When we test solid-state lighting products for longevity, the driver is often the first point of failure. When will be be able to project driver lifetimes with confidence?
At the LEDucation 8 conference in New York, Tom Ward of Finelite threw down a gauntlet at the feet of those who design and manufacture drivers for LED lighting. We know how to estimate the lumen depreciation, and therefore one important element of the working lifetime of a luminaire, using LM-80 testing extrapolated via TM-21 calculations. No such standards exist for driver longevity.
Accelerated aging tests of luminaires often show that the driver is either the first point of failure or among the first. For someone putting together a luminaire or a lighting solution, the expected lifetime of the driver under operating conditions is one crucial piece of data that is often difficult or impossible to find. "The data exist" internally at each driver manufacturer, Ward said. The time will come when driver makers will provide this data in the technical specs of their products.
Testing and analysis
Each manufacturer has its own methodology for testing its products, including lifecycle testing. Some probably use analytical methods involving component lifetimes to ballpark a mean time to failure for their drivers. They may figure in the reliability or process consistency of their suppliers. What are the best-practices? The industry hasn't decided. In fact, as far as I can determine, the players barely speak to one another on this subject. Some driver makers -- perhaps most -- may consider their testing or estimation methodology a trade secret.
The situation seems ripe for a conversation leading eventually to standardization, so that if three driver makers each say their product will last 50,000 hours in use, a luminaire designer can have some confidence they are all speaking the same language.
At LEDucation, I spoke with representatives of two companies that make drivers, and I got diverging viewpoints on this issue. TerraLUX's Eric Milz seemed open to such a conversation, while Lutron's Tom Shearer was happy to continue going it alone.
Milz cited one aspect of the testing situation I hadn't considered: Each company's culture, in the form of the industry background of its staffers, will go some way toward determining how testing is approached. TerraLUX is based in Colorado, and many of its engineers have a background in data storage. Other companies will differ.
Shearer said he considers matters such as how Lutron builds in quality and arrives at a conservative lifetime estimate to be part of its "secret sauce," wrapped up as it is in longstanding supplier relationships and the company's corresponding deep understanding of their capabilities and processes.
What body should convene a conversation and begin thinking about best-practices and standardization for claims of driver longevity? TerraLUX's Milz said that NEMA is the obvious candidate, but it has its hands full thinking about issues such as dimming and flicker at this time. Shearer said Lutron doesn't participate in NEMA.
What do you think? Would you like to see movement toward speaking a common language when we state a driver's expected lifetime?
— Keith Dawson , Editor-in-Chief, All LED Lighting