科技论文题目Display Physics of Passive LCDs系别尚德光伏学院专业微电子技术（液晶显示技术与应用）班级0902学生姓名赵俊学号090425指导教师丁兰2012年4 月Display Physics of Passive LCDs1. Working principles1.1 Twisted Nematic (TN)The principle of a TN-display has been explained in chapter2. For completeness, the essentials are repeatedhere. The schematic of a TN-display is shown in Figure1The situation on the left side represents an LCD in a state without an electric field over the LC. The LC is contained in a space between two glass plates. The glass plates have a thickness of 0.4 – 1.1 mm. Both glass plates have a transparent electrode, usually made of ITO (Indium Tin Oxide). These ITO layers have a thickness of up to 1000 Å. Between the LC and the ITO layer is a polymer layer (thickness < 1000 Å). This layer is often called the orientation layer: the polymer layer is rubbed in a certain direction and the director of the molecules close to the wall is oriented parallel to the rubbing direction. For the standard TN-effect the rubbing directions on both glass plates make an angle of 90° with each other. The result is that the director undergoes a 90° twist from one glass plate to the other, as indicated in the figure. In principle the liquid crystal does not automatically know which way to rotate, so it can also rotate over -90°instead of 90°. This is prevented by the addition of a small amount of acholesteric liquid crystal. A cholesteric liquid crystal has apart from a preferred direction as the nematic material, also an inherent twist in a certain direction. By addition of a small quantity of this cholesteric material with the correct sense of twist, the rotation over −90° is prevented.The thickness of the liquid crystal layer is of the order of 4 to 10 µm. The choice of the thickness is usually a compromise between the properties of the LCD and the available LCs. To fixate the thickness of the layer there are spacers between the glass plates, usually in the form of plastic balls or glass fibres.For the desired modulation of the light we need in addition two sheets of polariser material. For the discussion here we have placed the polarising direction of the sheets parallel to the belonging rubbing direction.The electrical properties of this system can be described accurately if the electrical and elastic properties of the materials used are known. The main parameters are:- Dielectric constants (two of them, ε// and ε⊥)- Elastic constants (three: splay, twist, bend, or K11, K22 and K33)- The total twist (φ) and tension of the helix, (thickness/natural pitch, or d/p0)- Surface tilt (θp)- Rotational viscosity γ1Since there is no basic difference between TN and STN in describing the director deformations, the TN case will be handled along with the STN configuration.1.2 Super Twisted Nematic (STN) Effect1.2.1 BackgroundAs the number of lines in passively addressed (multiplexed) displays increases, the difference between the ON and OFF voltages becomes verysmall. For this reason, the TN device is impracticalfor large information displays withconventional addressing schemes. This problem was solved in the mid 1980's with the invention of the super-twisted nematic (STN) display. In this device, the director rotates through an angle of up to 270 degrees, compared with the 90 degrees for the TN cell. The effect of twist angle on the electro-optical response curve is shown in Figure 2As most things in life, the improved steepness doesn’t come for free. With increased twist, the LCD tends to respond more slowly to an electric field. Also, the stability of the helix is considerably lower than for TN, causing defects like “fingerprint texture”, and worst of all, the black/white switching of the TN-effect is replaced by a coloured effect. In the next sections, the influence of display parameters on various display properties is evaluated. The occurrence of defects is discussed in section 3.To sustain a twist-angle greater than 90°requires a nematic liquid crystal with an intrinsically twisted structure known as “Chiral Nematic”. Chiral nematics are ordinary nematic liquid crystals doped with a small percentage of optically active (cholesteric) material. The “handedness“ (chirality, asymmetry1) of the dopant molecules results in an intrinsic, macroscopic twist of the whole nematic structure. The amount of twisting is characterised by the pitch length, p, which is the length along the helix axis required for a full 360° turn. The amount of chiral dopant can bechosen to make the natural pitch, p0, exactly fit the boundary conditions as defined by the rubbing directions of the LCD. However, as we will see later,usually conditions are chosen where the natural pitch is set to be larger than the boundary conditions. The value chosen is indicated as the d/p ratio.1.2.2 STN threshold voltage and curve steepnessThe main reason for STN displays to develop is, as mentioned above, the increase in effect steepness. As twist increases, so does steepness. Note that the change in the tilt angle can become very abrupt as the twist angle is increased above a certain value. The consequence of this increased steepness is that the off and on voltages are much closer together, as is shown in Figure 3.This in turn means that more lines can be addressed.Although it is desirable to obtain a sharp electro-optical transition, greyscale images require intermediate points along the curve. For this reason, many commercial STN displays use a twist angle around 240°.This broadens the transition region enough for greyscale, while still allowing for conventional multiplexed addressing.Parameters influencing the electrical response of an STN display are the elastic constants K11, K22 and K33, (splay, twist and bend), the dielectric ratio ∆ε/ε// , pre-tilt angle (θp), total twist (φ) and d/p. The resulting (normalised) threshold voltage of the electro-distortional curve (i.e. mid-plane tilt angle vs. applied voltage) for zero tilt boundary conditions can be calculated using the following formula [3].Or, taking pre-tilt into account [4]And the initial steepness (slope) of the curve is given byOf course, with many parameters involved, there is much room for optimisation. By changing thparameters mentioned, vastly different shapes of the electro-distortional curve can be generated. Figure 4 to 4f give an overview of the parameter variations: a: variation of total twistb: variation of pre-tiltc: variation of d/pd: variation of K33/K11e: variation of K22/K11f: variation of ∆ε/ε// (sometimes also confusingly called γ)1.2.3 Switching dynamicsAs mentioned, the response to an electric field of STN displays is slower than TN. For TN, the switching times can be described as:*****These relations also hold for STN displays, with the remarks that increased twist increases ton as well as toff. This clearly shows that:- There is a strong (quadratic) influence of cell gap on both switching times,- Viscosity and elastic constants play a secondary role,- These equations neglect the consequences of starting- and end-voltage inmultiplexed displays (driving force ∆εE is only considered at ton)This will also lead to the next obser-vation: Since viscosity and elastic constants are temperature dependent, switching times will also show this dependence. Viscosity as well as the elastic con- stants will change with temperature, so it is not immediately apparent which of the two will gain the upper hand. Upon closer examination it has been shown that the temperature dependence of the elastic constants is an order of magnitude lower than that of the viscosity, and in fact dependent upon the material used (Figure 5), so temperature dependence of switching times is governed by temperature dependence of viscosity effects.1.2.4 Temperature dependenceAbove, it already became apparent that a display property like switching times isstrongly temperature dependent. In some cases, the behaviour is predic- table, like temperature dependence of the order parameter S. Other para-meters are less predictable.An example is the ratio K33/K11, which plays an important role in threshold voltage and steepness.Depending on the type of material used, K33/K11 can increase with temperature, or decrease (Figure 5). Of course, the separate values of K33 and K11 will both decrease with increasing temperature, but not at the same rate.For most properties, general temperature dependencies are known and fixed, and the result on LCD performance can therefore be predicted. Others, like d/p, or rather p, because d can be assumed to be constant, can be tuned to requirements.The paramet er with the largest influence on electrical behaviour is ∆ε. There is a direct relation to the order parameter, and since the order parameter decreases quickly with increasing temperature, so does ∆ε.The result is a strong temperature dependence of the threshold voltage, especially near the clearing point(Figure 6).Since this behaviour is highly undesired, much is done to counteract this problem. One possible route is to choose the temperature dependence of the elastic constants in such a way that temperature dependence of ∆ε is counteracted. Another option is to use a combination of cholesteric dope to tune temperature dependence of d/p to the point of counteracting other temperature dependencies, and thus create a temperatureindependent threshold voltage.1.2.5 Frequency dependenceWhen AC-fields are used, it immediately becomes apparent that almost every electrical component behaves frequency dependent. Usually, this is caused by a combination of resistive and capacitive components in the electric circuit, such as:- ITO-track resistance- Contact resistance- Liquid crystal resistance- Driver output resistance- Parasitic capacitance- Pixel capacitanceThese parameters behave predictably, and, although they can pose problems, solutions usually exist. For further reading see section 3: crosstalk.Another ground for frequency dependence has a different cause: The dielectric constant of a liquid crystal is frequency dependent. This is best visible for ε//, which has the highest value and therefore relaxes at the lowest frequency. At a certain frequency, the changing dipoles of the LC-molecules can no longer keep pace with the electric field, and the dielectric contribution "relaxes" i.e. disappears (Figure 7). Ifε//disappears, this leaves ε⊥, which means that the sign of the dielectric anisotropy inverts: ∆ε becomes negative. In the past, attempts have been made to use this effect to quickly turn off an “on”-pixel. The idea was abandoned when it became apparent that a low relaxation frequency, which was needed for this “trick” also brought with is severe limitations in the addressability of these displays.To make matters worse, the relaxation frequency is temperature dependent, moving to lower frequencies at lower temperatures. At subzero temperatures, frequencies can be as low as a few hundred Hertz, making a display virtually unaddressable!See also chapter 3, section 5.3.2.2. Characterisation(Based on: IEC norm)Although the optical behaviour of an LCD may be calculated directly from its physical parameters and their temperature behaviour, calculation is usually not accurate enough to fully describe the resulting display. Optical characterisation is usually needed to describe or specify a display.Since the measurement result is depending on the conditions during measurement (illumination, type of electrical signal, temperature, measurement angle, detector optics, etc), many types of measurement have been meticulously specified, and laid down in measurement standards. In addition, the LCD industry has defined some “workable” combinations of parameters, conveniently describing a specific property of an LCD. Several of the most widely used measurements will be described here. Three different types of light measuring devices may be used as detector systems for the light transmitted and / or reflected by the 'device under test' (DUT): a luminance meter, tri-stimulus photometer or a spectro-radiometer. The optical system is schematically shown in Figure 8 and shall allow for measurement of well-defined spot sizes (field of view) on the DUT.When measuring matrix displays these meters should be set to a circular or rectangular field of view that includes more than 500 pixels on the display under normally incident observation (the standard measurement direction). The total acceptance angle of detection by these meters, θaccept shall be less than 5°(see Figure 8). This can e.g. be obtained by use of a measuring distance between the meters and display area center of 50 cm (recommended) and a diameter of the detector pupil of 4 cm. In case of measuring segment displays, the field of view should be set to a single segment, and not include any of its surroundings.Viewing direction and viewing angle range are given by polar coordinates θ and ϕ as defined in Figure 9. We refer to θφ = 0 as the 3 o'clock direction (the "right"), θφ = 90 as the 12 o'clock direction ("upward"), θφ = 180 as the 9 o'clock direction ("left") and θφ = 270 as the 6 o'clock ("bottom"). In the standard measurement direction, the ph otometer observes the DUT under vertical viewing angle (θ = 0o). While scanning θ and/or ϕ, the center of the measuring spot on the DUT shall stay fixedFigure 9: Definition of polar coordinates θ, ϕ. 12 o'clock pertains to the "top" of the display area,3 o'clock to the "right" of the display area (as viewed under "normal" viewing conditions).Any condition (either measuring spot on the DUT, meter aperture angle, viewing angle, meter spectral sensitivity, resolution etc.) that is not compliant to the required condition described in this shall be recorded in the detail specifications.If the DUT is not equipped with its own source of illumination, external illumination shall be supplied in either of two ways:- By means of an externally applied diffuse light source with specified (spatial and angular distributionof) luminance and spectrum. (This is e.g. used for measurements on direct view displays);- By means of an externally applied directional light source with calibrated spatial uniformity of illumination at the plane of the DUT, full opening angle of illumination at the location of the measuring spot in the plane of the DUT of less than 30 o, and (if needed) calibrated spectral intensity distribution in the visible wavelength range. (This is mostly used for measurements on projection-display modules).In both cases, records of the light source (intensity distribution, temporal stability, opening angle, etc.) and its distance to the device under test shall be added to the detail specification.The temporal drift in light intensity shall be less than 1% of the stabilized value per minute. Care shall be taken that the temperature of the DUT has stabilized and is notaffected by the illumination system. The temperature of the DUT shall be measured and specified.2.1 Transmission – voltage curvesCategory: world standard2.1.1 PurposeDetermination of the optical luminance response of an LCD when energised by an electric field.2.1.2 InstrumentationA luminance meter, power source, light source, driving signal generator and a recorder.2.1.3 Measurement methodMeasurements are conducted under dark conditions. The detector (luminance meter) is calibrated on maximum display luminance. The driving signal is applied to the LCD and increased from a starting value to an ending value. (an alternative is to start at a high voltage and decrease the voltage gradually)Care should be taken that the voltage increments are small enough not to let the switching time of theLCD influence the static luminance reading of the luminance meter.2.1.4 DefinitionsThe brightest condition at any given angle is defined as the 100% level, the darkest state 3is defined to be 0%. Between these values, the luminance levels are defined: The voltage where the first 10% point change occurs at increasing voltage is defined as V10 (not, as may be inferred for a normally white display, V90).The general notation for the voltages measured is of the form Vxx,yy,zz, where xx is the transmission level, yy is the measured angle and zz is the measurement temperature.2.1.5 Specified conditionsThe records of the measurement shall be made to describe deviations from the standard measurement conditions and include the following information:- Driving signals (waveforms, voltage limits and frequency).- Filtering applied to remove frame-response and refresh artifacts.- Data sampling and recording rate- Measurement angle, azimuth and DUT temperature2.2 Response characteristicsCategory: world standard2.2.1 PurposeDetermination of the time needed to change from light to dark (dark to light) by application of the driving voltage.2.2.2 InstrumentationThese typical times are measured using a luminance meter with sufficient frequency response, power source, driving signal generator, a trigger signal generator and a recorder.2.2.3 Measurement methodMeasurements are performed under darkroom standard measurement conditions. The electrical signal of the detector is conducted to the recorder. The display is driven by an invertible plain field signal 4 from a signal generator. Upon inverting, the signal must go from start level to end level without displaying any intermediate level on the display. The frequency of inversion must be low enough to allow the display to obtain optical equilibrium in each of the two states. A trigger signal is sent to the recorder upon inversion of the video at position 0. If the trigger signal has another timing (e.g. at the beginning of the field),corrections shall be made to allow for the scanning time t s needed for video-signal inversion to arrive at position 0. The luminance meter measures the optical response. Ripples in the detected signal due to not relevant effects (e.g. originating from the display frame-frequency) shall be eliminated from the response.The luminance in the WHITE (100 % input data-signal) mode is chosen as 100 % and in the BLACK (0 % input data-signal) mode as 0 %.2.2.4 DefinitionsThus, for all types of LCDs, the times needed for the luminance to change from 0%→90% ( t1) and from 100% →10% ( t2) are measured (see Figure 10 and/or chapter 2, section 4.5). Also, at the same position, the times needed for the luminance to change from 10% →90% ( τ1) and 90% →10% ( τ2) are measured. We now define the turn-on an turn-off time tOn and t off for the normally white (normally black) DUT according to.Both on- and -off times as well as rise- and fall-times are examples of (dynamic) response times, also called “switching” times; in other words, the (dynamic) response time and switching time are general terms that are NOT strictly defined. The difference between turn-on and turn-off times on the one hand and rise and fall times on the other is called the delay time.2.2.5 Specified conditionsThe records of the measurement shall be made to describe deviations from the standard measurement conditions and include the following information:- Driving signals (waveforms, voltage and frequency).- Filtering applied to remove frame-response and refresh artifacts.- Dynamic properties of light measuring device and data sampling and recording- If use is made of the term switching-time or (dynamic) response time, explanationof the use should be given in the detail specification and deviations from the above prescribed nomenclature should be given when using other names for any of these times.2.3 Operating voltages, Recommended V oltage Area (RVA)Category: Philips MDS2.3.1 PurposeDetermination of the driving voltage of a multiplexed passive matrix display required to obtain a specified contrast between specified angles over a specified temperature range.2.3.2 InstrumentationLuminance under driven conditions is measured using a luminance meter, power source/light source, thermostatic chamber, driving signal generator and a recorder. 2.3.3 Measurement methodTo start with, the angles at which the op-erating conditions will be specified are determined. Normally this would be something like perpendicular or 10°, as in Figure 11 and at, say, 40° in the optimal direction. In Figure 11: V on has to be higher than the Vrms at B and V off lower than the Vrms at A in order to obtain sufficient contrast in the specified viewing angle range.Then, the operating limits are set: As an example (Figure 11), the non-selected (“off”) pixels should not exceed V10 while selected (“on”) pixels should be at least above V50 . In the figure: VOn has to be higher than the Vrms at B and V off lower than the Vrms at A.Then, the operating voltages V op (see below) that correspond to these limits are measured, either by applying a multiplex signal and measuring the two conditions with off- and on-state signals respectively (sometimes, this is even determined v isually, hence “visual RVA”), or simply by measuring the limit conditions by applying a square wave, and calculating back to the multiplex-voltages by dividing by the applicable factors, thus using a signal with the same Vrms as a multiplex signal. This measurement is then performed for a number of temperatures. Since the optical properties of an LCD are not linear with temperature, two curved lines willresult if the measured limit conditions are graphed versus temperature. See Figure 12.Now, one of two possible situations is true:- The “on” condition dictates a lower voltage than the “off” condition. This is the desired situation, since a voltage exists where the off-segments are less visible than the limit (lower than the “upper limit”) while the o n-segments can be driven at a voltage higher than the minimum requirement (higher than the RVA “lower limit”).- The “on” condition dictates a higher voltage than the “off” condition. In this case, the display cannot be driven according to the specificati on, since the “off” segments will be strongly visible at the voltage required to drive the “on” segments at sufficient contrast..3.4 Definitions w.r.t.the recommended voltage area- The “upper limit” denominates the voltage at which the “off-pixels/segments become visible above a specified level.- The “lower limit” denominates the level at which the on-segments exceed a specified level of transmission.2.3.5 Specified conditions w.r.t.the recommended voltage area- The angle and transmission level fo r the “upper limit”.- The angle and transmission level of the “lower limit”- The temperature range.Summarising: A description is presented of a method to define a 'region' bordered by two viewing angles and two voltage curves, in which the diplay can be driven within specification. As well as the existence of an “RVA”, the curves in Figure 12 give an indication of the temperature compensation that should be used to obtain good performance over the whole temperature range.2.4 LUMINANCE CONTRAST RATIOCategory: world standardAs discussed in chapters 2 and 3, contrast ratio is nothing but the quotient of the display luminance in the “bright” and in the “dark” state. However, these luminances can be influenced by a lot of factors. Some of these are mentioned below (defects), some are the consequence of the display layout.Taking “global” defects into account, one can generally distinguish between “pixel contrast ratio”,defining the pure, electro-optic effect, offset by general cell-defects, and the “m ulti-pixel contrast ratio, taking the inter-pixel gaps into account.2.4.1 PurposeDetermination of the luminance contrast ratio (in short "contrast ratio") of the LCD device.2.4.2 InstrumentationLuminance meter, driving power source and driving signal generator.2.4.3 Measuring MethodThe contrast ratio of LCDs is not directly measurable with non-imaging detectors 5. It has to be evaluated as the quotient of two luminance values that are sequentially measured in the bright and the dark state of the display. It must be assured that all conditions besides the electrical driving remain constant during the measurement of both luminance values (e.g. temperature, illumination, etc.).Measurements are performed under dark room and standard measurement conditions. The luminance meter is positioned under θ = 0 °.2.4.4 Plain field contrast ratioThe module is driven by a plain field test pattern signal from a signal generator (every pixel the same driving signal). The luminance is measured at the center of the DUT in the WHITE mode and in the BLACK mode, leading to L white,and L black respectively. Care shall be taken that the reading of the luminance meter due to stray-light is less than 1% of L black.The plain-field contrast ratio CRPF is defined as.2.4.5 Specified conditionsThe records of the measurement shall be made to describe deviations from the standard measurement conditions and include the following information:- Driving signals (waveforms, voltage and frequency).3. DefectsIn general, d efects are “unwanted” visible imperfections in the display. Obviously, defects come in many ways. They can be caused by dust particles (visible to the micron level, if enough light is scattered off the dust particle), air bubbles, defective active matrix pixels, holes in layers or electrodes. Sometimes, defects are caused by the configuration of the display, if the molecules move in an unwanted direction. If this happens, usually a “disclination line” ( line where the macroscopic orientation of the liquid crystal suddenly changes) is formed between the correct and the incorrect configuration. It usually is this disclination which is the first visible defect.It is this kind of defect which is discussed here.3.1 Orientation defects (static)3.1.1 reversed tiltReverse tilt is the general name for a category of configuration defects which do not show up when the display is unenergised. Everything seems fine, until the display is driven. Then, parts of the display switch differently, where LC aligns to the field in different directions, generating disclination lines and areas within the pixel with opposite optimum viewing angles.The basic cause of the problem is usually an incorrect rubbing direction (180°off), which causes a net tilt of 0°in the middle of the cell gap. Alternative root causes can be found in non-parallel fields at the edge of pixels or across gate lines in active matrix displays (Figure 13).One of the annoying properties of a disclination line is that its energy content is very high. There is a strong tendency to make itself as short as possible. This means, that if a fringe field causes a very thin disclination line along two neighbouring edges of a pixel, this will grow into a line across the entire pixel,unless there is another factor preventing this from happening (Figure 14). This factor is usually the pre-tilt,but can also be a spacer, “pinning” the disclination in position.Reversed tilt can occur in almost any LCD system. In vertically aligned nematics, if no pre-tilt is created, pixels tend to switch in entirely random directions. As with the planar oriented counterpart, a minute pre-tilt can correct the situation. However, particles (spacers!!) can locally disturb the situation, again with a localised deviation from the intended configuration.3.1.2 Reverse twistAn LCD is fully defined by imposing the boundary conditions (rubbing directions, tilt) and the way the director changes between these boundary conditions. If left to itself, the director will move over the smallest angle from one boundary to the other. Imposing the correct twist angle has to be done by adding the correct amount of cholesteric liquid crystal, usually referred to as “dope”. If the concentration is too low, the liquid crystal will choose to rotate over an angle which is the intended angle minus πradians (180°). Notwithstanding the fact that this is only exactly the opposite situation of what is intended in the case of 90°TN (for a 240° STN, the “opposite”situation is 60° ), we still call this“reversed twist”, or sometimes“left-right regions”. The situation is very visible, since the configurations behave optically very differently. However, care should be taken, not to include all optical defects under this name: The situation where one of the glass plates locally orients homeotropically looks very similar to reversed twist, but has nothing to do with dope concentration!An interesting aspect of this is, that the pitch induced by a cholesteric dope is usually。
