Perspective of MEMS Based Raster Scanning Display and Its Requirements for Success

Invited Paper Perspective of MEMS Based Raster Scanning Display and Its Requirements for Success Yong-Hwa Park*, Jin-Ho Lee, Seong-Ho Shin and John Sunu Display Lab., Samsung Advanced Institute of Technology, Mt. 14-1, Nongseo-dong, Giheung-Gu, Yongin-City, Korea 449-712 * yongh.park@samsung.com; phone: +82-31-280-9464; fax: +82-31-280-9473 ABSTRACT The customers demand for real life-like display with natural colors and high definition is increasing and hence laser display with the best expression of natural color is being proposed as a way to realize this. In particular, the raster scanning display using the high-speed reflective MEMS scanner plus compact laser sources enables realization of ultrasmall optical engine with great optical efficiency. By the way, in recent years the emerging display systems including FPD (Flat Panel Display) and projection systems based on the microdisplay devices show rapid improvements in terms of picture quality, form factor as well as cost. The object of this paper is introducing a technology analysis of success factors of the MEMS based rater scanning display in order to get high-level development roadmap, through a comparison study with the conventional displays. Proper specifications of brightness, color, contrast, resolution, form factor, power consumption and cost-effectiveness are suggested for mobile projector application. The technical challenges toward achievement of the specifications are summarized. Keywords: laser display, mobile projector, raster scanning, MEMS mirror, compact laser 1. INTRODUCTION Different from analog displays, introducing the digital display technology leads changes of the information processing environments of office, home and personal uses. The digital displays are about to be the gateway of the bi-directional information exchanges for applications of entertainment, work, mobile communication, internet, commerce, broadcasting, digital photography and so on. Visual information flows forward and backward in the chain of the following three sub-domains: 1) visual source generation; 2) visual information storage/manipulation/transmission; and, 3) reproduction of visual information. Digital display technology makes all the three sub-domains full of innovative digital devices such as digital camera, camcorder for source generation; set top-boxes, small-size HDD, PVR (Personal Video Recorder), image editing S/W, high-speed wireless transmission of high resolution images for visual information manipulation; and, HD (high-definition) displays for vision reproduction. Digital display technology impacts consumer products to be more interactive, customized, portable, and have high-image quality. The consumers can easily create and distribute their own visual contents so that they become active prosumers who create and control the market, as Alvin Toffler expected in The Third Wave in early nineties. However, the digital display market is just opened nowadays and most of development activities so far have focused on replacing the CRT in home. The target is improving the image quality in order to get CRT-like image in terms of color, brightness, contrast, response time and image artifacts. Though various functionalities of display devices such as form-factor, ease of information manipulation, mobility are required in the next generation digital displays, the recent consumer survey shows that the picture quality is still the most concerning factor when they buy a next display device as shown in Figure 1 1. Hence, replacing the CRT is still a good strategy to win in the display market in coming several years until the progress of image quality is saturated. Along this line, the customers demand for real life-like display with natural colors and high definition is increasing and hence the laser display with the best expressions of natural color and high resolution is being proposed as a way to realize this. The applications of laser display are diverse ranging from ultra small, portable to large screen such as the HMD, mobile projector, projection TV and cinema projectors. In particular, the raster scanning display using the high-speed reflective MEMS scanner plus compact laser source enables realization of ultra-small optical engine with great optical efficiency so large screen video can be realized through low power consumption and small form-factor. MOEMS Display, Imaging, and Miniaturized Microsystems IV, edited by Hakan Ürey, David L. Dickensheets, Bishnu P. Gogoi, Proceedings of SPIE Vol. 6114, 611403, (2006) 0277-786X/06/$15 doi: 10.1117/12.659058 Proc. of SPIE Vol. 6114 611403-1

The positioning of the MEMS based raster scanning display in the market should be derived by comparing the several types of emerging display systems in terms of performance, features, and targeting market segments. Although it is quite early to clarify its development target, it is valuable to investigate the technical features, hurdles and requirements of the raster scanning display in order to initiate high-level product roadmap. Outline of this paper is such that a technology analysis of the key success factors of the MEMS based rater scanning display is shown through a comparison study with the conventional displays. Since the raster scanning display is a type of Microdisplays, the Microdisplay is compared to existing devices as a reference. Then, raster scanning display will be particularly investigated using its differentiation points apart from the existing Microdisplays and FPD (Flat Panel Display). As results, proper specifications of brightness, color, contrast, resolution, form factor, power consumption and cost-effectiveness are suggested for mobile projector application. The technical challenges toward achievement of the specifications are summarized. Design Brand 3% Function 10 /a 1 /o Price, 200/0 Picture Quality 66 /o Figure 1: Consumer survey result: key concerning factor when buy a new TV. The picture quality is the most concerning factor to customers. (Courtesy of Display Bank, Inc.) 2. EXISTING DISPLAY DEVICES TECHNOLOGY ANALYSIS In recent years, the emerging display systems including FPD and projection systems based on the microdisplay devices such as DMD (Digital Micromirror Device), HTPS (High Temperature Polysilicon), and LCOS (Liquid Crystal on Silicon) have shown rapid improvements in terms of picture quality, form factor as well as cost. This leads war-like competition between the emerging display devices for at least coming several years until the market shares are settled. Market perspective and comparison of devices from viewpoint of technology are summarized as follows. 2.1 Market perspectives Due to extensive investments to the manufacturing facilities in Gen 7, the price of LCD will drop very rapidly in coming couple of years. The price of LCD drives fierce cost competition between major display devices such as LCD, Plasma, and Microdisplays. Figure 2 shows an expectation of market windows of existing/emerging display devices in years 2005 and 2010. LCD s market window will deploy from under 40 inch screen diagonal sizes up to 60 inches in year 2010 by cost reductions. Due to this aggressive LCD cost drive, market windows for Plasma and Microdisplay will move from 30-70 inch to 40-80 inch, and 40-70 inch to 50-80 and more, respectively. Hot zone will be 45-55 inches, in which LCD, Plasma and Microdisplay probably have similar cost levels in year 2010. Full HD resolution (1920 by 1080p) will be common in most of existing devices in year 2010. The new technology such as OLED will emerge in the major TV market taking the market segment of LCD due to its similarity of manufacturing process. However, the material reliability limits its expansion of screen size over 30 inch and its cost competitiveness against existing LCD would not be enough by that time so that OLED s major market would be mobile display and personal uses. FED would appear in range of 30-40 inches. Though its low power consumption and CRT-like image quality, large screen size of Proc. of SPIE Vol. 6114 611403-2

FED more than 40 inch would be difficult due to its lower manufacturability than other devices arising from the vacuum packaging and uniformity issues. Figure 3 shows a street price expectation of average (42 inch) existing displays. Price destruction would happen by year 2006-2007, e.g., 42 Microdisplay projection TV will be around $1,000, which is about half of current price. The key point is that the material cost reduction will be one of the key success factors of existing displays including MEMS based Microdisplays, which conventionally have kept its price position at 2/3 of Plasma s. Year 2005 TFT-LCD TV PDP MD* PJTV Full HD (1920*1080) Year 2010 TFT-LCD TV PDP MD PJTV (SXGA) HD (WXGA) OLED TV FED TV FLAT CRT TV CRT PJTV SD (VGA) FLAT CRT TV CRT PJTV ~ 20 30 40 50 60 70 & Up ~ 20 30 40 50 60 70 & Up Figure 2: Expectation of market windows of competing display devices from viewpoint of screen size and resolution up to year 2010. (Source: JEITA, SRI, D/Search, JP Morgan, SEC-internal data) 2003 2004 2005 2006 2007 Large Size (>40 ) Display Trend CRT RPTV Microdisplay RPTV Plasma TFT LCD Appearance of major display product over 40 in the market (Market Size :1M set/year) Price Trend 8000 6000 8000 6000 TFT-LCD Market Average Price (Unit : US$) 4000 4000 Plasma 4000 2000 1500 MD RPTV CRT RPTV 3000 2000 800 2500 1800 1000 600 Figure 3: Expectation of street prices of major display products (average price of 42 inch screen). About 50% cost reduction is expected by year 2007. (Source: JEITA, SRI, D/Search, JP Morgan, SEC-internal data) Proc. of SPIE Vol. 6114 611403-3

2.2 Definition of high definition display The ultimate display (perfect reproduction of visual information of nature) is one of human s eventual engineering targets such as time machine. The standard of display quality has been evolved in line with human eyes standard and human need, closer to reality. The ultimate display can be defined with number of items such as: extreme contrast and brightness, acquisition of detail, natural colors, unlimited viewable region, natural motion, and infinite number of focal points (3D display). The display devices have been developed fulfilling the parts of the performances of the ultimate display. These ultimate display performances can be achieved by: 10,000:1 contrast level, bright light reflection by sunlight, expression of even minute particles, neither dragging nor flicker, color gamut covering more than 180% of NTSC standard, and full viewable region with two eyes. The term High Definition display in current market, can be defined conventionally as: screen size more than 40 inch, number of horizontal lines 720p and more, peak white level more than 1,000 nits, average white level 350 nits, black level less than 0.3 nits, color expression of NTSC 70% and more. 2.3 Technology comparison between existing devices As mentioned in Introduction, the performance comparison between existing major display devices such as LCD, Plasma, and Microdisplays is valuable since the raster scanning display is a type of Microdisplay. Hence, Pros and Cons of the Microdisplay are investigated to get initial referencing point of the development of the raster scanning display. Table 1 summarizes typical performances of FPD and rear projection Microdisplays. Device characteristics such as power consumption and set depth are added to image quality characteristics. Table 1: Performance comparison between existing major display devices such as LCD, Plasma, and Microdisplays as of 1Q 2005. 1) SONY 3p LCD, 2) Samsung DLP, 3) Philips LCOS Performances Plasma LCD Microdisplays (RPTV) HTPS 1) DLP 2) LCOS 3) Brightness 700 cd/m2 450 cd/m2 400 cd/m2 550 cd/m2 500 cd/m2 Ave. contrast 3,000:1 (dark room) 900:1 700:1 1,500:1 800:1 Resolution 1,366 by 768 1,920 by 1,080 1280 by 720 1280 by 720 1280 by 720 Screen size 30 ~80 10 ~57 42-70 42-70 42-70 Weight (kg) 60~70 60~70 ~35 ~35 ~35 Power Consumption High High Low Low Low Burn-in Yes No No No No Set Depth 3-6 2 13 ~20 7 ~20 24 ~30 Longevity (hrs) 25,000~30,000 50,000~75,000 (backlight) 8,000~10,000 (lamp) 8,000~10,000 (lamp) 8,000~10,000 (lamp) Viewing Angle 180 170 <150 <150 <150 Due to intensive technology innovations, the gaps between image qualities of LCD, Plasma, and Microdisplays are decreasing. The bright room contrast of Plasma has been the weakest point in its image quality, however it is getting improved by modifying the optical characteristics of electromagnetic shielding film in front of the screen. The motion dragging of LCD has been a painful image artifact toward the perfection of moving image reproduction, but nowadays response time less than 8 ms was achieved so that customers are difficult to see the tails of moving objects. The viewing angle of LCD has been greatly improved by special structures of LC plus color filters. Microdisplays have promising performance with lower price especially in large screen TVs since the material cost of the Microdisplay is still very low compared to FPD of the same screen size. Microdisplay shows excellent brightness and contrast. The brightness and contrast have been improved by using the high fill-factor MEMS mirror and large optical transmission ratio of HTPS Proc. of SPIE Vol. 6114 611403-4

supported by micro-lens arrays. Nowadays HD resolution (1280 by 720p) became standard of Microdisplays and even full HD (1920 by 1080p) has been introduced in the market. Also their weight and power consumption are lighter compared to FPD, ease installation and maintenance are attracting points to buy them especially in Northern America and Asian markets. However, the weak points of Microdisplays are thicker set depth, imperfect viewing angle, and relatively short life time of arc lamp. The slim projection technology continuously progresses by using aspheric mirrors in projection lens set to decrease the set depth less than 4-5. However the cost premium of slim projection is not attractive so far. Micro-lens based refracting technology could be adapted to get viewing angle closer to 180 but the mass-manufacturability and uniformity issues should be resolved before its adaptation. The arc lamp traditionally has been one of key components in projection displays for more than 20 years. It has been improved in its arc size down to 1mm, brightness level up to 13,000 lumens, and life time up to 10,000 hours 2. However, although the arc lamp achieved great improvements, needs of alternative light sources are emerging from the requirements of the next generation Microdisplay systems. 2.4 Thrust of solid-state new light sources for next generation projection displays As mentioned in section 2.2, the wide color expression is one of pursuing characteristics of next generation display systems toward ultimate performance. The current progresses of solid-state light sources such as LED and Lasers make display engineers explore new possibilities of color expression. The conventional arc lamp has limits in terms of color expression (NTSC 72%) due to its broad color spectrum. Thanks to the selectivity of particular RGB color points in the CIE chromaticity diagram as shown in Figure 4, the color expressions of LED and Laser can be approximately 105% and 130% of NTSC, respectively. Laser has wider color expression since it can generate pure single wavelength compared to LED. This single frequency characteristic of solid-state light sources yields additional benefits in viewpoint of system engineering. Table 2 summarizes the problems of current arc lamp based display systems and their solutions enabled by replacing the light source. Besides expansion of color gamut, solid-state light sources can bypass using color wheel so that the corresponding material cost and noise can be reduced. Complexity of illumination optics including color wheel can be released so that the illumination optics can be smaller and simpler, which yields lower operating failure rate in the market filed. Lifetime of arc-lamp, 10,000 hrs is not enough to satisfy customers. Customers should replace lamp in about every 2 years, which makes additional cost burden. Lifetime 50,000 hrs is quite excellent in display applications, however, additional improvements of solid-state light source should be involved to get this level of lifetime at this moment. Initial booting time of lamp-based projection displays due to pre-heating of arc lamp is annoying thing especially for first users. The instantaneous turn-on features of solid-state light source can help this. Due to wide spectrum of arc lamp, the optical throughput is inherently limited by filtering of unused spectrums of light, whereas solid-state light dramatically increases the optical efficiency along the light path from lamp-to-screen because particular wavelengths are generated from the light sources. This implies that thermal management of optical system is easier. Laser (130%) LED (105%) Arc Lamp (72%) (NTSC referenced) Figure 4: Color expressions of arc lamp, LED and Laser Proc. of SPIE Vol. 6114 611403-5

However the absolute optical power output of state-of-the-art solid state light is quite low compared with arc lamp by now. The lumen per watt of LED around 50 is achievable in the state of the art. For the promising applications, at least 100 lumen/watt is required. Insufficient optical power output, especially of LED for DLP projection, makes ironically thermal and longevity problems due to over driving of chips to get sufficient outputs. Most of LED and Laser based projection system suffer from the thermal management problem for this reason so far. One of key technology impacts of solid-state light source especially Laser is enabling applications of various types of MEMS based Microdisplays such as line type diffraction optics array (GLV: Grating Light Valve by Silicon Light Machine, patent transferred to SONY; GEMS: Grating Electromechanical System by Eastman Kodak), and point type raster scanning MEMS reflecting mirrors as well as conventional panel type devices (DMD, HTPS, LCOS). Because of Etendu restriction of line and point types SLM (Spatial Light Modulator), arc lamp and even LED are not sufficient to get enough optical throughputs in these devices (see Etendu and its definition in reference 2). Laser can only be a suitable light source for line type and point scanning devices. Table 2: Problems of arc lamp-based projection system and characteristics of solid state light sources as remedies. System characteristics Problems of arc lamp based projection system Characteristics of LED/Laser based projection system Color expression 70 % of NTSC LED: 105% / Laser: 130% of NTSC Color generation mechanism Rotating color wheel RGB photonic chips Life time 10,000 hrs ~ 50,000 hrs System booting time ~ 15 sec Instantaneous (except high power lasers) Optical efficiency Relatively low Relatively high Environment With mercury No mercury Matching Microdisplays Only panel type SLM (DMD, LCOS, HTPS) LED: Panel type SLM, raster scanning (near eye application) Laser: Panel type SLM, line type (GLV, GEMS), raster scanning 3. LASER DISPLAY AND ITS REQUIREMENTS FOR MOBILE APPLICATOIN As mentioned in section 2.1, the cost-reduction pressure of Microdisplay is getting heavier due to current display market situation. Application of lasers into RP(Rear Projection) TV should involve reasonable cost premium against arc lamp based RPTV to get into market field. However, the state-of-the-art laser sources do not meet the performance-cost requirements for big-screen RPTV, which will have average street price of under $1,000 by 2007. Especially optical power is not enough to have acceptable screen brightness at least around 350 nits at 50 inches with reasonable price. Hence more practical approach for laser display is starting with smaller screen, compact system applications. This Chapter summarizes performance of laser displays and its requirement especially for mobile projector application. 3.1 Attraction points Laser display includes not only the change of light source but also device changes. The light source technology and SLM technology are closely related. Table 3 lists comparison of display performances derived from the light characteristics plus SLM technology in general. Proc. of SPIE Vol. 6114 611403-6

As mentioned in section 2.4, wide color expression is an attractive point of laser display. The issue is nonmatching of laser colors with standard of current broadcasting system (SMPTE: Society of Motion Pictures and TV Engineers). Actually, the colors beyond SMPTE standard are not supported in the visual image capturing process of current standard. Wide colors out of SMPTE standard should be generated by expanding the captured standard colors and mapping to wide color gamut, which may results unrealistic colors in laser display. This color non-matching issue frequently ends up with compromising between TV engineers and marketers. Hence, to achieve a perfect reproduction of natural colors, improvement of color standardization and corresponding enhancement of image capturing devices should be involved along with expansion of color reproduction capability. Currently developed xvycc standard is very effective for laser display to have full-performance. Laser display using the line type SLM such as GLV and GEMS, and point scanning devices can increase the resolution more economically compared with panel type SLM such as DMD. The high yields and small cost increments are expected when the number of pixel increases due to 1-dimentional or 0-dimentional nature of pixel structures. However, additional improvements in laser modulation bandwidth and scanning speed/angle should be involved for higher resolutions of raster scanning display. The speckle noise is an inherent characteristic of coherent lasers. The coherency of laser should be loosened to get de-speckle image but the side effect is lost of beam quality. Vibrating screen, moving mirrors, DOE/HOE solutions have been introduced but raster scanning display requires a very high speed vibration up to 5X of 60 MHz is required for satisfactory removal of speckle noise of HD resolution. The small form factor of raster scanning display is one of the most attractive leveraging points, which enables mobile application of display. Figure 5 shows schematic of MEMS based raster scanning display. The number of optical components can be minimized in raster scanning scheme: collimating optics, diachronic mirrors and MEMS mirror constitute the minimal setup of optics. According to the application, projection optics such as F-theta lens can be incorporated at the backend of the optics setup but it is not a mandatory. Due to the simplicity of system, the system size can be drastically reduced down to around 100cc. Due to mobility of the display, the applications are diverse such as entertainment, office solution and communication. The reasonable cost could be achieved as long as the compact laser cost is reasonable Table 3: Comparison of display performances between conventional displays and MEMS-based raster scanning display. X:bad. A: fair, 0: good. 0: excellent) 1 display wi Laser Relennt feature Display Solutious FPD (Jut t)ve or of laser display PJ display wi arc laiup aneive) PJ display wi LED aneive) Naturalcolor A A 0 0 NTSCI3O% mlra-high resolution 0 0 0 0 Pointorlinescan Imageartilict A 0 0 A Speckle Small ibm' frctor A A 0 0 Cost-effective X A 0 MEMS scanner wi milumum optics MEMS scanner wi muumum optics as long as cost of laser is reasonable 3.2 Required specifications of components of MEMS based raster scanning display The raster scanning display has merits over conventional displays as well as LED based projection display as shown in Table 3 as long as required specifications are satisfied. The detailed requirements and technical hurdles of raster scanning display toward achievement of the display performances are summarized as follows. The requirements are: light budget/image size, resolution, efficiency of optics, electric power consumption, and form factor. The requirements Proc. of SPIE Vol. 6114 611403-7

Compact Lasers 1qEMS Scanner Micro-optics Figure 5: Schematic of MEMS based raster scanning laser display above cannot be decoupled and are incorporated to finalize the system performance. For the communication applications (projector built in hand held set), form factor and power consumption have priorities over brightness and screen size. So, small laser power with less than 10 screen is preferred. Contrarily, for office application, group of people could see the image in presentation, the screen size and lumen output have priorities over form factor and power consumption. Screen size over 40 is preferred. The personal entertainment application requires somewhere between those of communication and office applications. 3.2.1 Light budget and image size The image size, corresponding reasonable lumen output and required laser output are summarized in Table 4. For communication application, 10 would be promising for miniaturization and light power consumption less than 1W. The office applications, more than 100 mw laser outputs are required to get 40 screen size with 60 lm. Table 4: Required laser output power for various screen sizes and corresponding lumen outputs. Assuming that optical efficiency from laser to screen is 70% and direct modulation of laser is used. The required laser powers are varying depending on the white balancing point, so called color temperature. The reference color temperature is 10,000K. Required laser power (mw) Wavelength RED (635 nm) Green (532 nm) Blue (460 nm) Amplitude modulation Direct modulation Direct modulation Direct modulation Mobile projector application 6 lm (10 ) 13 10 11 (Optical system efficiency 30 lm (30 ) 64 52 55 of 70% assumed) 60 lm (40 ) 128 104 110 RP TV application (Optical system Efficiency of 70% assumed) 300 lm (60 ) 640 520 550 530 lm (80 ) 1,130 919 972 Proc. of SPIE Vol. 6114 611403-8

3.2.2 Efficiency of optical system The optical system efficiency is determined by integration of collimating efficiency, throughput of dichroic mirrors, reflectivity of MEMS mirror, duty of laser turn-on time in horizontal/vertical scanning period and possible scattering loss in the package of MEMS mirror. The optical efficiency of 70 % is very optimistic actually. If de-speckle element is added in the optical path, the resulting optical efficiency would be decreased by around 10-20% regarding the align mismatches and other irregularities of lenses. Anyway, optical efficiency of 70% is challenging but should be a reasonable target. 3.2.3 Power consumption Power consumption is determined by many factors. The most critical factors are specification of lumen output and WPE (Wall Plug Efficiency) of lasers especially green. Different from the red and blue diodes, green diode is still a missing link for miniaturization of lasers. The SHG (Second Harmonic Generation) green laser is most feasible in the state of the art. The green SHG has a quite low WPE compared to those of red and blue, so green took about more than 80% of total power consumption of lasers. WPE of green laser around 5% is promising to get reasonable power consumptions of HHP (<1W) and mobile projector applications (<5W). 3.2.4 Resolution The resolution is determined by MEMS mirror characteristics (spatial performance) plus laser modulation bandwidth (temporal performance). Assumption is that the beam quality of laser in terms of M-square (M 2 ) is 1.5-2.0. SVGA (800 by 600) is promising resolution of communication applications, whereas HD resolution (1280 by 720p) is promising to come up with entertainment and office applications in year 2007 and later time frame, in which HD contents will overspread broadcasting environment. The well established Theta-D relation 3 and restriction of dynamic deformation of mirror 4,5 gives design rules of the MEMS mirror. Most of raster scanning devices utilize resonance of fundamental mode, usually torsional mode of sprung mirror for horizontal scanning. For HD resolution, more than 22.5 KHz resonant frequency is required when bi-directional horizontal scanning is utilized. This high frequency, i.e., high torsional stiffness, constitutes trade-off situation between mirror diameter and maximum deflection angle. The eye-type mirror structure was invented to maximize the deflection angle preserving the mirror diameter 6,7. The modulation bandwidth of lasers directly related to the distinguishable pixel size on the screen. Table 5 summarizes required laser modulation bandwidth for various resolutions. 60 MHz modulation is required for realization of HD resolution. Implementation of this bandwidth is challenging especially for green laser since its current level is several amperes due to its low WPE, which requires special power electronics capable of high current. Resolution Table 5: Required laser modulation bandwidth for various resolutions. Modulation bandwidth (MHz) Rise/fall time (ns) Pixel duty time (ns) SD (480P), VGA 20 25 50 HD 720P 56 8.9 17.8 1080i 65 7.8 15.6 1080P 130 3.9 7.8 3.2.5 Gray scale and contrast The gray scale and contrast ratio are directly related to the precision of the laser modulation. 8 bits (256 steps) of gray scale is promising but 10 bits (1024 steps) gray scale would be required in most of display systems in year 2007 time frame. The stability of control input-laser output relationship should be guaranteed to get precise expression of gray scale. The contrast ratio is the ratio of full brightness to full darkness. 1000:1 is promising but 3000:1 will be dominant in display systems in 2007 time frame. The contrast ratio is sensitive to faithfulness of full-dark state. Lasers usually have threshold current to initiate the light generating. Identifying the stable threshold current are the key points to have full dark state in a stable manner Proc. of SPIE Vol. 6114 611403-9

3.2.6 Form factor The extreme case is in communication application (hand held set). Packing the entire display engine (lasers, optics, and scanner) into a hand held set is very challenging. Form factor of less than 10cc of entire display engine should be achieved for promising adaptation. Compact RGB laser sources, opto-electro-mechanical package and optimization are required. For the application of mobile projector, form factor less than 100 cc is promising including all electronics plus thermal solutions. 3.2.7 Thermal management The regulation of consumer electronics devices defines maximum operating temperature at the device surface as about 50 C depending on the surface material. This requires special thermal management solution in small projectors since there is not enough heat transfer path except directly to ambient. Also, intense heat generating component such as green laser should be properly cooled to get stable operation. The key issue is expanding heat-convecting area in a small volume. The optimization of heat transfer path and adaptation of micro cooling systems could give solutions 4. CONCLUSIONS The MEMS-based raster scanning display is an alternative solution of the existing lamp-based projection displays which overcomes weak points of arc lamp solution in terms of color expression, life time, optical efficiency, instantaneous turn-on, and simplicity of optical engine. The core competence of the raster scanning display is not only an alternative light solution but more importantly is a completely new kind of display solution which enables mobility of display with HD-image reproduction in screen size up to 40. Due to immaturity of compact laser source technologies, the miniaturization of the projection engine is limited so far. Improving WPE is a key to get sufficient optical output 100 lm with power consumption less than 5W for mobile projector applications. Developments of MEMS scanner and laser driving technology to support HD resolution are still on-going and need couple of years to have feasible manufacturing sample. As the successive work of last year, the VGA full color image was demonstrated in Samsung s Lab. The compact system was implemented by using directly modulated compact lasers. The system size is 120 cc excluding the power electronics. The demonstration revealed that lumen output and power consumption need to be improved for the feasible solution in the market. The future works will devote to miniaturizing the system size down to 100 cc. The 2- dimentional MEMS scanner will be adapted into the system targeting HD resolution. REFERENCES 1. Display Bank Inc., www.displaybank.com 2. Edward H. Stupp and Matthew S. Brennesholtz, Projection Displays, John Wiley & Sons, 1999. 3. Hakan Urey, David W. Wine and Thor D. Osborn, Optical performance requirements for MEMS-scanner based Microdisplays, Conf. in MOEMS and Miniaturized Systems, Proceedings of SPIE vol. 4178, pp. 176-196, 2000. 4. Robert. A. Conant, P.M. Hagelin, U. Krishnamoorthy, M. Hart, O. Solgaard, K.Y. Lau, and R.S. Muller, A rasterscanning full-motion video display using polysilicon micromachined mirrors, Sensors & Actuators A (Physical), vol. 83, no. 1-3, pp. 291-296, May 2000. 5. Robert A. Conant, Micromachined Mirrors, Ph. D dissertation, UC Berkeley, 2002. 6. Young-Chul Ko, Jin-Woo Cho, Hyun-Gu Jeong, Won-Kyoung Choi, Yong-Kweun Mun, Jin-Ho Lee, Design and fabrication of eye-type scanning mirror with dual vertical combs for laser display, 2004 IEEE/LEOS Intl. Conf. on Optical MEMS, Takamatsu, Japan, pp. 184-185, 2004. 7. Young-Chul Ko, Jin-Woo Cho, Yong-Kweun Mun, Hyun-Gu Jeong, Won-Kyoung Choi, Ju-Hyun Lee, Jeong-Woo Kim, Ji-Beom Yoo, Jin-Ho Lee, Eye-type scanning mirror with dual vertical combs for laser display, Conf. in MOEMS Display and Imaging Systems, Proceedings of SPIE vol. 5721, pp. 14-22, 2005. Proc. of SPIE Vol. 6114 611403-10