Measurement Instruments

After the fabrication of lightguides, its geometric characteristic and optical performances are very important to verify the simulation results. For the geometric analysis, optical microscope (OM) is utilized to observe the shape of micro-groove structures and to measure the size of lightguides.

The optical performances, such as angular distribution and uniformity of emitted light, are measured by Conoscope. Conoscope is a measurement system which utilizes optical lens based on Fourier transform to transfer the light beams transmitted (or reflected) from the platform of different angles to the CCD array, as shown in Fig. 3.4.

Each light beam transmitted from the platform with a incident angle θ will be focused on the focal plane at the same azimuth and at a position x=F(θ). Therefore, the angular properties of the sample are measured in a simple procedure without any mechanical movement. Besides, there are collimated and diffuse light sources depending on users’ requirements. Furthermore, moiré pattern regardless of the intensity distribution can be observed by Conoscope with the geometric analysis.

By equipped with a fast photometer system and a high sensitivity spectrometer, the functions of Conoscope are extended to compose of not only the simultaneous measurement of luminance and chromaticity versus viewing direction, evaluation of the data yields, i.e. luminance contras ratio, grey-scale inversion and reduction, color shift and many more parameters, but also the spectra and temporal luminance

variances.

Fig 3.4 Schematics of Conoscope.

3.5 Summary

The diamond turning of fabrication technology for lightguides has been introduced. Also, the double scanning driving system for resolving image crosstalk was proposed. The measurement of optical performances for lightguide is utilized by Conoscope whose principle has been introduced.

Chapter 4

Simulation Results and Discussions

4.1 Introduction

Based on the principle described in chapter 2, the simulation models are established for evaluating the performance of the dual directional backlight system. In this chapter, some parameters are obtained by Y. M. Chu [30] and others will be discussed respectively before the simulation. Then, some simulation models for maintaining the optical performances such as uniformity, angular distribution and suppressing moiré pattern will be discussed respectively. Finally, the moiré-free dual directional backlight system for time-multiplexed mobile display can be derived.

4.2 Simulation Software

The optical simulator Advanced Systems Analysis Program (ASAP^TM), developed by Breault Research Organization (BRO) was used to optimize the dual directional backlight system and simulate its angular distribution and light distribution on top surface of the backlight.

4.3 Simulation Model of Moiré-Free Dual Directional Backlight System

In order to evaluate the whole effect of 3D optical performances, a complete simulation model is needed. There are two main simulation models in the following simulation. One is parallel alignment between color filter and two lightguides as shown in Fig 4.1, the other one is vertical alignment between color filter and two lightguides as shown in Fig 4.2. The common points of these two simulation models

have two dual lightguides which are composed of numbers of micro-groove structures and the light source located to face the inclined surfaces of micro-groove structures.

Then, a detector which is to detect the optical performance including angular distribution and uniformity is set above two lightguides. Finally, a color filter which is to observe the moiré pattern with the lightguides by the geometric analysis positioned above the detector. The major different point of these two simulation models is the alignment between two lightguides and color filter. In the parallel alignment case, the panel is an portrait panel; in the vertical alignment case, the panel is a landscape panel.

Fig 4.1 Simulation model with parallel alignment between two lightguides and color filter.

Fig 4.2 Simulation model with parallel alignment between two lightguides and color filter.

4.3.1 Simulation Setting

Among the simulation settings, the light source is set as a lambersian distribution of a planar light source which is decayed as a factor of cos²θ from the normal direction [34]. The light source is to simulate the light bar composed of four LEDs and a lightguide. The refractive index of lightguides is set as 1.49 which is the same as that of PMMA. The lightguide and panel size is 36 × 32 × 1 mm³ and 35 × 28 × 1 mm³, respectively.

In chapter 2, the shape of each micro-groove structure which is a right triangle has been introduced, as shown in Fig 4.3. The inclined angle of each micro-groove is fixed as 38° by Y. M. Chu [30]. Other parameters including groove width, groove gap and groove pitch which are the variables will be discussed in the following simulation models.

Fig 4.3 The shape of micro-groove structures and the parameters related to groove.

There two kinds color filter built in the following simulation models. One is for parallel alignment between color filter and two lightguides, the other one is for vertical alignment between color filter and two lightguides, as shown in Fig 4.4. The simulated color filter is a simplified design, since it can provide the period to observe the moiré pattern in geometric analysis by ASAP. The practical configuration and size of color filter which can be observed by optical microscope (OM) will be illustrated in chapter 5.

(a) (b)

Fig 4.4 Color filter in the simulation model (a) parallel and (b) vertical alignment.

4.3.2 Crucial Properties in Simulation

distance, uniformity, light efficiency and moiré pattern are the most crucial issues for the simulation. These factors are discussed in detail below.

Angular Distribution

In order to display the parallax images correctly, the angular distribution of the dual directional backlight must be well controlled within each viewing direction, which depends on the viewer` s position. In the time-multiplexed 3D mobile display, the viewer` s position can be predicted roughly 10 to 40 cm normal from the display panel. The corresponding angular distribution of backlight is calculated as the range between ±4.7° to ±18° by assuming the distance between two eyes is 65 mm.

Therefore, most of light must be guided to the two symmetric viewing cones respectively and sequentially when the light sources are fast switched sequentially.

Furthermore, the crosstalk which is defined as the ratio of undesired image to desired image should be reduced as possible. In this case, the crosstalk can be written as equation (4-1). The crosstalk of less than 10% is the general criterion to form the 3D perception. In fact, the inclined angle of micro-groove is the main factor of the angular distribution. The variation of angular distribution by changing the micro-groove structures with groove width, groove gap and groove pitch is very small.

Viewing Angle and Viewing Distance

Viewing angle can be defined as the viewing region whose crosstalk is lower than 10%. By means of viewing angle, viewing distance can be calculated by the relationship of viewing angle and the distance of two eyes (65 mm).

Uniformity

Uniformity affects not only the image quality of single eye but also the whole 3D effect, defined as the ratio of the minimum to the maximum brightness by choosing the nine points of the panel. In general, the uniformity is considered as total hemisphere above the panel to be evaluated the general 2D wide viewing angle.

However, the viewing angle of this display can be roughly predicted among 0° to 25°.

The evaluation with total hemisphere above the panel may cause the inaccurate simulation. Therefore, uniformity here is set as the region between the angles 0° and 25° from the normal direction. Thus, the equation of uniformity can be written as:

cone

Light efficiency is a critical issue in time-multiplexed display. In general, each frame is divided into left-eye sub-frame and right-eye sub-frame, respectively. The brightness accepted by eyes can be predicted as a half of the general 2D panel.

Particularly, by the driving method mentioned in chapter 3, the light sources are in off state in half of each frame time. Therefore, the brightness is half than the original time-multiplexed display, i.e. the brightness in this case is only one-fourth than the general 2D panel. So the light efficiency can be defined as:

light

Moiré pattern which is an undesired phenomenon in the time-multiplexed display

can be observed while the simulation model is built up by the geometric observation mode of ASAP. Compared with the practical situation, moiré pattern by geometric observation mode of ASAP is strongly improved than the practical one. But the trend of moiré pattern distribution can be predicted. By means of this mechanism, moiré pattern can be reduced in advance.

4.4 Simulation Results

Several simulation results which include micro-groove structures distribution, angular distribution, uniformity, moiré pattern observation will be illustrated in the following.

4.4.1 Fixed Groove Width Design

This model is designed for parallel alignment between color filter and two lightguides as shown in Fig 4.4 (a) by Y. M. Chu [30]. The concept of maintaining uniformity is fixed groove width which is 25 μm. The gradually changed function of groove gap and x coordinate is shown in Fig 4.5. The light source is put in the left side of lightfuide. The groove gap is distributed between 368 μm and 25 μm from the groove nearest to the light source to the farthest one.

Fig 4.5 Distribution function of groove gap and x coordinate by the fixed groove width design.

After simulation, the light efficiency is about 60%. Simulation results including angular distribution and uniformity are shown in Fig 4.6.

Fig 4.6 Simulation results designed with fixed groove width by Y. M. Chu. (a) angular distribution and (b) uniformity.

Moiré pattern by geometric analysis is shown in Fig 4.7. By the theory of moiré pattern mentioned in chapter 2, moiré pattern has the maximum period when structures have the identical or double periods. The period of color by a portrait panel is about 70 μm. Therefore, when the periods of lightguide are close to 70 μm and 140 μm, moiré pattern has the maximum period, i.e. the serious moiré pattern is shown in the panel. Furthermore, the groove gap is gradually changed in a small amount, and then it causes the magnification of moiré pattern in the middle of panel because the groove gaps are almost the same in the middle of panel for two lightgudes.

Fig 4.7 Moiré pattern simulated results with the fixed groove width by the parallel alignment between color filter and two lightguides.

4.4.2 Discrete Micro-groove Distribution

In order to solve moiré pattern, the discrete micro-groove structures should be take into consideration. Here, we adopt the above-mentioned model designed by Y. M.

Chu, and then cut each continuous micro-groove structures into 20 discrete micro-groove structures for simulation. The design concept of discrete micro-groove structures is shown in Fig 4.8. The even parts of micro-groove structure are shifted to right hand side with a spacing of 0.7 mm. Particularly, the discrete parts in the region between 15 mm and 29 mm are also extended 1 mm along the top of each discrete micro-groove structure. The purpose of extending the discrete part is to break the period of the most serious moiré pattern which is happened in the middle of panel.

Moiré pattern is shown in Fig 4.9. Compared with Fig 4.8, moiré pattern is reduced by the discrete micro-groove distribution, but the region and width of moiré pattern are not reduced. The solution is to design a randomly discrete micro-groove distribution.

However, the discrete micro-groove distribution has the issue of fabrication process.

randomly discrete micro-groove structures are complex on both design and fabrication process. Therefore, a simple design of continuous micro-groove structures is needed.

Fig 4.8 Top view of the discrete micro-groove structures distribution design.

Fig 4.9 Moiré pattern simulated results with the discrete micro-groove distribution by the parallel alignment between color filter and two lightguides.

4.4.3 Fixed Groove Gap Design

The micro-groove distribution must be gradually changed for maintaining uniformity. There are three variables can be altered: groove width, groove gap, and groove pitch as mentioned before. In this model, groove gap is fixed and groove width is gradually changed by the distance between the micro-groove and light source. Here

we set the groove gap is 85 μm. The reason is to skip the regions of near 70 μm and 140 μm which are identical and double periods of color filter. The function of groove width and X coordinate is shown in Fig 4.10. The distribution function is divided into two parts to avoid the moiré pattern. The groove width is distributed between 5 μm and 40 μm from the groove nearest to the light source to the farthest one. Most of groove width is concentrated between 5 μm and 10 μm. Therefore, the light efficiency which is 54 % is slightly lower than the previous design with fixed groove width.

Fig 4.10 Distribution function of groove width and x coordinate by the fixed groove.

gap design

Simulation results including angular distribution, uniformity are shown in Fig 4.11.

Because the distribution function is discontinuous, the discontinuous place is expected to be very obvious while the viewer observes the lightguides. Moiré pattern is shown in Fig 4.12. Compared Fig 4.7 with Fig 4.9, moiré pattern by geometric observation is obviously narrowed. However, moiré pattern in the practical view is still not well enough (the practical view will be illustrated in chapter 5). There is a phenomenon of color separation because the colors on color filter are superimposed by moiré pattern.

lightguides as mentioned before. Also, the uniformity is needed to be improved.

Fig 4.11 Simulation results designed with fixed groove gap. (a) angular distribution and (b) uniformity.

Fig 4.12 Moiré pattern simulated results with the fixed groove gap by the parallel alignment between color filter and two lightguides.

4.4.4 Fixed Groove Pitch Design

The vertical alignment between color filter and two lightguides is adopted. Also, in order to keep uniformity, the design with fixed groove pitch is chosen. The distribution function of groove width and X coordinate is shown in Fig 4.13. The groove pitch is set as 150 μm and the groove width is distributed between 10 μm and 25 μm from the groove nearest to the light source to the farthest one. Simulation results of optical performances including angular distribution and uniformity are shown in Fig 4.14. The uniformity is obviously improved and has no discontinuous region. The angular distribution of light source 1 is not very ideal because the groove pitches of top and bottom lightguides are identical. Some emitted light of bottom lightguide is blocked or scattered by the micro-groove structure of top lightguide.

Therefore, the angular distribution of light source 1 is broader than the one of light source 2, i.e. some light is emitted to undesired direction and then not only causes crosstalk but also decreases the light efficiency. The results of image crosstalk will be measured in chapter 5. Besides, the average groove width is smaller than the design

on the bottom of lightguides. Therefore, the light efficiency is about 59% nearly the previous design by Y. M. Chu. Moiré pattern by geometric observation which is shown in Fig 4.15 is not very close to the practical case. From the analyses of moiré pattern introduced in chapter 2, there is magnification of moiré pattern period while the periods’ difference between two structures is very small. Besides, the inclined angle is also very sensitive in the small periods’ difference. Therefore, the alignment of two lightguides is very important. The color separation is solved by the vertical alignment between color filter and two lightguides.

Fig 4.13 Distribution function of groove width and x coordinate by the fixed groove pitch design.

Fig 4.14 Simulation results designed with fixed groove pitch. (a) angular distribution and (b) uniformity.

Fig 4.15 Moiré pattern simulated results with the fixed groove pitch by the parallel alignment between color filter and two lightguides.

4.5 Summary

By evaluating several models of dual directional backlight system, the fixed groove pitch should be the best solution for moiré pattern and uniformity. Also, the vertical alignment between color filter and two lightguides can solve the color separation phenomenon. The practical results mentioned above will be illustrated in chapter 5.

Chapter 5

Experimental Results and Discussions

5.1 Introduction

The optical performances including viewing angle, viewing distance and uniformity are examined by Conoscope and discussed in this chapter. The optimized directional lightguides with continuous micro-groove structures distribution were fabricated by diamond turning, whose specifications are the same as demonstrated in chapter 4. Moreover, the images are demonstrated to observe the moiré pattern in this chapter.

5.2 Measured Results of Lightguides

The designs of fixed groove gap and fixed groove pitch whose angular distribution on each point of panel is very similar to those introduced in chapter 4. Therefore, the angular distribution is measured at the center of panel. Uniformity is measured at five points, which are randomly chosen from the five specified regions on backlight, as shown in Fig. 5.1. The definitions of viewing angle and viewing distance are identical with the ones in chapter 4. Additionally, the practical photos of lightguides and moiré pattern comparison will be discussed in the following.

Fig. 5.1 Measurement of uniformity.

5.2.1 Fixed Groove Gap Design

The angular distribution of fixed groove gap design is measured as shown in Fig.

5.2. The viewing angle can reach ±7° to ±80°. The viewing distance can be calculated within the range of 0.57 cm and 25.47 cm. However, the viewing distance lower than 7 cm is hard to distinguish the image on the panel. The viewing distance can be roughly defined as the range of 7 cm and 25.47cm. Furthermore, the viewing distance of 7cm corresponds to the viewing angle of 25°. Therefore, the effective viewing angle is between ±7° to ±25°. Uniformity is calculated as 70%. The photos of lightguides and lightguides with panel are shown in Fig. 5.3. The lightguide for right eye image is positioned below the one for left eye image. Therefore, moiré pattern can be observed in the right image, as shown in Fig. 5.3(a). The micro-groove structure distribution is discontinuous, thus there are the discontinuities in both two images.

The panel includes color filter whose periodical structure causes the additional moiré pattern with two lightguides, as shown in Fig. 5.3(b). Furthermore, the color separation can be observed because colors on color filter superimpose the different colors by moiré pattern. Therefore, the rainbow-like color separation is occurred in the two sides of panel because the periods of moiré pattern are larger in these regions.

In order to improve the moiré pattern, color separation and uniformity, the 30%

haze diffuser is added on the lightguides. The angular distribution of fixed groove gap

design with 30% haze diffuser is shown in Fig. 5.4. The viewing angle is within the range of ±9° and ±39°. The viewing distance is between 7 cm and 20.52 cm.

Uniformity can reach 81%. The above-mentioned optical performances are acceptable for 3D display. The photos of lightguides with 30% diffuser and lighguides with panel and 30% diffuser are shown in Fig. 5.5. After adding 30% diffuser on lightguides, moiré pattern is nearly invisible. However, the discontinuity can still obviously be observed, as shown in Fig. 5.5(a). After that, the panel is added with the lightguides and 30% haze diffuser. The diffuser efficiently suppress moiré pattern and resolve the color separation, as shown in Fig 5.5(b). Similarly, the discontinuity can still be observed.

Fig. 5.2 The angular distribution of lightguides by fixed groove gap design.