Optimized Method of Addressing a Liquid-Crystal Screen and Device for Implementing It

The present invention relates to a method of addressing a liquid-crystal screen allowing a display of uniform quality over the entire line of the screen, as well as to a device for implementing this method.

A liquid-crystal screen consists of a set of image elements ("pixels", standing for picture element), each formed by an electrode and by a counter electrode framing the liquid crystal, the value of the field between these electrodes altering the optical properties of the liquid crystal. The voltage at the terminals of the electrodes of the pixels is delivered via addressing columns by peripheral circuits ("drivers") by virtue of the control transistors of these pixels, the conducting or non-conducting state of these transistors being determined by selection lines coming from other line drivers.

FIG. 1 represents a selection line Lj of a liquid-crystal screen with m lines and n columns, controlling the transistors T1 to Tn of the pixels P1 to Pn. This line is connected to a line driver which delivers, at A, the square selection signal V.sub.A (t) as represented in FIG. 2. The signal V.sub.A (t) causes the transistors T1 to Tn of the line L.sub.j to conduct, and thus allows the electrodes of the pixels P.sub.i to be polarized by the video signal coming from the columns C.sub.1 to C.sub.n. The capacitances C.sub.c1 represent the capacitive couplings between the line L.sub.j and the counter electrode CE through the liquid crystal. This line L.sub.j, the end of which is floating, constitutes a delay line which causes distortion of the selection signal at point B by comparison with point A; this signal V.sub.B (t) at point B is represented in FIG. 2. This is visible particularly when it is desired to display a uniform image and when the same voltage is applied to all the columns C.sub.1 to C.sub.n of the screen. At the instant t.sub.F, the voltage at the terminals of the capacitances C.sub.p formed by the electrodes of the pixels P.sub.i and the counter electrode CE is the same. However, after the instant t.sub.F this is no longer the case due to the difference between the shapes of the signals V.sub.A (t) and V.sub.B (t).

This is because, at point A, the voltage drop is very fast, the transistor T.sub.1 is therefore turned off immediately after t.sub.F. Moreover, a stray capacitance C.sub.p exists between the line L.sub.j and the pixels P.sub.i. The voltage drop .DELTA.V.sub.G at point A thus, by capacitive coupling, causes a voltage drop on the pixel which is:

If V.sub.1 is the voltage supplied to the pixel P.sub.1 by the column C.sub.1, the voltage drop .DELTA.V.sub.1 on the pixel at the instant when the transistor T.sub.1 ceases to conduct is illustrated by FIG. 3a, V.sub.ce being the voltage on the counter electrode.

At point B, the phenomenon of capacitive coupling is identical, but in this case the transistor T.sub.n continues to conduct as long as the voltage V.sub.B (t) is greater than V.sub.1 +V.sub.t, where V.sub.t is the threshold voltage of the transistor. The coupling .DELTA.V.sub.n between the line L.sub.j and the last pixel P.sub.n is therefore weaker than .DELTA.V.sub.1, since, as long as the transistor T.sub.n is conducting, the voltage at the terminals of the pixels remains equal to the voltage delivered by the column C.sub.n. The capacitive coupling thus causes a voltage drop for the pixel P.sub.n :

.DELTA.V' being the voltage drop at point B.

The voltage which allows the pixels to alter the optical properties of the liquid crystal is therefore V.sub.pix1 =V.sub.1 -V.sub.ce in the case of the pixel P.sub.1 and V.sub.pixn =V.sub.n -V.sub.ce in the case of the pixel P.sub.n, V.sub.pix1 being different from V.sub.pixn. It is this which is represented in FIG. 3b. The grey level is therefore not the same at the start and at the end of line. This problem called "horizontal shading" is particularly important in the case of large-size screens.

One solution frequently used, and described in the document SID 94 Digest, page 263, consists in using a counter-pulse to reduce this effect. This solution is expensive since it requires more complicated drivers to be produced.

Another solution frequently used consists in reducing the resistivity of the lines. However, this implies increasing the thickness of the metal used to produce the line, which renders the process more expensive and more difficult to keep control of.

The present invention proposes a simple and effective solution to this problem of "horizontal shading".

The method according to the invention in fact consists in periodically scanning each line with a signal of time-varying voltage, each period of which consists of a plateau and a preferably negative slope the value of which is less than the value of the characteristic slope of the delay line at the end of line.

These characteristics can easily be implemented by virtue of drivers having a VDD analogue input allowing the high level VH to be controlled, such as, for example, the Toshiba drivers of the T6A02/T6A03 type.

Moreover, this method also makes it possible to reduce the coupling and thus the stray voltages on a screen.

The present invention will be better understood and its additional advantages will emerge on reading the description which will follow, illustrated by the following figures:

FIG. 1, already described, is a diagram of an example of lines of a liquid-crystal screen.

FIG. 2, already described, represents the selection signal as it is received at the end of line and at the start of line, and illustrates the problem posed by the delay of the line,

FIGS. 3a and 3b represent the voltages of the pixels at the start and end of line,

FIGS. 4a and 4b represent the signals according to the invention respectively, received at the start and end of line respectively,

FIGS. 5a and 5b represent the voltages of the pixels controlled according to the invention at the start and end of line respectively,

and FIG. 6 represents the shape of the reference high level of a driver allowing the invention to be implemented.

An embodiment of the present invention is represented by FIG. 4a, and consists in altering the shape of the signal delivered by the selection circuit so as to compensate for the delay effect of the line responsible for the horizontal shading. After a plateau of a width, for example, of 28 .mu.s, and according to one important characteristic of the invention, the signal V.sub.A (t) does not decrease abruptly (after a plateau of duration t.sub.F -t.sub.i), but, from t.sub.F, with a slope .alpha. preferably less than or equal to the characteristic slope of the delay line at point B, that is to say that .alpha. is less than .DELTA.V/.tau., .tau. being the characteristic time of the delay line at B and .DELTA.V the potential drop at point A. An example of the value of .alpha. may be a few volts per .mu.s. This signal thus decreases until the voltage V.sub.A (t) is equal to V.sub.F', at which voltage the transistors T1 to Tn are turned off. From this instant t.sub.F' the signal drops instantaneously.

Thus, between t.sub.F and t.sub.F' (the duration t.sub.F' -t.sub.F may be equal to 3 .mu.s for 6 volts, for example), the signal is the same at point A and B, all the transistors of the line maintaining constant voltages on the pixels. The selection signal, with delay, complete with a slope .alpha. between t.sub.F and t.sub.F', is represented in FIGS. 4b.

From the instant T.sub.F', the transistors T.sub.1 and T.sub.n are turned off, the coupling is therefore .DELTA.V.sub.1 =.DELTA.V.sub.2 =C.sub.p /C.times..DELTA.V. The voltages at the terminals of the pixels P.sub.1 and P.sub.n are illustrated respectively by FIGS. 5a and 5b. It will be noted that the voltages on the pixels P.sub.1 to P.sub.n are equal and consequently that there is no horizontal shading.

A refinement of the method consists in using, between t.sub.F and t.sub.F', a curve which is not a straight-line portion but a portion of a function f(t) which remains unchanged by the transfer function of the delay line: applying f(t) to T.sub.1 results in applying f(t-T) on T.sub.n, T being a delay. f(t) may, for example, be a sinusoid or a sum of sinusoids.

This method according to the invention can be implemented by a driver having an input which makes it possible to control the output current. By severely limiting the output current between t.sub.F and t.sub.F', it is possible to alter the standard signal so as to obtain the desired waveform.

It is also possible to use drivers which have an analogue input which makes it possible to define the high level V.sub.H. The desired signal is obtained at the output of the driver by modulating this input in such a way as to obtain a wave V.sub.H having an inverse sawtooth shape as illustrated by FIG. 6. That is to say, at each line 1, 2, 3, 4, etc., the high level V.sub.H is maintained on a plateau over a line period up to the instant T.sub.F, then lowered linearly until the instant T.sub.F', when instantly raised back to the said plateau in order to scan the following line.

The present invention can be used for repairing flat liquid-crystal screens. In fact, known repair procedures exist, but they do not work as they increase the RC of the repaired line, which renders it visible since it does not experience the same coupling as the adjacent lines. By using the larger of the characteristic times of the repaired line or normal lines as .tau., the repaired lines become similar to the adjacent lines.

The present invention applies to the control of flat liquid-crystal screens including peripheral or integrated drivers, and in particular to large-size screens.