Solid-state lighting is making headlines as the next generation light source for lighting homes, offices, transportation, and others spaces, both indoors and out. Yet, challenges remain to create white LED lighting systems that offer better performance than some traditional light sources, including higher light output, improved efficacy, and better color.
In addition to its application studies and field evaluations of solid-state lighting, the Lighting Research Center has been exploring the basics of LED performance at the package and system levels. Two studies were published this fall by the International Society for Optical Engineering (SPIE) as part of its proceedings for the Sixth International Conference on Solid State Lighting, held August 14-17 in San Diego.
Measuring heat as a performance predictor
Heat affects the spectrum, light output, and life of LEDs. Yet, once LEDs are packaged into a lighting system, it is difficult to measure their operating temperature without disassembling the system, says Lalith Jayasinghe, Ph.D., senior research specialist at the LRC. That, he says, can alter the system’s performance. Consequently, researchers use indirect methods of measuring system heat. The most common method requires knowledge of the LED’s thermal resistance coefficient, a measure of how much resistance the heat flow encounters between the LED junction (where the semiconducting layers of the LED die meet and light is created) and the electronics board where the LED is attached..
“Most LED manufacturers characterize the thermal resistance of their products for only one drive current, ambient temperature, and junction temperature. But once LEDs are packaged into a system, such as an MR16 lamp, all of these parameters may change, which may alter the thermal resistance,” says Dr. Jayasinghe. The type of heat sink material and adhesive used in the package can also change the thermal resistance. “Without an accurate thermal resistance measurement, we can’t adequately estimate LED junction temperature or predict life.”
The LRC studied how the junction-to-board thermal resistance of high-power LEDs varies with changes in ambient temperature, power dissipation, and the amount of heat sink material. The research team found that increases in power dissipation increase the thermal resistance, while increases in heat sink material decrease the thermal resistance. Increases in ambient temperature have different effects, depending on the type of high-power LED. However, under operating conditions a linear relationship was found between junction temperature and board temperature.
“Our next step is to find a relationship between the outer surface temperature of an LED system and the junction temperature, which would provide a fast and easy way to determine the life of an LED system at any temperature,” said Jayasinghe.
More information, including a technical paper, about this study, “Characterization of Thermal Resistance Coefficient of High-power LEDs,” is available on the LRC Solid-State Lighting Web site. The study was funded by sponsors of ASSIST recommends (listed at www.lrc.rpi.edu/programs/solidstate/ASSISTSponsors.htm).
Dimming white LED systems
Dimming is a necessary function for many lighting applications, but it can change the appearance and performance of a light source. Color shift is the most common change that occurs with dimming, as found in a previous LRC study of white LED systems. However, efficacy can also be affected. The LRC explored how different dimming modes alter the peak wavelengths and luminous efficacies (lumens per watt) of white and colored high-power LEDs from three manufacturers.
LRC research specialist Yimin Gu says two types of LED dimming are available: continuous current reduction, which is a decrease in the forward current applied to the LED; and pulse-width modulation, which is a decrease in the duty cycle, or the on-off cycling rate.
As a result, she says, an RGB-based white LED system would experience a considerable color shift when dimmed if an electronic feedback control were not added to regulate the shift. Adding feedback systems increases the cost of RGB-based LED systems.
In terms of luminous efficacy, the LRC found that dimming by pulse-width modulation caused a significant efficacy loss for both white and RGB LED systems, especially at low dimming levels. “Therefore, a phosphor-converted white LED system with continuous current dimming would be a good solution for occasions where cost, color maintenance, and energy savings are critical,” says Gu.
This study was funded by sponsors of the Alliance for Solid-State Illumination Systems and Technologies (ASSIST). Sponsors are listed at www.lrc.rpi.edu/programs/solidstate/ASSISTSponsors.htm.
More information about this study, “Spectral and Luminous Efficacy Change of High-power LEDs Under Different Dimming Methods,” is available on the LRC Solid-State Lighting Web site.
For more about the LRC’s LED research, visit the Solid-State Lighting Web site (www.lrc.rpi.edu/programs/solidstate/).