Here's the reason the lamps in your living room probably have incandescent or fluorescent bulbs rather than LEDs: Incandescents and fluorescents still generate more visible light per watt than an equivalent LED. However, in applications such as brake and taillights, display panels, industrial controls, and traffic signals, LEDs outshine incandescents in energy efficiency.
One-third of all vehicle third brake lights are red LED clusters. Many carmakers use LEDs because the third brake light is often inaccessible and replacement is difficult. LEDs have a long life and probably will outlast the car.
It's helpful to compare a red LED with a red-filtered incandescent lamp. A red LED might be as much as 10 times more visible than the red-filtered incandescent lamp because the red filter reduces the incandescent's light output. Unlike an LED, which is monochromatic, an incandescent light emits the full spectrum and therefore requires filtering to produce light of a specific color.
Placed under a lens, an LED lamp cluster can suffer up to a 30% reduction in brightness depending on the lens color and thickness. The lighter the lens pigmentation, the brighter the LED appears. Additionally, most of an incandescent's energy, 80 to 90%, is wasted as heat.
One new application for LEDs is in computer-controlled lighting. The fact that LEDs are fundamentally digital in nature means they can be connected to a computer and controlled to produce amazing color and lighting effects. That's what one company is doing to gain a foothold in the lighting market. Boston-based Color Kinetics, (www.colorkinetics.com), developed a patented lighting technique it calls Chromacore. It uses a microprocessor to control the mixing of multicolored LEDs, generating millions of colors and an array of color-changing effects.
Chromacore integrates a controller, LED modulation and drive electronics, the LEDs themselves, and a network. The network can be any physical layer or protocol, but the initial choice was one called DMX because it was widely used in the theatrical industry. However, systems can interface with other networks such as Ethernet. There are also devices that take serial RS-232, parallel ports, and USB and connect to the lighting systems. Similarly, other architectural control systems such as Lutron, Vantage, and LiteTouch are compatible with the system.
The wizardry behind Chromacore starts with the light source, consisting of clusters of LEDs which are then arranged into channels of control. For example, a channel might contain a group of LEDs of one color, such as red. Others might be green or blue. A microprocessor controls individual channels.
Light intensity is controlled by rapidly switching the channels of LEDs on and off at a kilohertz rate via pulse-width modulation. Switch speeds in the kilohertz range are enough to avoid flicker. Varying the output of each LED channel creates a vast array of lighting effects. For instance, this method of additive light mixing will produce any of 16.7 million colors via three LEDs (red, green, blue) and 8-bit control for each color.
"Driving LEDs with constant not varying current prolongs lifetime and creates no color shift or adverse thermal issues," says Kevin Dowling, vice president of Strategy and Technology for Color Kinetics. The company adjusts the ratio of on-time to off-time to change apparent brightness.
More recently, Color Kinetics has developed Chromasic, a single ASIC that integrates its Chromacore system onto a single chip. And a recently awarded patent describes LED-based lighting networks as a means of wireless communication through a local-area network.
Light from nature
While LEDs have moved from indicator lights to sources of illumination, another quiet revolution is stirring in the world of displays. Organic LEDs, or OLEDs, have begun displacing older display technologies such as liquid-crystal displays (LCDs) and cathode-ray tubes (CRTs).
OLEDs are made by sandwiching layers of organic thin-film material between two conductors. An applied voltage recombines the positive and negative charges in the layers producing light through electroluminescence.
The big advantage over LCDs is that OLEDs are self-luminous and don't need backlights. This cuts power usage significantly and makes for a thinner overall package. OLEDs also have a wider viewing angle, up to 160°, compared with other flat-panel displays, and have faster response times to keep up with real-time video.
More importantly, OLEDs can be manufactured on flexible substrates such as plastic and metal foil. This means future displays can be bent or rolled up like a newspaper. For starters, however, OLEDs will most likely challenge LCDs in PDAs, digital cameras, computer monitors, and television screens.
There's quite a bit of activity in the OLED arena. One example is a venture called the Flexible Display Initiative (FDI), funded in part by the Army Research Laboratory. The aim is to research and develop a flexible display for use in military applications with target field demonstrations set for 2007-2008.
One thing is clear; the LED revolution has just begun. So sit back and enjoy the light show.
Sidebar: Measuring light
There are a few oddities and nothing standard about LED specifications. Some manufacturers provide output in candela, others in lumens or milliwatts. The lumen is a photometric measure, milliwatts are radiometric, and candela measures maximum intensity or brightness. But candela is a subjective, murky unit. For instance, suppose all the light from a source is focused down to a narrow beam. This doesn't change the actual light output or the number of photons, but the apparent brightness has increased.
The lumen/watt, the SI unit commonly used for measuring the total amount of light emitted by a light source per energy input, doesn't accurately reflect true efficiency. LEDs are directional light sources and emit little or no spherical light. But the lumen/watt unit takes into account all light emissions from a light source, including spherical, and not just what is actually accomplishing the required illumination task.
Unlike conventional incandescent and gaseous discharge sources, LEDs are semiconductor devices that convert electrical energy into a discrete color of light. The original gallium arsenide red LED was invented in the 1960s. Next came different colors such as amber and green. Then in the 90s came high-brightness LEDs, which yielded performance increases and let LEDs cross the threshold from indicator lights to sources of illumination.
LEDs are solid-state diodes made up of p-type and n-type semiconductors. The former have an excess of holes, or positive charges, while the latter have an excess of negative charges. Bringing the two materials together and applying a voltage of the right polarity causes electrons to flow into the p-type material and holes into the n-type material. But this combination has too much energy and is unstable, so some of the energy gets released as photons of light. The specific wavelength, and hence color, of light depends on the difference in energy levels as well as the type of semiconductor material used.
The standard LED chip mounts in a reflector cup of a lead frame and is encased in a solid epoxy lens. Chip packaging determines whether the light beam is narrow or wide.
Over the years, improved manufacturing techniques, packaging innovations, and better semiconducting materials have led to the development of brighter and multicolored LEDs. Once limited to simple status indicators, LEDs now play prominent roles in back lighting, panel indication, decorative illumination, emergency lighting, and animated signage.
Unlike incandescent bulbs that give off the full spectrum of light in a spherical pattern, LEDs emit a focused beam of a single wavelength. New techniques that dope semiconductors with more charge carriers increase LED light output as much as 20 times over earlier generations. Such techniques make possible daylight-visible LEDs in virtually any color of the spectrum. Still, LED production is a tricky process; devices from the same wafer can vary widely in color and light output.
Die manufacturers and packagers are often different companies. Many innovative LED designs depend heavily on die packaging, requiring a lot of cooperation to get an integrated design. Some newer designs use different reflector cups, die geometries, resin materials, bonding techniques, and thermal mechanisms.
The traditional T-1-3/4, or 5-mm, bullet shape is still what comes to mind when most people think about LEDs. But this package has poor thermal performance when used with high-output devices. LEDs can couple more intimately to the housing through use of gap pad materials, thermal epoxies, and convection currents. More recently, copper-based metal core boards have been used to further improve thermal transfer.
In addition to red, yellow, and amber, LEDs now sport colors ranging from green to ultrablue. LEDs even generate white light, long difficult to produce. The approach to generate it employs either a combination of red, blue, and green light, or shines light from a blue LED onto a phosphor layer, which then gives off white light. Indeed, the generation of blue light was a big hurdle for LEDs. The biggest difficulty was identifying a material with suitable bandgap energy. Use of gallium nitride (GaN) helped perfect not just blue but also green LEDs.