Background Electrochromic devices, such as information displays, smart windows, dimable mirrors, etc., comprise one or more electrochromic matrix-elements i. e. cells, segments, pixels, dots, etc. Each element is comprised of at least two electrodes that are electrically connected via an ion-conducting medium, i. e. an electrolyte.

The elements may change color if the potential difference between the electrodes is changed. This color change is caused by redox reactions, i. e. changes in the distribution of electric charge within the element. The color changes are caused by changes in the spectral absorption and/or emission properties of electromagnetic radiation. With'color'we mean also the spectral properties outside the visible region, e. g. in the infrared and the ultraviolet regions.

There are several types of electrochromic systems, such as: 1. Intercalation systems. Wherein thin metal-oxide films on conducting glass are commonly used as electrodes. The metal-oxide film (e. g. tungsten trioxide) on the color electrode colors at negative potentials by simultaneous injection of electrons and intercalation of small cations, i. e. small positive ions like protons and lithium ions. This reduction reaction is balanced by oxidation reactions at the counter electrode that in some cases contribute to the coloration.

2. Systems with dyed highly porous films. Wherein the color electrode is comprised of a highly porous film of sintered small metal-oxide particles on conductive glass. The surface of the particle network is covered by redox-active chromophore molecules (dye), which exhibit different colors in different oxidation states. Charge is guided to and from the dye molecules via the conductive particle network. As the film is highly porous the amount of dye molecules is high enough to cause strong coloration of the electrode. The redox reactions in the porous color electrode are balanced by reactions of redox-active species in the electrolyte or in a film at the counter electrode.

3. All-in-solution systems. Wherein both the color redox system and the counter redox system are dissolved in the electrolyte. The electrodes are in this case commonly made of plain conductive glass.

In an electrochromic system, a comparatively high electric charge is needed to switch between different color states. This situation differs from many other display technologies, e. g. liquid-crystal displays. Electrochromic devices gradually shift from one state to another over a system-related time period tswitch. The switch time twitch (or response time) for an electrochromic system is defined as the time required for a switch between the fully colored and the fully bleached state. Alternatively the switch time may be defined more rigorously in different ways, e. g. as the time when 90 % of the difference in transmission or reflectance between the fully bleached and colored states has been attained. This definition result in two switch times, from colored to bleached and from bleached to colored. Although these switch times may be different, it will be assumed in the following that they are equally long. The switch time may depend on the operation conditions, such as driving voltages, temperature, the number of matrix-elements that are switched simultaneously, etc.

Furthermore the switch time may be different for different matrix-elements in an array due to manufacturing variations, different sizes of the matrix-elements, the resistance in the power lines, etc.

As the present invention is mainly intended for electrochromic graphic displays and array devices of other kinds, it is useful to introduce the array switch time ta,, ay and the matrix-element switch time tm-etement. tarray is defined as the time required for a change between arbitrary images of the whole array, and tm-e) ement is defined as the time under which a single matrix-element has to be electrically connected in order to switch color state. Note that the definitions of tm-e) ement and tswitch are slightly different. The matrix-element switch time may be shorter than the real switch time (tm-element tswitch) In active-matrix displays, each matrix-element circuit comprises at least one active component, such as a field-effect transistor (FET) or the like. A'standard'active- matrix design with thin-film transistors as switches is shown in figure 1. The gates of the transistors in one row are connected to a common line, and the sources of the transistors in one column are connected to a common line. Each transistor drain is connected to a matrix-element electrode. In a simple row-by-row addressing scheme the rows are selected one at a time by signals on the respective select lines (row lines). Video information, that may contain grayscale information, is fed to the selected matrix-elements via the data lines (column lines). The array switch time for such a display is tarray N X tm-clement where N is the number of rows. Throughout this application all intermediate color- states between fully bleached and fully colored are referred to as grayscale-states (or grayscale levels) irrespective of color.

Prior art The switch times of electrochromic displays range from 10 ms to several minutes.

According to equation 1 the discussed row-by-row scheme would give an array switch time from 10 s up to several days for an electrochromic 1000x1000 matrix of the standard type. Fig 1 shows an example of a matrix element circuit 10 in an active matrix display, which comprises a drive transistor 12 having a source connected to a column line 14, a drain connected to a matrix element 16, and a gate connected to a row line 18. However, as shown in US 5,049,868, tarray is much shorter in existing modified active matrix displays 20 with one additional control transistor 22 and a control capacitor 24 in each matrix-element circuit 20, as shown in fig. 2. Such modified matrices are hereafter referred to as"active memory matrices". The thin-film transistors 12 in a standard active-matrix device 10 have both a selecting and a switching function. In active memory-matrices these functions are separated on two different transistors 12 and 22. The matrix-element electrodes 16 are, via drive-transistors 12, connected to a drive potential source that is common for all matrix-elements in the matrix. The drive-transistors 12 are either ON (fully conductive) or OFF (not conductive). The conduction of the drive- transistors 12 is controlled by the potentials of control-capacitors 24, and the charging states of the control-capacitors 24 are controlled by control-transistors 22.

Matrix-elements 16 are selected row-by-row by gating all control-transistors 22 in one row and charging the control capacitors 24 in this row to either ON (1) or OFF (0) potentials via signals on the column lines 14, as is schematically shown in fig. 3.

The process of selecting matrix-elements in the whole matrix is denoted SELECT and compared to the matrix-element switch times in an electrochromic display it is very fast: tSELEC1' « tm-element. (2) Deselecting all matrix-elements, DESELECT, is carried out by simultaneously gating all row lines with OFF potentials on the column lines. This process is even faster than SELECT: DESELECT < tsELECT << tm-etement. (3) As a conclusion, the time it takes to arbitrarily change the state of all drive- transistors (SELECT) is negligible in comparison with the matrix-element switch time (tm-element)- One possible arrangement of a display comprising a common back electrode, connected to ground, is shown in fig. 4. This display 30 comprises a number of electrochromic matrix-element electrodes 16, which are connected to a back electrode 32 via"matrix-element control circuits", i. e. drive-transistors 12. The electrolyte 34 and the front electrode 36 are common for all matrix-elements in this example. An external drive potential U is applied between the back electrode 32 and the front electrode 36 via an external"common switch"38. The device is operated in two basic modes: BREAK and LOAD. In BREAK the common switch 38 and all drive-transistors 12 are OFF. There are no external connections between the electrodes 32 and 36. Thus the imaged information is preserved if the color memory is long, i. e. if the degree of self-discharging is low. In the second operation mode, LOAD, the external switch 38 is ON and the drive-transistors 12 of selected matrix- elements are ON. Matrix-elements are selected in the absolute beginning of LOAD by SELECT, and all matrix-elements are deselected by DESELECT in the absolute end of LOAD. A current will flow through all selected matrix-elements until their potentials Um-element are equal to U or until the matrix-elements are deselected. To guarantee full charging the LOAD time is set equal to the matrix-element switch time: tload-tm-element- (4) In this way the potentials Um-element of all selected matrix-elements will attain the externally applied drive potential U during the LOAD period. It is not meaningful with longer LOAD-periods for switching purposes.

The process of arbitrarily changing the imaged information of the whole matrix is denoted ARRAY. An ARRAY is built up of several LOAD periods with different external drive potentials U. In the simplest case of only two gray shades, only black and white, an ARRAY consists of two LOAD periods with external drive potentials Ubteach and Udark (fig. 5). The array switch time becomes tal-lay = 2 x t", elene"t (5) which is independent of the array size N and a clear improvement as compared to the standard active-matrix (cf. eqn. 1). The general expression for ng, ay grayscale levels becomes tarray-ngray X tm-element. (6) Thus the existing active memory-matrix technology suffers from long array switch times for grayscale displays. For example, if the matrix-element switch time is 400 ms, which is normal for electrochromic displays of type 2 above, tarray becomes 6.4 s with 16 grayscale levels.

Therefore, it would in many situations be a great advantage if one could obtain means for significantly reducing the array switch time (tarray) for grayscale operation (eqn. 6).

Summary of the invention The object of the invention is to provide a method that overcomes the drawbacks of the prior art methods. This is achieved by the method of changing grayscale-states for a plurality of matrix-elements in an electrochromic matrix device as defined in claim 1.

Another object of the invention is to provide a device that overcomes the drawbacks of the prior art devices. This is achieved by the electrochromic matrix devices as defined in the claims 15,16 and 19.

One advantage with the method and the devices according to the invention is that the array switch time (tarray) may be reduced to the array switch time for electrochromic black-and-white arrays in prior art (eqn. 5) and preferably to the theoretical limit set by the matrix-element switch time (tm-element). The array switch time (tarray) may also be reduced for electrochromic color devices.

Another advantage with the method and the devices according to the invention is that they makes it possible to use a power supply with only two drive potentials for color changes in a monochrome grayscale device, and preferably to operate such devices without changing the applied drive potential.

Still another advantage is that the whole display is changed simultaneously, whereby enhanced visibility is achieved.

Embodiments of the invention are defined in the dependent claims.

Brief description of the figures Fig. 1 is a schematic representation of the circuit arrangement of an active matrix device.

Fig. 2 is a schematic representation of the circuit arrangement of an active memory matrix device according to prior art.

Fig. 3 shows selection of matrix-elements in an active memory matrix device.

Fig. 4 is a schematic representation of an active memory matrix device.

Fig. 5 illustrates the basic steps of controlling an active memory matrix device according to prior art.

Fig. 6 illustrates the basic steps of controlling an active memory matrix device according to the present invention.

Fig. 7 is a schematic representation of the circuit arrangement of an active memory matrix device according to the present invention.

Fig. 8 is a schematic representation of the circuit arrangement according to one embodiment of an active memory matrix device according to the present invention.

Fig. 9 is a schematic representation of the circuit arrangement according to another embodiment of an active memory matrix device according to the present invention.

Fig. 10 is a schematic representation of the circuit arrangement according to another embodiment of an active memory matrix device according to the present invention.

Fig. 11 is a schematic representation of the circuit arrangement according to another embodiment of an active memory matrix device according to the present invention.

Fig. 12 is a schematic representation of the circuit arrangement according to another embodiment of an active memory matrix device according to the present invention.

Detailed description of the invention The active memory-matrix technique described in prior art operates with basic periods (LOAD) of tm-efement to ensure that the potentials of all selected matrix- elements reach the same final potential. The drive transistor is ON during the whole LOAD period, meaning that the matrix-element current Im element (the current to the matrix-element electrodes) of selected matrix-elements is not actively regulated by the matrix-element circuit. The function of the control capacitors is only to store information about what matrix-elements are selected. The present invention provides means for actively regulating the current to the matrix-element electrodes during LOAD periods by controlling the charge in the control capacitors. Thus, compared to prior art, wherein the control capacitors only represent the selection status, i. e. selected/not selected, a new functionality is added to the control capacitors. The capacitor charges are individually controlled by either charging them differently in the absolute beginning of LOAD, or by changing their charges during LOAD (figure 6). The method according to the present invention comprises at least one LOAD (switch cycle), depending on embodiment, which LOAD as above has a duration that is equal to or less than switch. In a general representation each LOAD of the embodiments presented below comprises the steps of : selecting matrix-elements that are to be changed, including at least one matrix-element that is to be changed between different grayscale - states than the others connecting said selected matrix-elements to a drive-potential, and controlling the drive-potential for each of said selected matrix- elements, such that they reach their respective desired grayscale- states.

A more detailed description of the method and embodiments thereof are found below.

Fig. 7 shows a schematic representation of an electrochromic matrix device that may be controlled by the method according to the invention. In the electrochromic matrix device 40 the step of selecting matrix-elements 46 is performed by a matrix control unit 42, and the steps of connecting and controlling the drive-potential for a selected matrix-element 46 is performed by a matrix-element control-unit 44 connected to and associated with said selected matrix-element 46, each matrix- element control-unit 44 being connected to and controlled by the matrix control unit 42. The matrix-element control-unit 44 may be a matrix-element circuit of the type shown in fig. 2, but it may also be formed in alternative ways. As shown in fig 8, such a matrix-element control-unit 44 generally comprises : at least one drive transistor 12 having a source connected to a drive potential source 50, a drain connected to a matrix element 16, and a gate connected to a first node 52, said drive transistor 12 having an associated switch potential Uswitcil at which it changes conduction state, at least one control capacitor 24 connected between said first node 52 and ground, and a control-circuit 48 connected to the matrix control 42 unit and to said first node 52, wherein the steps of connecting and controlling the drive-potential for a selected matrix-element comprise controlling of the control circuit 48 to change the charging-state of said control capacitor 24 by changing the potential at the first node 52 such that the conduction state of the drive transistor 12 is changed. The control circuit 48 may, as shown in fig. 2, be comprised of a transistor having its drain coupled to said first node 52, and its gate and source coupled to row and column control lines, 18 and 14, respectively.

In a first embodiment, the method comprises two successive LOAD periods, one for each direction of change in grayscale state (i. e. bleaching and coloring). For each LOAD period the step of selecting matrix-elements comprises selecting all matrix- elements that are to be changed in that grayscale direction, the step of connecting said selected matrix-elements comprises connecting all selected matrix-elements to a common drive-potential (Ub, Cach or Udark) and the step of controlling the drive- potential for each of said matrix-elements, comprises disconnecting matrix- elements from the common drive-potential as they reach their respective desired grayscale-states.

In this embodiment, just like in the prior art system, an ARRAY-process comprises of two LOAD-periods, one for bleaching (U=Ubieac,,) and one for coloring (U=Udark), but in contrast to prior art the present invention the second LOAD-period is subdivided into gray-l grayscale periods, where ngray is the number of grayscale levels. Matrix elements that are to be colour-changed are fully bleached in the first LOAD and colored to the desired grayscale level in the second LOAD. To give a certain matrix- element a desired grayscale level, the drive transistor 12 is only ON during a predefined number of grayscale periods, whereby the charge of the control capacitor 24 may be varied in a controlled manner, and thus also the color (grayscale state) of the matrix-element. This method is applicable to a matrix device of the type shown in fig. 3 and described in detail above. However, the method of controlling the device is adapted to handle grayscale operation in a more efficient way. The array switch time (tarray) for this embodiment is equal to: tarray = 2 X tm-element 2 X tswitch irrespective of the number of grayscale levels.

For some applications the visual impression of the above embodiment might be considered disturbing, as all matrix elements that are to be changed first are bleached during the first LOAD. In an alternative embodiment both the first and the second LOAD periods are divided into (gray-1) grayscale periods. During the first LOAD the matrix elements that are to be bleached are bleached to their desired grayscale state whereas matrix elements that are to be colored are fully bleached.

During the second LOAD the matrix elements that are to be colored are colored to their desired grayscale state whereas matrix elements that are to be bleached are fully colored. Thus, matrix elements that are to be colored and bleached are first fully bleached and fully colored, respectively. Alternatively these first steps of full bleaching and coloration, respectively, may be eliminated in an operation scheme that is more demanding on the matrix control unit. In such a scheme only matrix elements that are to be bleached and colored are operated on in the first and second LOAD periods, respectively. That is, the matrix elements are not fully bleached or colored before they are charged to their desired grayscale states. This requires that information about the actual grayscale states of the matrix elements is accessible to the matrix control unit and, further, that the matrix control unit is able of processing this information such that the proper number of grayscale periods is calculated for arbitrary changes between different grayscale states. This may also mean that more than (ngray-1) grayscale periods are needed. The present invention does not comprise the details of how such grayscale information is accessed and processed.

The drive transistor 12 in fig. 3 is either ON or OFF, i. e. it is never in intermediate conduction states. The control capacitors 24 are also charged by one of two different potentials (0,1) via the column lines. The difference from prior art is that the drive transistor 12 of selected matrix-elements is not necessarily ON during the whole LOAD period. The ON-time is varied by SELECT processes for each grayscale period during LOAD (fig. 6). It is the ON-time that determines the final grayscale level. To have ng, ay grayscale states, (ng, ay-1) grayscale periods are needed, each comprising one SELECT process.

As discussed above there are several variants of the operation scheme for the first embodiment of the invention. This is the case also for the following embodiments below, although the operation schemes in most cases are not discussed in detail.

We emphasize however that similar variants as for the first embodiment are in several cases possible also for the following embodiments.

In an alternative embodiment of the method according to the invention, grayscale operation is achieved without several SELECT processes during LOAD, as the grayscale information is transferred to the matrix-element control-unit as a duration parameter. To achieve this, the matrix-element control-unit of fig. 2 is modified in that it has a resistance 54 added in parallel to the control capacitor 24 (fig. 9). The LOAD periods contain only one SELECT, but the control capacitors 24 are charged to different potentials via the column lines 14. The charges decrease because of the parallel resistances 54. In this way the drive transistors may change gradually between the ON and OFF conduction states during LOAD. The initial potentials after SELECT determine the final grayscale levels of the matrix-elements. ngray potential levels are needed on the column lines.

An alternative way to provide a matrix-element control-unit with the ability to control the potential on the matrix-element in accordance with a duration parameter, is to modify the matrix-element control-unit of fig. 2 such that the the gate-drain potential of the drive-transistors 56 determines the conduction state instead of the gate-substrate potential (fig 10). That is, the potential difference (Ucapacitor-Um-element) between the control capacitor 24 and the matrix-element electrode 16 determines if the drive transistor 56 is ON or OFF or between these states. Just like above, one ARRAY consists of two LOAD periods that contain only one SELECT each. In the first LOAD the control capacitors 24 of the selected matrix elements are charged such that the drive transistor 56 is fully conducting throughout the LOAD period, meaning that the selected matrix elements are fully bleached (or fully colored). In the second LOAD the selected matrix elements are colored (or bleached) to different grayscale states by charging the control capacitors 24 such that the conduction states of the drive transistors initially are ON but that they approach the OFF state during the LOAD period. The final grayscale state is determined by the pre-set potential of the capacitor 24, Ucapacitor.

In the examples the electrolyte is common for all matrix-elements, but it may also be separated in electrolyte baths for single matrix elements or groups of matrix elements to e. g. avoid or reduce cross-talk between matrix elements.

The back electrode may be segmented or replaced by row or column lines, whereby e. g. color displays may be constructed (see below). One such"back line"may further be shared by two neighbouring rows or columns (using N/2 lines for a NxN array). The front electrode may also be segmented in lines for e. g. color operation.

Color display operation may be achieved in a number of ways that are obvious to one skilled in the art. One way is to use color filters on different groups (e. g. rows or columns) of matrix elements. Alternatively the composition of the matrix elements may vary such that the"electrochromic color"of matrix elements in different groups is different. This may require that the drive potential source (e. g. Ugteen, Ublue, Ured) is different for the different color groups. One way of achieving this is to segment either the front or the back electrode in column or row lines, and by using different power sources for the different lines. It is also possible to use complex electrochromic systems (e. g. mixes of dyes in electrochromic systems of type 2) with several electrochromic colors in each matrix element. In this case the grayscale does not change monotonically between"fully colored"and"fully bleached". In spite of this we refer to the different color states of the matrix elements as grayscale states.

Regulation of the matrix-element current in prior art is obtained by changing the (constant) potential between the front and back electrodes at the same time as new matrix-elements are selected. The matrix-elements always obtain the applied potential. In the present invention the drive potentials of selected matrix elements are individually controlled. However, the matrix-element current can be regulated also by applying different kinds of applied potential functions (ramps, square waves, ...) in combination with the manipulation of the matrix-element transistor conduction state. One example is to have initially high potentials for decreasing the matrix-element switch time. Other examples may combine the potential function (e. g. ramp) and the matrix-element circuitry to attain the current regulation.

In another embodiment the array switch time (tarray) is reduced from 2xtm-e _element (cf. eq. 7) to the matrix-element switch time tarl ay = tm-eìementx (9) An ARRAY then consists of only one LOAD period of simultaneous bleaching and coloring. This is possible by further modifying the matrix-element control-unit of fig.

2 by adding a second drive transistor 58 and a second drive potential source 60 as shown in fig 11. Two external drive potentials, 50 and 60, (bleach and Udark) are connected to the matrix-element electrodes 16 via the two drive transistorsl2 and 58. The additional external drive potential 60 requires additional"back lines"to the matrix-elements. In figure 11 each back line is shared between two adjacent rows of matrix-elements. In this way N/2 lines are needed for Ubjeach and N/2 lines are needed for Udark, i. e. in total N back lines are needed for the drive potential connections in an NxN matrix. The external drive potentials, 50 and 60, may be held constant. Thus the drivers for the power supply to the matrix-elements become very simple. The first drive transistor 12 is ON if AU (Ucapacitor-Usubstrate) is higher than a certain limit, whereas the second drive transistor 58 is ON if AU is lower than a certain limit. The limits are chosen so that both transistors never conduct (significantly) at the same time. The control capacitors have three logic levels (-1, 0, 1), corresponding to: (-1) the first drive transistor 12 is in a conducting state, (0) neither one of the drive transistors, 12 and 58, are in a conducting state, and (1) the second drive transistor 58 is in a conducting state. The described operation scheme requires (as already discussed for one of the operation schemes for the first embodiment) that the matrix control unit can handle changes between arbitrary grayscale states in a single LOAD period. Note that the usage of two drive transistors, 12 and 58, and the simultaneous bleaching and coloring also reduces the array switch time for operation with only two grayscale levels (cf. eqns. 5 and 9).

In still a further embodiment, the array switch time (tarray) is equal to or less than the switch time of a single matrix-element (tm-element). In this embodiment the step of selecting matrix-elements comprises selecting all matrix- elements that are to be changed, and the step of connecting said selected matrix-elements comprises connecting selected matrix-elements that are to be changed to the same grayscale-state to a common drive-potential.

Whereby all selected matrix elements reach their desired grayscale-states simultaneously, and the step of controlling the drive-potential for each of said matrix-elements, comprises disconnecting all matrix-elements from their respective drive-potentials at the end of the switch cycle.

Fig. 12 shows a possible arrangement by which the method according to this embodiment may be performed. In this arrangement each matrix-element control- unit 100 is comprised of equally many matrix-element sub control-units 110 as the number of grayscales that are to be provided. Each of these matrix-element sub control-units 110 being comprised of : at least one drive transistor 12,70,72 and 74 having a source connected to a drive potential source 50,82,84 and 86, a drain connected to a matrix element 16, and a gate connected to a node 52,90,92 and 94 at least one control capacitor 24, 76,78 and 80 connected between said node 52, 90,92 and 94 and ground, and a control-circuit 48, 96,98 and 99 connected to the matrix control unit 42 and to said node 52,90,92 and 94.

Wherein the drive transistor 12,70,72 and 74 of each matrix-element sub control- unit 110 is connected to a drive potential different from drive potentials connected to drive transistors 12,70,72 and 74 of other matrix-element sub control-units 110. And wherein the step of connecting selected matrix elements 16 is performed by the matrix control unit 42, which is arranged to connect no more than one matrix-element sub control-unit 110 in each matrix-element control-unit 44 at a time.

This last embodiment combines fast switching, high grayscale accuracy, and a simple arrangement for the matrix control unit (no need of e. g. storing information about gray scale levels).

In fig. 4 the external drive potential U constitutes the drive potential source for all matrix elements. In this case it is not important if the back electrode 32 or the front electrode 36 is chosen as ground. However, if there are more than one external drive potential the choice of grounding point is less obvious. One can distinguish between two basic cases: Firstly, there may be more than one external drive potential U connected to each matrix-element control circuit (figs. 11-12). Then the natural choice of grounding is either at only one of these connections (like in fig. 11) or at the front electrode. Secondly, the front electrode 36 or the back electrode 32 may be segmented in different groups with different external drive potentials U, as has been discussed above. If the front electrode is segmented, then the natural choice of grounding point is the back electrode, and vice versa.

While the invention has been described in connection with what is presently considered to be the most practical and preferred embodiments, it is to be understood that the invention is not to be limited to the disclosed embodiments, but on the contrary, is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims.