Firstly, what do specifiers need to know about LEDs? "They need to know that the technology works, and the effects that can be achieved, but not the intimate details of LEDs," says Effron. "They specify a working product, not a technology." In part, this is because LED technology is highly complex; the devices are based around high-purity semiconductor crystals that are grown in a high-temperature environment, which is far removed from the traditional "glass, brass and gas" construction of conventional light sources.
Two other key questions are as follows: what does the LED industry give the lighting design community in terms of products and information; and what do lighting designers actually want?
"The key areas that specifiers worry about are predictability, reliability and stability," says Effron. From the point of view of predictability, the lack of standards in the LED industry continues be a cause of concern. However, work has been done in the automotive industry by the Society of Automotive Engineers (SAE); in the area of optical measurement by the International Commission on Illumination (CIE), and by other bodies such as IES and NEMA.
Specifiers are also concerned with performance, wondering if it will be consistent either within a batch, between batches from the same manufacturer, or from manufacturer to manufacturer. In terms of stability, specifiers want to understand that the product is well manufactured, so that it won't fall apart, and that the performance will be as stated.
Metrics
Appropriately-used metrics are obviously important, as Effron explains. "Specifiers are looking for the same kind of information that they would get from traditional light sources," he says. While this should be straightforward for white LEDs, photometric measurements incorporating lumens are problematic when considering colored LEDs. For example, a blue LED might be obviously bright to an observer, but a photometric measurement would have a low value due to bell-shaped V-lambda distribution curve that factors the photopic eye sensitivity. "The color issue might require the lighting community to rethink lumens," says Effron. "Many of the metrics are quite old and color rending index (CRI) is totally antiquated."
Binning is another issue with LEDs, although most specifiers have experience of seeing variations within batches of other lighting fixtures. Even so, they're not necessarily familiar with the 3D binning matrix of color, forward voltage and flux that occurs with LEDs, but they are critically aware of consistency.
The industry is still lacking a standard definition of lifetime, taking into account that although LEDs don't fail suddenly, their output degrades over time. Lifetime is usually defined as the time taken for the LED output to fall to a certain level e.g. 70% of its original luminous flux. Various parties have proposed different "lumen maintenance" values that could be used, but a figure that might be appropriate for a wall sconce application, where the lighting is more decorative than functional, could be completely unacceptable for a lighting consultant fitting out an office building. What is more, colored LEDs shift over time and there needs to be a metric for judging the useful life of these as well.
"LED companies either don't provide the information, or provide it in the wrong format, or both," says Effron, citing the example of an LED manufacturer supplying graphs showing the light distribution from some of its products - the graphs used Cartesian coordinates, whereas lighting consultants are used to looking at polar coordinates, which show the pattern of light.
Knowledge gap
Many of the points above illustrate the "knowledge gap" between LED manufacturers and the lighting design community. While some LED equipment manufacturers can act to bridge that gap, others lack the appropriate expertise and simply put LEDs into fixtures because it’s all the rage. "Companies such as Permlight have people like myself that come from a lighting design background," says Effron, "while other equipment manufacturers don't know anything more about lighting than the chip makers do."
The knowledge gap needs to be addressed in time for LEDs to start to compete in general lighting applications on a level playing field. "Information about the source needs to be available to those people on the early adopter curve," says Effron. "Lighting consultants and other specifiers need to be able to make informed decisions."
Effron also questions the availability of different LED products. "If I have an interesting idea, based around a dense array of low power, warm white LEDs, as opposed to less dense array of high power LEDs, where can I get these devices? With other lighting technologies, the choices are available. As a luminaire manufacturer, serving the specification community, I want to be able to make decisions based on the application, rather than being restricted by the product ranges of the LED makers."
Decisions regarding the available products tend to be made by the chip makers who have a focus on particular markets. For example, Lumileds has a strong focus on the lighting market, while rivals such as Nichia, Cree and Osram Opto Semiconductors are more deeply involved in other markets such as mobile phone handsets and automotive lighting. Yet everyone seems to agree that general lighting is the "Holy Grail" for chip makers.
Clearly, says Effron, if chip makers had engaged lighting designers in the process, then they might have gone down some different development roads. "The question is, at what point do you bring in the end user - the specifier, or early adopter - to help develop a product that might be accepted more rapidly in the lighting market."
Effron's presentation ended with a tantalizing question - will a "brand" name chip ever be specified by a lighting designer, interior designer or architect? This will depend on the LED makers doing a better job of addressing the needs of, and communicating with, the lighting design community.
***
We welcome your comments on this or any other LEDs Magazine article - please contact Tim Whitaker.
