P.81/T. Bignon SID 05 DIGEST  1 P.81: Image Sticking Cartography on PDP TV: A New Quantitative Measurement. Thibault Bignon, Pierre Boher, Véronique Gibour and Thierry Leroux R&D Department, ELDIM, 1185 rue d’Epron, 14200 Hérouville St Clair, France Abstract Image sticking phenomenon is currently one of the most serious problems of PDP for television applications. In this paper, we introduce a new automated solution to measure quantitatively the temporal image sticking on this type of display. The method is based on the use of a high quality and rapid 2D luminance meter that takes images of the PDP versus time. The panel is burned with a given test pattern during a fixed time and then switched to dark or bright state. The temporal evolution of the remaining image sticking level is followed in real time by the luminance meter. For quantitative cartography a chessboard is used as test pattern and the temporal evolution of each elemental pattern is deduced and adjusted with exponential decays. Quantitative information on the time constant and the amplitude of the effect are deduced on the entire surface of the panel. So the quality of each panel can be precisely determined which allows quality control at the end of the fabrication process and quantitative comparisons between different panels. 1. Introduction Rapid growing of the PDP (Plasma Display Panel) market for digital high definition television is now just at its beginning. However, there are still some issues related to the image quality of this type of display. Among them, the most critical is the image sticking phenomenon. This phenomenon has been studied by different authors both for PDP [1-3] and LCD [4-5]. It causes a deterioration of the image due to the residual image pattern still remaining on the subsequent images. It occurs when the same image pattern is displayed repeatedly over a short time (around 5 minutes) and has been related to the activation of MgO surface or the degradation of phosphor layers during the strong sustained discharges [1-2]. Special driving methods to reduce this parasitic effect have been proposed [3]. Nevertheless, the effect is never completely suppressed and accurate measurement methods are needed to quantify the phenomenon. Since it is mainly due to charge trap inside the structure of the panel it is certainly process dependent and then not necessarily homogeneous on the surface of the panel. A method to quantify imaging sticking on LCD panels has been proposed in 2002 [5]. It is based on the use of a 2D luminance meter and chessboard pattern generator. We have applied the same type of method – adding in addition the notion of temporal dependence - to PDP using our new accurate and rapid 2D luminance meter MURATest. An automated method to analyze the temporal dependence of the patterns and adjust exponentials decays models has been implemented. 2. Description of the Method 2.1 2D Luminance Meter Requirements Image sticking phenomenon requires a high signal/noise ratio for the measurements even if the effect is quite strong on PDP displays. Indeed, the human eye sensibility is in many aspects very high and residual image pattern less than 1/10000 of the original level can be easily detected depending on the conditions. The second main requirement is the measurement speed to be able to follow the temporal dependence of the effect as closely as possible. ELDIM MURATest has proven through years its excellent capabilities for 2D luminance (and color) imaging, thanks to a cooled CCD sensor and a high quality optics. In the third generation of the product, we use a Kodak CCD sensor with a real 16bits A/D converter and dark current compensation that ensures very low noise and high dynamic range. 2.2 Description of the Measurements The experiment is completely computer controlled at all its steps. A black screen is first applied to suppress any remaining image sticking of a previous experiment. The duration depends on display quality and on the previous experiments applied to the display. It can be as long as one hour for displays with very high image sticking. A specified pattern is then applied on the display during a given time (typically 5 to 15 minutes). The luminance of the panel is taken during this step and use as a reference for the temporal analysis. Afterwards the panel is switched to black or white state and real time luminance measurements of the entire surface of the panel are performed versus time. For white state switching the integration time of each measurement is fixed all along the experiment since the luminance of the panel is always high and around the same values. For black state switching the integration is adapted versus time for an optimize evaluation of the temporal dependence (as short as possible just after switching and as long as 30s at the end of the experiment). As an example, figure 1 shows some luminance measurements taken on one PDP using a chessboard pattern. On figure 1.a the original chessboard pattern is reported and its remaining image after 100 seconds due to image sticking for dark and bright configurations are reported in figures 1.b and 1.c respectively. The remaining image of the chessboard pattern is easily detected in both cases. One can notice also that white cells of the chessboard remain brighter than the black ones when switching to dark background (dark image sticking) whereas they appear darker by switching to bright background (bright image sticking).