標題: 發光二極體元件與系統之最佳化設計製作與應用
Optimal Design, Fabrication and Application of Light Emitting Diodes and Its Systems
作者: 蔡哲弘
Tsai, Che-Hung
趙昌博
Chao, Chang-Po
電控工程研究所
關鍵字: 發光二極體;基因演算法;背光模組;光子晶體;有限時域差分法;主動像素電;Light-emitting-diode;Genetic algorithm;Backlight unit;Photonic crystals;Finite-difference time-domain;Active pixel sensor
公開日期: 2013
摘要: 本論文主旨為研究發光二極體元件之最佳化設計製作與應用。首先,在發光二極體(Light emitting diodes, LEDs)元件上設計一創新錐形透鏡來改善每一顆發光二極體的光源強度分布與提高光源的均勻性。並透過TracePro光學軟體的模擬與基因演算法(Genetic algorithm, GA) 來估算每一顆RGB LED模組於整體背光模組(Backlight units, BLU)中最佳的放置角度,進而提高色彩飽度和降低色差,並將最佳化的結果與實驗做驗證。另一方面,本論文亦研究提升高分子發光二極體(Polymer light emitting diodes, PLEDs)之取光效率(Light-extraction-efficiency, LEE),透過製作光子晶體(Photonic crystals, PCs)結構於PLEDs上來增加其效率。而光子晶體之間距與半徑為影響增加效率的兩個重要因素,故本論文利用基因演算法(GA)針對光子晶體來做最佳化之設計,以尋求其最佳化之間距與半徑。此外為了精確地估測PLEDs之取光效率,使用3D有限時域差分法(Finite-difference time-domain, FDTD)之軟體來計算。對於計算PLEDs這種複雜多層的結構,FDTD的方法是很有效率的。除了光子晶體能提升PLEDs的取光效率外,側壁反射(Sidewall reflector)結構也是另一種提升效率的方法。故本論文針對兼具光子晶體與側壁反射結構之PLEDs作最佳化設計,以提升更多的取光效率。在製作具光子晶體結構之PLEDs,本論文採用奈米壓印(Nano imprint lithography, NIL)與聚焦離子束(Focus ion beam, FIB)這兩種方式來實現。並根據3D FDTD軟體之模擬結果,來闡明所設計之最佳化光子晶體結構能有效提升PLEDs之取光效率。最後在LED的應用方面,本論文將其應用於光近接感測面板系統上。此系統的前端是由發光二極體及光接收器(PD)組成的陣列,由光接收器接收發光二極體發光照射至待測物體所產生的反射光,再由反射光的強度來判斷與物體的距離。此研究提出一新式主動像素電路(Active pixel sensor, APS)並結合三維光學近接感測電路來消除背景光,以達到不需觸碰面板即可偵測出與物體的距離。此設計的主動式像素電路具有將光感測器偵測到的光轉為電壓值的作用,與一般常用的3T架構主動式像素電路(3T-APS)有較好的輸出解析度,而感測的結果經由計算3D物體距離的演算法與LABVIEW的程式設計來作驗證。
This dissertation aims to study the optimal design of fabrication and application in light emitting diode (LED). First of all, the novel cone-shaped cap is proposed to gain higher efficiency and provide satisfactory uniformity for each LED. The optimal design on angular placements of LEDs presented in this dissertation for satisfactory color-mixing and emission uniformity is achieved by necessary optics simulations via TracePro, followed by utilizing an intelligent numerical optimization technique, Genetic Algorithm (GA). The design parameters for GA optimization are different combinations of LED placement angles in a backlight unit (BLU) module. The experiments are conducted, which successfully validate the expected performance of color balance and emission uniformity for a novel cone-shaped LED lens with optimized angular placements in a large-area backlight module. In another way, to improve the light extraction efficiency (LEE) of polymer light-emitting-diodes (PLEDs) via photonic crystals (PCs) is presented in this dissertation. The pitch and radius of photonic crystals are two important factors for enhancing the light extraction efficiency of PLEDs. The intelligent numerical optimization method, GA is used to seek the optimal parameters of PCs, pitch and radius. Besides, in order to accurately predict the LEE of the PLEDs, the numerical simulation tool, the 3D finite-difference time-domain (FDTD) software is utilized. The FDTD method in modeling complex multilayer PLEDs device is very effective. In addition to increasing the LEE of PLEDs via photonic crystals, the sidewall reflectors structure is another way to improve the LEE. To achieve more light extraction efficiency, both PCs and sidewall reflectors are used in the simulation of the optimal design. In this dissertation, nano imprint lithography (NIL) and focus ion beam (FIB) are employed to fabricate the PLED with photonic crystals. The PLEDs with optimized PCs would enhance the LEE efficiently and this result could be explained by 3D FDTD calculation. Finally, optical proximity sensing panel system is proposed for application of LEDs in this dissertation. This frond-end of system is composed of LEDs and photo-detectors (PDs) array. The photo detectors receive the reflections of light from the measured object, and the system estimate the distance of the object via the intensity of light. A new active pixels and the associated 3D optical proximity sensing circuit with eliminating the effects of background lighting are proposed. The active pixel circuits for outputs of photo-detectors are designed to convert the light to the voltage, and it can improve the resolution output range, and reduce the noise, as compared to the conventional active pixel sensor (APS) with a 3T-structure (called 3T-APS). Toward success of the panel, background light is eliminated by specially-designed circuitry, while a computation algorithm for 3D object detection is developed based optics principles. Sensing results are finally illustrated by a LABVIEW program.
URI: http://140.113.39.130/cdrfb3/record/nctu/#GT079712801
http://hdl.handle.net/11536/73840
Appears in Collections:Thesis