Signs of the Times - April 2012 - (Page 36)

LED UPDATE “The LED industry has two segments – high-brightness (HB) LEDs with its many variations – and the rest.” manufacturing step would cut materials, processes and packaging costs. And, ideally, LED lamp-manufacturing standards would apply at the earliest stages; for example, a lessening of the various sizes and types of substrates. A masterpiece example, indeed, is reflected in some LED manufacturers’ ongoing attempt to determine whether sapphire or certain semiconductor substrates are best suited for low-cost and yet high-performance LEDs. Osram Opto Semiconductors recently reported successful results from GaN-based LEDs fabricated on 6-in.-diameter silicon substrates. It succeeded in manufacturing these GaN-on-Si LED devices for its standard Golden Dragon Plus packages, which produced 634 mW at a 438-nm dominant wavelength, for a drive electrical input power of 1.102 W – which yields an efficiency of 57.5% at the chip level. The results are on par with the company’s GaN-LEDs on sapphire substrates. Osram believes its projected, low-cost GaN-on-Si process could be production-ready within two years. The process could eventually increase the Si substrate size, which could cut material and processing costs. In contrast, Taiwan Semiconductor Mfg. Co., or TSMC, believes in growing GaN-based epitaxial layers first on 4- or 6-in.-diameter sapphire or other substrates, and then transferring them to different 8-in.-diameter wafers for subsequent fabrication steps. Such hybrid processes could be the optimized, low-cost solution because higher throughputs via an 8-in.-substrate manufacturing line would generate significantly more LED devices per run. Such dissimilar solutions pose corresponding challenges for other manufacturing stages, where frequent ideas, changes and innovations complicate the device-design layouts, fabrication procedures and standardpackaging testing methods and schemes. Chip production and inventory management The LED industry has two segments: high-brightness (HB) LEDs with its many variations, and the rest. All HB-LEDs eventually move to various applications, ranging from displayback illumination, refrigerator lighting, retail, street, automotive and general lighting. Many manufacturers don’t have systematic production lines or inventory-management processes that churn out different LED lamps for these applications. Instead, current selection process includes manual testing and matching of many HB-LEDs, which is both costly and time consuming. Ideally, perhaps, manufacturers would produce only a few types of LED modules, generated from correspondingly few production lines, and differentiated only by one or two parameters, such as their light-output level, efficacy or color temperatures. Further, if manufacturers placed tinting phosphors in the secondary optical portion instead of the initialsegment package, the color-temperature differentiation could be pushed closer to the final packaging stage. Such adoptions would make manufacturing processes more flexible and scalable, which could lead to faster turnaround times and improved inventory management. Compound semiconductor technology advancements The complexity challenge exists in generating light from compoundsemiconductor materials reproducibly, because of the numerous physical parameters’ interrelations and significant temperature dependencies. Extensive theoretical physics and practical engineering developments of compound-semiconductor materials and processing are needed to achieve fine control of these parameters in manufacturing environments and, thereby, improve LED light quality, efficiency and product yield. Compound semiconductors, GaN-based materials in particular, have higher defect densities that consequently require extensive engineering and stress-management processes for improving LEDs’ efficacy and reliability. The GaNLED side has yet to develop such a working infrastructure. Large-scale manufacturing process control If the core reasons for wide performance variability in GaN or other compound semiconductorbased LED chips are not resolved, large-scale manufacturing automation, process control and standardization will remain difficult. Engineers can only establish reliable transitions between subsequent processing stages if all variations or failures can be quantified and tolerated and, in turn, appropriate parameters can be adjusted to continue in-line processing for acceptable characteristics for final throughputs. Without such process-control manufacturing, the benefits of larger-diameter substrates will be few. Similarly, benefits from larger chip sizes, as well as batch tooling of material characterization equipment, will not significantly affect cost reductions for LED lamps. While the challenges for building well-functioning, large-scale LED manufacturing platforms may seem insurmountable, it’s encouraging to see the industry now acknowledges the unique difficulties of producing low-cost, high-quality LED lamps. When the manufacturing challenges are within control, the final outcome will present itself in higher binning yield and high-quality LED replacement lamps with much lower IFCs. n 36 SIGNS OF THE TIMES / APRIL 2012 / www.signweb.com http://www.signweb.com

Table of Contents for the Digital Edition of Signs of the Times - April 2012

Signs of the Times - April 2012
Contents
ST Update
Technology Update
Vinyl Apps
Strictly Electric
LED Update
Technology Review
Software Update
Sign Museum News
New Products
Times Square Update
That Sign is Fine (Art?)
Victory at Yorktown
Temporary, with Permanent Appeal
The Value of Signs
Anatomy of an M&A Failure
Industry News
Advertising Index
Editorially Speaking

Signs of the Times - April 2012