本論文建立一套暫態熱阻量測系統(thermal transient measurement)來探討發光二極體一階封裝的熱傳行為。暫態熱阻量測系統包含自動化量測系統與量測結果之數值分析處理，皆以JESD51-1與JESD51-14為量測與分析的基準。藉由量測發光二極體的降溫暫態響應，可得結構方程式，由此間接得知熱流在發光二極體封裝體內的行為，包含熱流之等效等溫面面積與其在熱傳路徑上的分佈與位置，由等溫面之間的關係定義熱阻值，進而得知各材料層的熱阻。本研究藉由發展出的暫態熱阻量測系統分析高亮度GaN-based 發光二極體封裝於不同尺寸的銅散熱基板之熱阻，銅基板包含直徑與厚度的變化，藉由結構方程式的結果，將銅基板到heat sink的熱阻分成兩部分，第一部分為熱流在基板內從晶片下方放射狀傳遞的區域，為基板的主要熱阻RsubP；第二部分為接續第一部分末的等溫面到heat sink與環境界面的等溫面間的熱阻，為基板後端經散熱膏層到heat sink之總合熱阻RαGS。相同厚度的基板，直徑的變化對RsubP沒有顯著影響；因散熱膏在小直徑基板與heat sink 間具有較好的接觸，因此散熱膏層的熱阻較小，但大直徑基板具有較好的熱擴散效應，此兩種效應交互影響RαGS。同直徑的基板，RsubP與基板厚度成正比關係，等溫面面積亦快速擴散，因此RαGS隨基板厚度增加快速遞減，使得整體熱阻RSGS (RsubP + RαGS)有效下降；當基板厚度過大時，RαGS遞減趨勢大幅減緩，熱流在基板內的擴散效應達到飽和，過厚的基板增加一維熱阻，反造成RSGS增加。;This study discusses the thermal transient measurement of first-level Cu substrate used in high-brightness LEDs. In the Chapter 3, the thermal transient measurement is introduced. The thermal transient measurement contains an automatic measurement system and a data processing system, and both of them are based on JESD51-1 and JESD51-14 standards. Measuring the transient variety of the junction temperature can obtain the structure function. The structure function indirectly shows the behavior of the heat flow in the LED package that includes the cross-sectional area of isothermal surface and its distribution. The thermal resistance of each layer is defined by the relationship between the isothermal surfaces. In the Chapter 4, the thermal performance of a series of Cu substrates with different geometric shapes is investigated by the present thermal transient measurement. The thermal resistance from the Cu substrate to the heat sink can be divided into two parts as the result of the structure function. The first part is the main thermal resistance of Cu substrate (RsubP) that is the radial transmission region of the heat flow from the top substrate to the bottom. The second part is the sum of the thermal resistance (RαGS) that includes the remaining Cu substrate, the thermal grease layer, and the heat sink. The varying sizes of diameters would not influence the RsubP obviously. Smaller substrates have lower thermal resistance of thermal grease layer, but the bigger substrates have better spreading effect. These two effects interact with the RαGS. The RsubP is proportional to the thickness of the substrate with the same diameter, and the cross-sectional area spreads effectively with the increasing thickness of the substrate. Therefore, the RαGS decreases rapidly with the thickness of the substrate, then, the sum of the thermal resistance (RSGS = RsubP + RαGS) reduces effectively. Nevertheless, if the substrate is too thick, the spreading effect approaches the saturation, then, the thick substrate adds the one-dimensional thermal resistance that also increases the RSGS.
