across the pixels to produce the desired image. The IC may receive image data by accessing a memory buffer that holds the image data, or may receive a signal for driving a conventional LCD panel, from which the image data is extracted. It may also receive any serial signal that holds the necessary pulse calculation information that it needs to perform. Alternatively, the timing controller may be implemented in software or incorporated as part of the CPU. The timing controller may also have the capability to measure any external parameter affecting the display operation, such as temperature.
The controller may operate as follows. A look-up table is stored in memory that the controller can retrieve. For each pixel in turn, all necessary initial, final and (optional) prior and physical state information is provided as input. These state information are then used to compute an index into the look-up table. In the case of quantized temperature or other corrections, the return from this query would be a voltage, or an array of voltages over time. The controller would repeat this operation for two scaled test temperatures in the look-up table and then interpolate between these values. For algorithmic temperature correction, the return value of the look-up table will have one or more parameters, which will then be substituted into an equation along with the temperature to determine the appropriate form of drive pulse, as described above. This procedure can be implemented in a similar manner for any other system variable that requires a real-time variation of the drive pulse. One or more of these system variables may be determined by, for example, the values of programmable resistors provided on the display panel at the time of construction or a memory stored in the EPROM in order to optimize the performance of the display.
An important feature of the display controller is that it does not require several complete display scans to complete an image update as in most displays and in most practical cases. Several scans required for one image update should constitute one indivisible unit. If the display controller and the image source operate synchronously, the controller must ensure that the data used to calculate the supplied pulses remains constant throughout the scan. This can be done in one or two ways. First, the input image data will be stored by the display controller in a separate buffer (optionally, if the display controller is accessing the display buffer via the dual port memory, it should block access from the CPU). Second, in the first scan, the controller may store the calculated pulses in a pulse buffer. The benefit of this second option is that the overhead for scanning the panel is only done once per transition and the data for keeping the scan is output directly from the buffer.
Alternatively, the image updates may be processed in a synchronized manner. In general, individual pixels may start transitions in the middle of a frame or flip an already started transition, although it typically takes several scans to produce a complete transition between two images at a time. To accomplish this, the controller must remember which portion of the overall transition has been completed for a given pixel. If a request is received to change the optical state of a pixel that is not in the current transition, the counter for that pixel is cleared and the pixel begins transitioning in the next frame. If the pixel is in active transition when a new request is received, the controller provides an algorithm to determine how to go from the current frame intermediate state to the new state. For a normal image stream of 1 bit, one possible algorithm is to simply provide a pulse of inverted polarity, which is amplified and has the same duration as the previous pulse portion already provided.
To minimize the power required to operate the display and maximize the image stability of the electro-optic medium, the display controller may stop the scanning of the display and reduce or approach zero the voltage applied to all pixels when no pixel is switching in the display. It is highly advantageous that the display controller can shut off power to its respective row and column drivers when the display is in the "hold" state, which minimizes power consumption. In this scheme, the driver is reactivated when the next pixel transition is requested.
Figure 1 schematically shows the device of the invention used with an associated device. All of the devices shown in fig. 1, generally designated 10, include an image source, such as a personal computer 12 shown outputting image data represented on a data line 14. The data line 14 may be of any conventional type and may be a single data line or a bus; for example, the data lines 14 may include Universal Serial Bus (USB), serial, parallel, IEEE-1394, or other lines. The data placed on line 14 may be in the form of a conventional bit mapped image, such as a Bitmap (BMP), a tagged image file format (TIF), a commutable image format (GIF), or a joint (joint) image experts group (JPEG) file. However, alternatively, the data placed on line 14 may be in the form of signals used to drive a video device; for example, many computers provide a video output for driving an external display, and signals on such an output may be used with the present invention. Those skilled in the art of image processing will appreciate that the apparatus of the present invention described below may perform basic file format conversion and/or decoding to use different types of available input signals, but such conversion and/or decoding is well known to those skilled in the art and, therefore, the apparatus of the present invention will be described only in this regard: the image data as its original input has been converted to a format that can be processed for the apparatus of the present invention.
As described in detail below, the data lines 14 extend to a controller unit 16 of the present invention. The controller unit 16 generates one set of output signals on a data bus 18 and a second set of signals on a separate data bus 20. The data bus 18 is connected to two row (or gate) drivers 22, while the data bus 20 is connected to a plurality of column (or source) drivers 24 (the number of column drivers shown in fig. 1 is greatly reduced for ease of illustration). Row and column drivers control the operation of bistable electro-optic display 26.
The apparatus shown in figure 1 was chosen to represent the various available units that are most suitable for one type of experimental "test plate" unit. In actual commercial manufacture, such as in conventional laptops and personal digital assistants equipped with LCDs, the controller 16 will typically become part of the same physical unit of the display 26 as will the image source. Similarly, the invention is shown in FIG. 1 and will be described below primarily in connection with an active matrix display structure having a single common transparent electrode on one side of the electro-optic layer, the common electrode extending across all pixels of the display. Typically, the common electrode is positioned between the electro-optic layer and the viewer and forms a viewing surface through which the viewer views the display. On the opposite side of the electro-optical layer a matrix of pixel electrodes arranged in rows and columns is placed, such that each pixel electrode is uniquely defined by the intersection of a separate row and a separate column. Thus, the electric field experienced by each pixel of the electro-optic layer is controlled by varying the voltage supplied to the corresponding pixel electrode relative to the voltage supplied to the common front electrode (generally denoted as "Vcom"). Each pixel electrode is connected to at least one transistor, typically a thin film transistor. The gates of the transistors on each row are connected to one of the row drivers 22 via a separate elongate row electrode. The sources of the transistors on each column are connected to one of the column drivers 24 via a separate elongate column electrode. The drain electrode of each transistor is directly connected to the pixel electrode. It will be appreciated that the gate to row and source to column assignments are random and, as with the source and drain assignments, can be reversed. However, the following description will assume a conventional allocation.
In operation, the row driver 22 applies voltages to the gates so that one and only one row of transistors is on at any given time. At the same time, a column driver 24 supplies a predetermined voltage to each column electrode. Thus, voltages applied to the column drivers are applied to only one row of pixel electrodes, which writes (or at least partially writes) a desired image on the electro-optic medium. The row driver then switches so that the transistors in the next row are turned on, applying a different set of voltages to the column electrodes, and writing the next row of images.
It is emphasized that the invention is not limited to such active matrix displays. Any switching scheme may be used to provide a waveform to a pixel, as long as the correct waveform for each pixel of the image is determined in accordance with the present invention. For example, the invention may use a so-called "direct drive" scheme, in which a separate drive line is provided for each pixel. In principle, the invention can also be used with a passive matrix driving scheme for some LCDs, but it is noted that many bistable electro-optic media are not suitable for passive matrix driving because they lack a switching threshold (i.e. they change optical state as long as a small electric field is provided for an extended period). However, as it appears that the invention finds its primary application in active matrix displays, the invention is described herein primarily with reference to such displays.
The controller unit 16 (fig. 1) has two main functions. First, using the method of the present invention, the controller calculates a two-dimensional matrix of pulses (or waveforms) that must be applied to the pixels of the display to change from the initial image to the final image. Second, the bistable electro-optic display is driven using conventional drivers designed for LCDs, and the controller 16 calculates from the pulse matrix all timing signals required to apply the desired pulses to the pixel electrodes.
As shown in FIG. 2, the controller unit 16 shown in FIG. 1 has two main parts, namely a frame buffer 16A that buffers the final image data representing the controller 16B to be written to the display 26 (FIG. 1), and the controller itself, labeled 16B. Controller 16B reads the data from buffer 16A on a pixel-by-pixel basis and generates various signals on data buses 18 and 20 as described below.
The signals shown in fig. 2 are as follows:
d0: D5-A six-bit voltage value for a pixel (obviously, the number of bits in the signal may vary depending on the particular row and column driver used)
POL-Pixel polarity relative to Vcom (see below)
START-A START bit is placed in the column driver 24 to initiate the loading of pixel values
Horizontal sync signal for HSYNC-latch column driver
PCLK-Pixel clock with switch Start bit along Row driver
VSYNC-vertical synchronization Signal that loads the start bits into the Row driver
OE-latch the output enable signal of the row driver.
Of these signals, VSYNC and OE supplied to the row driver 22 are substantially the same as the corresponding signals supplied to the row driver in a conventional active matrix LCD, since the row scanning method in the device shown in fig. 1 is in principle the same as the LCD scanning method, although the exact timing of these signals will of course vary depending on the precise electro-optic medium used. Similarly, for START, HSYNC and PCLK, although their timing will vary depending on the precise electro-optic medium used, the signals provided to the column drivers are substantially the same as the corresponding signals provided to the column drivers in a conventional active matrix LCD. Therefore, it may be considered unnecessary to further describe these output signals.
FIG. 3 illustrates in a highly schematic manner the method by which the controller 16B shown in FIG. 2 generates the D0: D5 and POL signals. As described above, the controller 16B stores data representing a final image 120 (which is intended to be written to the display), an initial image 122 previously written to the display, and optionally one or more previous images 123 written to the display prior to the initial image. The embodiment of the invention shown in fig. 3 stores two such prior images 123. (obviously, the storage of the necessary data may be in the controller 16B or in an external data storage device. the controller 16B uses the data of the initial, final and previous images 120, 122 and 123 for a particular pixel (shown in phantom in FIG. 3 as the first pixel in the first row) as pointers into a look-up table 124 that provides the values of the pulses that must be applied to the particular pixel to change the state of that pixel to the desired gray level in the final image. the resultant output from the look-up table 124 and the output from the frame counter 126 are provided to a voltage v. frame array 128 that produces D0: D5 and POL signals.
Controller 16B is designed for use with a TFT LCD driver that is equipped with a pixel switching circuit that typically changes the polarity of adjacent pixels relative to the top surface. The spaced pixels may be designed even and odd and connected to opposite sides of the voltage ladder. In addition, the driver inputs labeled "polarity" are used to switch the polarity of the even and odd pixels. The driver is provided in conjunction with four or more gamma voltage levels, the setting of which can determine the local slope of the voltage-level curve. A typical example of a commercial Integrated Circuit (IC) with these features is Samsung KS 0652300/309 channel TFT-LCD source driver. As mentioned above, the display to be driven uses a common electrode on one side of the electro-optic medium, to which a voltage reference such as the "top surface voltage" or "Vcom" is applied.
In one embodiment as shown in fig. 4, the reference voltage of the driver is arranged such that the top surface voltage is at half the maximum voltage (Vmax) that the driver can provide, i.e. the top surface voltage is at
Vcom＝Vmax/2
While the gamma voltage is arranged to vary linearly up and down the top surface voltage. (FIGS. 4 and 5 are plotted assuming an odd gamma voltage, so that, for example, gamma voltage VGMA (n/2+1/2) is equal to Vcom in FIG. 4. if there is an even gamma voltage, VGMA (n/2) and VGMA (n/2+1) are both set equal to Vcom. similarly, in FIG. 5, if there is an even gamma voltage, VGMA (n/2) and VGMA (n/2+1) are both set equal to Vs). The pulse length required to obtain a full transition is determined by assigning the maximum pulse required to build up a new image by Vmax/2. The pulses may be converted to frame numbers by multiplying by the display scan rate. The necessary number of frames is then multiplied by 2 to give a considerable number of even and odd frames. The even and odd frames are set high or low relative to the frame corresponding to the polarity bit. For each pixel in each frame, controller 16B must provide an algorithm to (1) whether the pixel is even or odd; (2) for the frame under consideration, whether the polarity bit is high or low; (3) whether the desired pulse is positive or negative; and (4) the magnitude of the desired pulse as its input. The algorithm then determines whether the pixel can be addressed with the desired polarity in this frame. If so, an appropriate drive voltage (pulse length) is supplied to the pixel. If not, the pixel is stopped at the top surface voltage (Vmax/2) to be in a hold state in which no electric field is applied to the pixel in the present frame.
For example, consider two adjacent pixels in a display, an odd pixel 1 and an even pixel 2. Also, it is assumed that when the polarity bit is high, the odd pixels can access the positive drive voltage range (i.e., above the top surface voltage), and the even pixels can access the negative voltage (i.e., below the top surface voltage). If both pixels 1 and 2 need to be driven with a positive pulse, then the following sequence must exist:
(a) in the positive polarity frame, pixel 1 is driven with a positive voltage, and pixel 2 is held at the top surface voltage; and
(b) in the negative polarity frame, pixel 1 is held at the top surface voltage, while pixel 2 is driven with a positive voltage.
Although frames will typically alternate in positive and negative polarities of 1: 1 (i.e., alternate with each other), this is not required; for example, all odd frames may be grouped together followed by all even frames. This results in alternating columns of the display being driven in two separate groups.
The main advantage of this embodiment is that the common front electrode does not have to be switched during operation. The main advantage is that the maximum drive voltage of the electro-optical medium that can be obtained is only half of the maximum voltage of the driver, whereas each row can be driven only 50% of the time. Thus, the refresh time of such a display is four times the electro-optic medium switching time at the same maximum drive voltage.
In a second embodiment of this form of the invention, the gamma voltages of the drivers are arranged as shown in figure 5, with the common electrode switching between V0 and V Vmax. The gamma voltages arranged in this way allow even and odd pixels to be driven in a single direction simultaneously, but require the common electrode to be switched to approximately opposite drive polarities. Furthermore, since this arrangement is symmetrical about the top surface voltage, a particular input to the driver will result in the same voltage being applied to either odd or even pixels. In this case, the inputs to the algorithm are the magnitude and sign of the desired pulse, and the polarity of the top surface. If the current common electrode is set corresponding to the sign of the desired pulse, this value is the output. If the desired pulse is in the opposite direction, the pixel is set to the top voltage so that no electric field is applied to the pixel in this frame.
As described in the previous embodiment, the necessary length of the drive pulse can be calculated in the present embodiment by dividing the maximum drive voltage by the maximum pulse, and the value is converted into the number of frames by multiplying the display refresh rate. The number of frames must be doubled again to account for the fact that the display can only be driven in one direction at a time corresponding to the top surface.
The main advantage of the second embodiment is that the full voltage of the driver can be used and all outputs can be driven at once. However, the two frames need to be driven in opposite directions. Thus, the refresh time of such a display is twice the switching time of the electro-optical medium at the same maximum drive voltage. The main disadvantage is the need to switch the common electrode, which may lead to undesired voltage noise in the electro-optical medium, the transistors connected to the pixel electrodes, or both.
In either embodiment, the gamma voltage is generally distributed with a linear slope between the maximum voltage of the driver and the top surface voltage. Depending on the driver design, one or more gamma voltages at the top surface value may be required to ensure that the driver is able to generate a true top voltage at the output.
Reference has been made above to the limitations that are required to adapt the method of the present invention to conventional drivers for LCD design. More specifically, the column drivers of conventional LCDs, and in particular Super Twisted Nematic (STN) LCDs (which can control higher voltages than other types of column drivers), can only apply one of two voltages to a drive line at any given time, since this is all that is required for a liquid crystal material that is not polarity sensitive. Conversely, a minimum of three driver voltage levels are required to drive a polarity-sensitive electro-optic display. The three required driver voltages are V-which drives the pixel negative with respect to the top voltage, V + which drives the pixel positive with respect to the top voltage, and 0V with respect to the top voltage which keeps the pixel in the same display state.
The method of the invention can, however, be implemented in a conventional LCD driver of this type, the controller provided being arranged to supply one or more column drivers and the row drivers associated therewith with an appropriate sequence of voltages in order to supply the necessary pulses to the pixels of the electro-optical display.
This method has two main variables. In a first variant, all the pulses supplied must have one of three values of + I, -I or 0, where:
+I＝-(-I)＝Vapp×tpulse
where Vapp is the voltage provided above the top surface voltage, and tpulseIs the pulse length in seconds. This variable only allows the display to operate in binary (black/white) mode. In a second variant, the pulses supplied may vary from + I to-I, but must be an integer multiple of Vapp/freq, where freq is the refresh frequency of the display.
This aspect of the invention takes advantage of the fact that, as already noted, conventional LCD drivers are designed to reverse polarity at frequency intervals to avoid certain undesirable effects that may occur in the display. Thus, such drivers are designed to receive a polarity or control voltage from the controller, which may be high or low. When a low control voltage is asserted, the output voltage at any given driver output line may take on one of two out of three voltages that may be required, say V1 or V2, whereas when a high control voltage is asserted, the output voltage at any given driver output line may take on a different one of three voltages that may be required, say V2 or V3. Thus, only if two out of the three required voltages can be addressed at any particular time, all three voltages can be achieved at different times. These three required voltages will typically satisfy the following relationships:
V2＝(V 3+V1)/2
v1 may be at or near logic ground.
In this method of the invention, the display would be scanned 2 × tpulseXfreq times. For half of these scans (i.e., for t)pulseX freq scans) the output of the driver will be set to V1 or V2, which will typically be equal to-V and Vcom, respectively. Thus, in theseDuring scanning, the pixels are either driven negative or remain in the same display state. For the other half of the scan, the output of the driver would be switched to V2 or V3, which would typically be equal to Vcom and + V, respectively. In these scans the pixels are either driven positive or remain in the same display state. Table 1 below illustrates how these options combine to produce either a drive or a hold state in either direction; of course a corresponding positive drive, which will result in a black state, and a negative drive, which will result in a bright state, is a function of the particular electro-optical medium used.
TABLE 1 drive sequence for achieving bidirectional drive pulses with retention of STN driver
There are a number of different ways to arrange the two parts of the drive scheme (i.e. two different types of scans or "frames"). For example, the two types of frames may alternate. In practice, when driven in opposite directions in alternating frames, the electro-optic medium will appear both clear and dark if done at a high refresh rate. Alternatively, all frames of one type may occur before any of the frames of the second type; this results in a two-step driving external characteristic. Of course other arrangements are possible; for example, two or more frames of one type followed by two or more frames of the opposite type. Furthermore, if no pixel needs to be driven in one of the two directions, the frame of that polarity is eliminated, reducing the drive time by 50%.
The second variable may provide an image with multiple gray levels when the first variable can only produce a binary image. This is achieved by combining the above-described pulse width modulated drive schemes for the different pixels. In this case, the display is scanned again by 2 × tpulseX freq times, but only any particular pixel is supplied with a drive voltage in enough of these scans to ensure that the desired pulse for the particular pixel is achieved. For example, for each pixel, the information mentioned can be recordedFor the total pulse, when the pixel reaches its desired pulse, the pixel will remain at the top surface voltage for all of the following scans. For pixels that need to be driven in less than the total scan time, the time of the drive portion (i.e., the portion of time during which a pulse is provided to change the display state of the pixel as opposed to the hold portion during which the applied voltage simply maintains the display state of the pixel) can be distributed over the total time in different ways. For example, the entire driving section may be set to start at the beginning of the total time, or the entire driving section may be set to finish at the end timing of the total time instead. If, as in the first variant, no more pulses of a particular polarity need to be supplied to any pixel at any time in the second variant, the scan for supplying pulses of that polarity can be cancelled. This may mean that the overall pulse is shortened, for example if the maximum pulse provided in the positive and negative directions is less than the maximum allowed pulse.
Assuming a highly simplified case for the purpose of illustration, it is assumed that the above-described gray scale scheme for a display has four gray scales, namely black (0 level), dark gray (1 level), bright gray (2 level) and white (3 level). One possible driving scheme for such a display is outlined in table 2 below.
TABLE 2
| Frame No. | 1 | 2 | 3 | 4 | 5 | 6 |
| Parity of parity | Magic card | Doll | Magic card | Doll | Magic card | Doll |
| Conversion | | | | | | |
| 0-3 | + | 0 | + | 0 | + | 0 |
| 0-2 | + | 0 | + | 0 | 0 | 0 |
| 0-1 | + | 0 | 0 | 0 | 0 | 0 |
| 0-0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 3-0 | 0 | - | 0 | - | 0 | - |
| 2-0 | 0 | - | 0 | - | 0 | 0 |
| 1-0 | 0 | - | 0 | 0 | 0 | 0 |
For ease of illustration, it is assumed that only six frames are used in the present drive scheme, although in practice a greater number of frames will typically be used. The frames alternate between odd and even. Transitions tending to white (i.e., transitions with increasing gray levels) are driven only in odd frames, while transitions tending to black (i.e., transitions with decreasing gray levels) are driven only in even frames. In any frame where the pixel is not driven, it is held at the same voltage as the common front electrode, as indicated by "0" in table 2. For a 0-3 (black-to-white) transition, a drive-to-white pulse (i.e., holding the pixel electrode at a voltage that tends to increase the pixel gray scale relative to the common front electrode) is provided in each odd frame, frames 1, 3, and 5. On the other hand, for 0-2 (black to light gray) transitions, only one pulse tending to white is provided in frames 1 and 3, while no pulse is provided in frame 5; this is of course random, for example a whitening pulse may be applied in frames 1 and 5 and no pulse applied in frame 3. For a 0-1 (black to dark gray) transition, only one pulse tending to white is applied in frame 1, and no pulses are applied in frames 3 and 5; in addition, this is also random, e.g. a whitening pulse may be applied in frame 3, while no pulse is applied in frames 1 and 5.
The trending black transitions are processed in a manner very similar to the corresponding trending white transitions, except that the trending black pulses are applied only in even frames of the present drive scheme. It is believed that one skilled in the art of driving electro-optic displays will readily understand from the foregoing description the manner in which transitions are not shown in table 2.
The groups of pulses may be independent (stand-along) transitions between the two images, or they may be part of a sequence of pulses designed to perform image transitions, such as in a slide-show waveform.
Whilst the emphasis has been placed on the above method of the invention which allows the use of conventional drivers designed for use with LCDs, the invention can also be used with customised drivers and a driver for enabling accurate control of grey scale states in electro-optic displays, and the implementation of fast writing of the display will now be described with reference to figures 6 and 7.
As described above, many electro-optic media first respond to a voltage pulse, which can be expressed as Vtiming t (or more conventionally, by the integral of V with respect to t), where V is the voltage applied to a pixel and t is the time elapsed for the application of that voltage. Thus, the grey state can be obtained by modulation of the length of the voltage pulse applied to the display, or by modulation of the applied voltage, or a combination of both.
In the case of pulse width modulation of an active matrix display, the achievable pulse width resolution is only the inverse of the display refresh rate. In other words, for a display with a 100Hz refresh rate, the pulse length may be subdivided into 10ms intervals. This is because each pixel in the display is addressed only once per scan, i.e. when the select lines of the pixels in that row are activated. For the remainder of the time, the voltage across the pixel may be maintained by a storage capacitor as described in the aforementioned WO 01/07961. As the response speed of the electro-optic medium becomes faster, the slope of the reflectivity curve with respect to time becomes steeper and steeper. Thus, to maintain the same gray scale resolution, the refresh rate of the display must be increased accordingly. The increase in refresh rate results in higher power consumption, which eventually becomes impossible as transistors and drivers are expected to charge pixel and line capacitances in shorter and shorter times.
In another aspect, in voltage modulated displays, the pulse resolution is determined solely by the number of voltage levels, independent of the speed of the electro-optic medium. The effective resolution can be increased by using a non-linear distribution of voltage levels, which are concentrated where the voltage/reflectivity response of the electro-optic medium is steepest.
Fig. 6 schematically illustrates the trade-off between Pulse Width Modulation (PWM) and Voltage Modulation (VM) approaches. The horizontal axis represents pulse width, and the vertical axis represents voltage. The reflectance of a particle-based electrophoretic display as a function of these two parameters is represented as a contour plot with regions and spaces representing a difference of 1L x in the reflected brightness of the display, where L x has the definition of a common ICE:
L*＝116(R/R0)1/3-16
where R is the reflectance and R0 is a standard reflectance value. (it has been empirically found that the difference in brightness at 1L is just significant for the mean subject in the dual excitation experiment.) the particular particle-based electrophoretic medium used in this experiment summarized in fig. 6 is shown to have a response time of 200ms at maximum voltage (16V).
The effect of the individual pulse width modulation can be determined by the tiles traversing horizontally along the top, while the effect of the individual voltage modulation is seen by examining the vertical edge on the right. From this plot it is clear that if a display using this particular medium is driven in a Pulse Width Modulation (PWM) mode at a refresh rate of 100Hz, it is not possible to obtain a reflectivity within ± 1L in the steepest intermediate gray regions of the profile. In the Voltage Modulation (VM) mode, achieving a reflectivity within ± 1L would require 128 equally spaced voltage levels while operating at a frame rate as low as 5Hz (assuming, of course, that the voltage holding capability provided by the capacitors is sufficiently high). Furthermore, the two methods can be combined to achieve the same accuracy with a smaller voltage level. To further reduce the required voltage steps, they may be concentrated in the middle steep part of the curve shown in fig. 6 and sparse in the outer regions. This can be done with a small number of input gamma voltages. To further reduce the required voltage levels, they can be concentrated on advantageous values. For example, if any desired grey state transition cannot be met using very small voltages within the allotted addressing time, such small voltages are not useful for achieving the transition. Selecting a voltage profile that excludes such small voltages enables the allowed voltages to have a more favorable profile.
As noted above, bistable electro-optic displays are not like the LCD in which the polarity of the drive voltages is reversed in successive frames (images) as is done in LCDs, and frame, pixel and line reversals are unnecessary and can actually be counterproductive. For example, an LCD driver with pixel inversion delivers voltages of alternating polarity in alternate frames. In this way it is only possible to deliver pulses of the appropriate polarity in half the frame. This is not a problem in LCDs, as liquid crystal materials are not polarity sensitive, but in bistable electro-optic displays twice as long is required to address the electro-optic medium.
Similarly, because bistable electro-optic displays are pulse sensors rather than voltage sensors, the displays integrate voltage errors over time, which can result in large shifts in the pixels of the display from their desired optical state. This makes it important to use drivers with high voltage accuracy, with a tolerance of + -3 mV or less being recommended.
In order for a driver to be able to address a single color XGA (1024 x 768) display panel at a 75Hz refresh rate, a maximum pixel clock rate of 60Hz is required; it is within the state of the art to obtain such a clock frequency.
As already mentioned, one of the main advantages of particle-based electrophoretic and other similar bistable electro-optic displays is their image stability, which in turn gives the opportunity to operate the display with very low power consumption. Taking maximum advantage of this opportunity, power to the driver can be disabled when the image is unchanged. Thus, the driver can be designed to power down in a controllable manner without generating any parasitic voltage on the output line. Because entering and leaving such a "sleep" mode can be a common event, the power down and power up sequences are as fast as possible and have minimal impact on the useful life of the drive.
In addition, there should be an input pin that makes all the output pins of the driver Vcom, which can keep all the pixels in their current optical state without the driver powering down.
The drive of the present invention is useful, inter alia, for driving high-resolution media, high information capacity portable displays, such as 7 inch (178mm) diagonal XGA monochrome displays. To minimize the number of integrated circuits required in such a high resolution panel, it is desirable to use drivers with a high number of (e.g., 324) outputs per chip package. It is also desirable that the driver have an option to operate in one or more other modes and have less output enable. The preferred method of placing the integrated circuit on the display panel is Tape Carrier Packaging (TCP), which requires the size and placement of the driver outputs to be arranged to facilitate use of the method.
The driver will typically be used to drive the dielectric active matrix plate with voltages as small as around 30V. Therefore, the driver needs to be able to drive a capacitive load of about 100 PF.
A block diagram of a preferred drive of the present invention, generally designated 200, is shown in fig. 7. The driver 200 includes a shift register 202, a data register 204, a data latch 206, a digital-to-analog converter (DAC)208, and an output buffer 210. This driver differs from those commonly used to drive LCDs in that it provides a polarity bit associated with each pixel of the display and produces an output above and below the top surface voltage through control of the corresponding polarity bit.
The signal description for this preferred driver is given in table 3 below:
the driver 200 operates in the following manner. First, shift register 202 is reset to a start state by setting, for example, DIO1 high to provide a start pulse. (it will be readily apparent to those skilled in the display driver art that the various DIOx inputs provided to the shift register enable the driver to be used with displays having different numbers of columns, but only one of these inputs is used for any given display, while the others are always limited to low.) now the shift register will operate in the conventional manner used in LCDs; at each CLK1 pulse, one and only one of the 162 outputs of shift register 202 goes high, the others remain low, and the high output shifts one position at each CLK1 pulse. As schematically represented in fig. 7, each of the 162 outputs of the shift register 202 is connected to two inputs, an odd input and an even input, of the data register 204.
The display controller (compare fig. 2) provides two six-bit pulse values D0(0:5) and D1(0:5) and two single-bit polarity signals D0POL and D1POL at the inputs of the data register 204. On the rising edge of each clock pulse CLK1, two seven-bit numbers (D0POL + D0(0:5) and D1POL + D1(0:5)) are written into the registers of the data register 204 in conjunction with the selected (high level) output of the shift register 202. Thus, after 162 clock pulses CLK1, 324 seven-bit numbers (relative to the pulse value for a complete row in a frame display) have been written into 324 of the data registers 204.
These 324 seven-bit numbers are transferred from the data register 204 to the data latch 206 on the rising edge of each clock pulse LCK 2. These numbers placed in the data latches 206 are read by the DAC208 and, in conventional manner, the corresponding analogue values are placed at the output of the DAC208 and fed via a buffer 210 to the column electrodes of the display where they are supplied to the pixel electrodes in a row selected in conventional manner by a row driver (not shown). It should be noted, however, that the polarity of each column electrode corresponding to Vcom is controlled by the polarity bit D0POL or D1POL written into the data latch 206 so that these polarities do not change between adjacent column electrodes as is conventional in the use of LCDs.
Fig. 8 is a flow chart illustrating a program that may be run by the controller unit shown in fig. 1 and 2. This program, generally designated 300, is for use with the look-up table method of the present invention (described in more detail below) in which all pixels of the display are erased and re-addressed each time an image is written or re-refreshed.
The process begins with a "power on" step 302 that initiates with the controller, typically as a result of a user input, such as the user pressing a power key of a Personal Digital Assistant (PDA). Step 302 may also be initiated by, for example, opening of a PDA case (which may be detected by a mechanical sensor or an electro-optical sensor), movement of the stylus off its shelf on the PDA, detection of movement when the user picks up the PDA, or detection of proximity when the user's hand is in proximity to the PDA. The next step 304 is a "reset" step in which all pixels of the display are alternately driven to their black and white states. It has been found that in at least some electro-optic media, such "flickering" of pixels is necessary to ensure accurate grey scale states in the sequential writing of images on a display. It has also been found that typically at least 5 flashes are required (one flash is calculated for each successive black and white state), or in some cases more. The more times the flicker is, the more time and energy is spent in this step and thus the longer the time has to elapse before the user can see the desired image on the display. It is therefore desirable to keep the number of flashes as small as possible consistent with accurate gray state delivery in subsequently written images. At the end of the reset step 304, all pixels of the display are in the same black or white state.
The next step is a write or "send image" step in which controller 16 sends signals to row and column drivers 22 and 24 (fig. 1 and 2), respectively, in the manner already described, thereby writing the desired image on the display. Because the display is bi-stable, rewriting need not be done immediately once the image is written, and thus the controller typically interrupts the writing of the display by the row and column drivers after the image is written, by setting a blanking signal (e.g., signal BL high in fig. 7).
The controller now enters a decision loop consisting of steps 308, 310 and 312. In step 308, controller 16 checks whether computer 12 (FIG. 1) requests the display of a new image. If so, the controller continues to erase the image written to the display in step 306 in an erase step 314, thereby returning the display substantially to the state reached at the end of the reset step 304. From the erase step 314, the controller returns to step 304, reset as previously described, and continues to write a new image.
If in step 308 there is no new image to be written to the display, the controller proceeds to step 310 where it is decided when the image that has been held on the display exceeds a predetermined period. As is known to those skilled in the display art, an image written onto a bi-stable medium is not maintained indefinitely and the image fades away (i.e., the contrast is reduced). Furthermore, in some types of electro-optic media, in particular electrophoretic media, there is often a trade-off between writing speed and bistability of the media, since a medium that remains bistable for hours and days has a substantially longer writing time than a medium that remains bistable for only a few seconds or minutes. Thus, although there is no need to continuously rewrite the electro-optic medium as is the case in an LCD, in order to provide an image with good contrast, the image needs to be refreshed at intervals of, say, a few minutes. Thus, in step 310 the controller determines whether the time that has elapsed since the image was written in step 306 exceeds a predetermined refresh interval, and if so, the controller proceeds to the erase step 314 and then to the reset step 304, resetting as described above, and continuing to rewrite the same image onto the display.
(the program shown in FIG. 8 may be changed to use both local and global overwrites, as described in more detail below; if so, step 310 is instead determined to require either local or global overwrite; if in the transformed program, the program determines at step 310 that the predetermined time has not expired, no action will be taken; however, if the predetermined time has expired, step 310 will not invoke erasure and overwrite of an image immediately, but will merely set a flag (a term generally in computer terms) indicating that updates to the next image are globally more effective than locally
If it is determined in step 310 that the refresh interval has not been exceeded, the controller proceeds to step 312 where it is determined whether it is time to turn off the display and/or image source. To conserve power in the portable device, the controller will not allow a single image to be refreshed indefinitely, terminating the process after an extended period of no operation as shown in fig. 8. Accordingly, at step 310 the controller determines whether a predetermined "off period (greater than the refresh interval mentioned above) has expired since a new image (rather than an overwrite of a previous image) was written to the display, and if so, the routine terminates as indicated at 314. Step 314 may include powering down the image source. That is, the user may also access the slowly fading image on the display after such a procedure terminates. If the off period has not been exceeded, control returns from step 312 to step 308.
Various possible waveforms for performing the look-up table method of the present invention will be described by way of example only. However, first, some general rules will be described as waveforms used in the present invention.
The waveforms of a bi-stable display exhibiting the memory effect described above can be divided into two main categories, namely compensated and uncompensated. In the compensated waveform, the entire pulse is fine-tuned to account for any memory effect in the pixel. For example, a pixel undergoing a series of gray levels 1-3-4-2 transitions will receive a slightly different pulse for the 4-2 transition than a pixel undergoing a 1-2-4-2 transition. Such pulse compensation may be performed by adjusting the pulse length, voltage or by other changes in the v (t) profile of the pulse. In the uncompensated waveform, there is no provision to take into account any prior state information (other than the initial state). In the uncompensated waveform, all pixels undergoing a 2-4 transition receive exactly the same pulse. In order for the uncompensated waveform to work successfully, one of two conditions must be met. One is that the electro-optic medium must exhibit no memory effect during its switching operation, or that any memory effect in the pixel must be effectively removed with each transition.
In general, uncompensated waveforms are best suited for systems that can only perform coarse pulse resolution. Such as a display with a three level driver or a display with only 2-3 bit voltage conversion capability. The compensation waveform requires fine pulse adjustment and is not possible with such a system. Clearly, while a coarse pulse system is best limited to uncompensated waveforms, a system with fine pulse adjustment can achieve both types of waveforms.
The simplest uncompensated waveform is a 1-bit normal image stream (1-bit GIF). In a 1-bit GIF, the display transition smoothly goes from one pure black-and-white image to the next. The conversion rule for such a sequence can be simply as follows: if the image switches from white to black, a pulse I is supplied. If it switches from black to white, a pulse of opposite polarity, -I is supplied. If the image remains in the same state, no pulse is supplied to the pixel. As previously stated, the mapping of pulse polarity to voltage polarity of the system may depend on the corresponding function (function) of the material.
Another non-compensated waveform capable of producing gray scale images is a non-compensated n-prepulse slide show (n-PPSS). This uncompensated slide show waveform has three basic parts. First, the pixel is erased to a unique optical state, usually white or black. The pixel is then driven back or forth between two optical states, typically white and black. Finally, the pixel is addressed to a new optical state, which may be a multi-gray state. The final (or write) pulse is referred to as the addressing pulse, while the other pulses (the first (or erase) pulse and the intermediate (or blanking) pulse) are collectively referred to as the pre-pulse. This type of waveform will be described below with reference to fig. 9 and 10.
The pre-pulse slide show waveform can be divided into two basic forms, with odd pre-pulses and with even pre-pulses. For the case of an odd pre-pulse, the erase pulse will be equal in pulse and opposite in polarity to the immediately preceding write pulse (see also FIG. 9 and description below). In other words, if a pixel is written from black to gray, the erase pulse will return the pixel to the black state. In the case of an even pre-pulse, the erase pulse will have the same polarity as the immediately preceding write pulse and the sum of the pulses of the preceding write pulse and the erase pulse will be equal to the pulse required for a complete transition from black to white. In other words, if a pixel is written from black in the case of an even pre-pulse, it must be erased to white.
After the erase pulse, the waveform includes zero or an even number of blanking pulses. These blanking pulses are typically equal but opposite polarity pulses arranged such that the first pulse is of opposite polarity to the erase pulse. These pulses are typically equal to the entire black-white pulse, but this is not a requirement. It may also be desirable only that the pulse pairs have equal but opposite polarity pulses-which may be pairs of widely varying pulses linked together, i.e., + I, -I, +0.1I, -0.1I, +4I, -4I.
The last pulse supplied is the write pulse. The selection of the pulse is based only on the desired optical state (independent of the current state, or any previous state). Typically, the pulse will increase or decrease monotonically with the grey state value, but this is not required. Since the waveform is specifically designed for coarse pulse system use, the selection of a write pulse will typically include a mapping of a set of desired gray states over a small number of possible pulse selections, e.g., 4 gray states over 9 possible applied pulses.
Examination of the even or odd version of the uncompensated n pre-pulse slideshow waveform will show that the write pulses always start in the same direction, i.e. from black or from white. This is an important feature of such waveforms. Since the principle of the uncompensated waveform is that the pulse length cannot be accurately compensated to ensure that the pixels reach the same optical state, it cannot be expected that a uniform optical state will be reached when starting from the opposite extreme optical state (black or white). Thus, for one of these forms, which may be labeled "from black" and "from white", there are two possible polarities.
One major drawback of this type of waveform is that there is greatly amplified optical flicker between the images. As described below with reference to fig. 9 and 10, this can be improved by shifting the update order by half the pixel at superframe time and interlacing the pixels at high resolution. Possible patterns include every other row, every other column, or an inspection plate pattern. Note that this does not mean that opposite polarities are used, i.e. "from black" versus "from white", since this would result in a mismatch of the grey levels on adjacent pixels. Alternatively, this can be achieved by delaying the start of the update by one "superframe" (a group of frames equal to the maximum length of a black-white update) for half the pixels (i.e., the first group of pixels completes the erase pulse, then the second group of pixels begins the erase pulse and the first group of pixels begins the first blanking pulse). This requires an additional superframe for the total update time, taking into account synchronization.
It will be seen first that the ideal method of the present invention will be referred to as "normal gray scale image streaming", in which the controller arranges the writing of each image so that each pixel can be switched directly from its initial gray scale level to its final gray scale level. However, in practice, ordinary grayscale image streams face problems from error accumulation. The pulses applied to any given greyscale transition must necessarily differ from the theoretical requirement due to the fact that unavoidable variations in voltage output are caused, for example, by drivers, variations in the production of the electro-optic medium in thickness, etc. The average error over each transition is assumed to be expressed as a difference term of + -0.2L of the display between the theoretical and actual frame reflectivities. Over 100 consecutive transitions, the pixels will exhibit an average shift from their desired 2L state; such shifts are evident to the ordinary viewer in certain types of images. To avoid this problem, the drive scheme used in the present invention needs to be arranged such that any given pixel can only undergo a predetermined maximum number of greyscale transitions once before going through one extreme optical state (black or white). These extreme optical states act as "fences" after a particular pulse has been applied to the electro-optic medium, which cannot become darker or whiter. Thus, the next transition always from an extreme optical state can start with an accurately known optical state, effectively canceling out any previously accumulated error. Various techniques for minimizing the optical effect of such a segment of pixels through extreme optical states are described below.
First, a simple driving scheme useful in the present invention will now be described with reference to a simple two-bit grayscale system with black (0), dark gray (1), bright gray (2) and white (3) optical states, using pulse width modulation techniques and a transition look-up table as shown in table 4 below to achieve the transition.
TABLE 4
| Conversion | Pulse of light | Conversion | Pulse of light |
| 0-0 | 0 | 0-0 | 0 |
| 0-1 | n | 1-0 | -n |
| 0-2 | 2n | 2-0 | -2n |
| 0-3 | 3n | 3-0 | -3n |
Where n is a number depending on the particular display and-n represents a pulse having the same length as pulse n but opposite polarity. It is further assumed that at the end of the reset pulse 304 in fig. 8, all pixels of the display are black (level 0). Thus, as described below, all transitions occur through an intermediate black state, only those transitions to or from the gray state are active. In this way the required look-up table size is significantly reduced, and it is clear that the scale factor in terms of the reduced look-up table size increases with the number of grey levels of the display.
Figure 9 shows a pixel transition associated with the drive scheme of figure 8. At the beginning of the reset step 304, the pixel is at some random gray level. In a reset step 304 the pixel is alternately driven to three black states and two intermediate white states, ending up in its black state. The appropriate gray scale for the first image, assuming 1 level, is then written to the pixel at 306. The pixel is held at this level for a period of time during the display of the same image; the length of the display period is greatly reduced in fig. 9 for ease of illustration. A new image is sometimes written to at which point the pixel returns to black (level 0) in the erase step 308 and then undergoes 6 reset pulses of alternating white and black in a second reset step designated 304 'so that at the end of the reset step 304' the pixel has returned to the black state. Finally, in a second writing state, designated 306', the pixel is written with a gray level appropriate for the second image, assuming 2 levels.
Many different drive schemes in fig. 9 are of course possible. One useful variation is shown in fig. 10. Steps 304, 306 and 308 shown in fig. 10 are the same as in fig. 9. However, in step 304 ' 5 reset pulses are used (obviously a different odd pulse can also be used), so that at the end of step 304 ' the pixel is in the white state (3 levels), while in a second writing step 306 ' the writing of the pixel is effected from the white state instead of from the black state as described in fig. 9. So that successive images are written alternately from the black state and the white state of the pixel.
In a variation of the further drive scheme shown in fig. 9 and 10, the erasing step 308 is not performed in terms of driving the pixel black, but white (3 levels). Then an even number of reset pulses are applied to the pixel terminals in the white state in a reset step, and a second image is written from the white state. As in the driving scheme of fig. 10, successive images are written alternately from the black state and the white state of the pixel in this scheme.
It will be appreciated that in all of the previous arrangements, the number and duration of the reset pulses may vary depending on the characteristics of the electro-optic medium used. Similarly, the pulses applied to the pixels can be varied by voltage modulation rather than pulse width modulation.
The black and white flashes represented on the display in the reset step of the above-described driving scheme are of course visible to the user and may be disliked by many users. To reduce the visual effect of such a reset step, it may be convenient to divide the pixels of the display into two (or more) groups and to provide different types of reset pulses to the different groups. More particularly, if it is desired to drive any given pixel alternating between black and white using a reset pulse, it may be convenient to divide the pixels into at least two groups and arrange the drive scheme so that one group of pixels is driven white while the other group is driven black. The spatial distribution of the two groups is provided carefully and the pixels are small enough that the user goes through a reset step like a greyscale interval (preferably some slight flicker) on the display, which is usually not obtrusive to a series of black and white flashes.
For example, in one form of such a "two-bank reset" step, pixels in odd columns may be assigned to one "odd" bank, while pixels in even columns are assigned to a second "even" bank. The odd pixels may then use the drive scheme shown in fig. 9, while the even pixels may use a variation of this drive scheme, wherein the pixels are not driven to the black state but to the white state in the erasing step. Both groups of pixels then experience an even number of reset pulses during the reset step 304', so that the reset pulses for both groups are substantially 180 degrees out of phase and the display appears gray throughout the reset step. Finally, during the second image writing of step 306', the odd pixels are driven from black to their final state and the even pixels are driven from white to their final state. To ensure that each pixel is reset in the same way over a long period of time (and so that the method of resetting does not cause any noise on the display), it is advantageous for the controller to switch the drive scheme between successive images so that each pixel is written alternately from a black and white state to its final state when writing a series of new images to the display.
It is clear that a similar scheme can be used in which the first group is formed by pixels in odd rows and the second group is formed by pixels in even rows. In yet another similar driving scheme, the first group comprises pixels in odd columns and odd rows, and even columns and even rows, and the second group comprises pixels in odd columns and even rows, and even columns and odd rows, such that the two groups are arranged in a checkerboard pattern.
Instead of or in addition to dividing the pixels into two groups and arranging the reset pulses in one group 180 degrees out of phase with the reset pulses in the other group, the pixels may be divided into groups using different reset pulses differing in number and pulse frequency. For example, one set may use a 6 pulse reset sequence as shown in fig. 9, while a second set may use a similar 12 pulses of twice the frequency. In a more sophisticated scheme the pixels may be divided into four groups, the first and second groups using a 6-pulse scheme but 180 degrees out of phase with each other, and the third and fourth groups using a 12-pulse scheme but 180 degrees out of phase with each other.
Another scheme for reducing the adverse effects of the reset step will now be described with reference to fig. 11A and 11B. In this scheme, the pixels are again divided into two groups, a first (even) group according to the drive scheme shown in fig. 11A and a second (odd) group according to the drive scheme shown in fig. 11B. Also in this scheme, the division of all gray levels in between black and white into a first group adjacent dark gray levels next to the black levels and a second group adjacent light gray levels next to the white levels is the same for both groups of pixels. It is desirable, but not necessary, to have the same number of gray levels in both sets; if there are an odd number of gray levels, then intermediate levels can be randomly assigned to either group. For ease of illustration, fig. 11A and 11B show that the present driving scheme provides 8-level gray scale display, designated as a level of 0 (black) to 7 (white); gray levels 1, 2 and 3 are dark gray levels, and gray levels 4, 5 and 6 are light gray levels.
In the driving scheme of fig. 11A and 11B, the transition between the grays is controlled according to the following rule:
(a) first, the even group of pixels, in the transition to the dark gray level, the last pulse supplied is always a pulse tending to white (i.e., a pulse having a polarity tending to drive the pixel from its black state to its white state), whereas in the transition to the light gray level, the last pulse supplied is always a pulse tending to black.
(b) Secondly, the odd groups of pixels, in the transition to the dark grey level, are supplied with the last pulse always a black going pulse, and in the transition to the light grey level, are supplied with the last pulse always a white going pulse.
(c) In all cases, the black-tending pulse may only follow the white-tending pulse when the white state has been attained, and the white-tending pulse may only follow the black-tending pulse when the black state has been attained.
(d) Even numbered pixels are not driven from a dark gray level to black by a single black-tending pulse, nor are odd numbered pixels driven from a light gray level to white by a single white-tending pulse. (obviously, in both cases, only one final white-going pulse can be used to obtain the white state, and only one final black-going pulse can be used to obtain the black state.)
The application of these rules allows the transition between grey scales to be achieved using the largest of three consecutive pulses. For example, fig. 11A shows an even pixel subjected to a transition from black (0 level) to gray level 1. This is achieved with a single whitening pulse, designated 1102 (which is of course represented as a positive slope in fig. 11A). The pixel is then driven to grey level 3. Since gray level 3 is a dark gray level, it must be realized by a whitening pulse according to rule (a), and thus the 1/3 level transition can be controlled by a separate whitening pulse 1104, which has a pulse difference with pulse 1102.
The pixel is now driven to grey level 6. Since this is a bright grey level, it must be achieved by a pulse tending to black according to rule (a). Thus, the application of rules (a) and (c) requires that the 3/6 level transition be achieved by a two pulse sequence, a first white-tending pulse 1106 which drives the pixel to white (7 levels), followed by a second black-tending pulse 1108 which drives the pixel from 7 levels to the desired 6 levels.
The pixel is then driven to grey level 4. Since this is a bright gray level, a 6/4 transition is achieved by a single dark pulse 1110 according to a very similar theory as described above for a 1/3 transition. The next transition is to level 3. Since this is a dark gray level, the 4/3 level transition is controlled by a two pulse sequence, a first tending black pulse 1112 driving the pixel to black (level 0), followed by a second tending white pulse 1114 driving the pixel from level 0 to the desired 3 levels, according to a very similar theory as described above for the 3/6 level transition.
The final transition shown in fig. 11A is a transition from 3 levels to 1 level. Since this is a dark grey level, it must be achieved by a trending white pulse according to rule (a). Thus, applying rules (a) and (c), the 3-level/1-level transition must be controlled by a three-pulse sequence comprising a first white-tending pulse 1116 to drive the pixel to white (7 levels), a second black-tending pulse 1118 to drive the pixel to black (0 levels), and a third white-tending pulse 1120 to drive the pixel from black to the desired 1-level state.
FIG. 11B shows the sequence of 0-1-3-6-4-3-1 gray states as for the even pixel in FIG. 11A for an odd pixel implementation. However, it can be seen that the pulse sequences used are very different. Rule (b) requires a dark gray level of 1 to be achieved by a going black pulse. Thus, a 0-1 transition is effected by a first white-tending pulse 1122 driving the pixel to white (7 levels), followed by a black-tending pulse 1124 driving the pixel from 7 levels to the desired 1 level. The 1-3 transition requires a three pulse sequence, a first black-tending pulse 1126 to drive the pixel to black (level 0), a second white-tending pulse to drive the pixel to white (level 7), and a third black-tending pulse 1130 to drive the pixel from level 7 to the desired level 3. The next transition is to 6 levels of bright gray levels, which is accomplished by a whitening pulse according to rule (b), and the 3/6 level transition is accomplished by a two pulse sequence including a whitening pulse 1132 driving the pixel to black (level 0) and a whitening pulse 1134 driving the pixel to the desired 6 levels. The 6 level/4 level transition is achieved by a three pulse sequence, a whitening pulse 1136 to drive the pixel to white (7 levels), a whitening pulse 1138 to drive the pixel to black (0 levels), and a whitening pulse 1140 to drive the pixel to the desired 4 levels. The 4-level/3-level transition is achieved by a two-pulse sequence comprising a white-going pulse 1142 to drive the pixel to white (7 levels), followed by a black-going pulse 1144 to drive the pixel to the desired 3 levels. Finally, the 3/1 transition is effected by a single blackening pulse 1146.
It can be seen from fig. 11A and 11B that this drive scheme ensures that each pixel follows a "zig-zag" shape in which each pixel transitions from black to white without changing direction (although it is clear that the pixel will be stationary for one or a short or long period at any intermediate grey level) and then from white to black without changing direction. Thus, the above rules (c) and (d) may be replaced by the following single rule (e):
(e) once a pixel has been driven from one extreme optical state (i.e. white or black) to the opposite extreme optical state by a pulse of a single polarity, the pixel no longer receives a pulse of the opposite polarity until it reaches the opposite extreme optical state as described previously.
Whereby the driving scheme ensures that a pixel only undergoes a number of transitions equal to (N-1)/2 times at most, where N is the number of grey levels before being driven to an extreme optical state; this prevents slight errors in the individual transitions (caused for example by unavoidable small fluctuations in the voltage supplied by the driver) from infinitely converging on the extent to which the series of distortions of a gray-scale image is noticeable to the viewer. Furthermore, the present drive scheme is designed so that even and odd pixels always reach a given intermediate grey level from opposite directions, i.e. the final pulse in the sequence tends to be white in one case and black in the other. If an active area of the display, which maintains a substantially equal number of even and odd pixels, is written to a single gray level, the "reverse" feature can minimize flicker in that area.
For other drive schemes that drive pixels in two discrete groups, for similar reasons as described above, when the zig-zag drive scheme of fig. 11A and 11B is implemented, attention is focused on arranging the pixels in even and odd groups. This arrangement is required to ensure that any substantially adjacent area of the display maintains a substantially equal number of odd and even pixels, and that the maximum size of the same set of adjacent blocks of pixels is small enough to be indistinguishable by an ordinary observer. As described above, arranging two pixel groups in a checkerboard pattern can satisfy these requirements. The random screen technique can also be used for the arrangement of two groups of pixels. However, in a sawtooth drive scheme, the use of a checkerboard pattern tends to increase the power consumption of the display. In any given column in such a pattern, adjacent pixels will belong to opposite groups, and in adjacent regions of basic size where all pixels are subject to the same grey level transition (as is the case), adjacent pixels tend to require pulses of opposite polarity at any given time. When each new row is to be written, supplying pulses of opposite polarity to successive pixels in any one column requires discharging and recharging the column (source) electrodes of the display. It is well known to those skilled in the art of driving active matrix displays that the discharge and recharge of the column electrodes is a major factor in the power consumption of the display. Thus, the checkerboard distribution tends to increase the power consumption of the display.
A reasonable compromise between power consumption and the desire to avoid large adjacent regions of the same set of pixels is to assign the pixels in each set to a rectangle in which the pixels all continue in the same column for only a few pixels along that column. With such an arrangement, when the overwrite regions have the same grey level, the discharge and recharge of the column electrodes is only required when switching from one rectangle to another. The ideal rectangle is 1 x 4 pixels and is arranged such that rectangles in adjacent columns do not end in the same row, i.e. rectangles in adjacent columns will have different "phases". The assignment of rectangles to phases in each column can be done in a random or cyclic way.
One benefit of the sawtooth drive scheme shown in fig. 11A and 11B is that any single color region of the image can be simply updated with a single black to white or white to black pulse as part of the overall update of the display. The maximum time taken to rewrite such a monochrome area is only half of the maximum time required to rewrite an area requiring an inter-gray conversion, and the use of this feature facilitates quick updating of image features such as characters entered by a user, pull-down menus, and the like. The controller may check whether any inter-gray transition is required for the image update; if not, the image area that needs to be rewritten can be rewritten using the fast monochrome update mode. Thus, the user may have a fast update of the input characters, pull-down menus and other user interactive features of the display that seamlessly coordinates with the slower update of the normal grayscale image.
As described in the above co-pending application serial numbers 09/561424 and 09/520743, in many electro-optic media, particularly particle-based electrophoretic media, it is desirable that the drive scheme used to drive such media have a Direct Current (DC) balance over the entire extended period, the algebraic sum of the currents through a particular pixel should be zero or as close to zero as possible, and the drive scheme of the present invention is designed to address this criterion. More specifically, the look-up table used in the present invention is designed such that any switching sequence that starts or ends at one extreme optical state (black or white) of the pixel should be dc balanced. From the foregoing, it can be seen first that such dc balancing may not be achieved because the current required to switch between any particular grey levels through the pixel is thus substantially constant due to the pulse. However, this is only true to a first approximation, and it has been found empirically that, at least in the case of particle-based electrophoretic media (and equally true in the case of other electro-optical media), the effect of providing a 50ms pulse of, say, 5 intervals to a pixel is different from the effect of providing a 250ms pulse of the same voltage to a pixel. Thus, there is some flexibility in the current through the pixel to achieve a given conversion, which can be used to assist in achieving dc balance. For example, a look-up table used in the present invention may store a number of pulses for a given transition along with the total current value provided by each of those pulses, and the controller may maintain a register for each pixel that stores the algebraic sum of the pulses provided to the pixel since some previous time (e.g., since the pixel was maintained in the black state). When a particular pixel is to be driven from a white or grey state to a black state, the controller may examine the register associated with that pixel, determine the current required for dc balancing throughout the sequence of transitions from the original black state to the upcoming black state, and select one of the stored pulses required for the white/grey to black transition, any of which will reduce the corresponding register to exactly zero, or at least to as small a remainder as possible (where the corresponding register will hold the remainder and add it to the current supplied in the next transition). It can be seen that repeated application of this process can achieve accurate long-term dc balance for each pixel.
It should be noted that the sawtooth drive scheme shown in fig. 11A and 11B is well suited to use with such a dc balancing technique because it ensures that only a significant number of transitions can be passed between successive passes of any given pixel through the black state, and that on average, in one-half of its transitions, there is actually one pixel that will pass through the black state.
The detrimental effect of the reset step can be further reduced by using local rather than global updating, i.e. by only overwriting parts of the display that only change between successive images, the parts to be overwritten can be selected on a "local" area or pixel-by-pixel basis. For example, when, for example, in a diagram illustrating part motion in a mechanical device or a diagram for contingency reconstruction, it is not difficult to find a series of images in which a relatively small object moves relative to a larger stationary background. To use local updates, the controller needs to compare the final image with the initial image and determine which areas differ between the two images and thus need to be overwritten. The controller may determine one or more local regions, typically rectangular regions having edges arranged in a grid of pixels holding pixels to be updated, or may determine only individual pixels that need to be updated. Thus any of the drive schemes already described can be used to update only the local areas or individual pixels determined to need overwriting. Such a local update scheme can substantially reduce the power consumption of the display.
The drive scheme described above can be varied in a number of ways depending on the characteristics of the particular electro-optic display being used. For example, many reset steps may be omitted in the above-described driving scheme in some cases. For example, if the electro-optic medium used is bistable for very long periods (i.e. the grey levels written to the pixels change only very slowly in time) and the impulse change required for a particular transition at that period for a pixel already in its initial grey state is not large, the look-up table may be arranged to effect the transition between grey states directly without intervention to return to a black or white state, after a substantial period, resetting of the display only being performed when a gradual "shift" of the pixels from their nominal grey levels has caused a significant error in the current image. Thus, for example, if a user uses the display of the present invention as an e-book reader that can display many screens of information before a reset of the display is required, it has been found empirically that with appropriate waveforms and drivers, as many as 1000 screens of information can be displayed before a reset is required, so that a reset is not actually required during normal reading of the e-book reader.
It will be readily apparent to those skilled in the display arts that the individual devices of the present invention can be used in a variety of conditions to accommodate a number of different drive schemes. For example, since in the drive schemes shown in fig. 9 and 10 the reset pulse consumes a small fraction of the total power consumption of the display, the controller may be provided with a first drive scheme in which the display is reset at frequency intervals, thus minimising greyscale errors, and a second scheme in which the display is reset only at longer intervals, thus allowing for larger errors but reducing power consumption. Switching between the two schemes can be done manually or by means of external parameters; for example, if the display is used as a laptop computer, the first drive scheme may be used when the computer is operating on mains power, and the second scheme may be used when the computer is operating on internal battery power.
From the foregoing it can be seen that the present invention provides a driver for operation control of an electro-optic display which is well suited to the features of particle-based bistable electrophoretic displays and similar displays.
From the foregoing it can be seen that the present invention provides a method and controller for controlling the operation of an electro-optic display which allows accurate control of grey scales without requiring inconvenient blinking of the entire display to its extreme states at frequency intervals. The invention also allows accurate control of the display despite variations in temperature and its operating time, while reducing power consumption of the display. These advantages can be achieved inexpensively since the controller can be constructed from commercially available components.
In the residual voltage method of the present invention, it is desirable that the measurement of the residual voltage is performed by a high-impedance voltage measuring device, such as a Metal Oxide Semiconductor (MOS) comparator. When the display is a matrix display having small pixels, e.g. 100 Dots Per Inch (DPI), where each pixel has 10-4Square inch or about 6 x 10mm2When the resistance of such a single pixel reaches 1012The comparator needs to have a very low input current at the magnitude of ohm. However, suitable comparators are readily available commercially; for example, the INA111 chip is suitable as a Texas instrument with an input current of only about 20 pA. (technically, this integrated circuit is an instrumentation amplifier, but it can be used as a comparator if its output goes into a schmitt trigger.) for displays with large individual pixels, such as large direct drive displays for signage (specified below), each pixel may have an area of a few square centimeters, the requirements for a comparator are not very high, and essentially all commercial FET input comparators can be used, such as LF311 comparators from national semiconductor corporation.
It will be readily apparent to those skilled in the art of electronic displays that, for cost and other reasons, mass-produced electronic displays will typically have drivers in the form of application-specific integrated circuits (ASICs), and in this type of display, the comparators will typically be provided as part of the ASICs. Although this approach may require a feedback circuit provided in the ASIC, it has the advantage of making the supply and oscillating parts of the ASIC simpler and smaller in area. This approach also makes the drive portion of the ASIC simpler and smaller in area if a 3-level normal image stream driver is required. Thus, this approach generally reduces the cost of the ASIC.
Conveniently, the drive pulse is provided by a driver which provides a drive voltage to electrically short or float the pixel. When such a driver is used, the pixel is addressed, electrically shorted, and then floated for each addressing period in which the dc balance correction is implemented. (as used herein, the term "addressing period" is used in its conventional meaning in electro-optic display technology to refer to the total period required to change from a first image to a second image on the display. As indicated above, an individual addressing period may comprise many complete display scans due to the relatively low switching speed of an electrophoretic display, typically on the order of tens to hundreds of milliseconds.) after a short delay time, a comparator is used to measure the residual voltage across the pixel and determine whether it is positive or negative in sign. If the residual voltage is positive, the controller may slightly extend the period of the negative address pulse (or slightly increase its voltage) in the next address period. But if the residual voltage is negative, the controller may slightly extend the period of the positive address voltage pulse (or slightly increase its voltage) in the next address period.
Thus, the residual voltage method of the present invention places the electro-optic medium in a switched feedback loop, driving the residual voltage toward zero by adjusting the length of the addressing pulses. As the residual voltage approaches zero, the media exhibits desirable performance and improved lifetime. In particular, the use of the invention allows an improved control of the grey scale. As previously indicated, it has been seen that the gray level achieved in an electro-optic display is a function of a starting gray level and the supplied pulse, as well as the previous state of the display. It is believed (although the invention is not limited by such belief) that one of the reasons for this "history" effect on gray scale is that the residual voltage acts on the electric field experienced by the electro-optic medium; the actual electric field that affects the state of the medium is the sum of the actual voltages applied by the electrodes and the residual voltage. The control of the residual voltage according to the invention thus ensures that the electric field experienced by the electro-optical medium corresponds exactly to the voltage supplied via the electrodes, thereby allowing an improved control of the grey scale.
The remnant voltage method of the present invention is particularly useful in displays of the so-called "direct drive" type, which are divided into a series of pixels each provided with separate electrodes, the display further comprising switching means for independently controlling the voltage applied to each separate electrode. Such direct drive displays are useful for the display of text or other limited character sets such as a number of digits, and are described, inter alia, in the aforementioned international application publication No. 00/05704. However, the remnant voltage method of the present invention may also be used in other types of displays, such as active matrix displays having a matrix of transistors, where at least one transistor is associated with each pixel of the display. The gate lines of the driving Thin Film Transistors (TFTs) used in such active matrix displays connect the pixel electrodes to the source electrodes. The residual voltage is smaller than the gate voltage (the absolute value of the residual voltage generally does not exceed 0.5V), so the gate drive voltage will always turn on the TFT. The source line may then be electrically floating and connected to a MOS comparator, thereby allowing the residual voltage of each individual pixel of the active matrix display to be read out.
It is noted that although the residual voltage across a pixel of an electrophoretic display may be closely related to the degree of current flow through a pixel that is dc balanced, zero residual voltage does not necessarily imply perfect dc balance. However, from a practical point of view, this makes little difference, since it is the residual voltage that is not the history of the dc balance that is responsible for the adverse effects indicated here.
It will be readily apparent to those skilled in the display art that since the purpose of the residual voltage method of the present invention is to reduce residual voltage and dc imbalance, the method need not be used in every addressing period of the display, it is provided at a suitable frequency to prevent dc imbalance from developing over a long period of time on a particular pixel. For example, if a display is required that uses "refresh" or "blanking" pulses at intervals, then all pixels are driven to the same display state, typically one of the extreme display states, in the refresh or blanking pulses (or, more generally, all pixels will be driven first to one of the extreme display states and then to the other extreme display state), this method of the invention may be used only in the refresh or blanking pulses.
Although the residual voltage method of the present invention has been generally described in the context of an encapsulated electrophoretic display, the present method may also be used in non-encapsulated electrophoretic displays, as well as other types of displays, such as electrochromic displays that exhibit residual voltage.
From the foregoing, it can be seen that the remnant voltage method of the present invention provides a method for driving electrophoretic and other electro-optic displays that reduces the cost of the equipment required to ensure pixel dc balance of the display, while providing increased display lifetime, enhanced operating window and long-term display optical performance.