Solid State Lighting

The promise of solid-state lighting (SSL) has been touted for nearly two decades with most industry experts and observers noting that once the light sources reached a benchmark of 50 lumens per Watt, a revolution in general lighting would begin. With the recent introduction of many high-power LEDs claiming outputs in excess of 50 lumens per Watt, the question is, "What are we waiting for?"

To adequately answer this question, we have to look at existing standards for competing light sources such as incandescent bulbs and compact fluorescent lamps (CFL). There are established terminology and photometrics with recognized standards that many lighting specifiers have come to depend upon. These standards are the benchmark by which a comparative analysis, formal or informal, is being made against SSL technologies. How does SSL efficacy compare against traditional efficacy? Are we comparing apples to oranges when we look at specifications? Is there a linear correlation between luminaire efficacy, light source efficacy and system efficacy?

First, high-power LEDs, such as the Cree Xlamp and the Lumileds Luxeon, measure the lumens output in a very narrow time window of 25 nanoseconds to be exact, and at a stable junction temperature of 25°C. When a lighting specifier sees 65 lumens output produced at a forward Voltage of 4.0 and a forward current of 350 mA, they would assume that when they turn on a light source they would get an efficacy of 46 lumens per Watt. The other assumption would be that a nine LED lamp using the same LED would produce an output of 585 lumens. It seems very comparable to a standard 50-Watt incandescent bulb. This is where what is being measure and what that measurement is being compared to diverge.

When a light bulb is screwed into a socket of an existing luminaire, the light bulb is measured independently of the fixture. The luminaire efficacy, or how much useful light exits from that fixture, is not taken into consideration when measuring the light source. A modern A19 bulb emits light in a nearly spherical pattern. What happens when that light source is placed in a recessed can? With most modern SSL luminaires, the light engine, or light source, and the driver are integrated into the luminaire, creating the need to measure the luminaire efficacy. Comparing the system efficacy, or the potential of a light source, against luminaire efficacy will not give an accurate picture of the difference between light sources. That is a real apples-to-oranges comparison.

What makes it even more difficult is that many SSL luminaire manufacturers create their output specifications based upon the LED manufacturer's specifications as noted above. This is a completely inaccurate method for determining the actual real-world output. We do not use light sources in 25 nanosecond bursts with a 25°C temperature. There is output to temperature degradation, there is a decline in output over time and there are fluctuations of input current that is correlative to temperature and input Voltage variations. It is critical to communicate these differences in a way that makes some sense to consumers and specifiers. Meeting the expectations of both is paramount.

Take packages from two standard CFL lamps manufactured by company ABC, and a package of 60-Watt incandescent bulbs as well from XYZ. The CFL lamp with a PAR shell claims 640 lumens (equal to a 65-Watt lamp -- manufacturer's claim), "warm white" color temperature, energy use of 14 Watts and a life expectancy of 8,000 hours. This lamp also carried the ENERGY STAR and UL marks. The Spiral CFL claims 800 initial lumens, 640 "mean" lumens, "warm white" color temperature of 3,000K, 13 Watts consumed and a life expectancy of 8,000 hours. The Spiral CFL has the UL mark, but no ENERGY STAR certification. The A19 standard incandescent claims 600 lumens, no claim on color temperature, 54 Watts consumed and a lifetime of 2,800 hours. There are no ENERGY STAR or UL marks.

Now, take a lamp from Infinilux that is yet to be released to the general public. The initial claims are 480 lumens, "daylight" 5,000K-color temperature, 9.2 Watts consumed and a lifetime greater than 50,000 hours. This product will bare the UL mark, but no ENERGY STAR standard has been agreed upon as of publication.

If we break that down the way a typical consumer would, we would find that the CFL PAR and Spiral CFL provide 45.7 lumens per Watt, the incandescent provides 11.1 lumens per Watt and the SSL lamp provides 52.2 lumens per Watt. Just by the simple numbers, one could assume that the SSL lamp is the most efficient of the group. But when you measure the output using a standard calibrated Lux meter and a Watt-Up Pro Watt meter, the results are surprising.

Imperially, each of these lamps can be powered up side by side and each lamp, while showing similar outputs on paper, will produce a radically different output characteristic. The CFL PAR has a type of PAR reflector and cover glass, giving it a narrow flood without the loss of light to the rear. The Spiral CFL emits light in a sphere and looks significantly brighter than the standard incandescent lamp even though the published output specifications are very similar. The A19 emits a warm light in a sphere, and side-by-side with both CFLs looks to be 30 percent of the brightness. The SSL lamp emits 100 percent of its output in a 100° flood. When I place these lamps into a common recessed can picked up at a local home improvement store, the variation is more radically pronounced. In open air, there is not as significant a difference in usable light to the naked eye.

This means that standards are necessary to define the unique output of SSL lamps and then a correlative must be created to bridge the gap between the outputs. There are standards groups right now pouring time, talent and treasure into helping the industry define these metrics (ANSI, NEMA, MIST and CIE are heavily invested in this process). There will be not only a standard for output (LM-79), but for color (ANSI C78.377A), lifetime (LM-80) and electrical performance (ANSI C78.XX3) as well. Uniform measurement standards and test procedures are being created as I write, and many are in final draft form. This will equip the SSL industry to communicate to the consumer and lighting designer what they can expect from a luminaire/light source.

The bottom line is that by the end of 2007, there will be accepted standards for measuring the light output of SSL luminaires and the communication of those measurements to the general public in a way that will allow them to use their existing preconceptions about lighting. We will have an appropriate answer to the common question, "How many Watts is that?" without having to re-educate the end user and the specifier. We will have determined how many input energy Watts will be required to generate the appropriate output per task, giving the consumer and specifier the power to make good decisions. The technology is there, the efficiency of the SSL solution cannot be disputed, but how the efficiency, efficacy and true cost per lumen over the lifetime of a light source will be communicated is still up in the air.

Anthony W. Vilgiate has been in the LED industry for 20 years and has numerous patents and awards for his designs. He currently is vice president of Infinilux Incorporated, focusing on the development of solid-state lighting for the general, commercial and industrial lighting markets.