DE69119771T2 - MULTIPLEX ADDRESSING OF FERROELECTRIC LIQUID CRYSTAL DISPLAYS - Google Patents

Description

The invention relates to multiplex addressing of ferroelectric liquid crystal displays. Such displays use a twisted chiral smectic D, I or F liquid crystal material.

Liquid crystal devices typically comprise a thin layer of liquid crystal material between two glass plates. Optically transparent electrodes are placed on the inner surfaces of both plates. When an electrical voltage is applied to these electrodes, the resulting electric field changes the molecular alignment of the liquid crystal molecules. The changes in molecular alignment are readily visible and form the basis for many types of liquid crystal displays.

In liquid crystal displays, the molecules switch between two different alignment orientations depending on the polarity of the applied electric field. These devices are bistable to some degree and tend to remain in one of the two states until they are switched to the other. This allows the multiplex addressing of fairly large displays.

A common multiplex display has display elements, ie pixels, arranged in the form of an x, y matrix as a display for eg alphanumeric symbols. The matrix format is obtained by arranging the electrodes on a disk as a sequence of column electrodes, and the electrodes on the other disk as a sequence of row electrodes. The intersections between each column and row form addressable elements or pixels. In addition to matrix arrangements, polar coordinates (r - θ) and numerical seven-segment displays are known.

There are many different types of multiplexed addressing. A common feature is the application of a voltage to each row or line in sequence, called a strobe voltage. Along with the strobe signal applied to each row, appropriate voltages, called data voltages, are applied to all column electrodes. The differences between the different types lie in the shape of the strobe signal and the data voltage waveforms.

European patent application 0 306 203 describes a multiplex addressing scheme for ferroelectric liquid crystal displays. In this application, the strobe signal is a unipolar pulse of alternating polarity and the two data waveforms are square waves of opposite sign. The strobe signal pulse width is half the data signal period. The interaction of the strobe signal and the corresponding one data voltage causes the liquid crystal material to switch.

Another type of addressing is described in EP 0 197 742, where the strobe signal and the data can be the same as in EP 0 306 203. In addition, an erase pulse of half the data waveform period with an amplitude that is larger than the strobe signal amplitude can be used. Erase occurs before the strobe signal is applied.

On the other hand, the erasure and the strobe signal can be combined into a single signal form, so that, for example, an erasure pulse with twice the period length and strobe signal amplitude followed by a strobe signal pulse of opposite polarity over a period of time followed by a strobe signal pulse of zero amplitude over a period of time. The time between application of the combined erase and strobe signal to successive electrodes is two period lengths. In one example, row n (n is an integer) is thus erased, while row n-1 receives the two strobe signal pulses. In another embodiment, the erase signal and the strobe signal are not applied to adjacent rows but to the row after but one.

Other addressing methods are described in GB 2 146 473-A; GB 2 173 336-A; GB 2 173 337-A; GB 2 173 629-A; WO 89/05025; Harada et al 1985 S.I.D. Digest Paper 8,4 pp. 131-134; and Lagerwall et al 1985 IEEE, IDRC, pp. 213-221; Proceedings 1988 IEEE, IDRC, pp. 98-101 "Fast Addressing for Ferro Electric LC Display Panels" by P. Maltese et al.

The material can be switched between its two states by means of two strobe pulses of opposite sign in conjunction with a data waveform. Alternatively, an erase pulse can be used to bring the material into one state and a single strobe pulse with an appropriate data pulse can be used to selectively return pixels to the other state. Periodically, the sign of the erase and strobe pulses is changed to maintain a net DC current of zero.

These erase pulses are normally larger in amplitude and length of application than the strobe signal pulses, so that the material changes regardless of which of the two data signals is present at any of the intersection points. Erase pulses can be pre-programmed line by line. the strobe signal, or the entire display can be cleared at a time, or a group of lines can be cleared at the same time.

One known type of erasure uses erasure pulses with the same product Vt of voltage (V) and time (t), but with opposite polarity to the product Vt of the strobe signal. The erasure pulse has an amplitude that is half as large and a duration that is twice as large as the strobe signal pulse. These values ensure that a net direct current of zero value flows during erasure and the strobe signal without periodic reversal of polarity. Experimentally, this type has shown poor performance.

Another known mode with a quench pulse is described in EP 0 378 293. This uses a conventional balanced DC strobe pulse (with equal periods of opposite polarity) with a similarly balanced DC quench pulse (with equal periods of opposite polarity) in which the width of the quench pulse can be several times that of the strobe pulse. Such a mode has a net DC current of zero value without periodic reversal of the polarity of the quench and strobe signals.

The DC balancing feature is particularly important in projection displays, since periodic reversal of polarity is not permitted if the gap between pixels is to be switched to an optical state.

A problem with previous displays is the time required to address complex displays. In order to control complex displays with a video frame rate, it is necessary to address the display quickly. Likewise, the Contrast ratio can be improved by fast addressing so that the column signal has a correspondingly high frequency. However, simply increasing the speed of addressing will not always result in correct switching. An object of the present invention is to reduce the time required to address a matrix display and to improve display contrast.

A method for multiplex addressing a ferroelectric liquid crystal display comprising a layer of chiral twisted smectic material between cell walls, the smectic material having a switching characteristic which depends both on the product of voltage and time and on the shape of this product, the cell walls carrying a first set of electrodes on one wall and a second set of electrodes on the other wall arranged to form a matrix (x, y) of addressable elements, the method comprising the steps of:

individual addressing of each electrode of the first electrode set, whereby the addressing is carried out either by applying two strobe signals of opposite polarity or by applying a clear signal followed by a strobe signal with periodic polarity reversal in order to maintain a net direct current of zero,

Applying one of two data signals to each electrode in the second set of electrodes, synchronously with the strobe signals, wherein both data signals alternately take on positive and negative values and one data signal is the inverse of the other data signal, and the period of the data signals, 2 ts, is twice as long as that of a single strobe pulse, ts, within the strobe signals,

wherein the resultant of one strobe signal and one data signal causes switching of the liquid crystal material and thereby addressing selected electrode intersections during a period equal to the data signal period, but the resultant of said one strobe signal and the other data signal at non-selected electrode intersections does not cause switching of the liquid crystal material during a period equal to the data signal period,

extending the end of the strobe signal to the selective addressing period of the next electrode in the first electrode set, while maintaining the same time period, 2 ts, between the application of the strobe signals to successive electrodes, whereby each electrode intersection is selectively addressed for a time greater than 2 ts,

and each electrode intersection is addressed with a resulting signal of appropriate sign, magnitude and time to bring the intersection (x, y) to the desired indicating state at a total net direct current of zero.

The strobe signal may first be zero during a first period ts, followed by a (main) voltage pulse which is not zero and is applied during a period of greater length than ts, e.g. (1.5, 2.0, 2.5, 3.0 or more) times ts. The strobe signal may have a non-zero voltage during the first ts period with the same or different polarity to the rest of the strobe signal; this first voltage pulse is variable in amplitude to allow temperature compensation. The strobe signal may be followed by a non-zero voltage during a period of time in which the polarity is opposite to that of the main voltage pulse, and which is e.g. greater than ts, ts, or smaller than ts.

The liquid crystal material can be switched between its two states by overlapping a strobe signal pulse and an appropriate data signal. Alternatively, the material can be switched to one of its states by an erase pulse and subsequently selected pixels can be switched back to the other state by overlapping a strobe signal pulse and an appropriate data signal.

The erase pulse can be in two parts; a first part and a second part, each with opposite polarity. Both parts of the erase pulse are selected to produce a product Vt of voltage and time, which in conjunction with the Vt product of the individual strobe signal produces a net direct current of zero value.

Temporally extending the length of the strobe signal pulse means an overlap in the addressing of successive electrodes in the first set of electrodes. Such an overlap increases the effective width of the switching pulse while leaving the other waveforms unaffected, thus reducing the total time required to address the entire display while maintaining a good contrast ratio between the elements in the two differently switched states.

Each strobe signal pulse may be immediately preceded by a smaller pre-pulse of the same or opposite sign with respect to the associated strobe signal pulse. This pre-pulse may be used to change the switching characteristics of the liquid crystal material. It may be used as part of a temperature compensation. In In this case, the temperature of the material is determined and the amplitudes of the pre-pulse are adjusted accordingly.

Each strobe signal pulse can be immediately followed by a pulse of the opposite sign.

According to this invention, a multiplexed addressed liquid crystal display comprises:

A liquid crystal cell comprising a layer of liquid crystal material between two cell walls, the liquid crystal material being a twisted chiral smectic material with negative dielectric anisotropy, the switching characteristics of which depend on both the product of voltage and time and the shape of such product, the cell walls carrying electrodes in the form of a first set of electrodes on one wall and a second set of electrodes on the other cell wall, the electrodes being arranged to together form a matrix of addressable (x, y) intersection points, at least one of the cell walls being surface treated so that the liquid crystal molecules align along the surface in a single direction;

Driver circuits for applying a strobe signal sequentially to each electrode of the first electrode set;

Driver circuits for applying data signals to the second set of electrodes;

Signal generators for generating a strobe signal and two data signals for controlling the driver circuits;

and means for controlling the sequence of data signals applied to the second set of electrodes when the strobe signal is applied sequentially to each electrode in the first set of electrodes to produce a desired display pattern, the time interval between the application of the strobe signal in sequence to each electrode being equal to the data signal period;

a data signal generator for generating two sets of signals of equal amplitude and frequency but opposite sign, each data signal form comprising DC pulses of alternating sign with a period of 2 ts,

and a strobe signal generator for generating a strobe signal of either a first pair of pulses of different amplitude and a second pair of pulses of different amplitude and opposite polarity to the first pair of pulses, or a clear pulse followed by a pair of pulses of different amplitude and periodic polarity reversal to maintain a net direct current of zero value, the end of the last pulse in the strobe signal pulse pair being of longer duration, > ts, than half the data signal period, 2 ts, and extending into an addressing period, 2 ts, of the next electrode, the arrangement being such that each intersection is addressed with a pulse of appropriate sign and magnitude to bring that intersection into the desired display state once during the complete display addressing period and at a total net direct current of zero value.

A simple analysis of the liquid crystal switching behavior (Liquid Crystal, 1989, Vol. 6, No. 3, pp. 341-347) yields the following equation for the field at which the Response time-voltage switching characteristics of the liquid crystal material result in a minimum for the response time.

Emin = Ps/ 3 ε₀ Δε sin² θ; (1)

where Emin is the field at which the response time- voltage switching characteristic of the liquid crystal material shows a minimum in the response time, where

ε0 is the dielectric constant in vacuum,

Δε is the (negative) dielectric anisotropy of the liquid crystal material,

θ is the cone angle of the liquid crystal material,

Ps is the spontaneous polarization.

This simple equation is only valid for some materials, and their values of Ps and Δε can be adjusted to achieve the desired operating voltages. Recent work (E.P. Raynes, The Physics of Displays for the 1990's in Fine Chemicals for the Electronics Industry II, Chemical Applications for the 1990's, pp. 130-146, Jones, Raynes and Towler, The Importance of Dielectric Biaxality for Ferroelectric Liquid Crystal Devices, 3rd International Conference on Ferroelectric Liquid Crystals, University of Boulder, Colorado, USA, June 24-28, 1991) has shown that dielectric biaxality is important for the existence of a minimum in the response time-voltage characteristic. The data for Figures 16-20 described below were determined experimentally.

The invention is described below by way of example only with reference to the accompanying drawings, of which:

Fig. 1 is a diagram of a time-multiplexed x,y matrix;

Fig. 2 is a cross-section of part of the display in Fig. 1 on an enlarged scale;

Fig. 3 is a plot of logarithm of time versus logarithm of voltage showing the circuit characteristics of a smectic material for two differently shaped addressing signals;

Fig. 4 to 8 show in diagrammatic form different strobe and data signals that can be used;

Fig. 9 shows signal diagrams that a strobe signal modified compared to Fig. 4 can take;

Fig. 10 shows diagrams of erase, strobe and data signals;

Fig. 11 shows strobe, data and addressing signals of a prior art display;

Fig. 12a, b show signal diagrams intended for addressing the 4 x 4 display elements in Fig. 13;

Fig. 13 illustrates an array of 4 x 4 elements forming some intersections that are in the ON state, with the rest in the OFF state;

Fig. 14, 15 show diagrams of the contrast ratio versus the applied voltage pulse width for two different materials;

Fig. 16 - 20 represent graphs plotting the logarithm of time against the logarithm of applied voltage, showing the switching characteristics of a material with different applied signal shapes;

Fig. 21 and 22 show different clear, strobe and data signals;

Fig. 23, 24 show row and column signals in a display according to the prior art;

Fig. 25, 26 show row and column signals in a modification of Fig. 6.

The display 1 in Fig. 1, 2 comprises two glass walls 2, 3, which are kept at a distance of about 1 - 6 µm by a spacer ring 4 and/or distributed spacers.

The electrode structures 5, 6 made of transparent tin oxide are located on the inner surface of the two walls. These electrodes are shown as rows and columns of an X, Y matrix but can also take other shapes. For example, radial and curved shapes for an r, θ display or segment shape for a digital seven-segment display.

A layer 7 of liquid crystal material is located between the walls 2, 3 and spacer ring 4.

Polarizers 8, 9 are provided in front of and behind the cell 1. Row drivers 10 and column drivers 11 apply voltage signals to the cell. Two sets of signals are generated to supply the row and column drivers 10, 11. A strobe signal generator 12 provides the row signals, and a data signal generator 13 provides the ON and OFF signals. OFF signals to the column drivers 11. All control of the timing and display format is carried out by a control logic unit 14. The temperature of the liquid crystal layer 7 is measured by a thermocouple 15, the output of which is applied to the strobe signal generator 12. The output of the thermocouple 15 can be applied directly to the generator or via a divider element 16, eg a programmed ROM chip, in order to vary a portion of the strobe signal pulse or the data signal.

Before assembly, the cell walls are surface treated in a known manner, e.g. by applying a thin layer of polyimide or polyamide, drying and if necessary cleaning and polishing with a cloth (e.g. rayon) in one direction, R1, R2. Alternatively, a thin layer of e.g. silicon monoxide can be vapor deposited at an oblique angle. These treatments ensure surface alignment of the liquid crystal molecules. The alignment/abrasion directions R1, R2 can be parallel or antiparallel. If necessary, unidirectional voltages are applied so that the molecular axis aligns along one of the two directions D1, D2, depending on the polarity of the voltage. An angle between D1, D2 of about 45º is ideal. If no electric field is applied, the molecules align in an intermediate direction between R1, R2 and the directions D1, D2.

The device can be operated in a transmissive or reflective mode. In the former, light passing through the device, eg from a tungsten bulb, is selectively transmitted or blocked to give the desired display. In the reflective mode, a mirror is arranged behind the second polarizer 9 to reflect ambient light back through the cell 1 and two polarizers. When using a partially reflective Mirror, the device can be operated in both transmissive and reflective mode.

Pleochroic dyes can be added to material 7. In this case, only one polarizer is needed and the layer thickness can be between 4 and 10 µm.

Suitable liquid crystal materials are:

Merck catalogue reference number SCE 8 (available from Merck Ltd. Poole, England) with a Ps of about 5 nC/cm2 at 30ºC, a dielectric anisotropy of about -2.0 and a phase sequence of -Sc 59ºC Sa 79ºC N 98ºC.

Mixture A contains 5% racemic dopant and 3% chiral dopant in the base material;

Mixture B contains 9.5% racemic dopant and 3.5% chiral dopant in the base material. Base material % by weight

Dopant (both racemic and chiral)

The * indicates chirality, without * the material is racemic.

Both mixtures A, B have a Ps of about 7 nC/cm2 at 30ºC and a dielectric anisotropy of about -2.3.

Mixture A has the phase sequence Sc 100ºC Sa 111ºC N 136ºC.

Mixture B has the phase sequence Sc 87ºC Sa 118ºC N 132ºC.

Liquid crystal material at the intersection of a row and column electrode is switched by applying an addressing voltage. This addressing voltage results from the combined application of a strobe signal Vs to the row electrode and a data signal Vd to the column electrode.

ie:- Vr = Vs - Vd,

where Vr = immediate value of the addressing signal,

Vs = immediate value of the strobe signal, and

Vd = immediate value of the data signal.

Chiral twisted smectic materials switch as a function of the product of voltage and time. This characteristic is shown in Fig. 3. Voltage-time products above the curve switch a material; below the curve is the non-switching region. Note that the switching characteristic is independent of the sign of the voltage; ie the material switches either at a positive or negative voltage of a given amplitude. The direction in which the material switches depends on the polarity of the voltage.

Two curves are shown in Fig. 3 because the switching characteristics depend on the shape of the combined addressing voltage pulses. The upper curve is obtained when the addressing voltage is immediately preceded by a small pre-pulse of opposite sign; e.g. a small negative pulse followed by a larger positive pulse. The material behaves in the same way when a small positive pulse is applied followed by a large negative pulse. This upper curve usually shows a turning point or minimum response time at a voltage. This is not evident from Equation 1 because the switching behavior is modified by the pre-pulse. The small pre-pulse can be referred to as the preceding pulse (Lp) and the larger addressing pulse as the following pulse (Tp). The upper curve is for a negative Lp/Tp ratio.

The lower curve is obtained when the addressing voltage is immediately preceded by a small prepulse of the same sign; i.e. a small positive pulse followed by a larger positive pulse. The same is true for a small negative pulse followed by a large negative pulse. The lower curve has a positive Lp/Tp ratio. This lower curve has a different shape than the upper curve; for some materials it does not need to have a minimum in a voltage-time curve.

The difference in the shape of the two curves allows unambiguous operation of the device over a fairly large time range. This is achieved by operating a device in a region between the two curves, e.g. as shown by the dashed lines. Intersections are addressed with an addressing voltage having a shape where the lower curve is applicable and where the voltage and pulse width are above the curve. Intersections that are not to be switched are either addressed with an addressing voltage having a shape where the upper curve is applicable and where the voltage and pulse width are below the curve, or just a data signal voltage. This is described in more detail below.

Figure 4 shows strobe, data and addressing signals according to an embodiment of the present invention. The strobe signal is initially zero for a time period ts, followed by +3 for two times ts. This repeats gradually on each line, i.e. during the time interval of a frame. The second part of the strobe signal is zero for a ts period, followed by -3 for twice ts. Again, this repeats gradually on each line during the time interval of a frame. The complete addressing of a display requires two time intervals of a frame. The values +3 and -3 are voltage units chosen for illustration, actual values will be given later for specific materials.

Data signals are arbitrarily defined as data-IN and data-OFF or D1 and D2. Data-IN initially has a value of +1 during a first time period of ts, followed by -1 during a time period ts. This repeats; i.e. data-IN is an alternating signal of amplitude 1 and period 2ts. Data-OFF is similar but has an initial value of -1 followed by +1; i.e. it is inverse to data-IN. The first part of the data waveform, e.g. for data-IN the value +1 during period ts, coincides with the first part of the strobe signal, i.e. zero during the period ts.

The addressing signal is the sum of the strobe signal and the data signal. The combination of a positive strobe pulse and data-ON is: -1, 4, 2, 1, -1, 1 etc. The value 4 immediately after -1 ensures that the switching characteristics of the material are determined by the upper curve in Fig. 3. The combination of a negative strobe pulse and data-ON is: -1, -2, -4, 1, -1, 1 etc. The combination of smaller pulses of the same sign as the large (-4) pulse ensures that the switching characteristics of the material are determined by the lower curve in Fig. 3. Similarly, a positive strobe pulse and data-OFF are: 1, 2, 4, -1, 1, etc; and a negative strobe pulse and data-OFF are composed of: 1, -4, -2, -1, 1, -1 etc.

When not receiving a strobe pulse, each row is grounded, i.e., at a voltage of 0. Each column is applied either data-ON or data-OFF. As a result, all intersections receive an alternating signal, due to the data signals when not addressed. This creates an AC bias at each intersection and helps to keep the material in the switched state. Larger AC biases result in improved contrast due to the well-known AC stabilization described in Proc 4th IDRC 1984, pages 217-220.

In addition, an AC bias, e.g. from a 50kHz source, can be applied directly to the lines that do not receive a strobe pulse. The effect of an AC bias on the contrast ratio, both in terms of magnitude and pulse width, is shown in the Figures 14 and 15 for materials SCE8 and Blend A. These have an inherent contrast ratio (CR) measured as a function of AC frequency when a cell is switched between its two bistable states, measured at different levels of AC bias.

Alternative strobe signals are shown in Figures 5 to 8. In Figure 5, the strobe signal is initially zero for 1 x ts and 3 for 3 x ts, followed by its inverse. In Figure 6, the strobe signals are initially zero for 1 x ts and 3 for 4 x ts, followed by their inverse. In Figure 7, the strobe signal is initially zero for 1 x ts, 3 for 2 x ts, and -3 for 1 x ts, followed by its inverse.

Figure 8 is a modification of Figure 4 in which a non-zero pre-pulse is used in the strobe signal. As shown, the first part of the strobe signal between -1 and 1 is not zero as in Figure 4. The remainder of the strobe signal is the same as in Figure 4, i.e. with amplitude 3 for twice ts. The resulting addressing signal is then a first pulse between -2 and -1 for both the first and second fields. The effect of this pre-pulse is to change the position of the switching curves in Figure 3 etc. Varying the value of the pre-pulse results in changing the shape and vertical position of the curves, as explained with reference to Figures 16 and 17 below. Table 8 below shows how the switching time varies with temperature. Such a change can be reduced by changing the pre-pulse amplitude.

Figure 9 shows a modification of Fig. 4. In this modification, the strobe signal is zero during the first ts and 3 during the next 1.5 ts. These 1.5 ts are only an example, since any larger value than ts up to 5 ts can be taken.

Figure 10 shows a single erase pulse of amplitude 4, which is applied for 4 ts. This brings all intersections into one switched state. A strobe signal is then used to bring selected intersections into the other switched state. Periodically the sign of the erase pulse and the strobe pulse is reversed so that an overall net DC voltage of zero is obtained. The use of an erase pulse and a single strobe pulse can be used in all figures 4 to 8. An advantage of the erase and strobe systems is that the entire display can be addressed in a single field time period.

For comparison, Figure 11 shows strobe, data and addressing signals of a state-of-the-art display using monopulse addressing.

Figures 21 and 22 show addressing modes of this invention using an erase pulse and a single strobe pulse at zero net DC current.

In Figure 21, the erase pulse is split into two, a pre-pulse of opposite sign with respect to the main pulse and an erase pulse. The function of the pre-pulse is to provide a DC balance of zero value. The pre-pulse has a value of 3 for 4 ts, immediately followed by -3 for 6 ts. The strobe pulse is first zero for 1 ts, immediately followed by 3 for 2 ts; this strobe signal is the same as the strobe signal in Figure 4. Data signals T1, T2 are also the same as in Figure 4. The combination of erase pulse and D1 or D2 gives a large negative Vt- Product which switches all pixels in the addressed row to the OFF state. The strobe pulse in combination with D2 switches the desired pixels to the ON state, as described above with reference to Fig. 4.

Figure 22 is similar to Fig. 21, but has a different shape of the erase pulse. This erase pulse has a pre-pulse of amplitude 3 for 4 ts, immediately followed by -4.5 for 4 ts. The strobe pulse has an amplitude 3 for 2 ts as in Fig. 4. The combination of the erase pulse and D1 and D2 gives a large negative Vt product which turns all addressed rows OFF. Again, selected pixels are turned ON by the strobe pulse and D2.

The erase pulses in Figures 21, 22 can be applied with other forms of strobe pulses as shown in Figures 5 to 9, the amplitude and/or the Vt product being selected to give a net direct current of zero value. In the example in Figure 8, where the first time period of the strobe signal is changed, e.g. as a function of the temperature, the amplitude of the pre- and/or main erase pulse is also adjusted to obtain a net direct current of zero value.

The erase pulse can lead the strobe pulse to varying degrees, but there is an optimal position for response time, contrast and visible flicker in the display. This is typically the case with erase pulses 6 lines ahead of the strobe pulse, but depends on material parameters and the details of the multiplexing type.

Figures 12a, b show the signals that occur when addressing a 4x4 matrix to display information as in Fig. 13. Filled circles have been chosen here to represent ON-switched electrode intersections, ie display elements, unmarked intersections are OFF. The addressing mode is that used in Fig. 4.

The positive or preceding strobe pulse is applied to each line 1 through 4 in alternation, forming the first field. After the last line is addressed by the preceding strobe pulse, the negative or succeeding strobe pulse is applied to each line 1 through 4 in alternation, forming the second field. Note that there is an overlap between the lines. For example, the third ts period on line 1 occurs at the same time as the first ts period on line 2. This overlap is more noticeable on displays using the strobe signals shown in Figures 5 and 6.

The data signal at data-ON at column 1 remains constant because every intersection in the column is always ON. Similarly, at column 2 the data signal is data-OFF and remains constant because all intersections in column 2 are OFF. For column 3 the data signal is data-OFF while rows 1 and 2 are addressed, alternating to data-ON while row 3 is addressed, then returning to data-OFF while row 4 is addressed. This means that column 3 receives data-OFF for 4 x ts, data-ON for 2 x ts, data-OFF for 2 x ts; one period of field time, the time required for the positive strobe pulse to address each row. Similarly, at column 4 the data signal is data-OFF for 2 ts, data-ON for 2 ts, data-OFF for 2 ts, and data-ON for 2 ts. This is repeated for another field period while the negative strobe pulse is applied. Two field periods are required for one image period and to fully address the display. The above is repeated until a new display pattern is required.

Resulting addressing signals are shown in Fig. 12b. At the row 1, column 1 (R1, C1) intersection, the material does not switch during the first field period because the switching of the material follows the upper curve in Fig. 3 and the time and applied voltage level are chosen to be below the switching curve. Instead, the material switches in the second field period when the material switches due to the lower voltage/time requirements represented by the lower curve in Fig. 3. A similar reasoning applies to the R1, C2 intersection where the material is switched during the first field period.

For the R3, C3 intersection, the material switches during the second field period because the selected time/voltage during the first field period does not reach the higher value required by the upper curve in Fig. 3. The R4, C4 intersection switches at the end of the second field period while a negative strobe pulse is applied.

The shape of the signals applied to column 4 causes difficulties. Due to the ON-OFF-ON-OFF pattern of the display, the data signal has a period twice as long as that of column 1, for example. This can mean a lower contrast ratio than that shown in Figures 14 and 15, where longer pulse widths (low frequency) give noticeably lower contrast ratios. In addition, the amplitude of the non-switching but large addressing pulse in the first field contrasts with a low amplitude of the switching pulse in the second field. Therefore, for reliable switching, a large Difference between the two switching curves e.g. in Fig. 3 necessary.

The contrast ratio (CR) curves in Fig. 14 (mixture A) and Fig. 15 (mixture SCE 8) indicate the inherent contrast of a device when switched between its two bistable positions with an AC bias applied. Of course, operation on the plateau of short pulse widths is desirable for both good contrast and uniform contrast. Since the multiplexed AC bias of the column waveform has varying frequency components depending on the pixel pattern, the contrast of the display can vary. This is particularly noticeable when all pixels are in one state (highest frequency components) and successive pixels are in the opposite state (lowest frequency components), where a factor of 2 difference in the column signal frequency results. These two cases are shown in Figures 12 and 13 for columns 1 and 4.

Figure 16 shows in logarithmic representation the time/voltage dependence of the switching characteristics for the material SCE 8 in a parallel rubbed cell with a layer thickness of 1.8 µm at a temperature of 25 ºC. The axes of the diagram are log ts and log pulse amplitude of the voltage.

The curves were recorded with a calibration cell simulating the addressing signals in Fig. 4. Two different addressing signals are used. The first, signal I, is a small negative pulse (-1) applied for a time ts, followed by a larger positive pulse (5) applied for a time 2 ts, ie the ratio Lp/Tp is -0.166. A period of 0 V is then followed by the inverse, ie a small positive pulse (1) and a negative larger pulse (-5). In addition, a 50kHz square wave signal is superimposed on the addressing to provide an AC bias and simulate a data signal. The small pulse is 0.166 times the value of the large pulse at all voltage levels when the trace was taken. This first addressing signal results in the upper trace. Time/voltage values above this curve cause the cell to switch, while values below the curve do not cause switching.

The second addressing signal II is first a positive small pulse of magnitude 1 which is applied for ts, immediately followed by a larger positive pulse of magnitude 4 which is applied for 2 ts. After a period of 0 V this is inverted. The small pulse is 0.25 times the value of the larger pulse, i.e. Lp/Tp = 0.25. Again a 50kHz signal is superimposed to produce an AC bias. This second addressing waveform gives the lower curve. Time/voltage values above this curve result in switching of the cell; while values below the curve do not result in switching. With a strobe signal voltage of Vs = 50 V, data voltage Vd = 10 V, the operating range is Vs - Vd = 40, switching at 52 µs, Vs + Vd = 60, switching at about 480 µs.

Figure 17 shows the time/voltage characteristics for the same addressing mode as in Fig. 16, particularly that in Fig. 4, but modified by a small pre-pulse in the strobe signal as in Fig. 8. Fig. 17 shows that the effect of the pre-pulse is to shift the vertical position of the curves. This is useful for temperature compensation, where shifts in the curves due to temperature variations are counteracted by changing the value of the pre-pulse.

In the upper trace, the simulation addressing signal first has a zero voltage for ts, followed by a larger positive pulse of 6 for 2 ts, i.e. Lp/Tp = 0. After a number of time intervals ts of zero volts, the inverse is applied to maintain a net zero DC voltage. A 50kHz signal is superimposed as an AC bias.

In the lower trace, the addressing signal first has a small positive pulse of 1 during ts, followed by a larger positive pulse of 2 during 2 ts, i.e. Lp/Tp = 0/5. This is later inverted in terms of polarity. A 50kHz signal is superimposed as an AC bias.

The operating range for Vs = 50, Vd = 10 is: lower curves, Vs - Vd = 40, switching at 42 µs, and for the upper curve Vs + Vd = 60, switching at about 500 µs.

Figure 18 is similar to Fig. 16 with respect to the identical cell, but uses simulations of the addressing signals in Fig. 5. Thus, the addressing signals are -1, 6, 4, 6 (Lp/Tp = -0.166) for the upper curve and 1, 4, 6, 4 (Lp/Tp = 0.25) for the lower curve. For Vs = 50, Vd = 10, the lower curve switches at 38 µs and the upper curve at about 210 µs.

Figure 19 is similar to Fig. 16 with respect to the identical cell, but uses simulations of the addressing signals of Fig. 7. The addressing signals are as shown, in particular for the curve whose points are marked with "+" the values -1, 6, 6, -6 (Lp/Tp = -0.166) and for the curve with the points marked with "0" the values 1, 4, 4, -4. The switching is complicated because the upper curve has a reentrant surface where the material switches at the follow-up pulse instead of the main pulse. For Vs = 50, Vd = 10, the lower curve, Vs - Vd = 40, switches at 58 to 240 and again at greater than 300 µs when switching to the follow-up pulse. The upper curve Vs + Vd = 60 shows no switching at 60 V. This results in multiplex operation at the main pulse between 58 and 240 µs and at the follow-up pulse at more than 300 µs.

For comparison, the logarithmic time/voltage characteristics in Figure 20 are given for a conventional monopulse addressing mode using a simulation of the strobe signal and data signals in Figure 11 in the same cell as Figure 11. For the upper trace, the simulation addressing signal is a negative pulse of 1 unit amplitude during ts, followed by positive 6 units during ts. For the lower trace, the addressing signal is a positive pulse of unit 1 during ts, followed by a positive pulse of 4 units during ts. The pulse amplitudes are described as units to indicate relative values. The traces are recorded at the voltages shown. For Vs = 50, Vd = 10 the lower curve Vs - Vd = 40 switches at 80 µs and the upper curve Vs + Vd = 60 switches at about 950 µs.

Details follow from the device characteristics for different liquid crystal materials and different addressing signals. A single pixel test cell was constructed and addressed with a simulation of a 50-line display. Different values for strobe, Vs and data Vd in terms of voltage amplitude were selected to achieve addressing voltage values so that switching voltages are above the lower curve in Fig. 3 and non-switching voltages are below the upper curve in Fig. 3, and the value of ts in µs was adjusted to achieve a clear switching display. This provided ensure that the cell was operated in the range indicated by dashed lines in Fig. 3. The contrast ratio value CR is the ratio of the light transmitted in one switched state to that transmitted in the other switched state; it is a measure of the clarity of the display. CR is measured at the extremes of the pulse width ts or at specified values of ts. CR was optimized by adjusting one of the switched positions of the axes in the liquid crystal so that it corresponded to the minimum transmittance.

In the following tables, the operating range of the time ts does not quite match the information given by the voltage/time graphs in Figures 16 to 20. There are three reasons for this. First, the simulations in Figures 16 to 20 are not without exception good for all display pattern possibilities. Second, with longer pulse widths and correspondingly long frame times, the operator can see flickering due to transient switching; this can be interpreted as non-multiplexing. Third, with longer pulse widths, the contrast ratio becomes lower, see Fig. 14, 15. For example, a CR of 2 results at 200 µs and it is therefore difficult to determine whether a material is switching or not.

Therefore, for practical displays, the upper time limit should be chosen as when an ad no longer switches properly. This can be much less than the actual switching time.

Material SCE 8 in a 1.8 µm thick layer at 25ºC. Table 1: Addressing type according to Fig. 4 Table 2: Addressing type according to Fig. 5 Table 3: Addressing type according to Fig. 7 Table 4: Addressing mode according to Fig. 11 (monopulse) Mixture B in a layer of 1.7 µm thickness at 30ºC Table 5: Addressing type according to Fig. 4 (at lowest ts) Table 6: Addressing type according to Fig. 5 Table 7: Addressing type according to Fig. 7 Table 8: Addressing type according to Fig. 4 at different temperatures Table 9: Addressing type according to Fig. 5 at different temperatures Table 10: Addressing mode according to Fig. 11 at 30ºC (monopulse) Material A in a layer of 1.7 µm thickness at 30ºC Table 11: Addressing type according to Fig. 4 Table 12: Addressing type according to Fig. 5 Table 13: Addressing type according to Fig. 7 Table 14: Addressing type according to Fig. 11 (monopulse) Material Merck, Catalogue No. 917 Temperature 30ºC; Vs = 60 V; Vd = 15 V Table 15 Addressing Type Fastest Time s Longest Time µs Operating Range (Time) Contrast Ratio (CR) Brightness (%) Fig. Operating range is: Longest Time/Fastest Time Brightness (%) compared to having no cell between parallel polarizers.

Material RSRE A206: temperature 30ºC, Vs = 30 V, Vd = 10 V. Table 16 Addressing type Fastest time µs Operating range (time) Contrast ratio (CR) Brightness (%) Fig.

Material RSRE A260 is: AS500:A151: 1:1 + 5 % dopant

AS 500 : A 151 1:1 + 5 % dopant

AS 500 is:

A 151 is: Dopant 2 % chiral 3 % racemic

The * indicates chirality, without * the material is racemic.

In nematic liquid crystal devices, it is known that row and column peak voltages can be reduced by applying additional waveforms to both row and column electrodes. For example, Figures 23, 24 show two different ways of reducing peak voltage in prior art monopulse drive systems in Figure 11. In Figure 23, a strobe (row) signal is alternatively zero for 1 ts and a positive pulse Vs for 1 ts in the first field, followed by zero for 1 ts and a negative pulse -Vs for 1 ts in the second field. The additional signal is positive Vs/2 during the first field, followed by -Vs/2 in the second field. The resulting strobe signal varies between Vs/2 and -Vs/2 as shown. The data (column) signals are alternatively Vd and -Vd, each for 1 ts. The additional signal applied to each column is Vs/2 for the first field, followed by -Vs/2 for a second field time. The resulting data signal varies between Vd + Vs/2 and -(Vs/2 + Vd) as shown. The effect of the additional signal is to reduce the peak voltage from, for example, 50 V to 35 V.

An alternative to Fig. 23 is shown in Fig. 24. As before, normal strobe pulses are zero for 1 ts and positive Vs for 1 ts in the first field time and zero for 1 ts and -Vs for 1 ts in the second field time. The additional signal is a square wave signal of period 2 ts, present during the first field time, followed by its inverse during the second field time, both varying between Vs/2 and -Vs/2. The resulting strobe (row) signals are as shown. Similarly, the data (column) signals are square, varying between +Vs and -Vd. The additional signal is the same as that present at the row electrodes. The resulting data (column) signal is as shown and varies between Vs/2 + Vd and -(Vs/2 + Vd). This in turn reduces the peak voltage required for the display drivers from e.g. 50 V to 35 V.

The same principles of Figures 23, 24 can be applied to the addressing mode of Figures 4 to 8 above. This is shown in Figure 25 which is a modification of the mode of Figure 5. Strobe pulses of zero for 1 ts are followed by Vs for 3 ts in the first field time. Strobe pulses of zero for 1 ts and then -Vs for 3 ts follow in the second field time. The strobe signal is shown for lines 1, 2, 3 and 4 of a four line display; two different strobes are shown for line 4 for reasons explained later. The additional signal applied to row (and also column) electrodes is shown as Vs/2 during the first field time, then -Vs/2 during the second field time. The resulting line signal for line 1 is shown as -Vs/2 for 1 ts, Vs/2 for 3 ts, -Vs/2 for 4 ts, Vs/2 for 1 ts, -Vs/2 for 3 ts, and Vs/2 for 4 ts in the first and second field times. The resultant of the strobe and additional signal and the line designated line 4a are shown with large peak values + and - 3 Vs/2. The reason for this is the extended strobe pulse length which extends into the adjacent field. To overcome this, line 4 is either hidden from view or addressed with a zero strobe voltage as indicated at line 4b. In a more practical example, e.g. a 128-line display, the generated signal was programmed for a 128-line display, but only 127 lines of the type shown in Fig. 25 were used. Even if a longer strobe pulse is used, eg as in Fig. 6, more lines will remain unused. Signals applied to column electrodes are shown in Fig. 26. Data 1 and its inverse data 2 are as in Fig. 5. The additional signal is Vs/2 for a field time and -Vs/2 for the second field time. The resulting column signal is shown alternating between +/- (Vd + Vs/2). Thus, for the mode of Fig. 5 with Vs = 50 V and Vd = 10 V, the mode of Figs. 25, 26 reduces the peak voltage to 35 V.

Claims (11)

1. Method for multiplex addressing a ferroelectric liquid crystal display made of a layer (7) of chiral twisted smectic material between cell walls (2, 3), the smectic material having a switching characteristic which depends both on the product of voltage and time and on the shape of such a product, the cell walls (2, 3) carrying a first set of electrodes (5) on one wall (20) and a second set of electrodes (6) on the other wall (3), which are arranged so as to result in a matrix (x, y) of addressable elements, which method comprises the steps: individual addressing of each electrode of the first electrode set (5), wherein the addressing is carried out either by applying two strobe signals of opposite polarity or by applying a clear signal followed by a strobe signal with periodic polarity reversal in order to maintain a net direct current of zero, Applying one of two data signals (D1, D2) to each electrode of the second electrode set (6) synchronously with the strobe signals, wherein both data signals (D1, D2) alternately assume positive and negative values and one data signal is the inverse of the other data signal and the period of the data signals, 2ts, is twice as long as that of a single strobe pulse, ts, within the strobe signals, wherein the resultant of one strobe signal and one data signal causes switching of the liquid crystal material and thereby addressing selected electrode intersections during a period equal to the data signal period, but the resultant of one strobe signal and the other data signal at non-selected electrode intersections does not cause switching of the liquid crystal material during a period equal to the data signal period, temporally extending the end of the strobe signal to the selective addressing period of the next electrode (5) of the first electrode set, while maintaining the same time period, 2ts, between the application of the strobe signals to successive electrodes (5), whereby each electrode intersection is selectively addressed for a time greater than 2ts, and each electrode intersection is addressed with a resulting signal of appropriate sign, magnitude and time to bring that intersection (x, y) into a desired display state at a total net direct current of zero value. 2. The method of claim 1, wherein the strobe signal is a zero voltage in a first ts time period and a non-zero voltage during a period greater than ts, followed by several periods ts of zero voltage, thereby representing a frame period followed by a similar signal with reversed polarity. 3. The method of claim 1, wherein the strobe signal is a non-zero voltage in the first time period ts, such voltage being less than the following voltage and variable in amplitude for compensating for material switching characteristics with temperature variations. 4. Method according to claim 1, in which an additional signal is applied to both sets of electrodes (5, 6) in order to bring about a reduction in the voltage peak applied to the electrodes (5, 6). 5. Method according to claim 1, in which the erase pulse is divided into two sections of opposite polarity, the voltage-time product (Vt) of which, combined with the voltage-time product of the strobe pulse, results in a net direct current of zero value. 6. A multiplex addressed liquid crystal display comprising: a liquid crystal cell (1) comprising a layer (7) of liquid crystal material between two cell walls (2, 3), the liquid crystal material (7) being a twisted chiral smectic material with negative dielectric anisotropy, the switching characteristics of which depend both on the product of voltage and time and on the shape of such a product, the cell walls (2, 3) having electrodes in Form of a first set of electrodes (5) on one wall (2) and a second set of electrodes (6) on the other cell wall (3), the electrodes (5, 6) being arranged such that together they form a matrix of addressable (x, y) intersection points, at least one of the cell walls (2, 3) being surface treated so that the liquid crystal molecules align along the surface in a single direction; Driver circuits (10) for applying a strobe signal to each electrode of the first electrode set (5) in succession; Driver circuits (11) for applying data signals (D1, D2) to the second set of electrodes (6); Signal generators (12, 13) for generating a strobe signal and two data signals (D1, D2) for controlling the driver circuits (10, 11); and means for controlling (14) the sequence of data signals applied to the second set of electrodes (6) when the strobe signal is applied to each electrode (5) of the first set of electrodes (5) in sequence to produce a desired display pattern, the time interval between the application of the strobe signal to each electrode (5) in sequence being equal to the data signal period; a data signal generator (13) for generating two sets of signals of equal amplitude and frequency, but opposite sign, each data signal comprising DC pulses of alternating sign with a period of 2 ts, and a strobe signal generator (10) for generating a strobe signal of either a first pulse pair with different amplitude and a second pulse pair with different amplitude and opposite polarity to the first pulse pair, or an erase pulse followed by a pulse pair with different amplitude and periodic polarity reversal to maintain a net direct current of zero, wherein the end of the last pulse in the strobe signal pulse pair is of longer duration, > ts, than half the data signal period, 2 ts, and extends into the addressing period, 2ts, of the next electrode (5), the arrangement being such that each intersection point is addressed with a pulse of corresponding sign and magnitude to bring that intersection point into a desired display state once during the complete display addressing period and at a total net direct current of zero. 7. A display according to claim 6, wherein the strobe signal generator (12) is for producing a zero voltage during a first time period ts, then a non-zero voltage during a period greater than ts, followed by a plurality of periods ts of zero voltage representing one frame period, followed by a similar signal of reverse polarity. 8. Apparatus according to claim 6, further comprising devices (12, 13) for generating an additional signal and applying such an additional signal to both sets of electrodes (5, 6). 9. A display according to claim 6, further comprising a temperature sensor (15) for sensing the display temperature and means for changing addressing voltages. 10. A display according to claim 9, wherein the strobe signal generator (12) serves to generate a non-zero voltage in the first time period ts, this voltage being lower than the adjacent following voltages and variable in amplitude and sign for compensating for material circuit characteristics depending on temperature variations. 11. Display according to claim 6, in which the strobe generator (12) serves to generate an erase pulse with two sections of opposite polarity, the voltage time product (Vt) of which, combined with the voltage time product of the strobe pulse, results in a net direct current of zero value.