Optical design tool for backlit display

Backlighting systems require some conversion of light from one or more sources to produce the desired distribution of light distribution in a region or at a fixed angle. The lighting design software must be capable of geometric modeling, setting optical characteristic parameters for different types of light sources and conversion units, and must be able to use optical tracing methods to evaluate the path of light through the model and calculate the final light distribution. Light distribution uses Monte Carlo simulation to calculate illuminance, brightness, or luminescence intensity for a particular area and/or angle. Light is emitted from the light source at random locations and angles, traced through the optical system, and received on the receiving surface. Illuminance can be calculated from the surface receiver and the intensity can be obtained from the far field receiver. By defining a luminance meter on the receiver surface, the distribution of brightness with space and angle can be calculated. In some cases, it may be important to analyze the color of the display. Specify the spectral energy distribution of the light source (such as the light-emitting diode), output the CIE coordinate value and the correlated color temperature (CCT), quantify the chromaticity of the display, and generate a RGB real light rendered graphic on the display. These analyses can be done in LightTools software.

The characteristics of the backlit display have special requirements for the lighting analysis software. As will be explained, the light emitted by the backlight depends on the distribution density of the printed dots, or the distribution pattern of the microstructure. Modeling a particular microstructure array can result in very large model sizes if the CAD model is used directly. LightTools software provides the ability to define 3D texture arrays for accurate ray tracing and rendering. Since there is no directly constructed geometry model, the model is smaller and ray tracing faster. Another aspect of backlight analysis involves the splitting and scattering of light on the surface of the light guide. Since the Monte Carlo method is used to simulate lighting effects, it is possible to use a large amount of ray tracing to obtain a design with sufficient accuracy. The most effective way is to trace the light with the highest energy. Trace the highest energy ray path by using the splitting probability and use the target area or scatter angle of the scatter surface to direct the scattered light to the "important" direction (eg, toward the viewer of the display).

What is backlighting?

A typical backlight consists of a light source such as a cold cathode fluorescent lamp (CCFL) or a light emitting diode (LED), and a rectangular light guide plate. Other available components include a diffuser plate to improve display uniformity and a brightness enhancing film (BEF) to increase the brightness of the display. The light source is typically located on one side edge of the light guide to reduce the thickness of the display. Sidelight illumination typically uses total reflection (TIR) ​​to conduct light in the display.

Backlight designers have multiple ways to model light sources in LightTools software. Different shapes of fluorescent light sources (such as straight, L-shaped, U-shaped or W-shaped, as shown in Figure 2) can be quickly defined using the fluorescent light creation tool. The reflector of the lamp can be defined by various geometric prototypes in the LightTools software, such as cylinders, elliptical grooves, extruded polygons. Reflectors defined in the CAD system can also be imported into LightTools software via standard data exchange formats (IGES, STEP, SAT and CATIA). If LEDs are used, designers can select the desired LED model from the product models of Agilent, Lumileds, Nichia, and OSRAM pre-stored in LightTools software. Once the light enters one side of the light guide, the problem becomes to extract light from the light guide perpendicular to the direction of propagation.

As shown in FIG. 3, the brightest light guide plate is a side close to the light source, and the farther the distance, the darker the brightness in the light guide plate. In order to obtain a uniform light output, the light extraction efficiency must increase as the distance increases. One of the main tasks of backlight design is to design a light guide that changes the efficiency of light extraction as needed. There are two extraction techniques available. The dot printing light extraction technique prints a dot matrix structure at the bottom of the light guide plate to scatter the light upward and to emit from the surface of the light guide plate. The second technique, compression molding light extraction, relies on total reflection (TIR) ​​of the bottom surface microstructure to cause light to exit the surface of the light guide.

LightTools software provides a backlight design tool to achieve the design of the light guide. This tool (Figure 4) assists the user in creating various components of the backlight. Other options include adding a light source/reflector component to the model, BEF modeling, and building a receiver to analyze the brightness. The backlight tool interface is a multi-label that is used to set and modify various types of light extraction mechanisms.

For backlights that use dot-printing light extraction methods, the backlight tool can set linear variations in the size and aspect ratio of the printed dots, as well as linear variations in dot spacing along the length of the light guide. This linearly varying structure is often a good starting point for display uniformity, but not sufficient to meet the final uniformity requirements. Further control of uniformity can use non-linearly varying light extraction parameters. One method that uses the fewest parameters and the control is very flexible is to define the parameter variables of the quadratic Bezier curve. The 2D area tool of LightTools software can be used to set up nonlinear structures. Figure 5 shows an example of the use of print extraction in which three parameters (print dot width, height and vertical spacing) are varied to achieve different extraction behavior. The output uniformity is shown in Figure 6. The image on the right shows that the average output brightness is a constant.

The second extraction method, the compression molding extraction technology, uses the three-dimensional texture function of LightTools software, which makes the ray tracing of the repetitive structure very effective and the storage information is very compact. Models created by non-three-dimensional textures have ray tracings that are more than 30 times slower than models created with three-dimensional textures, and files are more than 100 times larger. Three-dimensional textures are available in three basic shapes: spherical, prismatic, and pyramidal (Figure 7). The backlight tool can define a linearly variable microstructure. But the 3D texture tool can change the texture parameters nonlinearly using a quadratic Bezier curve. The example shown in Figure 8 is a trough-shaped microstructure (modeled using prismatic 3D texture) as the extraction mechanism. The resulting light guide plate and its simulation results are shown in FIG.

Optical calculation of backlight

The two most important optical quantities of a backlit display are the display brightness and illuminance uniformity of the surface on the light guide. It is also important to calculate the luminous intensity and various color metrics (CIE coordinates and correlated color temperature CCT). LightTools software has these built-in calculations and many other features to help understand the data generated by Monte Carlo simulation.

Monte Carlo simulation is the basis for the calculation of illumination by LightTools software. The random number generator is used to select the starting position, direction and wavelength of the light for sampling the light distribution on the receiving surface. The choice of "random" numbers can greatly affect the convergence of the simulation. Using a low difference (Sobol) number sequence (which is not completely random), the error can be reduced to 1/N, where N is the number of rays at the receiving end. A comparison of the chromaticity using a random number sequence (Fig. 10) and a Sobol number sequence (Fig. 11) can be seen. In this example, simulation results with a random 128,000 rays are equivalent to the accuracy of Sobol's 16,000 rays. It is important to compare the simulation convergence speeds of different software. What we are concerned with is the speed at which some kind of simulation accuracy is achieved, not the speed at which a certain amount of light is traced. In LightTools software, the receiver is used to collect ray data to calculate illuminance.

The ray data analyzed and displayed is collected from the data grid. Users can interactively control the size or number of data grids. For a given amount of light on the receiver, the fewer the number of grids, the lower the spatial and angular resolution, but the higher the relative accuracy (low error rate). Conversely, the more the number of meshes, the higher the spatial and angular resolution, but the lower the accuracy (high error rate). The estimated error rate is displayed on each grid to help the user decide if enough light is used for the trace simulation to meet the resolution and accuracy required by the design (Cassarly, WJ, Fest, EC, and Jenkins, DG, 2002). If more light is needed, the user can continue to simulate interactively until the target is reached.

An important aspect of backlight analysis is the splitting and scattering of the surface of the light guide. The function of the light guide plate is that light can be absorbed or emitted after being reflected on the inner surface multiple times. If the light is split into two parts, transmissive and reflective, on each contact surface, it will cause a very large amount of spectroscopic light, most of which does not carry much energy, thus slowing down the analysis speed. An example of this is shown in Figure 12, which shows a starting ray, many paths due to splitting.

The simulation below uses 2000 incident rays, and the receiver collects 277,948 rays due to splitting (Figure 13). Since most of the light reaching the receiver does not have much energy, the resulting error is 42%. Conversely, if the Fresnel loss factor and surface scattering characteristics are used to determine the probability of transmission and reflection of light, the likelihood of the path path is evaluated, and most of the time for ray tracing will be used to track the energy in the system, thus Speed ​​up the analysis. A simulation result of a 200,000 incident ray is shown in Figure 14. In this case, 118,969 rays arrive at the receiver with a calculation error of 6%. Using probabilistic mode ray tracing reduces the computational error by 7 times while reducing the computation time by 42%.

Conversely, if the Fresnel loss factor and surface scattering characteristics are used to determine the probability of transmission and reflection of light, the likelihood of the path path is evaluated, and most of the time for ray tracing will be used to track the energy in the system, thus Speed ​​up the analysis. A simulation result of a 200,000 incident ray is shown in Figure 14. In this case, 118,969 rays arrive at the receiver with a calculation error of 6%. Using probabilistic mode ray tracing reduces the computational error by 7 times while reducing the computation time by 42%.

Finally, in order to improve the uniformity of display, a diffusion plate is sometimes used on the top surface of the light guide plate. Since the diffuser diffuses the light to a wider angle, less light is scattered into the aperture of the luminance meter, and according to the conventional display brightness test method, a very large amount of light is required for the brightness calculation. LightTools software maps the target area or angle to the scattering surface so that the user can specify which scatter should be considered. This is an important form of sampling and another way to improve the convergence of Monte Carlo simulations. Figure 15 shows a luminance meter and a backlight with a diffuser plate with no specified target angle. Tracking 2,000 rays, the luminance meter receives 40 rays, and the raster of the spatial brightness is as shown.

The same example is shown in Figure 16, but sampled by significant values, the target angle is specified on the diffuser. The target angle matches the acceptance angle of the aperture of the luminance meter. When the light reaches the diffuser, LightTools software will generate scattered light (the luminous flux into the target area based on the angular distribution of the diffusion model) into the target angle, allowing the dither to collect all the scattered light, thus improving the convergence of the simulation. In this case, 2000 incident rays, 1416 rays (71%) are received by the luminance meter.

Other considerations

Backlights are widely used in liquid crystal displays (LCDs), which are a type of polarizing component. Modeling polarization components such as linear polarization, quarter-wavelength slices, and polarization tracking evaluation are critical factors for successful analysis. LightTools software provides a simple linear polarization and block model, as well as the Jones-Mueller matrix specification for polarization components. The user can use the polar ray tracing function when needed to track the polarization of the light according to the Stocks vector.

There are often optical coatings on the assembly that have various transparency, reflectance and polarization properties. Coatings are defined in LightTools software based on their performance, which is often the only information the user knows. The average or individual S or P values ​​of reflection and transmission can be specified by any of the following two parameters: angle of occurrence, wavelength, X position, or Y position. The system provides tools to convert the coated laminate into the LightTools software coating format.

While most backlights use dot printing or compression molding light extraction techniques, other methods are possible. One is to use particle scattering in the light guide plate. If the size and density of the particles are properly controlled, Mie scattering from the particles can effectively extract light from the light guide (Tagaya, et al., 2001: 6274). LightTools software can batch simulate the scattering of spherical particles according to Mie theory, or distribute the simulated scattering according to a user-defined angle.

Exporting a complete optical design to a CAD system is often a necessary step in the manufacture of a light guide. LightTools software supports standard format conversions such as STEP, SAT, orIGES. Since the data conversion standard only supports external geometric data, in the case of the compression molding design extraction, it is necessary to convert the three-dimensional texture definition into an external geometric data output. LightTools software supports standard formats and can selectively convert 3D textures into external geometry data, which includes the entire backlight design in the conversion file.

summary

Backlight design technology has been continuously improved and developed to provide better performance and lower cost in the market. This innovation requires lighting design software to continually add new features, especially to shorten the backlight design cycle. The main functions of LightTools software, such as model creation and file size, ray tracing and simulation time, and the ability to calculate a large number of optical parameters related to backlight design, have been recognized and verified by the industry.

The LightTools software version 5.0, released in 2004, includes noise-optimized illumination optimization, which is very practical in backlight design. This feature automatically defines light extraction templates for maximum efficiency and uniformity. In addition, LightTools software's backlight template optimization tool provides an efficient way to optimize the output distribution for backlights and light guides.