The Effects of Pressure, Moisture, and Temperature on the Reliability of High-Power LED Packages

• Main Authors: Wan Lin Tsai, 蔡宛霖
• Other Authors: M. Y. Tsai
• Format: Others
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/23311458225308610276

碩士 === 長庚大學 === 機械工程研究所 === 96 === The purpose of this study is to evaluate the effects of pressure, moisture and temperature on the reliability of the high-power COP (Chip on Plate) LED packages, by testing their packages and modules through three different standard tests: pressure cooker test (PCT), wet high temperature operating life (WHTOL), and thermal shock. First, the thermal behaviors of the LEDs in the test process are investigated by experimental measurement, and to observe the failure mode of the LEDs. Second, the effect of the die attach moisture absorption on the thermal resistances of the LED packages are studied through experiment and finite element method (FEM). Third, different environments in the test chamber (such as natural and forced convections) resulting in the change of thermal resistance of the LED packages will be evaluated by experiments and FEM. Finally, the luminous flux degradation of the LEDs after tests will be determined by experimental measurements. The PCT results show that all COP LED packages without silicone encapsulants pass for 96 hrs, but all those with silicone encapsulant fail after 56 hrs. However, replacing original Al wire with Au wire can make all those with silicone encapsulant pass in the pressure cooker test. The results of the WHTOL test indicate that all COP LED packages with phosphorus in the silicone encapsulant (i.e. white light) fail after 309 hrs, but all those without phosphorus (i.e. blue light) pass for 1008 hrs. Nevertheless, replacing original Al wire with Au wire can make all blue- and white-light LED packages pass the WHTOL test. Moreover, among the passing packages, their thermal resistances increase during WHTOL test by up to 12 oC/W. This is due to the decrease of the thermal conductivity of the die attach, resulting from the moisture absorption. Furthermore, the forced convection environment in the test chamber would lead to the decrease of the chip junction temperature and the junction-to-air thermal resistance of LED packages. The thermal shock results show that all blue- and white-light LED packages and their modules pass for 200 cycles. With the increase of test cycles, the thermal resistances of blue-light LED packages and their modules do not show much change. Besides, the luminous flux degradation of white-light LED packages and their modules are more obvious than blue-light LEDs.