May 27, 2010...In Part 1 of this series, we mentioned that one of the common questions we get from those outside of the "chip head" side of the industry is, "Why don't they just make the LEDs (and/or solar cells) cheaper? We offered the conclusion that at the lower level of the technology supply chain, rocket science is involved. Cost reductions are happening, but it is a process of innovation that follows an evolutionary path, helped along with some occasional breakthroughs. Here in the second of this two-part commentary, we'll cover what's happening to move those costs down in the middle of the supply chain which will help bring down the costs of the LED "bulbs", luminaires and even your flat screen TV.

LEDs, the other rocket science... It wasn't that long ago that packaged "lighting quality" LEDs were running at $10 for 100 lumens, or 10-cents per lumen (remember, blue and white weren't commercially available until around 2003). Announcements in the last few months have shown us 2-cents per lumen (Cree), then 1.5-cents (Bridgelux), and most recently less than 1-cent for warm white (Intematix, part of today's news). It's assured that Philips, Osram, Nichia and others out there aren't standing pat at 10-cents per, they just didn't happen to specifically promote the price in the their announcements. That's a factor of 10 decrease in something like 5 years.

A couple of key areas are driving that improvement, including manufacturing process efficiencies at the underlying equipment and material level (discussed in Part I) and progressive improvements in the LED designs, most notably in the area of light extraction.

It might be helpful for those that live higher on the chain, including luminaire manufacturers and lighting decision makers, to catch some basic LED tech-level stuff to understand why it's worth being patient, and to help decode some of the messaging that the LED manufacturers educate us with. Without going to too deep a level (trust me, I'm not qualified), there are two basic elements of making LEDs more efficient. One is generating photons, and the other is getting them out of there. Buzz words like "internal quantum efficiency" and others are used to describe some of these interactions. Much of the light generation hurdle has been accomplished in the material rocket-science that underpins this stuff, as witnessed by recent announcements of lab results for white LEDs of over 200 white lumens per watt. It's generally agreed that there is a little bit over 300 white lumens "hiding" in a watt of energy (we specify it as white, since lumens are a measure of the human eye's reactivity to the spectrum, and white balances that out, although there are ways to skew that, such as going towards the greenish which our eyes are more sensitive to). If we've made it to 200 out of the 300, it is assured that the material is generating more than 200 lumens worth since the "extraction" portion is far from perfectly efficient.

Getting the light out... When it comes to extraction, there are a number of avenues being pursued, including the use of more reflective materials "around" and under the light emitting surface, shaping surfaces, channeling (things like Luminus Devices PhlatLight with "Phlat" deriving from their "photonic lattice" approach) and integrating "optics" effectively down to the surface of the LED chip itself (LED producer Illumitex recently made their commercial debut citing that integration as their special mojo). There is also progress continuing in the phosphors that convert the typical blue LED source light into the other colors in the spectrum. New nanophosphors ("quantum dots") offer an interesting development, as their nanometer scale materials don't scatter back much, if any, of the photons that hit it, suggesting a 15+% improvement in conversion efficiencies right there. Improvements also continue in getting the heat away from the emitting surface. As long as there is less than 100% efficiency, the leftover is heat, and semiconductors don't like too much of it. The more more you can dissipate away from the emitting surface, the more light you can pack into a smaller space. It's all still rocket-science, and continuing with that same combination of steady progress and breakthroughs.

Drivers and power... LED lighting is not made up of just LEDs and a wire. LEDs run on low voltages, and most prefer DC, although AC LEDs are available and do save one step in the power-conditioning process. There's room for improvement in both drivers (which feed and control the LEDs) and power supplies (which feed the drivers, and can actually be integrated with the drivers into an amazingly small package - ref recent Lightfair commentary). From what I understand of it, this isn't an area that's rocket science, per se, but more driven by demands for high quality, but without the volume (right now) to match. Highly reliable power control chips aren't anything new, and have often been driven by a variety of military, medical and other high-precision, high-reliability applications. That obviously shifts the required value curve away from "lots and cheap" to "few and whatever saves lives". Progress will be natural, and as mentioned in my last column, we're already looking at capable solutions for "LED lightbulbs", such as NXP is offering, in the sub-$1 range. What needs to happen is for the lamp and luminaire manufacturers to employ nothing but high-quality solutions, and for the industry ("us-all", as we say in Texan), to start holding itself accountable with the, "Whose driver?" inquiry becoming as commonplace as, "Whose LEDs?" has become. Increasing volume and standard semiconductor creativity will solve this one.

Optics and heat... LED lighting also isn't just made up of electricity. There are also optics, heat management and environmental management (dust, humidity, water, etc.) issues that are all progressing in terms of both cost and capability. Companies like Carclo and Fraen are doing a lot to bring both standard and custom optic solutions to the SSL manufacturers. Need a 60-degree optic designed to work with the Luxeon Rebel? Got it. Need a bank of matching optics to fit an array of Cree's latest? No sweat. Volume and a natural tendency to settle on some more popular form factors will drive those costs right on down. Heat management is an interesting frontier, in that it can help address both the aspects of getting more light in less space (which is a selling point for LED lights) as well as assisting in the environmental packaging. Standard approaches use different metals to get the heat immediately away from the LEDs, and then to dissipate them into the surrounding air (which proves to be an interesting discussion in itself when you think about how to get heat out of electronics when you're in the near vacuum of space... but I diverge).

The challenge of metal is that the more "fins" or vents you add, the more opportunity there is for contamination and for those heat dissipating channels to get plugged up (look inside your computer sometime... and remember you're breathing part of that stuff in before it got sucked into there). Contamination leads to less ability to dissipate the heat, which decreases efficiency and lifetime. Two interesting approaches to helping manufacturers have hit our radar scope this last year. One is from Nuventix, a fairly new Austin, Texas-based company that has targeted SSL for it's Synjet active thermal management technology. Think of it as small, ultra-high frequency speakers that vibrate the air in a synchronized fashion to create an actual directional flow. Way more reliable, and quieter than fans, but still forcing heat where you want it go. In a different direction, we've had the chance for a little dialog with GrafTech International, which uses special graphite technologies as highly-efficient passive heat spreaders. Imagine a cobrahead streetlight, with a good sized surface area, and lots of exposure to the elements. With GrafTech's solution, you mount bond the heat generating LED array to the graphite surface, which wraps around the inside top and sides of the fixture. The magic is that spreads the heat pretty evenly over the whole of the graphite, which then transmits to the whole of the outer casing, suggesting the opportunity to drop the fins and other debris-susceptible exterior extrusions, as well as greatly simplifying the thermal design.

Integrated lamps, luminaires and even TVs... As we mentioned last time, we might never get to the $5 LED lightbulb, but not because we can't, but rather because we don't want to. If a solution offers more value than its predecessors, and improves on efficiency and/or saves on lifetime costs, why should it have to sell for a price that makes it a "no brainer" in terms of simple acquisition costs. I can buy a good ranch-capable horse for somewhat less than a used dirt bike and for a lot less than a decent 4-wheel drive truck (apples to apples with the horse, you know). The truck does almost every job better. PCs created a whole new set of efficiencies and entertainment possibilities, and manufacturers have figured out that they really didn't need to keep coming down in price much below $500 to keep them moving into households across the world. They hit the value point at $500 and have stayed there, with features and capabilities being added, rather than prices proceeding lower.

We can expect to see much the same approach in LED lighting, whether at the replacement lamp (aka "bulb") or at the luminaire ("fixture plus lamp") level. The first phase of that will be the cost reduction phase, driven by the underlying technology, as well as simple manufacturing efficiencies as production capacity and sales volumes both increase. Phase 2 will kick in when we hit "the price point", whatever that may be for the particular end product we're talking about. That's actually a series of price points, that will define "cheap" from "higher end", with each having it's appropriate value story and applications. For replacement lamps, $10-$20 will probably be as far as it needs to go, and then the features will start to pick up. More light, smaller package, better color. Heck, we heard a really interesting indication from one of the Japanese companies that they intend to bring out light bulbs with remote controls so that you can turn it on, then change the color at will, with the eventual intention for sensors in the house to decide your mood and adjust the ambiance for you. So do they feed you more blue when you're feeling blue, or do they brighten the light to brighten your mood? For now, we'll suffice it to say that we've not even tapped much of the end-products' innovation curve. The new light cometh...