CONVENED AND MODERATED BY PAUL O'SHEA
LED lighting continues to move in and take over from other forms of lighting systems as prices drop and advantages become clear. But the transition isn’t without design difficulty as you will see by the responses of our panel of experts.
Electronic Products : What are the near-term challenges for LED lighting system designers?
John Perry (Strategic Marketing Engineer, Lighting Power Products, Texas Instruments ): Staying ahead of ever-changing regulatory requirements! In other words, in which countries/regions/states is a power factor above 0.7 acceptable versus those where a higher 0.9 is needed for a luminaire? Am I going to sell products in both markets? If my product can achieve 0.9, it is compliant in both markets; however, striving to reach 0.9 adds unnecessary cost to my product in the 0.7 region and could make me uncompetitive. To achieve economy of scale, it is best if a manufacturer can have a limited portfolio of products to cover its major markets.
When good enough here is not good enough somewhere else, difficult engineering and business decisions have to be made. When regional regulations don’t solidify around reasonable, common requirements, confusion and frustration result.
Paul Scheidt (Product Marketing Manager, LED Components, Cree ): One of the barriers to LED lighting adoption continues to be cost. Many system designers tend to think that using a less expensive LED will be the main driver in keeping the cost of a lighting system design down. At Cree, we have a different attitude when it comes to cost control.
There are lots of costs associated with designing an LED lighting solution, and as LEDs have quickly become more efficient, their cost relative to the other components used to create a lighting solution has gone down. So we like to take the approach that to reduce the total system cost, we need to help lighting designers reduce their costs in other areas, such as drivers, optic design, heat sinks, etc. In order to do that, we need to make LEDs that are optimized for the application in which they will be used, whether that be by making them smaller, denser, or easier to control. The idea is to create an optimized LED solution that will enable lighting system designers to fine-tune their designs and expand into more lighting applications, all while achieving a low system cost.
Marc Dyble (Product Marketing Manager, Solid State Lighting, OSRAM Opto Semiconductors ): A system designer transitioning to working with LED-based lighting products will be faced with increased complexity of system design, compounded by the constant stream of new products from LED manufacturers which makes for more aggressive design cycles. Simply choosing the right LED solution given so many available options, with various attributes and levels of quality, is a primary challenge.
Designers now have to consider not only which type of lamp to integrate, but they also have to select the right correlated color temperature (CCT) — or binning options in some cases — electrical drivers, lens options, color consistency, LED replaceability, dimmer compatibility, and thermal management.
Using common baseline performance metrics to compare products among vendors is crucial for apples-to-apples comparisons, in addition to industry-standard testing reports such as LM-80 and TM-21.
In-situ testing is also important to determine how a lighting system will perform in the field. Elevated temperature and humidity can not only affect lifetime and lumen maintenance, but extraneous variables, such as water penetration, can cause system failures due to the corrosion of materials.
Finding scalable LED solutions is still challenging. Previously, when requiring brighter source lumens, a higher-wattage lamp could be substituted within the same luminaire, whereas a similar LED solution may require a different LED module, driver, and possibly a change in lens type. Integrating field-adjustable controls is necessary to seamlessly integrate LED solutions with incumbent technologies. Today, scalable lumen packages, such as OSRAM Opto Semiconductor’s Duris S 5 platform, can provide a reduction in a system’s LED count while using a common package size, footprint, and optical pattern. Further cost-down approaches are being implemented to enable broader adoption of LED lighting. Here, LED cost is getting secondary compared to heat sinks, drivers, and manufacturing costs.
LEDs now offer many options to meet various cost-to-performance ratios depending on the luminaire design parameters.
Kishore Manghnani (Vice President, Smart Energy Products at Marvell ): For residential lighting, the primary challenge is to develop lamps with high-quality light at an affordable price.
The metrics to be met are a color rendering index (CRI) above 90 and warm dimming similar to incandescent dimming. Secondly, the lamps need to be compatible with the dimmers already installed in the home by dimming flicker-free and without shimmer.
Brett Shriver (VP Sales and Marketing for Global Lighting Technologies ): There are many challenges for LED-lighting system designers at this early stage in development. The first for them is developing the technical knowledge and expertise necessary in order to properly adopt a new technology into an existing market. LEDs are very different from any source used previously in their design, requiring an understanding of light emission patterns and power requirements, as well as mechanical shapes and sizes.
A second challenge is in taking the light from the LED and creating distribution patterns from these point sources that are accepted in the market. There are many different thought groups on what the market needs, so it will be important to make a determination of the final design requirements. The third challenge is in educating the consumer properly about LED lighting systems. The designers must work to develop good reliable systems with high-quality components so that there is not a pullback in adoption due to low-quality products creating negative feelings in the market.
Scott Barney (VP Sales and Marketing at ERG Lighting ):
The near-term challenges for LED Lighting system designers is to make sure they properly choose an LED light engine, driver, and control system that will work in conjunction with each other without causing shortened life issues or poor lighting quality. It is very important to make sure that the driver and LEDs are kept cool in order to maximize the lifetime of the LED luminaire.
Electronic Products : What are the challenges for designers wanting to implement smart lighting concepts?
John Perry : Picking a winning standard! Currently there are too many choices. Wired, wireless, or hybrid of both? Commercial applications are slightly easier. Upgrading a lighting system in a factory, shopping center, or multi-story office building is going to be considered carefully. Most likely, one of the well-known lighting control companies will be awarded the contract, and it is going to be done once and last a very long time. Here it is not so important if the control system is an open standard or proprietary scheme
The consumer space is a totally different situation. I am not sure how much the “average” person really thinks about lighting today. Take LED retrofit lamps, for example. The value proposition of a 60-W light bulb retailing at $20 is actually pretty compelling if you do the math. Therein lies the problem: Most potential buyers are so put off by the sticker price they never considered the payback. Instead, they opt for the $0.50 incandescent they will replace 20 times over.
So what if residential lighting were smart? What features will consumers want — control by a mobile app, color changing, dimming, occupancy sensing, wall wash, disco lights, plays music, kill germs? That, as they say, is the $50,000 question. I don’t have a crystal ball on this either, but in my opinion, a common, open control standard and eco-system would go a long way in making smart lighting viable. It probably needs to be wireless for DIY installation, and it should not require extra equipment the user does not already have. In other words, I think anything needing gateways and complicated provisioning is not going to be successful. Few people have the patience for this.
Marc Dyble : The primary challenge in implementing smart lighting is balancing cost and function. The sky’s the limit with LEDs but smart- lighting concepts are complex and the solutions tend to be more expensive.
Some smart solutions have an ecosystem story with savings and efficiency only realized when you take into account the entire building or network of installed fixtures. We need to consider factors beyond fixture lm/W and lm/$. Also, the smartest solutions may mean changing the existing way we view luminaires and blending them with other functions (e.g., sensing, security systems, communication, etc.). Other smart systems allow for the ability for lamps to mimic the performance of their incandescent and halogen counterparts, such as a “Warm-Dim” effect by changing the CCT from 3,000K to 2,000K as the lamp is dimmed.
The currently installed control infrastructure is a roadblock to the deployment of smart lighting. Most LED systems cannot operate on, or offer reduced performance on, residential or commercial dimming systems, requiring new compatible controls to be installed which adds to the system cost. Riding on established communication networks, such as Wi-Fi, ZigBee, Bluetooth, or other RF protocols, LED systems are overcoming this obstacle, while at the same time becoming simpler to deploy, by allowing our lighting to be controlled not by a wall switch, but a mobile device, opening up new possibilities for remote operation. This deployment relies heavily on finding partners willing to integrate a common control protocol, giving specifiers a broad range of fixtures working with the same control system. Lamps become more than just a light source; they become a communication node on a mesh network, strengthening the network’s range and offering redundancy without depending on a standalone control system. By modulating the LED light source itself, your light fixtures could now also be your Internet gateway instead of using Wi-Fi.
Though adding convenience through networked lighting, these smart technologies are subject to the same security issues that all other connected products face, such as threats of security breaches and hacks, just like a computer network. Recent news of LED smart lights hacked by a security researcher caused an entire home to be blacked out. Secure, flexible lighting systems can now offer new ways to light spaces by unconventional means.
Kishore Manghnani : There are some common and unique challenges for residential and commercial smart lighting. The common challenge is a lack of standards for wireless control. When products are developed to comply to a single standard, it accelerates the adoption rate by eliminating confusion in the marketplace and reducing prices from increased competition.
The unique challenges for residential is developing a low-cost solution and small size for the bulb.
There are several challenges for commercial developers: 1) A long payback time on the investment due to the high price of equipment and installation costs driven by being both time consuming and requiring expensive, specialized technicians. 2) The lack of standards results in equipment from vendors not being interoperable. 3) Developers of commercial lighting need to maintain a large inventory of drivers SKUs in order to drive lamps with different output currents. 4) Also there are many new government regulations that apply to energy consumption in buildings they need to meet for permitting and inspections.
Brett Shriver : The largest challenge is the overall cost of implementation. In many applications, the cost and carbon footprint of implementing smart-lighting concepts cannot be recovered over the life of the product.
Scott Barney : The major challenge for designers wanting to implement “smart” lighting into the product is to be aware of the different options available that have been proven to be successful so far. The type of dimming they choose will depend heavily on the driver/control system working together. They must also understand what they are trying to accomplish with the product, whether it be energy savings, daylight harvesting, occupancy sensing, or individual task tuning. Once these questions have been answered, they can begin the investigation into what products are available to satisfy those needs.
Paul Scheidt : While I can’t address the specific challenges that designers face to implementing smart lighting concepts, what I do know is that the benefits that LED light sources provide, such as instant-on lighting, easily controlled dimming, and extended life cycles, have allowed for the design of more smart lighting concepts.
Electronic Products : What pointers do you have for designers needing to choose the best matched LED drivers for their designs?
John Perry : In my experience as a driver IC manufacturer, the LED is decided 90% of the time before the drive solution is considered. Many times we wish it weren’t true because the LED selection may conflict with the desired electrical performance expectations.
Best advice I can give…iterate a few times. Start with the desired light distribution, pick a suitable LED and optic, then match up the drive solution. If the complete design targets of the system are not met, reconsider the assumptions and try again.
Marc Dyble : It is important to choose drivers that best match your intended load to increase driver efficiency, reduce cost, and provide the desired lit appearance of the LED fixture. In an array of LEDs, matching the output current to the load, through a single high-current or multi-channel driver, will reduce the VF mismatch among the LED strings, providing a uniform intensity across an LED array. Paying particular attention to limiting the ripple current from the LED driver will help to reduce current fluctuations, which can cause visible flicker of the LEDs and additional heating of the LED system and driver.
The increasing availability of high-voltage dc LEDs allow designers to better match a driver’s input and output voltages, increasing driver efficiency and reducing system cost. Making sure these systems comply with UL Class 2 classification can be helpful in avoiding additional safety testing.
Other desired features such as dimming interfaces, short-circuit and polarity protection, and IP ratings (dry, damp, or wet location) should be evaluated. If dimming is required, determine the desired dimming protocol needed. Know that further analysis of dimmer switch compatibility may be required as not all dimmers work with all LED drivers, commonly leading to such undesirable consequences such as flicker, pop on, dead travel, and drop out.
Get in contact with LED suppliers for recommended driver manufacturers. For example, at OSRAM’s partner website, www.ledlightforyou.com, several trusted manufacturers are present to assist in the selection of electronic control gear (EGC) suitable for specific LED packages. Electronic distributors can also assist to find the best suited drivers using their technical resources and personnel.
Kishore Manghnani : Developers of commercial lighting need to select a solution with a platform that will allow them to take a modular approach, whether the lighting control they want to install is wired (DALI) or wireless (ZigBee or Wi-Fi). In this way, they will have greater flexibility and can significantly shorten their time to market. Additionally, using drivers that are software programmable can significantly reduce the number of SKUs.
Developers of retrofit lighting need to select a driver IC that is based on a digital architecture that can adapt to a wide range of dimmer types and ensure the highest dimmer compatibility. A driver should have the lowest component count possible, which results not only in lower system cost but also increases reliability. Finally, choosing a driver IC that supports a wide range of lamps from 4 to 40 W allows a manufacturer to develop a wide range of products around a common architecture.
Scott Barney : Designers must pay attention to the power that the light engine requires and the power capabilities of the driver without running either too close to the maximum ratings. The LED junction temperature and the driver thermals must be kept at a reasonable temperature in order to maximize the life of each. The driver must have the proper dimming options that the designer wants to implement into the LED system or they will see flicker or pulsing of the light engine. The designer must also understand the quality of the LED driver and the components that the manufacturer chooses to use, especially the capacitors inside the driver. The capacitor is viewed as the weakest link in the driver, although a correctly designed driver will not be affected by the type of capacitor used. A properly matched driver/light engine/control system will maximize the life and quality of the luminaire, giving the designer the best results for his product.
Paul Scheidt : As the industry moves toward more integrated solutions such as LED arrays and modules, designers no longer have to start from scratch and arrange discrete LEDs onto a printed circuit board and develop a driver to match it. The possibilities in that scenario can be overwhelming. By integrating the LEDs into arrays such as the Cree CXA family of LED arrays, you can simplify the process by giving them a well-defined set of specifications that will provide a clear path to driver compatibility. At Cree, we’ve taken this idea a step further with our Driver Compatibility Program (DCP) for our LED modules. The DCP helps us identify third-party drivers that are compatible with Cree LEDs. But we don’t just take a look at the specs and assume that the drivers are compatible; we actually test them with our LEDs to ensure they are, which allows us to extend our LED module warranty to our customers who are using DCP drivers.
Electronic Products : What are the thermal management challenges when designing for new LED lighting designs and for retrofit applications?
John Perry : New lighting designs: If the physical dimensions of the luminaire have some flexibility, a number of possibilities exist. Perhaps heat-sinking materials can be obscured by good industrial design, or even incorporated into it. Toshiba makes a very sleek street lamp in which the heat sink is a very prominent part of the luminaire, but they have done a very nice job of making the product look futuristic as a result. Others are very functional, but perhaps lack visual appeal.
Retrofit lamps: physics dictate the LED drive power level needed to produce “equivalent” light. Lamps have very specific form factors that constrain the area available for LEDs, optics, driver and heat sink, meaning compromises have to be made that may limit the environment where the lamp can be used. Some lamps will say “not for use in enclosed fixtures” like can lights. Reason being, can lights don’t provide any air flow and bulbs can overheat. Designing a heat sink that will work is more expensive than one which assumes a small amount of convective airflow. If can lights are not the target market for that lamp and manufacturer, it is easy to see why they made this tradeoff.
Marc Dyble : Thermal management requirements are application specific, therefore what may work for a retrofit lamp is different than what might work for a complete luminaire. In the case of retrofit lamps, both the size limitations and heat generated by the LED and driver limit the product’s rated life. As drivers become smaller and LEDs more efficient, higher wattage replacement lamps are achievable through the use of advanced cooling techniques. Due to the higher performance targets, most LED luminaires require more aggressive thermal management approaches to manage heat. Remotely mounted drivers can be used to mitigate heat transfer between LEDs and drivers leading to higher system reliability and lifetime.
The lighting industry is moving toward products that mimic the look of legacy products, despite the fact the lamp is hidden in many applications. This is particularly true with retrofit lamps. Previous lamp designs featured visible heat-sink fins that maximized thermal dissipation through convection. In some of the newer designs, heat sinks are occluded from view through use of plastic covers, producing products with a more conventional design and with no evident heat sink. Moldable thermoplastic materials, capable of conducting heat, have aided in this migration, offering lightweight thermal solutions for retrofit lamps with limited space constraints.
The use of low- or mid-power LEDs can significantly reduce the thermal load, compared to high wattage solutions, by spreading out the concentration of heat across a given area, thus reducing the amount of thermal mass required. This can enable lower cost designs through the use of less expensive FR4 material over costly metal-clad PCBs. Under-driving the LEDs is another approach to reduce the thermal load by limiting the LED temperature and increasing the system efficacy.
New LED luminaires offer more design flexibility, with trends to use Chip-on-Board (COB) LEDs for assembly simplicity, though the inherent high thermal density requires extra attention. Using good interface materials, proper pressure, and proper mechanical coupling to dissipative housing structures are key to well executed designs.
Kishore Manghnani : The temperature inside a bulb can easily exceed 125º. There are especially thermal challenges when designing for 75 to 100-W A19 bulb or for a 50-W MR16. For a solid thermal design, manufacturers need to consider an IC that has a high efficiency factor so less of the energy is dissipated as heat. The IC should have innate intelligence so it can react to overheating conditions by auto-correction via a thermal foldback. Also, they need to select components that are rated for higher temperatures. Manufacturers will have to design innovative heat sinks that are not only effective, but are aesthetically pleasing to consumers.
Brett Shriver : When working on new LED lighting designs and fixtures, there are very limited challenges with thermal management as they are being easily overcome. More of the issues with thermal management are created by trying to fit the new technology into the existing footprint of incandescent and fluorescent fixtures. The retrofit market continues to condense as many lumens as possible into the smallest spaces to mimic existing technology. This creates a need for thermal management.
Scott Barney : Thermal issues are the most important aspect of the lighting system's lifetime. This issue must be taken very seriously by the designer from the beginning of the design process. The major issues are keeping the LEDs and the driver cool in whatever environment the product is going to be exposed to. The LEDs will fail if the junction temperature is not kept within the specifications. A perfect example of this occurs in a T8-type LED tube product when the driver is incorporated within the tube itself. The heat of the driver at one end of the tube will cause discoloration and early failures of the LEDs that are sitting near the driver. The driver causes a temperature rise of anywhere between 10º and 40ºC, depending on the match between the driver and the light engine. The designer must consider the temperature rise of the system at ambient temperature along with the temperature rise of the environment of the end product, whether inside a cool office or outside in the elements.
Paul Scheidt : First let me start by saying that thermal management challenges in the LED industry are more difficult now than we expect they will be in the future. As LEDs become more efficient and start to put out more light with less power, thermal management will become comparatively easier. But that being said, probably one of the biggest factors affecting thermal management at this time is the fast adoption of LED arrays. LED arrays present a challenge for thermal management because heat is no longer dispersed across a range of smaller, discrete LEDs, but concentrated onto a single source. And as wattages increase to push the performance of LED arrays, the heat distribution will become even denser.
Electronic Products : What are the advantages/disadvantages of ceramic versus plastic packaging for LED lighting designs?
Paul Scheidt : Plastic packaging has become popular among LED lighting manufacturers because it is inexpensive and easy to assemble. But the concern with plastic-based LEDs is that they degrade rapidly when exposed to heat and short-wavelength blue light. Many lighting manufacturers are looking to cut cost by reducing the number of LEDs in their designs, which means using more current to produce more light, which creates more heat. That is a bad combination for plastic LEDs.
Ceramic-based LEDs are much more durable under high-heat and high-light output environments, which naturally gives them an advantage over plastic LEDs. This will be important as more LED manufacturers continue to push the boundaries of LED performance. Additionally, although ceramic material is traditionally viewed as more expensive, at Cree we’ve been able to reduce our cost of ceramic material relative to plastic because we are a vertically integrated company.
Marc Dyble : Ceramic based packages can be advantageous especially with respect to lumen maintenance, and use in extreme environments requiring more robust components. Ceramic packages typically have a lower thermal resistance based on the material substrate used (e.g., Al2O3), and their isolated thermal pads offer greater design freedom in PCB layouts by maximizing options for electrical isolation and heat dissipation. This is not possible with plastic, lead-frame-based packages. Most ceramic packages do not contain silver mirrors so they are a preferred choice in wet or corrosive environments or when high temperatures greater than 100°C are expected. When demanding optical designs are required for spot or street lighting applications, surface-emitting chips are preferred, which mainly come in ceramic packages, in contrast to volume-emitters that are mainly used in CoB or plastic packages. When mounted on MCPCBs, ceramic-based LEDs can experience greater thermomechanical stress as compared to plastic QFN packages, meaning temperature changes (thermal cycling) will result in stressing of the soldered joint, due to the difference in SMD package and PCB material (CTE mismatch). This stress can lead to gradual fatigue and aging of the solder joint and possible functional failure of the LED, owing to loss of mechanical, electrical, and thermal contact. When the LEDs are mounted to FR4 substrates, these conditions are lessened, resulting in a more stable and reliable system.
Plastic packages have the advantage of being very cost effective, making them suitable for a variety of lighting products, though their lifetime is limited due to package degradation caused by ultraviolet (UV) light. A balance between cost and performance comes from the use of advanced plastic packages, such as thermoplast or ceramic-based epoxy compounds, which lead to longer lifetimes as they are resistant to UV light due to increased package reflectivity and reduced absorption, and can withstand high operating temperatures and still provide a lower-cost solution than ceramic packages.
Choices have to be made on the best solution for a given application. Indoor versus exterior use, thermal resistance differences, and cost-performance tradeoffs need to be reviewed as one package style may not be best suited for every application.
Brett Shriver : Ceramic packages allow for better-quality fixtures to be created as the package can be electrically isolated while still being thermally conductive, allowing a more efficient transfer of thermals.
Scott Barney : We will leave the specifics of the LEDs to the LED manufacturers, but our understanding of the big differences between the two technologies is the thermals and the quality of light of each. There seems to be some concern in the marketplace about the ceramic LEDs being a little more fragile in the manufacturing process than there is with the Plastic LEDs.
Advertisement
