News for the supply chain and questions from an end user
[2008-5-15]
Tag: High Power LED Bulb
We're just back from the recent Blue 2008 LED supply chain conference that was held last week in Hsinchu (aka "LED-town") Taiwan, and with the 13-hour time change, we're compensating for creative-thinking gridlock by covering two distinct topics this week. We're often asked why LEDs can't simply just be made cheaper by taking advantage of what we know from silicon-based semiconductor manufacturing, and the the supply chain update will cover some of those concepts.We'll share the update at a bit of a primer level to help the readers that live more at the application or fixture level to maybe better relate to some of the challenges going on "down below".
Supply chain is alive, well and attacking costs
We'll likely share more about what we heard from an incredible line up of "power speakers" at Blue over the coming weeks, but when it came to the supply chain, an overriding theme this year were updates on options to continue to drive the costs out, as well as increase the overall throughput for the LED industry. Kyma Technologies CEO, Keith Evans, shared some thoughts with regard to ways to drive cost out of the epitaxial processes (basically analogous to properly matching and spreading the sauce on a pizza... differing materials in the substrate ("crust") and epitaxy ("sauce") create defects in the epitaxy that will compromise the ability to maximize the number of LEDs on a wafer (let's say that the pepperoni is only edible when it has been cooked on a perfectly smooth area of the sauce"). It's a 5-10 hour epi processing time, in which less incompatible layers are laid down progressively to lead in small steps to the final gallium-nitride (GaN) layers that the individual blue LED circuits can etched on (blue being the primary color underlying most white LED solutions today). In the process of innovating a way to produce a substrate that is extremely compatible, but currently $5K to $10K for a 2-inch wafer, they found a cost effective way to do those intermediate layers and trim roughly 20-40% off the processing time at costs at that stage, as well as increasing throughput by 30-70% which translates to more LEDs on the same equipment, and therefore less capital expense loaded up into the price.
Similarly, the CTO of Australia's BluGlass, Scott Butcher, shared the status of their new processes that produce that useful and consistent GaN layer on top of glass. It would appear to be cheaper and immediately scalable to larger diameter wafers. The move from the current 2-inch wafers to 4- or 6-inch wafers greatly increases the number of LEDs you can get from single processing run (the old pi times the radius squared thing as it relates to the amount of surface area you have to work with, and therefore less useless edge space, less handling equipment, etc, etc). Oleg Kachalov of Russia's Monocrystal shared a development path for using their sapphire starting substrates (by far the coolest, shiniest material amongst the sponsor showcase displays) in 4- and 8-inch processing systems. Sapphire represents the starting point for 80% of the high brightness LEDs out there, so the move from 2 to 4 to 8 inches represents the most apparent version of a future that ultimately drives the material and processing aspects of the LED costs to a fraction of what they are today. Key to that is the consistency and crystal uniformity in that substrate (bumpy crust leads to inconsistent sauce application, which ruins more of the light-emitting-pepperonis). Monocrystal claims it is able to get the same level of surface quality in the 4- and 8-inch substrates as it has been getting from the 2-inch models. Paraphrasing what Cree's Director of SSL Business Development, Mark McClear, shared with us in February, "2008 is the year we really begin to drive the costs out of the LEDs so they can get into all the applications they belong in."
End users encounter more sophisticated technology, more sophisticated terminology.
It was a fun call from a nice guy who is an electrician in the mid-section of the US. "I'm working with a contractor who's specifying these $75 LED replacements for fluorescent tubes and I'm trying to figure out what some of this terminology means, can you help me?" Hadn't even thought of making it an editorial discussion at the time, so this was just for fun, and it definitely got us thinking. Many of our spec sheets and discussions in the solid state lighting industry are at the more technical lighting level where beam angles and color rendering (CRI) are commonplace. But not every LED-based lighting installation is going to concepted and specified by a lighting designer. Builders and electricians are in on this deal two. I'm confident that the recessed fixtures or kitchen overhead in our last house weren't reviewed by anyone but the builder. While he may have been experienced as a builder (although naming the company "Waterloo" was apparently an experiment and the company lived up to it's name... exiled much like Napoleon), it's doubtful he carefully researched all the fixture options to choose one with specific radiation angles or "lux on the target" values.
The questions seemed kind of basic, since they are so important to understanding the characteristics of one LED lighting module or luminaire compared to another. Beam angle? My reply, "Hold your arms up in front of you in an A-frame shape and touch your middle fingers together. Now move your elbows in and out. The fluorescent fixture uses the sheet metal to vary how broadly or narrowly the light is cast. LEDs control it right there in what you would call the bulb." Lumens seem less than for the standard 40-watt tube; will it give me as much light? "In general, the LEDs can do more with less since they should only be generating light in that beam angle. The tubes you are used to have to bounce the majority of their photons off the reflector and you lose as many as 2/3 in the process, depending on the fixture." It says it's a cool white and says 5500, can we get warm white as well? "Manufacturers usually offer a range of 'color temperatures'. The trade off with LEDs are the warmer the light, typically moving from 'cool' at 5500K (for Kelvin) to 'warm' at something near 3300K. I'm noticing fluorescents seem to have the color temp labeled on them at least some of the time, so if you're not sure of what you want, try some of those tubes to see what fits the environment. There's also a pesky thing called CRI for 'color rendering index'. Above 90 is good, and less than that you need to see if it looks ok for what you're trying to do."
We're just back from the recent Blue 2008 LED supply chain conference that was held last week in Hsinchu (aka "LED-town") Taiwan, and with the 13-hour time change, we're compensating for creative-thinking gridlock by covering two distinct topics this week. We're often asked why LEDs can't simply just be made cheaper by taking advantage of what we know from silicon-based semiconductor manufacturing, and the the supply chain update will cover some of those concepts.We'll share the update at a bit of a primer level to help the readers that live more at the application or fixture level to maybe better relate to some of the challenges going on "down below".
Supply chain is alive, well and attacking costs
We'll likely share more about what we heard from an incredible line up of "power speakers" at Blue over the coming weeks, but when it came to the supply chain, an overriding theme this year were updates on options to continue to drive the costs out, as well as increase the overall throughput for the LED industry. Kyma Technologies CEO, Keith Evans, shared some thoughts with regard to ways to drive cost out of the epitaxial processes (basically analogous to properly matching and spreading the sauce on a pizza... differing materials in the substrate ("crust") and epitaxy ("sauce") create defects in the epitaxy that will compromise the ability to maximize the number of LEDs on a wafer (let's say that the pepperoni is only edible when it has been cooked on a perfectly smooth area of the sauce"). It's a 5-10 hour epi processing time, in which less incompatible layers are laid down progressively to lead in small steps to the final gallium-nitride (GaN) layers that the individual blue LED circuits can etched on (blue being the primary color underlying most white LED solutions today). In the process of innovating a way to produce a substrate that is extremely compatible, but currently $5K to $10K for a 2-inch wafer, they found a cost effective way to do those intermediate layers and trim roughly 20-40% off the processing time at costs at that stage, as well as increasing throughput by 30-70% which translates to more LEDs on the same equipment, and therefore less capital expense loaded up into the price.
Similarly, the CTO of Australia's BluGlass, Scott Butcher, shared the status of their new processes that produce that useful and consistent GaN layer on top of glass. It would appear to be cheaper and immediately scalable to larger diameter wafers. The move from the current 2-inch wafers to 4- or 6-inch wafers greatly increases the number of LEDs you can get from single processing run (the old pi times the radius squared thing as it relates to the amount of surface area you have to work with, and therefore less useless edge space, less handling equipment, etc, etc). Oleg Kachalov of Russia's Monocrystal shared a development path for using their sapphire starting substrates (by far the coolest, shiniest material amongst the sponsor showcase displays) in 4- and 8-inch processing systems. Sapphire represents the starting point for 80% of the high brightness LEDs out there, so the move from 2 to 4 to 8 inches represents the most apparent version of a future that ultimately drives the material and processing aspects of the LED costs to a fraction of what they are today. Key to that is the consistency and crystal uniformity in that substrate (bumpy crust leads to inconsistent sauce application, which ruins more of the light-emitting-pepperonis). Monocrystal claims it is able to get the same level of surface quality in the 4- and 8-inch substrates as it has been getting from the 2-inch models. Paraphrasing what Cree's Director of SSL Business Development, Mark McClear, shared with us in February, "2008 is the year we really begin to drive the costs out of the LEDs so they can get into all the applications they belong in."
End users encounter more sophisticated technology, more sophisticated terminology.
It was a fun call from a nice guy who is an electrician in the mid-section of the US. "I'm working with a contractor who's specifying these $75 LED replacements for fluorescent tubes and I'm trying to figure out what some of this terminology means, can you help me?" Hadn't even thought of making it an editorial discussion at the time, so this was just for fun, and it definitely got us thinking. Many of our spec sheets and discussions in the solid state lighting industry are at the more technical lighting level where beam angles and color rendering (CRI) are commonplace. But not every LED-based lighting installation is going to concepted and specified by a lighting designer. Builders and electricians are in on this deal two. I'm confident that the recessed fixtures or kitchen overhead in our last house weren't reviewed by anyone but the builder. While he may have been experienced as a builder (although naming the company "Waterloo" was apparently an experiment and the company lived up to it's name... exiled much like Napoleon), it's doubtful he carefully researched all the fixture options to choose one with specific radiation angles or "lux on the target" values.
The questions seemed kind of basic, since they are so important to understanding the characteristics of one LED lighting module or luminaire compared to another. Beam angle? My reply, "Hold your arms up in front of you in an A-frame shape and touch your middle fingers together. Now move your elbows in and out. The fluorescent fixture uses the sheet metal to vary how broadly or narrowly the light is cast. LEDs control it right there in what you would call the bulb." Lumens seem less than for the standard 40-watt tube; will it give me as much light? "In general, the LEDs can do more with less since they should only be generating light in that beam angle. The tubes you are used to have to bounce the majority of their photons off the reflector and you lose as many as 2/3 in the process, depending on the fixture." It says it's a cool white and says 5500, can we get warm white as well? "Manufacturers usually offer a range of 'color temperatures'. The trade off with LEDs are the warmer the light, typically moving from 'cool' at 5500K (for Kelvin) to 'warm' at something near 3300K. I'm noticing fluorescents seem to have the color temp labeled on them at least some of the time, so if you're not sure of what you want, try some of those tubes to see what fits the environment. There's also a pesky thing called CRI for 'color rendering index'. Above 90 is good, and less than that you need to see if it looks ok for what you're trying to do."
Related News »
In Focus »
footwear exports
Last month, European footwear manufacturers proposed extending anti-dumping measures against ..
B2B Keywords:
International market Chinese Importer Wholesale trade Wholesale products World trade Wholesale distributors International trade Foreign trade Wholesale distributor Importers Import export business Sell online Help u sell Global trade How to market a product Online supplier Wholesale product
International market Chinese Importer Wholesale trade Wholesale products World trade Wholesale distributors International trade Foreign trade Wholesale distributor Importers Import export business Sell online Help u sell Global trade How to market a product Online supplier Wholesale product
Feature Products»
Hot Products: A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z | 0-9
-
-
About Us -
Links -
Privacy Policy -
Terms of Use -
Site Map -
Contact Us -
Call Center: 0086-95105561 © 2007 Tootoo.com -
