LEDs bring a lot of creative potential to the lighting market. At this creative front will be artists who apply them in unique ways outside the mainstream general illumination marketplace. I’m one of them in fact. Gone are my days of fooling with halogen lamps and transformers – now its LEDs, drivers and power supplies… and heat sinks.
The first wall hit in application of LEDs is their output and thermal dynamic. In halogen lighting, the fixtures got hot, no big deal. As long as it wasn’t peeling flesh or raising a blister, its was fine. LEDs don’t work that way. Long before the lighting portion gets hot enough to raise a blister, the LED is fried, game over.
Ultimately, the goal of an artist in creating a new work of light is to generate as much light as possible in the smallest, most innocuous package. That makes LEDs at once attractive, and a problem. To illustrate, a 350mA 1W LED requires very little thermal management to survive. Glue or screw it to plate of aluminum or copper and its happy. Problem is, the amount of light is pretty weak, between 35 and 60 lumens if you want 3,000k color with good color rendering (yes, artists do care about color – a lot.) That’s a long way from the 600 to 800 lumens we were used to in halogen sources. Optimally, for a task light with a good optic (wide smooth beam), gettting to around 350 to 500 lumens is optimal. This means more energy. At 700mA, 100 lumens is attainable, and lands us in the 3 to 4 watt range. Three is always better than 10, so this sounds like the way to go. Problem is, a 3W LED is exponentially more demanding thermally than that 1W device. These higher energy devices demand heat sinks, real life thermal management, with airflow and everything.
So far, we are in the neighborhood of around 35 lumens per watt for a 90+ CRI, 3,000k, single point LED device. That means for 350 lumens we will have 10 watts of energy to deal with. The heat sink and associated airflow around it demands as much space as we once committed to a 20W or 50W halogen lamp – so not much gain there.
This is where efficiency comes into play. If we project forward, and say the 35 l/w of 2009 imporves to 120 l/w in time, the amount of energy within the same 500 lumen design drops from 10 watts to a little over 4. If this were delivered in the 1W 350mA package, we’d be down from 10 LEDs to 3 or 4, and back to virtually no heat sink demand, beyond a little conductive material to transfer heat to the body of the finished piece. Cool!
Creative arts lead to creative new applications of technology, and new design concepts. At this moment in time, LED performance is very good, much better than the halogen lamp for the same general poitn source illumination. However, the thermal issues are creating as much of an issue as the low lumen output of the individual sources. This is stiffling creative adoption of LEDs somewhat. As efficiency improves, the artistic potential of LEDs surpasses anything available today. We are at a threshold now. For some designs, the 1W LED serves well, and thermal management of a 4W device is near tolerable. As this improves, we will eventually reach a point where the creative potential and pain of adoption lines cross. At that point, there will be an explosion of new, quality artistic designs, creative works, and sculpture, incorporating LEDs in all new ways.
I, for one, intend to be there for every step. While the optical issues remain challenging (color separation being a big one for white sources), and controlling color in a self contained product a dilemma (expensive and complex), and the electronic gadgetry fussy (less user friendly than halogen by a fair margin), and the technology lost on gallery owners (oh well…), the potential for creative use of these devices is greater than anything any sculpturist in light has had to date.
Efficiency is more than an energy power density variable, it is also a factor in the creative adoption of the technology. So, counter to what many may think, artists do care about getting more light for less energy, even if for reasons beyond the engineering agenda.