Abstract
Through the technology developing rapidly, the concept of environmental friendly would be in our mind. The world faced the problem of globe warming. In daily life, light-emitting diode(LED) had characteristics of low pollution, low power consumption, and long lifetime. The reliability test of IES LM80-08 was in common use to the manufacturers of LED. It would cost too much time and prolonged the time-to-market. This situation can be regarded as an obstacle to the LED research and development. The purpose of this research would observe the factor of current stress and temperature stress in accelerated aging test, and aimed at (i) fitting the appropriate life prediction model, propose the accurate lifetime, (ii) investigating the accelerated factor in the test with current stress and temperature stress. (iii) proposing an available method to shorten the reliability test standards, promoted the light-emitting diode industry. The research would measure the junction temperature of LED at ambient temperature. Building the finite element model by the sample to understand the transportation of the heat in the packaging. The simulation model would assist the accelerated aging test design. The result of the simulation would compare with experimental junction temperature measurement to validate the model. The sample consist of white LEDs, and blue LEDs in order to separate the influence of phosphor. The test would use the finite element model to predict the target junction temperature which is consider both of temperature and current joule heating. The test would carry out based on the standard IES LM-80-08 and predict the lifetime by TM-21-11. The degradation of light output was detected during aging. The research analysis the experiment result in blue LEDs and white LEDs, respectively. It could propose the influence of current effect in HP-LEDs. Finally, the result would compare with the previous research which is only consider the temperature stress. The result reveal the current stress is an essential factor in life prediction model.