Development of a DLPTM Common Engine Platform
چکیده
A DLPTM projector common engine platform allows the same engine design to be shared by several companies to develop different projectors. This approach also allows some of the components to be standardized and the economy of scale can be applied to reduce the cost of the engine. The collaboration between Texas Instruments (TI) and Opto-Electronics & Systems Laboratories of Industrial Research Institute (OES/ITRI) to develop a common DLP engine platform for Taiwanese projector manufacturers is described. INTRODUCTION The projection display, mainly the front projector for business and presentation applications, has grown into a multi-billion dollar industry over the past few years. In the meantime, strong market demand and fierce competition among projector manufacturers have created some very dramatic improvements in the performance and cost effectiveness of data projectors [1]. However, as the price comes down due to competition, a larger economic scale is needed to keep manufacturers’ profit margin. Projector manufacturers need to find every possible approach to further improve the cost-performance ratio of their products, in order to explore new market segments such as home theater, rear projection TV, and SOHO applications to fuel an even more dramatic growth in market size. It is every major projector manufacturer’s goal to put a projector in every conference room and in every household. Single chip DLP projection technology has been proven to offer high contrast, high brightness, and sharp images for both computer data presentation and consumer video applications. With the latest 12° DDR DMD panels, even higher contrast and better optical efficiency could be achieved to develop projection displays with competitive edges in terms of image quality, in addition to DLPTM projection technology’s advantages in compactness and lightweight. The inherent simplicity of a single chip optical system offers the potential to develop low cost projector systems. To further improve the cost effectiveness of DLPTM projectors by reducing the component and development costs, TI proposed the concept of common engine design where different size panels and resolutions could be used. It is one of the unique features of DLPTM technology that a well-planned engine design can be used to develop projectors with diversified characteristics. Also, if the engine design is shared among several manufacturers, the development cost can be shared also. Thus not only the development cost can be reduced, critical components such as the color wheel, integrator rod, and TIR prism can be standardized, and the economy of scale can be applied to reduce the cost of components. This paper describes the collaborative project between TI and OES/ITRI to develop a common engine platform for local projector manufacturers in Taiwan. The optical design and the performance results of the prototype engine are also presented. The Common Engine Platform To fulfill the above described functions, the design of DLPTM common engine should provide following flexibilities: 1. The engine can be adapted for both front and rear applications. 2. The design should be able to accommodate both 12° 0.7” XGA and 0.55” SVGA DDR DMD panels. 3. The design of the illumination relay should be compatible with the optical requirements of SCR (Sequential Color Recapture) technology [2] to provide an easy path for future upgrades. With such a flexible platform the same engine can be used to develop different projectors to meet the requirements of various market segments. The third feature could prolong the life span of the engine, which also helps to reduce the cost of the future products. TI/OES Common Engine Project A collaborative project was established by TI and OES/ITRI to carry out the concept of a common DLPTM engine platform. OES/ITRI is a nonprofit industrial technology research institute, with a good track record of developing industrial technologies for private industries in Taiwan. Working Model Figure 1 shows the working model of the project. At the incubation stage of the project, TI and OES worked together to define the product specifications and a business model for the project. A proposal was prepared to solicit the local display-related companies to join the project and to financially support the project. Frequent discussion and intensive exchange of technical information occurred between TI and OES during the design stage of the development process. It was a common goal of both organizations to have a successful product at the end of the project. During the design stage of the project, we planned to get suppliers of components such as the projection lens, color wheel, integrator rod, and TIR prism, involved in the development process. The goal was to localize the supply of optical components and in the meantime, to help develop the DLPTM optical components infrastructure in Taiwan. After the design stage and during the prototyping of the engine, the optical and mechanical design data were transferred to projector companies who signed on the project. The information obtained from various stages of verification tests with mockups was also provided to the partner companies to help them modify the designs for mass production. Finally, a prototype projector with machined mechanical parts was delivered to TI and to each partner company. Fig. 1 Working model of TI/OES Common engine development project. Project Management Both TI and OES dedicated a team of engineers for the project. The project was managed by the first author on the TI side and the OES team was led and managed by the second author. With guidance from TI team members, the actual optical and mechanical design work of the common engine was carried out by the OES team. TI team members’ many years of experience with the various aspects of the DLPTM projection technology were a great help towards achieving the excellent design the project has produced. The communication of technical issues between two teams was without any hurdle. Many e-mails and late night (or early morning) telephone correspondences were exchanged to keep the project on track and on schedule. The members of both teams also met about once a month to review the project schedule and design work. Major Milestones The idea of a common engine was proposed to the Taiwan projector industry in December of 2001. The project did not get started until early April of 2002 because a long period of time was spent defining a business model acceptable to all of the participating parties. The table below summarizes the major milestones of the project. The first image “on the wall” was demonstrated with an engine mockup and the evaluation electronics supplied by TI during one of the review meetings at TI’s Plano site. Using TI’s DLPTM Composer TM and the evaluation kits, high quality images could be demonstrated easily on a well designed optical engine. Table 1 Major milestones of the common engine project. Milestones Time Project kick off Apr 1 ’02 Illumination system design finalized Jun 23 ’02 Projection lens design finalized Jul 30 ’02 First image on wall Nov 19 ’02 First image with reference system Jan 15 ’03 Delivery of prototype projector Feb 28 ’03 During the development stage of the project, very thorough analysis were performed on the design of illumination and projection systems, including evaluations such as performance simulations, stray light analysis, and ghost image analysis. The purpose was not only to make sure that the design met the design targets, but also to let OES team members learn as many details of the DLPTM projection technology as possible. Optical Design of the Common Engine One of the design goals of the common engine is that it must be compatible with the optical requirements of SCR for future upgrades. As such, the optical design of the engine was targeted at a 12° 0.7” XGA DDR DMD system using SCR technology, with attentions paid to ensure the design’s compatibility with field sequential systems using 0.7” XGA or 0.55” SVGA DDR panel. It was also decided that a high output lumen front projector would be developed as a reference system to demonstrate the performances of the common TI OES ITRI Projection Lens Manufacturer Color Wheel Manufacturer TIR Prism Manufacturer Light Pipe Manufacturer OEM System Manufacturer Projector Manufacturer Engine Manufacturer engine. The optical specifications of the reference system were jointly defined by TI and OES at the early stage of the project. TIR Prism Telecentric Architecture To meet the design goals of the common engine, a telecentric architecture using a TIR prism to separate the illumination and projection paths was chosen as the common engine’s optical system architecture. The TIR prism architecture offers following virtues: 1. Can be used for both front and rear projection applications with projection lens change. 2. Good architecture for SCR. 3. Low illumination overfill for high optical efficiency. 4. Illumination and projection path can be designed separately. 5. Best architecture for standardization of optical components. 6. Quick path to production. Illumination Relay Based on a reference design provided by TI, a F/2.0 relay system was designed to image the dichroic filter pattern of the SCR color wheel onto the DMD mirror array, with the integrator rod as the illuminating light source. The relay optics consisted of three condenser lenses with two aspheric surfaces (Fig. 2). The relay optics were optimized to have small enough rms spot sizes over all field positions so that high quality image of the SCR color wheel’s spiral pattern could be formed on the DMD array. As shown in Fig. 3, the small rms spot sizes also produced a uniform and well focused illumination area for FSC (Field Sequential Color) applications. The magnification of the relay system was about -2, which provided an overfill of approximately 10% around the active area of the mirror array (Fig. 4) to allow for mechanical and optical tolerances. Fig.2 Layout of the illumination optics of the common engine Two high reflectance mirrors were used to fold the optical path to give a compact system layout. The optical axes of illumination and projection path were folded to have a large angle between them, to prevent the back end of the lamp from extruding too far from the last surface of the projection lens. The purpose of such a configuration was to try to achieve a differentiated ID design (size and weight, aesthetics) which meets all of the project’s objectives. Fig.3 ASAP simulated illumination area on the surface of DMD array. Fig.4 ASAP simulated overfill pattern around micro mirror array. Projection Lens To match with the design of the illumination system, a F/2.0 zoom lens was designed to give a maximum output lumen for the reference system. The effective focal lengths of the projection lens are about 23mm and 28 mm, respectively at wide and tele positions, which gives a zoom ratio of 1:1.2. The short focal length at wide position also provided a fairly small ANSI throw ratio for the reference system. At a projection distance of two meters, the lens can project an image larger than 60” in diagonal. This specification compares very favorably with most of the projectors on the market. The projection lens consists of 13 elements in 5 groups, with the last group used for focusing. With a carefully selected architecture, the projection lens offers excellent design performances at various working configurations. As shown in Fig. 5, the MTF is well above 50% at both wide and tele positions when the projection distance is at 2m. Also shown in Fig. 6, the lateral colors between green and blue, and between green and red are below 3/4 pixel of the DMD array, which is 13.68μm. The projection lens has an object field size of 24.6mm in diameter. This field size allows a 90% offset in the projected image. fold mirrors