HK1093254A1 - Methods for driving electro-optic displays - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an appropriate method for driving an electro-optic display. <P>SOLUTION: The electro-optic display having a plurality of pixels divided into a plurality of groups is driven by selecting each of the plurality of groups of pixels in succession and applying to each of the pixels in the selected group either a drive voltage or a non-drive voltage, the scanning of all the groups of pixels being completed in a first frame period; repeating the scanning of the groups of pixels during a second frame period, and interrupting the scanning of the groups of pixels during a pause period between the first and second frame periods, this pause period being not longer than the first or second frame period. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

Method for driving electro-optic display

The present invention relates to a method for driving an electro-optic display. The method of the present invention is particularly, although not exclusively, designed for driving a bi-stable electrophoretic display.

This application is closely related to International patent application PCT/US02/37241 (publication No. WO03/044765) and PCT/US2004/10091, and the following description will assume that the reader is familiar with the contents of these documents.

The term "electro-optic" as used in materials or displays, and used herein in its conventional sense in imaging technology, refers to a material having first and second display states differing in at least one optical characteristic, the material changing from its first display state to its second display state by the application of an electric field to the material. Although this optical property is typically a color perceptible to the human eye, it may be another optical property, such as light transmittance, reflectance, luminescence, or in the case of a display designed for machine reading, a false color in the sense that the reflectance of electromagnetic wavelengths outside the visible range changes.

The term "grey state" is used herein in its conventional sense in imaging technology to refer to an intermediate state between two extreme optical states of a pixel, not necessarily meaning a black-to-white transition between the two extreme states. For example, several patents and published patent applications relate to electrophoretic displays described below in which the extreme states are white and deep blue, so that the intermediate "grey state" is actually light blue. Indeed, as already indicated, the transition between the two extreme states may not be a color change at all.

The terms "bistable" and "bistable" are used herein in the conventional sense of imaging technology to include displays whose display elements have first and second display states differing in at least one optical characteristic such that any given element, after being driven with an addressing pulse of finite duration, assumes either its first or its second display state, which state, after the addressing pulse has ended, will continue to change the state of that display element by at least a multiple, for example at least four times, the minimum duration of the addressing pulse required. It is shown in international patent application WO 02/079869 that certain particle-based electrophoretic displays can have a gray scale which is stable not only in their extreme black and white states, but also in intermediate gray states, as is the case with certain other types of electro-optic displays. This type of display may be referred to as "multi-stable" where appropriate, rather than bi-stable, although for convenience the term "bi-stable" may be used herein to encompass both bi-stable and multi-stable displays.

The term "pulse" is used herein in its conventional sense in imaging technology to refer to the integral of voltage over time. However, some bistable electro-optic media function as charge transducers and so with such media an alternative definition of pulse can be used, namely the integral of current over time (equal to the total charge applied). The proper definition of the pulse should depend on whether the medium functions as a voltage-time pulse transducer or a charge pulse transducer.

As described in the above-mentioned WO03/044765 and PCT/US2004/10091, several types of electro-optic displays are known, such as rotating bichromal film types, for example, as described in US patent nos. 5,808,783; 5,777,782, respectively; 5,760,761, respectively; 6,054,071, respectively; 6,055,091; 6,097,531, respectively; 6,128,124, respectively; 6,137,467, respectively; and 6,147,791, and electrochromic types; see, e.g., O' Regan, B, et al, Nature 1991, 353, 737; and Wood, d., Information Display, 18(3), 24(March 2002). See also adv.mater, 2002, 14(11), 845 by Bach, u. For another example, U.S. patent No.6,3019038, international patent application publication No. wo 01/27690 and in U.S. patent application 2003/0214695 also describe the Nanochromic film.

Another type of electro-optic display, which has been the subject of intense research and development for many years, is the particle-based electrophoretic display. Many patents and applications assigned to or filed in the name of the Massachusetts Institute of Technology (MIT) and e.ink Corporation describe such displays; see, for example, U.S. Pat. Nos. 5,930,026; 5,961,804; 6,017,584; 6,067,185, respectively; 6,118,426, respectively; 6,120,588; 6,120,839, respectively; 6,124,851, respectively; 6,130,773, respectively; 6,130,774, respectively; 6,172,798; 6,177,921, respectively; 6,232,950, respectively; 6,249,721, respectively; 6,252,564, respectively; 6,262,706, respectively; 6,262,833; 6,300,932, respectively; 6,312,304, respectively; 6,312,971, respectively; 6,323,989, respectively; 6,327,072, respectively; 6,376,828, respectively; 6,377,387, respectively; 6,392,785, respectively; 6,392,786, respectively; 6,413,790, respectively; 6,422,687, respectively; 6,445,374, respectively; 6,445,489, respectively; 6,459,418, respectively; 6,473,072, respectively; 6,480,182, respectively; 6,498,114, respectively; 6,504,524; 6,506,438, respectively; 6,512,354, respectively; 6,515,649, respectively; 6,518,949, respectively; 6,521,489, respectively; 6,531,997, respectively; 6,535,197, respectively; 6,538,801, respectively; 6,545,291, respectively; 6,580,545, respectively; 6,639,578, respectively; 6,652,075, respectively; 6,657,772, respectively; 6,664,944, respectively; 6,680,725, respectively; 6,683,333, respectively; 6,704,133, respectively; 6,710,540, respectively; 6,721,083, respectively; 6,724,519, respectively; and 6,727,881; and U.S. patent application publication No. 2002/0019081; 2002/0021270, respectively; 2002/0053900, respectively; 2002/0060321, respectively; 2002/0063661, respectively; 2002/0063677, respectively; 2002/0090980, respectively; 2002/0106847, respectively; 2002/0113770, respectively; 2002/0130832, respectively; 2002/0131147, respectively; 2002/0145792, respectively; 2002/0171910, respectively; 2002/0180687, respectively; 2002/0180688, respectively; 2002/0185378, respectively; 2003/0011560, respectively; 2003/0011868, respectively; 2003/0020844, respectively; 2003/0025855, respectively; 2003/0034949, respectively; 2003/0038755, respectively; 2003/0053189, respectively; 2003/0102858, respectively; 2003/0132908, respectively; 2003/0137521, respectively; 2003/0137717, respectively; 2003/0151702, respectively; 2003/0189749, respectively; 2003/0214695, respectively; 2003/0214697, respectively; 2003/0222315, respectively; 2004/0008398, respectively; 2004/0012839, respectively; 2004/0014265, respectively; 2004/0027327, respectively; 2004/0075634, respectively; and 2004/0094422; and international patent application publication No. wo 99/67678; WO 00/05704; WO 00/38000; WO 00/38001; WO 00/36560; WO 00/67110; WO 00/67327; WO 01/07961; WO 01/08241; WO 03/092077; WO 03/107315; WO 2004/017035; and WO 2004/023202.

Many of the above patents and applications recognize that the walls surrounding discrete microcapsules in an encapsulated electrophoretic medium, which comprises a plurality of discrete electrophoretic droplets and a continuous phase of polymeric material, may be replaced with a continuous phase, resulting in a so-called "polymer dispersed electrophoretic display", and that the discrete electrophoretic droplets may be viewed as capsules or microcapsules in such a polymer dispersed electrophoretic display, although there is no capsule film associated with each individual droplet; see, for example, 2002/0131147, supra. Accordingly, for the purposes of this application, such polymer dispersed electrophoretic media are considered to be a subclass of encapsulated electrophoretic media.

A related type of electrophoretic display is also known as a "microcell electrophoretic display". In a microporous electrophoretic display, the charged particles and suspending liquid are not encapsulated, but rather are held in a plurality of cavities formed in a carrier medium, typically a polymer film. See, for example, international patent application publication No. wo 02/01281 and U.S. patent application publication No.2002/0075556, both assigned to SipixImaging, Inc.

Although electrophoretic media tend to be opaque (because, for example, in many electrophoretic media, the particles substantially block visible light from passing through the display), and operate in a reflective mode, many electrophoretic displays can be operated in a so-called "shutter mode" in which one display state is substantially opaque and the other is light-transmissive. See, for example, U.S. patent nos. 6,130,774 and 6,172,798, and U.S. patent No.5,872,552; 6,144,361, respectively; 6,271,823, respectively; 6,225,971, respectively; and 6,184,856. Dielectrophoretic (Dielectrophoretic) displays, which are similar to electrophoretic displays but rely on changes in the electric field strength, can be operated in a similar mode; see U.S. patent No.4,418,346. Other types of electro-optic displays may also operate in shutter mode.

The bi-or poly-stable variation characteristics of particle-based electrophoretic displays, and other electro-optic displays exhibiting similar variation characteristics (such displays may be referred to hereinafter for convenience as "impulse-driven displays"), are in sharp contrast to conventional liquid crystal ("LC) displays. Twisted nematic liquid crystals do not act as bi-or poly-stable but as voltage transducers, so that the application of a given electric field to a pixel of such a display produces a particular grey scale at that pixel, independent of the grey scale previously presented by that pixel. In addition, LC displays are driven in only one direction (from opaque or "dark" to transparent or "bright"), and the transition from the lighter state to the darker state is made by reducing or eliminating the electric field. Finally, the grey scale of a pixel of an LC display is not sensitive to the polarity of the electric field, but only to its amplitude, and indeed, for technical reasons, commercial LC displays typically reverse the polarity of the driving electric field at frequent intervals. In contrast, bistable electro-optic displays, most closely, function as pulse transducers, such that the final state of a pixel depends not only on the applied electric field and the time at which the field is applied, but also on the state of the pixel prior to application of the field.

It may be thought first of all that the ideal way to deal with such an impulse driven electro-optic display would be to say a so-called "general gray scale image stream", in which the controller configures each writing of an image such that each pixel transitions directly from its initial gray scale to its final gray scale. However, writing an image on an impulse driven display inevitably has some errors. Some such errors encountered in practice include:

(a) previous state dependencies; for at least some electro-optic media, the pulse required to switch a pixel to a new optical state depends not only on the current and the desired optical state, but also on the previous optical state of the pixel.

(b) Residence time dependence; for at least some electro-optic media, the pulse required to switch a pixel to a new optical state depends on the time it takes for the pixel to be in its different optical state. The precise nature of this dependence is not readily apparent, but in general, the longer the pixel stays in its current optical state, the more pulses are required.

(c) Temperature dependence; the pulse required to switch the pixel to a new optical state is strongly dependent on temperature.

(d) Dependence on humidity; for at least some types of electro-optic media, the pulse required to switch a pixel to a new optical state depends on the ambient humidity.

(e) Mechanical uniformity; the pulse required to switch the pixel to a new optical state can be affected by mechanical variations in the display, for example variations in the thickness of the electro-optic medium or associated lamination adhesive. Other types of mechanical non-uniformity may be due to unavoidable variations between different manufacturing batches of media, manufacturing tolerances, and material variations.

(f) A voltage error; the actual pulses applied to the pixels will inevitably differ slightly from the theoretically applied pulses, since a slight error in the voltage supplied by the driver is unavoidable.

As described in the aforementioned WO03/044765 and PCT/US2004/10091, the general flow of grayscale images has problems caused by "error accumulation" phenomena, which can produce, on certain types of images, grayscale deviations that are noticeable to the ordinary observer. This error accumulation phenomenon applies to all types of errors listed above. As described above at 2003/0137521, it is possible to compensate for such errors, but only to a limited degree of accuracy. Thus, the typical gray scale image stream must be very closely controlled with respect to the applied pulses to achieve good results, and it has been empirically found that the typical gray scale image stream is not feasible on commercially available displays in the current state of the art electro-optic displays.

Almost all electro-optic media have a built-in reset (error limiting) mechanism, that is, their extreme (typically black and white) optical states, acting as an "optical fence". After a particular pulse is applied to a pixel of an electro-optic display, the pixel cannot become whiter (or blacker). For example, in an encapsulated electrophoretic display, after a particular pulse is applied, all of the electrophoretic particles are forced against each other or against the capsule wall and can no longer move, thus creating a restrictive optical state or optical fence. Because in such a medium there is electrophoretic particle size and charge distribution, some particles hit the pen before others, creating a "soft pen" phenomenon, thus reducing the required pulse accuracy when the final optical state of a transition approaches the extreme black and white states, and thus reducing the required optical accuracy when the final optical state of a transition approaches the extreme black and white states, and dramatically increasing the required optical accuracy when the end point of the transition approaches the middle of the optical range of the pixel.

Different types of drive schemes using optical fences for electro-optic displays are known. For example, figures 9 and 10 of the above-mentioned WO03/044765 and the related description describe a "slide show" drive scheme in which the entire display is driven to two optical fences before any new images are written. Such a slide show drive scheme produces accurate gray scale, but when driven into the optical enclosure, the flickering of the display annoys the viewer. It has been proposed (see us patent No.6,531,997) to use a similar driving scheme in which only pixels whose optical state needs to be changed in the new image are driven into the optical fence. However, this type of "limited slide show" drive scheme is even more annoying to the viewer, if any, because the overall (solid) flicker of a normal slide show drive scheme is replaced by a flicker dependent on the picture, where the features of the old and new pictures on the screen flicker in opposite colors before writing the new picture.

Clearly, a purely gray scale image stream driving scheme cannot look to avoid gray scale errors with the optical fence, as any given pixel may experience an infinite number of changes in gray scale without touching any optical fence in such a driving scheme.

In one aspect, the present invention seeks to provide a method for achieving gray scale control in an electro-optic display that achieves gray scale stabilization similar to that achieved by a slide-down drive scheme, but without the annoying flicker problems of slide-down drive schemes. The preferred method of the present invention provides the viewer with a visual experience similar to that provided by a purely generic grayscale video stream driving scheme.

In other aspects, the invention seeks to provide methods of achieving fine control of grey scale in displays driven by pulse width modulation.

When driving an active matrix display having a bistable electro-optic medium for writing gray scale images thereon, it is desirable to be able to apply a precise amount of pulse to each pixel in order to achieve accurate control of the displayed gray scale. The driving method used may rely on modulation of the voltage applied to each pixel and/or modulation of the "width" (duration) of the applied voltage. Pulse width modulation is commercially attractive because of the relatively high cost of voltage modulated drivers and their associated power supplies. However, in an active matrix display using such pulse width modulation, conventional drive circuitry only allows a single voltage to be applied to any given pixel during a scan of any one of the matrices. Thus, pulse width modulated driving of an active matrix display is performed by scanning the matrix a number of times, the drive voltages applied during zero, a few or all scans being dependent on the required change in grey scale for a particular pixel. Each scan can be viewed as a frame drive waveform and the complete addressing pulse is a super-frame formed by a number of successive frames. It should be noted that for one row address time during each scan, although the drive voltage is only applied to any particular pixel electrode, during the time between successive selections of the same row the drive voltage remains on that pixel electrode, only decaying slowly so that the pixel is driven between successive selections of the same row.

As already noted, each row of the matrix needs to be individually selected during each frame, so that the frame rate cannot exceed about 50-100 Hz in practice for high resolution displays (e.g., 800 x 600 pixel displays); thus, each frame typically lasts 10 to 20 ms. Frames of such length make it difficult to finely control the grey scale of many rapidly switching electro-optic media, for example, some encapsulated electrophoretic media are substantially in their extreme optical states (about 30L per transition)^*Unit) is completed within about 100ms, while a 20ms frame for such a medium corresponds to about 6L^*The gray scale of the unit is shifted. Such a shift is too large for accurate control of the gray scale; for human eye about 1L^*The unit gray scale is sensitive and is only equivalent to 6L^*Controlling the pulse on a unit scale is likely to cause visible distortions such as "ghosting" caused by the previous state dependence of the electro-optic medium. More specifically, ghosting may be experienced because, as discussed, in some of the above patents and applications, the variation in applied pulse gray scale is not linear, and the total pulse required for any particular gray scale change may vary with the time of application of the pulse and the gray scale in between. For example, in a simple 4 gray (2-bit) display with gray scales 0 (black), 1 (dark gray), 2 (light gray), and 3 (white), driven by a simple pulse width modulation drive scheme, these non-linearities may cause the actual gray scale achieved after a nominally 0-2 transition to be different from the gray scale achieved after a nominally 1-2 transition, while producing highly undesirable gray scalesThe visual distortion of (1). The present invention provides a method for achieving fine control of gray scale in a display driven by pulse width modulation, thereby avoiding the above-mentioned problems.

Accordingly, in one aspect, the present invention provides an electro-optic display having at least one pixel capable of achieving any one of at least four different gray levels, including two extreme optical states. The method comprises the following steps:

displaying a first image on the display; and

rewriting the display to display a second image thereon,

wherein, during rewriting of the display, any pixel having a number of transitions exceeding a predetermined value, the predetermined value being at least one, without reaching an extreme optical state, is driven to at least one extreme optical state before the pixel is driven to its last optical state in the second image.

For convenience, this process may be referred to as the "limited conversion process" of the present invention.

In one form of the finite transition method, the rewriting of the display is performed such that once a pixel is driven from one extreme optical state to the opposite extreme optical state by a pulse of one polarity, the pixel does not receive a pulse of the opposite polarity until the opposite extreme optical state is reached.

In addition, in the finite transition method, a predetermined value (predetermined number of transitions) is not more than N/2, where N is the total number of gradations that one pixel can display. The limited transition method may be performed using a three level driver, i.e. the rewriting of the display may be performed by applying any one or more of the voltages-V, 0 and + V to the or each pixel. Alternatively, the finite transition method may be DC balanced, i.e. the rewriting of the display may be such that the integral of the applied voltage over time is bounded for any series of transitions experienced by a pixel.

With the limited transitions of the invention, the rewriting of the display can be done such that the pulses applied to a pixel during a transition depend only on the initial and final grey levels of the transition. Alternatively, the method may be designed to take into account other states of the display, as described in more detail below. In a preferred form of the finite transition method, for at least one transition from grey scale R2 to grey scale R1 experienced by at least one pixel, a sequence of pulses of the form:

TM(R1，R2)IP(R1)-IP(R2)TH(R1，R2)

where "ip (rx)" represents the correlation value from a pulse potential matrix having one value for each gray level, and TM (R1, R2) represents the correlation value from a transfer matrix having one value for each R1/R2 combination. (for convenience, this type of pulse sequence may be abbreviated hereinafter as an "x/Δ IP/x" sequence.) such an-x/Δ IP/x sequence may be applied to all transitions where the initial and final gray levels are different. In addition, in such an-x/IP/x sequence, the last "x" segment may take more than half of the maximum update time. The TM (R1, R2) or x values may be chosen such that the sign of each value depends only on R1; in particular, these values may be chosen to be positive for one or more bright gray levels and negative for one or more dark gray levels such that gray levels outside the two extreme optical states approach from a direction closer to the extreme optical states.

The above-described-x/Δ IP/x sequence may contain additional pulses. In particular, such a sequence may include an additional pair of pulses of the form [ + y ] [ -y ], where y is a negative or positive pulse value, the [ + y ] and [ -y ] pulses being inserted into the-x/Δ IP/x sequence. The sequence may also include a second additional pair of pulses of the form [ + z ] [ -z ], where z is a pulse value different from y and may be negative or positive, the [ + z ] and [ -z ] pulses being inserted into the-x/Δ IP/x sequence. The-x/Δ IP/x sequence may also include a period in which no voltage is applied to the pixel. The "no voltage" period may occur between two elements of the-x/Δ IP/x sequence, or within a single element thereof. The-x/Δ IP/x sequence may include two or more "no voltage periods.

With the above-described-x/Δ IP/x sequence, the display may comprise a plurality of pixels divided into a plurality of groups, and the transition may be made by: (a) selecting each of a plurality of phase pixel groups, and applying either a driving voltage or a non-driving voltage to each pixel in the selected group, the scanning of all the pixel groups being completed in a first frame period; (b) repeating the scanning of the group of pixels during a second frame period; and (c) interrupting the scanning of the group of pixels during a pause period between the first and second frame periods, the pause period being no longer than the first and second frame periods.

In the limited transition method, the rewriting of the display can be performed such that the transition to a given gray level is always performed by a last pulse of the same polarity. In particular, the grey levels outside the two extreme optical states may approach from the direction of the closer extreme optical state.

The invention also provides a method of driving an electro-optic display having a plurality of pixels divided into groups. The method comprises the following steps:

(a) selecting each of a plurality of successive groups of pixels and applying either a drive voltage or a non-drive voltage to each pixel in the selected group, the scanning of all the groups of pixels being completed in a first frame period;

(b) repeating the scanning of the group of pixels during a second frame period; and

(c) the scanning of the group of pixels is interrupted during a pause period between the first and second frame periods, which pause period is not longer than the first or second frame period.

For convenience, this method may be referred to hereinafter as the "interrupted scanning" method of the present invention.

In such an interrupted scanning method, the first and second frame periods are generally equal in length. The length of the pause period may be a fraction (sub-multiple) of the length of one of the first and second frame periods. The interrupt scanning method may include a plurality of pause periods; thus, the method may comprise scanning the group of pixels during at least first, second and third frame periods and interrupting the scanning of the group of pixels during at least first and second pause periods between successive frame periods. The first, second and third frame periods may be substantially equal in length, with the total length of the pause periods being equal to one frame period or one frame period minus one pause period. In general, in the interrupted scanning method, the pixels are arranged in a matrix having a plurality of rows and a plurality of columns, wherein each pixel is defined by the intersection of a given row and a given column, and each group of pixels comprises one row or one column of the matrix. The interrupted scan method is preferably DC balanced, i.e. the scanning of the display is preferably performed such that the integral of the applied voltage over time is bounded for any series of transitions experienced by a pixel.

In other aspects, the invention provides a method for driving an electro-optic display having a plurality of pixels driven with a pulse width modulated waveform capable of applying a plurality of different pulses to each pixel. The method comprises the following steps:

(a) storing data indicating whether applying a given pulse to a pixel produces a gray scale above or below a desired gray scale;

(b) detecting when two adjacent pixels both require the same gray level; and

(c) the pulses applied to the two pixels are adjusted so that one pixel is below the desired gray level while the other pixel is above the desired gray level.

For convenience, this method may be referred to hereinafter as the "balanced gray scale method" of the present invention.

In this approach, the pixels may be divided into two groups such that each pixel has at least one opposite group of neighbors, and different drive schemes are used for the two groups.

As described above, each of the methods of the present invention may be applied to any of the above-described types of electro-optic media. Thus, the method of the present invention may be used with electro-optic displays comprising electrochromic or rotating bichromal thin film electro-optic media, encapsulated electrophoretic media, or microporous electrophoretic media. Other types of electro-optic media may also be used.

FIGS. 1A and 1B illustrate two portions of the finite transition drive scheme of the present invention;

FIG. 2 illustrates a preferred-x/Δ IP/x sequence that can be used in the method of the invention;

FIG. 3 schematically illustrates how the waveforms shown in FIG. 2 may be modified to include an additional pair of drive pulses;

FIG. 4 illustrates the modification of the waveforms of FIG. 2 to produce a waveform in the manner illustrated in FIG. 3;

FIG. 5 illustrates modifying the waveforms of FIG. 2 to produce a second waveform in the manner illustrated in FIG. 3;

FIG. 6 schematically illustrates how the waveforms shown in FIG. 5 may be further modified to include an additional pair of drive pulses;

FIG. 7 illustrates the modification of the waveforms of FIG. 5 to produce a waveform in the manner illustrated in FIG. 6;

8-10 illustrate three modifications of the waveforms shown in FIG. 2 to include a period of zero voltage;

from the foregoing it will be apparent that the present invention provides several different improvements in the driving method of electro-optic displays. In the following description, the various improvements provided by the present invention will generally be described separately, although those skilled in the imaging arts will appreciate that in practice a single display may utilize more than one of these primary aspects; for example, displays using the limited transition method of the present invention may also utilize an interrupt scan method. In addition, because the improvements provided by the present invention can be applied to the various methods of driving electro-optic displays described in WO03/044765 and PCT/US2004/10091 above, the following description will assume (the reader) familiarity with the basic driving methods and associated descriptions shown in FIGS. 1-10 of WO 03/044765. Specifically, FIGS. 9 and 10 of this application describe a so-called uncompensated n-pre-pulse slide show (n-PP SS) waveform having three elementary sections. First, the pixel is erased to a uniform optical state, typically either white or black. The pixel is then driven back and forth between two optical states, typically again white and black. Finally, the pixel is brought to a new optical state, which may be one of several gray states. The last (or write) pulse is called the address pulse and the other pulses (the first (or erase) pulse and the intermediate (or blank) pulse) are collectively called the pre-pulse.

One major drawback of this type of waveform is that it has large amplitude optical flicker between several images. This can be improved by shifting the update sequence by the superframe time for half the pixels and interleaving the pixels at high resolution as discussed in WO 0303044765 with reference to figures 9 and 10 thereof. Possible patterns include every other row, every other column or a checkerboard pattern. Note that this is not meant to use the opposite polarities, i.e., "from black" and "from white", as this would result in non-matching gray levels on adjacent pixels. But half of the pixels can be completed by delaying the start of the update by one "super frame" (a set of several frames equivalent to the maximum length of a black-and-white update) (i.e., the first group of pixels completes the erase pulse and the second group of pixels starts the erase pulse when the first group of pixels starts the first blank pulse). This would require a superframe of total update time added to allow them to synchronize.

Limited conversion method of the present invention

To avoid the above-mentioned flicker problem of the drive scheme shown in figures 9 and 10 of WO03/044765, while also avoiding the general gray scale image flow problem discussed earlier, it is preferred according to the present invention that the limited transition method be configured such that any given pixel can only undergo a predetermined maximum number (at least one) of gray scale transitions before passing through an extreme optical state (black or white). A transition from the extreme optical state starts with an accurately known optical state, virtually canceling any previously accumulated error. Different techniques for minimizing such optical effects of passing pixels through extreme optical states, such as flicker of the display, are discussed in WO 03/044765.

The black and white flashes appearing on the display during the reset step of such a driving scheme as described above are of course visible to the user and annoying to many users. To reduce the visual effect of such a reset step, it is convenient to divide the pixels of the display into two (or more) subgroups and to apply different types of reset pulses to the different subgroups. More specifically, if it is necessary to use a reset pulse that drives any given pixel alternately black and white, it is convenient to divide the pixels into at least two sub-groups and to configure the drive scheme so that one group of pixels is driven to white while another group of pixels is driven to black. Assuming careful selection of the spatial distribution of the two subgroups and sufficiently small pixels, the user will experience the reset step on the display as if it were a greyscale interval (perhaps with some slight snowfall) and such a greyscale interval is generally less objectionable than a series of black and white flashes.

For example, in one form of such a "two-set reset" step, pixels on odd-numbered columns may be grouped into an "odd" set, while pixels on even-numbered columns are grouped into a second "even" set. The odd pixels can then be driven to the black state during the erase step using the drive scheme shown in figure 9, while the pixels can be driven to the white state during the erase step using a variation of this drive scheme. Both groups of pixels are then subjected to an even number of reset pulses during the reset step, so that the reset pulses for the two subgroups are substantially 180 degrees out of phase, and the display appears grey throughout the reset step. Finally, during the second image writing of this step, the odd pixels are driven from black to their final state, while the even pixels are driven from white to their final state. To ensure that each pixel is reset in the same manner for a long period of time (and thus the reset manner does not introduce any distortion on the display), the controller preferably switches the drive scheme between successive images so that each pixel is alternately written from black and white states to its final state as a series of new images are written to the display.

Obviously, a similar scheme may be used, wherein pixels of odd-numbered lines form the first group and pixels of even-numbered lines form the second group. In another similar driving scheme the first group comprises odd columns and odd rows and even columns and even rows of pixels and the second group comprises odd columns and even rows and even columns and odd rows such that the two groups are arranged in a checkerboard fashion.

Instead of (or in addition to) dividing the pixels into two subgroups and arranging the reset pulses in one group 180 degrees out of phase with the other group, the pixels may be divided into subgroups using different reset steps, differing in the number and frequency of pulses. For example, one set may reset the sequence with six pulses, while a second set may use a similar sequence with 12 pulses at twice the frequency. In a more elaborate scheme, the pixels may be divided into four groups, with a six-pulse scheme for the first and second groups, the phases being 180 degrees from each other, while the third and fourth groups use a 12-pulse scheme, but the phases being 180 degrees from each other.

Another approach to reducing the flicker problem, in accordance with the limited transition method of the present invention, can be performed using a driving scheme that allows any given assumption to be non-zero but a limited number of successive gray states before touching an optical fence. In such a driving scheme, when the display is adapted to display a new image thereon, any pixel that has undergone a number of transitions exceeding a predetermined value without touching one of the extreme optical states is driven to at least one of the extreme optical states before the pixel is driven to its final optical state. In a preferred form of such a drive scheme, a pixel driven to one of the extreme optical states is driven to the extreme optical state whose grey scale is closer after the transition, assuming of course that the required optical state is not one of the extreme optical states. In addition, in a preferred form of such a drive scheme using a look-up table as described above, the maximum number of transitions a pixel is allowed to undergo without touching the optical fence (the extreme optical state) is set equal to the number of previous optical states considered at the transition matrix; such an approach does not require additional control logic or memory.

The driving method that limits the maximum number of transitions before touching an optical fence does not have to significantly increase the time it takes to complete the rewriting of the display. For example, consider a four gray scale (2-bit) display, where a transition from white to black or vice versa takes 200 milliseconds, so that a typical gray scale video stream drive scheme takes this time to completely overwrite the display. The only situation that requires modification for the transition on such a display is when the pixel is repeatedly switched between two central grey levels. If the number of transitions such a pixel switches between two central grey levels exceeds the predetermined number, the limited transition method of the present invention requires that the next switching be performed through the optical fence (extreme optical state). It has been found that in such a case, it takes 70 milliseconds to go to the optical pen while it takes about 13 milliseconds to go to the gray level later, resulting in a total transition time of about only 200 milliseconds. Therefore, the present limited transition method does not need to extend the transition time at all compared to the general gray-scale image stream.

A limited transition driving method that reduces the annoying effects of the reset step will now be described with reference to fig. 1A and 1B. In this scheme the pixels are again divided into two sub-groups, the first (even) group according to the driving scheme shown in fig. 1A and the second (odd) group according to the driving scheme shown in fig. 1B. In addition, in this scheme, the total gray levels between black and white are divided into a first group of adjacent dark gray levels adjacent to the black level and a second group of adjacent light gray levels adjacent to the white level, thus being divided as two pixel groups. Preferably, but not essentially, there are the same number of grey levels in both subgroups; if there are odd gray levels, the center level can be arbitrarily given to any group. For ease of illustration, fig. 1A and 1B show drive schemes applied to an eight-level gray scale display, the levels being labeled 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 fig. 1A and 1B graphics driving scheme, the grayscale to grayscale transition is handled according to the following rules:

(a) in the first even group of pixels, the last pulse applied is always a white-going pulse (i.e. its polarity tends to drive the pixel from its black state to its white state) in the transition to the dark grey level, whereas in the transition to the bright grey level, the last pulse applied is always a black-going pulse;

(b) in the second odd-numbered pixel group, the last pulse applied is always a black pulse in the transition to the dark gray scale, and the last pulse applied is always a white pulse in the transition to the light gray scale;

(c) in all cases, a black-tending pulse may only follow a white-tending pulse after the white state is reached, and a white-tending pulse may only follow a black-tending pulse after the black state is reached; and

(d) even pixels cannot be driven from a dark grey level to black with a single black-tending pulse, nor can odd pixels be driven from a bright grey level to white with a single white-tending pulse.

(obviously, in all cases, the white state can only be achieved with the last white-going pulse, while the black state can only be achieved with the last black-going pulse.)

Applying these rules allows each gray-to-gray transition to be made with up to three consecutive pulses. For example, fig. 1A shows that even-numbered pixels undergo a transition from black (level 0) to gray level 1. This is achieved with a single whitening pulse (represented by the positive gradient in fig. 1A, of course) indicated at 1102. The pixel is then driven to grey level 3. Since gray level 3 is a dark gray level, according to rule (a), it must be reached by a white going pulse, and thus a level 1/level 3 transition can be handled by a single white going pulse 1104, which has a different pulse than pulse 1102.

The pixel is now driven to grey level 6. Since this is a bright grey scale, it must be achieved by a black going pulse according to rule (a). Accordingly, applying rules (a) and (c) requires that a level 3/level 6 transition be made through a sequence of two pulses, namely a first white-going pulse 1106 which drives the pixel to become white (level 7) followed by a second black-going pulse 1108 which drives the pixel from level 7 to the desired level 6.

The pixel is then driven to grey level 4. Since this is a bright gray level, the level 6/level 4 transition is made by a single black going pulse 1110, with a parameter just similar to that used for the level 1/level 3 transition discussed earlier. The next transition is to level 3. Since this is a dark grey level, performed by a parameter just similar to that used for the level 3/level 6 transition, the level 4/level 3 transition is processed by two pulse sequences, namely a first black-tending pulse 1112, which drives the pixel black (level 0), followed by a second white-tending pulse 1114, which drives the pixel from level 0 to the desired level 3.

The final transition shown in fig. 1A is from level 3 to level 1. Since level 1 is a dark gray level, it must be approached by a going white pulse according to rule (a). Accordingly, applying rules (a) and (c), the level 3/level 1 transition must be processed through a sequence of three pulses, including a first white-going pulse 1116 that drives the pixel to white (level 0), a second black-going pulse 1118 that drives the pixel to black (level 7), and a third white-going pulse 1120 that drives the pixel from black to the desired level 1 state.

FIG. 1B shows odd pixels undergoing the same 0-1-3-6-4-3-1 gray state sequence as the even pixels of FIG. 1A. However, it will be seen that the pulse sequences used are very different. Rule (b) requires level 1, a dark gray level, to be approached by the going black pulse. Thus, the 0-1 transition is made by driving the pixel to a first white-going pulse 1122 at level 7 followed by a black-going pulse 1124 driving the pixel from level 7 to the desired level 1. The 1-3 transition requires three pulse sequences, a first black-tending pulse 1126 to drive the pixel to black (level 0), a second white-tending pulse 1128 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 level 6, which is a bright gray level, which is approached by a white-going pulse according to rule (b), and the level 3/level 6 transition is made by a sequence of two pulses, including a black-going pulse 1132 driving the pixel to black (level 0), and a white-going pulse 1134 driving the pixel to the desired level 6. The level 6/level 4 transition is made by a sequence of three pulses, namely a white-going pulse 1136 to drive the pixel to white (level 7), a black-going pulse 1138 to drive the pixel to black (level 0) and a white-going pulse 1140 to drive the pixel to the desired level 4. The level 4/level 3 transition is made by a sequence of two pulses, including a white-going pulse 1142 driving the pixel to white (level 7), followed by a black-going pulse 1144 driving the pixel to the desired level 3. Finally, the level 3/level 1 transition is made by a single black going pulse 1146.

As will be seen from fig. 1A and 1B, this drive scheme ensures that each pixel follows a sawtooth pattern in which the pixel goes from black to white without changing direction (obviously, although the pixel may stay at any intermediate grey level for a short or long period), and then goes from white to black without changing direction. Thus, the above rules (c) and (d) may be replaced by a single rule (e) as follows:

(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 one polarity, the pixel may not receive a pulse of the opposite polarity until it has reached the above-mentioned opposite extreme optical state.

Thus, the driving scheme is a "barrier stable gray" or "RSGS" driving scheme. Such an RSGS driving scheme is a special case of a limited transition driving scheme, which ensures that a pixel can only undergo transitions at most a number of transitions equal to N/2 (or more precisely (N-1)/2), where N is the total number of grey levels that can be displayed, without requiring the transition to occur through an optical fence. Such a driving scheme avoids slight errors in the individual transitions (e.g., caused by the inevitable small fluctuations in the voltages applied by the drivers), accumulating indefinitely until severe distortion of the grayscale image appears to the viewer. In addition, the drive scheme is designed so that the even and odd pixels always approach a given intermediate grey scale from opposite directions, i.e. in one case the last pulse of the sequence is a white-going pulse and in the other case a black-going pulse. If a large area of the display contains substantially equal numbers of even and odd pixels to write a single gray scale, the "reverse direction" feature minimizes flicker in that area.

When implementing the zigzag drive scheme of fig. 1A and 1B, the pixels are divided into two discrete subgroups, and careful attention should be paid to the arrangement of the pixels in the even and odd subgroups for reasons similar to those discussed above in relation to the other drive schemes. Such an arrangement will preferably ensure that any substantially continuous display area will contain substantially equal numbers of odd and even pixels, and that the maximum size of a continuous block of pixels of the same group is sufficiently small to be imperceptible to an ordinary observer. As already discussed, the checkerboard pattern configuration of the two pixel groups meets these requirements. In addition, stochastic screening techniques may be used to configure the pixels of the two subgroups.

However, in this sawtooth drive scheme, the use of a checkerboard pattern tends to increase the power consumption of the display. At any given column of such a pattern, adjacent pixels will belong to opposite subgroups, and in a continuous region where all pixels are subject to the same greyscale transition (an unusual case) of considerable size, adjacent pixels will tend to require pulses of opposite polarity at any given time. The application of pulses of opposite polarity to successively occurring pixels on any column requires the column electrodes (sources) of the display to be discharged and recharged each time a new row is written. It is well known to the skilled person that in the driving of active matrix displays, the discharge and recharge of the column electrodes is a major factor in the power consumption of the display. Thus, the checkerboard arrangement tends to increase the power consumption of the display.

A reasonable compromise between power consumption and the desire to avoid large contiguous areas of pixels of the same group is to give each group of pixels a rectangle in which the pixels all lie in the same column, but extend a few pixels along that column. With such an arrangement, when overwriting an area with the same grey level, the column electrodes must only be discharged and recharged when moving from one rectangle to the next. The rectangles are preferably 1 x 4 pixels and are arranged such that the rectangles in adjacent columns do not end in the same row, i.e. the rectangles in adjacent columns have different "phases". The rectangles in each column give each phase that can be done randomly or in a cyclic manner.

One advantage of the sawtooth drive scheme shown in fig. 1A and 1B is that any single color region of the image is simply updated with a single pulse, either black to white or white to black, as part of the overall update of the display. The maximum time for rewriting such a monochrome area is only half of the maximum time required for area rewriting required for a grayscale-to-grayscale transition, and this feature can be advantageously used to quickly update image features such as user character entry, drop-down menus, etc. The controller may check whether a request for an image requires a gray-to-gray transition; if not, the image area to be rewritten can be rewritten in the quick monochrome updating mode. Thus, the user can quickly update the display's entered characters, drop-down menus, and other user interactive features, seamlessly overlaying the slower updates of the generic grayscale image.

A limited transition drive scheme does not necessarily require the use of a counter to measure the number of transitions experienced by each pixel of the display and does not preclude the use of a drive scheme (such as the cyclic RSGS drive scheme already described with reference to fig. 1A and 1B) that requires some transitions to occur through the optical fences even if the predetermined number of transitions has not been reached, provided that the algorithm used to determine the manner in which the transitions are made does not allow any pixel to undergo more than the predetermined number of transitions without touching the optical fences. In addition, it will be appreciated that the number of transitions a given pixel undergoes without touching the optical fence does not have to be made each time the display image is rewritten, especially if the display is updated at frequent intervals. For example, the check may only be done on alternate updates, assuming that all pixels either exceed a predetermined number of transitions, or may exceed that number after the next update is driven into the optical fence.

Another preferred limited conversion method of the present invention will now be described, although by way of illustration only. The preferred method is for operating a four gray scale (2-bit) active matrix display that uses a transition matrix that takes into account only the initial and final gray levels of the transition to be made (labeled "R2" and "R1", respectively) and does not take into account additional prior states. The display controller is a three-level Pulse Width Modulation (PWM) controller capable of applying-V, 0 or + V to each pixel electrode with respect to a common front electrode held at 0 (level).

The display controller includes two RAM image buffers. A buffer ("A") stores the current image on the display. Typically, the controller is in a sleep mode, saving data in the RAM and keeping the display driver inactive. The bi-stability of the electro-optic medium preserves the same image on the display. When a video update command is received, the controller loads a new video into the second buffer ("B"). The controller then looks (in flash) at the multi-frame drive waveform for each pixel of the display in accordance with the last state R1 required for that pixel (from buffer "B") and the current, initial state R2 for each pixel (from buffer "a").

The data in the flash file is organized into an array of three-dimensional voltage values V (R1, R2, frames), where as already indicated R1 and R2 are each integers from 1 to 4 (corresponding to four available gray levels), and "frame" is the number of frames, i.e., the number of associated frames in the superframe for each transition. In general, a superframe may be 1 second long while each frame takes 20ms, so that the number of frames may be in the range of 1 to 50. Thus, the array has 800 entries, 4 × 4 × 50. Since each entry in the array must be able to represent any of the voltage values-V, 0 and + V, two bits are typically used to store each voltage value (array value).

It is clear that since there can be any of the three possible voltage values at the top of each 800 array, there will be a large number of possible arrays (waveforms) that are too large to be exhaustive to search. In theory, there is 3⁸⁰⁰Or about 5X 10³⁸¹A plurality of possible arrays; since the universe has about 1078 atoms, the average life of a human is 10⁹Second, the practical capability is at least 200 orders of magnitude, and is not exhaustive of the search. Fortunately, the existing knowledge of the changing characteristics of electro-optic displays, particularly the need for DC balancing thereon, imposes additional constraints on the possible waveforms and allows for optimization or optimizationThe search for near-optimal waveforms is limited to the extent feasible.

As discussed in the aforementioned U.S. patent nos. 6,504,524 and 6,531,997 and the aforementioned WO03/044765, it is known that most, if not all, electro-optic media require Direct Current (DC) balanced waveforms, otherwise deleterious effects may occur. When using an unbalanced DC waveform, such effects may include damage to the electrodes and over several L^*Long-term drift of grey states in the range of units (period over one hour). It therefore appears that a scheme that strives to use DC balanced drive waveforms should be followed in three thoughts.

It may be appreciated from the above that such a DC-balancing approach may not be achievable because the pulse through the pixel, and hence the current, required for any particular grey-to-grey transition is substantially constant. However, this is only true to the greatest approximation and it has been found empirically that the effect of applying five 50 ms-spaced pulses to a pixel is different from applying a 250 ms pulse of the same voltage, at least in the case of particle-based electrophoretic media (as is the case with other electro-optic media). Accordingly, there is some flexibility in the current flowing through a pixel to achieve a given transition, and this flexibility can be used to assist in achieving DC balance. For example, the look-up table may store a number of pulses for a given transition, together with the value of the total current provided by each of the pulses, and the controller may maintain a register for each pixel configured to store the algebraic sum of the pulses applied to the pixel since some previous time (e.g. since the pixel was maintained in a 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 of the overall sequence of transitions from the previous black state to the upcoming black state, and select one of a plurality of stored pulses for the white/grey to black transition that will not accurately reduce the associated register to zero, but will at least reduce the remainder to as small as possible (in which case the associated register will maintain the value of the remainder and add it to the current applied during the subsequent transition). Clearly, repeated application of this process can achieve accurate long-term DC balance for each pixel.

The precisely defined waveform of the DC balance in the waveform must be examined. To determine whether a waveform is DC balanced, a resistive model of the electro-optic medium is typically used. Such a model is not completely accurate, but may be assumed to be sufficiently accurate for the present purpose. With such a model, the characteristic that defines a DC-balanced waveform is that the integral of the applied voltage over time (applied pulse) is bounded. Note that this definition requires that the integral be "bounded" and not "zero". To illustrate this, consider a monochrome addressing waveform that drives a white to black transition using a 300ms x-15V rectangular pulse and drives a black to white transition using a 300ms x 15V rectangular pulse. The waveform is apparently DC balanced, but the integral of the applied voltage at each instant is not zero; the integral varies between 0 and ± 4.5V-sec. However, to the extent that the integral is bounded, the waveform is DC balanced; for example, the integral never reaches 9 or 18V-sec.

To further consider the DC-balanced waveform, it is advisable to make certain definitions of terms. The term "pulse" has been defined as a definite integral (in M-sec) of the applied voltage over time during a specific interval, typically an addressing pulse or pulse element. The term "pulsed potential" will be used to refer to the sum of all pulses applied to the display from an arbitrary starting point (the starting point of a series of transitions generally considered). At this starting point, the pulse potential is arbitrarily set to zero, and the pulse potential rises and falls when a pulse is applied.

With these terms, DC balance is defined as the waveform being DC balanced if and only if the pulse potential is bounded. Having a bounded pulsed potential means one must be able to say that the pulsed potential will be in each of a limited number of possible situations.

For time-independent controllers (i.e., controllers whose pulse shape is affected only by the initial and final states of the transition under consideration, and not by dwell times, temperature, and other factors, such as the R1/R2 controller described above), to indicate that a waveform is DC balanced, it must be possible to demonstrate that the pulse potential will be bounded after each transition of any infinitely long sequence of optical states. A sufficient condition for such a demonstration is that the pulsed potential can be expressed as a function of a fixed number of previous states and that a DC balanced working concept is provided for the electro-optic display controller, i.e. the pulsed potential can be expressed as a function of a limited number of previous and current optical states. Note that the pulse potential of any pixel of the display is constant from the end of one image update to the start of another image update, since no voltage is applied during this period.

The controller applies fixed pulses (pulses determined by the data in the flash memory described above) for each combination of a (finite) number of previous states, and these fixed pulses may be enumerated. To enumerate them, it must be determined by at least the number of previous states used in the controller (i.e., for the R1/R2 controller, the number used for the enumeration of previous states needs to be defined for the combination of all previous states of the two back-ends).

To define the pulse potential at the end of the update, knowing the fixed pulse applied during the pulse, it must be possible to determine the pulse potential at the beginning of the update for all states in the enumeration. This means that the net pulse applied by a waveform must be a function of a smaller number of previous states than is required to uniquely define the pulse potential at the end. In order to translate this into a determination of the optimum waveform to be applied by a controller, this means that the pulse potential of a waveform must be a function of a smaller number of previous states than the number of states used to determine the waveform. For example, if a controller has pulse data determined by three states R1, R2, and R3 (where R3 is the gray level immediately preceding the initial gray level of the transition under consideration), each combination of R1 and R2 must hold the electro-optic medium at the same pulse potential independent of R3.

In other words, the controller must "know" the pulse potential of the electro-optical medium when the considered transition starts, so that it can apply the correct pulse to generate the appropriate value of the pulse potential following the transition. If the pulse potential is allowed to vary according to all of R1, R2, and R3 in the above example, there will be no way for the controller to "know" the starting pulse potential in the next transition, since the previously used R3 information may have been discarded.

As has been shown, the limited transition method of the present invention is preferably performed using an R1/R2 controller (i.e., where the pulses applied during any transition depend only on the initial and final gray levels of the transition), and as will be appreciated from the above discussion, the pulse potential must be uniquely defined as a function of R1 alone in such a controller.

A more complex situation arises when determining the optimum waveform from such a phenomenon which can be referred to as "pulse lag". Except in rare cases of extreme overdrive under the optical pen, the electro-optic medium driven by a voltage of one polarity always becomes darker, while the electro-optic medium driven by a voltage of the opposite polarity always becomes whiter. However, for some electro-optic media, and in particular for some encapsulated electro-optic media, the optical state exhibits hysteresis as a function of the pulse; as the medium is driven further to white, the optical change per unit of applied pulses decreases, but if the polarity of the applied voltage is suddenly reversed so that the display is driven to the opposite direction, the optical change per unit of pulses increases suddenly. In other words, the optical change per pulse unit depends not only strongly on the current optical state, but also on the direction of change of the optical state.

This pulse lag creates an inherent "restoring force" that tends to shift the electro-optic medium toward intermediate gray levels, defeating the efforts of driving the medium from one state to another with unipolar pulses (as in a typical gray scale image flow), while still maintaining DC balance. Upon application of the pulse, the medium floats on the three-dimensional R1/R2/pulse hysteresis surface until it reaches an equilibrium. The balance is fixed for each pulse length and is generally at the center of the optical range. For example, it has been found empirically that driving an encapsulated four gray scale electro-optic medium from black to dark gray requires a 100ms x-15V unipolar pulse, and driving it from dark gray back to black requires a 300ms x 15V unipolar pulse. For obvious reasons, the waveform is not DC balanced.

One solution to the problem of pulse lag is to use bipolar driving, that is, driving the electro-optic medium from one grey level to the next on a (potentially) indirect path, first applying a pulse to drive the pixel optical rail into the optical rail as necessary to maintain DC balance, and then applying a second pulse to achieve the desired optical state. For example, in the above case, one can change from black to dark gray by applying a 100ms x-15V pulse, but change from dark gray back to white by first applying an additional negative voltage, then applying a positive voltage, floating the R1/R2 pulse curve down on the black state. As already discussed, such indirect transitions also avoid the error accumulation problem by the barrier stabilizing the gray scale.

The pulse hysteresis and previous state dependence of the electro-optic medium, as discussed above, in the above patents and applications, requires that the waveform of each transition vary with the previous state history of the pixel under consideration. As described in the aforementioned WO03/044765, the optimum waveform for each transition may be determined by building up a waveform (which is used to complete the series of optical states to access the electro-optic medium through a fixed, typically pseudo-random or previous state) using an initial "guess" transition matrix (i.e. the transition table corresponding to the aforementioned data array may be "tuned"). A program subtracts the actual optical states achieved in the same combination from the target gray state for each previous state to calculate an error matrix that is the same size as the transition matrix. Each element of the error matrix corresponds to an element of the transition matrix. If an element in the transition matrix is too high, the corresponding element in the error matrix is also pushed higher. The error matrix can then be driven towards zero with PID (proportional-integral-derivative) control. There are cross terms (each element in the transition matrix affects more than one element in the error matrix), but these effects are small, often decreasing as tuning proceeds through iterations, decreasing as the magnitude of the values in the error matrix decreases. (Note that sometimes the I or D constant of a PID controller can be set to 0, resulting in PI, PD, or P control.)

When the tuning process is complete, it is found that some number of previous optical states should be in the transition matrix in order to achieve some gray scale accuracy performance. For example, with the process the particular encapsulated electro-optic medium results in a waveform in which the controller records a more previous optical state than in the transfer matrix and algorithmically calculates the pulses in the first segment of the waveform to ensure DC balance. In the waveform, the pulse potential is allowed to be different for each combination of previous states covered by the transfer matrix.

The correlation between the number of transfer matrix sizes ("TM size") and the maximum optical error of the waveform is found, as set forth in table 1 below:

TABLE 1

TM dimension	Maximum optical error (L))
		1	10.6
2	3.8
		3	2.1
4	1.7

Because the limit of the visual perception of the ordinary observer is about 1L^*The data in the table indicate that it is useful to have more than one dimension in the transition matrix, two-dimensional matrices over one-dimensional, three-dimensional matrices over two-dimensional, etc.

Note that for all the points mentioned above, a preferred waveform is designed for the R1/R2 two-bit gray scale controller mentioned above. The waveform maintains a fixed pulse potential for each of the final optical states R1, but uses a two-dimensional transfer matrix. The barrier is stabilized to reduce the accumulation of errors and is designed to have low divergence during flipping because it takes into account the pulse-lag curve.

In the symbols used below, the numbers represent pulses. The amplitude of the volt-time product is made equal to the amplitude of the negative pulse by applying a negative pulse at a given time (i.e., -15V) and a positive pulse at a given time by applying a positive pulse at + V (i.e., the waveform is pulse width modulated). Voltage modulation may be used as another scheme.

In the preferred waveform, the following pulse sequence is applied during each update, reading in time from left to right:

-TM(R1，R2)IP(R1)-IP(R2)TM(R1，R2)

where "ip (rx)" represents the correlation value (in this case, a vector) from the pulse potential matrix having one value per gray level, and TM (R1, R2) represents the correlation value from the transfer matrix having one value per R1/R2 combination. Of course, TM (R1, R2) may be negative for certain values of R1 and R2. (As already noted, pulse sequences of this type may be abbreviated hereinafter for convenience as "-x/Δ IP/x" sequences.)

The values in the transfer matrix can be adjusted as needed without having to worry about DC balance because the net pulses of the first and third segments of the waveform are always 0. The difference in pulse potential between the initial and final states is applied in the middle segment of the waveform.

It has been found empirically that the last drive pulse has almost always a greater effect on the last grey level than the initial pulse, so that the transfer matrix for the waveform can be tuned with the same PID method as described above. For a fixed final grey level, the value set for the pulse potential influences the update speed of the waveform. For example, all pulse potentials may be set to zero, but it results in a long update time, since the last drive pulse (third segment) is always opposite to the initial pulse (first segment) of equal length. Thus, in this case, the last drive pulse cannot be longer than half the total update time. By careful selection of the pulse potential, it is possible to use a much larger percentage of the total update time for the last pulse; for example, more than half of the last drive pulse may be reached, and up to 80% of the maximum total update time.

The lengths of the different pulses are preferably selected by a computer, using a gradient-following optimization method, similar PID control, a combination of finite differences evaluation, etc.

As WO03/044765 indicates, transitions in electro-optic media are generally temperature sensitive, and it has been found that the uncompensated stability of grey levels to temperature increases when all transitions to a particular grey level always come from the same optical fence. The reason for this is simply that the switching speed of the electro-optical medium becomes faster or slower with temperature changes. If, in a 2-bit grayscale display, dark grayscale to light grayscale transitions are popped from the black bar, but white to light grayscale transitions are popped from the white bar. If the switching speed of the medium becomes slower, the bright gray state from black addressing will become darker, but the bright gray state from black addressing will become brighter. Thus, it is important that for a temperature stable waveform, a given gray level always approaches from the same "side", i.e. the last pulse of the waveform always has the same polarity. In the preferred drive scheme described above using the-TM (R1, R2) IP (R1) -IP (R2) TM (R1, R2) sequence, this requires that the TM (R1, R2) values be chosen such that the sign of each value depends only on R1, at least for some gray levels. One preferred method is to allow the TM value to be the sign of either of the black and white states, but only positive for bright gray shades and only negative for dark gray shades, thus only approaching the intermediate gray shades from the closer optical fence.

The preferred waveform is sufficiently compatible with techniques such as the insertion of short pause periods into the waveform to increase pulse resolution, as described below.

As already indicated, the above-described-x/Δ IP/x pulse sequence may be modified to include additional pulses. One such modification allows for the inclusion of an additional class of pulses, hereinafter referred to as "y" pulses. "y" pulse is characterized as being of the form [ + y ] [ -y ], where y is a pulse value, which may be negative or positive (in other words, the form [ -y ] [ + y ] is equally valid). The y pulse is distinct from the previously described "x" pulse in that half of the "x" pulse pair [ -x ] and [ + x ] are set before and after the Δ IP pulse, however the "y" pulse may be set elsewhere in the pulse sequence.

A second such modification is to apply a 0V "pulse" (i.e. a period when no voltage is applied to the relevant pixel) at any point within the pulse sequence to improve the performance of the sequence, for example by shifting the grey level caused by the transition up or down by a small amount, or by reducing or changing the effect of previous state information on the final state of the pixel. Such a 0V segment can be inserted either between different pulse elements or between single pulse elements.

A preferred method for constructing a fence stabilizing waveform, using the transition table described in WO03/044765, is as follows:

(a) setting the value of the pulse potential for each gray level (generally empirically derived) and inserting the appropriate Δ IP pulse of the transition table for each transition;

(b) for each transition, pick a value for x and insert a-x pulse before the Δ IP pulse and a + x pulse after the Δ IP pulse (as already noted, the value of x may be negative, so the-x and + x pulses may have any polarity);

(c) for each transition, a y value is picked and a-y and + y pulse is inserted into the pulse sequence. The-y/+ y pulse combination may be inserted into the pulse train at any pulse boundary, e.g., before the-x pulse, before the Δ IP pulse, before the + x pulse, or after the + x pulse;

(d) for each transition, inserting n frames 0V at any point or points within the sequence, where n is 0 or greater; and

(e) the above steps are repeated as many times as necessary until the waveform performance reaches the desired level.

This process will now be illustrated with reference to the figures. FIG. 2 shows the basic-x/Δ IP/+ x structure of a transition waveform, assuming for purposes of illustration that both the values of x and Δ IP are positive. Unless a 0V interval is required between the Δ IP and + x pulses, the voltage applied across the junction between the two pulses does not have to be reduced so that the Δ IP and + x pulses actually form a long positive pulse.

FIG. 3 illustrates symbolically an [ -y ] [ + y ] pulse pair inserted into the basic-x/Δ IP/+ x waveform shown in FIG. 2. The-y and + y pulses do not have to occur sequentially, but may be interpolated from different positions in the original waveform. There are two particular advantageous special cases.

In the first special case, the "-y, + y" pulse pair is placed at the beginning of the-x/Δ IP/+ x waveform, before the-x pulse, to produce the waveform shown in FIG. 4. It has been found that when the y and x signs are opposite, as illustrated in fig. 4, this last optical state can be fine tuned by even moderately coarse adjustment of the duration y. Thus, the value of x may be adjusted for coarse control and the value of y may be adjusted for final control of the final optical state of the electro-optic medium. This is believed to occur because the y pulse augments the-x pulse, thus changing the degree to which the electro-optic medium is pushed into one of its optical pens. The degree of pushing into one of the optical pens is known to derive a fine adjustment of the last optical state after a pulse away from the optical pen (in this case, provided by an x-pulse).

In a second special case, illustrated in FIG. 5, the-y pulse is again placed at the beginning of the-x/Δ IP/+ x waveform, before the-x pulse, but the + y pulse is placed at the end of the waveform, after the + x pulse. In this type of waveform, the last pulse provides coarse tuning because the last optical state is very sensitive to the amplitude of y. The x-pulse provides finer tuning because the final optical state is generally less strongly dependent on the amplitude of the drive into the optical pen.

As has been shown, more than one pair of "y" pulses may be inserted into the basic-x/Δ IP/+ x waveform to allow "fine tuning" of the gray scale level of the electro-optic medium, and such pairs of "y" pulses may be different from each other. Fig. 6 illustrates symbolically, a second pair of y-type pulses (labeled "-z", "+ z") inserted into the waveform of fig. 5 in such a manner similar to that of fig. 3. It will be readily apparent that since the-z and + z pulses may be introduced at any of the pulse boundaries of the waveforms shown in fig. 5, a large number of different waveforms may result from the introduction of the-z and + z pulses. A preferred resulting waveform is shown in fig. 7; this type of waveform is useful for fine tuning of the final optical state for the following reasons. Consider the case where there are no-z and + z pulses (i.e., the fig. 5 waveforms discussed above). The x pulse elements are used for fine tuning, and the final optical state can be decreased by increasing x and increased by decreasing x. However, reducing x beyond a certain point is undesirable because then the electro-optic medium is close enough to an optical fence that is needed to stabilize the waveform without the electro-optic medium. To avoid this problem, the-x pulse can be (actually) increased without changing the + x pulse by adding a-z, + z pulse pair as shown in fig. 7, making z have the opposite sign to x, instead of decreasing x. The + z pulse augments the-x pulse while the-z pulse maintains the transition at the desired net pulse, thus maintaining an overall DC balance transition table.

In the limited transition waveform scheme of the present invention, for "diagonal elements" (the transition table elements correspond to zero transitions where the initial and final gray levels are the same, so called because such elements are on the leading diagonal in a common matrix representation of a transition table; such diagonal elements have Δ IP equal to 0) to contain both x and y pulses. Any given transition table element may contain zero or more sets of x and/or y pulses.

The limited transition method of the present invention may also utilize the frame period. The scanning of groups of pixels is repeated during the course of a frame set (conveniently called a super-frame) required to completely overwrite the display, and on a typical electro-optic display, the scanning will be repeated more than once. Typically, a fixed scan rate is used for updates, e.g., 50Hz, which is allowed for 20ms frames. However, this frame length may provide insufficient resolution for optimizing waveform performance. In many cases, a frame of length t/2 is preferred, for example in a waveform of a typical 20 millisecond frame length. It is possible to combine frames of different delay times to produce a pulse resolution of n/2. To take into account a particular situation, a single frame of length 1.5 t may be inserted at the start of the waveform, while a similar frame is inserted at the end of the waveform, with a similar frame at the end of the waveform (immediately before the end of the 0V frame, which should occur at the normal frame rate, and which is typically used at the end of the waveform to avoid unwanted effects caused by residual voltage on the pixels). The two longer frames can be achieved by simply adding a 0.5 x t delay time between the scans of two adjacent frames. The waveform would then have the following structure:

tms frame, t/2ms delay time tms frame (all output to 0V)

For a normal frame of length 20 milliseconds, the initial and final frames plus their respective delays amount to 30 milliseconds.

With this waveform, structure, the initial and final pulses are allowed to vary in length by 10 milliseconds by the following algorithm:

(a) if the length of the initial pulse is divisible by t, the first frame consists of a 0V drive, and a corresponding number of tms frames are activated to achieve the desired pulse length; or

(b) If the length of the initial pulse divided by t leaves a remainder of t/2, then the first frame, 1.5 x t, is activated, followed by the initial frame by a corresponding number of t ms frames to achieve the desired pulse length.

The final pulse follows the same algorithm. Note that for the algorithm to work properly, the initial and final pulses must be aligned at the start and end points, respectively. In addition, to maintain DC balance, the initial and final pulses may be a-x/+ x pair of corresponding segments.

Regardless of whether a pause period is employed, it has been found that the effect of the waveform to effect the transition is modified by the presence of a period of zero voltage (in effect a time delay) during or before any pulse in the waveform, and the finite transition method of the present invention may include periods of zero voltage within or between successive pulses within the waveform, i.e. the waveform may be "non-continuous", as that term is used above and in the above-mentioned PCT/US 2004/010091. Fig. 8 to 10 illustrate variations of the basic-x/Δ IP/+ x waveform of fig. 2 including such a zero voltage period. In the waveform of fig. 8, a time delay is inserted between the-x pulse and the Δ IP pulse. In the waveform of fig. 9, a time delay is inserted within the Δ IP pulse, or, again, the Δ IP pulse is split into two separate pulses separated by the time delay. The waveform of fig. 10 is similar to that of fig. 9, except that the time delay is inserted within the + x pulse. The time delay may be included in a waveform to achieve an optical state that cannot be achieved without such a delay. The time delay can also be used to fine tune the final optical state. The fine tuning capability is important because in an active matrix drive the time resolution of each pulse is defined by the scan rate of the display. The time resolution provided by this scanning rate may be so coarse that the exact final optical state cannot be reached without some additional means of fine tuning.

Interrupt scanning method of the present invention

As already indicated, the present invention provides an "interrupted scanning" method for driving an electro-optic display having a plurality of pixels divided into a plurality of groups. The method includes selecting each of a plurality of successive groups of pixels and applying either a drive voltage or a non-drive voltage to each pixel in the selected group, and completing scanning of all the groups of pixels during a first frame period. The scanning of the group of pixels is repeated during a second frame period (it will be appreciated that any particular pixel may have the drive voltage applied during the first frame period and the non-drive voltage applied during the second frame period, or vice versa). In the interrupted scanning method of the present invention, the scanning of the pixel group is interrupted in a pause period between the first and second frame periods, the pause period being not longer than the first or second frame period. In the method, the first and second frame periods are generally equal in length, and the pause period is generally a fraction of the length of one of the frame periods (preferably 1/2, 1/4, etc.).

The interrupted scanning method may include a plurality of pause periods between different pairs of adjacent frame periods. Such a plurality of pause periods are preferably substantially equal in length, and the total length of the plurality of pause periods is preferably equal to either a full frame period or a frame period minus a pause period. For example, as will be discussed in more detail below, one embodiment of the first method may use multiple 20ms frame periods, and either three or four 5ms pause periods.

In the interrupted scan method, the pixel groups are of course typically each row of a conventional row/column active matrix pixel array. The interrupted scanning method includes selecting each of a plurality of successive groups of pixels (i.e., typically, scanning the rows of the matrix) and applying either a driving voltage or a non-driving voltage to the selected group, the scanning of all the groups of pixels being completed in a first frame period. The scanning of the pixel groups is repeated and, on a typical electro-optic display, the scanning is repeated more than once during the super-frame required to overwrite the display. The scanning of the group of pixels is interrupted during a pause period between the first and second frame periods, the pause period being no longer than the first or second frame period.

Although the drive voltage is applied to any particular pixel electrode only during each scan for a row addressing time, the drive voltage is sustained at the pixel electrode only slowly during the time between successive selection of the same row, so that the pixel continues to be driven during that time when the other rows of the matrix are being selected, whereas the interrupted scanning method relies on this continued driving of the pixel. Ignoring this moment during its "non-selected" time during which the slow decay of the voltage on the pixel electrode will occur, during the frame period immediately preceding the pause period, the pixel set to the drive voltage will continue to experience the drive voltage during the pause period, so that for such a pixel the preceding frame period is effectively extended by the length of the pause period. On the other hand, a pixel set to the non-driving (typically zero) voltage will continue to experience zero voltage during the pause period immediately preceding x during the frame period preceding the pause period. It may be desirable to adjust the length of the pause period to allow a slow decay of the voltage on the pixel electrode to ensure that the total pulse delivered during the pause period has the required value.

To illustrate a simple example of the interrupted scanning method for illustrative purposes, consider a pulse width modulation driving scheme having a superframe consisting of multiple (e.g., 10) 20ms frames. Typically, the last frame of the super-frame will set all pixels to the non-driving voltage, since the bistable electro-optic display will generally only be driven for a relatively long interval when the displayed image changes or when it is deemed preferable to refresh the displayed image, so that each super-frame will generally be followed by a long period in which the display is not driven, and it is very good to set all pixels to the non-driving voltage at the end of the super-frame in order to avoid rapid changes at some pixels during this long non-driving period. To change such a driving scheme in accordance with the interrupted scanning method of the present invention, a 10ms pause period may be inserted between two successive 20ms frames, and this simple modification halves the maximum possible difference between the applied pulse and the pulse ideally required to complete a given transition, thereby in practice approximately halving the maximum deviation in achieved grey levels. The 10ms pause period is conveniently inserted after the penultimate frame in each superframe, but may be inserted at other points in the superframe if desired.

In practice, in this example, it is preferable to insert not only a 10ms pause period, but also an additional 20ms frame into each superframe. The unmodified drive scheme enables one to apply such a pulse to any given pixel:

160, 180 units

One pulse unit is defined as a pulse caused by applying a driving voltage for 1 ms. Thus, the maximum difference between the available pulses and the ideal pulses for a given transition is 10 units. As already explained, any pixel set to the drive voltage on a frame preceding the pause period continues to experience the drive voltage for a period equal to the frame period plus the pause period, and thus, for that frame, a 30 unit pulse is experienced instead of 20 units.

0, 20, 30, 40, 50, 60 units, etc

The additional frame is preferably inserted into the superframe to allow the modified drive scheme to transmit a burst of exactly 180 units. Since any pulse that is a multiple of exactly 20 units requires that the relevant pixel be set to the non-drive voltage during the frame preceding the pause period, a pulse up to exactly 180 units requires an 11-frame super-frame, so that any pixel that is to receive the 180 pulses can be set to the drive voltage during the 9 frames, to the non-drive voltage during the frame preceding the pause period, and (as always) to the non-drive voltage during the last frame of the super-frame. Thus, when using this modified driving scheme, the maximum difference between the available pulses and the ideal pulses for a given transition is reduced to 5 units. (although the modified drive scheme is not capable of applying a 10 unit pulse, in practice the consequences are small in order to produce reasonably consistent gray scale levels, the number of available pulses must be significantly greater than the number of gray scale levels of the display, so that any gray scale transition is unlikely to require a pulse as small as 10 units.)

Of course the pause period may be of any number and length required to achieve the required control of the applied pulses. For example, rather than changing the above-described drive scheme to include a 10ms pause period, the drive scheme could be modified to include three 5ms pause periods after a different 20ms drive frame, preferably with the drive scheme being further supplemented with three further 20ms drive frames, not followed by a pause period. This modified drive scheme allows for the application of some such pulses to any given pixel:

0, 20, 25, 30, 35.. 170, 175, 180 units

Thereby reducing the maximum difference between the available pulses and the ideal pulses for a given transition to 2.5 units, a factor of four compared to the original unmodified drive scheme.

The interrupted scanning method the preceding discussion has ignored the problem of the polarity of the applied pulses. As discussed above and in the aforementioned WO03/044765, a bistable electro-optic medium requires pulses of both polarities to be applied. In some drive schemes, such as a slide-show drive scheme, all pixels of the display are first driven to an extreme optical state, either black or white, before a new image is written to the display, and the pixels are then driven to their final gray state by a single polarity pulse. Such a driving scheme may be modified in accordance with the interrupted scanning method in the manner already described. Other drive schemes require that pulses of both polarities be applied to drive the pixels to their final grey state. The pulses of both polarities may be applied in separate frames or the pulses of both polarities may be applied on the same frame, for example using a three level driving scheme where the common front electrode is held at one voltage V/2 while the respective pixel electrode is held at 0, V/2 or V. When pulses of two polarities are applied to a single frame, it is desirable to effect the interrupted scanning method by providing at least 2 single pause periods, one pause period following the frame in which one polarity pulse is applied and the other pause period following the frame in which the opposite polarity pulse is applied. However, when using a drive scheme in which pulses of both polarities are applied to the same frame, the interrupted scanning method can use only a single pause period, since as will be apparent from the foregoing, the inclusion of a pause period after a frame has the effect of increasing the amplitude of the pulse applied to any pixel to which a drive voltage has been applied in that frame, regardless of the polarity of the drive voltage.

As also discussed in the above-mentioned WO03/044765 and above, many bistable electro-optic media are preferably driven with a drive scheme that achieves long-term Direct Current (DC) balance, and such DC balance is conveniently achieved with a drive scheme in which a DC balance segment, the gray level of the pixel of which is substantially unchanged, is applied before the main drive segment in which the gray level is to be changed, the two segments being selected such that the algebraic sum of the applied pulses is 0 or at least very small. If the main drive section is modified according to the interrupted scanning method, it is strongly recommended to modify the DC balance section to avoid additional pulses caused by inserting pause periods that cumulatively cause a DC severe imbalance. However, the DC-balanced section does not have to be modified in a way that is an exact mirror image of the main drive section modification, since the DC-balanced section can have gaps (zero voltage frames) and most electro-optic media are not harmed by short-time DC imbalances. Thus, in the driving scheme discussed above, which utilizes a single 10ms pause period inserted between 10 20ms frames, DC balance can be achieved by making the duration of the first frame of the driving scheme 30 ms. Applying or not applying a drive voltage to a pixel during the frame will make the total pulse equal to several times 20 units so that the pulse can then be easily balanced. In a drive scheme using three 5ms pause periods, the first two frames of the drive scheme may similarly be 25 and 30ms in duration (in either order), again with the total pulse equaling several times 20 units.

From the above it will be seen that the interrupted scanning method of the present invention requires a trade-off between the need to extend the addressing time due to the need to include an additional frame for each intervening pause period in each super-frame, and the improved control of the pulses and hence the grey scale produced by the method. However, the interrupted scanning method provides a very significant improvement in the control of the pulses as long as the addressing time is not extended too much; for example, a driving scheme that includes 10 20ms frame modifications to include three 5ms pause periods for one superframe as described above would yield a four-fold improvement in pulse accuracy at the expense of a 40% extension in addressing time.

Balanced gray scale method of the present invention

As already indicated, the present invention also provides a balanced gray scale method for driving an electro-optic display having a plurality of pixels arranged in an array. The pixels are driven with a pulse width modulated waveform capable of applying a plurality of different pulses. The drive circuit stores data indicating whether a given pulse is applied, which will produce a gray level that is higher or lower than the desired gray level. When two adjacent pixels are both required to be at the same gray level, the pulses applied to the two pixels are adjusted so that one pixel is below the required gray level while the other pixel is above the required gray level.

In a preferred form of the method, the pixels are divided into two subgroups, hereinafter denoted "even" and "odd". The two groups of pixels may be arranged in a checkerboard pattern (so that the pixels of each row and column alternate between the two subgroups), or in other arrangements, as described in WO03/044765 above, assuming that each pixel has at least one neighbour of the opposite subgroup and that the two subgroups use different drive schemes. If the stored data indicates that one of the available pulses will produce a transition that is substantially a desired gray level, the pulses applied for that transition are applied to both even and odd pixels. However, if the stored data indicates that the pulses required for a particular grey scale transition are substantially halfway between the two available pulses, one of the pulses is used for the transition in even pixels and the other of the pulses is used for the transition in odd pixels. Thus, if two adjacent pixels are intended to be in the same gray state (the state where accurate control of the gray scale is most important), one of these pixels will have a gray scale level slightly above the desired level, while the other will have a gray scale level slightly below the desired level. The result of the visual and optical averaging will be to see an average of the two grey levels, thus producing an approximate grey level that is closer to the desired level than the level achieved with the available pulses. In practice, the balanced gray scale method uses a small signal spatial dither (applied to correct the error of the applied pulse) superimposed on the large signal true gray scale to increase the factor of the two available pulse levels. The effective resolution of a practical display is not compromised because each pixel is still at approximately the correct grey level.

The necessary calculations for a complete implementation are given in MATHLAB pseudo-code as follows. The floor function is rounded down to the nearest integer, and the mod function computes the remainder of its first argument divided by its second argument:

quotient＝floor(desired_impuslse)

remainder＝mod(desired_impulse，1)

ifremainder<＝0.25

even_parity_impulse＝quotient

odd_parity_impulse＝quotient

else if remainder<＝0.75

even_parity_impulse＝quotient+1

odd_parity_impulse＝quotient

else

even_parity_impulse＝quotient+1

odd_parity_impulse＝quotient+1

end.

in some of the previously described drive schemes, such as the cyclic RSGS drive scheme described above with reference to fig. 1A and 1B, the pixels of the display have been divided into two groups and different drive schemes are applied to the two groups so that the amplitude of the pulses required to achieve the required grey level will be different for the two groups. Such a "two-group" driving scheme may be modified according to a balanced gray scale method, but the implementation details of this method are slightly different from the simple case discussed above. Rather than simply comparing the available pulses with what is required for the desired transition, the errors for the two sets of gray levels are calculated separately, the errors are arithmetically averaged, and a determination is made as to whether moving a small group to a different available pulse will reduce the arithmetical average. Note that in this case, the reduction of the arithmetic mean may depend differently on which group is moved to a different pulse and obviously which is moved to produce the smaller mean.

Again, this method can be imagined as a small signal spatial dither realized on top of the large signal internal grey scale, which is used to correct the pulse error caused by the limitations of the pulse width modulation used. Because in this scheme each pixel is still approximately at the correct grey level and the correction is only to correct for rounding errors, the actual resolution of the display is not compromised. Put another way, the method achieves small-signal spatial dithering on top of the true gray scale of the large signal.

The different methods of the present invention may utilize various additional variations and techniques described in the above-mentioned applications, particularly the above-mentioned WO03/044765 and PCTUS 2004/010091. It will be appreciated that in the overall waveform used to drive an electro-optic display, at least in some cases, some of the transitions may be made in accordance with a different method of the invention, while other transitions may not make use of the method of the invention, but may make use of other types of transitions described below. For example, the various methods of the present invention may utilize any one or more of:

non-sequential addressing (see above PCT/US2004/010091, paragraphs [0142] to [0234] and FIGS. 1-12);

DC balanced addressing, as discussed in part above (see also PCT/US2004/010091, supra, paragraphs [0235] through [0260] and FIGS. 13-21);

a defined region update (see PCT/US2004/010091, paragraphs [0261] to [0280], supra);

offset voltage addressing (see PCT/US2004/010091, paragraphs [0284] to [0308] and FIG. 22, supra);

DTD integral reduction addressing (see PCT/US2004/010091, supra, paragraphs [0309] to [0326] and FIG. 23); and

the remaining voltage addresses (see WO03/044,765, pages 59 to 62, mentioned above).

Claims (10)

1. A method for driving a bistable electro-optic display having a plurality of pixels divided into a plurality of groups, the method comprising:

(a) selecting each of a plurality of successive groups of pixels and applying either a drive voltage or a non-drive voltage to each pixel in the selected group, the scanning of all the groups of pixels being completed in a first frame period; and

(b) selecting each of the plurality of successive groups of pixels and applying either the drive voltage or the non-drive voltage to each pixel in the selected group, repeating the scanning of all the groups of pixels for a second frame period,

the method is characterized by interrupting the scanning of the group of pixels in a pause period between the first and second frame periods, the pause period being no longer than the first or second frame period.

2. A method according to claim 1, wherein the first and second frame periods are of equal length.

3. A method according to claim 2, wherein the length of the pause period is a fraction of the length of one of the first and second frame periods.

4. The method according to claim 1, wherein the method further comprises:

selecting said each of said plurality of successive groups of pixels and applying either said drive voltage or said non-drive voltage to each pixel in the selected group, repeating the scanning of all said groups of pixels for a third frame period;

the pixel groups are scanned during at least first, second and third frame periods, and the scanning of all the pixel groups is interrupted in at least first and second pause periods between successive frame periods.

5. A method according to claim 4, wherein the first, second and third frame periods are substantially equal in length, and the total length of said at least first and second pause periods is equal to one frame period or one frame period minus one pause period.

6. A method according to claim 1, wherein the pixels are arranged in a matrix having a plurality of rows and a plurality of columns, each pixel being defined by the intersection of a given row and a given column, and wherein each group of pixels comprises one row or one column of the matrix.

7. A method according to claim 1, wherein the scanning of the display is performed such that the integral of the applied voltage over time is bounded for any series of transitions the pixel undergoes.

8. A method according to claim 1 wherein the electro-optic display comprises an electrochromic or rotating bichromal thin film electro-optic medium.

9. A method according to claim 1 wherein the electro-optic display comprises an encapsulated electrophoretic medium.

10. A method according to claim 1 wherein the electro-optic display comprises a microporous electrophoretic medium.