Extrapolation of lifetime of high power LEDs under temperature-humidity conditions

Lifetime of high power LEDs as lighting source is needed for its widespread adoption, and extrapolation of lifetime from accelerated stress is needed. Current practice of extrapolation is to use TM-21 and Peck model. In this work, we show that the use of Peck model is not correct, and the use of TM-21 must be done with cautions for the temperature range and the identification of possible mixture degradation mechanisms is needed before performing the extrapolation. Extrapolation of lifetime beyond 10% degradation may not be trivial due to the change in the degradation mechanism.