DMC009 file://\\fileserver\ 함께갖다 \[[X 논문 ]]\DMC\009\proceedings.htm 페이지 1 / 7 010-01- Welcome Acknowledgement Committees Chairperson Proceedings Author ndex Search Home PROCEEDNGS Keynote Speeches Wed-KN-01 : FDP Materials Development in organic materials for FPDs Dr. Alasdair Jelfs, Merk (Germany) Wed-KN-0 : OLED AMOLED Large Size Technology Dr. Duglas(Sung Soo) Park, Chi Mei EL Corp. (Korea) Wed-KN-03 : 3D Systems and Applications The Latest Trends in 3D Entertainment Dr. Joshua Greer, REALD (USA) Session01 OLED (1) AM/Backplane Technologies for AMOLEDs Tue-S01-01 (nvited) External Compensation of Non-Uniform mage Quality and mage Sticking in Active Matrix Organic Light Emitting Diode Displays Oh-Kyong Kwon, Hai-Jung n (Hanyang Univ., Korea) Tue-S01-0 (nvited) Manufacturing Poly-Si Backplanes for Low cost and Large AMOLEDs Dong-Hoon Shin, S.K. Lee, J.M Lee, K.K. Park, H.J. Kim (Viatron Technologies Corp., Korea), J.W. Hong, K.Y. Lee (Samsung SD, Korea) Tue-S01-03 (nvited) AdMo TM and AdMo-p TM backplane technologies for AMOLED Corbin Church, G. Reza Chaji, Muahel Tabet, Stefan Alexander, Arokia Nathan (gnis novation nc, Canada)
External Compensation of Non-Uniform mage Quality and mage Sticking in Active Matrix Organic Light Emitting Diode Displays Oh-Kyong Kwon and Hai-Jung n Division of Electronics and Computer Engineering Hanyang University, 17 Haengdang-dong, Seongdong-gu, Seoul, 133-791, Korea ABSTRACT The variation of electrical characteristics of thin film transistors (TFTs) causes non-uniform image quality problem and the differential aging of organic light emitting diode (OLED) devices causes image sticking problem in active-matrix OLED (AMOLED) displays. n this paper, various driving methods with internal and external compensation of threshold voltage and mobility variations of driving TFTs and degradation of OLED are reviewed. External compensation methods are expected as an effective solution for high image quality AMOLED displays. NTRODUCTON An organic light emitting diode (OLED) display is expected as the most promising next generation flat panel display due to strong points such as fast response time, ultra-lightweight form factor, and excellent color reproducibility. 1 However, nonuniformity of luminance due to electrical characteristic variation of thin film transistors (TFTs) used as active matrix backplane and image sticking due to the differential aging of OLED devices 3 cause serious problems of image quality in active-matrix OLED (AMOLED) displays. Fig. 1 shows a simple AMOLED pixel structure with TFTs and 1 capacitor. 4 D (na) T1 is used as current source for OLED, so it is called driving TFT. T is used as switch for programming data voltage to gate node of T1. C1 is used as storage capacitor and holds the gate voltage of T1 during a frame time. Fig. 1 shows D -V SG characteristic of T1 in 4 different pixels on polycrystalline silicon (poly-si) TFT backplane. 500 400 300 00 Pixel 1 Pixel Pixel 3 Pixel 4 100 0 0 1 3 4 5 6 7 8 V SG (V) Fig. 1. A simple AMOLED pixel structure with TFTs and 1 capacitor and measured D -V SG characteristic of T1 in 4 different pixels on poly-si TFT backplane. Because electrical characteristics of TFTs are different from pixel to pixel, D of each pixel is different even same V SG is applied. This current deviation directly affects to the non-uniform luminance of OLEDs in pixels because the luminance of OLED is in proportion to current density. As a result, the compensation of electrical characteristic variation of T1 is required for uniform luminance of AMOLED displays. Even though we can make uniform current source in pixel, there remains image sticking problem due to differential aging of OLED devices. Fig. shows normalized luminance of OLED devices in constant current stress. Because the luminance of OLED in each pixel is different due to display image, the aging of OLED in each pixel is different. Normalized luminance 1.0 0.9 0.8 0.7 0.6 0.5 0 100 00 300 400 500 600 700 800 900 1000 Time (h) Fig.. Measured luminance of OLED in constant current stress condition. For the acceleration of degradation high level current, 11.3 ma/cm, is applied to the OLED. n this paper, previously proposed internal and external compensation methods are reviewed and the consideration factors for each driving method are discussed. NTERNAL COMPENSATON METHODS Most of the approach to compensate electrical characteristic variation of TFTs and degradation of OLEDs was compensating them inside the pixel using various pixel structures. These internal compensation methods can be divided into voltage programming methods 5,6, current programming methods 7,8,9, electrical feedback methods 10,11, optical feedback methods 1, digital driving methods 13, clamped inverter methods 14, degradation compensation methods using
capacitive coupling 15. data line. Voltage Programming Current Mirror Structure Voltage programming current mirror pixel structure 5 as shown in Fig 3 used mirroring TFTs to compensate variations of threshold voltage of T1. Fig. 3. Schematic diagrams of voltage programming current mirror pixel and voltage programming current source pixel. The current can be expressed as ( ( ) ) = VDD Vdata Vth Vth1, (1) and β 1 is W = μ 1C ox, () L 1 where V data, V th, V th1, μ 1, C ox, (W/L) 1 are applied data voltage to gate node of T1, threshold voltage of T, threshold voltage of T1, mobility of T1, gate capacitance, the ratio of channel-width to channellength of T1, respectively. t is shown that the mismatch of V th1 and V th causes deviation of current. Also, mobility variation of T1 affect to the current. Voltage Programming Current Source Structure Voltage programming current source pixel 6 as shown in Fig. 3 used self-v th compensation technique using diode connected structure of T1. The current of this structure can be expressed as 1 W ( ) = μ 1C ox Vdata V, (3) SUS L where V SUS is the DC reference voltage. This structure could compensate threshold voltage variation, but the mobility variation of T1 could not be compensated. Current Programming Current Source Structure Current programming current source pixel 7 as shown in Fig. 4 was proposed. The current does not depend on threshold voltage and mobility variations of T1. However, this method had a problem of programming low level current during row line time due to large parasitic capacitance of (c) Fig. 4. Schematic diagrams of current programming current source pixel, current programming current mirror pixel, and (c) current programming voltage boosting pixel. Current Programming Current Mirror Structure Current programming current mirror pixel 8 as shown in Fig. 4 was proposed to overcome charging problem in low gray level. The current can be expressed as β =, (4) 1 data VDD Vth Vth1 β 1 and β is W β = μ C ox, (5) L where μ and (W/L) are mobility and the ratio of channel-width to channel-length of T, respectively. By using the ratio of (W/L) 1 to (W/L), high-level programming current can be scaled down to lowlevel current. However, mismatch of threshold voltage and mobility of mirroring TFTs causes current deviation. Current Programming Voltage Boosting Structure Current programming pixel with voltage boosting 9 as shown in Fig. 4(c) used voltage boosting technique to scale down the high level programming current to low level current. The current of this method can be expressed as β 1 data ΔVscan C =, (4) C1 + C
where Δ V scan is voltage difference of high level and low level of scan signal. This structure could overcome data-charging problem and compensated threshold voltage variation of T1, but voltage boosting causes current error due to the mobility variation of T1. Electrical Feedback Method Current-sensing voltage-feedback method 10 and voltage programming with transimpedance feedback method 11 could compensate both threshold voltage and mobility variation and did not have data charging problem, but the pixel structure and driving method were too complex. Digital Driving Method The pixel structure of digital driving method 1 is same as Fig. 1, but T1 is used as just a switch instead of a current source circuit. Gray level is expressed using pulse width modulation (PWM) method. Because T1 is used in linear region the electrical characteristic variation of T1 rarely affects to the deviation of current. However, the luminance is sensitive to OLED degradation and power line R-drop because the variation of applied voltage to OLED directly changes the current of OLED. Moreover, it has color breaking problem because PWM method is used. Clamped nverter Driving Method Clamped inverter driving method 13 used an inverter in a pixel to control the time, so the electrical characteristic variation of TFTs rarely affected to the deviation of current. However, it was sensitive to OLED degradation and power line R-drop and also had color breaking problem like digital driving methods. Furthermore, because of distortion of ramp signal due to RC delay of signal line, it was hard to apply this method to large size AMOLED displays. Optical Feedback Method Optical feedback method 14 as shown in Fig. 5 was proposed to compensate electrical characteristic variation of TFTs and degradation of OLEDs. T was used as photo TFT. Generated photo current of T changed the gate voltage of T1. The idea of using optical feedback was good, but the problem was the photo TFT itself had a variation in sensitivity from pixel to pixel. Fig. 5. Schematic diagrams of optical feedback pixel structure and degradation compensation pixel using capacitive coupling. Degradation Compensation Using Capacitive Coupling Degradation compensation pixel using capacitive coupling 15 as shown in Fig. 5 could partially compensate luminance deviation due to differential aging of OLED devices. By using capacitive coupling, gate voltage of T1 was boosted by the increased anode voltage of OLED. However, the pixel structure was too complex, and it was hard to compensate degradation of every gray level using just simple capacitive coupling. Furthermore, this method could not compensate mobility variation of TFTs. EXTERNAL COMPENSATON METHODS Adding many functional TFTs in AMOLED pixel to improve image quality is not a good approach because increased number of TFTs causes another additional error due to parasitic capacitance and many switching operations. Furthermore increased number of TFTs in a pixel decreases the yield of the panel because nowadays high resolution display consists of millions number of pixels. We have to compensate not only the electrical characteristic variation of driving TFTs but also the degradation of OLED and still retain simple pixel structure. A feasible solution is an external compensation method. By sensing and compensating all of deviation factors externally, we can achieve highimage quality and simple pixel structure together. Several external compensation methods such as adjustment method using light intensity 16, current comparing method using feedback iteration 17, and parameter extraction method using voltage sensing 18 were proposed. n external compensation methods, sensing time and compensation algorithm are important because additional sensing time and external calculation for compensation are required. Adjustment Method Using Light ntensity Adjustment method using light intensity was proposed as shown in Fig. 6. 16 This method applied reference voltage to gate node of T1 and sensed the luminance of the pixel. By comparing
luminance of target pixel one by one adjusting reference voltage, the luminance (c) deviation of panel Fig. can 6. be Schematic sensed diagrams and memorized. of external However, compensation this pixels method took with adjustment a long time method to sense using the light luminance intensity, deviation current comparing of each method pixel because using feedback luminance iteration, is and (c) parameter extraction method using voltage sensing. sensed pixel by pixel. Current Comparing Method Using Feedback teration Fig. 6 shows the schematic diagram of current comparing method using feedback iteration. 17 This method senses the current of T1 through T3 and T4 and compares it with data current in the external circuit. The adjustment applied voltage to the gate of T1 is performed by the feedback iteration. This method is suffered from long sensing time because many numbers of feedback iteration are needed for fine accuracy. Parameter Extraction Method Using Voltage Sensing Parameter extraction method using voltage sensing was proposed as shown in Fig. 6(c). 18 By sensing voltage with sinking or sourcing reference current, the electrical characteristics of T1 and degradation of OLED can be sensed and stored as parameters. Because only two points of -V curve of T1 and one point of -V curve of OLED are required for parameter extraction, the sensing time is fast. EXTERNAL COMPENSATON ALGORTHM The operation of external compensation method with parameter extraction method can be divided into TFT sensing step, OLED sensing step, and display step. TFT sensing is performed only once after panel is fabricated, and sensed data is stored to nonvolatile memory. During this step, T, T3, and T4 turn on in Fig. 7(c) and MAX current sinks at data line, where MAX is the maximum required current for maximum luminance. n this time the gate voltage of T1 can be expressed as MAX VG1 = VDD Vth1, (5) and stored to external memory. Once more the (1/56) MAX current sink at data line, then the sensed voltage at the data line can be expressed as 1 MAX VG = VDD Vth1, (6) 16 and stored to external memory. OLED sensing is performed when user pushes the turn-on switch of the display system. During this step, T and T3 turn off and T4 turns on, and ref current applied to OLED from data line, where ref is the reference DC current. As shown in Fig., the anode voltage of OLED increases as OLED is degraded. By sensing this voltage at data line, we can determine the degradation percentage of OLED and store it to external memory. Because this operation is performed when the display system is turned on, the sensing time of each pixel should be fast. The sensing time of every pixel on panel is 384 msec when analog-to-digital converter (ADC) is used in every channel on 14.1-inch HD resolution (1366 X 768) condition. After OLED sensing is performed, the system starts to display step. TFT sensing and OLED sensing steps are not performed during normal display operation. During this step video data is modulated using stored V G1, V G and OLED degradation. T3 and T4 turn off, and T turns on and the modulated data voltage, V data, is applied to gate node of T1. V data can be expressed as 16 100 i Vdata VG1 ( VG - VG1 ) 1, 15 100 - α n = (7) where α, i, and n are the stored degradation percentage of OLED, selected gray level from video data, and the total number of gray level, respectively. When T turns off and T3 turns on, the OLED on that pixel start to. n this time the current can be expressed as = ( VDD Vdata Vth1 ), (7) = 100 100 - α i n MAX t is shown that the current does not depend on threshold voltage and mobility of T1 and increases as OLED is degraded. EXPERMENTAL RERULTS Fig. 7 and show the measured current of 4 pixels with and without external compensation, respectively. t is shown that the current deviation due to electrical characteristic variation of TFTs is successfully compensated by the external compensation algorithm using parameter extraction method. Emission current (na) 1000 Pixel 1 Pixel 800 Pixel 3 600 Pixel 4 400 00 0 0 3 64 96 18 160 19 4 56 Gray level Fig. 8 and show the photograph of OLED devices of 0% and 50% degradation with and Emission current (na) 600 500 400 300 00 100. Pixel 1 Pixel Pixel 3 Pixel 4 0 0 3 64 96 18 160 19 4 56 Gray level Fig. 7. Measured current of 4 pixels without and with external compensation algorithm using parameter extraction method.
without external compensation, respectively. t is shown that the luminance degradation of OLED is also successfully compensated by the external compensation algorithm using parameter extraction method. Fig. 8. Photograph of 50% and 0% degraded OLEDs without and with external compensation algorithm using parameter extraction method. CONCLUSONS The internal and external compensation methods for AMOLED displays are reviewed and the consideration factors for each driving methods are discussed. Experimental results show that the external compensation method using parameter extraction successfully compensates not only the electrical characteristic variations of TFTs but also the degradation of OLED. External compensation methods are expected to be an effective solution for non-uniform luminance and image sticking problems with simple pixel structure. Proceedings EURODSPLAY, p. 613 (00). 6. S. M. Choi, O. K. Kwon, and H. K. Chung, Proceedings SD, p. 60 (004). 7. M. Ohta, H. Tsutsu, H. Takahara,. Kobayashi, T. Uemura, and Y. Takubo, Proceedings SD, p. 108 (003). 8. T. Sasaoka, M. Sekiya, A. Yumoto, J. Yamada, T. Hirano, Y. wase, T. Yamada, T. shibashi, T. Mori, M. Asano, S. Tamura, and T. Urabe, Proceedings SD, p. 384 (001). 9. Y. W. Kim, O. K. Kwon, K. N. Kim, D. Y. Shin, B. H. Kim, and H. K. Chung, Proceedings SD, p. 367 (00). 10. H. J. n, B. D. Choi, H. K. Chung, and O. K. Kwon, Jpn. J. Appl. Phys. Vol. 45, p. 4396 (006). 11. H. J. n, P. S. Kwag, J. S. Kang, O. K. Kwon, and H. K. Chung, J. SD Vol. 14, p. 665 (006). 1. A. Tagawa, T. Numao, and T. Ohba, Proceedings DW, p. 79 (004). 13. H. Akimoto, H. Kageyama, Y. Shimizu, H. Awakura, N. Kasai, N. Tokuda, and T. Sata, Proceedings SD, p. 1550 (005). 14. M. Hack, M. Lu, R. Kwong, M. S. Weaver, J. J. Brown, J. A. Nichols, and T. N. Jackson, Proceedings EURODSPLAY, p. 1 (00). 15. S. M. Choi: Korea Patent 0873075 (008). 16. W. Edward. Naugler Jr. and Damoder: US Patent 705390 (006). 17. G. Reza Chaji, S. Alexander, A. Nathan, and C. Church, Proceedings SD, p. 119 (008). 18. H. J. n and O. K. Kwon, EEE Electron Device Lett. Vol. 30, p. 377 (009). Acknowledgement This work was partially supported by Korea Research Foundation Grant (KRF-006-005- J04101). References 1. S. Forrest, P. Burrows, and M. Thompson, EEE SPECTRUM Vol. 37, p. 9 (000).. X. Guo and S. R. P. Silva, EEE Trans. Electron Devices Vol. 5, p. 379 (005). 3. P. E. Burrow, V. Bulovic, S. R. Forrest, L. S. Sapochak, D. M. McCarty, and M. E. Thompson, Appl. Phys. Lett. Vol. 65, p. 9 (1994). 4. R. M. A. Dawson, M. G. Kane, Z. Shen, D. A. Furst, S. Connor, J. Hsu, R. G. Stewart, A. pri, C. N. King, P. J. Green, R. T. Flegal, S. Pearson, W. A. Barrow, E. Dickey, K. Ping, S. Robinson, C. W. Tang, S. Van Slyke, F. Chen, J. Shi, J. C. Sturm, and M. H. Lu, EEE Lasers and Electro-Optics Society Annual Meeting, Vol. 1, p. 18 (1998). 5. Y. W. Kim, S. R. Lee, and O. K. Kwon,