Life time comparison of LED package and the self-ballasted LED lamps by simple linear regression analysis

The energy efficient long-life self-ballasted LED lamps are designed considering the reliability test result of LED package. We also try to predict the lumen maintenance life for the self-ballasted LED lamps because there were a lot of failure mechanisms for a various environment stress. Many studies have investigated the environmental stress effects on LED packages [2], [3],4, and 5]. However, comparing to LED package, there is not enough data to study the lumen maintenance life for the self-ballasted LED lamps. IES LM-80-08 is the approved method for measuring lumen maintenance of LED light sources [1]. It suggests using the Arrhenius Equation to calculate the interpolated lumen maintenance life. We followed this method and obtained lumen degradation patterns from three different thermal stresses as indicated in the IEC 62621 standard. Under the three different thermal stresses, the on-off test (30 on and 30 s off) was conducted on the self-ballasted LED lamps which were made of the same selected LED package (3 V, 64 mA). The intent of the research isn't to set up a new curve-fit for non-linear lumen degradation pattern. Our research goals are to find methods to compare the lumen degradation pattern of LED package and the self-ballasted LED lamp using simple linear regression analysis. As simple straight linear line is assumed for lumen degradation after ALT testing We found the slope of the mathematical model is easily comparable for three different temperatures and for different number of LED whether it is single or consists of forty or eighty LED. By comparing each values of the slope the effect of increase in number of LED can be shown at each thermal stress. This study is the first step towards the presentation on the lumen maintenance life of the self-ballasted LED lamp on three different temperatures. Additional information must be accumulated in the future to further study on the LED lighting product reliability.